Short- and Long-Term Effects on Mother's and Her Children's Health

Pregnancy and Lifestyle: Short- and Long-Term Effects on Mother's and Her Children's Health

Guest Editors: Riitta Luoto, Michelle F. Mottola, and Leena Hilakivi-Clarke Pregnancy and Lifestyle: Short- and Long-Term Effects on Mother’s and Her Children’s Health Journal of Pregnancy

Pregnancy and Lifestyle: Short- and Long-Term Effects on Mother’s and Her Children’s Health Guest Editors: Riitta Luoto, Michelle F. Mottola, and Leena Hilakivi-Clarke Copyright © 2013 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in “Journal of Pregnancy.” All articles are open access articles distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Editorial Board

Vincenzo Berghella, USA Sinuhe Hahn, Switzerland Vorapong Phupong, Thailand Sean Blackwell, USA Mordechai Hallak, Israel Fulgencio Proverbio, Venezuela Hein Bruinse, The Netherlands Jeﬀrey Keelan, Australia M. Rogers, Hong Kong Rosa Corcoy, Spain Justin Konje, UK Daniel S. Seidman, Israel R. L. Deter, USA David F. Lewis, USA Lee P. Shulman, USA Graziano Di Cianni, Italy E. R. Lumbers, Australia J. L. Simpson, USA Gian Carlo Di Renzo, Italy Sam Mesiano, USA Marlene Sinclair, UK Keith A. Eddleman, USA Richard K. Miller, USA Mark S. Sklansky, USA Fabio Facchinetti, Italy Roberta B. Ness, USA Deborah A. Wing, USA Antonio Farina, Italy Cees Oudejans, The Netherlands Tamas Zakar, Australia Albert Fortuny, Spain Attila Pal, Hungary Contents

Pregnancy and Lifestyle: Short- and Long-Term Effects on Mother’s and Her Children’s Health, Riitta Luoto, Michelle F. Mottola, and Leena Hilakivi-Clarke Volume 2013, Article ID 537526, 2 pages The Heart-Placenta Axis in the First Month of Pregnancy: Induction and Prevention of Cardiovascular Birth Defects, Kersti K. Linask Volume 2013, Article ID 320413, 11 pages Prevalence of Metabolic Syndrome One Year after Delivery in Finnish Women at Increased Risk for Gestational Diabetes Mellitus during Pregnancy, Jatta Puhkala, Tarja I. Kinnunen, Tommi Vasankari, Katriina Kukkonen-Harjula, Jani Raitanen, and Riitta Luoto Volume 2013, Article ID 139049, 7 pages Limiting Excess Weight Gain in Healthy Pregnant Women: Importance of Energy Intakes, Physical Activity, and Adherence to Gestational Weight Gain Guidelines, Tamara R. Cohen and Kristine G. Koski Volume 2013, Article ID 787032, 6 pages Current Thoughts on Maternal Nutrition and Fetal Programming of the Metabolic Syndrome, Bonnie Brenseke, M. Renee Prater, Javiera Bahamonde, and J. Claudio Gutierrez Volume 2013, Article ID 368461, 13 pages Physical Activity during Pregnancy: Impact of Applying Different Physical Activity Guidelines, Katie M. Smith and Christina G. Campbell Volume 2013, Article ID 165617, 9 pages Stages of Change Model for Participation in Physical Activity during Pregnancy, Lene Annette Hagen Haakstad, Nanna Voldner, and Kari Bø Volume 2013, Article ID 193170, 7 pages Kinematic Analysis of Gait in the Second and Third Trimesters of Pregnancy, Marco Branco, Rita Santos-Rocha, Liliana Aguiar, Filomena Vieira, and Antonio´ Veloso Volume 2013, Article ID 718095, 9 pages Prepregnancy Physical Activity in relation to Offspring Birth Weight: A Prospective Population-Based Study in Norway-The HUNT Study, Silje Krogsgaard, Sigridur L. Gudmundsdottir, and Tom I. L. Nilsen Volume 2013, Article ID 780180, 6 pages Use of Medicines with Unknown Fetal Risk among Parturient Women from the 2004 Pelotas Birth Cohort (Brazil), Andrea´ Damasoˆ Bertoldi, Tatiane da Silva Dal Pizzol, Aline Lins Camargo, Alu´ısio J. D. Barros, Alicia Matijasevich, and InaS.Santos´ Volume 2012, Article ID 257597, 11 pages Effects of Tobacco Smoking in Pregnancy on Offspring Intelligence at the Age of 5, Hanne-Lise Falgreen Eriksen, Ulrik Schiøler Kesmodel, Theresa Wimberley, Mette Underbjerg, Tina Røndrup Kilburn, and Erik Lykke Mortensen Volume 2012, Article ID 945196, 9 pages Preventing Long-Term Risk of Obesity for Two Generations: Prenatal Physical Activity Is Part of the Puzzle, Stephanie-May Ruchat and Michelle F. Mottola Volume 2012, Article ID 470247, 33 pages Differential Effects of Chronic Pulsatile versus Chronic Constant Maternal Hyperglycemia on Fetal Pancreatic β-Cells, Mackenzie S. Frost, Aqib H. Zehri, Sean W. Limesand, William W. Hay Jr., and Paul J. Rozance Volume 2012, Article ID 812094, 8 pages Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 537526, 2 pages http://dx.doi.org/10.1155/2013/537526

Editorial Pregnancy and Lifestyle: Short- and Long-Term Effects on Mother’s and Her Children’s Health

Riitta Luoto,1,2 Michelle F. Mottola,3 and Leena Hilakivi-Clarke4

1 UKK Institute for Health Promotion Research, 33501 Tampere, Finland 2 Department of Children and Families, National Health and Welfare Institute, 00271 Helsinki, Finland 3 R. Samuel McLaughlin Foundation-Exercise and Pregnancy Lab, 3-M Centre, The University of Western Ontario, London, ON, Canada N6A 3K7 4 GeorgetownUniversity,ResearchBuilding,RoomE407,3970ReservoirRoad,NW,Washington,DC20057,USA

Correspondence should be addressed to Riitta Luoto; [email protected]

Received 8 May 2013; Accepted 8 May 2013

Copyright © 2013 Riitta Luoto et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Pregnancy and the foetal environment can have a profound included in this issue, it was reported that women at high influence on many chronic diseases, such as diabetes obesity, risk of gestational diabetes were at increased risk of devel- breast cancer, and cardiovascular diseases in both the mother oping metabolic syndrome, which includes central obesity, and her offspring. Much of the influence of the intrauterine dyslipidemia, and blood glucose abnormalities, already one milieu is transmitted to the next generation through epige- year postpartum. netic mechanisms. However, the specific pathways affected Lifestyle modifications during or after pregnancy have through these mechanisms by lifestyle during pregnancy are been suggested to be valuable in chronic disease preven- largely unknown and deserve further study. tion.Further,thesemodificationsmayhavetotakeplace The special issue presents the latest findings and reviews before pregnancy. Prenatal education programs have been from all over the world concerning pregnancy and lifestyle, studied in Ontario. Canadian researchers recommend that especially nutrition, physical activity, and weight gain during prenatal education programs should include information pregnancy. The special issue includes three reviews, twoof on appropriate gestational weight gain targets that can be which cover the role of maternal physical activity, under- achieved by lowering energy intakes and being physically and overnutrition, and intake of specific nutrients during active. According to the Norwegian HUNT study, women pregnancy on long-term risk of chronic diseases among with high prepregnancy body mass index have an increased mother and her offspring. The third review article concen- risk of having macrosomic offspring, if their physical activity trates on induction and prevention of cardiovascular birth was low as well. defects. The first month of gestation is the most important According to researchers from Iowa, in the USA, exist- for development of cardiovascular birth defects that may be ingphysicalactivityguidelinesduringpregnancyvary,and preventable by folate acid supplementation. further research is warranted to identify health (mother and The original articles in this special issue cover topics from offspring) enhancing physical activity and relevant pregnancy lifestyles (physical activity and nutrition) to chronic diseases outcomes. This is important, since pregnant women in a Nor- and predisease states, such as metabolic syndrome, during wegian hospital-based study had high motivational readiness pregnancy. The only animal study included in the special or intention to increase their physical activity level. Preg- issue describes findings showing how maternal hypergly- nancy can therefore be considered a “window of opportunity” caemia is related to foetal pancreatic function, resulting in for long-term physical activity habits. New information on inadequate insulin secretion postnatally. In a human study pregnant women’s physical activity was reported by using 2 Journal of Pregnancy kinematic analysis in Portugal. Pregnant women need to maintain greater stability of body and increase the efficiency of physical activity. Finally, pregnant women are exposed to other lifestyle- related threats besides poor nutrition and lack of physical activity, which should be monitored as well. A Brazilian birth cohortstudyreportedthathalfofpregnantwomenusedat least one medicine with an unknown fetal risk. It is well reported that smoking during pregnancy is harmful for a fetus, but a Danish National Birth Cohort study was unable to establish a connection between smoking during pregnancy and offspring’s intelligence at the age of 5. Later consequences of lifestyle during pregnancy may be difficult to separate from genetic profiles and other determinants, which should be investigated in future studies. Riitta Luoto Michelle F. Mottola Leena Hilakivi-Clarke Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 320413, 11 pages http://dx.doi.org/10.1155/2013/320413

Review Article The Heart-Placenta Axis in the First Month of Pregnancy: Induction and Prevention of Cardiovascular Birth Defects

Kersti K. Linask

USF Children’s Research Institute, CRI #2007, Department of Pediatrics, 140-7th Avenue South, St. Petersburg, FL 33701, USA

Correspondence should be addressed to Kersti K. Linask; [email protected]

Received 8 November 2012; Revised 4 March 2013; Accepted 13 March 2013

Academic Editor: Riitta Luoto

Copyright © 2013 Kersti K. Linask. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Extrapolating from animal studies to human pregnancy, our studies showed that folate (FA) deficiency as well as one-time exposure to environmental factors in the first two to three weeks of human gestation can result in severe congenital heart defects (CHDs). Considering that approximately 49% of pregnancies are unplanned, this period of pregnancy can be considered high- risk for cardiac, as well as for neural, birth defects, as the woman usually is not aware of her pregnancy and may not yet be taking precautionary actions to protect the developing embryo. Using avian and mouse vertebrate models, we demonstrated that FA supplementation prevents CHD induced by alcohol, lithium, or elevation of the metabolite homocysteine, a marker for FA deficiency. All three factors affected the important Wnt signaling pathway by suppressing Wnt-mediated gene expression inthe heart fields, resulting in a delay of cardiomyocyte migration, cardiomyogenesis, and CHD. Optimal protection of cardiogenesis was observed to occur with FA supplementation provided upon morning after conception and at higher doses than the presently available in prenatal vitamin supplementation. Our studies demonstrate pathways and cell processes that are involved with protection of one-carbon metabolism during heart development.

1. Environmental Influences fetal development, including variants associated with adult- onset disease. The finding that a large number of disease- Extrapolation of experimental results using mouse and avian associated variants are located in regulatory DNA regions embryonic models demonstrates that environmental factors that are active during embryonic development suggests that present in utero during the second to third week of human environmental exposures affecting the intrauterine milieu gestation (human, 16 to 19 days after fertilization; mouse, during early embryonic stages can influence risk for not embryonic days ED 6.75 to 7.5; avian, HH stage 4 to stage only birth defects, but also a large number of adult diseases. 5-) can alter early developmental processes resulting in This concept of developmental origins of health and adult severe cardiac anomalies [1–5]. Some perturbations may disease arose with data that the intrauterine milieu influences alter development in a manner that may not be clinically cardiovascular and mental diseases long into adulthood evident at birth, but result in increased susceptibility to [7–9]. cardiac problems after birth, or increasingly as we age. In Early in human gestation following conception, the fer- a recent study published in September 2012 [6], NIH Com- tilizedembryoimplantsintheuterinewall.Thesuccessful mon Fund researchers reported on genome-wide association implantation requires adequate maternal uterine perfusion studies (GWASs) of genetic variants, specifically pertaining and endogenous hormone preparation to allow initial embry- to noncoding regions of DNA that actively regulate gene onicsurvivalandlaterorgandevelopmentandgrowthfor expression. They found that 88 percent of GWAS variants a term gestation. Shortly after implantation, cardiomyocyte are in regulatory DNA regions of genes that are active in specification and commitment take place between 16–19 days 2 Journal of Pregnancy postconception (pc). The circulation and a beating tubular to develop. The statistic that 49% of all pregnancies are heart are established by 21 days pc in human pregnancy. unplanned [25],wouldindicatethefirstmonthisahighrisk In mouse development, cardiac specification corresponds period for the early embryo and specifically for cardiovascu- toembryonicday(E)6.75toE7.5wherebythecardiac lar development, neurogenesis, and for placentation in that in crescent forms. Next, a beating, linear, and tubular heart all three organ systems important steps in cell specification forms that then loops and septates (E10.5) to form a four- and early differentiation processes are taking place that are chambered heart (4–6 weeks of human gestation). Human regulated by the same signaling pathways, specifically as we cardiac morphological development is complete by 8 weeks have reported, Wnt/𝛽-catenin signaling. of age. The focus of my research has been on early stages of heart Abnormal cardiac development leading to congenital development, specifically regulation of cardiac cell specifica- heartdiseasecanbeassociatedwithabnormalplacen- tionanddifferentiationusingthemouseandavianembryonic tal development with abnormal trophoblast invasion and models. In the human embryo, a beating, tubular heart with remodeling and the resultant abnormal transfer of nutrients extra- and intraembryonic circulations bringing blood to the and oxygen from the mother to the fetus [10, 11]. Studies relat- heart is present at 21 days. Because cardiac cell fate decisions ing to germ-line ablation of specific genes, as p38𝛼-mitogen are taking place several days earlier, approximately at 16 to activated protein (MAP) kinase or of peroxisome proliferator- 19 days of human pregnancy, these early periods of human activated receptors (PPARs), demonstrated they were critical gestation are not possible to analyze on a cell and molecular for proper placental function and, when ablated only in the level. Thus, the mouse model is an excellent mammalian placenta, resulted in cardiac defects. When the placental model to analyze developmental processes leading to a function was rescued via aggregation with tetraploid embryos functional heart. With present day technologies, both cardiac (tetraploid morulas failed to contribute to embryonic tissues, structural and functional relationships can be monitored in only placental), this corrected the cardiac defect [10, 11]. These the mouse embryo using Doppler ultrasound (Figure 1). studies highlight the existence of a critical heart-placental axis Noninvasive evaluation of fetal heart function during that is just beginning to be recognized. early human pregnancy using clinical echocardiography is The heart-placental axis is associated with parallel devel- challenging. The earliest variables of early fetal cardio- opment of the placenta and heart that utilizes many com- circulatory dynamics have been established for 11–14 weeks mon molecules and genes [11–16] and reflects intimate and of gestation [26]. Those critical developmental processes synergistic growth of both organs. Two important pathways are already well underway in cardiogenesis by the time critical to placental development, implantation, and cardiac pregnancy is usually confirmed at weeks 5/6 after-fertilization development during the same period of early development 𝛽 which may relate to why congenital heart defects arise in 1% are canonical Wnt/ -catenin signaling and pathways asso- of births world-wide and of all birth defects are the most ciated with folate metabolism [17–21]. If these pathways are common. As the heart is the first organ-system to develop, altered, developmental anomalies ensue. Failure to maintain followed closely by the nervous system, our recent research a pregnancy or have normal cardiac development may be on early development using environmental exposure during due to abnormal placental function or cardiac function. Fetal gastrulation may help to explain why cardiac birth defects are loss can occur prior to usual detection of pregnancy at 5 to so prevalent and are the leading cause of infant mortality. In 6weeksofgestation(estimatedtobeasignificantpercent the first month, women may be continuing to use prescribed of all fertilized embryos) or occurs later after initiation of medications or leisure drugs, to smoke, and to drink alcohol, the embryonic heartbeat. Placental volume blood flow is may be type-2 diabetic or obese, and not be following neither a major determinant of early cardiac output, fetal growth, healthy lifestyles or diets, nor protecting their embryos with and well-being [22, 23]. The associated placental abnormality prescribed perinatal folate supplementation. during gestation of the fetus with congenital heart disease Multiple early processes are associated with similar sig- has deleterious effects on the outcome of pregnancy. There naling pathways in heart, placental, and neural develop- may be such severe starvation of the fetal circulation and ment, and thus development of different tissues in the same growth failure that growth arrest, severe hypoxemia and embryos can be affected with the same exposures [27]. In marginal cardiac output occur with spontaneous loss of the pregnancy. Premature delivery may result in birth of a recent study of infants in the Atlanta metropolitan area, a neonate with poor prognosis related to the prematurity 28.7% of infants with CHDs also had associated major and low birth weight. Management of the neonate with noncardiac malformations [28].Ifonereadstheliterature prematurity and congenital heart disease is associated with provided to women upon learning of their pregnancy usually a very high perinatal mortality. at5to6weeksafterconception,pamphletspredominantly The underlying mechanisms by which environmental focusonneuraldevelopmentandtheuseoffolateduring exposures can result in congenital birth defects, as well as pregnancy with targeting the second to third month of disease much later in life, have had no clear experimental gestation. Possibly due to a lack of knowledge on effects of definitions, but there is evidence that cellular processes asso- environmental exposures during early embryonic develop- ciated with cell fate decisions and cell lineage modulation are ment, generally there is little discussion of adverse effects on involved [4, 24]. In the case of cardiomyocyte specification, cardiac outcomes, importance of planned pregnancy, and the it is among the earliest cell-fate decisions to occur, because useoffolatebeforebecomingpregnantandcontinuinguntil the heart-cardiovascular system is the first organ-system pregnancy is confirmed, so as to protect the earliest stages Journal of Pregnancy 3

IF-OF

DV DA

UC UC (UA-UV)

Figure 1: Color Doppler (top left) and pulsed directed Doppler were used to obtain blood waveform patterns. Normal ventricular inflow and outflow (IF-OF) pattern is shown in top right. Locations of waveforms obtained for ductus venosus (DV), descending aorta (DA), and umbilical cord (UC) for umbilical artery (UA) and umbilical vein (UV) are indicated. Permission is granted for the modification of a figure from Merck Source Resource Library, an online Elsevier publication (http://www.mercksource.com).

of embryonic development, specifically of the cardiovascular indicates that the observed cardiac defects are preventable system [5, 18], placentation [29], and neurogenesis [30]. with folate supplementation that is initiated early after conception and at higher doses than presently available in 2. Cardiac Birth Defects periconceptional vitamins. Our results with animal models demonstrate that even before pregnancy is realized, many different forms of cardiac birth 3. Common Cardiac Birth Defects Arise with defects can arise with just a single environmental exposure Acute Environmental Exposure to Lithium, occurring during gastrulation. Cardiac structural and func- Homocysteine, or Alcohol tional defects in mouse embryos we have shown can result with one-time exposure to the drug lithium, alcohol, or The causes of congenital heart disease (CHD) can be many, homocysteine, the latter a marker of folate deficiency [5, 18]. butitisgenerallyassumedthatthemajorityofmutationsis We analyzed different parameters of the cardiac cycle and not in the fetal organs but in the mother resulting in an altered umbilical blood flow using ultrasound (Vevo 770, Visual- environment for the developing embryo that predisposes to Sonics, Inc., Toronto, Ca, instrumentation) with a 40 mHz fetal malformations [32–36]. As changes in cellular processes transducer or a clinical instrument (Philips Sonos 5500, that relate to cell fate decisions and determination of cell Andover, MA) with a 12 mHz transducer. Both transducers lineages are thought to underlie many birth defects, it is demonstrated similar blood flow patterns. Maternal uterine important to have an understanding of the pathways involved artery blood flow velocity waveforms were obtained, and during specification and early differentiation stages ofthe the pulsatility index was calculated. Blood flow in the heart heart-placental axis. What has not been explored are the andbloodvesselswasdetectedineachembryousingtwo- etiological or simultaneous environmental modulation of dimensional real-time echocardiography. Color Doppler was embryonic cardiac and placental functional proteins and used to identify the embryonic heart and directed pulse developmental pathways. Little research has been done to Doppler to obtain blood flow velocity waveformsFigure ( 1). assay the biomarkers of CHD at a high-risk period of cardiac For venous hemodynamics, we used the presence of flow development, that is, during 3–6 weeks of human gestation, reversal during atrial contraction (A wave) in the ductus since most women do not know they are pregnant until after venosus and the presence of umbilical venous pulsations. The the abnormal placental/cardiac microenvironment already abnormal echocardiographic patterns relating to myocardial has had its deleterious effect on the early stages of the and valve defects have been published [17, 18]. Our research developing embryo. 4 Journal of Pregnancy

(a) (b)

Figure 2: Diagrams depicting the temporal sequence of vertebrate cardiomyocyte differentiation. (a) In the gastrula stage embryo, in the most undifferentiated part of the cardiogenic crescent 𝛽-catenin localization, an important intermediary of Wnt signaling appears at cell boundaries within the mesoderm. (b) Within this region soon round 𝛽-catenin-expressing cells are seen. (c) Near these round cells, pericardial coelomata (depicted as black areas) form, as cardiomyocytes change from a mesenchymal organization and elongate to form an epithelial, ventral cardiac compartment. A more ventral population not part of the epithelium now move more ventrally to a fibronectin-rich matrix (green), where these cells (orange) differentiate into endothelial endocardial cells. See Linask review,31 2003[ ].

It has been demonstrated that normal heart function development [24, 44–47],andinplacentation[48, 49]and relates to formation of normal cardiac structure [37, 38]. Mul- neurogenesis [50–52]. Alterations that occur in the placenta tiple mechanotransducing structures and molecules coordi- andintheheartfieldsinresponsetoacuteexposureof nate detection of blood flow forces with morphogenesis [39– mouse embryos on ED 6.0 at 6 PM, that is, ED 6.75, to Li, 42]. In our mouse studies, we observed that even one-time alcohol, or elevated homocysteine levels that correlate with environmental exposure during gastrulation (ED 6.75 to 8.0; changes ∼ a week later on ED15.5 in umbilical blood flow, extrapolating to human gestation, 16–19 days postconcep- abnormal heart (myocardial) function and formation, and tion) is associated with abnormal umbilical artery blood flow, valve anomalies in 63.2%, 86.6%, and 66.1% of the embryos, lower weight fetuses, and both structural and functional car- respectively [4, 5, 18]. In comparison to control embryos diac anomalies [4, 5, 18]. Although the mammalian embryo demonstrating normal cardiac development, there was a is well protected in the uterus, environmental chemicals, high incidence of semilunar valve abnormalities with our drugs, and maternal nutritional imbalances can interfere timing of the acute exposure. The one-time exposure at E6.75 with signaling pathways directing placental and embryonic corresponds to embryonic gastrulation, when the embryo development early in gestation. These environmental factors is comprised of only the three germ layers (Figure 2(a)), are thought now to cause at least 7% to 10% of all congenital the ectoderm, mesoderm, and endoderm. This is a time- anomalies [43]. Because biochemical differentiation precedes period during which the primary and second heart fields are morphological outcome often by daysFigure ( 2), the period being specified (Figures 2(b)–2(d)) in the bilateral anterior of susceptibility to environmental chemicals expectedly pre- mesoderm regions of the embryo [31, 53–57]. Both Hex cedes visible morphogenic effects. expression, an important inducer of primary heart field In our studies, we have focused on embryonic exposure specification and Islet-1,amarkerofthesecondheartfield to the drug lithium (Li) widely used for bipolar disorder, that leads to the development of the outflow [58, 59], were the metabolic intermediary homocysteine (HCy) in the folate suppressed with the acute environmental exposure [5, 18]. pathway, or alcohol (ethanol; EtOH). With the same timing of The delay in regulatory gene expression led to a delay inthe acute exposure, all three induce similar gene misexpression migration of cardiac precursors to the embryonic midline of Wnt-mediated genes Hex and Islet-1 in the cardiogenic resulting in a lethal condition known as cardiabifida and/or crescent during specification. Subsequently, all exposures to abnormal differentiation of part of the heart wall [4]. result in heart, valve, and placental abnormalities [4, 24]. Dependent on length of exposure and the dose, variability in The exposure effects appear to intersect with an early, critical heart development is seen in regards to degree of cardiabifida signaling pathway, the canonical Wnt/𝛽-catenin pathway, and the part of the myocardium that is affected (see Figure 3). important in the prechordal plate, in cardiac induction, valve Effects of exposure could lead to early embryonic demise, Journal of Pregnancy 5

Experimental: cardiabifida Control heart

(a) (b)

Figure 3: In contrast to the control embryo showing a looping, single tubular heart (right column of figure panels), a lithium-exposed embryo demonstrates a delay in the bilateral heart fields coming together at the midline, and thus a condition of cardiabifida is observed (left column of panels). The experimental, cardiabifida, heart also shows bilateral regions of the heart in which cardiomyocyte differentiation, asdefined by MF-20 staining for sarcomeric myosin heavy chain expression, has been suppressed (see areas delineated by white arrows). A–C shows sections through heart from anterior to posterior for the experimental heart. D. A higher magnification of boxed-in region in C is shown here. After the effect of acute environmental exposure dissipates, cardiomyocyte differentiation reappears posteriorly. E–H depict anterior to posterior sections of the control heart. Approximate regions of sections that cut through the heart are shown by lines in the whole mounts of the respective hearts on top.

to cardiac anomalies relating to the induction of valves or cardiovascular malformations [60]. Our published results the conduction system, or to eventual myocardial disease as on lithium exposure administered during gastrulation sug- adults. gested an additional possible basis for the high incidence of Defects in cardiac valves are the most common sub- valve defects with early embryonic exposure. We made the type of cardiac defects, and account for 25%–30% of all observation that early cell populations extrinsic to the heart 6 Journal of Pregnancy

Maternal decidua Spongiotrophoblast Labyrinth Fetal side

(a) (b) (c) (d)

(e) (f) (g) (h)

(i) (j) (k) (l)

(m) (n) (o) (p)

Figure 4: In comparison to placental cells of control embryo ((a)–(d)), one exposure of embryos to a binge-level of alcohol during gastrulation resulted in an upregulation of NMHC-IIB in all cell populations of the placenta more than a week after exposure, as observed at mid-gestation (embryonic day 15.5; (e)–(h)). Folate supplementation prevents the upregulation and maintains normal protein expression ((i)–(l)). Negative control is shown in bottom row ((m)–(p)). Figure from Han et al., 2012 [17].

fields arising during gastrulation, specifically the prechordal valve abnormalities, although there is evidence from patho- plate cells [61–64], were susceptible to environmental lithium logical analysis that early cell populations are contributing to exposure and failed to contribute in normal numbers to a cell valves [65–67]. Valve structural morphogenesis itself takes population that localizes to the ventral floor of the foregut place much later, that is, after a tubular heart has formed and that contributes to the dorsal mesenchymal protrusion [68, 69]. (DMP) region [65]. Kirby et al. showed that these cells eventually move into the endocardium [61]. Our lithium exposure studies suggested that this cell population within 4. Environmental Effects on the Placenta the DMP arises initially in the prechordal plate, and after migration into the endocardium may become associated with This same time period of embryonic exposure during gas- endocardial cushions contributing to valvulogenesis. If the trulation also relates to the formation of the placenta [15, prechordal plate cells are inhibited in their migration as they 16, 70].Werecentlypublishedthatexposuretolithium, are by lithium exposure, then cells in the Wnt-expressing homocysteine, or alcohol modulate gene and protein expres- valveregionsaremuchmoresparsethanseenincontrol sion in human HTR-8/SVneo extravillous trophoblasts in embryos and abnormal valvular function was observed in the culture and in vivo in the mouse placentas of the same experimental embryos [4]. Gastrulation is an earlier period of embryos that demonstrated cardiac anomalies [71]. Central development than generally has been considered to relate to totheplacenta’sfunctionisitsvascularlabyrinth,aparallel Journal of Pregnancy 7 series of finely branched blood vessels and trophoblasts that The folate pathway not only relates to purine and regulate nutrient, gas, and waste exchange while keeping the pyrimidine synthesis important in DNA synthesis and cell maternal-fetal blood supplies separate. After implantation, proliferation, but also to the synthesis of the primary angiogenic-vasculogenic processes promote formation of the methyl donor S-adenosylmethionine (SAM) important in highly arborized labyrinth vascular bed [72]. In addition to methylation reactions of cellular lipids, proteins, RNA, and the role of trophoblasts in spiral artery remodeling, migration DNA. DNA methylation is critical to epigenetic regulation of of trophoblasts in response to chemokines [73]isimportant gene expression [90–92]. Epigenetic factors that predispose in the colonization of the maternal decidua [74] between ED to congenital heart disease, and placental dysfunction are 7. 5 −10.5 of gestation [75].TheWntantagonistDickkopf-1’s suspectedtobethecauseofanincreaseintherecurrencerisk modulation of Wnt/𝛽-catenin signaling plays a critical role of CHD after one affected child. During cell differentiation, in regulating the equilibrium between active and inactive gene expression patterns are controlled through two principle Wnt signaling in the utero-placental interaction [76, 77]. mechanisms.Thebestunderstoodisimposedthroughthe Although human placentation differs from that of the rodent DNA sequence itself. The second or “epigenetic” relates model, there is homology regarding trophoblast invasion to heritable changes in gene function that occur indepen- and in the important regulatory elements [78]. We found dently of alterations in primary DNA sequence. The best- thatalcoholexposureaswellaslithiumandhomocysteine characterized epigenetic modifications are DNA methylation did not affect cell proliferation, but rather trophoblast cell and histone modifications, both of which function in Wnt migration [17]. All three environmental exposures mod- signaling, a critical pathway in early cardiomyogenesis [24, ulated placental nonmuscle-myosin-heavy-chain-(NMHC-) 44, 45, 47, 93], neurogenesis [50, 94–96], and placentation IIA and NMHC-IIB expression (Figure 4). NMHC-IIA has [48]. Vertebrate DNA methylation is in general restricted to beenreportedtohaveauniqueroleintheplacentaand cytosine (C) nucleotides in the sequence CG, known as CpG when deleted embryonic lethality ensued [79]. NMHC-IIB, islands. DNA methyltransferases (Dnmts) are the enzymes a protein associated with cell motility, also has a role in heart responsible for DNA methylation. A principle source of looping and trabeculation [39, 80, 81]. methyl groups in the cell is S-adenosyl methionine (SAM) synthesized by the FA metabolic cycle. In the early embryo, DNA after fertilization undergoes progressive demethyla- 5. Folate Protection of Normal tion and becomes hypomethylated during the pluripotential Heart-Placental Axis Formation and stages [97]. As cells differentiate, a more hypermethylated Prevention of Birth Defects stateisattained,ascells“lock-in”theirownspecificfate and other genes are silenced through methylation [98]. Thus, There is a general acceptance that folate aids in the prevention specificity of DNA methylation patterns is expected for of neural tube defects. There is an increasing evidence, specific cell types. However, recent evidence indicates that including our own, that folate supplementation can prevent neither all genes are methylated with differentiation nor all or reduce the risk and severity of congenital heart disease genes that are methylated are necessarily silenced. Epigenetics induced by an abnormal uterine microenvironment. Prenatal are important areas for analysis in relation to birth defects. treatment with folate has been shown to prevent neural tube Canonical Wnt signaling is necessary for the induction of defects and reduce the severity of CHD in clinical studies cardiogenesis [44] to occur, but if potentiated or prolonged [82–88]. In animal analyses using chick and mouse models, beyond when it is normally downregulated in the heart fields supplementation of folate or simultaneous supplementation by Wnt antagonists such as Dkk-1 and by Wnt 11, it becomes of folate and myo-inositol have been shown to prevent the inhibitory to differentiation [24, 47, 99]. In the human DKK- teratogenic effects of a number of environmental molecules 1 gene expression involves regulation by DNA methylation that can affect human gestation, including alcohol5 [ , 18]. In within its promoter [100]. Recent reports demonstrated human epidemiological studies, folate doses of 10 mg/kg have that alcohol exposure altered DNA methylation profiles in proven effective in preventing cardiovascular defects87 [ ]. mouse embryos at early neurulation [101]andthepatterning In our mouse studies, a metabolic weight dose equivalent of 5-methylcytosine expression during neurogenesis [102]. of 10.5 mg/kg maternal weight completely prevented cardiac Another study reported on a correlation between chronic defects induced during gastrulation [5, 18], but a more moderate dose of 6.2 mg/kg provided only partial protection. alcohol use, for example, and demethylation of normally Nodeleteriouseffectsonnormaldevelopmentwerenoted hypermethylated imprinted regions in sperm DNA [103]. with these doses. These results suggest that a slightly higher Thus, as based on our studies and that of others, analyses dose of folate may be necessary for the prevention of cardiac of one-carbon metabolism and specific epigenetic effects of birth defects than for neural tube defects where 1 mg/kg FA embryonic environmental exposure in the induction of pla- maybegivenforhighriskpregnancies.Amousestudyusinga cental and cardiac anomalies and of folate in the prevention folate dose 4 times higher than ours, that is, 40 mg/kg, showed of abnormal development are warranted. Given that a high that such a high dose was deleterious to normal embryonic percentage of pregnancies are unintended, the mechanism of development [89]. It appears that these investigators were action of the folate and intersecting pathways during early using a dose within a toxic range. It is suggested that more gestation together with prophylactic mechanisms involving clinical studies are needed to define an optimal FA dose for epigenetic effects are critical to define in the placenta-heart the prevention of CHD in human pregnancy. axis. 8 Journal of Pregnancy

There is an increasing awareness in the last 4 to 5 years [11] Y. Barak, M. C. Nelson, E. S. Ong et al., “PPAR𝛾 is required for that increased levels of attention are necessary to be brought placental, cardiac, and adipose tissue development,” Molecular to environmental exposures occurring in the first month of Cell,vol.4,no.4,pp.585–595,1999. pregnancy in the etiology of congenital heart defects [104]. [12] J. C. Cross and L. Mickelson, “Nutritional influences on implan- Epidemiological studies and statistics provide the basis for the tation and placental development,” Nutrition Reviews,vol.64, necessitytostudyearlyembryonicdevelopmentinrelationto no. 5, pp. S12–S18, 2006. the uterine environment and the formation of birth defects. [13] M. Hemberger and J. C. Cross, “Genes governing placental 120,000 to 160,000 children are born with major birth defects development,” Trends in Endocrinology and Metabolism,vol.12, eachyear(3%–5%ofalllivebirths).Twelvepercentof no. 4, pp. 162–168, 2001. admissions to pediatric hospitals relate to children with birth [14] J. Clark, K. E. Anderson, V. Juvin et al., “Quantification of defects [105]. Not withstanding the stress and toll of birth PtdInsP3 molecular species in cells and tissues by mass spec- trometry,” Nature Methods,vol.8,no.3,pp.267–272,2011. defects to the child and family, fifteen of the most significant birth defects have a cost to society of $8 billion annually. [15]S.J.Monkley,S.J.Delaney,D.J.Pennisi,J.H.Christiansen,and B. J. Wainwright, “Targeted disruption of the Wnt2 gene results Genetic mutations and chromosomal abnormalities account in placentation defects,” Development,vol.122,no.11,pp.3343– for approximately 28% of birth defects among affected indi- 3353, 1996. viduals [106]. More than 66% of birth defects, however, relate [16] S. Peng, J. Li, C. Miao et al., “Dickkopf-1 secreted by decidual to unknown, most likely multifactorial causes, including cells promotes trophoblast cell invasion during murine placen- environmental influences. Understanding how and when tation,” Reproduction,vol.135,no.3,pp.367–375,2008. in utero environmental exposures affect embryonic gene [17] M. Han, A. L. Neves, M. Serrano et al., “Effects of alcohol, regulation and which pathways are perturbed will provide lithium, and homocysteine on nonmuscle myosin-II in the insights into ways to prevent disease and birth defects by mouse placenta and human trophoblasts,” American Journal of reducing effects of exposures early in life, instead of treating Obstetrics and Gynecology, vol. 207, no. 2, pp. 140.e7–140.e19, the anomaly often by surgery or the disease when symptoms 2012. occur in adulthood. [18] M. Han, M. C. Serrano, R. Lastra-Vicente et al., “Folate rescues lithium-, homocysteine- and Wnt3A-induced vertebrate cardiac anomalies,” Disease Models and Mechanisms,vol.2,no.9- References 10, pp. 467–478, 2009. [19] K. K. Linask and J. Huhta, “Folate protection from congenital [1] B. A. Bailey and R. J. Sokol, “Pregnancy and alcohol use: evi- heart defects linked with canonical Wnt signaling and epige- dence and recommendations for prenatal care,” Clinical Obstet- netics,” Current Opinion in Pediatrics,vol.22,no.5,pp.561–566, rics and Gynecology,vol.51,no.2,pp.436–444,2008. 2010. [2] H. J. Bruyere Jr. and C. E. Stith, “Strain-dependent effect of [20] M. Serrano, M. Han, J. C. L. Huhta, R. Lastra, B. Garita, and ethanol on ventricular septal defect frequency in White K. K. Linask, “Folate suppresses Wnt/𝛽-catenin inhibition of Leghorn chick embryos,” Teratology,vol.48,no.4,pp.299–303, cardiogenesis to protect against induction of congenital heart 1993. defects,” in Proceedings of the American Heart Association [3]S.M.Smith,“Theavianembryoinfetalalcoholresearch,” Meeting, New Orleans, La, USA, 2008. Methods in Molecular Biology,vol.447,pp.75–84,2008. [21] M. C. Serrano, K. K. Linask, G. Acharya, J. Chen, M. Han, and [4] J. Chen, M. Han, S. M. Manisastry et al., “Molecular effects J. C. Huhta, “One-time lithium dose in early gestation causes of lithium exposure during mouse and chick gastrulation and placental and cardiac dysfunction,” American Journal of Obstet- subsequent valve dysmorphogenesis,” Birth Defects Research A, rics & Gynecology,vol.195,no.6,p.S210,2007. vol.82,no.7,pp.508–518,2008. [22] G. Acharya, T. Erkinaro, K. Makikallio,¨ T. Lappalainen, and J. Rasanen, “Relationships among Doppler-derived umbilical [5]M.Serrano,M.Han,P.Brinez,andK.K.Linask,“Fetalalcohol artery absolute velocities, cardiac function, and placental vol- syndrome: cardiac birth defects in mice and prevention with ume blood flow and resistance in fetal sheep,” American Journal folate,” American Journal of Obstetrics and Gynecology,vol.203, of Physiology,vol.286,no.4,pp.H1266–H1272,2004. no. 1, pp. 75.e7–75.e15, 2010. [23] G. Acharya, T. Wilsgaard, G. K. R. Berntsen, J. M. Maltau, and [6]M.T.Maurano,R.Humbert,E.Rynesetal.,“Systematiclocal- T. Kiserud, “Reference ranges for serial measurements of blood izationofcommondisease-associatedvariationinregulatory velocity and pulsatility index at the intra-abdominal portion, DNA,” Science,vol.337,pp.1190–1195,2012. and fetal and placental ends of the umbilical artery,” Ultrasound [7] D.J.P.Barker,“Theoriginsofthedevelopmentaloriginstheory,” in Obstetrics and Gynecology,vol.26,no.2,pp.162–169,2005. Journal of Internal Medicine,vol.261,no.5,pp.412–417,2007. [24]S.M.Manisastry,M.Han,andK.K.Linask,“Earlytemporal- specific responses and differential sensitivity to lithium and [8] D. Barker, “Human growth and cardiovascular disease,” Nestle Wnt-3A exposure during heart development,” Developmental Nutrition Workshop Series,vol.61,pp.21–33,2008. Dynamics,vol.235,no.8,pp.2160–2174,2006. [9] D. J. Barker, R. L. Bergmann, and P. L. Ogra, “Concluding [25] D. S. Walker, C. S. Fisher, A. Sherman, B. Wybrecht, and K. remarks. The window of opportunity: pre-pregnancy to 24 Kyndely, “Fetal alcohol spectrum disorders prevention: an months of age,” Nestle Nutrition Workshop Series,vol.61,pp. exploratory study of women’s use of, attitudes toward, and 255–260, 2008. knowledge about alcohol,” JournaloftheAmericanAcademyof [10] R. H. Adams, A. Porras, G. Alonso et al., “Essential role of p38𝛼 Nurse Practitioners,vol.17,no.5,pp.187–193,2005. MAP kinase in placental but not embryonic cardiovascular [26] W. Rozmus-Warcholinska, A. Wloch, G. Acharya et al., “Ref- development,” Molecular Cell,vol.6,no.1,pp.109–116,2000. erence values for variables of fetal cardiocirculatory dynamics Journal of Pregnancy 9

at 11–14 weeks of gestation,” Ultrasound in Obstetrics and expression of shear stress responsive genes KLF-2, ET-1, and Gynecology,vol.35,no.5,pp.540–547,2010. NOS-3 in the developing cardiovascular system of chicken [27] T. Kaukola, J. Ras¨ anen,R.Herva,D.D.Patel,andM.Hallman,¨ embryos,” Developmental Dynamics,vol.230,no.1,pp.57–68, “Suboptimal neurodevelopment in very preterm infants is 2004. related to fetal cardiovascular compromise in placental insuffi- [42] B. C. W. Groenendijk, B. P. Hierck, J. Vrolijk et al., “Changes in ciency,” American Journal of Obstetrics and Gynecology,vol.193, shear stress-related gene expression after experimentally altered no. 2, pp. 414–420, 2005. venous return in the chicken embryo,” Circulation Research,vol. [28]A.Miller,T.Riehle-Colarusso,C.J.Alverson,J.L.Fr´ıas, and A. 96, no. 12, pp. 1291–1298, 2005. Correa, “Congenital heart defects and major structural non- [43] K. L. Moore and T. V. N. Persaud, The Developing Human: cardiacanomalies,Atlanta,Georgia,1968to2005,”Journal of Clinically Oriented Embryology, Saunders, Philadelphia, Pa, Pediatrics,vol.159,no.1,pp.70–78.e2,2011. USA, 2003. [29] M. Han, A. L. Neves, M. Serrano et al., “Comparison of effects [44] C. Kwon, J. Arnold, E. C. Hsiao, M. M. Taketo, B. R. Conklin, of alcohol, lithium, and elevated homocysteine exposure on and D. Srivastava, “Canonical Wnt signaling is a positive nonmuscle myosin IIB expression in the mouse placenta and regulator of mammalian cardiac progenitors,” Proceedings of the human trophoblasts,” American Journal of Obstetrics & Gyne- National Academy of Sciences of the United States of America,vol. cology, vol. 207, no. 2, pp. 140.e7–140.e19. 104, no. 26, pp. 10894–10899, 2007. [30] S. Giles, P. Boehm, C. Brogan, and J. Bannigan, “The effects of [45] C. Kwon, K. R. Cordes, and D. Srivastava, “Wnt/𝛽-catenin ethanol on CNS development in the chick embryo,” Reproduc- signaling acts at multiple developmental stages to promote tive Toxicology,vol.25,no.2,pp.224–230,2008. mammalian cardiogenesis,” Cell Cycle,vol.7,no.24,pp.3815– [31] K. K. Linask, “Regulation of heart morphology: current molec- 3818, 2008. ular and cellular perspectives on the coordinated emergence of [46]K.K.Linask,K.A.Knudsen,andY.H.Gui,“N-Cadherin- cardiac form and function,” Birth Defects Research C,vol.69,no. catenin interaction: necessary component of cardiac cell com- 1, pp. 14–24, 2003. partmentalization during early vertebrate heart development,” [32] E. Goldmuntz, “The epidemiology and genetics of congenital Developmental Biology,vol.185,no.2,pp.148–164,1997. heart disease,” Clinics in Perinatology,vol.28,no.1,pp.1–10, [47] S. Ueno, G. Weidinger, T. Osugi et al., “Biphasic role for 2001. Wnt/𝛽-catenin signaling in cardiac specification in zebrafish [33]A.K.Nath,M.Krauthammer,P.Lietal.,“Proteomic-based and embryonic stem cells,” Proceedings of the National Academy detection of a protein cluster dysregulated during cardiovas- of Sciences of the United States of America,vol.104,no.23,pp. cular development identifies biomarkers of congenital heart 9685–9690, 2007. defects,” PLoS ONE,vol.4,no.1,ArticleIDe4221,2009. [48] J. Pollheimer, T. Loregger, S. Sonderegger et al., “Activation of [34]E.Bovill,S.Westaby,S.Reji,R.Sayeed,A.Crisp,andT.Shaw, the canonical wingless/T-cell factor signaling pathway pro- “Induction by left ventricular overload and left ventricular motes invasive differentiation of human trophoblast,” American failure of the human Jumonji gene (JARID2) encoding a protein Journal of Pathology, vol. 168, no. 4, pp. 1134–1147, 2006. that regulates transcription and reexpression of a protective [49] S. Sonderegger, H. Husslein, C. Leisser, and M. Knofler,¨ “Com- fetal program,” Journal of Thoracic and Cardiovascular Surgery, plex expression pattern of Wnt ligands and frizzled receptors in vol. 136, no. 3, pp. 709–716, 2008. human placenta and its trophoblast subtypes,” Placenta,vol.28, [35] J.Long,P.J.Lupo,E.Goldmuntz,andL.E.Mitchell,“Evaluation pp. S97–S102, 2007. of heterogeneity in the association between congenital heart defects and variants of folate metabolism genes: conotruncal [50] M. I. Garc´ıa-Castro, C. Marcelle, and M. Bronner-Fraser, “Ecto- and left-sided cardiac defects,” Birth Defects Research A,vol.91, dermal Wnt function as a neural crest inducer,” Science,vol.297, pp. 879–884, 2011. no.5582,pp.848–851,2002. [36] E. Goldmuntz, S. Woyciechowski, D. Renstrom, P. J. Lupo, and [51] A. K. Knecht and M. Bronner-Fraser, “Induction of the neural L. E. Mitchell, “Variants of folate metabolism genes and the risk crest: a multigene process,” Nature Reviews Genetics,vol.3,no. of conotruncal cardiac defects,” Circulation,vol.1,no.2,pp.126– 6, pp. 453–461, 2002. 132, 2008. [52] S. Ruffins and M. Bronner-Fraser, “A critical period for conver- [37] J.R.Hove,R.W.Koster,A.S.Forouhar,G.Acevedo-Bolton,S.E.¨ sion of ectodermal cells to a neural crest fate,” Developmental Fraser, and M. Gharib, “Intracardiac fluid forces are an essential Biology,vol.218,no.1,pp.13–20,2000. epigenetic factor for embryonic cardiogenesis,” Nature,vol.421, [53] K. L. Waldo, M. R. Hutson, C. C. Ward et al., “Secondary heart no.6919,pp.172–177,2003. field contributes myocardium and smooth muscle to the arterial [38] D. Sedmera, “Form follows function: developmental and physi- pole of the developing heart,” Developmental Biology,vol.281, ological view on ventricular myocardial architecture,” European no. 1, pp. 78–90, 2005. Journal of Cardio-thoracic Surgery, vol. 28, no. 4, pp. 526–528, [54] K. L. Waldo, D. H. Kumiski, K. T. Wallis et al., “Conotruncal 2005. myocardium arises from a secondary heart field,” Development, [39]B.Garita,M.W.Jenkins,M.Hanetal.,“Bloodflowdynamics vol. 128, no. 16, pp. 3179–3188, 2001. of one cardiac cycle and relationship to mechanotransduction [55] M. V. de la Cruz and C. Sanchez-Gomez, “Straight tube heart. and trabeculation during heart looping,” American Journal of Primitive cardiac cavities vs. primitive cardiac segments,” in Physiology, vol. 300, no. 3, pp. H879–H891, 2011. Living Morphogenesis of the Heart,M.V.delaCruzandR. [40] K. K. Linask and M. Vanauker, “A role for the cytoskeleton in R. Markwald, Eds., pp. 85–98, Birkhauser, Boston, Mass, USA, heart looping,” The Scientific World Journal, vol. 7, pp. 280–298, 1998. 2007. [56] R. R. Markwald, T. Nakaoka, and C. H. Mjaatvedt, “Overview: [41] B.C.W.Groenendijk,B.P.Hierck,A.C.Gittenberger-DeGroot, formation of the primary heart tube,” in Etiology and Morpho- and R. E. Poelmann, “Development-related changes in the genesisofCongenitalHeartDisease:TwentyYearsofProgressin 10 Journal of Pregnancy

Genetics and Developmental Biology,E.B.Clark,M.Nakazawa, [72]J.Cross,M.Hemberger,Y.Lu,T.Nozaki,andK.Whitely, andA.Takao,Eds.,pp.103–114,Futura,NewYork,NY,USA, “Trophoblast functins, angiogenesis and remodeling of the 2000. maternal vasculatrue in the placenta,” Molecular and Cellular [57]K.Linask,Y.H.Gui,R.Rasheed,andL.Kwon,“Patterndevel- Endocrinology, vol. 187, pp. 207–212, 2002. opment during pericardial coelom formation and specification [73] N. J. Hannan and L. A. Salamonsen, “Role of chemokines in the of the cardiomyocyte cell population by N-cadherin and the endometrium and in embryo implantation,” Current Opinion in Drosophila armadillo protein homologue in the early chick Obstetrics and Gynecology,vol.19,no.3,pp.266–272,2007. embryo,” Molecular Biology of the Cell,vol.3,p.206A,1992. [74] L. K. Harris, “Review: trophoblast-vascular cell Interactions in [58] C. L. Cai, X. Liang, Y. Shi et al., “Isl1 identifies a cardiac pro- early pregnancy: how to remodel a vessel,” Placenta,vol.31,pp. genitor population that proliferates prior to differentiation and S93–S98, 2010. contributes a majority of cells to the heart,” Developmental Cell, [75]S.Afonso,L.Romagnano,andB.Bablarz,“Theexpressionand vol.5,no.6,pp.877–889,2003. function of cystatin C and cathepsin B and cathepsin L during [59] M. P. Verzi, D. J. McCulley, S. De Val, E. Dodou, and B. L. mouse embryo implantation and placentation,” Development, Black, “The right ventricle, outflow tract, and ventricular vol.124,no.17,pp.3415–3425,1997. septum comprise a restricted expression domain within the [76] A. Cattelino, S. Liebner, R. Gallini et al., “The conditional secondary/anterior heart field,” Developmental Biology,vol.287, inactivation of the 𝛽-catenin gene in endothelial cells causes no.1,pp.134–145,2005. a defective vascular pattern and increased vascular fragility,” [60] C. A. Loffredo, “Epidemiology of cardiovascular malforma- Journal of Cell Biology, vol. 162, no. 6, pp. 1111–1122, 2003. tions: prevalence and risk factors,” American Journal of Medical [77] O. A. Mohamed, M. Jonnaert, C. Labelle-Dumais, K. Kuroda, Genetics,vol.97,pp.319–325,2000. H.J.Clarke,andD.Dufort,“UterineWnt/𝛽-catenin signaling is [61] M. L. Kirby, A. Lawson, H. A. Stadt et al., “Hensen’s node required for implantation,” Proceedings of the National Academy gives rise to the ventral midline of the foregut: implications of Sciences of the United States of America,vol.102,no.24,pp. for organizing head and heart development,” Developmental 8579–8584, 2005. Biology,vol.253,no.2,pp.175–188,2003. [78] S. Caluwaerts, L. Vercruysse, C. Luyten, and R. Pijnenborg, [62] E. M. Pera and E. M. De Robertis, “A direct screen for secreted “Endovascular trophoblast invasion and associated structural proteins in Xenopus embryos identifies distinct activities for the changes in uterine spiral arteries of the pregnant rat,” Placenta, Wnt antagonists crescent and frzb-1,” Obstetrics and Gynecology, vol.26,no.7,pp.574–584,2005. vol.96,no.3,pp.183–195,2000. [79] M. A. Conti, S. Even-Ram, C. Liu, K. M. Yamada, and R. S. [63]A.F.Schier,S.C.F.Neuhauss,K.A.Helde,W.S.Talbot,andW. Adelstein, “Defects in cell adhesion and the visceral endoderm Driever, “The one-eyed pinhead gene functions in mesoderm following ablation of nonmuscle myosin heavy chain II-A in and endoderm formation in zebrafish and interacts with no tail,” mice,” Journal of Biological Chemistry,vol.279,no.40,pp. Development,vol.124,no.2,pp.327–342,1997. 41263–41266, 2004. [64] R. Seifert, M. Jacob, and H. J. Jacob, “The avian prechordal head region: a morphological study,” Journal of Anatomy,vol.183, [80] A. N. Tullio, D. Accili, V. J. Ferrans et al., “Nonmuscle myosin part1,pp.75–89,1993. II-B is required for normal development of the mouse heart,” Proceedings of the National Academy of Sciences of the United [65]A.Wessels,R.H.Anderson,R.R.Markwaldetal.,“Atrial States of America,vol.94,no.23,pp.12407–12412,1997. development in the human heart: an immunohistochemical study with emphasis on the role of mesenchymal tissues,” The [81]W.Lu,S.H.Seeholzer,M.Hanetal.,“Cellularnonmuscle Anatomical Record, vol. 259, pp. 288–300, 2000. myosins NMHC-IIA and NMHC-IIB and vertebrate heart [66] R. H. Anderson, S. Webb, N. A. Brown, W. Lamers, and A. looping,” Developmental Dynamics,vol.237,no.12,pp.3577– Moorman, “Development of the heart: (3) Formation of the 3590, 2008. ventricular outflow tracts, arterial valves, and intrapericardial [82] S. Cains, A. Shepherd, M. Nabiuni, P. J. Owen-Lynch, and J. arterial trunks,” Heart, vol. 89, no. 9, pp. 1110–1118, 2003. Miyan, “Addressing a folate imbalance in fetal cerebrospinal [67]B.S.Snarr,E.E.Wirrig,A.L.Phelps,T.C.Trusk,andA. fluid can decrease the incidence of congenital hydrocephalus,” Wessels, “A spatiotemporal evaluation of the contribution of Journal of Neuropathology and Experimental Neurology,vol.68, the dorsal mesenchymal protrusion to cardiac development,” no. 4, pp. 404–416, 2009. Developmental Dynamics,vol.236,no.5,pp.1287–1294,2007. [83] L. M. De-Regil, A. C. Fernandez-Gaxiola, T. Dowswell, and J. P. [68] J. Lincoln, A. W. Lange, and K. E. Yutzey, “Hearts and bones: Pena-Rosas, “Effects and safety of periconceptional folate sup- shared regulatory mechanisms in heart valve, cartilage, tendon, plementation for preventing birth defects,” Cochrane Database and bone development,” Developmental Biology,vol.294,no.2, of Systematic Reviews, vol. 10, Article ID CD007950, 2010. pp. 292–302, 2006. [84] A. Rosano, D. Smithells, L. Cacciani et al., “Time trends in [69] J. Lincoln and K. E. Yutzey, “Molecular and developmental neural tube defects prevalence in relation to preventive strate- mechanisms of congenital heart valve disease,” Birth Defects gies: an international study,” Journal of Epidemiology and Research A, vol. 91, no. 6, pp. 526–534, 2011. Community Health,vol.53,no.10,pp.630–635,1999. [70] J. S. Fitzgerald, A. Germeyer, B. Huppertz et al., “Governing the [85]S.Thompson,M.Torres,R.Stevenson,J.Dean,andR.Best, invasive trophoblast: current aspects on intra- and extracellular “Periconceptional vitamin use, dietary folate and occurrent regulation,” American Journal of Reproductive Immunology,vol. neural tube defected pregnancies in a high risk population,” 63, no. 6, pp. 492–505, 2010. Annals of Epidemiology,vol.10,no.7,p.476,2000. [71]L.Carrel,P.A.Hunt,andH.F.Willard,“Tissueandlineage- [86]R.Ionescu-Ittu,A.J.Marelli,A.S.Mackie,andL.Pilote,“Pre- specific variation in inactive X chromosome expression of the valence of severe congenital heart disease after folic acid murine Smcx gene,” Human Molecular Genetics,vol.5,no.9, fortification of grain products: time trend analysis in Quebec, pp. 1361–1366, 1996. Canada,” The British Medical Journal,vol.338,p.b1673,2009. Journal of Pregnancy 11

[87] A. Czeizel, “Reduction of urinary tract and cardiovascular [104] L. Burd, E. Deal, R. Rios, E. Adickes, J. Wynne, and M. G. Klug, defects by periconceptional multivitamin supplementation,” “Congenital heart defects and fetal alcohol spectrum disorders,” American Journal of Medical Genetics,vol.62,pp.179–183,1996. Congenital Heart Disease,vol.2,no.4,pp.250–255,2007. [88] A. E. Czeizel, E. H. Puho,´ Z. Langmar, N. Acs,´ and F. Banhidy,´ [105]P.W.Yoon,R.S.Olney,M.J.Khoury,W.M.Sappenfield, “Possible association of folic acid supplementation during G. F. Chavez, and D. Taylor, “Contribution of birth defects pregnancy with reduction of preterm birth: a population- and genetic diseases to pediatric hospitalizations: a population- based study,” European Journal of Obstetrics Gynecology and based study,” Archives of Pediatrics and Adolescent Medicine,vol. Reproductive Biology,vol.148,no.2,pp.135–140,2010. 151, no. 11, pp. 1096–1103, 1997. [89] L. Pickell, K. Brown, D. Li et al., “High intake of folic acid dis- [106]W.J.Larsen,Human Embryology, Churchill Livingstone, New rupts embryonic development in mice,” Birth Defects Research York, NY, USA, 1993. A,vol.91,no.1,pp.8–19,2011. [90]O.Aguilera,A.F.Fernandez,A.Munoz,andM.F.Fraga, “Epigenetics and environment: a complex relationship,” Journal of Applied Physiology,vol.109,no.1,pp.243–251,2010. [91] V.Bollati and A. Baccarelli, “Environmental epigenetics,” Hered- ity,vol.105,no.1,pp.105–112,2010. [92] L. M. Villeneuve and R. Natarajan, “The role of epigenetics in thepathologyofdiabeticcomplications,”American Journal of Physiology,vol.299,no.1,pp.F14–F25,2010. [93] A. C. Foley and M. Mercola, “Heart induction by Wnt antagonists depends on the homeodomain transcription factor Hex,” Genes and Development,vol.19,no.3,pp.387–396,2005. [94] M. Broun, L. Gee, B. Reinhardt, and H. R. Bode, “Formation of the head organizer in hydra involves the canonical Wnt pathway,” Development,vol.132,no.12,pp.2907–2916,2005. [95] C. Carmona-Fontaine, H. K. Matthews, S. Kuriyama et al., “Contact inhibition of locomotion in vivo controls neural crest directional migration,” Nature,vol.456,no.7224,pp.957–961, 2008. [96] M. A. Chacon,´ L. Varela-Nallar, and N. C. Inestrosa, “Frizzled- 1 is involved in the neuroprotective effect of Wnt3a against𝛽 A oligomers,” Journal of Cellular Physiology,vol.217,no.1,pp.215– 227, 2008. [97] T. M. Nafee, W. E. Farrell, W. D. Carroll, A. A. Fryer, and K. M. K. Ismail, “Epigenetic control of fetal gene expression,” British Journal of Obstetrics and Gynaecology ,vol.115,no.2,pp.158– 168, 2008. [98] J. C. Kiefer, “Epigenetics in development,” Developmental Dynamics,vol.236,no.4,pp.1144–1156,2007. [99] S. Gessert and M. Kuhl,¨ “The multiple phases and faces of Wnt signaling during cardiac differentiation and development,” Circulation Research,vol.107,no.2,pp.186–199,2010. [100] O. Aguilera, M. F. Fraga, E. Ballestar et al., “Epigenetic inactivation of the Wnt antagonist DICKKOPF-1 (DKK-1) gene in human colorectal cancer,” Oncogene,vol.25,no.29,pp.4116– 4121, 2006. [101]Y.Liu,Y.Balaraman,G.Wang,K.P.Nephew,andF.C.Zhou, “Alcohol exposure alters DNA methylation profiles in mouse embryos at early neurulation,” Epigenetics,vol.4,no.7,pp.500– 511, 2009. [102] F. C. Zhou, Y. Chen, and A. Love, “Cellular DNA methylation program during neurulation and its alteration by alcohol exposure,” Birth Defects Research A,vol.91,no.8,pp.703–715, 2011. [103] L. A. Ouko, K. Shantikumar, J. Knezovich, P. Haycock, D. J. Schnugh, and M. Ramsay, “Effect of alcohol consumption on CpG methylation in the differentially methylated regions of H19 and IG-DMR in male gametes—implications for fetal alcohol spectrum disorders,” Alcoholism,vol.33,no.9,pp.1615–1627, 2009. Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 139049, 7 pages http://dx.doi.org/10.1155/2013/139049

Clinical Study Prevalence of Metabolic Syndrome One Year after Delivery in Finnish Women at Increased Risk for Gestational Diabetes Mellitus during Pregnancy

Jatta Puhkala,1 Tarja I. Kinnunen,2 Tommi Vasankari,1,3 Katriina Kukkonen-Harjula,1 Jani Raitanen,1,2 and Riitta Luoto1,3

1 UKK Institute for Health Promotion Research, P.O. Box 30, FI-33501 Tampere, Finland 2 School of Health Sciences, University of Tampere, FI-33014 University of Tampere, Finland 3 NationalInstituteforHealthandWelfare,P.O.Box30,FI-00271Helsinki,Finland

Correspondence should be addressed to Jatta Puhkala; [email protected]

Received 20 November 2012; Accepted 25 February 2013

Academic Editor: Leena Hilakivi-Clarke

Copyright © 2013 Jatta Puhkala et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. Women with a history of gestational diabetes mellitus (GDM) are at increased risk for metabolic syndrome (MeS) after delivery. We studied the prevalence of MeS at one year postpartum among Finnish women who in early pregnancy wereat increased risk of developing GDM. Methods. This follow-up study is a part of a GDM prevention trial. At one year postpartum, 2 150 women (mean age 33.1 years, BMI 27.2 kg/m ) were evaluated for MeS. Results. The prevalence of MeS was 18% according tothe International Diabetes Federation (IDF) criteria and 16% according toNational Cholestrol Education Program (NCEP) criteria. Of MeS components, 74% of participants had an increased waist circumference (≥80 cm). Twenty-seven percent had elevated fasting plasma glucose (≥5.6 mmol/L), and 29% had reduced HDL cholesterol (≤1.3 mmol/L). The odds ratio for the occurrence of MeS at one year postpartum was 3.0 (95% CI 1.0–9.2) in those who were overweight before pregnancy compared to normal weight women. Conclusions. Nearly one-fifth of the women with an increased risk of GDM in early pregnancy fulfilled the criteria of MeS at one year postpartum. The most important factor associated with MeS was prepregnancy overweight. Weight management before and during pregnancy is important for preventing MeS after delivery.

1. Introduction complications in pregnancy [10]. Depending on the population and the diagnostic criteria used, the prevalence is Metabolic syndrome (MeS) is defined as a cluster of athero- roughly 1%–14% of pregnancies [10, 11]; and the occurrence sclerotic risk factors, including abdominal obesity, elevated is increasing worldwide [12, 13]. The most important risk serum triglycerides, decreased HDL cholesterol, elevated factors for GDM are prepregnancy overweight, high maternal blood pressure, and elevated serum plasma glucose [1–3]. age and a family history of type 2 diabetes [14]. Women Insulin resistance is a central feature in the pathogenesis with a history of GDM are at increased risk of developing of MeS [4] in addition to an unhealthy diet and physical type2diabetesandalsoMeSafterdelivery[15–17]. Among inactivity promoting overweight and genetic factors [1, 5– CanadianwomenwithahistoryofGDM,theprevalence 7]. As obesity increases worldwide, this leads to an increased of MeS was 20% at as early as three months postpartum incidence and an earlier onset of MeS [3, 8, 9]. [18]. According to studies from the USA and Denmark, Gestational diabetes mellitus (GDM), a disorder in glu- approximately 30%–40% of women with a history of GDM cose and insulin metabolism, is one of the most common develop MeS by ten years postpartum [19, 20]. 2 Journal of Pregnancy

The aim of this study was to determine the prevalence and a two-hour OGTT was performed at 8–12 and 26–28 of MeS and its components at one year postpartum among weeks’ gestation and one year postpartum. All blood analysis Finnish women who in early pregnancy were at increased risk was performed at the UKK Institute. For glucose and lipid of developing GDM. A secondary aim was to characterize risk analysis, venous blood was drawn into citric acid/fluoride factors associated with the development of MeS. andEDTAtubes.DuringtheOGTT,bloodsampleswere taken between 60 and 120 min after the participants had drunk 75 g glucose in 330 mL water (Glucodyn, Ultimed, 2. Subjects and Methods Finland). Plasma glucose concentrations were measured fresh within 24 hours after the OGTT, but plasma samples ∘ The study is a part of a cluster-randomized controlled for lipid analysis were stored frozen at −80 Cuntilanalysed. trial, NELLI (counseling and lifestyle during pregnancy, Glucose, total cholesterol, HDL cholesterol, and triglyceride ISRCTN33885819) [21]. A detailed description of the design concentrations were determined in enzymatic assays using a and methods has been published previously [22]. The pri- Roche Cobas Mira Plus analyser. All testing was performed in maryaimofthetrialwastopreventGDMamongpregnant duplicate. MeS was diagnosed according to the International women who were assessed in early pregnancy to have an Diabetes Federation (IDF) [6]andtheNationalCholesterol increased risk of GDM. The study was conducted in primary Education Program Adult Treatment Panel (NCEP ATP- health care maternity clinics in Western Finland in 2007– III) [23] criteria. At one year postpartum, oral glucose 2009. The intervention included structured individual coun- tolerance was evaluated according to the American Diabetes selling on weight gain, diet, and physical activity by public Association (ADA) [24] and the World Health Organization health nurses during five routine visits to maternity clinics. (WHO) [25] criteria. The primary outcome of this study The women in the control clinics received the usual maternal was the prevalence of MeS and its components at one year care, including some lifestyle advice. postpartum. Pregnant women were recruited by nurses at their first The background characteristics and descriptive informa- visit (8–12 weeks’ gestation) in maternity clinics. Women tion on components of metabolic syndrome are reported here were eligible if they had at least one of the following GDM as means and standard deviations (SDs) or frequencies and risk factors: age ≥ 40 years, prepregnancy body mass index proportions. A multivariate logistic regression model was 2 (BMI) ≥ 25 kg/m , GDM or any sign of glucose intolerance, used to obtain odds ratios (ORs) and their 95% confidence amacrosomicbaby(≥4500 g) in any previous pregnancy, intervals (95% CIs) to study associations between metabolic or diabetes in first- or second-degree relatives. The main syndrome and its explanatory variables. Explanatory vari- exclusion criteria were age < 18 years, a GDM diagnosis at 8– ables included were age (continuous), group (intensified 12 weeks’ gestation, twin pregnancy, physical restrictions that counselling, usual care, and early GDM); and five GDM risk precluded exercise, or a clinical history of chronic disease. A factors (as used in entrance criteria to the study, that is, BMI 2 diagnosis of GDM was based on a two-hour 75-gram oral ≥ 25 kg/m ,age≥ 40 years, GDM or any sign of glucose glucose tolerance test (OGTT) whose results met at least intolerance in any previous pregnancy, a macrosomic baby one of the following criteria: a fasting plasma glucose of [≥4500 g] in any previous pregnancy, and diabetes in first- or ≥5.3 mmol/L, >10.0 mmol/L at one hour, or >8.6 mmol/L at second-degree relatives). The results were considered to be two hours [14]. statistically significant if 𝑃 < 0.05.Allanalysiswasperformed Six hundred forty pregnant women participated in base- with SPSS software (version 20.0). line assessment at 8–12 weeks’ gestation (Figure 1). Of them, 442 (69%) were eligible for the randomized clinical trial (RCT; intensified counselling or usual care), while 198 (31%) 3. Results 𝑛 = 174 were excluded, most of them ( , 88%) due to a GDM 3.1. Background Characteristics. Before pregnancy, self- diagnosis at 8–12 weeks’ gestation. At postpartum followup, reported weight was 74.2 kg (range 50.0–120.0 kg), which MeS component data were available for 150 women. The was 0.9 kg less than the weight measured at the first maternal intensified counselling, the usual care, and the early GDM clinical visit at 8–12 weeks’ gestation (75.1 kg, 𝑛 = 127). (originally excluded from the RCT) groups were merged for 2 The mean prepregnancy BMI was 26.7 kg/m (range 18.1– present analysis. 2 Information on maternal measurements was obtained 39.5 kg/m ). At followup measurement; one year after delivery, the mean weight increase was 1.4 kg (range from from the standard maternity cards. Height was measured − at the first maternity care visit, and weight was measured 16.1 to 18.0 kg) compared with prepregnancy weight. Table 1 at each maternity care visit and one year postpartum. shows that one year after delivery, the mean age of the women Waist circumference was measured (the average of three was 33.1 years (range 20–49 years) and the mean number of measurements) at one year postpartum. Blood pressure was deliveries was 2.0 (range 1–8). The most common inclusion criteria for the study were prepregnancy overweight (66%) measured in duplicate at each maternity care visit and one 𝑛=30 year postpartum. Because 15% of the weight data from the and diabetes in relatives (53%). Twenty-one percent ( ) first visit were missing, self-reported prepregnancy weight smoked frequently or occasionally before pregnancy. wasusedasthebaselineweight. Blood specimens were taken for glucose, cholesterol, 3.2. Metabolic Syndrome and Its Components at One-Year HDL cholesterol, and triglyceride analysis after a 12-hour fast, Postpartum. The prevalence of MeS and its components Journal of Pregnancy 3

All Intensified counselling Usual care Early GDM 640 343 women agreed to participate 297 women agreed to participate - 16 had a miscarriage before - 8 had a miscarriage before baseline measurements baseline measurements 1 1 - 81 had abnormal OGTT at baseline - 93 had abnormal OGTT at baseline

1 616 174 had abnormal OGTT at baseline 2 and received usual care of GDM

442 246 received intensified counselling 196 received usual care - 21 did not participate in assessments - 8 did not participate in assessments at 26–28 weeks’ gestation at 26–28 weeks’ gestation - 6 had a miscarriage - 8 had a miscarriage

399 219 finished allocated counselling 180 finished allocated usual care

150 56 participated in follow-up 62 participated in follow-up 32 participated in follow-up measurements at 1 year postpartum measurements at 1 year postpartum measurements at 1 year postpartum

1 OGTT: oral glucose tolerance test 2 GDM: gestational diabetes mellitus

Figure 1: Flow diagram of the study, ending in followup assessments at one year postpartum. The two intervention groups (intensified counselling and usual care) and those with GDM diagnosed at early pregnancy (8–12 weeks’ gestation) were invited for followup measurements at one year postpartum.

oneyearafterdeliveryinallwomenandintheintensified compared to the intensified counselling group (OR 3.4, counselling, usual care, and abnormal OGTT groups is 95% CI 1.0–11.3, 𝑃 = 0.051)(Table 3). When analysed presented in Table 2. Three out of four women exceeded the by trial inclusion criteria (GDM risk factors at baseline), 2 waist circumference limit of 80 cm, and about half reached prepregnancy overweight (BMI > 25 kg/m )wasastrong the limit of 88 cm. Compared to the intensified counselling predictor for developing MeS (OR 3.0, 95% CI 1.0–9.2, 𝑃= group, there tended to be more abdominally obese (waist 0.053). circumference ≥ 88 cm) women in the early GDM and usual care group. More than one-fourth of all and half of the 3.3. Dropout Analyses. Measurements of MeS components at women with early GDM had elevated fasting plasma glucose ≥ one year postpartum were available for 24% of the women ( 5.6 mmol/L) at one year postpartum. One-fifth of all and whoparticipatedinthebaselinemeasurementsandfor32% one-fourth of women with early GDM also had elevated ≤ of those who participated in the followup. Compared with blood pressure. HDL cholesterol was reduced ( 1.3 mmol/L) the participants for whom the data for MeS diagnosis at among more than one-fourth of all women. followup were not available (𝑛 = 466), participants with On the other hand, almost one-third (31%) according to data for MeS (𝑛 = 150) were more likely to belong to NCEP criteria and one-sixth (16%) according to IDF criteria the usual care group (55% versus 45%, 𝑃 = 0.032)and did not meet the criteria for any MeS components. Most were less likely to be frequent smokers before pregnancy women (52% according to IDF and 63% according to NCEP) (10% versus 22%, 𝑃 = 0.013). There were also some more had one or two MeS components, while 8% (according to women with GDM diagnosed at 26–28 weeks’ gestation both IDF and NCEP) had four or five components. According (abnormal OGTT) among women with MeS data available to IDF criteria, the prevalence of MeS among the participants atoneyearpostpartumthanamongthosewithoutMeSdata 𝑛=27 was 18% ( ) and according to NCEP criteria 16% (18% versus 10%, 𝑃 = 0.064), but there were no differences 𝑛=24 ( ). In OGTT, at followup, 8 women (5%) had impaired in the occurrence of GDM at 8–12 weeks’ gestation (20% glucose tolerance according to both WHO and ADA cri- versus 24% had early GDM, 𝑃 = 0.393). Neither were teria. Seven had MeS according to both IDF and NCEP there differences between these two groups in any other criteria. background characteristic (weight or BMI, age, parity, or In multivariate logistic regression analysis, the risk of education) or laboratory analysis other than OGTT at 26–28 MeS in the group with early GDM tended to be higher weeks’ gestation. 4 Journal of Pregnancy

Table 1: Background characteristics of women at one year postpar- In 2007, the prevalence of obesity among women aged tum. Means and standard deviations or frequencies (and propor- 25–34 years in Finland was 11.1% [28]; in our subjects, one tions) of participants (𝑛 = 150). year after delivery, it was 2.5 times higher at 27%. Half of ± the subjects were abdominally obese (waist circumference Age (years) 33.1 4.9 ≥ <30 37 (25) 88 cm). Abdominal obesity is one of the most important 30–34 60 (40) independent factors in development of metabolic syndrome ≥35 53 (35) [29]. Although overweight among young adults is on the rise Weight (kg) 75.6 ± 15.3 worldwide [30], there are few studies of the prevalence of MeS BMIa (kg/m2)27.2± 5.0 among young adults. BMI ≥ 25 kg/m2 88 (64) The NELLI study [21] is one of the largest random- BMI ≥ 30 kg/m2 37 (27) ized controlled trials about preventing the development Education of gestational diabetes. The NELLI trial showed that the lifestyle counselling was effective in controlling the propor- Basic or secondary education 48 (32) tion of large-for-gestational-age newborns and improving Polytechnic education 60 (40) the women’s diet and had a minor effect on gestational University degree 41 (28) weight gain and decrease in physical activity [21, 31–33]. Parity 2.0 ± 1.2 Since participation in followup measurements at one year 156(37) postpartum was low (24% of the original cohort) and the 2-3 82 (55) total number of participants was modest, the results must be ≥412(8) interpreted carefully. Because of low number of participants, GDMb risk criteria (at 8–12 weeks’ pregnancy, 𝑛 = 148) some findings were only borderline statistically significant. BMIa ≥ 25 kg/m2 97 (66) There are many possible reasons for the loss. Among Macrosomic child in any previous pregnancy 8 (5) followup participants, there were more women from the usual GDMb in any previous pregnancy 25 (17) care group and fewer frequent smokers before pregnancy. Diabetes in first- or second-degree relatives 78 (53) Participants in followup also were more likely to have been Age ≥ 40 years 6 (4) diagnosed with GDM in midpregnancy (26–28 weeks) as aBMI: body mass index. b compared with the women who did not participate in MeS GDM: gestational diabetes mellitus. testing at followup. Women with small children may have founditdifficulttofindtimetocomeinfortesting,especially as it included the two-hour OGTT. Some subjects who 4. Discussion showed up for followup testing may have been more health conscious, which is advocated by the fact that there was a At one year postpartum, MeS was diagnosed in nearly one- smaller proportion of smokers among them. Some may have fifth of the women who had an increased risk of gestational been concerned about ill health due to their GDM diagnosis diabetes at the beginning of their pregnancy. Especially; during pregnancy. Nevertheless, some women with a GDM prepregnancy overweight was associated with a higher risk of diagnosis were already being monitored by the healthcare developing MeS. In a prospective Finnish population study system, and they may have refused to participate in our among nonpregnant women aged 30–33, the prevalence of testing for that reason. Another reason for refusal was a new MeS was 7%–14% according to the IDF definition and 4%– pregnancy, but the number of these women was unclear. This 11% according to the NCEP definition [26]. The prevalence study was limited to Finnish women, and the results can only of MeS was more common in our study due to our risk- beextrapolatedtoCaucasianpopulations. group approach. According to other studies, the prevalence One limitation of this study is that, at the followup, the of MeS among parous women aged 30–40 is no more than subjects were not queried about hormonal contraception or 10% depending on country, criteria, and time after delivery medications for hypertension or dyslipidemia, which influ- [18, 19, 27]. ence the components of MeS. However, when medications Our study is one of the first followup studies on the were queried in the baseline, none of the women reported prevalence of MeS after delivery among young women with taking any cardiovascular medication. risk factors for GDM. According to earlier studies, a history We used self-reported weight before pregnancy, which of GDM is strongly associated with a higher prevalence of is not as reliable a measurement as using a scale. When MeS [16, 20, 27]. The present study suggests that the risk self-reported weight was compared to the available weight is high, also among women with risk factors for GDM in measurement at the first maternity clinical visit, the mean early pregnancy, but does not necessarily lead to GDM. GDM difference was less than 1 kg, which could easily be explained andMeSsharesomeriskfactors,suchasoverweightand by a minimal early pregnancy weight gain. a genetic tendency towards impaired glucose metabolism. Still, the development of MeS after glucose disorders during pregnancy has not been given as much attention as the 5. Conclusion increased risk of type 2 diabetes after GDM. In any case, the cardiometabolic risk factors in women at increased risk of Our study suggests that MeS at one year postpartum seems GDMshouldalsobefollowedafterdelivery. to occur more often among women who in early pregnancy Journal of Pregnancy 5

Table 2: Components of metabolic syndrome (MeS) and prevalence of MeS by two criteria in all women and in the intensified counselling, usual care, and abnormal OGTT groups at one year postpartum. Means and standard deviations or frequencies (and proportions) of participants.

All Intensified counselling Usual care Early GDMa (𝑛 = 135–150) (𝑛=49–56) (𝑛=56–62) (𝑛=30–32) Waist circumference (cm) 88.8 ± 11.4 86.7 ± 11.5 88.8 ± 10.2 92.7 ± 12.7 Waist ≥ 80 cm 102 (74) 36 (69) 43 (77) 23 (77) Waist ≥ 88 cm 69 (50) 20 (39) 30 (54) 19 (63) Fasting glucose (mmol/L) 5.4 ± 0.4 5.3 ± 0.4 5.3 ± 0.4 5.7 ± 0.5 Fasting glucose ≥ 5.6 mmol/L 43 (29) 11 (20) 16 (26) 16 (50) Systolic pressure (mmHg) 116 ± 11 114 ± 9116± 12 118 ± 12 Diastolic pressure (mmHg) 74 ± 972± 776± 10 75 ± 10 Blood pressure ≥ 130 or ≥85 mmHg 23 (17) 4 (8) 12 (21) 7 (23) HDL cholesterol (mmol/L) 1.50 ± 0.34 1.43 ± 0.30 1.57 ± 0.37 1.49 ± 0.30 HDL cholesterol ≤ 1.3 mmol/L 40 (27) 17 (32) 14 (23) 9 (28) Triglycerides (mmol/L) 0.97 ± 0.43 0.93 ± 0.37 0.94 ± 0.44 1.06 ± 0.49 Triglycerides ≥ 1.7 mmol/L 12 (8) 3 (6) 4 (7) 5 (16) b Metabolic syndrome (IDF) 27 (18) 6 (11) 11 (18) 10 (31) c Metabolic syndrome (NCEP) 24 (16) 4 (7) 10 (16) 10 (31) aGDM: gestational diabetes mellitus. bInternational Diabetes Federation. cNational Cholesterol Education Adult Treatment Panel III.

Table 3: Occurrence of metabolic syndrome by its explanatory variables (group, age, and five GDM risk factors). Odds ratios (ORs), 95% confidence intervals (CIs), and 𝑃 values, 𝑛 = 150. Multivariate logistic regression model.

OR (95% CI) 𝑃 value Group (reference = the intensified counselling) Usual care 1.5 (0.5 to 4.6) 0.48 Early GDMa 3.36 (1.00 to 11.4) 0.051 Age (continuous) 1.0 (0.9 to 1.2) 0.44 BMIb ≥ 25 kg/m2 (prepregnancy) 3.0 (1.0 to 9.2) 0.053 A macrosomic baby (≥4500 g) in any previous pregnancy 0.9 (0.1 to 6.5) 0.91 GDMa or any sign of glucose intolerance in any previous pregnancy 2.6 (0.9 to 7.6) 0.077 Type 1 or 2 diabetes in first- or second-degree relatives 1.6 (0.6 to 4.00) 0.32 Age ≥ 40 years 0.9 (0.1 to 11.8) 0.92 aGDM: gestational diabetes. bBMI: body mass index. had an increased risk of GDM. The most important factor Conflict of Interests associated with MeS seemed was prepregnancy overweight. This study suggests that especially women with an increased The authors have no conflict of interests. riskofGDMshouldbefolloweduponforcardiometabolic risk factors after delivery. Weight management or reduction before pregnancy and prevention of excessive weight gain during pregnancy are important for the prevention of GDM Acknowledgments and of MeS. Overweight and obesity among pregnant women may increase, as the average maternal age is rising along with This study was funded by the Competitive Research Funding the obesity epidemic, and this represents an even greater of the Tampere University Hospital, the Juho Vainio Foun- challenge in following up on and managing risk factors for dation, and the Finnish Ministry of Education and Culture. chronic diseases. There is a need for larger population studies are thankful to Tiina Solakivi, Ph.D., Associate Professor; on the prevalence of MeS among young women, especially at the University of Tampere, who was responsible for the among those who are at an elevated risk of GDM. laboratory testing. 6 Journal of Pregnancy

References [17] B. Akinci, A. Celtik, S. Genc et al., “Evaluation of postpartum carbohydrate intolerance and cardiovascular risk factors in [1] M.A.Pereira,T.E.Kottke,C.Jordan,P.J.O’Connor,N.P.Pronk, women with gestational diabetes,” Gynecological Endocrinology, and R. Carreon,´ “Preventing and managing cardiometabolic vol. 27, no. 5, pp. 361–367, 2011. risk: the logic for intervention,” International Journal of Environ- [18] R. Retnakaran, Y. Qi, P. W. Connelly, M. Sermer, B. Zinman, mental Research and Public Health,vol.6,no.10,pp.2568–2584, and A. J. G. Hanley, “Glucose intolerance in pregnancy and 2009. postpartum risk of metabolic syndrome in young women,” The [2] B. Balkau, P. Valensi, E. Eschwege,andG.Slama,“Areviewof` Journal of Clinical Endocrinology and Metabolism,vol.95,no.2, the metabolic syndrome,” Diabetes and Metabolism,vol.33,no. pp. 670–677, 2010. 6, pp. 405–413, 2007. [19] A. Verma, C. M. Boney, R. Tucker, and B. R. Vohr, “Insulin [3] S. M. Grundy, “Metabolic syndrome pandemic,” Arteriosclero- resistance syndrome in women with prior history of gestational sis, Thrombosis, and Vascular Biology,vol.28,no.4,pp.629–636, diabetes mellitus,” The Journal of Clinical Endocrinology and 2008. Metabolism,vol.87,no.7,pp.3227–3235,2002. [4] D. Lann and D. LeRoith, “Insulin resistance as the underlying [20] J. Lauenborg, E. Mathiesen, T. Hansen et al., “The prevalence of causeforthemetabolicsyndrome,”Medical Clinics of North the metabolic syndrome in a Danish population of women with America,vol.91,no.6,pp.1063–1077,2007. previous gestational diabetes mellitus is three-fold higher than [5]M.R.Carnethon,C.M.Loria,J.O.Hill,S.Sidney,P.J.Savage, in the general population,” The Journal of Clinical Endocrinology and K. Liu, “Risk factors for metabolic syndrome: the Coronary and Metabolism, vol. 90, no. 7, pp. 4004–4010, 2005. Artery Risk Development in Young Adults (CARDIA) study, [21] R. Luoto, T. I. Kinnunen, M. Aittasalo et al., “Primary preven- 1985–2001,” Diabetes Care, vol. 27, no. 11, pp. 2707–2715, 2004. tion of gestational diabetes mellitus and large-for-gestational- [6]K.G.M.M.Alberti,R.H.Eckel,S.M.Grundyetal.,“Harmo- age newborns by lifestyle counseling: a cluster-randomized nizing the metabolic syndrome: a joint interim statement of the vontrolled trial,” PLoS Medicine,vol.8,no.5,ArticleID International Diabetes Federation Task Force on Epidemiology e1001036, 2011. and Prevention; National Heart, Lung, and Blood Institute; [22] R. M. Luoto, T. I. Kinnunen, M. Aittasalo et al., “Prevention of American Heart Association; World Heart Federation; Inter- gestational diabetes: design of a cluster-randomized controlled national Atherosclerosis Society; and International Association trial and one-year follow-up,” BMC Pregnancy and Childbirth, for the Study of Obesity,” Circulation,vol.120,no.16,pp.1640– vol. 10, article 39, 2010. 1645, 2009. [7]C.L.Edwardson,T.Gorely,M.J.Daviesetal.,“Association [23] J. I. Cleeman, “Executive summary of the third report of of sedentary behaviour with metabolic syndrome: a meta- the National Cholesterol Education Program (NCEP) expert analysis,” PLoS ONE,vol.7,no.4,ArticleIDe34916,2012. panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III),” Journal of the [8]E.S.Ford,W.H.Giles,andA.H.Mokdad,“Increasing American Medical Association,vol.285,no.19,pp.2486–2497, prevalence of the metabolic syndrome among U.S. adults,” 2001. Diabetes Care, vol. 27, no. 10, pp. 2444–2449, 2004. [24] American Diabetes Association, “Standards of medical care in [9] G. Hu, J. Lindstrom,¨ P. Jousilahti et al., “The increasing diabetes—2012,” Diabetes Care, vol. 35, supplement 1, pp. S11– prevalence of metabolic syndrome among finnish men and S63. women over a decade,” The Journal of Clinical Endocrinology and Metabolism,vol.93,no.3,pp.832–836,2008. [25] Definition and Diagnosis of Diabetes Mellitus and Intermediate Hyperglycemia: Report of a WHO/IDF Consultation,World [10] American Diabetes Association, “Gestational diabetes mellitus,” Health Organization, 2006. Diabetes Care, vol. 27, supplement 1, pp. S88–S90, 2004. [11] S. Schneider, C. Bock, M. Wetzel, H. Maul, and A. Loerbroks, [26] J. R. H. Raiko, J. S. A. Viikari, A. Ilmanen et al., “Follow-ups of “The prevalence of gestational diabetes in advanced economies,” the cardiovascular risk in young Finns study in 2001 and 2007: Journal of Perinatal Medicine,vol.40,no.5,pp.511–520,2012. levels and 6-year changes in risk factors,” Journal of Internal Medicine, vol. 267, no. 4, pp. 370–384, 2010. [12] D. Dabelea, J. K. Snell-Bergeon, C. L. Hartsfield, K. J. Bischoff, R. F. Hamman, and R. S. McDuffie, “Increasing prevalence [27] B. Akinci, A. Celtik, S. Yener, and S. Yesil, “Prediction of devel- ofgestationaldiabetesmellitus(GDM)overtimeandby oping metabolic syndrome after gestational diabetes mellitus,” birth cohort: Kaiser Permanente of Colorado GDM screening Fertility and Sterility,vol.93,no.4,pp.1248–1254,2010. program,” Diabetes Care, vol. 28, no. 3, pp. 579–584, 2005. [28] M. Peltonen, K. Harald, S. Mannist¨ oetal.,“TheNational¨ [13] A. Ferrara, “Increasing prevalence of gestational diabetes mel- FINRISK 2007 Study. Statistics,” Publications of the National litus: a public health perspective,” Diabetes Care,vol.30,no.2, Public Health Institute B35/2008, 2008. pp. S141–S146, 2007. [29] S. A. Ritchie and J. M. C. Connell, “The link between abdom- [14] American Diabetes Association, “Diagnosis and classification inal obesity, metabolic syndrome and cardiovascular disease,” of diabetes mellitus,” Diabetes Care, vol. 29, supplement 1, pp. Nutrition, Metabolism and Cardiovascular Diseases,vol.17,no. S43–S48, 2006. 4, pp. 319–326, 2007. [15] S. Bo, L. Monge, C. Macchetta et al., “Prior gestational hyper- [30] M. M. Finucane, G. A. Stevens, M. J. Cowan et al., “National, glycemia: a long-term predictor of the metabolic syndrome,” regional, and global trends in body-mass index since 1980: Journal of Endocrinological Investigation,vol.27,no.7,pp.629– systematic analysis of health examination surveys and epi- 635, 2004. demiological studies with 960 country-years and 9⋅1 million [16]S.D.Sullivan,J.G.Umans,andR.Ratner,“Gestationaldia- participants,” The Lancet,vol.377,no.9765,pp.557–567,2011. betes: implications for cardiovascular health,” Current Diabetes [31] T. Kinnunen, J. Puhkala, J. Raitanen et al., “Effects of dietary Report,vol.12,no.1,pp.43–52,2012. counselling on food habits and dietary intake of Finnish Journal of Pregnancy 7

pregnant women at increased risk for gestational diabetes— a secondary analysis of a cluster-randomized controlled trial,” Maternal & Child Nutrition,2012. [32] T. I. Kinnunen, M. Pasanen, M. Aittasalo et al., “Preventing excessive weight gain during pregnancy—a controlled trial in primary health care,” European Journal of Clinical Nutrition,vol. 61,no.7,pp.884–891,2007. [33]M.Aittasalo,J.Raitanen,T.I.Kinnunen,K.Ojala,P.Kolu,and R. Luoto, “Is intensive counseling in maternity care feasible and effective in promoting physical activity among women at risk for gestational diabetes? Secondary analysis of a cluster randomized NELLI study in Finland,” The International Journal of Behavioral Nutrition and Physical Activity,2012. Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 787032, 6 pages http://dx.doi.org/10.1155/2013/787032

Research Article Limiting Excess Weight Gain in Healthy Pregnant Women: Importance of Energy Intakes, Physical Activity, and Adherence to Gestational Weight Gain Guidelines

Tamara R. Cohen and Kristine G. Koski

School of Dietetics and Human Nutrition, Macdonald Campus, McGill University, 21,111 Lakeshore Road, Ste-Anne-de-Bellevue, QC,CanadaH9X3V9 Correspondence should be addressed to Kristine G. Koski; [email protected]

Received 13 November 2012; Accepted 16 January 2013

Academic Editor: Riitta Luoto

Copyright © 2013 T. R. Cohen and K. G. Koski. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Few studies have investigated if compliance with energy intakes, physical activity, and weight gain guidelines attenuate postpartum weight retention (PPWR) in mothers attending prenatal classes. We investigated whether (a) daily energy intakes within 300 kcal of estimated energy requirements (EERs), (b) walking more than 5000 steps/day, (c) targeting the recommended weight gain goals for prepregnancy BMI, and/or (d) achieving weekly or total gestational weight gain (GWG) recommendations minimized PPWR in 54 women attending prenatal classes in Montreal/Ottawa, Canada. Participants completed a validated pregnancy physical activity questionnaire (PPAQ), 3 telephone-validated 24-hr dietary recalls, and wore a pedometer for one week. PPWR was measured 6 weeks after delivery. Results showed that 72% had healthy prepregnancy BMIs. However, 52% consumed >300 kcal/day in excess of their EER, 54% exceeded recommended GWG, and more overweight (93%) than normal weight women (38%) cited nonrecommended GWG targets. Following delivery, 33% were classified as overweight, and 17% were obese. Multiple logistic regressions revealed that women targeting “recommended weight gain advice” were 3 times more likely to meet total GWG recommendations (OR: 3.2, 𝑃 < 0.05); women who complied with weekly GWG goals minimized PPWR (OR: 4.2, 𝑃 < 0.02). In conclusion, appropriate GWG targets, lower energy intakes, and physical activity should be emphasized in prenatal education programs.

1. Introduction when compared against recommendations released in 1999 [9, 10]orthenewerrevised2010guidelines[11, 12], with nearly Research describes increasing rates of obesity in women of 50% exceeding gestational weight gain (GWG) guidelines. childbearing age [1, 2]. With more than 40% gaining in excess The Canadian GWG guidelines [12]wereharmonizedwith of Institute of Medicine recommendations [1], pregnancy the US guidelines [13, 14]in2010. is now emerging as an important risk factor for excessive Despite the existence of GWG guidelines since the 1990’s weight gain [3] and an important target for obesity prevention [10, 15], recent studies show that most women either receive studies [4]. Although most studies have focussed on multi- no GWG advice [16–22] or target nonrecommended GWG ethnic, socioeconomically disadvantaged obese women [1, 5– goals [18, 23, 24]. Moreover, women no longer cite health 7], there is a growing concern that many healthy, university- professionals as their main source of information but identify educated, nonobese women may also gain excess weight the internet and family and friends as main sources of during pregnancy, leading to postpartum weight retention information [6, 24–27]. (PPWR)andobesitylaterinlife[3, 8]. In Canada, national Because targeted goals [28], dietary and exercise inter- statistics from the Maternity Experiences Study show that ventions [29], and physical activity (PA) levels [30]have primiparous, university-educated women with medium- to been associated with limiting excessive GWG, our objec- high-household incomes gain more than is recommended tives were (1) to investigate in healthy nonobese women 2 Journal of Pregnancy attending prenatal classes in Ottawa/Montreal, Canada, if one discussed weight gain with me.” Responses reported as Health Canada’s GWG targets [12] were being followed by ranges (e.g., between 11.3 and 15.9 kg) were recorded as the our college-educated prenatal class attendees and (2) to mean of a weight range category. The “recommended advice” identify specific behaviours that might be associated with was calculated using each women’s individual prepregnancy achieving current Health Canada’s GWG recommendations BMI and was compared to the following Health Canada and and/or minimizing postpartum weight retention (PPWR). Institute of Medicine recommendations: BMI < 18.5: 12.5– We investigated the likelihood of achieving a healthy GWG 18 kg; BMI = 18.5–24.9: 11.5–16 kg; BMI = 25.0–29.9: 7–11.5 kg; and/or minimizing PPWR for the following four behaviours: BMI ≥ 30: 5–9 kg [12, 13]. (a) daily energy intakes within 300 kcal of estimated energy requirements (EERs); (b) walking more than 5000 steps/day; 2.3. Weight Assessment. Self-reported pre-pregnancy weight (c) targeting Health Canada’s recommended GWG goals and height were obtained at the first home visit as previously based on the mother’s prepregnancy body mass index (BMI); described [24]; actual body weight using a Tanita scale was and (d) achieving weekly or total GWG recommendations alsomeasuredatthefirsthomevisit.WeeklyGWGwascal- established by Health Canada [12] for healthy nonobese culated using current pregnancy weight minus prepregnancy mothers. weight (kg) divided by week of gestation minus twelve, as previously reported [24]. Self-reported pre-pregnancy weight 2. Methods was defined as weight at conception, and gestational age was based on date of last menstruation. Women were also 2.1. Recruitment. From 18 prenatal classes held in either telephoned at delivery and 6-week postpartum and were Ottawa,Ontario,orMontreal,Quebec,betweenAugust asked to report both their measured weight recorded by the and December 2008, 142 women were approached in their physician at the time of delivery and their 6-wk postpartum second and third trimesters to participate in this study. weight recorded at this routine doctor’s visit. This 6-wk time Ethics approvals were obtained from McGill University, point has also been previously used to represent the maximal Ottawa Public Health Ethics Board, and Centre de Sante´ fat mass gained during pregnancy [32–34]. Specifically, this de Services Sociaux (CSSS) Montreal (West Island and 6-wk postpartum measurement is considered a valid early Cavendish boards). Researchers who were trained clinical indicator of adipose tissue accumulation during pregnancy nutritionists briefly described the study elements in a five- because, at this time point, maternal weight is no longer influ- minute presentation at the beginning of each prenatal class, enced by changes in blood volume arising from pregnancy and interested clients provided their contact information at and/or planned weight loss. PPWR was calculated based on the end of class. In Canada, prenatal classes are voluntary and the difference between this 6-wk postpartum weight and the encouragedforfirst-timeparentstolearnaboutallaspectsof women’s self-reported pre-pregnancy weight. pregnancy, delivery, and how to care for your newborn. These sessions are offered in both English and French and are free of charge. 2.4. Physical Activity and Dietary Intake. During a home visit, Signed consent was obtained for 81 mothers; 54 mothers mothers were instructed on how to wear the pedometer by completed all phases of the study that included a “GWG a certified sports nutritionist. PA was assessed using a New advice” questionnaire, 3 24-hr dietary recalls obtained by Lifestyles Digi-Walker SW-200 [Step Into Health, Plainfield, phone on nonconsecutive days, and wearing a pedometer for IL, USA] pedometer for one week. Women also completed a 7consecutivedaysduringthesameweekinwhichthedietary PA logbook that included wear time, total steps/day and total records were completed; 27 were excluded because they did time bathing, swimming or napping times, and how these not complete 3 dietary recalls or wear the pedometer for 7 compared with pedometer values. Values were compared consecutive days. Final inclusion criteria were for women >12 to public health recommendations where less than 5000 weeks of pregnancy, free of medical risks for PA, as described steps/day is classified as a sedentary lifestyle [35]. in the Physical Activity Readiness Medical Examination for Women also completed 3 nonconsecutive telephone Pregnancy (PARmed-X for Pregnancy) [31], and women who dietaryrecallsduringtheweektheyworethepedometer. 2 Dietary interview kits and training were provided to assist were not underweight (prepregnancy BMI < 18 kg/m )or 2 women with estimating food portion sizes during the tele- obese (prepregnancy BMI > 30 kg/m )anddidnothavea phonerecalls.TheCanadianNutrientFile2007[36]and multiple pregnancy, which is a contraindication as per the ESHA Research Food Processor (version 9.1) (Salem, OR, PARmed-X. USA) were used to analyze food recalls for total energy (kcals) [37]. These were compared to the estimated energy 2.2. Provider Advice Questionnaire. During the first visit, requirements (EERs) (kcal/day) [37]. EER were calculated women were interviewed by the principle researcher, a using the formula from the Dietary Reference Intake which licensed dietitian, on sources of GWG advice. Categories estimates the EER based on age, PA level, height, weight, included physician and/or other health professionals, family and the additional requirement associated with pregnancy. andfriends,andinternetorbooks.Theamountofweight For all energy calculations, weights measured during the eachpregnantwomenwasadvisedtogainwasalsorecorded. first home visit were used as previously described [24]. In Responses were categorized into the following weight ranges: Canada, in contrast to USA, no increased energy intakes <6.8 kg, 9.1–10.9 kg, 11.3–13.1 kg, 13.6–15.9 kg, >15.9 kg, or “no are recommended for the first trimester, but an additional Journal of Pregnancy 3 )

𝑛=54 2 Table 1: Population characteristics ( ). 34 32 Characteristic (mean (SD)) 30 ± 28 Age (y) 32.0 4.3 26 Height (m) 1.7 ± 0.1 24 Gestational age (weeks) 26.8 ± 6.3 22 20

Pregnancy weight assessments Body mass index (kg/m 18 2 a ± Prepregnancy body mass index (kg/m ) 23 3 Prepregnancy BMI Estimated energy requirements (kcal/day)b 2341 ± 151 Postpartum BMI Average steps per day (steps/day)c 6133 ± 2203 Figure 1: Comparison of Individual Mother’s Pre-Pregnancy with d ± Average energy expenditure (MET-hrs/day) 6.6 2.6 their Post-Partum BMI. The differences in the individual weight Total gestational weight gain (kg) 17.1 ± 6.4 gains for our 54 mothers are described. The shaded box represents 2 Weekly gestational weight gain (kg) 0.71 ± 0.44 the normal BMI category (18.8–24.9 kg/m ). Postpartum weight assessments Weight retention at 6-wk postpartum (kg)e 10.9 ± 4.5 < 2 > ± Postpartum body mass index (kg/m2)25± 4 25 kg/m (54%) had retained 4.5 kg (mean 5.9 4.9 kg), a whereas 60% of overweight women (pre-pregnancy BMI = Based on self-reported weight. 2 > ± bCalculated using self-reported prepregnancy weight, self-reported height, 25.0–29.9 kg/m )hadretained 4.5 kg (mean 8.2 6.0 kg). age, and physical activity level. Thus, although 72% began pregnancy having a normal pre- cSteps per day determined by 7-day wear time of pedometer. pregnancy BMI, based on their 6-week postpartum weight, dMET-hrs/day determined by the pregnancy physical activity questionnaire 50% of study mothers were now classified as overweight or [24]. obese (𝑃 > 0.001)(Figure 1). eWeight retention at 6-wk postpartum was determined by subtracting self- reported pre-pregnancy weight from the 6-wk value measured at the time of their routine 6-wk postpartum doctor’s visit. 3.2. Information Sources. The majority (76%) received advice about GWG: 49% from books/internet, 29% from physicians, 10% from other health care professionals (dietitian, nurse, 340 kcal is recommended for second trimester and an addi- and midwife), and 10% referenced all 3 sources. Twenty four tional 450 kcal for the third trimester [10, 12]; these later two percent of all study participants reported receiving no GWG values were added to the following equation: EER = 354 − advice. None cited their prenatal course as a source for their (6.91 × age[y])+PA×{(9.36 × weight [kg]) + (726 × height targeted GWG. Those who received practitioner advice or [m])} [38]. obtained information from books or the internet most often cited 13.6–15.9 kg as their targeted normal healthy GWG 2.5. Statistical Analyses. Data was analyzed using Statistical regardless of prepregnancy BMI. AnalysisSoftware[Version9.2,2002-2003,SASInstituteInc, Cary, NC, USA]. Univariate logistic regressions were used to 3.3. Adherence to “Recommended” GWG Targets. The major- compute odds ratios (OR) for (a) achieving recommended ity (61%) of study participants followed incorrect GWG total GWG and (b) carrying less than 4.5 kg (10 lbs) of advice for their prepregnancy BMI. As well, 52% exceeded additional weight 6-week postpartum for each of our 4 recommended weekly rates of GWG, normal prepregnancy behaviors. Statistical significance was 𝑃 < 0.05. BMI averaged 0.6 ± 0.2 kg/week; overweight pre-pregnancy BMIaveraged0.7± 0.3 kg/week. However, univariate logistic regressions showed that women who followed the correct 3. Results total weight gain for their prepregnancy BMI were three timesmorelikelytoachieveHealthCanada’sGWGrecom- 3.1. Population Characteristics. Table 1 describes our pop- 𝑃 < 0.05 ulation characteristics. Participants were mostly Caucasian mendations (OR: 3.2, )(Table 2). Women who (85%), nulliparous (79%), and married (72%) women, with achieved their recommended weekly rate of GWG were four times more likely to have retained less than 4.5 kg at 6- college (74%) or university degrees (26%) and household 𝑃 < 0.02 incomes >$50,000 (82%) who maintained “low active” week postpartum (OR: 4.2, )(Table 2). Finally, < 2 lifestyles (6120 ± 2185 steps/day, range 840–11090 steps/day). women with a pre-pregnancy BMI 25 kg/m were nine Postpartum assessment revealed that the majority delivered times more likely to target and to achieve Health Canada healthy infants at term (39.3 ± 1.5 weeks) (3450 ± 494 g) either recommendations for GWG of 11.5–16 kg (Table 2). byvaginaldelivery(68%)orplannedcaesareansection(32%). Energy intakes ranged from 1080 kg to 3760 kcal/day (mean 4. Discussion 2240 ± 448); 54% exceeded their EER, individualized for their trimester, by more than 300 kcal per day. Total GWG Pregnancy is now considered obesogenic [3], but preventing ranged from 7.5 kg–35 kg (mean 17.3 kg). By 6 weeks post- excessive weight gain is proving to be a challenge [28–30]. partum, the majority of women with a pre-pregnancy BMI In our study, we explored compliance with 3 measurable 4 Journal of Pregnancy

Table 2: Odds ratios identifying behaviours associated with achieving a healthy GWG and minimizing PPWR in pregnant women (𝑛=54) attending prenatal classes.

Odds ratio 95% Confidence interval 𝑃 value Behaviours associated with achieving an appropriate total gestational weight gain (GWG) Normal prepregnancy BMI (18.5–24.9 kg/m2) 9.6 1.88–48.99 0.0065 Consuming within 300 kcal/day of EER 1.1 0.36–3.16 0.9005 Walking >5,000 steps/day 1.4 0.35–3.78 0.8271 Following “correct” total weight gain guidelines 3.2 1.04–9.85 0.0426 Behaviours associated with achieving <4.5 kg (10 lbs) postpartum weight retention (PPWR) Normal pre-pregnancy BMI (18.5–24.9 kg/m2) 1.3 0.38–4.31 0.6839 Consuming within 300 kcal/day of EER 1.3 0.45–3.84 0.6257 Walking >5,000 steps/day 1.1 0.34–3.74 0.8385 Following “correct” total weight gain guidelines 1.3 0.45–3.84 0.6257 Achieving Health Canada’s recommended average weekly GWG rate (kg/week) 4.2 1.33–13.27 0.0147 behaviouralobjectivesandtheirimpactonGWGandPPWR attending prenatal classes who are not advised to meet GWG in women attending prenatal classes. GWG exceeded Health recommendations are at an increased risk of exceeding weight Canada recommendations in 52% of our study population, gain recommendations and of retaining in excess of 4.5 kg which is similar to other Canadian studies [9, 11, 39]. Our postpartum if they are sedentary throughout pregnancy as resultsshowedthatmeetingGWGtargetswereassociated our study did. with the following modifiable conditions and behaviours. Strengths of our study include the fact that we measured 2 First, a normal weight BMI < 25 kg/m increased the like- energy intakes using 3 24-hr recalls and PA by both pedome- lihood of mothers complying with Health Canada GWG ters and a validated PPAQ [24]. Our previous modeling guidelines, but a normal prepregnancy BMI did not prevent paper had identified interrelated causal pathways among excessive PPWR, suggesting that behaviours associated with energy intakes and PA and pregnancy outcomes that were GWG and PPWR were not related to one another, which has not supported by direct relationships [33]. It is possible that been suggested [33]; secondly in both situations—for exam- increased PA and lower energy intakes were not associated ple, achieving an appropriate GWG or avoiding PPWR— with GWG and PPWR, which might be explained by the neither energy intakes within 300 kcal of a mother’s EER or sedentary behavior of our mothers coupled with their high becoming active and walking more than 5000 steps per day energy intakes. As well, pedometers are considered an accept- directly increased the likelihood of mothers achieving GWG able method of assessing PA in pregnant women [42], their recommendations or avoiding PPWR. Previous research has sensitivitytotiltanglemaybemorepronouncedaspregnancy shown that purposeful walking, as measured by pedometer progresses and could affect recording of steps [43], thus steps per day is related to weekly rate of GWG [24, 40], underestimating PA. whereas higher energy intakes predict PPWR [33]. Thirdly Despiteoursmallsamplesize,webelieveourfindings our evidence strongly points to the important role for emphasizeanimportantroleforprenatalclassestoeducate understanding and targeting recommended GWG guidelines pregnant women. Development of a public health promotion and the importance of health care professionals conveying the strategy for women attending prenatal classes should empha- message. It is well established that provider advice strongly size correct GWG based on the mother’s prepregnancy BMI, impacts what expectant mothers actually gain [16, 22]. Should appropriate energy intakes, and a nonsedentary lifestyle. In no advice or inappropriate weight gain advice be followed, this study, university-educated women who received and previous studies show that women will exceed their GWG followed the correct weight advice were more likely to avoid recommendations [9, 16, 22, 24]. Only 61% followed GWG excessive GWG, and those who complied with weekly GWG guidelines for their prepregnancy BMI established by Health which was achieved through “walking” [40], minimized Canada. Others have reported lower rates [9, 11, 39]. PPWR, but none described their prenatal classes as having Our GWG results were greater than a recent Australian these objectives. This study demonstrates that by providing study that reported that only 30% of normal weight women correct GWG targets in prenatal classes early in pregnancy exceeded Institute of Medicine recommendations [41]. Simi- more women can more easily achieve the Health Canada and larly, recent Canadian studies identified 47% of primiparous Institute of Medicine recommendations for GWG and avoid mothers and 43% and 38% of college- and university- PPWR. educated women, respectively, had excessive GWG; there were no differences for mothers from low (43%) versus References high (41%) household incomes [9]. A second study reported [1] E. M. Davis, K. C. Stange, and R. I. Horowitz, “Childbearing, nearly 50% exceeded GWG, but they found an association stress and obesity disparities in women: a public health perspec- with income, ethnicity, and health status of the mother [11]. tive,” Maternal and Child Health Journal,vol.16,pp.109–118, Taken together, these studies show that pregnant women 2012. Journal of Pregnancy 5

[2]J.YehandJ.A.Shelton,“Increasingprepregnancybodymass about weight gain during pregnancy,” American Journal of index: analysis of trends and contributing variables,” American Obstetrics and Gynecology,vol.205,no.333,pp.e1–e6,2011. Journal of Obstetrics and Gynecology,vol.193,no.6,pp.1994– [19] S. D. McDonald, E. Pullenayegum, K. Bracken et al., “Com- 1998, 2005. parison of midwifery, family medicine, and obstetric patients’ [3] K. Melzer and Y. Schutz, “Pre-pregnancy and pregnancy pre- understanding of weight gain during pregnancy: a minority of dictors of obesity,” International Journal of Obesity,vol.34,no. women report correct counseling,” JournalofObstetricsand 2, pp. S44–S52, 2010. Gynaecology Canada,vol.34,no.2,pp.129–135,2012. [4] R. L. Atkinson, A. Pietrobelli, R. Uauy, and I. A. Macdonald, [20] N. E. Stotland, P. Gilbert, A. Bogetz, C. C. Harper, B. Abrams, “Are we attacking the wrong targets in the fight against obesity? and B. Gerbert, “Preventing excessive weight gain in pregnancy: The importance of intervention in women of childbearing age,” how do prenatal care providers approach counseling?” Journal International Journal of Obesity,vol.36,pp.1259–1260,2012. of Women’s Health,vol.19,no.4,pp.807–814,2010. [5]S.J.Herring,D.B.Nelson,A.Daveyetal.,“Determinantsof [21] N. Stotland, Y. Teoh, and B. Gerbert, “Prenatal weight gain: who excessive gestational weight gain in urban, low-income women,” is counseled?” Journal of Women’s Health,vol.21,no.6,pp.695– Women’s Health Issues,vol.22,no.5,pp.e439–e446,2012. 701, 2012. [6]M.R.Stengel,J.L.Kraschnewski,S.W.Hwangetal.,“Whatmy [22]S.Phelan,M.G.Phipps,B.Abrams,F.Darroch,A.Schaffner, doctor didn’t tell me: examining health care provider advice to and R. R. Wing, “Practitioner advice and gestational weight overweight and obese pregnant women on gestational weight gain,” Journal of Women’s Health,vol.20,no.4,pp.585–591,2011. gain and physical activity,” Women’s Health Issues,vol.22,no.6, [23] S.J.Herring,D.N.Platek,P.Elliott,L.E.Riley,A.M.Stuebe,and pp. e535–e540, 2012. E. Oken, “Addressing obesity in pregnancy: what do obstetric [7] A. F. L. Bogaerts, R. Devlieger, E. Nuyts et al., “Effects of providers recommend?” Journal of Women’s Health,vol.19,no. lifestyle intervention in obese pregnant women on gestational 1,pp.65–70,2010. weight gain and mental health: a randomized controlled trial,” [24] T.R.Cohen,H.Plourde,andK.G.Koski,“AreCanadianwomen International Journal of Obesity,pp.1–8,2012. achieving a fit pregnancy? A pilot study,” Canadian Journal of [8] U. M. S. Koepp, K. Dahl-Jorgensen, H. Stigum et al., “The Public Health,vol.101,no.1,pp.87–91,2010. associations between maternal pre-pregnancy body mass index, [25] J. Huberty, D. Dinkel, M. W.Beets, and J. Coleman, “Describing gestational weight change during pregnancy and body mass the use of the internet for health, physical activity, and nutrition indexofthechildat3yearsofage,”International Journal of inormation in pregnant women,” Maternal and Child Health Obesity,vol.36,pp.1325–1331,2012. Journal,2012. [9] H. Lowell and D. C. Miller, “Weight gain during pregnancy: [26] C. Sparud-Lundin, A. Ranerup, and M. Berg, “Internet use, adherence to Health Canada’s guidelines,” Health Reports,vol. needs and expectations of web-based information and com- 21, no. 2, 2010, Statistics Canada, Component of Statistics munication in childbearing women with type 1 diabetes,” BMC Canada Catalogue no 82-003-XPE. MedicalInformaticsandDecisionMaking, vol. 11, pp. 49–55, [10] Health Canada, Nutrition for a Healthy Pregnancy: National 2011. Guidelines for the Childbearing Years,, Minister of Public Works [27]J.A.Herman,K.Mock,D.Blackwell,andT.Hulsey,“Useofa and Government Services Canada, Ottawa, Canada, 1999. pregnancy support web site by low-income African American [11] C. Kowal, J. Kuk, and H. Tamim, “Characteristics of weight gain women,” Journal of Obstetric, Gynecologic, and Neonatal Nurs- in pregnancy among Canadian women,” Maternal and Child ing,vol.34,no.6,pp.713–720,2005. Health Journal, pp. 1–9, 2011. [28] M. J. Brown, M. Sinclair, D. Liddle et al., “A systematic review [12] Health Canada, “Canadian gestational weight gain recom- investigating healthy lifestyle interventions incorporating goal mendations,” http://www.hc-sc.gc.ca/fn-an/nutrition/prenatal/ setting strategies for preventing excess gestational weight gain,” qa-gest-gros-qr-eng.php,2010. PLOS One,vol.7,no.7,ArticleIDe39503,2012. [13] Institute of Medicine, “Weight gain during pregnancy: reex- [29] H. Skouteris, L. Hartley-Clark, M. McCabe et al., “Preventing amining the guidelines,” http://www.iom.edu/Reports/2009/ excessive gestational weight gain: a systematic review of inter- Weight-Gain-During-Pregnancy-Reexamining-the-Guidelines ventions,” Obesity Reviews, vol. 11, no. 11, pp. 757–768, 2010. .aspx,2009. [30] I. Streuling, A. Beyerlein, E. Rosenfeld, H. Hofmann, T. Schulz, [14] A. M. Siega-Riz, A. Deierlein, and A. Stuebe, “Implementation and R. Von Kries, “Physical activity and gestational weight gain: of the new institute of medicine gestational weight gain guide- a meta-analysis of intervention trials,” BJOG,vol.118,no.3,pp. lines,” JournalofMidwiferyandWomen’sHealth,vol.55,no.6, 278–284, 2011. pp.512–519,2010. [31] L. A. Wolfe and M. Mottola, “Physical Activity Readiness [15] Institute of Medicine, Nutrition During Pregnancy, National Medical Examination for Pregnancy [PARmed-X for preg- Academies of Sciences Press, Washington, DC, USA, 1990. nancy],” Canadian Society for Exercise Physiologists, pp. [16]M.E.Cogswell,K.S.Scanlon,S.B.Fein,andL.A.Schieve, 1-4, http://uwfitness.uwaterloo.ca/PDF/parmed-xpreg 000.pdf, “Medically advised, mother’s personal target, and actual weight 2002. gain during pregnancy,” Obstetrics and Gynecology,vol.94,no. [32] L. O. Walker, B. S. Sterling, and G. M. Timmerman, “Retention 4, pp. 616–622, 1999. of pregnancy-related weight in the early postpartum period: [17] P.Brawarsky, N. E. Stotland, R. A. Jackson et al., “Pre-pregnancy Implications for women’s health services,” Journal of Obstetric, and pregnancy-related factors and the risk of excessive or Gynecologic, and Neonatal Nursing,vol.34,no.4,pp.418–427, inadequate gestational weight gain,” International Journal of 2005. Gynecology and Obstetrics, vol. 91, no. 2, pp. 125–131, 2005. [33] A. E. Montpetit, H. Plourde, T. R. Cohen, and K. G. Koski, [18] S. D. McDonald, E. Pullenayegum, V. H. Taylor et al., “Despite “Modeling the impact of prepregnancy BMI, physical activity, 2009 guidelines, few women report being counseled correctly and energy intake on gestational weight gain, infant birth 6 Journal of Pregnancy

weight, and postpartum weight retention,” JournalofPhysical Activity and Health,vol.9,pp.1020–1029,2012. [34] A. Sohlstrom and E. Forsum, “Changes in adipose tissue volume and distribution during reproduction in Swedish women as assessed by magnetic resonance imaging,” American Journal of Clinical Nutrition,vol.61,no.2,pp.287–295,1995. [35] C. Tudor-Locke and D. R. Bassett, “How many steps/day are enough? preliminary pedometer indices for public health,” Sports Medicine,vol.34,no.1,pp.1–8,2004. [36] Health Canada, “Canadian Nutrient File,” http://www.hc- sc.gc.ca/fn-an/nutrition/fiche-nutri-data/index-eng.php,2010. [37] Institute of Medicine (US) Panel on Macronutrients, Standing Committee on the Scientific Evaluation of Dietary Reference Intakes., Dietary Reference Intakes For Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids, vol. xxv, National Academies Press, Washington, DC, USA, 2005. [38] WHO and FAO, Human Energy Requirements Report of a Joint FAO/WHO/UNU Expert Consultation Food and Agriculture Organization, Food and Nutrition Technical Report Series, Energy Requirements of Pregnancy, Rome, Italy, 2004. [39] F. Begum, I. Colman, L. J. McCargar, and R. C. Bell, “Gesta- tional weight gain and early postpartum weight retention in a prospective cohort of Alberta women,” Journal of Obstetrics and Gynaecology Canada,vol.34,no.7,pp.637–647,2012. [40] T.R.Cohen,H.Plourde,andK.G.Koski,“Useofthepregnancy physical activity questionnaire (PPAQ) to identify behaviours associated with appropriate gestational weight gain during pregnancy,” JournalofPhysicalActivityandHealth.Inpress, http://www.ncbi.nlm.nih.gov/pubmed/23132834. [41] S.J.deJersey,J.M.Nicholson,L.K.Callaway,andL.A.Daniels, “A prospective study of pregnancy weight gain in Australian women,” Australian and New Zealand Journal of Obstetrics and Gynaecology,vol.52,no.6,pp.545–551,2012. [42] D. S. Downs, G. C. LeMasurier, and J. M. DiNallo, “Baby steps: pedometer-determined and self-reported leisure-time exercise behaviors of pregnant women,” JournalofPhysicalActivityand Health,vol.6,no.1,pp.63–72,2009. [43] C. L. Harrison, R. G. Thompson, H. J. Teede, and C. B. Lombard, “Measuring physical activity during pregnancy,” International JournalofBehavioralNutritionandPhysicalActivity,vol.8,no. 19, 2011, http://www.ijbnpa.org/content/8/1/19. Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 368461, 13 pages http://dx.doi.org/10.1155/2013/368461

Review Article Current Thoughts on Maternal Nutrition and Fetal Programming of the Metabolic Syndrome

Bonnie Brenseke,1,2 M. Renee Prater,1,3 Javiera Bahamonde,1 and J. Claudio Gutierrez1

1 Department of Biomedical Sciences and Pathobiology, Virginia Tech, Blacksburg, VA 24061, USA 2 Department of Pathology, Campbell University School of Osteopathic Medicine, Buies Creek, NC 27506, USA 3 Department of Biomedical Sciences, Edward Via College of Osteopathic Medicine, 2265 Kraft Drive, Blacksburg, VA 24060, USA

Correspondence should be addressed to M. Renee Prater; [email protected]

Received 13 October 2012; Accepted 3 January 2013

Academic Editor: Riitta Luoto

Copyright © 2013 Bonnie Brenseke et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Chronic diseases such as type 2 diabetes and cardiovascular disease are the leading cause of death and disability worldwide. Although the metabolic syndrome has been defined in various ways, the ultimate importance of recognizing this combination of disorders is that it helps identify individuals at high risk for both type 2 diabetes and cardiovascular disease. Evidence from observational and experimental studies links adverse exposures in early life, particularly relating to nutrition, to chronic disease susceptibility in adulthood. Such studies provide the foundation and framework for the relatively new field of developmental origins of health and disease (DOHaD). Although great strides have been made in identifying the putative concepts and mechanisms relating specific exposures in early life to the risk of developing chronic diseases in adulthood, a complete picture remains obscure. To date, the main focus of the field has been on perinatal undernutrition and specific nutrient deficiencies; however, the current global health crisis of overweight and obesity demands that perinatal overnutrition and specific nutrient excesses be examined. This paper assembles current thoughts on the concepts and mechanisms behind the DOHaD as they relate to maternal nutrition, and highlights specific contributions made by macro- and micronutrients.

1. Introduction no specific treatment for metabolic syndrome. Therapeutics include lifestyle changes (e.g., weight reduction and increased The metabolic syndrome has become a major public health physical activity) and pharmaceutical agents, but prevention challenge with an estimated 22% of US adults having this con- would be preferred. A mounting body of evidence indicates dition [1]. A consensus group for the International Diabetes that certain adverse exposures during the perinatal period Federation defines metabolic syndrome as central obesity, contribute to the development of the metabolic syndrome. plus any two of the following: raised triglycerides, reduced This early life stage may offer an attractive point in the disease high-density lipoprotein (HDL) cholesterol, raised fasting process for prevention and intervention strategies. plasma glucose, and raised blood pressure [2]. The consen- In the 1970s Forsdahl, using official statistical data on sus group also recommends additional criteria that should Norwegian counties, reported that poverty during adoles- be part of further research into the metabolic syndrome cence, followed by prosperity, was positively correlated with including tomographic assessment of visceral adiposity and risk of death from coronary heart disease [4]. Although liver fat, biomarkers of adipose tissue (leptin, adiponectin), no biological mechanisms were identified, Forsdahl spec- andglucosetolerancetesting.Thecauseofthesyndrome ulated that some form of permanent damage caused by remains obscure but the pathophysiology seems to be largely the nutritional deficit may be involved4 [ ]. In 1986, Barker attributable to insulin resistance, excessive flux of fatty and colleagues began publishing reports on the association acids, and a chronic proinflammatory state [3]. There is between an adverse intrauterine environment, as determined 2 Journal of Pregnancy primarily by low birth weight, and an increased risk of and assembles thoughts on how alterations in maternal coronary heart disease later in life. These studies involved consumption of specific nutrients may irreversibly direct fetal examining men and women in middle and late life whose programming of the metabolic syndrome (Figure 1). body measurements at birth were recorded in the archives and records offices of Britain [5, 6]. Upon further inves- 2. Concepts and Mechanisms tigation, it was found that the correlation between low birthweightandheartdiseasewasalsopresentbetween 2.1. Developmental Plasticity, Programming, and Mismatch. low birth weight and type 2 diabetes with the prevalence For most organs and systems the critical period of plasticity of impaired glucose tolerance and type 2 diabetes falling is during intrauterine development [17]. Developmental plas- progressively with increasing birth weight [7]. Moreover, in a ticity is the ability of an organism to change its phenotype in follow-up study, an association between birth weight and the response to changes in the environment [18]. If this change or presence of metabolic syndrome was discovered [8]. In this adaptation is permanent, it is considered a “programming” study metabolic syndrome was defined as impaired glucose change and is associated with persistent effects in structure tolerance, hypertension, and elevated triglycerides. In both and/or function [18, 19]. Programming is an established men and women, the prevalence of metabolic syndrome fell biological phenomenon that is exemplified in nature. Gluck- progressively with increasing birth weight. Of the 64-year-old man and Hanson use the example of the meadow vole men whose birth weights were 2.95 kg (6.5 pounds) or less, (Microtus pennsylvanicus) to illustrate programming [16]. 22% had metabolic syndrome and their risk of developing Meadowmolesborninautumnhaveathickerhaircoatthan the syndrome was more than 10 times greater than that of thoseborninthespring[20]. This occurs as a response of men whose birth weights were more than 4.31 kg (9.5 pounds) the fetus to maternally derived signals for day length [21]. A [8, 9]. The association between metabolic syndrome and thick coat reflects an adaptive programming response to help low birth weight prompted the authors to suggest that the ensure survival in the cold environment that was predicted metabolic syndrome, also called syndrome X, be renamed while in utero [16]. Barker uses the functional activity of the “the small-baby syndrome” [8]. Collectively, these studies sweat glands to illustrate plasticity and programming [17]. generated the Forsdahl-Barker hypothesis, recognizing Fors- Humans have similar numbers of sweat glands at birth, but dahlastheoriginalsourceoftheideaandBarkerasthe they are essentially nonfunctional. In the first few years of developer of the concept [10]. In the decades following life a proportion of the glands become functional depending their initial publications, the Forsdahl-Barker hypothesis has on the temperature to which the child is exposed. The become more well known as the “fetal origins” hypothesis hotter the conditions, the greater number of functional sweat and has produced a new branch of scientific knowledge and glands. After a few years the process is complete and the inquiry known as the developmental origins of health and number of functional sweat glands is fixed [17]. Programming disease (DOHaD). changes can also be experimentally-induced as demonstrated Historically, investigations on the fetal origins hypoth- by examination of adult (100-day-old) female rats that were esis have focused on maternal undernutrition and specific given a single subcutaneous injection of testosterone at 2, 5, nutrient deficiencies. Today, the world faces the dual bur- 10,or20daysofage[22]. Those that were injected between den of malnutrition that encompasses both under- and birth and 10 days of age had lasting changes in organ weight, overnutrition [11]. It is estimated that maternal and child specifically larger adrenal and pituitary weights and smaller undernutrition is the underlying cause of 3.5 million deaths ovarian and uterine weights, as well as absent corpora lutea, globally [12]. While at the same time overweight and obesity failure to develop normal female sexual behavior, and per- are major public health challenges in both developed and manent sterility [22, 23]. Theses aberrations in morphology developing regions of the world. In 2005, 33% of the world’s andphysiologyweremuchreducedorabsentinthoserats adult population (1.3 billion people) was overweight or obese, injected at 10 or 20 days of age, indicating that the period of and it is estimated that by 2030, up to 57.8% of the world’s plasticity during which rats are androgen sensitive is between adult population (3.3 billion people) will be overweight or birth and 10 days of age [22]. obese [13]. Adding to the complexity of the dual burden of In most cases programming is beneficial for the health malnutrition is the influence of socioeconomic status. By and and survival of the organism. However, the problem of “mis- large, in developed countries lower socioeconomic status is match” occurs when individuals developmentally adapted associated with both lower birth weights among the offspring to one environment are exposed to another [24]. Revisiting of women and a higher prevalence of obesity among women the functional activity of the sweat glands to provide an [14]. As a result, failure to control for social position could example, the child who has experienced hot conditions will obscure a relationship between high birth weight and subse- be better adapted to such conditions in later life because more quent obesity and obesity-related disorders [15]. Incidentally, functioning sweat glands provide the ability to cool down although the vast majority of the literature concerning the faster. While the child who has experienced cool conditions DOHaD is focused on the relationship between small size at faces the problem of mismatch and will be unable to cool birth and increased incidence of disease in adult life, it is now down as easily in hot conditions because of fewer functioning recognized that higher incidences of disease occur in both sweat glands. Other examples of this mismatch phenomenon those born small and those born large, thus reflecting a U- includepeoplewhosebirthweightsweretowardsthelower shaped curve [15, 16]. This paper provides a brief overview of end of normal that subsequently grow up in affluent societies the putative concepts and mechanisms behind the DOHaD being at increased risk for hypertension, type 2 diabetes, and Journal of Pregnancy 3

Nutrient Nutrient deﬁciencies excesses Fetal plasticity and programming

Intrauterine growth and development

Thrifty phenotype, epigenetic modiﬁcations, altered HPA axis activity, altered appetite regulation, problems relating to mismatch and Glucocorticoid catch-up growth Oxidative exposure stress

Time/age

Metabolic syndrome, type 2 diabetes, cardiovascular disease

Figure 1: Schematic representation of the relationship between nutrient exposures and the concepts and mechanisms underlying the developmental origins of health and disease (DOHaD). cardiovascular disease [16, 25]. The problem of mismatch is deemed an adaptive mechanism to ensure the survival of the thought to be involved in the current “epidemic” of type 2 fetus. Alternatively, the changes may reflect developmental diabetes and cardiovascular disease in the young adult and malformations analogous to teratogenesis [16]. The thrifty middle-aged populations of India [26]. Several countries, phenotype hypothesis is based on a study of 468 men in which including India, are undergoing swift economic and nutri- thepercentageofthosewithimpairedglucosetolerance tional transitions, exposing individuals to conditions that or type 2 diabetes fell progressively with increasing birth promote weight gain [27]. The current surge in metabolic weight and weight at 1 year [30]. From this research it was and cardiovascular disease in India may be being fueled by a hypothesized that poor intrauterine nutrition results in the combination of undernutrition in early life and overnutrition growth of vital organs, specifically the brain, at the expense in later life. In a study of Indian children, those born small of other organs (i.e., endocrine pancreas resulting in 𝛽 cell and where relatively fat and tall by 8 years of age had the most hypoplasia). Such adaptations may increase the chance of adverse risk profiles for cardiovascular disease, including fetal survival by means of “brain sparing” but result in components of the metabolic syndrome. The authors com- difficulty in coping with nutritional abundance as an adult. mented on the possible importance of preventing childhood Interestingly, the thrifty phenotype hypothesis has been obesity in the prevention of disease in the low birth weight challenged by the “fetal salvage” hypothesis which offers members of the Indian population [26]. It is important to note a different explanation for the insulin resistance seen in thatIndiansasaracearemorepronetothedevelopmentof those affected by intrauterine growth restriction31 [ ]. In this the metabolic syndrome, compared to Caucasians, as a result alternate explanation it is not hypoplasia of the pancreatic 𝛽 ofhavingaphenotypeofhighfatmassandlowleanmass[28]. cells that leads to impaired glucose tolerance, but rather it is that the fetus develops peripheral insulin resistance. It is 2.2. Thrifty Phenotype Hypothesis. The “thrifty phenotype” this peripheral insulin resistance that ensures that adequate hypothesis, put forth by Hales and Barker, proposes that poor amounts of glucose are delivered to essential organs such as fetal and early postnatal nutrition imposes mechanisms of the brain with subsequent reduced delivery to nonessential nutritional thrift upon the growing individual. In conditions tissues such as skeletal muscle. Support for the fetal sal- of severe intrauterine deprivation the developing fetus may vage hypothesis comes from intrauterine growth restriction lose functional and structural units such as pancreatic 𝛽 cells, studies in rats. In these rodent models, glucose transport nephrons, and cardiomyocytes [29]. Such changes have been (glucose uptake and glucose transporter mRNA and protein 4 Journal of Pregnancy levels) is reduced in lung and skeletal muscles of growth- insulin resistance. Thus, in this proposed scenario, insulin restricted fetuses, whereas glucose transport is unaffected in resistance is serving as a metabolic defense mechanism to brain [32, 33]. protect the organism against hypoglycemia [39]. In addition The kidney is another organ thought to be affected to insulin resistance and type 2 diabetes, accelerated weight by nutritional thrift. In both human and animal studies gain in early life has been associated with elevated blood small offspring, due to maternal undernutrition and other pressure [44, 45]andcoronaryheartdisease[46]. causes, have reduced numbers of nephrons which has been The implication of catch-up growth is that rapidly strongly linked to hypertension [16, 34, 35]. In the human, enhancing early childhood growth by a nutrient enriched nephrogenesis is completed in utero, and no new nephrons diet may cause harm overtime and that encouraging slower are formed after birth; therefore a nephron deficit present at growth rates may actually be beneficial. In a study that pro- birth persists throughout life [29, 36]. In a study of human vides evidence of the benefits of steady (i.e., not accelerated infants affected by intrauterine growth restriction, nephron by dietary means) growth, investigators measured fasting number estimates were below the control group [35]. A concentrations of 32-33 split proinsulin, a marker of insulin similar study found a 20% reduction in nephron number resistance, in adolescents born preterm who had participated in human neonates with low birth weights as compared in randomized intervention trials of neonatal nutrition [47]. to normal birth weight neonates. Additionally, researchers Fasting 32-33 split proinsulin concentration was greater in found a positive relationship between weight at birth and the those given a nutrient-enriched diet than those given the number of glomeruli as well as a negative correlation between lower-nutrient diet and was associated with greater weight weight at birth and glomerular volume [37]. This reduction in gain in the first 2 weeks of life47 [ ]. nephrons may reflect an adaptation that has the immediate advantage of energy and resource conservation but no long- 2.4. Oxidative Stress. Excessive reactive oxygen species can term advantages [16, 34]. cause modulation of gene expression and/or direct damage to cell membranes and other molecules at critical devel- 2.3. Catch-Up Growth. Catch-up growth, also known as com- opmental windows. Many believe that oxidative stress is pensatory growth, is where children return to their genetic the primary link between adverse fetal growth and later trajectoryforsizeafteraperiodofgrowthdelayorarrest. elevatedrisksofthemetabolicsyndrome,type2diabetes,and It may occur at any stage of growth but is most commonly other disorders [40]. Smoking, hypertension/preeclampsia, observed in the first 2 years of life38 [ ]. Studies have found inflammation/infection, obesity, and malnutrition are com- that catch-up growth often results in overcompensation, mon causes of preterm and/or low birth weight as well as whereby the organism exceeds normal weight and often has known sources of oxidative stress. Malnutrition can directly excessive fat deposition. This rapid and excessive growth has lead to a pro-oxidative state by means of creating protein been associated with the development of adult obesity, insulin and micronutrient deficiencies. Proteins provide amino acids resistance, metabolic syndrome, and type 2 diabetes [38–40]. needed for antioxidant synthesis, such as glutathione and In the Avon Longitudinal Study of Pregnancy and Child- albumin, and many micronutrients themselves are antioxi- hood (ALSPAC) birth cohort, children who showed catch- dants, such as vitamins A, C, and E [40, 48]. Pancreatic 𝛽 cells up growth, as calculated by specified gains in standard are particularly sensitive to reactive oxygen species because deviation scores, between 0 and 2 years of age, were heavier, they are low in enzymatic antioxidant defense equipment taller, and fatter (body mass index, percentage body fat, and [49]. It has been demonstrated that oxidative stress can blunt waist circumference) at 5 years of age than other children insulin secretion [50]. With the susceptibility of pancreatic 𝛽 [38]. A study that evaluated glucose tolerance and plasma cells to oxidative stress, it is believed that early and ongoing insulin concentrations in more than 1400 young adults exposures to oxidative insults could result in the eventual (26 to 32 years old) who had grown up in the city of Delhi, manifestations of the metabolic syndrome and related disor- India, found an association between thinness in infancy and ders [40]. the presence of impaired glucose tolerance or diabetes in Elevated oxidative stress has been observed among young adulthood [41]. As a group, the young adults in the human infants born small for gestational age (SGA) as study who had impaired glucose tolerance or diabetes were compared to those appropriate for gestational age (AGA) and overweight. They were not, however, overweight or obese in among preterm as compared to term infants. It is important childhood. Instead, they were characterized by their high rate to note that preterm (birth at <37 weeks of pregnancy) infants of gain in body mass after the age of 2 years41 [ ]. are generally of low birth weight but are not necessarily Cianfarani and colleagues speculate that the tremendous growth restricted or small for gestational age. A study using effort to recover lost growth shortly after birth involves cord blood to compare the status of oxidative stress between insulin and insulin-like growth factors (IGF). Infants affected SGA infants born to undernourished mothers and AGA by intrauterine growth restriction have low concentrations infants born to healthy mothers found elevated oxidative of insulin and IGF-1 at birth, and normalization of these stress, as determined by increased quantities of malondi- parameters occurs in the postnatal period [42, 43]. It is aldehyde (one of the major products of lipid peroxida- thought that tissues chronically depleted of insulin and tion), reduced quantities of the antioxidant glutathione, and IGF-1 throughout fetal life and then suddenly exposed to decreased activity of the antioxidants superoxide dismutase increased concentrations of the two hormones shortly after and catalase in the SGA infants as compared to the AGA birth may counteract the actions of insulin by developing controls [51]. Journal of Pregnancy 5

A human study determined vitamin A, C, and E levels blood pressures. Such studies suggest that excess glucocorti- in the umbilical cord blood of term and preterm infants and coid exposure at certain developmental stages or “windows” the blood of their mothers taken at the time of delivery [52]. programs higher blood pressure [57]. Another study of Maternal vitamin A and E levels were higher than cord values, pregnant sheep found that severe brief undernutrition in yet there was a positive correlation between maternal and lategestationalteredthefunctionoftheHPAaxisofadult cord levels of these two vitamins. Contrary to vitamins A and offspring. Those exposed to gestational undernourishment E, cord vitamin C levels were higher than maternal levels, and for 10 days demonstrated altered steroid levels including an no significant correlation between cord and maternal vitamin increased adrenocorticotropic hormone (ACTH) response as C levels was present. In comparing vitamin levels in term and compared to offspring from dams fed ad libitum or offspring preterm infants, term babies had significantly higher levels of from dams undernourished for 20 days [58]. vitaminsA,C,andE.Theauthorsconcludedthatforvitamins An epidemiological study in humans involving three A and E, neonatal levels are dependent on maternal levels populations found that adults who had lower birth weights and that preterm babies have fewer lipid-soluble antioxidant had elevated fasting plasma concentrations of cortisol and vitamins in their serum compared to term infants thus may that cortisol concentrations positively correlated with current be more susceptible to oxidative stress [52]. blood pressure [59]. The authors concluded that increased Gestational diabetes, a common cause of macrosomia, activity of the HPA axis may link low birth weight with isassociatedwithoxidativestresswhichcouldhaveeffects raised blood pressure in adult life [59]. In a study of over on the developing infant. A study evaluating oxidative and 300 men, plasma cortisol concentrations fell progressively antioxidative status in pregnant diabetic (gestational diabetes with increasing birth weight; this trend was independent of or type 1 diabetes) women between 26 and 32 weeks gestation age and body mass index [60]. Raised plasma cortisol con- demonstrated increased oxidative stress in those who were centrations were significantly associated with higher blood diabetic [53]. Levels of malondialdehyde were significantly pressure as well as plasma glucose concentrations and insulin higher, and vitamin A and E levels were significantly lower resistance, suggesting that programming of the HPA axis may in all diabetic women than controls. Moreover, glutathione be a mechanism underlying the association between low birth peroxidase and superoxide dismutase activities were signif- weight and the metabolic syndrome in adult life [60]. icantly reduced in diabetic women; glutathione peroxidase was reduced in both those with gestational and type 1 2.6. Neuropeptides. The hypothalamus plays a critical role in diabetes, and superoxide dismutase was reduced only in type the regulation of appetite and body composition by way of 1diabetics[53]. responding to cues from neuropeptides. A series of studies, primarily in rodents, have explored the possibility that mater- 2.5. Hypothalamic-Pituitary-Adrenal Axis. An increase in nal nutrition during pregnancy may alter the level of energy circulating glucocorticoids may play a role in early program- intake in the offspring through inducing changes in the ming of adult disease. Experimental studies in animals and hypothalamic circuitry and in the expression, localization, epidemiological studies in humans have suggested an altered and action of specific neuropeptides. Central and periph- set-point of the hypothalamic-pituitary-adrenal (HPA) axis eral neuropeptides function in the regulation of appetite. as an important long-term change that occurs in association Appetite stimulating neuropeptides include neuropeptide Y with reduced fetal growth. Pregnant animal models have (NPY), agouti-related peptide (AgRP), and ghrelin. Con- shown that exposure to a variety of stressors, including versely, appetite suppressing neuropeptides include cocaine nutrient restriction, results in the birth of offspring with and amphetamine-related transcript (CART), melanocyte- elevated basal or stress-induced glucocorticoid secretion [54, stimulating hormone (MSH), serotonin, insulin, and leptin 55]. It is thought that maternal exposure to stressors during [61, 62]. Of note is that the overweight and obese tend pregnancy subsequently leads to excessive fetal exposure to to develop resistance to insulin and leptin; therefore these glucocorticoids resulting in persistent alterations in HPA two neuropeptides are usually elevated in such groups. The axis activity. In support of this hypothesis are studies con- hypothalamus has multiple nerve centers, or nuclei, that are ducted in rats in which fetoplacental exposure to maternally sensitive to a variety of physiologic changes. Hypothalamic administered steroids throughout gestation reduced birth nuclei include the arcuate nucleus (ARC), paraventricular weightandproducedhypertensiveadultoffspring.Inone nucleus (PVN), and ventromedial nucleus (VMN) which are such study, dexamethasone administration to pregnant rats involved in a number of biological activities with a fair degree on days 15–20 of gestation resulted in offspring with reduced of overlap [63]. In general, the ARC is a conduit of many birth weight, elevated blood pressure, increased basal plasma diverse signals involved in various functions such as energy corticosterone, lower mRNA expression of hippocampal homeostasis [64], while the PVN is associated with blood neuronal glucocorticoid receptor, and decreased gene expres- pressure and stress responses [65] and the VMN with satiety sion of hippocampal mineralocorticoid receptor [56]. In a [63]. Each also plays a role in the central nervous system sheep study, dams were treated with dexamethasone over 2 regulation of food intake and body weight. days; treatment group 1 was treated during 22–29 days of Studies in rats have found that exposure to overnutrition pregnancy and treatment group 2 was treated during 59– in the fetal or neonatal period can result in permanent 66 days of pregnancy (term 145 days) [57]. Offspring from changes in body fat mass and in the hypothalamic neuronal dams that had received dexamethasone during 22–29 days circuitry regulating appetite in the adult brain. Dorner,¨ gestation, but not days 59–66 of gestation, had elevated Plagemann, and colleagues found immunocytochemical and 6 Journal of Pregnancy morphometric aberrations in the hypothalamus of offspring of evidence supports the role of environmentally-induced of diabetic dams or in those exposed to milk from diabetic epigenetic changes in disease susceptibility. Experimental dams. In one of their early studies, offspring of diabetic dams studies in agouti mice suggest a role for maternal diet in had permanent hypoplasia of the VMN, decreased insulin inducing epigenetic changes in the offspring [75, 76]. Agouti responsiveness to glucose, impaired glucose tolerance, and mice are so named because they carry the agouti gene 𝑦 𝑣𝑦 increased susceptibility to diabetes [66]. A later study exam- [77]. Those carrying the dominant agouti alleles 𝐴 or 𝐴 ined the effects of exposure to milk from diabetic dams. In develop the complex set of traits collectively referred to this study, offspring of control dams cross-fostered to diabetic as the yellow obese syndrome. The yellow obese syndrome dams developed early postnatal growth delay and showed encompassesthepleiotropiceffectsofyellowfur,obesity, structural and functional hypothalamic “malprogramming” hyperinsulinemia, hyperglycemia, and increased suscepti- as it relates to appetite stimulation and suppression [67]. bility to neoplasia [77]. Female a/a mice fed a methyl- Exposure to milk from a diabetic dam resulted in an upregu- supplemented diet with extra folate, vitamin B-12, choline, 𝑣𝑦 lationoftheappetitestimulantsNPYandAgRPandadown- and betaine 2 weeks prior to mating with male 𝐴 /𝑎 agouti regulation of the appetite suppressant MSH. Although the mice and throughout pregnancy and lactation passed along researchersexpectedtofindchangesintheVMNasintheir the agouti gene to their offspring intact, yet demonstrated a previous study, they found increased total number of neurons shift in distribution towards having more offspring with the in the PVN which, as mentioned previously, is associated with pseudoagouti phenotype [76]. Pseudoagouti mice are brown, regulation of blood pressure [67]. Incidentally, their research lean, healthy, and longer lived than their yellow siblings [75]. team performed a human, clinical study and reported an This shift in the distribution in coat color was mediated 𝑣𝑦 association between the neonatal ingestion of breast milk by CpG methylation at the at the 𝐴 locus [76]. Another from diabetic mothers and increased blood pressure during example of environmentally-induced epigenetic changes is childhood [68]. reduced pancreatic and duodenal homeobox 1 (Pdx1), also Leptin has received a considerable amount of attention known as insulin promoter factor 1, in a rat model of in regard to early programming of appetite and body com- intrauterine growth restriction. Pdx1 is a transcription fac- position. In a rodent study to establish whether neonatal tornecessaryfordevelopmentandfunctionoftheinsulin (postnatal day 3–13) leptin treatment can alleviate postnatal producing pancreatic 𝛽 cell. Rats faced with intrauterine obesity and the associated metabolic sequelae that occur growth restriction had permanently reduced expression of in the offspring of undernourished dams, leptin treatment Pdx1 in 𝛽 cells and developed type 2 diabetes in adulthood reversed the programmed phenotype [69]. The benefits of [78, 79]. Reduced Pdx1 transcription was mediated through leptin treatment in these rats included a slowing of neonatal a cascade of epigenetic modifications characterized by loss weight gain and normalized caloric intake, body weight, fat of upstream stimulatory factor-1 binding at the proximal mass, and fasting plasma glucose and insulin [69]. Studies promoter of Pdx1, recruitment of the histone deacetylase 1 in leptin deficient (Lepob/Lepob) mice have demonstrated and the corepressor Sin3A, and deacetylation of histones H3 permanently disrupted neural projection pathways from the and H4 which culminated in the eventual silencing of Pdx1 ARC [70]. In the adult Lepob/Lepob mice, leptin treatment [78]. did not reverse these neuroanatomical defects; however The Dutch famine of 1944 has been used by various treatment of neonatal Lepob/Lepob mice with exogenous investigators as an equivalent to an experimental study to leptin rescued the development of ARC projections [71]. investigate the effects of perinatal undernutrition in humans [80]. The ongoing Hunger Winter Families Study contributed empirical support for the hypothesis that early-life environ- 2.7. Epigenetics. Epigenetics refers to all modifications to mental conditions cause epigenetic changes in humans that genes other than changes in the DNA sequence itself and persist throughout life. Individuals who had been exposed includes DNA methylation and histone modifications [18, to famine during periconception had, 6 decades later, less 72]. DNA methylation is a post-replication modification DNA methylation of the insulin-like growth factor II (IGF2) that is predominantly found in the cytosines of the dinu- gene compared with their unexposed, same-sex siblings cleotide sequence cytosine phosphate guanine (CpG) [73]. [81]. The Dutch famine cohort has also been used to study DNA methylation stabilizes gene expression in cells. Changes the transgenerational effects of famine exposure. Women in methylation status (hyper- or hypomethylation) have exposed to famine while in the womb later had offspring with been associated with various health conditions including birth weights lower than offspring of women not exposed to malignancies [73]. Histones are proteins that permit the famine. This effect of in utero exposure to famine on birth packaging of DNA into nucleosomes, the fundamental units weight in the subsequent generation persisted after control of chromatin [74]. Histones are subject to a large number of for potential confounding and intervening variables [82]. post-translational modifications which are likely to control In a study of 2 prospective cohorts, Godfrey and col- the structure and/or function of the chromatin fiber. Different leagues used DNA extracted from umbilical cord tissue modifications may have distinct consequences such as tran- obtained at birth in children who were later assessed for adi- scriptional activation or DNA repair [74]. posity at 9 years of age to measure methylation status of CpGs Everycellinthebodyhasthesamegeneticinformation; in the promoters of candidate genes [83]. Five candidate genes what makes cells, tissues, and organs different is that different were selected based on a number of criteria which included sets of genes are turned on or expressed. An increasing body animal data [83–85] and correlations with overall gene Journal of Pregnancy 7 methylation status and adiposity. Methylation of retinoid X animal protein was less than 50 g daily, a higher carbohydrate receptor-𝛼 (RXRA) chr9: 136355885+ (in cohort 1 and 2) and intake was associated with higher blood pressures in the off- endothelial nitric oxide synthase (eNOS) chr7: 150315553+ (in spring. Conversely, when daily protein intake was above 50 g, cohort 1 only) at birth was correlated with greater adiposity lower carbohydrate intake was associated with higher blood in later childhood. Although the data is correlative and pressure. An additional finding was that increases in blood does not prove causality between DNA methylation at birth pressure were associated with decreased placental size. Their andadiposityinchildhood,theobservationsuggeststhat conclusions were that blood pressure in adulthood may be epigenetics is involved in fetal programming of later obesity influenced by maternal intakes of protein and carbohydrates [83]. during pregnancy and that this may be mediated through effects on placental growth [89]. 3. Macronutrient Contributions 3.2. Carbohydrates. All dietary carbohydrates can be con- 3.1. Protein. A sizable number of experimental studies, verted into glucose [90]. In both animal and human studies, predominantly in rats, have explored the effects of pro- intrauterine exposure to high sugar diets and/or hyper- tein restriction on fetal growth and later health. As dis- glycemia has been found to increase the risk of the metabolic cussed in the section on the thrifty phenotype hypothe- syndrome. In rats, a high fructose diet for 2 weeks resulted sis, protein restriction during pregnancy has been found in increases in systolic blood pressure as well as plasma to produce small offspring that have reduced numbers of insulin and triglyceride concentrations as compared to rats nephrons which potentially contributes to the development fed a normal chow [91]. In another rat study early-and long- of hypertension in adulthood [16, 34, 35]. Low protein term exposure to a high sucrose diet (HSD) was assessed to exposure in fetal rats has also been shown to result in determine whether such exposure alleviates the detrimental reduced pancreatic 𝛽 cell mass at birth and reduced insulin effects of sucrose feeding in later life [92]. Dams were fed secretion in later life presumably due to dietary-induced either standard or HSD (70% calories as sucrose) starting 1 reduction in proliferation rate and increased apoptosis of week before breeding and throughout gestation and lactation. 𝛽 cells [86]. Weanling (26-day-old) rats exposed to low After weaning, all male offspring were fed HSD until the protein during gestation and lactation were smaller and had age of 20 weeks, then detailed metabolic and morphometric reduced serum insulin and increased serum glucose and profiles were ascertained. Offspring of sucrose-fed dams triglycerides [87]. In addition, low protein-exposed offspring displayed higher adiposity and increases in triglyceride liver had increased hepatic triglycerides and hepatic expression content together with higher low-density lipoprotein (LDL) of lipogenic enzymes favoring fat synthesis. The authors cholesterol concentrations. Although the significance and suggested that the increased expression of fat-synthesizing mechanisms are not clear, a somewhat perplexing observa- enzymes may account for the increased levels of serum tion was that of substantial increases in the insulin sensitivity and liver triglycerides seen in those exposed to low protein of skeletal muscle together with higher concentrations of and that these increases may predispose the offspring to adiponectin in the offspring of sucrose-fed dams compared excessive accumulation of fat and ultimately obesity and with the offspring of standard diet-fed dams. Triglycerides, insulin resistance [87]. Protein restriction may be involved free fatty acids, overall glucose tolerance, and the insulin in programming food preferences. Rat studies indicate that sensitivity of adipose tissue were comparable in both groups gestational exposure to low levels of protein due to maternal [92]. protein restriction establishes a preference for high fat foods Data derived from the Camden Study of adolescent [88]. When given the choice of selecting high fat, high pregnancyimplicatedhighsugarconsumptionwithatwofold protein, or high carbohydrate foods, 12-week-old offspring of increased risk for delivering small for gestational age infants dams fed a low protein diet throughout gestation consumed [93]. Additionally, evaluation of the data based on ethnicity significantly more of the high fat food and significantly less of led to the detection of a substantial decrease in the duration the high carbohydrate food than their control counterparts. of gestation among Puerto Rican adolescents with high sugar At30weeksofage,therewasnodifferenceinthepatternof diets, whereas there was no effect of a high sugar diet on food selection between the two groups. Their results suggest gestation duration among black and white adolescents. The that early exposure to undernutrition programs a preference authors commented that other studies involving comparisons for fatty foods, and thus maternal nutrition may influence among ethnicities have found Hispanic women to be more pathways involved in control of appetite or the perception of likely to have abnormalities of carbohydrate metabolism dur- palatability in the offspring88 [ ]. ing pregnancy as evidenced by abnormal glucose tolerance Although the literature on protein restriction and dis- tests and increased incidence of gestational diabetes [93– ease risk is quite extensive, it is largely based on rodent 95]. Pregnancies complicated by maternal diabetes, in any models with few having assessed the specific role of protein form, place the offspring at risk for developing obesity and restrictioninhumanriskofdisease.Onestudyofmenand glucose intolerance [29, 96]. In a study of 18- to 27-year- women in Scotland sought to determine how diet of the old women born to diabetic (gestational diabetes or type 2 mother in pregnancy influences blood pressure in the 40- diabetes) mothers, the risk of overweight was doubled in year-old adult offspring [89]. The authors found that the the offspring of diabetic mothers as compared with offspring relations between the diet of mothers and the blood pressure from a background population [97]. Moreover, the risk of of their offspring were complex. When maternal intake of the metabolic syndrome was increased 4-fold for offspring of 8 Journal of Pregnancy mothers with gestational diabetes and 2.5-fold for offspring of a significant inverse relationship was established between the mothers with type 1 diabetes. Offspring risk of the metabolic risk of spontaneous abortion and the consumption of green syndrome increased significantly with increasing maternal vegetables, fruits, milk, cheese, eggs, and fish [103]. Results fasting blood glucose as well as 2-hour blood glucose fol- of a comprehensive dietary review of women with a history lowing an oral glucose load. Their findings indicate that of gestational diabetes found that women with recurrence intrauterine exposure to hyperglycemia contributes to the of gestational diabetes in a subsequent pregnancy had sig- pathogenesis of the metabolic syndrome, thus offspring of nificantly higher fat intakes as a percentage of total energy diabetic mothers are risk groups for this condition [97]. than women who did not have recurrence [104]. Lipids and lipid disorders play a central role in the development of 3.3. Fats. Our laboratory has performed multiple experimen- the metabolic syndrome and its associated diseases [105]. tal studies aimed at determining the role of high fat diet Atherosclerotic lesions are known to begin in early life [106]. in maternal-fetal health. In one such study, mice exposed Fatty streaks in the aorta have been observed in children as to a high saturated fat diet during gestation and lactation young as 3 years old [107] and autopsies of young soldiers developed adult obesity, hyperglycemia, insulin resistance, killed in the Korean and Vietnam wars revealed advanced and hypertension, despite being fed a standard rodent diet coronary artery lesions [108, 109]. Similarly, the Pathobio- post-weaning [98]. Given that oxidative stress is a potential logical Determinants of Atherosclerosis in Youth (PDAY) mechanism connecting fetal exposures to the development study investigated atherosclerosis in autopsied persons 15 to ofthemetabolicsyndrome,agroupofpregnantmicewere 34 years old that died from various traumas and observed supplemented with quercetin, a powerful antioxidant, which advanced lesions of atherosclerosis in adolescents and young mitigated the detrimental effects of high fat diet exposure adults and that serum lipoprotein concentrations were a risk [98]. Similarly, Bouanane and colleagues used a rat model factor [110]. to determine the effects of a cafeteria diet, that had 42% of energy from fat, on oxidant/antioxidant status as well 4. Micronutrient Contributions as variety of metabolic markers [99]. Dams fed the cafeteria diet developed increased body weight, hyperglycemia, Although micronutrient deficiencies are known causes of hyperinsulinemia, hyperleptinemia, hyperlipidemia, and an several well-characterized diseases, such as scurvy and rickets imbalance between oxidants/antioxidants favoring oxidative [111], the role of micronutrients in the fetal origins of adult stress. Male and female offspring displayed similar effects, disease has yet to be explored in detail. However, existing which persisted after birth, suggesting an unrelenting effect evidence points to their potential importance [112]. Many of maternal diet on metabolism and body habitus of sur- micronutrients are antioxidants or are components of the viving offspring99 [ ]. In the development of a dietary- antioxidant defense system. The section on oxidative stress induced gestational diabetes mouse model, our laboratory provided examples on how micronutrients, as antioxidants, demonstrated that consumption of a high saturated fat diet may be associated with various short- and long-term effects prior to conception and throughout pregnancy can result in on the offspring. In relating maternal micronutrient intake insulin resistance and placental vascular damage and that with fetal glucocorticoid exposure, a mouse model of mater- these abnormalities could be a result of oxidative stress [100]. nal dietary restriction of copper, zinc, and vitamin E demon- Frias and colleagues, using a nonhuman primate model to strated reduced activity level of placental 11𝛽-hydroxysteroid determine the effect of chronic high fat diet on pregnancy dehydrogenase-2, an enzyme that protects the fetus from outcome, found that consumption of such a diet, independent overexposure to maternal glucocorticoids [113]. As men- ofmaternalobesity,ledtoasignificantreductioninuterine tioned in the section on the hypothalamic-pituitary-adrenal blood flow, a rise in placental inflammation, and an increase axis, fetal exposure to glucocorticoids is associated with infetalriskofnonalcoholicfattyliverdisease,asevidenced small size at birth and insulin resistance and hypertension by increased levels of liver triglycerides and increased hepatic in adult life. Indeed, in this study offspring exposed to oxidative stress [101, 102]. Offspring of high fat diet dams the micronutrient restricted diet had significantly reduced also exhibited elevated hepatic expression of gluconeogenic body weight and crown-to-rump length at birth as well as enzymes and transcription factors. Moreover, reversing the increased systolic blood pressure and insulin levels post- maternal high fat diet to a low fat diet during a subsequent weaning as compared to those exposed to a control diet. As pregnancy improved fetal hepatic triglyceride levels and discussed in the section on epigenetics, experimental studies partially normalized gluconeogenic enzyme expression. The on the impact of methyl-supplemented diets containing authors concluded that a developing fetus is highly vulnerable added folate and vitamin B-12 revealed the potential role of to excess lipids, independent of maternal diabetes and/or these micronutrients in inducing major changes in offspring obesity and that such exposure may increase the risk of phenotype, including the ability to impact the development of pediatric fatty liver disease [102]. chronic diseases. Studies on maternal dietary zinc restriction Humanobservationalstudieshavealsofoundhighfat in rats indicate that zinc deficiency during intrauterine and intakes to be related to pregnancy and birth outcome as well postnatal growth can induce elevations in blood pressure as long-term health of the offspring. In a case-control study and renal lesions in adulthood [114, 115]. Concerning the of 912 women admitted to obstetric hospitals for spontaneous renal alterations identified, early exposure to zinc deficiency abortion, the risk of miscarriage was directly associated with resulted in a decrease in glomerular filtration rate which consumption of the main types of fats, butter and oil, while wasassociatedwithareductioninthenumberandsizeof Journal of Pregnancy 9 nephrons. These animals also had proteinuria, higher lipid humans particularly relating to early development. Rodents peroxidation end products, and evidence of increased renal are altricial species, and their organs mature after birth, apoptosis and fibrosis [114, 115]. which explains why postnatal factors may be more perti- Few human studies have explored the role of early nent in rodent species, whereas pre- and perinatal factors micronutrient deficiencies in the development of adult play a vital role in human development. Human infants metabolic disorders. Of the existing reports, results are often of hyperglycemic mothers develop macrosomia, whereas conflicting and the significance is ambiguous. In two ran- hyperglycemic rodent dams typically have normally sized or domized controlled trials, both in Nepal, one found that daily small pups. The more precocious human 𝛽 cells produce fetal multiple micronutrient supplementation in pregnant women insulin prior to birth, thus increasing glucose transport and resulted in slightly lower blood pressure in the offspring at growth prenatally, which does not occur in the rodent, whose 2.5 years of age [116], while the other found no effects on endocrine pancreas becomes functional postnatally [62, 124, blood pressure with maternal multiple micronutrient supple- 125]. Second, studies relating early exposures to later life risk mentation among 6- to 8-year-old offspring [117]. As part of of disease have been largely based on body weight and/or thePune(India)MaternalNutritionStudy,highermaternal body mass index; however, these measurements provide erythrocyte folate concentrations at 28 weeks gestation were no information on body composition. More recent studies associated with higher adiposity and insulin resistance in have included the assessment of body composition, and children 6 years of age [118]. Conversely, a study in England evidence suggests that lower birth weights reflect increased foundthatmaternalfolateintakeat18or32weeksgestation relative adiposity, while higher birth weights reflect higher was not associated with any measures of body composition in lean mass [15]. The increased use of body composition children at 9 years of age [119]. Clearly more work is needed to analysis is likely to offer a more complete understanding disentangle the intricate interactions among micronutrients as to why small babies are prone to the development of and to determine how these interactions may be involved in the metabolic syndrome, an association that likely involves the initiation and progression of chronic disease. increased adiposity and decreased lean mass at birth. Finally, studies in twins, especially monozygotic twins, have caused 5. Conclusions considerable contention in the field15 [ , 29]. Twins face both physical and nutrient constraints and are generally It is now widely accepted that certain chronic diseases of smaller than singletons. The fetal origins hypothesis indicates adulthood may have their origins in the womb. Studies that such individuals should have increased morbidity and discussed in this paper provide evidence that a mother’s diet mortality from metabolic and cardiovascular disease, but this during pregnancy can exert major effects on the short- and has not been proven to be necessarily the case [15, 29, 126]. long-term health of her children including programming The DOHaD is a relatively new field of research and in time of the metabolic syndrome. The challenges at present are such discrepancies may be resolved. With a more complete to identify common mechanisms and pathways involved understanding of the role of maternal health and nutrition in disparate perinatal malnutrition paradigms, deciphering in the initiation and progression of the metabolic syndrome physiological and/or pathological roles of specific nutrients, and other disorders comes the hope of prevention of chronic and to determine which components of the maternal diet diseases at their earliest beginnings. may be best modified to optimize maternal health, placental integrity, birth outcome, and lifelong health of the offspring. The implications of this avenue of research, particularly to Authors’ Contributions obstetric and preventative medicine, dictate that effective B.BrensekeandM.R.Pratercontributedequallytothiswork. interdisciplinary communication and knowledge transfer occur and that the information generated is disseminated to the general public. A recent review provides encouraging References evidence that prenatal nutritional advice appears effective in reducing the risk of preterm birth and that balanced energy [1] E. S. Ford, W. H. Giles, and W. H. Dietz, “Prevalence of the and protein supplementation seem to improve fetal growth metabolic syndrome among US adults: findings from the Third and may reduce the risk of stillbirth and infants born small National Health and Nutrition Examination Survey,” Journal of for gestational age [120]. the American Medical Association,vol.287,no.3,pp.356–359, 2002. Whilst the evidence in support of the DOHaD is com- pelling, there are many caveats and controversies in the [2]K.G.M.M.Alberti,P.Zimmet,andJ.Shaw,“Metabolic field including species differences, body weight verses body syndrome: a new world-wide definition. A consensus statement composition, and twin studies. First, species differences from the International Diabetes Federation,” Diabetic Medicine, (e.g., maternal age, parity, gestation length, fetal number, vol.23,no.5,pp.469–480,2006. and variability in fetoplacental development) should be [3]R.H.Eckel,S.M.Grundy,andP.Z.Zimmet,“Themetabolic considered when results derived from animal studies are syndrome,” The Lancet,vol.365,no.9468,pp.1415–1428,2005. extrapolated to humans. Rodents are particularly important [4] A. Forsdahl, “Are poor living conditions in childhood and to reproductive and obstetric studies because of the strong adolescence an important risk factor for arteriosclerotic heart similarities between human and murine placentas [121–123]; disease?” British Journal of Preventive and Social Medicine,vol. however, there are many differences between rodents and 31,no.2,pp.91–95,1977. 10 Journal of Pregnancy

[5] D. J. Barker and C. Osmond, “Low birth weight and hyperten- [24] P. Bateson, D. Barker, T. Clutton-Brock et al., “Developmental sion,” British Medical Journal, vol. 297, no. 6641, pp. 134–135, plasticity and human health,” Nature,vol.430,no.6998,pp.419– 1988. 421, 2004. [6]D.J.P.Barker,P.D.Winter,C.Osmond,B.Margetts,andS.J. [25] C. Kanaka-Gantenbein, “Fetal origins of adult diabetes,” Annals Simmonds, “Weight in infancy and death from ischaemic heart of the New York Academy of Sciences,vol.1205,pp.99–105,2010. disease,” The Lancet,vol.2,no.8663,pp.577–580,1989. [26] A. Bavdekar, C. S. Yajnik, C. H. D. Fall et al., “Insulin resistance [7]C.N.Hales,D.J.P.Barker,P.M.S.Clarketal.,“Fetaland syndrome in 8-year-old Indian children: small at birth, big at 8 infant growth and impaired glucose tolerance at age 64,” British years, or both?” Diabetes,vol.48,no.12,pp.2422–2429,1999. Medical Journal,vol.303,no.6809,pp.1019–1022,1991. [27] S. A. Norris, C. Osmond, D. Gigante et al., “Size at birth, weight [8]D.J.P.Barker,C.N.Hales,C.H.D.Fall,C.Osmond,K.Phipps, gain in infancy and childhood, and adult diabetes risk in five and P. M. S. Clark, “Type 2 (non-insulin-dependent) diabetes low- or middle-income country birth cohorts,” Diabetes Care, mellitus, hypertension and hyperlipidaemia (syndrome X): vol. 35, no. 1, pp. 72–79, 2012. relation to reduced fetal growth,” Diabetologia,vol.36,no.1,pp. [28]S.A.Lear,S.Kohli,G.P.Bondy,A.Tchernof,andA.D. 62–67, 1993. Sniderman, “Ethnic variation in fat and lean body mass and [9] S. E. Ozanne and C. N. Hales, “Early programming of glucose- the association with insulin resistance,” Journal of Clinical insulin metabolism,” Trends in Endocrinology and Metabolism, Endocrinology and Metabolism,vol.94,no.12,pp.4696–4702, vol. 13, no. 9, pp. 368–373, 2002. 2009. [10] S. Vangen, R. Nordhagen, and K. K. Lie, “Revisiting the [29] I. C. McMillen and J. S. Robinson, “Developmental origins of Forsdahl-Barker hypothesis,” JournaloftheNorwegianMedical the metabolic syndrome: prediction, plasticity, and program- Association,vol.125,no.4,pp.451–453,2005. ming,” Physiological Reviews,vol.85,no.2,pp.571–633,2005. [11] M. Jehn and A. Brewis, “Paradoxical malnutrition in mother- [30] C. N. Hales and D. J. P.Barker, “Type 2 (non-insulin-dependent) child pairs: untangling the phenomenon of over- and under- diabetes mellitus: the thrifty phenotype hypothesis,” Diabetolo- nutrition in underdeveloped economies,” Economics and gia,vol.35,no.7,pp.595–601,1992. Human Biology,vol.7,no.1,pp.28–35,2009. [12] R. E. Black, L. H. Allen, Z. A. Bhutta et al., “Maternal and [31] P.L.Hofman,W.S.Cutfield,E.M.Robinsonetal.,“Insulinresis- child undernutrition: global and regional exposures and health tance in short children with intrauterine growth retardation,” consequences,” The Lancet,vol.371,no.9608,pp.243–260,2008. Journal of Clinical Endocrinology and Metabolism,vol.82,no.2, pp.402–406,1997. [13]T.Kelly,W.Yang,C.S.Chen,K.Reynolds,andJ.He,“Global burden of obesity in 2005 and projections to 2030,” International [32] R. A. Simmons, A. S. Gounis, S. A. Bangalore, and E. S. Journal of Obesity,vol.32,no.9,pp.1431–1437,2008. Ogata, “Intrauterine growth retardation: fetal glucose transport is diminished in lung but spared in brain,” Pediatric Research, [14] J. Sobal and A. J. Stunkard, “Socioeconomic status and obesity: vol.31,no.1,pp.59–63,1992. areviewoftheliterature,”Psychological Bulletin,vol.105,no.2, pp.260–275,1989. [33] R. A. Simmons, A. S. Flozak, and E. S. Ogata, “The effect of [15] I. Rogers, “The influence of birthweight and intrauterine insulin and insulin-like growth factor-I on glucose transport in environment on adiposity and fat distribution in later life,” normal and small for gestational age fetal rats,” Endocrinology, International Journal of Obesity and Related Metabolic Disorders, vol. 133, no. 3, pp. 1361–1368, 1993. vol. 27, no. 7, pp. 755–777, 2003. [34] H. S. Mackenzie and B. M. Brenner, “Fewer nephrons at birth: [16] P.D. Gluckman and M. A. Hanson, “The developmental origins a missing link in the etiology of essential hypertension?” The of the metabolic syndrome,” Trends in Endocrinology and American Journal of Kidney Diseases, vol. 26, no. 1, pp. 91–98, Metabolism,vol.15,no.4,pp.183–187,2004. 1995. [17] D. J. P.Barker, “The origins of the developmental origins theory,” [35] S. A. Hinchliffe, M. R. J. Lynch, P.H. Sargent, C. V.Howard, and Journal of Internal Medicine,vol.261,no.5,pp.412–417,2007. D. Van Velzen, “The effect of intrauterine growth retardation on [18] B. M. Cota and P.J. Allen, “The developmental origins of health the development of renal nephrons,” British Journal of Obstetrics and disease hypothesis,” Pediatric Nursing,vol.36,no.3,pp.157– and Gynaecology,vol.99,no.4,pp.296–301,1992. 167, 2010. [36] G. B. Haycock, “Development of glomerular filtration and [19] P.D. Gluckman, M. A. Hanson, and C. Pinal, “The developmen- tubular sodium reabsorption in the human fetus and newborn,” tal origins of adult disease,” Maternal and Child Nutrition,vol. British Journal of Urology,vol.81,no.2,pp.33–38,1998. 1, no. 3, pp. 130–141, 2005. [37]R.Manalich,L.Reyes,M.Herrera,C.Melendi,andI.Fundora, [20] T. M. Lee and I. Zucker, “Vole infant development is influenced “Relationship between weight at birth and the number and perinatally by maternal photoperiodic history,” The American size of renal glomeruli in humans: a histomorphometric study,” Journal of Physiology, vol. 255, no. 5, p. 24/5, 1988. Kidney International,vol.58,no.2,pp.770–773,2000. [21] T. M. Lee, N. Spears, C. R. Tuthill, and I. Zucker, “Maternal [38] K. K. L. Ong, M. L. Ahmed, D. B. Dunger, P.M. Emmett, and M. melatonin treatment influences rates of neonatal development A. Preece, “Association between postnatal catch-up growth and of meadow vole pups,” Biology of Reproduction,vol.40,no.3, obesity in childhood: prospective cohort study,” British Medical pp.495–502,1989. Journal,vol.320,no.7240,pp.967–971,2000. [22] C. A. Barraclough, “Production of anovulatory, sterile rats by [39] S. Cianfarani, D. Germani, and F. Branca, “Low birthweight single injections of testosterone propionate,” Endocrinology,vol. and adult insulin resistance: the “catch-up growth” hypothesis,” 68, pp. 62–67, 1961. Archives of Disease in Childhood,vol.81,no.1,pp.F71–F73,1999. [23] J. E. Harding, “The nutritional basis of the fetal origins of adult [40] Z. C. Luo et al., “Tracing the origins of “fetal origins” of disease,” International Journal of Epidemiology,vol.30,no.1,pp. adult diseases: programming by oxidative stress?” Medical 15–23, 2001. Hypotheses, vol. 66, pp. 38–44, 2006. Journal of Pregnancy 11

[41]S.K.Bhargava,H.S.Sachdev,C.H.D.Falletal.,“Relation [55] D. I. Phillips, “Fetal growth and programming of the hy- of serial changes in childhood body-mass index to impaired pothalamic-pituitary-adrenal axis,” Clinical and Experimental glucose tolerance in young adulthood,” The New England Pharmacology and Physiology,vol.28,no.11,pp.967–970,2001. Journal of Medicine,vol.350,no.9,pp.865–875,2004. [56]N.S.Levitt,R.S.Lindsay,M.C.Holmes,andJ.R.Seckl,“Dex- [42] J. Leger, M. Noel, J. M. Limal, and P. Czernichow, “Growth amethasone in the last week of pregnancy attenuates hippocam- factors and intrauterine growth retardation. II. Serum growth pal glucocorticoid receptor gene expression and elevates blood hormone, insulin-like growth factor (IGF) I, and IGF-binding pressure in the adult offspring in the rat,” Neuroendocrinology, protein 3 levels in children with intrauterine growth retardation vol.64,no.6,pp.412–418,1996. compared with normal control subjects: prospective study from [57] M.Dodic,C.N.May,E.M.Wintour,andJ.P.Coghlan,“Anearly birth to two years of age,” Pediatric Research,vol.40,no.1,pp. prenatal exposure to excess glucocorticoid leads to hypertensive 101–107, 1996. offspring in sheep,” Clinical Science,vol.94,no.2,pp.149–155, [43] S. Cianfarani, D. Germani, P. Rossi et al., “Intrauterine growth 1998. retardation: evidence for the activation of the insulin-like [58] F. H. Bloomfield, M. H. Oliver, C. D. Giannoulias, P. D. Gluck- growth factor (IGF)-related growth-promoting machinery and man, J. E. Harding, and J. R. G. Challis, “Brief undernutrition the presence of a cation-independent IGF binding protein-3 in late-gestation sheep programs the hypothalamic-pituitary- proteolytic activity by two months of life,” Pediatric Research, adrenal axis in adult offspring,” Endocrinology,vol.144,no.7, vol.44,no.3,pp.374–380,1998. pp. 2933–2940, 2003. [44] C. M. Law, A. W. Shiell, C. A. Newsome et al., “Fetal, infant, [59] D. I. W. Phillips, B. R. Walker, R. M. Reynolds et al., “Low and childhood growth and adult blood pressure: a longitudinal birth weight predicts elevated plasma cortisol concentrations in study from birth to 22 years of age,” Circulation,vol.105,no.9, adults from 3 populations,” Hypertension,vol.35,no.6,pp.1301– pp.1088–1092,2002. 1306, 2000. [45] L. S. Adair, R. Martorell, A. D. Stein et al., “Size at birth, weight [60]D.I.W.Phillips,D.J.P.Barker,C.H.D.Falletal.,“Ele- gain in infancy and childhood, and adult blood pressure in 5 vated plasma cortisol concentrations: a link between low birth low- and middle-income-country cohorts: when does weight weight and the insulin resistance syndrome?” Journal of Clinical gain matter?” The American Journal of Clinical Nutrition,vol. Endocrinology and Metabolism, vol. 83, no. 3, pp. 757–760, 1998. 89,no.5,pp.1383–1392,2009. [61] S. Arora and Anubhuti, “Role of neuropeptides in appetite [46]D.J.P.Barker,C.Osmond,T.J.Forsen,E.Kajantie,and´ regulation and obesity: a review,” Neuropeptides,vol.40,no.6, J. G. Eriksson, “Trajectories of growth among children who pp.375–401,2006. have coronary events as adults,” The New England Journal of [62]I.C.McMillen,C.L.Adam,andB.S.Muhlh¨ ausler,¨ “Early ori- Medicine,vol.353,no.17,pp.1802–1809,2005. gins of obesity: programming the appetite regulatory system,” [47]A.Singhal,M.Fewtrell,T.J.Cole,andA.Lucas,“Lownutrient Journal of Physiology,vol.565,no.1,pp.9–17,2005. intake and early growth for later insulin resistance in adoles- [63]J.K.Elmquist,C.F.Elias,andC.B.Saper,“Fromlesionsto cents born preterm,” The Lancet,vol.361,no.9363,pp.1089– leptin: hypothalamic control of food intake and body weight,” 1097, 2003. Neuron,vol.22,no.2,pp.221–232,1999. [48] J. K. Willcox, S. L. Ash, and G. L. Catignani, “Antioxidants and [64]R.D.Cone,M.A.Cowley,A.A.Butler,W.Fan,D.L.Marks, prevention of chronic disease,” Critical Reviews in Food Science andM.J.Law,“Thearcuatenucleusasaconduitfordiverse and Nutrition,vol.44,no.4,pp.275–295,2004. signals relevant to energy homeostasis,” International Journal of [49] S. Lenzen, “Oxidative stress: the vulnerable 𝛽-cell,” Biochemical Obesity,vol.25,supplement5,pp.S63–S67,2001. Society Transactions,vol.36,no.3,pp.343–347,2008. [65] P. G. Guyenet, “The sympathetic control of blood pressure,” [50] A. Ceriello and E. Motz, “Is oxidative stress the pathogenic Nature Reviews Neuroscience,vol.7,no.5,pp.335–346,2006. mechanism underlying insulin resistance, diabetes, and car- [66] G. Dorner,¨ A. Plagemann, J. Ruckert et al., “Teratogenetic diovascular disease? The common soil hypothesis revisited,” maternofoetal transmission and prevention of diabetes suscep- Arteriosclerosis, Thrombosis, and Vascular Biology,vol.24,no. tibility,” Experimental and Clinical Endocrinology and Diabetes, 5, pp. 816–823, 2004. vol. 91, no. 3, pp. 247–258, 1988. [51] P. Gupta, M. Narang, B. D. Banerjee, and S. Basu, “Oxidative [67] S. Fahrenkrog, T. Harder, E. Stolaczyk et al., “Cross-fostering stress in term small for gestational age neonates born to to diabetic rat dams affects early development of mediobasal undernourished mothers: a case control study,” BMC Pediatrics, hypothalamic nuclei regulating food intake, body weight, and vol. 4, article 14, 2004. metabolism,” Journal of Nutrition,vol.134,no.3,pp.648–654, [52]G.Baydas,F.Karatas,M.F.Gursuetal.,“Antioxidantvitamin 2004. levels in term and preterm infants and their relation to maternal [68] T. Harder, K. Franke, A. Plagemann, and R. Kohlhoff, “Early vitamin status,” Archives of Medical Research,vol.33,no.3,pp. nutrition and later blood pressure: effect of maternal diabetes,” 276–280, 2002. Journal of Pediatrics,vol.139,no.6,pp.905–906,2001. [53] E. Peuchant, J. L. Brun, V. Rigalleau et al., “Oxidative and [69] M. H. Vickers, P. D. Gluckman, A. H. Coveny et al., “Neona- antioxidative status in pregnant women with either gestational tal leptin treatment reverses developmental programming,” or type 1 diabetes,” Clinical Biochemistry,vol.37,no.4,pp.293– Endocrinology,vol.146,no.10,pp.4211–4216,2005. 298, 2004. [70] S. G. Bouret, S. J. Draper, and R. B. Simerly, “Formation of [54]R.Lingas,F.Dean,andS.G.Matthews,“Maternalnutrient projection pathways from the arcuate nucleus of the hypotha- restriction (48 h) modifies brain corticosteroid receptor expres- lamus to hypothalamic regions implicated in the neural control sion and endocrine function in the fetal guinea pig,” Brain of feeding behavior in mice,” Journal of Neuroscience,vol.24,no. Research,vol.846,no.2,pp.236–242,1999. 11, pp. 2797–2805, 2004. 12 Journal of Pregnancy

[71] S. G. Bouret, S. J. Draper, and R. B. Simerly, “Trophic action of a low-protein diet in utero,” Obesity Research, vol. 11, no. 3, pp. leptin on hypothalamic neurons that regulate feeding,” Science, 461–468, 2003. vol.304,no.5667,pp.108–110,2004. [88] L. Bellinger, C. Lilley, and S. C. Langley-Evans, “Prenatal expo- [72] A. P.Wolffe and M. A. Matzke, “Epigenetics: regulation through sure to a maternal low-protein diet programmes a preference repression,” Science,vol.286,no.5439,pp.481–486,1999. for high-fat foods in the young adult rat,” British Journal of [73] R. Jaenisch and A. Bird, “Epigenetic regulation of gene expres- Nutrition,vol.92,no.3,pp.513–520,2004. sion: how the genome integrates intrinsic and environmental [89]D.M.Campbell,M.H.Hall,D.J.P.Barker,J.Cross,A.W. signals,” Nature Genetics,vol.33,pp.245–254,2003. Shiell, and K. M. Godfrey, “Diet in pregnancy and the offspring’s [74] C. L. Peterson and M. A. Laniel, “Histones and histone modifi- blood pressure 40 years later,” British Journal of Obstetrics and cations,” Current Biology,vol.14,pp.R546–551,2004. Gynaecology,vol.103,no.3,pp.273–280,1996. [75]G.L.Wolff,R.L.Kodell,S.R.Moore,andC.A.Cooney, [90] D. S. Ludwig, “The glycemic index: physiological mechanisms “Maternal epigenetics and methyl supplements affect agouti relating to obesity, diabetes, and cardiovascular disease,” Journal (𝑣𝑦) gene expression in A /a mice,” Journal of the Federation of of the American Medical Association,vol.287,no.18,pp.2414– American Societies for Experimental Biology,vol.12,no.11,pp. 2423, 2002. 949–957, 1998. [91]I.Hwang,H.Ho,B.B.Hoffman,andG.M.Reaven,“Fructose- [76] R. A. Waterland and R. L. Jirtle, “Transposable elements: targets induced insulin resistance and hypertension in rats,” Hyperten- for early nutritional effects on epigenetic gene regulation,” sion,vol.10,no.5,pp.512–516,1987. Molecular and Cellular Biology, vol. 23, no. 15, pp. 5293–5300, [92] L. Sedovˇ a,´ O. Seda,ˇ L. Kazdova´ et al., “Sucrose feeding during 2003. pregnancy and lactation elicits distinct metabolic response in [77]R.J.Miltenberger,R.L.Mynatt,J.E.Wilkinson,andR.P. offspring of an inbred genetic model of metabolic syndrome,” Woychik,“Theroleoftheagoutigeneintheyellowobese The American Journal of Physiology,vol.292,no.5,pp.E1318– syndrome,” Journal of Nutrition,vol.127,no.9,pp.1902–1907, E1324, 2007. 1997. [93] C. M. Lenders, M. L. Hediger, T. O. Scholl, C. S. Khoo, G. B. [78]J.H.Park,D.A.Stoffers,R.D.Nicholls,andR.A.Simmons, Slap, and V. A. Stallings, “Gestational age and infant size at “Development of type 2 diabetes following intrauterine growth birth are associated with dietary sugar intake among pregnant retardation in rats is associated with progressive epigenetic adolescents,” Journal of Nutrition, vol. 127, no. 6, pp. 1113–1117, silencing of Pdx1,” Journal of Clinical Investigation,vol.118,no. 1997. 6, pp. 2316–2324, 2008. [94] G. S. Berkowitz, R. H. Lapinski, R. Wein, and D. Lee, [79] R. A. Simmons, L. J. Templeton, and S. J. Gertz, “Intrauterine “Race/ethnicity and other risk factors for gestational diabetes,” growth retardation leads to the development of type 2 diabetes The American Journal of Epidemiology,vol.135,no.9,pp.965– in the rat,” Diabetes,vol.50,no.10,pp.2279–2286,2001. 973, 1992. [80]T.J.Roseboometal.,“Hungryinthewomb:whatarethe [95] J. R. Green, I. G. Pawson, L. B. Schumacher, J. Perry, and N. consequences? Lessons from the Dutch famine,” Maturitas,vol. Kretchmer, “Glucose tolerance in pregnancy: ethnic variation 70, pp. 141–145, 2011. and influence of body habitus,” The American Journal of Obstet- [81] B. T. Heijmans, E. W. Tobi, A. D. Stein et al., “Persistent rics and Gynecology,vol.163,no.1I,pp.86–92,1990. epigenetic differences associated with prenatal exposure to [96] G. Dorner¨ and A. Plagemann, “Perinatal hyperinsulinism famine in humans,” Proceedings of the National Academy of as possible predisposing factor for diabetes mellitus, obesity Sciences of the United States of America,vol.105,no.44,pp. and enhanced cardiovascular risk in later life,” Hormone and 17046–17049, 2008. Metabolic Research,vol.26,no.5,pp.213–221,1994. [82] L. H. Lumey, “Decreased birthweights in infants after maternal [97] T. D. Clausen, E. R. Mathiesen, T. Hansen et al., “Overweight in utero exposure to the Dutch famine of 1944-1945,” Paediatric and the metabolic syndrome in adult offspring of women with and Perinatal Epidemiology,vol.6,no.2,pp.240–253,1992. diet-treated gestational diabetes mellitus or type 1 diabetes,” [83] K. M. Godfrey, A. Sheppard, P. D. Gluckman et al., “Epigenetic Journal of Clinical Endocrinology and Metabolism,vol.94,no. gene promoter methylation at birth is associated with child’s 7, pp. 2464–2470, 2009. later adiposity,” Diabetes, vol. 60, no. 5, pp. 1528–1534, 2011. [98] C. Liang, M. E. Oest, and M. R. Prater, “Intrauterine exposure [84] G. C. Burdge and K. A. Lillycrop, “Nutrition, epigenetics, to high saturated fat diet elevates risk of adult-onset chronic and developmental plasticity: implications for understanding diseases in C57BL/6 mice,” Birth Defects Research B,vol.86,no. human disease,” Annual Review of Nutrition,vol.30,pp.315– 5, pp. 377–384, 2009. 339, 2010. [99]S.Bouanane,N.B.Benkalfat,F.Z.BabaAhmedetal.,“Time [85] K. A. Lillycrop, J. Rodford, E. S. Garratt et al., “Maternal protein course of changes in serum oxidant/antioxidant status in restriction with or without folic acid supplementation during overfed obese rats and their offspring,” Clinical Science,vol.116, pregnancy alters the hepatic transcriptome in adult male rats,” no. 8, pp. 669–680, 2009. British Journal of Nutrition,vol.103,no.12,pp.1711–1719,2010. [100] C. Liang, K. DeCourcy, and M. R. Prater, “High-saturated-fat [86]J.Petrik,B.Reusens,E.Aranyetal.,“Alowproteindietalters diet induces gestational diabetes and placental vasculopathy in the balance of islet cell replication and apoptosis in the fetal C57BL/6 mice,” Metabolism,vol.59,no.7,pp.943–950,2010. andneonatalratandisassociatedwithareducedpancreatic [101]A.E.Frias,T.K.Morgan,A.E.Evansetal.,“Maternalhigh-fat expression of insulin-like growth factor-II,” Endocrinology,vol. diet disturbs uteroplacental hemodynamics and increases the 140, no. 10, pp. 4861–4873, 1999. frequency of stillbirth in a nonhuman primate model of excess [87]C.A.Maloney,A.K.Gosby,J.L.Phuyal,G.S.Denyer,J. nutrition,” Endocrinology,vol.152,no.6,pp.2456–2464,2011. M. Bryson, and I. D. Caterson, “Site-specific changes in the [102]C.E.McCurdy,J.M.Bishop,S.M.Williamsetal.,“Maternal expression of fat-partitioning genes in weanling rats exposed to high-fat diet triggers lipotoxicity in the fetal livers of nonhuman Journal of Pregnancy 13

primates,” Journal of Clinical Investigation,vol.119,no.2,pp. reduces metabolic syndrome in 6- to 8-year-old children in 323–335, 2009. rural nepal,” Journal of Nutrition,vol.139,no.8,pp.1575–1581, [103] E. Di Cintio, F. Parazzini, L. Chatenoud et al., “Dietary factors 2009. and risk of spontaneous abortion,” European Journal of Obstet- [118] C. S. Yajnik, S. S. Deshpande, A. A. Jackson et al., “Vitamin rics Gynecology and Reproductive Biology,vol.95,no.1,pp.132– B12 and folate concentrations during pregnancy and insulin 136, 2001. resistance in the offspring: the Pune Maternal Nutrition Study,” [104]R.G.Moses,J.L.Shand,andL.C.Tapsell,“Therecurrenceof Diabetologia,vol.51,no.1,pp.29–38,2008. gestational diabetes: could dietary differences in fat intake be an [119] S. J. Lewis, S. Leary, G. Davey Smith, and A. Ness, “Body explanation?” Diabetes Care,vol.20,no.11,pp.1647–1650,1997. composition at age 9 years, maternal folate intake during preg- [105] S. M. Grundy, “Hypertriglyceridemia, atherogenic dyslipi- nancy and methyltetrahydrofolate reductase (MTHFR) C677T demia, and the metabolic syndrome,” The American Journal of genotype,” British Journal of Nutrition,vol.102,no.4,pp.493– Cardiology,vol.81,no.4,pp.18B–25B,1998. 496, 2009. [106] G. S. Berenson, W. A. Wattigney, R. E. Tracy et al., “Atheroscle- [120]E.Ota,R.Tobe-Gai,R.Mori,andD.Farrar,“Antenataldietary rosis of the aorta and coronary arteries and cardiovascular advice and supplementation to increase energy and protein risk factors in persons aged 6 to 30 years and studied at intake,” Cochrane Database of Systematic Reviews,vol.9,Article necropsy (The Bogalusa Heart Study),” The American Journal of ID CD000032, 2012. Cardiology,vol.70,no.9,pp.851–858,1992. [121] P. Georgiades, A. C. Fergyson-Smith, and G. J. Burton, “Com- [107]R.L.Holman,H.C.McGill,J.P.Strong,andJ.C.Geer,“The parative developmental anatomy of the murine and human natural history of atherosclerosis: the early aortic lesions as seen definitive placentae,” Placenta,vol.23,no.1,pp.3–19,2002. in New Orleans in the middle of the of the 20th century,” The [122]B.Cox,M.Kotlyar,A.I.Evangelouetal.,“Comparativesystems Americanjournalofpathology,vol.32,no.2,pp.209–235,1958. biology of human and mouse as a tool to guide the modeling of [108]W.F.Enos,R.H.Holmes,andJ.Beyer,“Coronarydisease human placental pathology,” Molecular Systems Biology,vol.5, among United States soldiers killed in action in Korea: prelim- article 279, pp. 1–15, 2009. inary report,” Journal of the American Medical Association,vol. [123] A. M. Carter, “Animal models of human placentation: a review,” 256, no. 20, pp. 2859–2862, 1986. Placenta, vol. 28, pp. S41–S47, 2007. [109] J. J. McNamara, M. A. Molot, J. F. Stremple, and R. T. Cutting, [124]M.M.Elahi,F.R.Cagampang,D.Mukhtar,F.W.Anthony,S.K. “Coronary artery disease in combat casualties in Vietnam,” Ohri, and M. A. Hanson, “Long-term maternal high-fat feeding Journal of the American Medical Association,vol.216,no.7,pp. from weaning through pregnancy and lactation predisposes 1185–1187, 1971. offspring to hypertension, raised plasma lipids and fatty liver [110] J. P. Strong, “Natural history and risk factors for early human in mice,” British Journal of Nutrition,vol.102,no.4,pp.514–519, atherogenesis. Pathobiological Determinants of Atherosclerosis 2009. in Youth (PDAY) Research Group,” Clinical Chemistry,vol.41, [125] M. E. Symonds, S. P. Sebert, and H. Budge, “The impact of diet no.1,pp.134–138,1995. during early life and its contribution to later disease: critical [111] L. Goldsmith, S. Katz, B. Gilchrest, A. Paller, D. Leffell, and K. checkpoints in development and their long-term consequences Wolff, “Micronutrients,”in Fitzpatrick’S Dermatology in General for metabolic health,” Proceedings of the Nutrition Society,vol. Medicine, pp. 1264–1264, McGraw-Hill, 8th edition, 2012. 68,no.4,pp.416–421,2009. [112] P. Christian and C. P. Stewart, “Maternal micronutrient defi- [126] D. Vager˚ o¨ and D. Leon, “Ischaemic heart disease and low birth ciency, fetal development, and the risk of chronic disease,” weight: a test of the fetal-origins hypothesis from the Swedish Journal of Nutrition,vol.140,no.3,pp.437–445,2010. twin registry,” The Lancet, vol. 343, no. 8892, pp. 260–263, 1994. [113] J. F. Rosario, M. P. Gomez, and P. Anbu, “Does the maternal micronutrient deficiency (copper or zinc or vitamin E) modulate the expression of placental 11 𝛽 hydroxysteroid dehydrogenase-2 per se predispose offspring to insulin resistance and hypertension in later life?” Indian Journal of Physi- ology and Pharmacology,vol.52,no.4,pp.355–365,2008. [114] A. L. Tomat, F. Inserra, L. Veiras et al., “Moderate zinc restric- tionduringfetalandpostnatalgrowthofrats:effectsonadult arterial blood pressure and kidney,” The American Journal of Physiology,vol.295,no.2,pp.R543–R549,2008. [115] A. L. Tomat, M. D. L. T. Costa, and C. T. Arranz, “Zinc restriction during different periods of life: influence in renal and cardiovascular diseases,” Nutrition,vol.27,no.4,pp.392–398, 2011. [116]A.Vaidya,N.Saville,B.P.Shrestha,A.M.deLCostello,D. S. Manandhar, and D. Osrin, “Effects of antenatal multiple micronutrient supplementation on children’s weight and size at 2 years of age in nepal: follow-up of a double-blind randomised controlled trial,” The Lancet,vol.371,no.9611,pp.492–499, 2008. [117] C.P.Stewart,P.Christian,K.J.Schulze,S.C.LeClerq,K.P.West, and S. K. Khatry, “Antenatal micronutrient supplementation Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 165617, 9 pages http://dx.doi.org/10.1155/2013/165617

Research Article Physical Activity during Pregnancy: Impact of Applying Diﬀerent Physical Activity Guidelines

Katie M. Smith and Christina G. Campbell

InterdepartmentalGraduatePrograminNutritionalSciences,DepartmentofFoodScienceandHumanNutrition, Iowa State University, 220 MacKay Hall, Ames, IA 50011, USA

Correspondence should be addressed to Christina G. Campbell; [email protected]

Received 12 October 2012; Revised 10 December 2012; Accepted 18 December 2012

Academic Editor: Michelle F. Mottola

Copyright © 2013 K. M. Smith and C. G. Campbell. is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Multiple guidelines and de�nitions of physical activity (PA) have been used to study the bene�ts of activity during pregnancy. e diﬀerent guidelines lead to a wide range of prevalence estimates and this has led to con�icting reports about activity patterns during pregnancy. A longitudinal study was conducted to assess PA using a pattern-recognition monitor for a 7-day period at week 18 ( )andweek35( ) of pregnancy. e amount of activity performed and the number of women meeting six diﬀerent PA guidelines were evaluated. Adherence to PA guidelines ranged from 5 to 100 and 9 to 100 at weeks 18 and 35, respectively. All women𝑛𝑛 푛푛푛 achieved the 500𝑛𝑛 MET-minute 푛푛푛 guideline and nearly all women accumulated 150 minutes of weekly moderate-vigorous physical activity (MVPA) at both time points. Only 22 and 26 participated in%3 sessions of MVPA% lasting 30 minutes at both time points and this further declined to 5 and 9 when the guideline was increased to≥ 5 sessions of 30 minutes. e amount of PA during pregnancy varied drastically depending on% which guideline% was used.≥ Further research is warranted≥ to clearly identify the patterns of activity that are associated% with healthy% pregnancy outcomes. ≥

1. Introduction program while pregnant. Considerable evidence was published regarding the safety of maternal exercise between the Views on physical activity and exercise during pregnancy 1980s and early 1990s supporting the need for updated and have taken on new meanings and implications throughout revised exercise guidelines [3, 5, 6]. Consequently, ACOG history. e importance of maternal physical activity dates as responded in 1994 by eliminating the constraints on heart far back as the third century BC when Aristotle eluded to the rate and exercise duration, stating that exercise can be done difficulty endured during childbirth as a result of a sedentary in moderation but not to exhaustion [7]. Finally, in 2002, maternal lifestyle [1]. However, society and expert opinions ACOG issued a statement promoting the health bene�ts and have not always supported the prenatal exercise since that safety of exercise in pregnancy for both previously active time. For many years maternal, exercise was thought to harm and inactive women (assuming medical clearance and no the fetus or promote adverse pregnancy outcomes such as contraindications are present) [2]. ese recommendations preterm delivery and fetal growth restriction or small for of 30 minutes or more of moderate exercise on most, if not gestational age infants [2, 3]. In 1985, the American College all, days of the week were reaffirmed by ACOG in 2009 [2]. of Obstetricians and Gynecologists (ACOG) published the Most recently, physical activity recommendations for �rst exercise guidelines for pregnant women. ese included pregnant women were also included in the �rst ever Physical limitations on heart rate and duration, restricting heart rate Activity Guidelines for Americans published in 2008 by the to 140 beats per minute, and exercise to no more than 15 United States Department of Health and Human Services minutes at a time [4]. Furthermore, women that were inactive (DHHS) [8]. In this document, pregnant women (previously prior to pregnancy were not advised to begin an exercise active and inactive) are encouraged to engage in at least 150 2 Journal of Pregnancy minutes of moderate-intensity aerobic activity each week. of this study was twofold: (1) to evaluate the difference in Women already doing regular activity of vigorous intensity the minutes of physical activity that women participate in may continue provided that they remain healthy and discuss during pregnancy depending upon what guideline is used and their activity with their healthcare provider over time [8]. (2) to compare the percentage of women that meet physical Recommendations for the nonpregnant population are very activity guidelines during pregnancy depending on what similar; however, they speci�cally state that the activity can be guideline is used. Data were evaluated for the second and accumulated in minimum bouts of 10 minutes. e recom- third trimesters to demonstrate the impact of these guidelines mendation for shorter sustained bouts of activity transpired across pregnancy. from a summary of experimental �ndings in nonpregnant adults suggesting that activity performed at a level of at least moderate intensity and sustained for at least 10 minutes at 2. Materials and Methods a time was as effective as single, longer bouts of activity 2.1. Participants. Eighty-nine healthy pregnant women prior in lowering chronic disease risk [9]. e Physical Activ- to 18-week gestation were enrolled for a larger longitudinal ity Guidelines for all American adults, including pregnant study analyzing the relationship between maternal exercise women, encourage the activity to be spread throughout the and fetal docosahexaenoic acid status. Participants were week. �owever, speci�c minimum bouts of activity clarifying recruited via local obstetric clinics, �iers placed in town, and “what counts” towards meeting activity guidelines during online and campus-wide emails. All women were screened pregnancy, such as the 10-minute bouts for nonpregnant to ensure they met the study’s inclusion criteria (singleton adults, are not explicitly stated in the pregnancy guidelines pregnancy and maternal age of 18–45 years of age) and for Americans [8]. Conversely, the 2011 Canadian Physical exclusion criteria (smoker or history of chronic disease) Activity Guidelines suggest that accumulating 150-minutes which was veri�ed by each participant’s primary obstetric of weekly moderate- to vigorous-intensity aerobic physical medical provider. Due to the observational design of the activity in bouts of 10-minutes or more may be appropriate study, no additional medical prescreens for exercise were forpregnantwomen[10]. needed. Nineteen women did not complete the study due Previous studies have evaluated the prevalence of activ- to pregnancy complications or personal time constraints. ity during pregnancy using multiple interpretations of the Additionally, 13 other women at week 18 and 1 woman at ACOG guidelines. Some studies have focused on accumu- week 35 had incomplete datasets, while 2 other women at lating at least 30 minutes of moderate-vigorous physical week 18 and 3 women at week 35 had substantial off-body activity (MVPA) throughout the day [11, 12]. For example, time (further described in “Data processing”); thus data were McParlin et al. assessed the percentage of overweight and analyzed for 55 women at week 18 and 66 women at week obese pregnant women accumulating at least 30 minutes 35. e protocol was approved by the campus’ Institutional of MVPA per day [11]. eir results revealed 63 , 62 , Review Board. and 71 of women accumulating the recommended amount of activity in the 1st, 2nd, and 3rd trimester, respectively.% % Conversely,% Chandonnet et al. compared the amount of 2.2. Study Design. Participants visited the research facility accumulated MVPA performed by obese pregnant women to for 2 data collection periods lasting 7 days each at 18 MVPA performed in at least 10-minute bouts [12]. Average and 35 weeks ( 1 week) of pregnancy. During the week total daily activity was drastically reduced by 66 minutes 18 appointment, participants provided written informed perdaywhenonlytheactivitythatlastedforatleasta10- consent and completed± a medical history questionnaire minute bout was counted. ese two studies demonstrate that indicating their age, ethnicity, education, parity, height, pre- how the guideline is interpreted in�uences the number of pregnancy weight, and due date. At the beginning of both data women meeting the ACOG recommendations. As described, collection periods, participants were weighed without shoes multiple guidelines have been used to assess levels of physical using a Sunbeam analog scale (2008 Sunbeam Products, Inc., activity and exercise throughout pregnancy. e differences Boca Raton, Florida). A SenseWear Mini Armband monitor in guidelines have led to widely disparate estimates of the (Model Name: MF) (SWA) (BodyMedia, Inc., Pittsburgh, prevalence of pregnant women achieving physical activity Pennsylvania) was then con�gured for each woman to quan- guidelines during pregnancy—ranging from 3 to 78.4 [11, tify physical activity. e monitor was initialized according 13–20]. Furthermore, the use of multiple guidelines has to her height, weight, age, and handedness and placed on her contributed to con�icting evidence regarding the role% of upper le arm per manufacturer’s instructions. Participants physical activity to improve certain pregnancy outcomes, were instructed to wear the monitor for the subsequent 7-day such as healthy gestational weight gain, insulin sensitivity, period, 24 hours a day except during any water submersion and preeclampsia. �o date, no study has speci�cally evaluated activities such as showering and swimming. Additionally, the impact of different physical activity guidelines on the all the daily activity was to be documented in a provided reported patterns of physical activity using an objective physical activity record to con�rm activity while the monitor assessment tool evaluated for use in pregnancy. Understand- was not worn (e.g., swimming and bathing). Staff members ing the implications of the subtle, but distinct, differences instructed each woman to participate in her normal daily between guidelines is an important consideration to explain activity and return the SWA and physical activity record to the inconsistency of previous studies and to improve the the research facility at the end of the 7-day data collection reporting of physical activity during pregnancy. e purpose period. e amount of moderate and vigorous activity was Journal of Pregnancy 3 assessed by the SWA except for activity endured during [2, 7, 8, 11–15, 17–20, 24]. e guidelines included (1) 150- water submersion (e.g., swimming or water aerobics). Water minutes of accumulated moderate-vigorous physical activity activity was con�rmed by the physical activity record. (MVPA) [2, 11, 17, 19], (2) 150 minutes of MVPA performed in periods of at least 10 minutes [12], (3) 150 minutes of MVPA performed for at least 10 minutes with 1 minute of 2.3. Physical Activity Armband. e validity of the SWA vigorous activity equivalent to 2 minutes of moderate activity to predict energy expenditure in pregnant women has (M2VPA) [8], (4) at least 3 sessions of MVPA [7, 24] sustained been previously assessed and correlated well with indirect for at least 30 minutes at a time [14, 16, 23], (5) at least 5 calorimetry, [21]. e monitor is a pattern- sessions of MVPA sustained for at least 30-minutes at a time recognition monitor2 with a triaxial accelerometer, heat-�ux [13, 15, 20], and (6) at least 500 MET minutes accumulated thermometers,𝑅𝑅 a galvanic= 0.86 skin response sensor, and a skin throughout the week [18]. Four of these guidelines included temperature sensor. e SWA uses these sensors via the use the use of a minimum bout of activity, 10 (guideline 2 and of proprietary algorithms to predict the energy expenditure. 3) or 30 minutes (guideline 4 and 5). When analyzing data e monitor records data in 1-minute epochs and provides a for these 4 guidelines, interruptions of 1 or 2 minutes below metabolic equivalent (MET) value for each minute of activity the moderate intensity threshold within a 10-minute period using the equation METS = kcal hour kg . e raw SWA were allowed, as has been previously reported with analysis of �les were sent to the manufacturer (BodyMedia,−1 −1 Inc.) and physical activity data in the nonpregnant population [25, 26]. processed with algorithm 5.2. ⋅ ⋅ e percentage of women that met the physical activity guidelines were classi�ed according to three categories: (1) 2.4. Data Processing. SWA data �les returned from the sufficient activity to meet the guideline, (2) insufficient manufacturer were exported into Microso Office Excel 2007 activity to meet the guideline, and (3) no activity. Sufficient, (Microso, Redmond, WA,USA). Excel code was written insufficient, and no activity are de�ned for each of the six to identify and categorize total minutes spent in moderate guidelines in Table 1. Because it is not physically possible or vigorous activity. Previous research with the SWA in to accumulate zero MET minutes of activity throughout the this population demonstrated an overestimation of energy entire week, the “no activity” category was not used for the MET minute de�nition (guideline 6). expenditure by approximately 20 [21]. To adjust for this overestimation, 20 was added2 to the standard MET thresholds for moderate (3–5.9% (𝑅𝑅 METs)= 0.86) and vigorous 2.6. Statistical Analyses. Descriptive characteristics of the activity ( 6 METs) such that the% SWA MET thresholds used participants are summarized using mean standard devi- in the current study were 3.6–7.1 METs (moderate) and 7.2 ation (SD). Physical activity data were tested for normality METs (vigorous).≥ Similarly, total accumulated MET minutes prior to any analysis using the D’Agostino-Pearson± test and were reduced by 20 (independently of categorizing METs≥ visually analyzed with the use of normality plots and his- into increased intensity thresholds). SWA data �les were tograms to evaluate the distribution of the data. Total minutes thoroughly reviewed% to identify periods of nonwear time of weekly physical activity according to the parameters of to ensure �les were as close to 24-hours of wear time as each physical activity guideline are described as medians and possible. Two women did not wear the monitor at night. interquartile ranges (IQR). e number of women meeting is particular nonwear time activity was con�rmed to be each physical activity guideline is represented graphically and spent sleeping by checking the physical activity record and as percentages. All statistical analyses were performed in Med subsequently this time was �lled as sedentary time, equivalent Calc version 12.3 (MedCalc Soware, Mariakerke, Belgium). to 0.95 METs [22]. Twenty-six women participated in water activities such as swimming or aqua aerobics. Nonwear 3. Results and Discussion time spent doing water activities was accounted for using corresponding MET values from the 2011 Compendium of 3.1. Results: Participant Characteristics. On average, the Physical Activities [22]. To evaluate the remaining nonwear women were years old, primarily Caucasian (94 , time, it was assumed that nearly 1 hour of self-care per ), had previous pregnancy (not including the day would result in approximately 500 minutes of nonwear current pregnancy),29 ± 4.2 had live births, and had a pre-% time per week; thus �les with more than 500 minutes per 𝑛𝑛pregnancy𝑛𝑛 BMI1 ± of 1.4 kg m . week of off-body time were deemed as noncompliant and 0.8 ± 1.3 2 excluded from the analysis for that data collection period. 3.2. Results: Minutes24.9 of ± Moderate-Vigorous 4.7 ⋅ Physical Activity. is excluded 2 �les at week 18 and 3 �les at week 35 as Results from the D’Agostino-Pearson test and normality plots previously mentioned in paragraph 2.1. revealed a nonnormal distribution (week 18: and week 35: ) when all guidelines were applied to the 2.5. Physical Activity Guideline. e armband data �les were data except weekly accumulated MVPA at week𝑃𝑃 35 (guideline푃 𝑃�푃 processed and evaluated using six different physical activity 1) ( 𝑃𝑃 푃 𝑃𝑃𝑃)andMETminutesatbothweek18and35 guidelines. ese guidelines were either previously used to (guideline 6) ( and , resp.). Minutes of de�ne women that were exercising regularly throughout MVPA𝑃𝑃 𝑃𝑃𝑃 as assessed by each guideline are reported in Table pregnancy [16, 23] or to assess prevalence of women meeting 2. e very small𝑃𝑃 𝑃𝑃 differences푃� in𝑃𝑃 minutes𝑃𝑃𝑃 of activity between weekly physical activity recommendations during pregnancy MVPA bouts (wk 18 : 141 min; wk 35 : 118 min) and M2VPA 4 Journal of Pregnancy

T 1: �e�nitions of physical activity guidelines and activity categori�ations.

PA guideline MVPA MVPA bouts M2VPA bouts 3 30 5 30 MET PA 150 min 150 min MVPA 5 sessions 150 min MVPA performed in 10 min 3 sessions∗ MVPA MVPA∗ Sufficient accumulated 500 MET min performed≥ in bouts≥ with 1 min VPA = sustained for 30 mins sustained≥ for ≥MVPA 10 min bouts 2 min MPA≥ ≥ 30 mins ≥ 1–149 min 1–149 min MVPA ≥ 1–4 sessions 1–149 min ≥ MVPA performed in 10 min 1-2 sessions MVPA ≥MVPA 1–499 MET Insufficient accumulated performed in bouts with 1 min VPA = sustained for 30 mins sustained for minutes MVPA 10 min bouts 2 min MPA≥ 30 mins ≥ 0 sessions 0 min 0 min MVPA 0 min MVPA performed ≥ 0 sessions MVPA ≥MVPA None accumulated performed in in 10 min bouts with N/A sustained for 30 mins sustained for MVPA 10 min bouts 1 min VPA = 2 min MPA 30 mins ≥ ≥ Sufficient activity was de�ned as enough≥ activity to meet the guideline. MVPA: moderate-vigorous physical activity; M2VPA: 1 minute of vigorous physical activity is equivalent to 2 minutes of moderate activity. N/A: e “none” category is not applicable to this guideline. ≥

T 2: Weekly minutes spent in moderate-vigorous physical 100 activity according to multiple guidelines. 90

Week 18 Week 35 80 Median min wk Median min wk 70 (IQR) −1 (IQR) −1 MVPA 455 (351–585)⋅ 468 (240–644)⋅ 60 MVPA bouts 𝑛𝑛 푛푛푛 𝑛𝑛 푛푛푛 141 (79–199) 118 (31–257) 50 M2VPA bouts 145 (86–221) 125 (32–268) 40 3 30 1 bout (0–2) 1 bout (0–3)

5 30 1 bout (0–2) 1 bout (0–3) Percent of participants 30 ∗ MET PA 10664 10433 20 ∗ (10052–11228) (9587–11288) MVPA: moderate-vigorous physical activity accumulated throughout the 10 week; MVPA bouts: moderate-vigorous physical activity performed in bouts 0 of at least 10 minutes; M2VPA bouts: moderate-vigorous physical activity Week 18 (n = 55) Week 35 (n = 66) performed in bouts of at least 10 minutes with 1 minute of vigorous physical activity equivalent to 2 minutes of moderate activity; 3 30: At least 3 ∗ sessions of moderate-vigorous physical activity sustained for at least 30 MVPA 3 30 ∗ minutes; 5 30: At least 5 sessions of moderate-vigorous∗ physical activity MVPA bouts 5 30 sustained for at least 30 minutes; MET PA: total accumulation of weekly MET M2VPA bouts MET PA minutes. ∗ F 1: Percentage of pregnant women meeting physical activity guidelines. MVPA: moderate-vigorous physical activity accumulated throughout the week; MVPA bouts: moderate-vigorous physi- bouts (wk 18 : 145 min; wk 35 : 125 min) are explained by the cal activity performed in bouts of at least 10 minutes; M2VPA bouts: very small amount of vigorous activity performed by this moderate-vigorous physical activity performed in bouts of at least 10 population. minutes with 1 minute of vigorous physical activity equivalent to 2 minutes of moderate activity; 3 30: at least 3 sessions of moderate- vigorous physical activity sustained for at least 30 minutes; 5 30: 3.3. Results: Adherence to Physical Activity Guidelines. e at least 5 sessions of moderate-vigorous physical activity sustained percentage of women meeting physical activity guidelines ∗ for at least 30 minutes; MET PA: total accumulation of weekly MET∗ ranged from 5 to 100 atweek18and9 to 100 minutes. during week 35 (see Figure 1). e percentage of women participating in% “no activity”% ranged from 0 to 42% at week% 18 and 0 to 44 at week 35 (see Table 3). All women met the MET minute guideline of at least 500 MET minutes% of 3.4. Discussion. e current study revealed wide ranges in accumulated weekly% activity at both week 18 and week 35. both amount of physical activity performed during preg- Guidelines requiring 3 or 5 sustained bouts of 30 minutes nancy and percentage of women meeting physical activity (guidelines 4 and 5) were met by the fewest number of women guidelines depending upon which guideline is used and at both week 18 (22 and 5 , resp.) and week 35 (26 and how guidelines are interpreted. e evaluation of multiple 9 , resp.). guidelines, the use of an objective monitor that has been % % % % Journal of Pregnancy 5 0 30 MET PA ∗ ( 𝑛𝑛 )) 30 5 % ( ∗ 44 (29) 44 (29) N/A 30 (20) 47 (31) 26 (17) 9 (6) 100 (66) 𝑛𝑛 푛 푛푛 5 (3) 52 (34) 44 (29) 30: at least 3 sessions of moderate-vigorous physical activity ∗ Week 35 of pregnancy, 5 (3) 56 (37) 39 (26) 0 9 (6) 0 30 MET PA MVPA MVPA bouts M2VPA bouts 3 ∗ ( 𝑛𝑛 )) 30 5 ( % ∗ 42 (23) 42 (23) N/A 36 (20) 53 (29) 22 (12) 5 (3) 100 (55) 91 (60) Distribution of pregnant women meeting physical activity guidelines. 𝑛𝑛 푛 푛푛 3: T 4 (2) 47 (26) 49 (27) Week 18 of pregnancy, 30: At least 5 sessions of moderate-vigorous physical activity sustained for at least 30 minutes; MET PA: total accumulation of weekly MET minutes. 4 (2) 49 (27) 47 (26) ∗ 0 MVPA MVPA bouts M2VPA bouts 3 None Insufficient 5 (3) Sufficient 95 (52) MVPA: moderate-vigorous physical activity accumulatedphysical throughout activity the performed week; in bouts MVPA of bouts: at moderate-vigorous least physical 10 activity minutes performed with 1 in minute bouts of of vigorous physical at activity least equivalent 10 to 2 minutes; minutes M2VPA of bouts: moderate moderate-vigorous activity; 3 sustained for at least 30 minutes; 5 6 Journal of Pregnancy evaluated for use in pregnancy, and the longitudinal design health outcomes in pregnant women. Rather, it is recom- contribute novel research �ndings regarding the assessment mended as a safe amount of activity assuming a healthy of prenatal physical activity. e terms “exercise” and “phys- pregnancy with no contraindications. Despite this fact, this ical activity” are commonly used interchangeably despite amount of activity has been applied as the basis of several having distinct de�nitions [27]. Guidelines set forth by prenatal physical activity interventions and may or may not ACOG encourage exercise [2] and the Department of Health be the appropriate volume to reduce the incidence of adverse and Human Services recommends physical activity [8]. e pregnancy outcomes. ese outcomes include, but are not considerable variability in the literature as to how to de�ne limited to, excessive gestational weight gain, gestational recommended activity during pregnancy has contributed, diabetes, and preeclampsia. As a result, the current literature among other factors, to multiple inconsistent conclusions reports considerable discrepancies as to whether or not the as to whether or not physical activity and/or exercise is an maternal physical activity is an effective approach to prevent effective way to reduce adverse pregnancy outcomes such such complications [28–34]. as excess gestational weight gain, gestational diabetes, and To advance research on physical activity in pregnancy, a preeclampsia. more systematic epidemiological approach is needed. Welk One potential reason for the inconsistencies in de�ning outlined a physical activity epidemiology model that shows physical activity during pregnancy and the prevalence of how different types of physical activity research interact to women meeting physical activity guidelines in the literature collectively advance the science [35]. In this model (see is the subtle differences in the wording of the ACOG exercise Figure 2), health outcomes research de�nes the appropriate guideline [2]. e abstract of the guideline recommends volume (duration, frequency, intensity) and type of activity that pregnant women engage in 30 minutes or more of related to speci�c health bene�ts for multiple populations moderate exercise a day on most, if not all, days of the all across the life span. is type of research is pertinent week. e opening paragraph of the ACOG guideline [2] to the development of physical activity guidelines and rec- then states that pregnant women can adopt the 1995 Centers ommendations because it focuses on relating the effects of for Disease Control and Prevention (CDC) and American a speci�c volume of physical activity to multiple indicators College of Sports Medicine’s (ACSM) recommendation for of health. e identi�ed volumes of activity associated with exercise for nonpregnant adults [27]. is recommendation particular health outcomes can then be incorporated into is to accumulate 30 minutes or more of moderate exercise public guidelines and recommendations for improved health, onmost,ifnotall,daysoftheweek.Incontrasttothe or in this case, improved pregnancy outcomes such as ACOG guidelines, the 1995 CDC/ACSM recommendation reduced excess gestational weight gain and prevalence of did not explicitly state a minimum length of time that activity gestational diabetes. should be sustained to count towards meeting the suggested e previously described studies used a variety of meth- 30 minutes a day. In the 2007 ACSM/American Heart ods (both subjective and objective) to assess physical activity Association’s updated recommendation on physical activity data and a variety of de�nitions to determine the adherence to and health, bouts of at least 10 minutes were recommended physical activity guidelines. Additional research is needed to [9]. However, pregnancy guidelines for Americans have not determine the volume and type of physical activity necessary yet adopted this part of the recommendation. to promote optimal health outcomes for both the mother It is possible that a lack of understanding regarding andthebaby.us,itisimportanttonotethatwhilethese the maternal and fetal bene�ts of accumulated total activity studies may not all be directly comparable to one another, versus the bene�ts of activity sustained for a minimum period the current study provides evidence to explain some of the of time (e.g., 10, 20, or 30 minutes) has contributed to the dis- inconsistencies in the literature and demonstrates the need crepancies. It has yet to be determined if pregnancy outcomes for health outcomes research to provide pregnancy-speci�c differ when activity is evaluated as the total accumulated the guidelines for particular maternal and fetal outcomes. Fur- activity above a speci�c threshold compared to activity above thermore, it is imperative to use an assessment tool, whether a speci�c threshold sustained for a certain period of time (e.g., it be subjective or objective, which has been validated for use 10, 20, or 30 minutes) [12]. Furthermore, it is possible that the in pregnant women. e activity monitor used in the current appropriate amount of physical activity during pregnancy is study is a sensitive tool that is effective at capturing physical outcome speci�c, similar to physical activity recommenda- activity [14] and additionally has been previously evaluated tions for nonpregnant adults for reduced risk of premature in pregnant women [21]. e monitor was worn 24 hours a death, cardiovascular disease, and type 2 diabetes mellitus day (with the exception of water activities) for a 7-day period. (150 minutes weekly), mental health bene�ts (3–5 times/wk Previous studies have also used accelerometry to assess for 30–60 minutes), and weight loss or weight maintenance physical activity during pregnancy, but data were commonly (at least 300 minutes) [8]. processed for partial days (e.g., 8 or 10 hours of wear time Recommendations to perform 150 minutes of moderately over a 24-hour period) [11, 36]. us, while amounts of intense activity, such as the DHHS guidelines, have been accumulated physical activity by the participants in this study adopted for use in pregnancy from nonpregnant physical may be higher than some previous reports [11, 13, 18, 36], activity guidelines. is amount of activity was not speci�- the accumulated moderate activity data shown here represent cally selected as the recommended amount of weekly activity common activity for pregnant women accrued in the course during pregnancy because of its association with particular of daily living because a tool previously evaluated in pregnant Journal of Pregnancy 7

Surveillance research Guidelines and recommendations

Theory and Physical activity Health outcomes correlates research research

Mechanistic or basic Intervention research research

F 2: A model of physical activity epidemiology reproduced from [35]. women was used, data were appropriately adjusted for previ- and activity in bouts). Furthermore, interventions have used ously reported overestimation of the assessment tool when these guidelines as a target level of physical activity for used in this population, and participants wore the monitor women to engage in during pregnancy in order to promote nearly for 97 of the 7-day monitoring period. speci�c pregnancy outcomes (e.g., healthy gestational weight A limitation of the current study is the diﬀering number gain and improved glucose tolerance) and have provided of women analyzed% at week 18 and week 35 of pregnancy. inconsistent results. Future recommendations should incor- SWA data was not available for 11 women at week 18 and porate previous and future �ndings of improved pregnancy some of these women may have exercised during their preg- outcomes with speci�c volumes of physical activity. Addi- nancy. Additionally, participation in a prenatal aqua aerobics tionally, further research is warranted to identify positive class is common in the local community. Due to the large pregnancy outcomes associated with clearly identi�ed de�- popularity of this class, there is a waiting list to participate nitions of physical activity and/or exercise. and women are typically able to begin participating in this class near weeks 20–22 of pregnancy. Since our observation Con�ict of �nterests of activity occurred near week 18 of pregnancy, it is likely that some women were not participating in this class at None of the authors report any con�ict of interests with any week 18 but were by the time physical activity was assessed of the companies listed in this paper. at week 35. erefore, the higher percentage of women meeting guidelines 4 and 5 (3 or 5 sessions of 30 minutes) Acknowledgments atweek35thanweek18islikelyduetothecombinationof these two circumstances. Lastly, physical activity was de�ned A special recognition is given to the women that participated in the current study by any activity of at least moderate in this study and the undergraduate and graduate students intensity. It was not categorized by leisure time activity of Dr. Christina Campbell’s Laboratory that assisted with the versus volitional exercise. is allowed all physical activity data collection. Considerable appreciation is also given to Dr. of at least moderate intensity to be detected by an objective Gregory J. Welk for his assistance in reviewing this paper. monitor and eliminated the potential for the recall bias or the e funding for this work was provided by a USDA Special incomplete reporting of the physical activity in the physical Grant (no. 416-28-28) provided to the Iowa State University’s activity record. It also provided the advantage to capture Nutrition and Wellness Research Center; a portion was then all accumulated, shorter bouts of at least moderate intensity awarded to C. G. Campbell. rather than just longer sustained bouts of activity. References

4. Conclusions [1] K. Vaughan, Exercises before Childbirth, Faber & Faber, London, UK, 1951. e amount of time pregnant women spend in moderate- [2] American College of Obstetrics and Gynecology (ACOG), vigorous physical activity or volitional exercise varies dras- “Committee opinion number 267: exercise during pregnancy tically depending upon what guideline is used. Previous and the postpartum period,” Obstetrics and Gynecology, vol. 99, reports regarding the prevalence of physical activity during no. 1, pp. 171–173, 2002. pregnancyhaverangedfromaslowas3 to as high as [3] G. A. L. Davies, L. A. Wolfe, M. F. Mottola, and C. MacKin- 78 . e large range is due in part to the multiple diﬀerent non, “Joint SOGC/CSEP clinical practice guideline: exercise in guidelines that have been used in these% studies and the pregnancy and the postpartum period,” Canadian Journal of interpretation% of these guidelines (e.g., accumulated activity Applied Physiology, vol. 28, no. 3, pp. 330–341, 2003. 8 Journal of Pregnancy

[4] American College of Obstetrics and Gynecology (ACOG), [20] A. M. Petersen, T. L. Leet, and R. C. Brownson, “Correlates of Technical Bulletin: Exercise During Pregnancy and the Postnatal physical activity among pregnant women in the United States,” Period, ACOG, Washington, DC, USA, 1985. Medicine and Science in Sports and Exercise, vol. 37, no. 10, pp. [5] R. G. McMurray, M. F. Mottola, L. A. Wolfe, R. Artal, L. Millar, 1748–1753, 2005. and J. M. Pivarnik, “Recent advances in understanding maternal [21] K. M. Smith, L. M. Lanningham-Foster, G. J. Welk, and C. G. and fetal responses to exercise,” Medicine and Science in Sports Campbell, “Validation of the SenseWear armband to predict and Exercise, vol. 25, no. 12, pp. 1305–1321, 1993. energy expenditure in pregnant women,” Medicine and Science [6] L. A. Wolfe and G. A. L. Davies, “Canadian guidelines for in Sports and Exercise, vol. 44, no. 10, pp. 2001–2008, 2012. exercise in pregnancy,” Clinical Obstetrics and Gynecology, vol. [22] B. E. Ainsworth, W. L. Haskell, S. D. Herrmann et al., “2011 46, no. 2, pp. 488–495, 2003. compendium of physical activities: a second update of codes [7] American College of Obstetrics and Gynecology (ACOG), and MET values,” Medicine and Science in Sports and Exercise, “ACOG Technical Bulletin Number 189: exercise during preg- vol. 43, no. 8, pp. 1575–1581, 2011. nancy and the postpartum period,” International Journal of [23] A. D. Stein, J. M. Rivera, and J. M. Pivarnik, “Measuring Gynaecology and Obstetrics, vol. 45, no. 1, pp. 65–70, 1994. energy expenditure in habitually active and sedentary pregnant [8] U.S. Department of Health and Human Services, “2008 Phys- women,” Medicine and Science in Sports and Exercise, vol. 35, ical Activity Guidelines for Americans,” ODPHP Publication no. 8, pp. 1441–1446, 2003. UOO36, U.S. Department of Health and Human Services, [24] American College of Sports Medicine, “Exercise prescription Washington, DC, USA, 2008, http://www.health.gov/PAGuide- for healthy populations and special considerations,” in Guide- lines/. lines for Exercise Testing and Prescription, W.R. ompson, Ed., pp. 183–187, Wolters Kluwer, Lippincott Williams & Wilkins, [9] W. L. Haskell, I. M. Lee, R. R. Pate et al., “Physical activity Philadelphia, Pa, USA, 8th edition, 2010. and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart [25] R. P.Troiano, D. Berrigan, K. W.Dodd, L. C. Mâsse, T.Tilert, and Association,” Circulation, vol. 116, no. 9, pp. 1081–1093, 2007. M. Mcdowell, “Physical activity in the United States measured by accelerometer,” Medicine and Science in Sports and Exercise, [10] Canadian Society for Exercise Physiology, “Canadian Physical vol. 40, no. 1, pp. 181–188, 2008. Activity Guidelines: 2011 Scienti�c Statements,” http://www .csep.ca/CMFiles/Guidelines/CanadianPhysicalActivityGuide- [26] J. M. Tucker, G. J. Welk, and N. K. Beyler, “Physical activity in linesStatements_E 203.pdf. U.S. adults: compliance with the physical activity guidelines for Americans,” AmericanJournalofPreventiveMedicine, vol. 40, [11] C. McParlin, S. C. Robson, P. W. G. Tennant et al., “Objectively no. 4, pp. 454–461, 2011. measured physical% activity during pregnancy: a study in obese [27] R. R. Pate, M. Pratt, S. N. Blair et al., “Physical activity and public and overweight women,” BMC Pregnancy and Childbirth, vol. health: a recommendation from the Centers for Disease Control 10, article 76, 2010. and Prevention and the American College of Sports Medicine,” [12] N. Chandonnet, D. Saey, N. Alméras, and I. Marc, “French Journal of the American Medical Association, vol. 273, no. 5, pp. Pregnancy Physical Activity Questionnaire compared with an 402–407, 1995. accelerometer cut point to classify physical activity among [28] B. Muktabhant, P. Lumbiganon, C. Ngamjarus, and T. pregnant obese women,” PLoS One, vol. 7, no. 6, Article ID Dowswell, “Interventions for preventing excessive weight gain e38818, 2012. during pregnancy,” Cochrane Database of Systematic Reviews, [13] K. M. Borodulin, K. R. Evenson, F. Wen, A. H. Herring, and no. 4, Article ID CD007145, 2012. A. M. Benson, “Physical activity patterns during pregnancy,” [29] B. Gardner, J. Wardle, L. Poston, and H. Croker, “Changing diet Medicine and Science in Sports and Exercise, vol. 40, no. 11, pp. and physical activity to reduce gestational weight gain: a meta- 1901–1908, 2008. analysis,” Obesity Reviews, vol. 12, no. 7, pp. e602–e620, 2011. [14] K. M. Smith, R. C. Foster, and C. G. Campbell, “Accuracy of [30] H. Skouteris, L. Hartley-Clark, M. McCabe et al., “Preventing physical activity assessment during pregnancy: an observational excessive gestational weight gain: a systematic review of inter- s tu d y,” BMC Pregnancy Childbirth, vol. 11, Article ID 86, 2011. ventions,” Obesity Reviews, vol. 11, no. 11, pp. 757–768, 2010. [15] K. R. Evenson, D. A. Savitz, and S. L. Huston, “Leisure-time [31] I. Streuling, A. Beyerlein, E. Rosenfeld, H. Hofmann, T. Schulz, physical activity among pregnant women in the US,” Paediatric and R. von Kries, “Physical activity and gestational weight gain: and Perinatal Epidemiology, vol. 18, no. 6, pp. 400–407, 2004. a meta-analysis of intervention trials,” An International Journal [16] J. Zhang and D. A. Savitz, “Exercise during pregnancy among of Obstetrics and Gynaecology, vol. 118, no. 3, pp. 278–284, 2011. US women,” Annals of Epidemiology, vol. 6, no. 1, pp. 53–59, [32] S. angaratinam, E. Rogozinska, K. Jolly et al., “Eﬀects of 1996. interventions in pregnancy on maternal weight and obstetric [17] K. R. Evenson and F. Wen, “National trends in self-reported outcomes: meta-analysis of randomised evidence,” British Med- physical activity and sedentary behaviors among pregnant ical Journal, vol. 344, Article ID e2088, 2012. women: NHANES 1999–2006,” Preventive Medicine, vol. 50, no. [33] D. K. Tobias, C. Zhang, R. M. van Dam, K. Bowers, and F. B. 3, pp. 123–128, 2010. Hu, “Physical activity before and during pregnancy and risk of [18] T.R. Cohen, H. Plourde, and K. G. Koski, “Are Canadian women gestational diabetes mellitus: a meta-analysis,” Diabetes Care, achieving a �t pregnancy� A pilot study,” Canadian Journal of vol. 34, no. 1, pp. 223–229, 2011. Public Health, vol. 101, no. 1, pp. 87–91, 2010. [34] American College of Sports Medicine, “Impact of physical [19] M. A. Pereira, S. L. Rifas-Shiman, K. P. Kleinman, J. W. Rich- activity during pregnancy and postpartum on chronic disease Edwards, K. E. Peterson, and M. W. Gillman, “Predictors of risk: roundtable consensus statement,” Medicine and Science in change in physical activity during and aer pregnancy. Project Sports and Exercise, vol. 38, no. 5, pp. 989–1006, 2006. viva,” AmericanJournalofPreventiveMedicine, vol. 32, no. 4, pp. [35] G. J. Welk, Physical Activity Assessments for Health-Related 312–319, 2007. Research, Human Kinetics Books, Champaign, Ill, USA, 2002. Journal of Pregnancy 9

[36] K. R. Evenson and F. Wen, “Prevalence and correlates of objectively measured physical activity and sedentary behavior among US pregnant women,” Preventive Medicine, vol. 53, no. 1-2, pp. 39–43, 2011. Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 193170, 7 pages http://dx.doi.org/10.1155/2013/193170

Research Article Stages of Change Model for Participation in Physical Activity during Pregnancy

Lene Annette Hagen Haakstad,1 Nanna Voldner,2 and Kari Bø1

1 Department of Sports Medicine, Norwegian School of Sports Sciences, P.O. Box 4014, Ullevål Stadion, 0806 Oslo, Norway 2 Department of Obstetrics and Gynecology, Oslo University Hospital, Rikshospitalet, 0424 Oslo, Norway

Correspondence should be addressed to Lene Annette Hagen Haakstad; [email protected]

Received 26 September 2012; Revised 25 November 2012; Accepted 18 December 2012

Academic Editor: Michelle F. Mottola

Copyright © 2013 Lene Annette Hagen Haakstad et al. is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. e transtheoretical model (TTM) has been successful in promoting health behavioral change in the general population. However, there is a scant knowledge about physical activity in relation to the TTM during pregnancy. Hence, the aims of the present study were (1) to assess readiness to become or stay physically active according to the TTM and (2) to compare background and health variables across the TTM. Methods. Healthy pregnant women ( )wereallocatedtothestudyfrom �slo �niversity Hospital, Norway. e participants �lled in a validated self-administered questionnaire, physical activity pregnancy questionnaire (PAPQ) in gestation, weeks 32–36. e questionnaire contained 53 questions𝑛𝑛 with one푛 푛푛푛 particular question addressing the TTM and the �ve stages: (1) precontemplation stage, (2) contemplation stage, (3) preparation stage, (4) action stage, and (5) maintenance stage. Results. More than half of the participants (53 ) were involved in regular exercise (stages 4-5); however, only six speci�ed that they had recently started an exercise program (stage 4). About 33 reported engaging in some physical activity, but not regularly (stage 3). e results showed that receiving advice% from health professionals to exercise during pregnancy increased the likeliness of being in stages 4-5, while higher age, multiparity, pregravid overweight,% unhealthy eating habits, pelvic girdle pain, and urinary incontinence were more prevalent with low readiness to change exercise habits (stages 1–3). Conclusion. According to the TTM, more than half of the participants reported to be physically active. Moreover, most of the participants classi�ed as inactive showed a high motivational readiness or intention to increase their physical activity level. Hence, pregnancy may be a window of opportunity for the establishment of long-term physical activity habits.

1. Introduction 30 min of daily moderate intensity physical activity [18, 19]. However, studies have shown that most pregnant women do To date, intervention studies show that exercise during preg- not exercise on a regular basis [20, 21], and that only 5–20 nancy may enhance quality of life and wellbeing, improve follow current exercise guidelines [22, 23]. Hence, pregnant self-image and �tness, prevent excessive maternal weight gain, low back pain, pelvic girdle pain, and urinary incon- women may have a great potential to increase physical activity% tinence, as well as decrease the risk of depression during and reduce the risk of inactivity related complications and pregnancy and postpartum [1–9]. Some observational studies illnesses. have also reported associations between regular exercise dur- Antenatal care is part of public health promotion and ing pregnancy and gestational diabetes, preeclampsia, shorter prevention programs in most western countries, with preg- labor, fewer birth complications, and caesarean sections [10– nant women advised to attend between 5–8 visits through- 17]. out pregnancy [24]. Backe [25]foundthattheNorwegian Present recommendations for exercise during pregnancy antenatal healthcare system reaches almost 100 of pregnant suggest that, in the absence of medical and obstetric com- women. Consequently, health care providers are in a unique plications, pregnant women should aim to perform at least position to inform and encourage pregnant women% to start 2 Journal of Pregnancy or continue speci�c and general exercise programs. Such level (commuting activities, occupational activities, house- programs may also help to establish long-term PA habits. work, and family care activities), sedentary activities, recre- In several settings, the transtheoretical model (TTM) of ational exercise, exercise motivation/barriers, and social sup- change has been successful in promoting behavioral change port. More details of the questionnaire have been described [26, 27]. According to this model, a speci�c health behavior elsewhere [30, 31]. develops over time and progresses through �ve stages which e stages of change towards physical activity were may be used to examine readiness to become and stay assessed by a particular question aimed to classify the physically active (1) precontemplation, (2) contemplation, participants to one of �ve categories adapted from �odin (3) preparation, (4) action, and (5) maintenance [26, 27]. and Shephard [32] and further developed by Prochaska et al. However,thereisascantknowledgeaboutthestagesof [33]. Table 1 shows the TTM scoring system, questionnaire change towards physical activity among pregnant women, categories, and motivational readiness to modify behavior. and search on PubMed revealed no studies on this topic. e participants were asked to pick the response category Hence, the speci�c aims of the present study were to that most accurately described their current physical activity (1) assess perceptions regarding readiness to become or behavior or their interest for physical activity. Due to low stay physically active using the TTM and (2) to compare response rate in the precontemplation stage and the action background and health variables across the �ve stages of the stage, and in agreement with a previous Norwegian study, TTM among Norwegian pregnant women. the �ve stages were merged into two new variables in some supplementary statistical analyses [34]. Hence, participants 2. Materials and Methods in the precontemplation, contemplation, and preparation (stages 1–3) were classi�ed as physically inactive (insu�- 2.1. Study Design and Participants. is study was part of ciently PA), and participants in action and maintenance a larger prospective study of determinants of macrosomic (stage 4-5) were classi�ed as physically active (currently PA). infants in Norway (STORK). Results from the main study Concurrent validity for TTM has been demonstrated with have been published previously [16, 28]. Data collection a signi�cant association with the seven-day physical activity to answer the present research questions was conducted recall questionnaire [35], and the kappa index of test-retest through a self-administered questionnaire, (PAPQ) [29]. intrareliability over a two-week period was 0.78 [35].Forthe Healthy pregnant women were allocated to the study from purpose of the present study, we used a translated Norwegian the application form for birth at Oslo University Hospital version, previously used in a study to assess motivational readiness to stay or increase physical activity level [34]. between 2002 and 2005. Inclusion criteria were enrolment to the project before week 14–16 of gestation, having a singleton Maternal prepregnant weight was self-reported. Maternal fetus, ability to answer PAPQ in gestation week 32–36, and weight gain was calculated as the diﬀerence between self- being of Scandinavian origin. Exclusion criterion was preges- reported prepregnancy weight and the weight measured at the last clinical visit prior to delivery (pregnancy week 40.2, tational diabetes or other serious diseases due to the primary SD 1.3). e responsible midwife used a digital beam scale aim of the main study. Of the 2145 women invited to par- to measure the participant�s body weight (kg). Classi�ca- ticipate in STORK, 678 accepted the invitation. However, 90 tion of maternal weight gain and prepregnancy body mass withdrew before inclusion. Fourteen women were excluded index (BMI, kg/m ) was according to recommendations aer routine ultrasound at gestation week 17-18, due to from the IOM: 12.7–18.2 kg weight gain for underweight congenital disorders ( ) and twin births ( ). Further 2 women (prepregnancy BMI 18.5), 11.4–15.9 kg weight exclusions were two stillbirths, eleven relocations, and births gain for normal weight women (prepreg BMI of 18.5–24.9), at another hospital, and𝑛𝑛푛푛 eight participants chose𝑛𝑛푛푛 to withdraw. 6.8–11.4 kg weight gain for< overweight women (prepreg Consequently, 553 women were invited to participate in the BMI of 25.0–29.9), and 5.0–9.1 kg weight gain for obese present study. Of these, 467 (84.4 ) �lled in the PAPQ at women (prepreg BMI 30) [36]. In the present study, 15 home and returned the surveys at the last consultation with women had a prepregnancy BMI 18.5, and 33 women the midwife (NV), at mean pregnancy% week 36.4 (SD = had a prepregnancy BMI≥ 30. ese women were classi�ed 1.7). Not all the participants answered every question, and as either normal weight or overweight,< and corresponding therefore individual questions had varying response rates. weight gain recommendations≥ were used in the statistical e STORK project followed the Helsinki declaration, analysis. It was presumed that more women in the precon- and all the women gave written informed consent to par- templation, contemplation, and preparatory groups would ticipate. e Regional Committee for Medical and Health have less favorable weight gain compared to the action and Research Ethics, Southern Norway, Oslo, and the Norwegian maintenance groups. Social Sciences Data Services approved the project. 2.3. Statistical Analysis. All statistical analyses were con- 2.2. Assessment Procedures and Outcome Measures. e ducted with SPSS statistical soware version 18.0 for win- PAPQ is a validated twelve-page questionnaire speci�cally dows. Background variables are presented as frequencies, designed to assess physical activity behavior in pregnant percentages, or means with standard deviations (SDs). e women [29] and includes information on background vari- relationship between the TTM model and the selected variables, health status and complaints, total physical activity ables, including health variables were assessed by one-way Journal of Pregnancy 3

T 1: e readiness for physical activity stages of change scale (TTM).

Stage Questionnaire response categories Motivational readiness for change (1) I am currently not physically active and do not intend to engage in physical Pre-Contemplation (PC) activity in the next six months (Physically inactive, no intentions to change) (2) I am currently not physically active, but I am thinking about getting more Contemplation (C) physically active in the next six months (Physically inactive, intentions to change) Preparation (P) (3) I currently do some physical activity, but not regularly (Physically active, but not regularly) (4) I am currently physically active, but have only begun doing so within the last Action (A) six months (Regularly active, but only recently) Maintenance (M) (5) Iamcurrentlyphysicallyactiveandhavedonesoformorethansixmonths (Regularly active)

Anova, independent t-tests, or as appropriate. In addition, T 2: Proportions of women classi�ed across the stages of change we compared the TTM with present2 weight gain recom- (TTM) based on self-report ( ). mendationsbytheIOM[36]. e Spearman correlation 𝜒𝜒 Motivational readiness for change n coefficient was used to evaluate self-reported physical activity 𝑛𝑛 � 푛푛� levels in the 3rd trimester (de�ned as vigorous leisure time Pre-Contemplation (stage 1) 6 1.3 Preaction % physical activity 20 minutes once a week) and the TTM. (Inactive) Contemplation (stage 2) 55 11.8 �evel of statistical signi�cance was set at . Preparation (stage 3) 152 32.5 ≥ Postaction Action (stage 4) 6 1.3 3. Results 𝑃𝑃 푃 푃푃푃푃 (Active) Maintenance (stage 5) 241 51.6 Mean age of the participants was 31.6 years (range 20–49), Missing 7 1.5 mean prepregnancy BMI 23.6 (SD 3.7), and mean parity 1.3 (SD 0.5). e women were generally well educated, and 83 had education from college or university 4years.estudy stages of change merged into two groups; insufficiently PA or group did not differ from nonparticipants giving birth% at currently PA. is did not change the overall results. the same hospital, in mean maternal age,≥ parity, gestational e majority of the participants (65.3 ) had gained age at delivery, educational level, or marital status. Further weight above the present recommendations from IOM. information about background variables of the cohort has Comparing the participants and merging the% �ve stages into been presented elsewhere [30]. two groups, we found that the difference between women in e distribution of participants within each stage of stages 1–3 (insufficiently PA) was not different in excessive change is summarized in Table 2. According to the TTM, a weight gain compared to Current PA (stages 4-5) (72 versus large proportion of the participants reported to be somewhat 67 , ). Mean maternal weight gain was nearly or currently physically active, with 86.7 categorized in similar across the �ve groups, regardless of what% stage of stages 3–5. Most participants were in maintenance (stage 5), physical% 𝑃𝑃 activity 푃푃푃�푃� the women were located in ( ). followed by preparation (stage 3). Six women% speci�ed that they had recently started an exercise program (stage 4). 4. Discussion 𝑃𝑃 푃 푃푃푃푃푃 As shown in Table 3, women de�ned in stages 1–3 (insufficiently PA) were somewhat older, multiparous, and Asfaraswehaveascertained,thisisthe�rststudyto pregravid overweight (BMI 25) and reported to be suffering examine pregnant women’s motivation for physical activity from pelvic girdle pain and urinary incontinence. In total, according to the TTM. In addition, several demographic 29.2 of the women de�ned≥ their eating habits and nutri- and health indicators among women in the different stages tional status as unhealthy. A signi�cantly higher proportion of physical activity were compared. More than half of the of these% (64.7 versus 35.3 , ) were categorized participants were in stages 4-5, categorized as regularly in stages 1–3 (insufficiently PA) as compared to stages 4-5, active according to the TTM (Table 1). More than 32 respectively. In% contrast, more% 𝑃𝑃 women (푃 푃푃푃푃 ) of those were in stage 3, preparation. Only, 1.3 of the participants having received advice to exercise during pregnancy were in reported to be in stage 1 and had no intention to modify% the higher stages in the TTM ( ). No𝑛𝑛 � differences푛푛푛 were their physical activity behavior. Women% being older, having found when comparing the stages of exercise with education, children, suffering from pelvic girdle pain, reporting urinary being sick listed in 3rd trimester,𝑃𝑃 or daily 푃 푃푃푃� smoking. incontinence, unhealthy eating habits, pregravid BMI 25, Since several cells in the crosstabs had expected count less and not receiving advice from health care providers on how than 5, caused by the low percentage of participants located to perform PA during pregnancy were more oen located≥ in in the precontemplation stage and action stage (Tables 3 and stages 1–3 (insufficiently PA) versus 4-5 (currently PA). Our 4), we also performed additional tests analyzed with the �ve study suggests that pregnancy may be a good time to guide 4 Journal of Pregnancy

T 3: Comparison of background and health variables between the �ve stages of change (TTM). Results are presented as means with standard deviation (SD), in addition to number and percentages. With the exception of age and prepreg BMI, all variables are yes/no responses. Missing data are reported for each outcome as there are diﬀerent response rates for several variables.

Stages of exercise PC C P A M Missing P value Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Age 36.6 (4.2) 30.9 (4.1) 31.4 (4.0) 29.3 (4.1) 31.8 (3.9) 8 0.009 College/university education 4 (80 ) 43 (78.2 ) 129 (84.9 ) 6 (100 ) 198 (82.8 )9 0.2 Multiparous 5 (100 ) 33 (60 ) 74 (48.7 ) 1 (16.7 ) 97 (40.2 ) 7 0.005 Pelvic girdle pain 3 (60%) 40 (72.7%) 72 (48 %) 3 (50 %) 142 (59.2%) 10 0.037 Urinary incontinence 4 (80 %) 19 (34.5% ) 44 (28.9%) 0% 49 (20.3%) 7 0.004 Sick-listed 3rd trimester 2 (40%) 0 % 59 (39.9% ) 0% 78 (32.9 %) 20 0.5 Daily smokers 3rd trimester 0% 1 (1.8 %) 4 (2.6 %) 0 7 (2.9 %)7 1.0 Prepreg BMI 24.4 (3.6)% 22.9 (3.7) 24.7 (4.0)% 23.7 (4.6) 23.0 (3.2)% 9 0.001 Prepreg BMI 25 0 12 (21.8% ) 64 (42.1% ) 1 (16.7 ) 52 (21.6% ) 7 0.001 “I consider my eating habits unhealthy” 1 (20 ) 24 (45.3 ) 61 (40.4 ) 1 (16.7 ) 46 (19.2 ) 11 0.000 “I have received≥ advice about PA in present pregnancy” 3 (75 ) 14 (25.9%) 40 (26.8%) 1 (16.7%) 109 (46.0% ) 16 0.001 PC: precontemplation, C: contemplation, P: preparation, A: action, and% M: maintenance.% % % % ∗ % % % % %

T 4: e relationship between maternal weight gain parameters and the TTM. Results are presented as means with standard deviation (SD), in addition to number and percentages.

Stage of exercise Maternal weight gain parameters Missing P value PC C P A M Weight gain (kg) 12.4 (7.4) 14.0 (4.2) 14.2 (5.7) 11.8 (2.7) 13.8 (5.2) 29 0.558 Exceeding IOM recommendations 2009 3 (60 ) 38 (70.4 ) 103 (72.5 ) 4 (66.7 ) 152 (67.9 ) 29 0.634 PC: precontemplation, C: contemplation, P: preparation, A: action, and M: maintenance. ∗ % % % % % and encourage women to be more physically active, as most well as was available to answer any questions at the time when women categorized as “insuﬃciently PA” reported a high the participants handed in the questionnaire. motivational intention to become or increase their physical An obvious limitation of cross-sectional surveys is that activity level. the design precludes the establishment of causation between e strength of the present study is the high response variables and that most data are self-reported. Also, it is rate among the women receiving the PAPQ questionnaire. only a snapshot of the situation and may be biased by In addition, the population in STORK was similar in marital socially desirable responses. From 2002 to 2005, a total of status, educational level, mean maternal age, parity, gesta- 2145 women were randomly invited to participate in the tional age at delivery, and the baby’s birth weight as compared STORK project. Unfortunately, due to logistic limitations, to nonparticipants giving birth at Oslo University Hospital. not all eligible women were approached, and unfortunately When compared to the general Scandinavian pregnant pop- only about one fourth of the women accepted the invitation. ulation giving birth at Ullevål University Hospital, another Hence, although the response rate of eligible women to our major hospital in Oslo, the STORK participants included study may be considered high, the representativeness of the more nonsmokers, but were otherwise similar. Hence, the STORK study can be questioned. e TTM was originally survey participants in the present study may be considered to developed to be used in promoting or stopping a certain be fairly representative for an urban Norwegian population behavior [32]. In the present study, the model was used of Scandinavian origin [16, 28]. We used a validated form as a measure of pregnant women’s readiness to become or of the TTM, and the association between self-reported stay physically active, as it has been used in other study physical activity levels and the stages of change found in the populations [34]. present study, may provide some evidence for the concurrent Several studies have shown a decline in physical activity validity of the measure. e PAPQ included a broad range levelbeforeandthroughoutthecourseofpregnancy,andthat of determinants, ranging from demographic characteristics only 5–20 follow the present exercise guidelines [18, 21– to lifestyle habits (such as smoking and diet), pregnancy 23, 30, 37]. Hence, this is in contrast to the participants of the complaints, and social support, including the physician’s role present study% and how they perceive their physical activity to in�uence on physical activity level during pregnancy. In level according to the TTM. We found that more than 50 addition, the same midwife (NV) completed all weighing of of the pregnant women reported that they were currently the participants and calculated total maternal weight gain, as regularly active. % Journal of Pregnancy 5

e high amount of exercisers in the present study may mean maternal weight gain or weight gain above the IOM ref- be due to the main objective of the primary study, including erences. is may be because maternal weight gain has been evaluation of nutritional intake and physical activity on fetal found to be independent of exercise [36].Wehavenotbeen macrosomia. Hence, the women who chose to participate able to �nd other studies examining the same relationship. may have had more interest in general health compared en again, data from published studies yield con�icting to nonparticipants. In addition, most of the participants results regarding the impact of physical activity to control reported a high educational level. Statistics Norway’s survey maternal weight gain [45, 46], and it has been suggested that of the living conditions in 2008, found that those with a high participating in a physical activity may improve pregnancy level of education were more physically active than those with outcomes independent of weight changes, and that even a a low level of education. However, in the present study, no small increase in activity level has a positive effect on mother signi�cant association ( ) was found between the and fetus [47]. participants’ level of education and stages of exercise during Among the participants in the present study, about 1/4 pregnancy. ese results𝑃𝑃 differ from푃 푃푃푃푃 other studies, �nding that educational attainment is a strong determinant of stage reported to have a problem with urinary incontinence, and for physical activity [38, 39]. the majority of those women did not exercise regularly according to the TTM. is condition has been found to be a In our study, signi�cantly more women receiving advice from health care providers on physical activity during preg- barrier to participation in physical activity due to embarrass- nancy reported to be in the higher stages of the TTM. Hence, ment, feeling of discomfort, and possible increased leakage our �nding highlights the importance of precise and updated [48]. It is strong evidence that pelvic �oor muscle training information, based on the current ACOG guidelines, to be can prevent and treat urinary continence, and this should distributed by health care professionals to their pregnant be taken into account when designing exercise program clients. Only 37 of the participant in the present study for women in all age groups, with extra emphasis during reported to have received advice about exercise. Considering pregnancy [49]. that most pregnant% women visit their health care provider on a regular basis, this may be a window of opportunity 5. Conclusion for providing information of the bene�ts of regular exercise during pregnancy. Hence, midwives and physicians should be Receiving advice from health professionals to exercise during encouraged to promote physical activity in pregnancy. is is pregnancy increased the likeliness of being in the action supported by several studies reporting that pregnant women and maintenance stages. Higher age, multiparty, pregravid tend to follow the advice of health care providers regarding overweight, pelvic girdle pain, and urinary incontinence maternal weight gain [40, 41]. were more prevalent with lower readiness to change exercise Previous studies have found that being sedentary before habits. ere is a need for more research to evaluate whether the onset of pregnancy is a risk factor not to start exercising a TTM-based intervention is useful in promoting physical when pregnant [30, 38, 42]. Our results support these �nd- activity during pregnancy. ings, con�rming that women who are accustomed to exercising prior to pregnancy are more likely to maintain this habit Acknowledgments and that those not physically active prepregnancy do not start during pregnancy. Hence, to achieve higher rates of exercise e authors thank the project leaders of STORK: Professors during pregnancy, health promotion programs should target Tore Henriksen and Jens Bollerslev, Oslo University Hospital. the general female population in their childbearing years. ere is no con�ict of interests or �nancial disclosure. According to the review of Gaston and Cramp [38], being nulliparous has been a consistent predictor of regular exer- References cise. In this study, more than 60 of the women categorized in the currently PA group (stages 4-5) were nulliparous. [1] A. V. M. Arizabaleta, L. O. Buitrago, A. C. A. de Plata, M. M. Escudero, and R. Ramírez-Vélez, “Aerobic exercise during erefore, to increase the level of% exercise among multiparous women, activities allowing for personal time management pregnancy improves health-related quality of life: a randomised trial,” Journal of Physiotherapy, vol. 56, no. 4, pp. 253–258, 2010. and �exibility in terms of place and type of activity should be stimulated. Moreover, initiating supervised group activities [2] I. A. Santos, R. Stein, S. C. Fuchs et al., “Aerobic exercise and social support in a safe setting with quali�ed instructors and submaximal functional capacity in overweight pregnant women: a randomized trial,” Obstetrics and Gynecology, vol. may aid compliance to an exercise program. 106, no. 2, pp. 243–249, 2005. An increasing proportion of women are overweight or [3] A. B. Granath, M. S. E. Hellgren, and R. K. Gunnarsson, obese at the start of their pregnancy [43],anditisassumed “Water aerobics reduces sick leave due to low back pain during that this group is less likely to adapt and maintain the pre g n an c y,” Journal of Obstetric, Gynecologic, and Neonatal recommended levels of physical activity [44].Wefoundthat Nursing, vol. 35, no. 4, pp. 465–471, 2006. fewer women with a high prepregnancy BMI ( 25) reported [4] S. Mørkved, K. Å. Salvesen, B. Schei, S. Lydersen, and K. Bø, to be regularly active during pregnancy. “Does group training during pregnancy prevent lumbopelvic No group differences were found between women≥ report- pain? A randomized clinical trial,” Acta Obstetricia et Gyneco- ing to be insufficiently PA or currently PA with respect to logica Scandinavica, vol. 86, no. 3, pp. 276–282, 2007. 6 Journal of Pregnancy

[5] S. Mørkved, K. Bø, B. Schei, and K. �. Salvesen, “Pelvic �oor [20] J. Zhang and D. A. Savitz, “Exercise during pregnancy among muscle training during pregnancy to prevent urinary inconti- US women,” Annals of Epidemiology, vol. 6, no. 1, pp. 53–59, nence: a single-blind randomized controlled trial,” Obstetrics 1996. and Gynecology, vol. 101, no. 2, pp. 313–319, 2003. [21] K. R. Evenson, D. A. Savitz, and S. L. Huston, “Leisure-time [6] A. F. Robledo-Colonia, Y. F. Mosquera-Valderrama, C. Escob- physical activity among pregnant women in the US,” Paediatric ar-Hurtado, and R. Ramírez-Vélez, “Aerobic exercise training and Perinatal Epidemiology, vol. 18, no. 6, pp. 400–407, 2004. during pregnancy reduces depressive symptoms in nulliparous [22] K. M. Owe, W. Nystad, and K. Bø, “Correlates of regular women: a randomised trial,” Journal of Physiotherapy, vol. 58, exercise during pregnancy: the Norwegian Mother and Child no. 1, pp. 9–15, 2012. C o h or t Stu d y,” Scandinavian Journal of Medicine and Science in [7] K. M. Songøygard, S. N. Stafne, K. A. I. Evensen, K. A. Salvesen, Sports, vol. 19, no. 5, pp. 637–645, 2009. T. Vik, and S. Mørkved, “Does exercise during pregnancy [23] A. M. Petersen, T. L. Leet, and R. C. Brownson, “Correlates of prevent postnatal depression?: a randomized controlled trial,” physical activity among pregnant women in the United States,” Acta Obstetricia et Gynecologica Scandinavica, vol. 91, no. 1, pp. Medicine and Science in Sports and Exercise, vol. 37, no. 10, pp. 62–67, 2012. 1748–1753, 2005. [8] E. Norman, M. Sherburn, R. H. Osborne, and M. P. Galea, “An [24] J. Villar, G. Carroli, D. Khan-Neelofur, G. Piaggio, and M. exercise and education program improves weil-being of new Gülmezoglu, “Patterns of routine antenatal care for low-risk mothers: a randomized controlled trial,” Physical erapy, vol. pre g n an c y,” Cochrane Database of Systematic Reviews,no.4, 90, no. 3, pp. 348–355, 2010. article CD000934, 2001. [9] L. A. H. Haakstad and K. Bø, “Eﬀect of regular exercise on [25] B. Backe, “Overutilization of antenatal care in Norway,” Scan- prevention of excessive weight gain in pregnancy: a randomised dinavianJournalofPublicHealth, vol. 29, no. 2, pp. 129–132, controlled trial,” European Journal of Contraception and Repro- 2001. ductive Health Care, vol. 16, no. 2, pp. 116–125, 2011. [26] J. O. Prochaska, C. C. DiClemente, and J. C. Norcross, “In search [10] J. C. Dempsey, T. K. Sorensen, M. A. Williams et al., “Prospec- of how people change: applications to addictive behaviors,” tive study of gestational diabetes mellitus risk in relation American Psychologist, vol. 47, no. 9, pp. 1102–1114, 1992. to maternal recreational physical activity before and during [27] B. H. Marcus and L. R. Simkin, “e transtheoretical model: pre g n an c y,” American Journal of Epidemiology, vol. 159, no. 7, applications to exercise behavior,” Medicine and Science in pp. 663–670, 2004. Sports and Exercise, vol. 26, no. 11, pp. 1400–1404, 1994. [11] J. C. Dempsey, C. L. Butler, T. K. Sorensen et al., “A case- [28] N. Voldner, K. F. Frøslie, K. Bø et al., “Modi�able determinants control study of maternal recreational physical activity and risk of fetal macrosomia: role of lifestyle-related factors,” Acta of gestational diabetes mellitus,” Diabetes Research and Clinical Obstetricia et Gynecologica Scandinavica,vol.87,no.4,pp. Practice, vol. 66, no. 2, pp. 203–215, 2004. 423–429, 2008. [12] C. B. Rudra, M. A. Williams, I. M. Lee, R. S. Miller, T. K. Soren- sen, and C. L. Butler, “Perceived exertion during prepregnancy [29] L. A. H. Haakstad, I. Gundersen, and K. Bø, “Self-reporting physical activity and preeclampsia risk,” Medicine and Science in compared to motion monitor in the measurement of physical Sports and Exercise, vol. 37, no. 11, pp. 1836–1841, 2005. activity during pregnancy,” Acta Obstetricia et Gynecologica Scandinavica, vol. 89, no. 6, pp. 749–756, 2010. [13] T. K. Sorensen, M. A. Williams, I. M. Lee, E. E. Dashow, M. L. ompson, and D. A. Luthy, “Recreational physical activity [30] L. A. H. Haakstad, N. Voldner, T. Henriksen, and K. Bø, “Phys- during pregnancy and risk of preeclampsia,” Hypertension, vol. ical activity level and weight gain in a cohort of pregnant Nor- 41, no. 6, pp. 1273–1280, 2003. wegian women,” Acta Obstetricia et Gynecologica Scandinavica, vol. 86, no. 5, pp. 559–564, 2007. [14] J. F. Clapp, “e course of labor aer endurance exercise during pre g n an c y,” American Journal of Obstetrics and Gynecology, vol. [31] L. A. H. Haakstad, N. H. T. Voldner, and K. Bø, “Why do preg- 163, no. 6 I, pp. 1799–1805, 1990. nant women stop exercising,” Acta Obstetricia et Gynecologica Scandinavica. In press. [15] K. R. Kardel, B. Johansen, N. Voldner, P. O. Iversen, and T. Henriksen, “Association between aerobic �tness in late [32] G. Godin and R. J. Shephard, “A simple method to assess exer- pregnancy and duration of labor in nulliparous women,” Acta cise behavior in the community,” Canadian Journal of Applied Obstetricia et Gynecologica Scandinavica, vol. 88, no. 8, pp. Sport Sciences, vol. 10, no. 3, pp. 141–146, 1985. 948–952, 2009. [33] J. O. Prochaska, W. F. Velicer, J. S. Rossi et al., “Stages of [16] N. Voldner, K. F. Frøslie, L. A. H. Haakstad, K. Bø, and change and decisional balance for 12 problem behaviors,” T. Henriksen, “Birth complications, overweight, and physical Health Psychology, vol. 13, no. 1, pp. 39–46, 1994. inactivity,” Acta Obstetricia et Gynecologica Scandinavica, vol. [34] A. K. Jenum, C. Lorentzen, S. A. Anderssen et al., “Promoting 88, no. 5, pp. 550–555, 2009. physical activity in a multi-ethnic district—methods and base- [17] T. J. Bungum, D. L. Peaslee, A. W. Jackson, and M. A. Perez, line results of a pseudo-experimental intervention study,” Euro- “Exercise during pregnancy and type of delivery in nulliparae,” pean Journal of Cardiovascular Prevention and Rehabilitation, Journal of Obstetric, Gynecologic, and Neonatal Nursing, vol. 29, vol. 10, no. 5, pp. 387–396, 2003. no. 3, pp. 258–264, 2000. [35] B. H. Marcus, V. C. Selby, R. S. Niaura, and J. S. Rossi, “Self- [18] ACOG. Committee Opinion, “Exercise during pregnancy and eﬃcacy and the stages of exercise behavior change,” Research the postpartum period,” Obstetrics and Gynecology, vol. 99, no. Quarterly for Exercise and Sport, vol. 63, no. 1, pp. 60–66, 1992. 1, pp. 171–173, 2002. [36] Institute of Medicine, “Weight gain during pregnancy: reex- [19] R. Artal and M. O’Toole, “Guidelines of the American College of amining the guidelines,” 2009, http://www.iom.edu/pregnancy Obstetricians and Gynecologists for exercise during pregnancy weightgain. and the postpartum period,” British Journal of Sports Medicine, [37] Y. Ning, M. A. Williams, J. C. Dempsey, T. K. Sorensen, I. O. vol. 37, no. 1, pp. 6–12, 2003. Frederick, and D. A. Luthy, “Correlates of recreational physical Journal of Pregnancy 7

activity in early pregnancy,” Journal of Maternal-Fetal and Neonatal Medicine, vol. 13, no. 6, pp. 385–393, 2003. [38] A. Gaston and A. Cramp, “Exercise during pregnancy: a review of patterns and determinants,” Journal of Science and Medicine in Sport, vol. 14, no. 4, pp. 299–305, 2011. [39] C. E. Garber, J. E. Allsworth, B. H. Marcus, J. Hesser, and K. L. Lapane,“Correlatesofthestagesofchangeforphysicalactivity in a population survey,” American Journal of Public Health, vol. 98, no. 5, pp. 897–904, 2008. [40] M. E. Cogswell, G. S. Perry, L. A. Schieve, and W. H. Dietz, “Obesity in women of childbearing age: risks, prevention, and treatment,” Primary Care Update for Ob/Gyns,vol.8,no.3,pp. 89–105, 2001. [41] C. M. Olson, “Achieving a healthy weight gain during preg- n an c y,” Annual Review of Nutrition, vol. 28, pp. 411–423, 2008. [42] K. Gjestland, K. Bø, K. M. Owe, and M. Eberhard-Gran, “Do pregnant women follow exercise guidelines? Prevalence data among 3482 women, and prediction of low-back pain, pelvic girdle pain and depression,” British Journal of Sports Medicine. In press. [43] N. Heslehurst, L. J. Ells, H. Simpson, A. Batterham, J. Wilkinson, and C. D. Summerbell, “Trends in maternal obesity incidence rates, demographic predictors, and health inequalities in 36 821 women over a 15-year period,” British Journal of Obstetrics and Gynaecology, vol. 114, no. 2, pp. 187–194, 2007. [44] M. E. Leijon, P. Bendtsen, A. Ståhle, K. Ekberg, K. Festin, and P. Nilsen, “Factors associated with patients self-reported adherence to prescribed physical activity in routine primary health care,” BMC Family Practice, vol. 11, article 38, 2010. [45] L. Poston and L. C. Chappell, “How should women be advised on weight management in pregnancy?” British Medical Journal, vol. 344, p. e2774, 2012. [46] A. K. Ronnberg and K. Nilsson, “Interventions during pregnancy to reduce excessive gestational weight gain: a systematic review assessing current clinical evidence using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) system,” British Journal of Obstetrics and Gynaecol- ogy, vol. 117, no. 11, pp. 1327–1334, 2010. [47] T. L. Weissgerber, L. A. Wolfe, G. A. L. Davies, and M. F. Mottola, “Exercise in the prevention and treatment of maternal- fetal disease: a review of the literature,” Applied Physiology, Nutrition and Metabolism, vol. 31, no. 6, pp. 661–674, 2006. [48] K. Bø, “�rinary incontinence, pelvic �oor dysfunction, exercise and sport,” Sports Medicine, vol. 34, no. 7, pp. 451–464, 2004. [49] J. Hay-Smith, S. Mørkved, K. A. Fairbrother, and G. P.Herbison, “Pelvic �oor muscle training for prevention and treatment of urinary and faecal incontinence in antenatal and postnatal women,” Cochrane Database of Systematic Reviews, no. 4, article CD007471, 2008. Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 718095, 9 pages http://dx.doi.org/10.1155/2013/718095

Research Article Kinematic Analysis of Gait in the Second and Third Trimesters of Pregnancy

Marco Branco,1,2 Rita Santos-Rocha,1,2 Liliana Aguiar,1,3 Filomena Vieira,1,3 and António Veloso1,3

1 CIPER, Faculdade de Motricidade Humana, Estrada da Costa, Dafundo, 1495-688 Cruz Quebrada, Portugal 2 ESDRM-IPS, Escola Superior de Desporto de Rio Maior, Avenida Dr. Mario´ Soares, 2040-413 Rio Maior, Portugal 3 FMH-UTL, Faculdade de Motricidade Humana, Estrada da Costa, Dafundo, 1495-688 Cruz Quebrada, Portugal

Correspondence should be addressed to Rita Santos-Rocha; [email protected]

Received 11 October 2012; Revised 6 December 2012; Accepted 28 December 2012

Academic Editor: Michelle F. Mottola

Copyright © 2013 Marco Branco et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The kinematic analysis of gait during pregnancy provides more information about the anatomical changes and contributes to exercise and rehabilitation prescription. The purposes were to quantify the lower limb kinematics of gait and to compare it between the second and third trimesters of pregnancy and with a control group. A three-dimensional analysis was performed in twenty- two pregnant women and twelve nonpregnant. Repeated Measures and Manova tests were performed for comparisons between trimesters and between pregnant and controls. The walking speed, stride width, right-/left-step time, cycle time and time of support, and flight phases remain unchanged between trimesters and between pregnant and controls. Stride and right-/left-step lengths decreased between trimesters. Double limb support time increased between trimesters, and it increased when compared with controls. Joint kinematics showed a significant decrease of right-hip extension and adduction during stance phase between trimesters and when compared with controls. Also, an increase in left-knee flexion and a decrease in right-ankle plantarflexion were found between trimesters. The results suggested that pregnant women need to maintain greater stability of body and to become more efficient in locomotion. Further data from the beginning of pregnancy anthropometric data may contribute to the analysis.

1. Introduction 15 women. They reported that overall gait kinematics were unchanged during pregnancy. However, significant increases The third trimester of pregnancy is characterized by a rapid in hip and ankle kinetics were found. Their findings indicate growth in size and weight of the fetus, so that an additional that during pregnancy there may be an increased demand 50% increase in fetal weight is observed in this trimester placed on hip abductor, hip extensor, and ankle plantar flexor [1].Thisinturncausesanincreaseinabdominalweight muscles during walking. Lymbery and Gilleard [4]investi- and volume in pregnant woman, which is associated to an gated the temporospatial and ground reaction forces (GRF) increase in the weight of growing breasts and an increase in variables in the stance phase of walking during late pregnancy lumbar lordosis, resulting in a superior and posterior shifts of of13womenat38weeks’gestationand8weeksafterbirth. the woman’s center of gravity [2]. These changes, occurring in They concluded that in late pregnancy, there was a wider the body of the pregnant woman, lead to many complains of step width, and mediolateral GRF tended to be increased in a discomfort and pain in lower limbs. Few studies describe the medial direction. They suggested that women may adapt their kinematic motion on the lower limb of the pregnant women, gait to maximize stability in the stance phase of walking and particularly in a longitudinal perspective. to control mediolateral motion. Huang et al. [5]comparedthe In previous studies, Foti et al. [3] performed a three- natural pattern of walking of 10 nulligravidae and 10 pregnant dimensional (3D) analysis of gait during the second half of women, divided into three groups, respectively, at 12 weeks, the last trimester of pregnancy and one year postpartum on at13–28weeks,at29–40weeksofgestationalageandtested 2 Journal of Pregnancy only one time. They reported significant differences between the pregnant and nonpregnant women, especially in knee abduction angle, knee and hip internal rotation angles. Also, as gestational age increases, the experimental group increased hip extension moment, decreased knee extension moment, increased knee adduction moment, and decreased ankle plantar flexion moment, and these changes were related with sacroiliac pain. The authors suggested that the hip is the main work-loading area. Little published data in this field analyzed what changes occur in each trimester or associated the kinematic and kinetic variables for each phase of the gait cycle.Otherreasontostudythegaitaspregnancyprogresses is to associate the gait variables with the increased prevalence of back and foot pain and other clinical complications. Also, the analysis of both sides of the body might be interesting in order to understand potential imbalances. Thepurposeofthisstudywastodescribespatialand temporal parameters and quantify the kinematic variables on the structures of the lower limb during gait and compare it Figure 1: Spherical reflective markers placed with double-sided adhesive tape on the skin, in both sides of the lower body of a between the later stages of second and third trimesters of pregnant women. pregnancy and with a control group of nonpregnant women.

2. Materials and Methods measurement protocol [6] by ISAK certified anthropome- 2.1. Participants. Twenty-two pregnant women, between the trists, with exception of the abdominal girth [7]. ages of 27 and 38 years and with no history of foot, ankle, Inordertocollectstaticanddynamicdatatrials,spherical knee, musculoskeletal, and neuromuscular trauma or dis- reflective markers were placed with double-sided adhesive ease, participated in this study. Pregnant participants were tape on the skin, in both sides of the lower body (Figure 1). recruited via direct contact and flyers placed in gym and Markers setup is in agreement with the suggestion of health centers and have volunteered to participate in the Cappozzo et al. [8], for lower limb segments, and CODA study. Twelve healthy nulligravidae women participated in (Charnwood Dynamics Ltd, Leicestershire, United Kingdom) the study as controls. None of the participants had contraindi- protocols for model of pelvis segment. Thereby, in the static cation to physical exercise. All subjects gave written informed trial,forfootmodel,markerswereplacedonthefifthmeta- consent prior to participation in the study. tarsal head, first metatarsal head, posterior proximal top The pregnant participants presented the following char- of calcaneus, posterior distal top of calcaneus, lateral top acteristics: mean (± sd) age of 32.5 ± 2.6 years (range: 27.0– of calcaneus. For shank model construction, markers were 38.0); height of 1.62 ± 0.06 m(range:1.50–1.76);numberof placed on the lateral malleolus, medial malleolus, lateral gestational weeks of 27.0 ± 1.3 weeks (range: 25.0–29.1) in femoral epicondyle, medial femoral epicondyle, and a cluster the second trimester (2T); mass of 67.1 ± 6.9 Kg (range: 55.5– with 3 markers in the lateral of shank. For the thigh model 2 85.0) in the 2T; body mass index (BMI) of 25.6 ± 2.9 Kg/m construction, markers were placed on the lateral femoral (range: 21.4–33.2) in the 2T; number of gestational weeks of epicondyle, medial femoral epicondyle, and a cluster with 3 36.3±1.0 weeks (range: 34.6–38.4) in the third trimester (3T); markers in the lateral of thigh. To define the CODA pelvis mass of 71.4 ± 6.7 Kg (range: 59.0–87.0) in the 3T; BMI of model, markers were placed in both anterior superior iliac 2 27.3 ± 2.8 Kg/m (range: 22.8–34.0) in the 3T. spine and posterior superior iliac spine. For the dynamic The nonpregnant group presented the following charac- trials, at least 3 markers were left in each segment, as reference teristics: mean (± sd) age of 20.58 ± 1.73 years (range: 18.0– to static markers setup. Planar motion of the hip, knee, and 23.0); height of 1.64±0.07 m (range: 1.54–1.73); mass of 58.33± anklejointwascalculatedwithVisual3Dsoftware(C-Motion 2 8.71 Kg (range: 45.0–73.5); BMI of 21.5 ± 2.4 Kg/m (range: Inc., Germantown, USA) by a computational procedure 18.1–25.7). implementing the dot product between the skeletal segments articulated by these joints. 2.2. Procedures prior to Motor Task. The study was approved by the ethical committee of the faculty, and data were 2.3. Motor Task. The motor task was to walk barefoot a collected at the Laboratory of Biomechanics and Functional distance of 10 meters between two points, in a straight line Morphology, in two times: during the later stages of the at a natural and comfortable speed, as suggested in previous second trimester (2T) and third trimester (3T). studies [5, 9], for 3 minutes, with a time break of 1 minute Before performing the motor task, anthropometric data between each trial. The floor had no specific patterns or was measured, according to the International Society for irregularities, and the participants had no knowledge of the Advancement of Kinanthropometry (ISAK) standardized the location of force platforms. Participants were allowed Journal of Pregnancy 3 to get familiar with the laboratory system, and no fatigue (v) time of support and flight phases in both lower limbs; occurrence was reported. (vi) stride width; (vii) stride length; 2.4. Kinematic Data Collection. Kinematic data were collected through ten infrared high-speed cameras (Oqus-300, (viii) right- and left-step length; Qualisys, Sweden) at a rate of 200 Hz and two Kistler force (ix) joint angles in the sagittal plane of the hip, knee, and platforms (Kistler AG, Winterthur, Switzerland) of 0.60 m × ankle for right and left lower limbs; 0.40 m (length, width), at a rate of 1000 Hz. The capture (x) in the hip joint two peaks were considered in the hardware was connected to Qualisys USB Analog Acquisition sagittal plane: the first peak represents the maximum interface in order to synchronize kinetic and kinematic hip extension which occurs in the toe off event; the data with software Qualisys Track Manager (QTM; Qualisys second peak occurs some instants before heel strike AB, Gothenburg, Sweden). Data sequences, of cameras and and represents the maximum hip flexion; force platforms, were recorded in the same file. System wascalibratedbywandtype,withanexactwandlengthof (xi) in the knee joint four peaks were considered in the 751.4 mm moved randomly across the recorded field, before sagittal plane:thefirstpeakoccursaftertheheel each participant data collection. Calibration was accepted strike and represents a slightly flexion to absorb the if the standard deviation of the wand’s length measures contact with the floor; the second peak represents was below 0.5 mm. Cameras were positioned statically to the slight knee extension near the late mid stance minimize light reflection artifacts and to allow recording phase; the third peak occurs in the mid swing phase of at least two consecutive walking cycles, defined as the as the maximum knee flexion; and the fourth peak, time between two consecutive initial ground contacts of the represents the maximum extension of the knee and heel strike for each side. The last cycles performed by each occurs instantly before heel strike; participant were considered for the analysis. Digital images (xii) in the ankle joint four peaks were considered in the (of the markers) were collected at same time as the GRF. sagittal plane: the first peak occurs immediately after heel strike with a sudden decrease of dorsiflexion 2.5. Kinematic Data Analysis. A three-dimensional (3D) ofthefoot;thesecondpeakoccursapproximately analysis was performed including both sides of the body and at contralateral heel strike; the third peak indicates also in the transverse plane. Gait events and walking cycles the maximum plantarflexion at toe off event with were manually defined based on the vertical trajectory of a decrease of plantarflexion; the fourth peak occurs the proximal end of the foot segment and on the vertical inmidswingphaseandrepresentsthemaximum GRF curve. Collected data were interpolated using a Cubic dorsiflexion of the foot in preparation for contact with Spline Interpolation as suggested by Robertson et al. [10], for the ground; a maximum of 10 frames gap. The trajectory of the reflective (xiii) joint angles in the frontal plane of the hip and ankle markers and the kinetic data were filtered with a Butterworth for right and left lower limbs; digital lowpass filter, at 10 Hz cutoff frequency, as suggested (xiv) in the hip joint two peaks were considered in the by Robertson and Dowling [11]. All data were normalized in frontal plane: the first peak occurs in the mid-stance time. phase and represents the maximum value of hip Considering the two trimesters in analysis, kinematic adduction; the second peak occurs after toe off event pattern curves (angular displacement in ankle, knee and hip and represents the maximum value of hip abduction; in degrees) were estimated relative to the walking stride cycle. The data curves and the peak angles values were estimated, (xv) in the ankle joint four peaks were considered in the for left and right side, with visual 3D. The mean and standard frontal plane:thefirstpeakoccursinmidstancein deviations were analyzed in IBM SPSS Statistics (version 20). the maximum ankle eversion phase; the second peak The range of motion of each joint was also analyzed in represents the maximum value of ankle inversion IBM SPSS Statistics (version 20). For kinematic and kinetic during toe off event; the third peak occurs in the mid parameters, initial foot contact was collected at the time swing phase close to the neutral position; the fourth corresponding to first contact of the foot on the floor. For peak occurs instantly at the end of the swing phase in kinematicparameters,theendofthestridecorresponded inversion; to the next contact with the same foot. Four strides of each (xvi) joint angles in the transverse plane of the hip and subject were considered in the two trimesters. ankle for right and left lower limbs; (xvii) in the hip joint two peaks were considered in the 2.6. Variables Analyzed. The independent variables were the transverse plane: the first peak represents the maxi- 2T and 3T. The following dependent variables were analyzed: mum value of internal rotation, and it occurs instantly (i) walking speed; before the toe off event; the second peak represents the maximum value of external rotation, and it occurs (ii) cycle time; in a late swing phase; (iii) right- and left-step time; (xviii) in the ankle joint two peaks were considered in the (iv) double limb support time; transverse plane: the first peak occurs at the beginning 4 Journal of Pregnancy

of the mid-stance phase, and the second peak occurs no influence between experimental variables, and the results in the mid swing phase; are dependent on the trimester which they relate. However, (xix) kinetic pattern of gait-ground reaction forces (GRF). between the group of nonpregnant and pregnant women in the second trimester, differences were observed in double The determination of angle peaks was performed accord- support time. The same analysis between the group of ing to Rose and Gamble [12]. However, few more peaks were nonpregnant and pregnant women in the third trimester also also included in the analysis: in the sagittal plane, two has shown that there was no influence between experimental more peaks in the knee were included, which represent the variables, and the results are dependent on the group they extension of the shank. In the frontal plane, four peaks were belong to. Significant differences were found in the stride calculated in the ankle joint. The first peak occurs between length, in the right- and left-step length and in double theheelstrikeandthecontralateralfoottoeoffandrepresents support time. the largest peak of the foot eversion. The second peak occurs immediately before the toe off of the first foot and represents thehighestpeakofthefootinversion.Thethirdpeakoccurs 3.2. Joint Kinematics. The joints range of motion was ana- during the swing phase of the first foot and represents the lyzed in all motion planes, and data are presented in Tables returning to the neutral position, and, finally, the fourth peak 1, 2,and3. occurs immediately to heel strike and represents an inversion peak. In the transverse plane, two peaks were calculated at the 3.3. Joint Kinematics: Sagittal Plane. The kinematic pattern ankle joint. Both peaks represent an abduction of the foot. of the gait in sagittal plane is represented in Figure 2.The The first peak occurs after the heel strike, and the second quantitative data are presented in Table 2. occurs during the swing phase. The Repeated Measures analysis has shown that the angular data are dependent on pregnancy trimester, and 2.7. Statistical Procedures. All statistical procedures were there was no angular dependence between angle peaks. The conducted using IBM SPSS Statistics (version 20) software first peak of the hip joint, presented in the sagittal plane, a for Windows. Shapiro-Wilk normality test was conducted significant decrease in its magnitude, keeping the thigh close and not assumed for all cases. The Mauchly’s test of spheric- to the neutral position at the end of the stance phase (𝐹= ity was performed before Repeated Measures analysis and 6.390; 𝑃 = 0.001;power=0.233).Thepeaksoftheknee was assumed. For pairwise analysis, the Repeated Measures joint remain with similar magnitude from the second to the analysis was performed between second and third trimesters. third trimester; however, the third peak performs a significant All the requirements for application of the Repeated Mea- increase of 1.2 degrees of knee flexion during the swing phase. sures and MANOVA analyses were calculated and assumed. The analysis of the angular displacement between second MANOVAwas applied between each of the trimesters and the and third trimesters showed that most of the peaks angles of group of nonpregnant, to verify what was the level of change the ankle remain unchanged. However, the third peak of the between nonpregnant and the pregnant participants. Bonfer- rightanklehasshownasignificantreductionofitsangular roni confidence interval adjustment was applied to allow an magnitude, signifying a decrease in plantar flexion performed adjustment to the confidence intervals and significance values in the third trimester, of about 1.4 degrees. The remaining formultiplecomparisons.AssuggestedbyVincent[13], for all peaksdidnotshowsignificantchangesbetweentrimesters. cases, the level of statistical significance was set at 𝑃 ≤ 0.05. In multivariate analysis between the group of nonpreg- nantandpregnantwomeninthesecondtrimester,therewere 3. Results significant differences only in the first peak of the hip joint (𝐹 = 18.697; 𝑃 = 0.000; power = 0.369). Between the group 3.1. Spatiotemporal Parameters. Spatiotemporal data are pre- of nonpregnant and those in the third trimester, in the sagittal sented in Table 1. After performing the Repeated Measures plane, also significant differences in the first peak of the hip analysis between the second and third trimesters, it was found joint (𝐹 = 36.922; 𝑃 = 0.000; power = 0.536) were found. that the results are influenced by the effect of the trimester to which they relate. However, most of the spatiotemporal parameters remain unchanged between trimesters. Thereby 3.4. Joint Kinematics: Frontal Plane. The kinematic analysis no significant differences were found in walking speed, stride of the gait in frontal plane is represented in Figure 3.The width, right and left step time, cycle time, and in the time quantitative data are presented in Table 3. of support and flight phases in both lower limbs. Significant The Repeated Measures analysis of the angular displace- differences were observed in right- and left-step length, stride ment of the second to the third trimester of pregnancy, in the lengththatdecreasedfrom2Tto3T,andindoublelimb frontal plane, revealed that the results are dependent on the supporttimethatincreasedbetweentrimesters(𝐹 = 122.342, trimester to which it relates. However, in the hip joint there 𝑃 = 0.000; power = 0.853). Among these variables there was a significant change in the first peak, which represents was no difference between the left- and right-step length a decrease of magnitude of abduction of the thigh of about variables,pointingoutthatalthoughtherearedifferences 1.4 degrees from the second to the third trimesters. The same between trimesters these differences do not occur laterally. peak revealed differences between the group of nonpregnant The MANOVAanalysis between the second trimester and andthepregnantwomeninthesecondtrimester(𝐹 = 5.412; the group of nonpregnant women has shown that there was 𝑃 = 0.026; power = 0.145) and the women in late pregnancy Journal of Pregnancy 5

Sagittal Plane Right Le Pelvis posterior (+) anterior ( −) tilt 1

−9.5 2nd (degrees) 1st

−20 010050 Gait cycle (%)

Hip ﬂexion (+) extension (−) Hip ﬂexion (+) extension (−) 50 2nd 50 2nd

17.5 17.5

∗ (degrees) 1st (degrees) 1st

−15 −15 010050 010050 Gait cycle (%) Gait cycle (%)

Knee extension (+) ﬂexion ( −) Knee extension (+) ﬂexion ( −)

0 4th 0 4th

2nd 2nd −35 1st −35 1s t

(degrees) (degrees) ∗ 3rd 3rd −70 −70 010050 010050 Gait cycle (%) Gait cycle (%)

Ankle D. ﬂexion (+) extension (−) Ankle D. ﬂexion (+) extension (−) 25 25

2nd 4th 2nd 4th 0 1s t 0 1s t

∗ 3rd 3rd

(degrees) (degrees)

−25 −25 010050 010050 Gait cycle (%) Gait cycle (%) Figure 2: Kinematic parameters (sagittal plane) of gait of pregnant women in the later stages of second trimester (dashed line) and third trimester (solid line) and of nulliparous controls (dot line). Mean joint angles of the pelvis, hip, knee, and ankle, for right and left lower limbs, ∗ in degrees. The curve peaks are indicated by numbers: first, second, third and fourth and( ) points the significant differences.

(𝐹 = 12.876; 𝑃 = 0.001; power = 0.287). In the ankle there third trimesters and also no changes between pregnant and were no significant changes in angular peaks. control group.

3.5. Joint Kinematics: Transverse Plane. The kinematic anal- 4. Discussion ysisofthegaitintransverseplaneisrepresentedinFigure 4. The quantitative data are presented in Table 4. The knowledge of the kinematic parameters associated to gait In the transverse plane, the ankle, knee, and hip joints and other motor tasks performed by the pregnant woman, have shown no significant changes between the second and during the three trimesters of pregnancy and postpartum, 6 Journal of Pregnancy

Table 1: Spatiotemporal parameters of gait during the later stages of the second and third trimesters of pregnancy (𝑁=22) and nulliparous controls (𝑁=22). Units (mean ± sd) of mass (Kg), velocity (m/s), length (m), and time (s).

Second trimester (2T) Third trimester (3T) Nonpregnant (NP) Mass (Kg) 67.082 ± 6.946 71.368 ± 6.652 ↑ 58.333 ± 8.711 Velocity (m/s) 1.159 ± 0.125 1.127 ± 0.128 1.243 ± 0.089 Stride width (m) 0.096 ± 0.025 0.101 ± 0.027 0.078 ± 0.024 3 Stride length (m) 1.260 ± 0.098 1.234 ± 0.088 ↓ 1.316 ± 0.099 ↑ 3 Left-step length (m) 0.630 ± 0.051 0.616 ± 0.044 ↓ 0.657 ± 0.053 ↑ 3 Right-step length (m) 0.630 ± 0.049 0.618 ± 0.045 ↓ 0.659 ± 0.049 ↑ Cycletime(s) 1.081± 0.054 1.091 ± 0.063 1.048 ± 0.040 Left-step time (s) 0.541 ± 0.030 0.545 ± 0.032 0.523 ± 0.020 Right-step time (s) 0.540 ± 0.026 0.545 ± 0.033 0.525 ± 0.021 Left-stance time (s) 0.640 ± 0.040 0.655 ± 0.046 0.616 ± 0.026 Left-swing time (s) 0.439 ± 0.018 0.435 ± 0.021 0.433 ± 0.018 Right-stance time (s) 0.645 ± 0.042 0.656 ± 0.042 0.616 ± 0.029 Right-swing time (s) 0.438 ± 0.019 0.436 ± 0.025 0.433 ± 0.020 2,3 Double limb support time (s) 0.208 ± 0.029 0.219 ± 0.030 ↑ 0.183 ± 0.018 ↓ Bold: significant differences with 𝑃 < 0.05. 3 ↑significance only with third trimester. 2,3 ↓ significance with second and third trimesters.

Table 2: Joint kinematic peak values (mean) of gait in the sagittal plane, during the second and third trimesters of pregnancy (𝑁=22)and nulliparouscontrols(𝑁=12). Units (mean ± sd) are in degrees.

Joint Side Peak Second trimester Third trimester Nonpregnant 1st −14.874 ± 3.653 −15.195 ± 4.141 −10.355 ± 2.879 Pelvis 2nd −18.036 ± 3.566 −18.706 ± 3.523 −7.127 ± 2.515 1st −2.301 ± 5.282 −0.295 ± 4.507 −10.072 ± 4.437 Right 2nd 41.510 ± 4.413 43.168 ± 3.897 34.030 ± 2.998 Hip 1st −2.903 ± 6.347 −1.963 ± 5.150 −10.710 ± 4.388 Left 2nd 41.853 ± 3.793 43.628 ± 3.488 34.498 ± 2.786 1st −16.427 ± 6.689 −18.175 ± 7.152 −18.862 ± 5.770 2nd −2.620 ± 5.161 −4.464 ± 6.521 −5.861 ± 4.825 Right 3rd −60.516 ± 14.282 −61.864 ± 14.219 −62.815 ± 5.575 4th −1.035 ± 5.535 −3.350 ± 7.188 −5.879 ± 5.262 Knee 1st −16.402 ± 6.651 −17.665 ± 7.478 −19.248 ± 6.194 2nd −2.444 ± 7.224 −3.091 ± 7.028 −5.297 ± 4.281 Left 3rd −63.781 ± 3.828 −65.074 ± 3.441 −63.992 ± 2.731 4th −3.047 ± 5.706 −4.058 ± 6.609 −6.737 ± 3.793 1st −4.972 ± 2.653 −4.921 ± 5.064 −1.508 ± 1.850 2nd 12.062 ± 3.782 12.729 ± 5.017 15.308 ± 2.571 Right 3rd −17.960 ± 4.849 −16.536 ± 4.416 −15.093 ± 5.363 4th 6.737 ± 1.818 6.375 ± 3.692 8.952 ± 2.739 Ankle 1st −4.132 ± 4.560 −4.668 ± 3.645 −0.024 ± 3.228 2nd 12.731 ± 3.899 12.659 ± 3.466 14.982 ± 2.575 Left 3rd −16.066 ± 5.241 −15.946 ± 4.432 −14.921 ± 6.725 4th 6.850 ± 3.128 6.568 ± 2.007 8.673 ± 2.700 Bold: significant differences with 𝑃 < 0.05. Journal of Pregnancy 7

Frontal plane

Right Le

Hip adduction (+) abduction (− ) Hip adduction (+) abduction (− ) 18 18 2nd ∗ 0 1st 0 2nd

(degrees) 1st

(degrees)

−18 −18 010050 010050 Gait cycle (%) Gait cycle (%)

Ankle inversion (+) eversion (− ) Ankle inversion (+) eversion (−) 15 5 3rd 1st 5 4th −5 2nd 3rd

(degrees) (degrees) 4th 1st 2nd −5 −15 010050 010050 Gait cycle (%) Gait cycle (%)

Figure 3: Kinematic parameters (frontal plane) of gait of pregnant women in the later stages of second trimester (dashed line) and third trimester (solid line) and of nulliparous controls (dot line). Mean joint angles of the hip and ankle, for right and left lower limbs, in degrees. ∗ The curve peaks are indicated by numbers: first, second, third and fourth and( ) points the significant differences.

Sagittal plane

Right Le

Hip internal rotation (+) external rotation ( −) Hip external rotation (+) internal rotation ( − ) 15 20

1st −2.5 2.5 2nd

(degrees) (degrees) 1st 2nd −20 −15 0 50 100 010050 Gait cycle (%) Gait cycle (%)

Ankle FF adduction (+) abduction (− ) Ankle FF adduction (+) abduction (− ) 1 25 2nd 12 −12 1st 1st

(degrees) (degrees) 2nd

−25 −1 0 50 100 010050 Gait cycle (%) Gait cycle (%)

Figure 4: Kinematic parameters (transverse plane) of gait of pregnant women in the later stages of second trimester (dashed line) and third trimester (solid line) and of nulliparous controls (dot line). Mean joint angles of the hip and ankle, for right and left lower limbs, in degrees. ∗ The curve peaks are indicated by numbers: first and second and( ) points the significant differences. 8 Journal of Pregnancy

Table 3: Joint kinematic peak values (mean) of gait in the frontal plane, during the second and third trimesters of pregnancy (𝑁=22)and nulliparouscontrols(𝑁=12). Units (mean ± sd) are in degrees.

Joint Side Peak Second trimester Third trimester Nonpregnant 1st 11.681 ± 3.705 10.282 ± 3.346 14.805 ± 3.808 Right 2nd −7.774 ± 4.454 −8.946 ± 3.966 −9.142 ± 2.998 Hip 1st 9.553 ± 4.031 11.086 ± 2.687 14.403 ± 3.183 Left 2nd −9.821 ± 4.195 −7.920 ± 3.534 −9.768 ± 5.746 1st −5.117 ± 3.646 −5.479 ± 6.513 −3.825 ± 2.401 2nd 8.220 ± 5.194 8.020 ± 7.902 11.465 ± 4.102 Right 3rd −0.177 ± 4.074 −0.532 ± 6.053 0.661 ± 3.896 4th 4.455 ± 4.198 3.927 ± 4.973 5.089 ± 2.982 Ankle 1st −4.662 ± 2.317 −5.253 ± 2.202 −4.367 ± 2.040 2nd 9.271 ± 5.613 8.808 ± 4.445 11.214 ± 3.089 Left 3rd 0.298 ± 3.894 0.276 ± 3.205 1.350 ± 3.711 4th 4.594 ± 2.869 4.183 ± 3.597 6.525 ± 3.369 Bold: significant differences with 𝑃 < 0.05.

Table 4: Joint kinematic peak values (mean) of gait in the transverse plane, during the second and third trimesters of pregnancy (𝑁=22) and nulliparous controls (𝑁=12). Units (mean ± sd) are in degrees.

Joint Side Peak Second trimester Third trimester Nonpregnant 1st −17.839 ± 8.948 −19.745 ± 7.116 −18.654 ± 5.926 Right 2nd −20.365 ± 10.228 −21.871 ± 7.276 −20.277 ± 6.903 Ankle 1st −17.611 ± 6.492 −19.253 ± 9.174 −22.699 ± 5.876 Left 2nd −20.951 ± 6.242 −21.670 ± 7.943 −21.050 ± 7.410 1st 8.932 ± 7.699 8.083 ± 6.416 11.250 ± 5.580 Right 2nd −10.377 ± 7.711 −9.104 ± 6.871 −8.269 ± 5.824 Hip 1st 7.908 ± 7.948 10.132 ± 6.534 12.355 ± 8.342 Left 2nd −10.099 ± 7.887 −7.751 ± 5.256 −5.350 ± 5.965

provides more information about the effect of pregnancy in to protect themselves from falling and possibly injuring the a range of performance conditions. This kind of information fetus. will be helpful for prescribing exercise programs and rehabil- In the analysis of joint kinematics, the range of motion in itation programs and preventing musculoskeletal injuries. thetransverseplaneoftherighthipwastheonlyparameterto The unchanged results found in most temporal param- experience significant changes, with a reduction in its ampli- eters of walking in the third trimester are similar to results tude, possibly due to increased volume in the abdominal found by Foti et al. [3] and Lymbery and Gilleard [4]. Our region or to the lateral dominance; however, this data were results showed an increased time of double support between not collected. the group of nonpregnant and pregnant women and between The analysis of angular peaks revealed that most of the second and third trimesters. Similar results were also the peaks remain unchanged during pregnancy. However, found by Foti et al. [3]. Furthermore, in spatial parameters, between trimesters significant differences were found in the it was found a significant decrease in the length of right and extension and abduction peaks of the right thigh, in the left step and therefore the size of the gait stride, from the maximum flexion peak of the left knee, and in the plantarflex- nonpregnant group and the second trimester of pregnancy. ion peak of the right ankle. When those peaks were com- Both spatial and temporal parameters corroborate that, while pared between nonpregnant group and the groups in both walking at a self-selected pace, the pregnant woman needs trimesters, there was a significant reduction in the extension to promote stability of the body. The observed decrease in and abduction of the right thigh. These results highlight stride length, while the double limb support time increased, that the hip joint, possibly because it is near the pelvic between the second and third trimesters, might be related region, carries more angular adjustments, especially during to the fact that pregnant women experience an altered eye the stance phase. contact with the floor due to abdomen volume. Also, if the Considering the variables analyzed, in the majority, the pregnant women became heavier by the end of their preg- samebehaviorwasobservedinrightandleftlowerlimbs. nancy,theyaresupposedtobemorecarefulwhenwalking Differences between right and left sides of the body were not Journal of Pregnancy 9 expected. However, few differences that were found between [4]J.K.LymberyandW.Gilleard,“Thestancephaseofwalking both sides and the potential imbalances related to these cases, during late pregnancy: temporospatial and ground reaction need further analysis. force variables,” Journal of the American Podiatric Medical Association,vol.95,no.3,pp.247–253,2005. [5]T.H.Huang,S.C.Lin,C.S.Ho,C.Y.Yu,andY.L.Chou,“The 5. Conclusion gait analysis of pregnant women,” Biomedical Engineering— In conclusion, considering all planes of motion we find that Applications, Basis & Communications,vol.14,no.2,pp.67–70, 2002. most of the studied parameters remain unchanged between the second and third trimester of pregnancy. However, [6]M.Marfell-Jones,T.Olds,A.Stewart,andJ.E.L.Carter,Inter- national Standards for Anthropometric Assessment,ISAK,2006. parameters related to the stance, and corresponding time, suggested that participants need to maintain greater stability [7] T. G. Lohman, A. F. Roche, and R. Martorell, Anthropometric Standardization Reference Manual, Human Kinetics, Cham- of body. Nevertheless, it may induce discomfort and pain in paign, Ill, USA, 1988. the lower limbs often reported by pregnant women. These changesalsomaypromotethepregnantwomentobecome [8]A.Cappozzo,A.Cappello,U.D.Croce,andF.Pensalfini, “Surface-marker cluster design criteria for 3-d bone movement more efficient in locomotion. Much of the differences found reconstruction,” IEEE Transactions on Biomedical Engineering, during pregnancy are dependent to which trimester they vol.44,no.12,pp.1165–1174,1997. belong, and we believe that these changes may happen [9] J. M. Falola, P. Gouthon, F. E. Koussihoued´ e,´ B. Agossa, and J. from the beginning of pregnancy to the end of it with Brisswalter, “Gait coordination in pregnancy: a study in a rural greater magnitude, because when compared with the group of population in Africa,” Science and Sports,vol.24,no.1,pp.49– nonpregnant, greater magnitude of differences were verified. 51, 2009. However,furtherdatafromthebeginningofpregnancyare [10]D.G.E.Robertson,G.E.Caldwell,J.Hamill,G.Kamen,and required. The inclusion of anthropometric data may also S. N. Whittlesey, Research Methods in Biomechanics,Human contribute to the analysis of its influence on biomechanical Kinetics,Champaign,Ill,USA,2004. parameters. The literature primarily analyzes the changes [11] D. G. E. Robertson and J. J. Dowling, “Design and responses between the end of pregnancy to postpartum; however, it of Butterworth and critically damped digital filters,” Journal of may be wise to assume that pregnancy induces changes Electromyography and Kinesiology,vol.13,no.6,pp.569–573, that remain in the postpartum period, in a way justifying 2003. that much of the studied parameters remain unchanged [12] J. Rose and J. G. Gamble, Human Walking, Williams & Wilkins, as was reported by Foti et al. [3]. Further data from the Philadelphia, Pa, USA, 3rd edition, 2006. beginning of pregnancy are needed, and also the inclusion [13] W. J. Vincent, Statistics in Kinesiology,HumanKineticsBooks, of anthropometric data may also contribute to the analysis Champaign, Ill, USA, 3rd edition, 2005. of gait during pregnancy and its influence on biomechanical parameters.

Conflict of Interests The authors declare no commercial relationships or conflict of interests.

Acknowledgments Theauthorswishtothankallparticipantsinthestudyand Filomena Carnide (Ph. Degree), researcher of CIPER, for statistical guidance. This study was supported by FCT— Fundac¸ao˜ para a cienciaˆ e a Tecnologia/Portuguese Foun- dation for Science and Technology (http://alfa.fct.mctes.pt/), Project no. PTDC/DES/117031/2010, principal researcher: R. Santos-Rocha.

References

[1] ACOG, Your Pregnancy and Childbirth: Month to Month,Amer- ican College of Obstetricians and Gynecologists, Washington, DC, USA, 5th edition, 2010. [2]T.W.WangandB.S.Apgar,“Exerciseduringpregnancy,” American Family Physician,vol.57,no.8,pp.1846–1852,1998. [3] T. Foti, J. R. Davids, and A. Bagley, “Abiomechanical analysis of gait during pregnancy,” Journal of Bone and Joint Surgery A,vol. 82,no.5,pp.625–632,2000. Hindawi Publishing Corporation Journal of Pregnancy Volume 2013, Article ID 780180, 6 pages http://dx.doi.org/10.1155/2013/780180

Research Article Prepregnancy Physical Activity in relation to Offspring Birth Weight: A Prospective Population-Based Study in Norway—The HUNT Study

Silje Krogsgaard, Sigridur L. Gudmundsdottir, and Tom I. L. Nilsen

Department of Human Movement Science, Norwegian University of Science and Technology, 7491 Trondheim, Norway

Correspondence should be addressed to Tom I. L. Nilsen; [email protected]

Received 12 October 2012; Revised 7 December 2012; Accepted 10 January 2013

Academic Editor: Michelle F. Mottola

Copyright © 2013 Silje Krogsgaard et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. The objective was to examine the association between prepregnancy physical exercise and offspring birth weightand to assess the combined association of pre-pregnancy body mass index (BMI) and physical exercise on birth weight. Methods.The study included 2,026 women aged 20–39 years participating in the Norwegian HUNT study and linked with the Medical Birth Registry. We calculated mean differences in birth weight and odds ratios (ORs) for a macrosomic infant (i.e., birth weight > 4000 g) using linear and logistic regression analysis. Results. There was no clear association between leisure time physical exercise andmean birth weight. Women who reported no exercise had reduced risk of a macrosomic infant (OR, 0.6; 95% confidence interval (CI), 2 0.4–0.9) compared to women with a high exercise level. Overweight (BMI ≥ 25.0 kg/m ) was associated with an OR of 1.9 (95% CI, 1.2–2.9) for a macrosomic infant among women who reported low exercise levels, whereas the OR was 1.2 (95% CI, 0.8–1.8) among women with higher exercise levels. Conclusion. There was some evidence that women who reported no exercise before pregnancy had lower risk for a macrosomic infant than women who exercised. Pre-pregnancy BMI was positively associated with birth weight and risk of macrosomia but only among the least active women.

1. Introduction pregnancy and birth weight [19–22]. Although women who exercise regularly before pregnancy are more likely to con- The proportion of women giving birth to large infants has tinue to exercise during pregnancy [23–25], few studies have increasedaroundtheworld[1, 2], most likely because of examined the associations between prepregnancy physical the rising rates of maternal overweight and obesity [3–7]. activity and birthweight, and the results have been inconsis- Whereas consequences of low birth weight may include tent [18, 20, 26, 27]. infant mortality and morbidity [8], high birth weight has been In this prospective study in Norway, we have utilized data related to increased risk for caesarean section, chorioam- on women participating in a population-based health study nionitis, fourth degree perinatal lacerations, postpartum linked with information from the Medical Birth Registry haemorrhage, shoulder dystocia [9–11], and low Apgar score to study the association between maternal pre-pregnancy [12]. Additionally, high birth weight has been positively physical exercise and offspring birth weight. Additionally, we associated with obesity [13]andtype2diabetes[14]in explored the combined association of pre-pregnancy body adulthood. mass index and physical exercise on birth weight. Previous studies have reported that physical activity in pregnancy is related to foetal growth rate and birth weight 2. Materials and Methods [15, 16], and that physically active women have a reduced risk of delivering a large infant [17, 18], possibly by increased 2.1. Study Population. The Nord-Trøndelag health study (The insulin sensitivity [6]. However, not all studies have reported HUNT study) is a large population-based health study con- consistent inverse associations between physical activity in ducted in the county of Nord-Trøndelag, Norway. The HUNT 2 Journal of Pregnancy

Table 1: Characteristics of the study population (N = 2,026) according to total leisure time physical exercise.

Total physical exercise level a Characteristic No activity Low Medium High No participants (% of total) 147 (7.3) 647 (31.9) 641 (31.6) 591 (29.2) Mean age at baseline, y 26.8 27.2 26.7 26.6 2 Mean body mass index, kg/m 22.4 22.3 22.5 22.4 Parity (% primiparous) 20.4 28.2 28.5 40.6 Smoking (% current smoking) 53.1 44.0 38.8 31.3 Education (% college/university) 8.1 16.1 20.6 25.4 Alcohol (% not drinking last 2 weeks) 49.7 45.7 46.0 43.1 Marital status (% married) 53.1 55.6 49.5 46.7 a Based on a summary score of frequency, duration, and intensity of exercise. study is a collaboration between HUNT Research Centre duration (<15,15–30,31–60,and>60 minutes; coded 1–4) and (Faculty of Medicine, Norwegian University of Science and intensity (light, moderate, and to exhaustion; coded 1–3) of Technology NTNU), Nord-Trøndelag County Council, Cen- the activity. Among participants who reported exercising at tral Norway Health Authority, and the Norwegian Institute of least once a week, a summary score of frequency, duration, Public Health. It constitutes three consecutive cross-sectional and intensity was calculated according to the following equa- waves; the first was conducted in 1984–1986 (HUNT 1), the tion: 1/5 ∗ frequency + 1/4 ∗ duration + 1/3 ∗ intensity. This second in 1995–1997 (HUNT 2), and the third in 2006– procedureintendedtogiveequalweighttoeachcomponent 2008 (HUNT 3). For the purpose of the present study, we of physical activity and resulted in a maximum score of 1.0 have used information from the first wave (HUNT 1). In foreachofthethreecomponentsofthesummaryscore.The HUNT 1, 87,285 persons aged ≥ 20 years were invited to median score value of 1.97 (range, 1.2–3.0) was then used participate, and 77,216 (88.5%) accepted the invitation, filled as a cut-off to classify women into two categories of score in questionnaires, and attended a clinical examination (37,826 values (±median). This information was used to construct a men and 39,390 women). A more detailed description of variable of total physical exercise with four unique categories: participation, method, and procedures of the HUNT study (1) no activity, (2) low activity (<1 session per week), (3) canbefoundelsewhere[28]. medium activity (44 weeks) delivery, 32 women with multiple births, 10 women with gestational diabetes, 118 women with preeclamp- 2.3. Statistical Analysis. We used linear regression to analyze sia, 831 women without information of physical activity, body the association between measures of leisure time physical mass index, or gestational age, and 490 women who gave birth exercise and mean birth weight. We also calculated odds within 10 months after participation (i.e., possibly pregnant at ratios (OR) for having a macrosomic infant (≥4,000 g) in the time of participation). This left 2,026 women available for different categories of leisure time physical exercise using statistical analysis. logistic regression. Precision of the estimated associations was assessed by a 95% confidence interval (CI). Women 2.2. Study Variables. Information on the offspring was who reported the highest activity level were used as the obtained by a linkage to the Medical Birth Registry of referencecategoryinallanalysis.Thefollowingvariableswere Norway. These data were obtained for the first child born considered as potential confounders in the analysis; age (20– during five years after participation in HUNT. The main 24, 25–29, 30–34, and 35–39 years), smoking (never, former, outcome variable was newborn birth weight, measured in and current), frequency of alcohol consumption during the grams (g), first analyzed as a continuous variable and then as past 2 weeks (none, 1–4 times, ≥5 times, abstainer, and a dichotomized variable using 4,000 g as cutoff. Macrosomia unknown), education (<10, 10–12, >12 years, and unknown), was defined as birth weight at or above 4,000 g29 [ ]. marital status (unmarried, married, and previously married), Leisure time physical exercise was assessed using three and parity (primiparous, 1-2 children, and 3–6 children). questions. In the first question, the participants were asked Covariates were removed from the model if there was no to report how many exercise sessions (e.g., walking, skiing, meaningful difference between adjusted and unadjusted esti- swimming, or other sports) they usually had during a week, mates. All estimates were adjusted for maternal age, smoking, with five response options (0, <1, 1, 2-3, and ≥4times; and parity. Tests for trend across categories of leisure time coded 1–5). If the participants reported exercising at least physical exercise were conducted by treating the categories once a week, they were also asked to report the average as an ordinal variable in the regression model. Journal of Pregnancy 3

Table 2: Maternal pre-pregnancy leisure time physical exercise and mean offspring birth weight from linear regression analyses.

Physical exercise No. of persons Mean birth weight (g) Crude difference Adjusteda difference (95% CI) P trendb Sessions per week None 147 3589.0 −25.2 −26.9 (−137.9to84.2) <1 647 3632.5 18.3 21.7 (−64.8 to 108.3) 1 591 3639.7 25.5 31.0 (−56.1 to 118.1) 2-3 491 3590.4 −23.7 −24.4 (−113.2 to 64.4) ≥4 150 3614.1 0.0 0.0 (reference) 0.49 Total exercise c No activity 147 3589.0 −38.9 −53.0 (−141.8 to 35.8) Low 647 3632.5 4.6 −3.5 (−58.3 to 51.4) Medium 641 3606.9 −21.0 −36.9 (−91.5 to 17.8) High 591 3627.9 0.0 0.0 (reference) 0.56 CI: confidence interval. aAdjusted for maternal age (20–24, 25–29, 30–34, and 35–39 years), smoking (never, former, current, and unknown), and parity (primiparous, 1-2 children, and 3–6 children). bP value from trend test when categories were entered as an ordinal variable in the regression model. cBased on a summary score of frequency, duration, and intensity of exercise.

Since maternal BMI could be both an effect modifier and Table 3: Odds ratio (OR) from logistic regression for giving birth to on the causal pathway between exercise and birth weight, a marcosomic infant (i.e., birth weight > 4000 g) in association with BMI was not included as a confounder in the primary maternal pre-pregnancy leisure time physical exercise. analyses. However, additional analysis was conducted for the No. No. of Crude Adjusteda OR combined associations of prepregnancy BMI and total leisure Physical exercise of P trendb persons OR (95% CI) time physical exercise in relation to birth weight, using linear cases and logistic regression as described earlier. We also included 2 Sessions per week aproducttermofBMI(<25 versus ≥25 kg/m ) and exercise level (no or low activity versus medium or high activity) to None 147 20 0.7 0.7 (0.4–1.3) assess possible interaction between the two variables and as <1 647 148 1.3 1.3 (0.8–2.1) well as stratified the analyses of physical exercise on the two 1 591 122 1.1 1.2 (0.7–1.8) BMI groups. 2-3 491 98 1.2 1.1 (0.7–1.8) All statistical analyses were performed using the statisti- ≥ cal software SPSS for Windows, version 17.0. 4 150 28 1.0 1.0 (reference) 0.89 c The study was approved by the Regional Committee for Total exercise Ethics in Medical Research. All eligible participants received No activity 147 20 0.6 0.6 (0.4–0.9) a written invitation with information about the study, and Low 647 148 1.2 1.1 (0.9–1.5) all participants gave their consent by filling in and returning Medium 641 128 1.0 0.9 (0.7–1.2) the first questionnaire that was mailed together with the invitation. High 591 120 1.0 1.0 (reference) 0.65 CI: confidence interval. aAdjusted for maternal age (20–24, 25–29, 30–34, and 35–39 years), smoking (never, former, current, and unknown), and parity (primiparous, 1-2 chil- 3. Results dren, and 3–6 children). bP value from trend test when categories were entered as an ordinal variable Descriptive characteristics of the study population are pre- in the regression model. sented in Table 1.Meanbaselinematernalageamongthe cBased on a summary score of frequency, duration, and intensity of exercise. 2,026 women in the study was 26.9 years, whereas mean birth weight of their offspring was 3,620 g (SD, 502), and a total of 416 (20.5%) newborns weighed 4,000 g or more (i.e., had physical exercise and risk of a macrosomic infant. However, macrosomia). women who reported being inactive before pregnancy had Table 2 presents results from linear regression show- alowerriskofgivingbirthtoaninfantwithexcessivebirth ing that there was no clear association between maternal weight (OR, 0.6; 95% CI, 0.4–0.9) compared to women with leisure time physical exercise and mean birth weight of a high total exercise level (Table 3). their offspring, neither in relation to number of exercise The combined association of pre-pregnancy BMI and sessions per week (𝑃 trend, 0.49) nor in relation to total leisure time physical exercise in relation to birth weight is amount of exercise (𝑃 trend, 0.56). Correspondingly, results showninTable4. Women who were overweight (BMI ≥ 2 from logistic regression presented in Table 3 provide no 25.0 kg/m ) before pregnancy and reported no or low leisure consistent evidence for an association between maternal time physical exercise gave birth to infants with significantly 4 Journal of Pregnancy

Table 4: Maternal pre-pregnancy body mass index (BMI) and leisure time physical exercise related to mean offspring birth weight from linear regression and odds ratio (OR) from logistic regression for a macrosomic infant (i.e., birth weight > 4000 g).

Combined categories of No. of Crude mean Adjustedb mean No. of cases Adjustedb OR Crude OR BMI and exercisea persons difference difference (95% CI) (BW > 4000 g) (95% CI) <25 kg/m2 and 1,046 0.0 0.0 (reference) 206 1.0 1.0 (reference) medium/high level <25 kg/m2 and no/low level 676 −8.6 −8.4 (−55.6 to 38.8) 131 1.0 1.0 (0.8–1.2) ≥25 kg/m2 and 186 29.8 33.1 (−42.5 to 108.7) 42 1.2 1.2 (0.8–1.8) medium/high level ≥25 kg/m2 and no/low level 118 129.7 134.1 (41.0 to 227.3) 37 1.9 1.9 (1.2–2.9) CI: confidence interval; BW: birth weight. aBased on a summary score of frequency, duration, and intensity of exercise. bAdjustedformaternalage(20–24,25–29,30–34,and35–39years),smoking(never,former,current,andunknown),andparity(primiparous,1-2children, and 3–6 children). higher mean birth weight (134 g; 95% CI, 41.0–227.3)and had women in other studies. Studies have shown that a sedentary ahigherriskforamacrosomicinfant(OR,1.9;95%CI,1.2– lifestyle in pregnancy is associated with lower birth weight 2 2.9), compared to women with BMI < 25.0 kg/m who had a [32] and an increased risk of a very low birth weight infant medium or high exercise level. In additional analysis stratified [16]. It has been observed that mothers of very low birth 2 by body mass index, overweight women (BMI ≥ 25.0 kg/m ) weight infants were less likely to be physically active during who reported no or low exercise levels had significantly their pregnancy [33]. It has been suggested that both excessive higher offspring birth weight (132 g; 95% CI, 20.4–243.7) and and insufficient physical activities in pregnancy are related higheroddsratioforamacrosomicinfant(OR,2.0;95%CI, to an inadequate fetal growth [34], although some women 1.1–3.5) than women who reported a medium or high exercise mightbeadvisedtobeinactiveandatresttoreducetherisk level (data not shown). However, there was no statistically of adverse pregnancy outcomes. significant interaction between BMI and total exercise level Unlike the present study, some previous studies have (𝑃 = 0.08). shown inverse associations between maternal pre-pregnancy exercise behaviors and offspring birth weight or risk of having 4. Discussion excess weight [17, 18], although the results are not entirely consistent [21]. A recent study of leisure time physical activity In this large prospective study of Norwegian women, we during pregnancy is more in agreement with the results of the foundnoclearassociationbetweenreportedleisuretime present study. Hegaard et al. [22] found no association with < physical exercise level before pregnancy and offspring birth mean birth weight or the risk of giving birth to low ( 2,500 g) ≥ weight. There was some evidence that inactive women had or high ( 4,500 g) birth weight infants. The inconsistent a slightly lower likelihood of giving birth to a child with results in these studies could be due to different measures excessive birth weight than more physically active women, of physical activity, in addition to the possibility for chance but the small numbers of inactive women call for a cautious findings in the smaller studies. interpretation. Analysis of the combined association of BMI There is growing evidence that overweight or obesity andexerciseonbirthweightshowedthatwomenwitha before pregnancy is a risk factor for macrosomia [3, 12, 35, 36]. 2 BMI ≥ 25.0 kg/m gave birth to infants with significantly The results from the present study suggest that the effects 2 higher birth weight than women with a BMI < 25.0 kg/m , of maternal pre-pregnancy overweight were associated with butonlyiftheyalsoreportednoorlowlevelsofphysical higher birth weight only among women who reported no or exercise. This could suggest that physical exercise to some low level of activity. This is contradictory to the findings by extent reduces the effect of maternal adiposity on offspring Lof¨ et al. [27] who showed that a high pre-pregnancy activity birth weight. level did not reduce the risk of high birth weight infants The suggestive evidence that women who were inactive in women who were overweight or gained much weight before pregnancy had lower risk for delivering a macrosomic in pregnancy. Nevertheless, physical activity may improve infant is contradictory to some previous studies. Voldner maternal weight control, both before and during pregnancy et al. [20] reported that inactive women (defined as <1h [37, 38]. per week) had almost a threefold higher odds ratio for fetal The strengths of the present study include the prospective macrosomia than physically active women (>1 h per week). design, the large sample size of women reporting physical However, another Norwegian study found no association activity before pregnancy, and the standardized measures of between frequency of regular exercise before pregnancy and size at birth obtained from the Medical Birth Registry of offspring with excessive birth weight (≥90th percentile) [18]. Norway. However, some of the categories of physical exercise In the present study, those who were classified as inactive (e.g., inactive) suffered from small samples size, and this reported never engaging in physical exercise, and these couldresultinchancefindings.Moreover,asinanyobser- women could be more extremely sedentary than inactive vational study, residual confounding due to unmeasured Journal of Pregnancy 5

(e.g., smoking during pregnancy and gestational weight [10]N.E.Stotland,A.B.Caughey,E.M.Breed,andG.J.Esco- gain) or unknown factors cannot be ruled out. Since the bar, “Risk factors and obstetric complications associated with information on physical activity was self-reported, it could be macrosomia,” International Journal of Gynecology and Obstet- subject to misclassification, although a validation study has rics,vol.87,no.3,pp.220–226,2004. shown acceptable agreement with objective measures [39]. [11] N. Voldner, K. F. Frøslie, L. A. H. Haakstad, K. Bø, and Althoughpre-pregnancyphysicalactivityhasbeenassociated T. Henriksen, “Birth complications, overweight, and physical with physical activity level during pregnancy [23–25], it has inactivity,” Acta Obstetricia et Gynecologica Scandinavica,vol. alsobeenshownthatexerciselevelsdeclineinpregnancy 88, no. 5, pp. 550–555, 2009. [24, 25, 40, 41]. Unfortunately, our data did not allow us to [12] M. C. Jolly, N. J. Sebire, J. P. Harris, L. Regan, and S. Robinson, examine if pre-pregnancy physical exercise was related to the “Risk factors for macrosomia and its clinical consequences: a activity level during pregnancy. study of 350,311 pregnancies,” European Journal of Obstetrics Gynecology and Reproductive Biology, vol. 111, no. 1, pp. 9–14, 2003. 5. Conclusions [13] W. McGuire, L. Dyson, and M. Renfrew, “Maternal obesity: consequences for children, challenges for clinicians and carers,” There was no clear association between maternal prepreg- Seminars in Fetal and Neonatal Medicine,vol.15,no.2,pp.108– nancy leisure time physical exercise and offspring birth 112, 2010. weight. However, the result may indicate that inactive women [14] T. Harder, E. Rodekamp, K. Schellong, J. W. Dudenhausen, and gave birth to infants with lower birth weight and had A. Plagemann, “Birth weight and subsequent risk of type 2 lower risk for delivering an infant with macrosomia. The diabetes: a meta-analysis,” American Journal of Epidemiology, results also show that high maternal pre-pregnancy BMI was vol. 165, no. 8, pp. 849–857, 2007. associated with higher mean birth weight and increased risk [15] J. F. Clapp III, H. Kim, B. Burciu, and B. Lopez, “Beginning of macrosomia in offspring of physically inactive women, regular exercise in early pregnancy: effect on fetoplacental whereas pre-pregnancy BMI was not associated with birth growth,” American Journal of Obstetrics and Gynecology,vol. weight and risk of macrosomia among more active women. 183, no. 6, pp. 1484–1488, 2000. [16] J. A. Leiferman and K. R. Evenson, “The effect of regular leisure Conflict of Interests physical activity on birth outcomes,” Maternal and Child Health Journal,vol.7,no.1,pp.59–64,2003. The authors declare that they have no conflict of interests. [17]B.W.Alderman,H.Zhao,V.L.Holt,D.H.Watts,andS.A.A. Beresford, “Maternal physical activity in pregnancy and infant size for gestational age,” Annals of Epidemiology,vol.8,no.8,pp. References 513–519, 1998. [1] M. S. Kramer, I. Morin, H. Yang et al., “Why are babies getting [18] K. M. Owe, W.Nystad, and K. Bø, “Association between regular bigger? Temporal trends in fetal growth and its determinants,” exercise and excessive newborn birth weight,” Obstetrics and Journal of Pediatrics, vol. 141, no. 4, pp. 538–542, 2002. Gynecology,vol.114,no.4,pp.770–776,2009. [2]J.Ørskou,U.Kesmodel,T.B.Henriksen,andN.J.Secher,“An [19] A. M. Z. Jukic, K. R. Evenson, J. L. Daniels, A. H. Herring, increasing proportion of infants weigh more than 4000 grams A. J. Wilcox, and K. E. Hartmann, “A prospective study at birth,” Acta Obstetricia et Gynecologica Scandinavica,vol.80, of the association between vigorous physical activity during no. 10, pp. 931–936, 2001. pregnancy and length of gestation and birthweight,” Maternal [3] S. Bhattacharya, D. M. Campbell, W. A. Liston, and S. Bhat- and Child Health Journal,vol.16,no.5,pp.1031–1044,2012. tacharya, “Effect of body mass index on pregnancy outcomes [20] N. Voldner, K. F. Frøslie, K. Bø et al., “Modifiable determinants in nulliparous women delivering singleton babies,” BMC Public of fetal macrosomia: role of lifestyle-related factors,” Acta Health,vol.7,article168,2007. Obstetricia et Gynecologica Scandinavica,vol.87,no.4,pp.423– [4] H. M. Ehrenberg, L. Dierker, C. Milluzzi, and B. M. Mercer, 429, 2008. “Prevalence of maternal obesity in an urban center,” American [21]M.Juhl,J.Olsen,P.K.Andersen,E.A.Nøhr,andA.M.N. Journal of Obstetrics and Gynecology,vol.187,no.5,pp.1189– Andersen, “Physical exercise during pregnancy and fetal growth 1193, 2002. measures: a study within the Danish National Birth Cohort,” [5] S. Phelan, C. Hart, M. Phipps et al., “Maternal behaviors during American Journal of Obstetrics and Gynecology,vol.202,no.1, pregnancy impact offspring obesity risk,” Experimental Diabetes pp. 63.e1–63.e8, 2010. Research, vol. 2011, Article ID 985139, 9 pages, 2011. [22] H. K. Hegaard, K. Petersson, M. Hedegaard et al., “Sports and [6] S.M.Nelson,P.Matthews,andL.Poston,“Maternalmetabolism leisure-time physical activity in pregnancy and birth weight: a and obesity: modifiable determinants of pregnancy outcome,” population-based study,” Scandinavian Journal of Medicine and Human Reproduction Update,vol.16,no.3,pp.255–275,2010. Science in Sports,vol.20,no.1,pp.e96–e102,2010. [7] http://www.who.int/mediacentre/factsheets/fs311/en/index. [23] L. A. H. Haakstad, N. Voldner, T. Henriksen, and K. Bø, “Why html do pregnant women stop exercising in the third trimester?” [8] M. S. Kramer, “The epidemiology of adverse pregnancy out- Acta Obstetricia et Gynecologica Scandinavica,vol.88,no.11,pp. comes: an overview,” JournalofNutrition,vol.133,no.5, 1267–1275, 2009. supplement 2, pp. 1592S–1596S, 2003. [24] K. M. Owe, W.Nystad, and K. Bø, “Correlates of regular exercise [9] T. Henriksen, “The macrosomic fetus: a challenge in current during pregnancy: the Norwegian Mother and Child Cohort obstetrics,” Acta Obstetricia et Gynecologica Scandinavica,vol. Study,” Scandinavian Journal of Medicine and Science in Sports, 87,no.2,pp.134–145,2008. vol. 19, no. 5, pp. 637–645, 2009. 6 Journal of Pregnancy

[25] H. K. Hegaard, P. Damm, M. Hedegaard et al., “Sports and [41] D. B. Fell, K. S. Joseph, B. A. Armson, and L. Dodds, “The leisure time physical activity during pregnancy in nulliparous impact of pregnancy on physical activity level,” Maternal and women,” Maternal and Child Health Journal,vol.15,no.6,pp. Child Health Journal,vol.13,no.5,pp.597–603,2009. 806–813, 2011. [26] C. Fleten, H. Stigum, P. Magnus, and W. Nystad, “Exercise during pregnancy, maternal prepregnancy body mass index, and birth weight,” Obstetrics and Gynecology,vol.115,no.2,part 1, pp. 331–337, 2010. [27] M. Lof,¨ L. Hilakivi-Clarke, S. Sandin, and E. Weiderpass, “Effects of pre-pregnancy physical activity and maternal BMI on gestational weight gain and birth weight,” Acta Obstetricia et Gynecologica Scandinavica,vol.87,no.5,pp.524–530,2008. [28] S. Krokstad, A. Langhammer, K. Hveem et al., “Cohort profile: the HUNTStudy, Norway,” International Journal of Epidemiol- ogy.Inpress. [29] S.L.Boulet,G.R.Alexander,H.M.Salihu,andM.Pass,“Macro- somic births in the United States: determinants, outcomes, and proposed grades of risk,” American Journal of Obstetrics and Gynecology, vol. 188, no. 5, pp. 1372–1378, 2003. [30] T. I. L. Nilsen, P.R. Romundstad, and L. J. Vatten, “Recreational physical activity and risk of prostate cancer: a prospective population-based study in Norway (the HUNT study),” Inter- national Journal of Cancer, vol. 119, no. 12, pp. 2943–2947, 2006. [31] http://www.who.int/mediacentre/factsheets/fs311/en/index. html [32] M. I. Both, M. A. Overvest, M. F. Wildhagen, J. Golding, and H. I. J. Wildschut, “The association of daily physical activity and birth outcome: a population-based cohort study,” European Journal of Epidemiology,vol.25,no.6,pp.421–429,2010. [33] W. F. Schramm, J. W. Stockbauer, and H. J. Hoffman, “Exercise, employment, other daily activities, and adverse pregnancy outcomes,” American Journal of Epidemiology,vol.143,no.3,pp. 211–218, 1996. [34] M. Y. Takito, M. H. Ben´ıcio, and L. C. Neri, “Physical activity by pregnant women and outcomes for newborns: a systematic review,” Revista de saude´ publica´ ,vol.43,no.6,pp.1059–1069, 2009. [35] I. O. Frederick, M. A. Williams, A. E. Sales, D. P. Martin, and M. Killien, “Pre-pregnancy body mass index, gestational weight gain, and other maternal characteristics in relation to infant birth weight,” Maternal and Child Health Journal,vol.12,no.5, pp. 557–567, 2008. [36] H. M. Ehrenberg, B. M. Mercer, and P. M. Catalano, “The influence of obesity and diabetes on the prevalence of macrosomia,” American Journal of Obstetrics and Gynecology,vol.191,no.3, pp.964–968,2004. [37] K. Melzer, Y. Schutz, M. Boulvain, and B. Kayser, “Physical activity and pregnancy: cardiovascular adaptations, recommendations and pregnancy outcomes,” Sports Medicine,vol.40,no. 6, pp. 493–507, 2010. [38] Z. Sui, R. M. Grivell, and J. M. Dodd, “Antenatal exercise to improve outcomes in overweight or obese women: a systematic review,” Acta Obstetricia et Gynecologica Scandinavica,vol.91, no. 5, pp. 538–545, 2012. [39] N. Kurtze, V. Rangul, B. E. Hustvedt, and W. D. Flanders, “Reliabilityandvalidityofself-reportedphysicalactivityinthe Nord-Trøndelag Health Study (HUNT 2),” European Journal of Epidemiology,vol.22,no.6,pp.379–387,2007. [40] L. A. H. Haakstad, N. Voldner, T. Henriksen, and K. Bø, “Physicalactivitylevelandweightgaininacohortofpregnant Norwegian women,” Acta Obstetricia et Gynecologica Scandi- navica, vol. 86, no. 5, pp. 559–564, 2007. Hindawi Publishing Corporation Journal of Pregnancy Volume 2012, Article ID 257597, 11 pages doi:10.1155/2012/257597

Research Article Use of Medicines with Unknown Fetal Risk among Parturient Women from the 2004 Pelotas Birth Cohort (Brazil)

Andrea´ Damasoˆ Bertoldi,1 Tatiane da Silva Dal Pizzol,2 Aline Lins Camargo,3 Aluısio´ J. D. Barros,1 Alicia Matijasevich,1 and InaS.Santos´ 1

1 Programa de Pos-Graduac´ ¸ao˜ em Epidemiologia, Universidade Federal de Pelotas, Rua Marechal Deodoro, 1160, 3 Piso, 96020-220 Pelotas, RS, Brazil 2 Programa de Pos-Graduac´ ¸ao˜ em Epidemiologia, Universidade Federal do Rio Grande do Sul, Rua Ramiro Barcelos, 2400, 2 Andar, 90035-003 Porto Alegre, RS, Brazil 3 Departamento de Ciˆencias Basicas´ da Saude,´ Universidade Federal de Ciˆencias da Saude´ de Porto Alegre, Rua Sarmento Leite, 245, 90050-170, Porto Alegre, RS, Brazil

Correspondence should be addressed to Andrea´ Damasoˆ Bertoldi, [email protected]

Received 8 September 2012; Revised 25 November 2012; Accepted 26 November 2012

Academic Editor: Riitta Luoto

Copyright © 2012 Andrea´ Damasoˆ Bertoldi et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. To estimate the exposure to medicines with unknown fetal risk during pregnancy and to analyze the maternal characteristics associated with it. Methods. A questionnaire was administered to 4,189 mothers of children belonging to the 2004 Pelotas (Brazil) birth cohort study about use of any medicine during gestation. We evaluated the associations between use of medicines with unknown fetal risk and the independent variables through logistic regression models. Unknown fetal risk was defined as medicines in which studies in animals have revealed adverse effects on the fetus, and no controlled studies in women, or studies in women and animals, are available. Results. Out of the 4,189 women, 52.5% used at least one medicine from unknown fetal risk. Use of these medicines was associated with white skin color, high schooling, high income, six or more antenatal care consultations, hospital admission during pregnancy, and morbidity during gestation. Conclusion. The use of unknown fetal risk medicines is high, suggesting that their use must be addressed with caution with the aim of restricting their use to cases in which the benefits are greater than the potential risks.

1. Background Prescription of medicines during pregnancy is challenging for doctors because they need to balance the benefits Studies have demonstrated high prevalence of medicine use of the treatment against possible damages to the fetus [14]. during pregnancy. The exposure to at least one medicine dur- Most studies evaluating the risk that medicines pose to the ing pregnancy is most often exceeding 70% [1–10]. A recent fetus used the US Food and Drug Administration (FDA) systematic review, which evaluated medicine use prescribed classification [15]. Such studies have shown that exposure during pregnancy in developed countries, showed that 27 to to category A medicines (considered safe during pregnancy) 93% of the pregnant women received at least one prescrip- occurred in 49% to 100% of the participating women [5, tion of medicine, excluding vitamins and minerals [11]. 7, 16, 17]. Medicines with evidence of fetal risk (category In many situations the use may be unnecessary and even D and X) were used by less than 12% of pregnant women dangerous for the fetus, although one should consider that in [1, 5, 7, 17–20], with the exception of a study conducted in most cases the use of medicines to treat chronic conditions France, in which the percentage of exposure to category D and problems associated with pregnancy can be vital to the drugs was 59% [16]. Many studies focused on the factors health and well-being of the woman and the offspring [12, associated with the use of products from categories D or X 13]. [18, 19, 21, 22]. 2 Journal of Pregnancy

The most challenging category of medicines in this case, mentioned by commercial name, the dosage use indicated by however, is the one in which the risk to the fetus is unknown, their manufacturer label was considered to establish the risk. classified by the FDA in Category C. This group represents Regarding the association of medicines, the evidence of an important fraction when considering the quantity of fetal risk was investigated to each individual component of prescribed drugs and the number of exposed women. Studies the formulation. The formulation was classified as medicine indicate that the use of unknown fetal risk medicines during with known fetal risk when at least one of the components of pregnancy ranges from 19% to 40% [1, 5, 6, 23]. The present the formulation was identified with evidence of risk. study aimed to verify the prevalence of use of medicines Statistical analyses were performed using SPSS version with unknown fetal risk during pregnancy and to analyze 18.0. We first analyzed the association between medicine maternal characteristics associated with this outcome. use in general and the use of unknown fetal risk medicines. We then repeated the analyses after excluding vitamins and iron supplements. The variables of interest were: trimester in which the medicine was used, age of the mother, skin 2. Methods color, years of schooling, marital status, per capita income, birth order, number of prenatal consultations, gestational In 2004, the third birth cohort in the city of Pelotas, southern week of the first prenatal consultation, hospitalization during Brazil was initiated. In this study, all mothers of babies pregnancy, and the presence of the following self reported born in the city’s five maternities (99% of all births) were clinical conditions: hypertension, diabetes, depression, ane- interviewed within the first 24 hours that succeeded child- mia, threatened abortion, threatened premature birth, and birth. A standardized precoded questionnaire was applied, urinary tract infection. Chi-square tests were used for including questions about the use of any medicine during comparing proportions. Logistic regression was used for pregnancy and the initiation month and termination of adjusted analysis, adjusting all the variables of interest. A this usage. The question about medicine use was: “Please, significance level of 5% was employed. let us know the name of all medicines that you have The study was approved by the School of Medicine Ethics used during pregnancy. Please remember to consider those Committee of the Federal University of Pelotas. used for nausea, heartburn, anemia, treatment of urinary tract infection, vaginal infection, high blood pressure, or diabetes.” Further details on the methodology of the cohort 3. Results can be found elsewhere [24]. Medicines with unknown fetal risk were classified as Out of a total of 4,189 interviewed pregnant women, 92.7% those in which studies in animals have revealed adverse reported the use of some medicine during pregnancy, total- effects on the fetus (teratogenic or embryocidal or other), ing the use of 11,425 medicines. The number of medicines and there are no controlled studies in women or studies in used during gestation ranged from 1 to 10 medicines by preg- women and animals are not available. nant women with an average of 2.9 (SD 1.6). Considering the We decided to use criteria adopted by the FDA [15]for pregnancy trimesters, the average number of medicines used the category C, in order to enable possible comparisons since was 1.3 (SD 0.6), 1.4 (SD 0.7), and 1.7 (SD 0.9) during the this is the most frequent classification found in different first, second, and third trimesters, respectively. Among the studies. Medicines showing evidence about risk, leading to women who reported the use of medicines, 61.2% used one damage or safety of the fetus, were classified as medicines or two medicines, 28.7%, three or four medicines and 10.3%, with known fetal risk. five or more. If vitamins and iron are excluded from the Firstly, the drug monograph was queried in the DRUGD- analyses, 1,829 pregnant women (43.7%) reported using at EX System [25], followed by consultation of the Guide to least one medicine, adding up to the use of 7,860 medicines. Fetal and Neonatal Risk by Briggs et al. [26]. When the Out of all medicines used, 38.9% were identified as information was not located in any of the two sources, a unknown fetal risk. More than half of all mothers used at search in PubMed was performed by intersecting the drug least one product with unknown fetal risk (52.5%). name with keywords related to pregnancy and teratogenic- Table 1 summarizes the results for the trimester of the ity/adverse effects. Medicines without any information about first usage for each medicine as well as the period of their the teratogenic risk were classified as unknown fetal risk [26], use. The number of medicines with usage initiated in each in accordance with previous studies [5, 23]. Medicines were trimester was very homogeneous (around 30% for each also classified as unknown fetal risk when information was trimester) and the largest proportion of medicines was used available on the risk of the therapeutic group, but not the exclusively in the third trimester. Only 5.1% of the medicines medicine itself (e.g., benzodiazepines versus bromazepam). used in the first trimester were also used during the second When information was available on the risk of the active trimester and 14.6% of the medicines were used during the substance, but not for the salts, the classification of the three trimesters. main element was adopted (e.g., when there was information The most commonly used medicines with unknown fetal for iron, but not for the iron polymaltose). In the case of risk are presented in Table 2, by the gestational trimester of medicines for which the risk category changes depending on use. Multivitamins and the association between scopolamine the dose (e.g., ferrous sulphate) the classification was based butyl bromide and dipirone accounted for more than 50% of on the most common prescribed dose. When drugs were the medicines used in any time during pregnancy. Journal of Pregnancy 3

Table 1: Medicines used by the mothers participating in the 2004 Birth Cohort by the gestational trimester and percentages of mothers who used at least one medicine in the respective periods. Birth Cohort of Pelotas, RS 2004.

Medicines (n = 11,425) Mothers (n = 3,883) Trimester n % n % Start of the use of medicinesa 1st 3,241 29.9 1,909 50.0 2nd 3,847 35.6 2,377 62.2 3rd 3,738 34.5 2,260 59.1 Total 10,826 100.0 3,821ce Period of the use of medicinesb 1st 1,093 10.1 848 22.2 2nd 1,884 17.5 1,366 35.8 3rd 3,734 34.7 2,259 59.2 1st and 2nd 545 5.1 467 12.2 2nd and 3rd 1,939 18.0 1,504 39.4 1st, 2nd and 3rd 1,575 14.6 1,170 30.7 Total 10,770 100.0 3,814de a 599 missing; b655 missing; c62 missing; d69 missing; ethe percentages do not add up to 100% since each mother who used any medicine could have used more than one medicine in the same period or in diﬀerent periods.

Table 2: Most used medicines by the mothers participating in the 2004 Birth Cohort classiﬁed as unknown fetal risk, by the gestational trimester of medicine use. Birth Cohort of Pelotas, RS 2004.

Trimester of use a Total Medicines used at any time during 1st 2nd 3rd 1st and 2nd 2nd and 3rd 1st, 2nd, and 3rd (N = 3,505) pregnancy (N = 292) (N = 527) (N = 1,236) (N = 160) (N = 764) (N = 526) %%%%%% % Multivitamins 31.2 17.4 20.0 26.6 35.9 46.8 44.8 Scopolamine Butylbromide + 27.7 37.6 31.7 32.0 23.7 19.6 21.7 dipyrone Aluminium hydroxideb 7.4 4.5 4.6 3.9 5.3 8.9 6.5 Piperidolate + hesperidine + ascorbic 6.5 12.0 6.4 5.1 16.8 6.1 7.2 acid Nystatin 5.2 7.4 11.5 7.7 0.8 1.3 0.7 Sodium bicarbonateb 3.3 2.9 1.1 0.8 0.8 2.7 4.8 Ferrous sulphate + ascorbic acid + 3.0 2.5 1.6 0.8 4.6 4.1 8.7 folic acid Salbutamol 2.6 0.4 2.3 5.2 0.0 2.0 1.3 Isoxsuprine 2.2 1.2 2.5 3.5 1.5 2.1 0.7 Miconazole 1.9 1.7 4.4 2.8 0.0 1.1 0.0 Ascorbic acid 1.6 2.5 0.9 1.9 0.8 1.8 1.3 Diclofenac 1.4 2.9 1.8 1.6 3.8 0.4 0.4 Norﬂoxacin 1.2 0.8 3.0 1.8 0.0 0.1 0.2 Dimethicone 0.8 0.0 0.7 0.8 1.5 0.9 1.1 Terconazole 0.8 0.8 3.4 0.8 0.8 0.0 0.0 Promethazine 0.8 0.4 0.9 1.3 2.3 0.4 0.4 Betamethasone 0.6 0.0 0.5 1.4 0.0 0.3 0.2 Trimethoprim-sulfamethoxazole 0.6 1.7 1.1 0.5 0.0 0.4 0.0 Pipemidic acid 0.6 0.8 1.6 0.8 0.0 0.1 0.0 Fluoxetine 0.6 2.5 0.0 0.5 1.5 0.7 0.0 Others 8.9 17.1 17.3 20.2 18.1 7.7 12.5 a There is no information about in which gestational trimester was the use of 177 medicines classiﬁed as unknown fetal risk. bMedicine used individually or in association. 4 Journal of Pregnancy

Table 3 presents demographic, socioeconomic, and and middle income countries is likely misleading due to huge health characteristics of the mothers during pregnancy. Data heterogeneity in pharmaceutical policies. on the total prevalence of use of medicines (92.7%) and The prevalence of the use of at least one medicine during on the use of medicines with unknown fetal risk (52.5%) pregnancy was high; however, it is consistent with other according to such variables are presented. The following studies conducted in Brazil with prevalence over 90% [4, 9]. variables remained associated to the use of medicines with This usage is also quite high in other countries. However, unknown fetal risk after adjusted analyzes: white skin color, most studies investigated only the use of prescription education and per capita income (in direct proportion to medicines [5, 16, 18, 28], potentially resulting in estimates use), six or more prenatal consultations, hospitalization, that are lower than the prevalence identified in this study. and morbidity during the pregnancy (depression, anemia, Other fact to be considered is the type of question used in threat of abortion, threat of premature labor, and urinary our study. We opted to use an open-ended question followed infection). by selected indication-orientation (the question used in our Table 4 presents the same analysis conducted in Table 3, study is described in the methods section). de Jong-van but excluding vitamin and iron supplements. The total den Berg and colleagues [29] evaluated the influence of the prevalence of use of medicines excluding vitamins and type of questions used in interviews about medicine use in iron was 43.7%. The prevalence of use of medicines with pregnancy, and concluded that the prevalence of medicine unknown fetal risk was 25.6%. The following variables were use increases considerable when the question about medicine associated with the use of medicines with unknown fetal risk use includes indication-orientation and drug-orientation to in the adjusted analyses: per capita income, hospitalization, the open-ended questioning. The authors also concluded threat of abortion, and threat of premature labor. that these higher prevalence figures are more comparable In Table 5 we analyzed only women who reported to with pharmacy records. have diabetes or hypertension. Medicines used for treating It was observed that the prevalence of medicine use diabetes were insulin and metformin, and, for hypertension by pregnant women exceeds the prevalence for women in the most frequent medicine used was methyldopa followed general [30–32]. In studies carried out in Brazil to evaluate by others with low frequencies classified as “unknown risk”. medicine utilization in the general population, women in The prevalence of use of antidiabetics was 13.8% and reproductive age (18 to 49 years old) presented prevalence of the prevalence of use of antihypertensives medication was medicine use from 42% [33] to 60.4% [30]. In a European 25.8%. Among diabetic women, no antidiabetic medicine of study it was found a slightly higher prevalence (68.8%), unknown fetal risk was reported to be used, whereas 38% of considering women aged 25 to 44 years old. [32]. The use the other medicines used were classified as of unknown fetal of medicines indicated during the gestational period, such as risk. Among hypertensive women, 9.6% of the hypertension vitamins and iron derivatives, can explain these differences, medication used was classified as of unknown fetal risk, which is attenuated when these groups of medicines are whereas 41.3% of the other medicines used were classified excluded from the analysis. as such. The present study reports the gestational trimester in which the medicine use started, while available data in the literature reports only whether the pregnant women were exposed to medicines in different periods [3, 5, 8, 10, 34]. The 4. Discussion largest number of used medicines was reported as first used during the second trimester (35.6%) and a slightly smaller In this study the frequency of exposure to medicines by proportion (29.9%) was reported as first used during the gestational trimesters, the identification of the medicines first trimester, which may indicate an issue of concern about with unknown information about fetal risk, and the main exposure to medicines in early pregnancy where the risks to factors associated to these medicines were examined. Some the fetus are known to be higher [35]. features of this study are distinctive from other similar Only 14.6% of the medicines were used throughout the studies in the literature: the inclusion of prescription and entire gestational period, indicating that situations in which over-the-counter medicines, since the majority of studies do the continued use of medicines is necessary are few. However, not consider medicines used for self-medication or those this percentage does not necessarily indicate that medications prescribed during hospitalizations [16, 27], and collection are unnecessary—the medication may be warranted, but of data by interview with the mothers instead of using sec- women may not continue use, due to nonadherence or ondary data. Data presented here are directly generalizable concerns about use after realizing that they are pregnant. to the population of the city, because virtually all mothers In addition, pregnant women are usually younger and who gave birth in 2004 were included. Findings are also likely healthier that women in general, which leads to the use of valid for other areas of Brazil, which similarly to Pelotas have medicines for acute situations, such as infection, the control huge socioeconomic disparities leading to inequities in access of symptoms commonly associated with the pregnancy itself to health care. Therefore, we believe data may be extrapolated (such as heartburn) or prophylactically indicated during the to other municipalities with similar characteristics thus pregnancy (such as folic acid). contributing to the improvement of knowledge regarding Most studies classifying medicines in risk categories used factors associated to medicine use with unknown risk during the FDA criteria [1, 4–7, 16, 18–21, 26, 28, 36]. Some studies pregnancy in Brazil. Extrapolating our findings to other low excluded vitamins and minerals [1, 18] from the analyses Journal of Pregnancy 5

Table 3: Description of the sample and the health conditions of the mothers participating in the 2004 Birth Cohort (N), prevalence (P%) and adjusted analyses of the use of medicines during pregnancy considering all medicines and the medicines with unknown fetal risk used. Birth Cohort of Pelotas, RS 2004. N Use of medicines (total) Use of medicines with unknown fetal risk Characteristic Adjusted analyses Adjusted analyses P% PR (IC 95%) P value P% PR (IC 95%) P value Total of participants 4,189 92.7 — — 52.5 — — Age 0,062 0.076 <20 796 90.8 1.00 43.6 1.00 20 to 29 2,085 93.1 1.58 (1.03–2.43) 53.2 1.09 (0.90–1.32) 30 to 39 1,170 92.9 2.01 (1.19–3.38) 56.9 1.31 (1.05–1.64) 40 or more 136 94.1 2.35 (0.83–6.62) 55.1 1.18 (0.76–1.84) Skin color 0,08 <0.001 White 2,555 93.8 1.31 (0.97–1.78) 57.7 1.36 (1.17–1.58) Non-white 1,586 90.9 1.00 44.1 1.00 Education in years 0.026 <0.001 0 to 4 647 87.3 1.00 32.8 1.00 5 to 8 1,711 91.6 1.50 (1.02–2.23) 47.0 1.70 (1.35–2.13) 9 to 11 1,396 95.1 1.90 (1.19–3.03) 62.7 2.57 (2.01–3.30) 12 or more 435 97.2 2.66 (1.22–5.79) 70.6 2.74 (1.95–3.85) Marital status 0.064 0.92 With the baby’s father 3,502 93.1 1.44 (0.98–2.11) 53.5 0.99 (0.81–1.21) Without 687 90.5 1.00 47.0 1.00 Income per capita 0.09 <0.001 1st quintile 830 88.4 1.00 38.2 1.00 2nd quintile 834 89.2 0.83 (0.55–1.25) 41.7 1.03 (0.82–1,30) 3rd quintile 886 94.2 1,48 (0.93–2.36) 52.7 1.29 (1.03–1.62) 4th quintile 849 94.7 1.11 (0.68–1.82) 61.1 1.79 (1.41–2.27) 5th quintile 790 96.8 1.55 (0.81–2.95) 69.2 1.94 (1.47–2.55) Parity 0.025 0.52 primipara 1,658 94.8 2.45 (1.34–4.46) 55.9 1.12 (0.81–1.56) 1 or 2 1,086 93.0 1.65 (0.96–2.82) 54.8 1.06 (0.78–1.45) 3 or 4 1,026 91.6 1.36 (0.83–2.23) 48.6 0.95 (0.70–1.27) 5 or more 419 86.2 1.00 42.2 1.00 Prenatal consultation 0.004 0.001 <6 689 86.6 1.00 36.0 1.00 6 or more 3,250 94.9 1.65 (1.18–2.32) 56.8 1.39 (1.14–1.70) Startofprenatalcare 0.70 0.32 Before the 20th week 3,765 94.0 1.10 (0.68–1.80) 54.8 1.18 (0.85–1.64) After the 20th week 283 84.8 1.00 31.4 1.00 Hospitalization No 3,722 91.9 1.00 0.002 50.9 1.00 0.05 Yes 467 98.9 6.50 (1.99–21.21) 64.7 1.28 (1.00–1.62) High Blood Pressure 0.012 0.87 No 3,189 92.3 1.00 52.6 1.00 Yes 992 94.0 1.60 (1.11–2.31) 52.3 0.90 (0.59–1.37) Diabetes 0.76 0.63 No 4,062 92.6 1.00 52.4 1.00 Yes 124 96.0 0.86 (0.32–2.30) 56.4 0.90 (0.59–1.37) Depression 0.034 <0.001 No 3,138 92.1 1.00 51.0 1.00 Yes 1049 94.3 1.49 (1.03–2.16) 56.9 1.38 (1.16–1.64) Anemia <0.001 <0.001 No 1,402 83.0 1.00 42.7 1.00 Yes 2756 97.8 10.79 (7.68–15.16) 57.7 1.53 (1.31–1.78) 6 Journal of Pregnancy

Table 3: Continued. N Use of medicines (total) Use of medicines with unknown fetal risk Characteristic Adjusted analyses Adjusted analyses P% PR (IC 95%) P value P% PR (IC 95%) P value Threat of abortion 0.15 <0.001 No 3,736 92.2 1.00 50.0 1.00 Yes 449 96.4 1.55 (0.85–2.83) 73.0 2.40 (1.84–3.11) Threat of premature labor <0.001 <0.001 Nao˜ 3,416 91.7 1.00 48.0 1.00 Sim 771 97.0 2.72 (1.57–4.73) 72.5 2.65 (2.15–3.27) Urinary Infection <0.001 0.012 Nao˜ 2,623 89,8 1.00 52.5 1.00 Sim 1,552 97.7 5.53 (3.58–8.54) 52.5 0.82 (0.71–0.96) ∗ Chi square test; ∗∗1st quintile relates to the most deprived quintile ∗∗∗PR: prevalence ratio. while others evaluated the use of the medicines in part of B12 (2.2–2.6 µg), vitamin C (70 mg), vitamin E (10 mg), and the gestation period (up to the gestational age at the time nicotinamide (17 mg). The following vitamins show positive of the interview) [23, 36]. Only studies that evaluated the use evidence of fetal risk in concentrations higher than 400 IU of medicines throughout the whole gestational period and and 8,000 IU/day, respectively, for Vitamin D (and D3) and included vitamins and minerals were selected for the purpose vitamin A. of comparisons with our data [4–7, 16, 18, 21, 28]. The use of multivitamins during pregnancy is often The frequency of the use of medicines with unknown suggested as an intervention aimed at improving maternal fetal risk (38.9%) is consistent with a study conducted with and fetal health. However, substantive evidence regarding the parturients interviewed after childbirth in a hospital from effectiveness of multiple-micronutrient supplements during another Brazilian city (42.4%) [4], however, higher than the pregnancy is not available. Its use during the prenatal period reported frequency in a study conducted in a city in southern is associated with controversial results in the literature which Brazil (24.5%) [6] and in Ethiopia (15.2%) [7]. points to positive results [38] and lack of effect [39]in The prevalence of women exposed to the medicines with terms of the benefit of multivitamins in the outcome of unknown fetal risk (52.5%) was higher than that observed low birth weight. There is not sufficient evidence for other in most studies (14.7 to 19%) [5, 7, 28] but lower than the relevant clinical outcomes. Therefore, the use of this type one reported in the study by Lacroix (85.0%) [16]. The study of medicine, advocated in the obstetric practice, should be by Lacroix presents the peculiarity of inclusion of products assessed considering the different existing formulations in that generally were not included in other studies, such as the market and acknowledging that, depending on the dose homeopathic remedies [37], thus leading to its elevated per- present in the formulation its use is only justified if the centages. Other studies do not mention having investigated potential benefits overcome the potential risks for the fetus. the primary literature about teratogenic risk, which may have The analysis of the variables potentially related to a led to an underestimated classification of many medicines greater use of medicines with unknown fetal risk was also in this group [5, 7, 18, 21, 28]. According to the FDA’s performed for the use of all medicines during pregnancy. recommendations, the use of these medicines should only White women with higher education and greater income occur when the potential benefits justify the potential risks used more medicines with unknown fetal risk. In general, to the fetus. As a consequence, the uncertainty about the higher education alone was associated with greater use of risks associated with this category can generate doubts and medicines, which is in agreement with a study conducted in anxiety in doctors and pregnant women. However, medical France and others in several Brazilian cities [4, 23, 36, 40]. information justifying the use of these medicines by each Primiparous status increases the proportion of the use of pregnant woman is not available for analysis, which would medicines (excluding medicines with unknown fetal risk) allow a better assessment of the impact from the use of these in this study and other studies in Brazil [23, 36]. Mothers medicines. who had a greater number of prenatal consultations used Formulations with multiple vitamins stood out in the more medicines in general and specifically medicines with analyses of the medicines with unknown fetal risk during unknown fetal risk. However, the use of medicines is not pregnancy. The risk of some vitamins during gestation associated to the initiation of prenatal care occurring either is dose-dependent. The multivitamin’s components were before or after the 20th week of gestation. In the study by analyzed individually and classified according to their dosage Guerra et al., prenatal care in the 1st trimester was associated in the formulation. According to Briggs et al. [26], the with a higher use of medicines [36]. following vitamins, in concentrations higher than the ones Generally speaking, the presence of morbidity was described as follows, present unknown fetal risk: vitamin B1 associated to an increased risk of the use of medicines in (1.5 mg), vitamin B2 (1.6 mg), vitamin B5 (10 mg), vitamin general and also medicines with unknown fetal risk. This Journal of Pregnancy 7

Table 4: Description of the sample and the health conditions of the mothers participating in the 2004 Birth Cohort (N), prevalence (P%) and adjusted analyses of the use of medicines during pregnancy considering all medicines and the medicines with unknown fetal risk used, excluding vitamin and iron supplements. Birth Cohort of Pelotas, RS 2004.

Use of medicines (total) Use of medicines with unknow fetal risk Characteristic NP Adjusted analyses P Adjusted analyses % PR (IC 95%) P value % PR (IC 95%) P value Total of participants 4,189 43.7 — — 25.6 — — Age 0.275 0.883 <20 796 38.8 1.00 24.5 1.00 20 to 29 2,085 44.6 1.19 (0.98–1.44) 26.1 0.98 (0.78–1.23) 30 to 39 1,170 44.8 1.21 (0.98–1.50) 25.7 0.93 (0.70–1.23) 40 or more 136 47.8 1.25 (0.82–1.88) 23.5 0.85 (0.51–1.40) Skin color 0.038 0.154 White 2,555 45.8 1.17 (1.01–1.35) 27.0 1.12 (0.96–1.32) Non-white 1,586 40.2 1.00 23.1 1.00 Education in years 0.002 0.075 0 to 4 647 40.6 1.00 19.8 1.00 5 to 8 1,711 40.6 1.09 (0.88–1.35) 25.0 1.28 (1.02–1.61) 9 to 11 1,396 47.9 1.42 (1.12–1.79) 28.2 1.39 (1.08–1.78) 12 or more 435 46.9 1.56 (1.14–2.13) 28.5 1.36 (1.00–1.89) Marital status 0.615 0.331 With the baby’s father 3,502 44.2 1.05 (0.86–1.29) 26.0 1.11 (0.90–1.39) Without 687 41.0 1.00 23.3 1.00 Income per capita 0.009 0.020 1st quintile 830 39.0 1.00 20.5 1.00 2nd quintile 834 40.3 1.03 (0.83–1.28) 24.8 1.26 (1.00–1.59) 3rd quintile 886 45.8 1.37 (1.10–1.70) 24.8 1.21 (1.00–1.53) 4th quintile 849 46.5 1.37 (1.09–1.72) 28.5 1.45 (1.14–1.84) 5th quintile 790 46.6 1.30 (1.01–1.68) 29.5 1.48 (1.14–1.92) Parity 0.991 0.830 primipara 1,658 43.5 0.99 (0.72–1.37) 26.5 0.94 (0.67–1.32) 1 or 2 1,086 44.5 1.01 (0.75–1.37) 25.9 0.91 (0.66–1.26) 3 or 4 1,026 44.4 0.98 (0.74–1.30) 24.5 0.88 (0.65–1.19) 5 or more 419 40.3 1.00 23.9 1.00 Prenatal consultation 0.747 0.277 <6 689 42.8 1.00 24.2 1.00 6 or more 3,250 43.9 0.96 (0.77–1.20) 26.2 0.88 (0.70–1.11) Start of prenatal care 0.861 0.517 Before the 20th week 3,765 43.7 0.97 (0.71–1.33) 26.2 1.12 (0.80–1.58) After the 20th week 283 40.6 1.00 20.8 1.00 Hospitalization <0.001 <0.001 No 3,722 42.1 1.00 24.3 1.00 Yes 467 56.3 1.51 (1.20–1.89) 35.5 1.48 (1.19–1.84) High Blood Pressure <0.001 0.305 No 3,189 40.9 1.00 25.1 1.00 Yes 992 52.5 1.62 (1.38–1.90) 27.2 1.10 (0.92–1.31) Diabetes 0.995 0.390 No 4,062 43.5 1.00 25.4 1.00 Yes 124 50.8 1.00 (0.67–1.50) 32.3 1.20 (0.79–1.81) 8 Journal of Pregnancy

Table 4: Continued. Use of medicines (total) Use of medicines with unknow fetal risk Characteristic NP Adjusted analyses P Adjusted analyses % PR (IC 95%) P value % PR (IC 95%) P value Depression <0.001 0.066 No 3,138 41.5 1.00 24.8 1.00 Yes 1049 50.1 1.45 (1.24–1.70) 28.1 1.18 (0.99–1.40) Anemia <0.001 0.366 No 1,402 64.4 1.00 24.2 1.00 Yes 2756 32.9 0.23 (0.20–0.26) 26.4 1.08 (0.92–1.26) Threat of abortion 0.003 <0.001 No 3,736 42.3 1.00 24.4 1.00 Yes 449 54.6 1.39 (1.11–1.73) 35.6 1.48 (1.19–1.83) Threat of premature labor <0.001 <0.001 Nao˜ 3,416 41.0 1.00 23.5 1.00 Sim 771 55.3 1.46 (1.22–1.76) 34.9 1.50 (1.25–1.80) Urinary Infection <0.001 0.434 Nao˜ 2,623 38.6 1.00 24.6 1.00 Sim 1,552 52.2 2.10 (1.81–2.42) 27.2 1.06 (0.91–1.24) ∗ Chi square test; ∗∗1st quintile relates to the most deprived quintile, ∗∗∗PR: prevalence ratio.

Table 5: Medicine use among hypertensive and diabetic women. Birth Cohort of Pelotas, RS 2004.

Women with diabetes (n = 124)∗ Women with hypertension (n = 992)∗∗ Medicines N % % with unknown risk Medicines N % % with unknown risk Antidiabetics 15 13.8 — Antihypertensives 219 25.8 9.6 Other 94 86.2 38.0 Other 630 74.2 41.3 ∗ Casos validos:´ 109 mulheres com informac¸ao˜ sobre o nome dos medicamentos. ∗∗Casos validos:´ 849 mulheres com informac¸ao˜ sobre o nome dos medicamentos.

was verified among mothers who were hospitalized during of use of medicines with unknown fetal risk in pregnancy. the pregnancy and those who presented depression, anemia, Another issue to be considered is that due to the low cost threat of premature birth and urinary tract infection. A of iron supplements, these products are widely used by greater risk of the use of medicines with unknown fetal all socioeconomic groups, and therefore, exclusion of them risk was not observed specifically by the mothers suffering from the analyses maximizes the differences between low from high blood pressure since, methyldopa was among one and high income women. The association of number of of the most used antihypertensive medications by pregnant antenatal consultations and use of medicines was also no women. Methyldopa has not been confirmed to present a longer significant when vitamin and iron supplements were risk to the embryo or fetus. The increased use of medicines excluded from the analysis. The use of such products is more among mothers with chronic diseases was also verified in common among low risk pregnant women and those with studies by Costa da Fonseca et al. and Guerra et al. [4, 36]. higher number of antenatal care consultations, because the In order to have some information about use of prescription of antianemic products is highly prevalent in medicines to treat chronic diseases in pregnancy, we analyzed Brazil despite some controversies in the literature [41]. pregnant women with diabetes and hypertension due to We opted for the use of logistic regression in order to the high burden of disease caused by these two illnesses maintain comparability with most of the previous published in Brazil. It was observed that all antidiabetics and most studies in the subject. However, it should be noted that, antihypertensive products are of known risk, with only a because the outcome “use of medicines” has high prevalence, small percentage of antihypertensive medicines classified the odds ratios estimated by logistic regression are well above as unknown fetal risk. This might be explained by more their respective prevalence ratios. rigorous prescription patterns for treating these two diseases The absence of a risk assessment in accordance with the due to their high burden. medicine doses used by mothers is a common problem in It was possible to identify that vitamins and iron most studies including this one. It is known that teratogenic supplements play a key role at increasing the prevalence and fetotoxic effects are usually dependent on dosage and Journal of Pregnancy 9 the length of usage time [37]. The fact that the investigation Authors’ Contribution about the use of medicines occurred in a single moment (at the interview time), after childbirth, can lead to recall errors A. D. Bertoldi, T. S. Dal Pizzol, and A. L. Camargo performed as a function of the long period investigated (nine months). the statistical analysis and drafted the paper. A. J. D. Barros, In addition to this, the fact that the mothers did not have the A. Matijasevich and I. S. Santos were responsible for the packaging of the used medicines at the time of the interview, conception, design, acquisition of data, and general super- which often made it impossible for the identification of the vision of the research group. Also, they have been involved exact name of the medicine used and consequent inability in revising the paper critically for important intellectual to assign it in the proper risk category due to its unknown content. All authors have given final approval of the version composition must also be considered in the overall analyses. to be published. The risk evaluation by the FDA and other similar classifications has been widely criticized, especially as they Acknowledgments lead to an incorrect impression that the risk increases from A to X and that drugs from the same category present the same This paper is based on data from the study “Pelotas birth potential risk. Further, the system does not address potential cohort, 2004” conducted by the Postgraduate Program in developmental adverse events on the basis of expected Epidemiology at Federal University of Pelotas, Brazil. The incidence, severity, or reversibility, nor whether there are 2004 birth cohort study is currently supported by the degrees of risk based on the dose, duration, frequency, Wellcome Trust Initiative entitled Major Awards for Latin route, or gestational timing of exposure to a given product. America on Health Consequences of Population Change. The FDA was led to propose, in 2008, a review to present Previous phases of the study were supported by the World information about medicine use during pregnancy included Health Organization, National Support Program for Centers in the package insert of the medicines. This proposal about of Excellence (PRONEX), the Brazilian National Research the new regulation has been discussed, but until now the new Council (CNPq), the Brazilian Ministry of Health, and the labeling requirements have not yet been initiated [26, 42, 43]. Children’s Mission. The FDA initiative to revise this classification contributes to an important and necessary reflection about the safety of medicines used during gestation, considering that the classi- References fication adopted for many years confounded the prescribers [1] S. E. Andrade, J. H. Gurwitz, R. L. Davis et al., “Prescription in the moment to decide about the use or not of some drug use in pregnancy,” American Journal of Obstetrics and medicines. The majority of the medicines do not clearly state Gynecology, vol. 191, no. 2, pp. 398–407, 2004. the risk or safety associated with the use of the product. This [2]M.K.Bakker,J.Jentink,F.Vroom,P.B.vandenBerg,H.E.K. fact lead the category C to an uncertain category, where data de Walle, and L. T. W. de Jong-Van den Berg, “Drug prescrip- available do not state whether the medicines are safe or not. tion patterns before, during and after pregnancy for chronic, occasional and pregnancy-related drugs in the Netherlands,” International Journal of Obstetrics and Gynaecology, vol. 113, 5. Conclusion no. 5, pp. 559–568, 2006. [3] A. Engeland, J. G. Bramness, A. K. Daltveit, M. Rønning, S. Cohort studies aimed at the measurement of the conse- Skurtveit, and K. Furu, “Prescription drug use among fathers quences of the use of medicines during pregnancy are and mothers before and during pregnancy. A population- needed. For the advancement of knowledge in this subject, based cohort study of 106 000 pregnancies in Norway 2004– the evaluation of dosages and duration of the use of each 2006,” British Journal of Clinical Pharmacology, vol. 65, no. 5, medicine by the mothers during pregnancy should be cor- pp. 653–660, 2008. related to the health outcomes of the mother and newborn [4] M. R. C. Costa da Fonseca, E. da Fonseca, and G. Bergsten- in different phases of development. Health professionals Mendes, “Prevalence of drug use during pregnancy: a phar- should be frequently educated on relationships between macoepidemiological approach,” Revista de Saude Publica, vol. teratogenic agents and maternal and fetal health in order 36, no. 2, pp. 205–212, 2002. to make appropriate recommendations regarding the use [5] J. J. Gagne, V. Maio, V. Berghella, D. Z. Louis, and J. of some medicines based on the best information available S. Gonnella, “Prescription drug use during pregnancy: a population-based study in Regione Emilia-Romagna, Italy,” and not solely based on classifications that are subject to European Journal of Clinical Pharmacology, vol. 64, no. 11, pp. significant errors, especially for taking into account a short 1125–1132, 2008. term follow-up of the offspring. Little is known about long ff [6] L. T. C. Geib, E. F. Vargas Filho, D. Geib, D. I. Mesquita, and M. term consequences for the o spring and for that reason these L. Nunes, “Prevalence and determinants of medication during classification systems have limitations. pregnancy by risk class among mothers of liveborn infants,” In addition, regulatory changes obligating pharma- Cadernos de Saude Publica, vol. 23, no. 10, pp. 2351–2362, ceutical companies to include more detailed information 2007. about the risks associated with the use of medicines and [7] B. Kebede, T. Gedif, and A. Getachew, “Assessment of drug health professionals’ training are needed, particularly during use among pregnant women in Addis Ababa, Ethiopia,” pregnancy, therefore empowering health professionals to Pharmacoepidemiology and Drug Safety, vol. 18, no. 6, pp. 462– prescribe effective and safe medicines to patients. 468, 2009. 10 Journal of Pregnancy

[8] C. Olesen, H. T. Sørensen, L. de Jong-Van den Berg et al., [26] G. G. BRIGGS, R. K. FREEMAN, and S. J. YAFFE, Drugs “Prescribing during pregnancy and lactation with reference in Pregnancy and Lactation, Lippincott Williams & Wilkins, to the Swedish classification system. A population-based Philadelphia, 9th edition, 2011. study among Danish women,” Acta Obstetricia et Gynecologica [27] F. Haramburu, G. Miremont-Salame,´ and N. Moore, “Good Scandinavica, vol. 78, no. 8, pp. 686–692, 1999. and bad drug prescription in pregnancy,” Lancet, vol. 356, no. [9]C.G.Osorio-de-Castro,V.L.Pepe,V.L.Luizaetal.,“Pre- 9243, p. 1704, 2000. scribed and reported drug use during pregnancy,” Cadernos [28] T. Yang, M. C. Walker, D. Krewski et al., “Maternal character- de Saude´ Publica´ , vol. 20, supplement 1, pp. S73–82, 2004. istics associated with pregnancy exposure to FDA category C, [10] D. K. Rohra, N. Das, S. I. Azam et al., “Drug-prescribing D, and X drugs in a Canadian population,” Pharmacoepidemi- patterns during pregnancy in the tertiary care hospitals ology and Drug Safety, vol. 17, no. 3, pp. 270–277, 2008. of Pakistan: a cross sectional study,” BMC Pregnancy and [29] L. T. de Jong-van den Berg, C. M. Waardenburg, F. M. Childbirth, vol. 8, article 24, 2008. Haaijer-Ruskamp, M. N. G. Dukes, and H. Wesseling, “Drug [11] J. R. Daw, G. E. Hanley, D. L. Greyson, and S. G. Morgan, “Pre- use in pregnancy: a comparative appraisal of data collecting scription drug use during pregnancy in developed countries: methods,” European Journal of Clinical Pharmacology, vol. 45, a systematic review,” Pharmacoepidemiology and Drug Safety, no. 1, pp. 9–14, 1993. vol. 20, no. 9, pp. 895–902, 2011. [30] A. D. Bertoldi, A. J. D. Barros, P. C. Hallal, and R. C. [12] G. Koren, A. Pastuszak, and S. Ito, “Drugs in Pregnancy,” New Lima, “Drug utilization in adults: prevalence and individuals England Journal of Medicine, vol. 338, no. 16, pp. 1128–1137, determinants,” Revista de Saude Publica,vol.38,no.2,pp. 1998. 228–238, 2004. [13] F. Banhidy,´ R. Lowry, and A. Czeizel, “Risk and benefit of [31] P. Carrasco-Garrido, R. Jimenez-Garc´ ´ıa, V. Hernandez´ Bar- drug use during pregnancy,” International Journal of Medical rera, and A. Gil de Miguel, “Predictive factors of self- Sciences, vol. 2, no. 3, pp. 100–106, 2005. medicated drug use among the Spanish adult population,” [14] H. Malm, J. Martikainen, T. Klaukka, and P. J. Neuvo- Pharmacoepidemiology and Drug Safety, vol. 17, no. 2, pp. 193– nen, “Prescription drugs during pregnancy and lactation— 199, 2008. a Finnish register-based study,” European Journal of Clinical [32] S. Sans, G. Paluzie, T. Puig, L. Balana, and I. I. Balaguer- Pharmacology, vol. 59, no. 2, pp. 127–133, 2003. Vintro, “Prevalence of drug utilization in the adult population [15] Food and Drug Administration, “Code of Federal Regulati- of Catalonia, Spain,” Gaceta Sanitaria, vol. 16, no. 2, pp. 121– ons,” Title 21, Vol. 4, 2012, http://www.accessdata.fda.gov/ 130, 2002. scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=201. [33]K.S.Costa,M.B.A.deBarros,P.M.S.B.Francisco,C.L.G. [16] I. Lacroix, C. Damase-Michel, M. Lapeyre-Mestre, and J. Cesar,´ M. Goldbaum, and L. Carandina, “Use of medication L. Montastruc, “Prescription of drugs during pregnancy in and associated factors: a population-based study in Campinas, France,” Lancet, vol. 356, no. 9243, pp. 1735–1736, 2000. SaoPauloState,Brazil,”˜ Cadernos de Saude Publica, vol. 27, no. [17] A. Basgul,A.Akici,A.Uzuneretal.,“Drugutilizationand¨ 4, pp. 649–658, 2011. teratogenicity risk categories during pregnancy,” Advances in [34] V. Egen-Lappe and J. Hasford, “Drug prescription in preg- Therapy, vol. 24, no. 1, pp. 68–80, 2007. nancy: analysis of a large statutory sickness fund population,” [18] S. Andrade, M. Raebel, A. Morse et al., “Use of prescription European Journal of Clinical Pharmacology,vol.60,no.9,pp. medications with a potential for fetal harm among pregnant 659–666, 2004. women,” Pharmacoepidemiology and Drug Safety, vol. 15, no. [35] P. Wiffen, M. Mitchell, and M. Snelling, Oxford Handbook of 8, pp. 546–554, 2006. Clinical Pharmacy Oxford, Oxford University Press, Oxford, [19] E. H. Riley, E. Fuentes-Afflick, R. A. Jackson et al., “Corre- UK, 2007. lates of prescription drug use during pregnancy,” Journal of [36]G.C.Guerra,A.Q.daSilva,L.B.Franca,P.M.Assuncao,R. Women’s Health, vol. 14, no. 5, pp. 401–409, 2005. X. Cabral, and A. A. Ferreira, “Drug use during pregnancy in [20] J. R. Daw, B. Mintzes, M. R. Law, G. E. Hanley, and S. G. Natal, Brazil,” Revista Brasileira de Ginecologia e Obstetr´ıcia, Morgan, “Prescription drug use in pregnancy: a retrospective, vol. 30, no. 1, pp. 12–18, 2008. population-based study in British Columbia, Canada (2001– [37] A. Addis, N. Magrini, and P. Mastroiacovo, “Drug use during 2006),” Clinical Therapeutics, vol. 34, no. 1, pp. 239–249, 2012. pregnancy,” Lancet, vol. 357, no. 9258, p. 800, 2001. [21] C. Irl and J. Hasford, “The PEGASUS project—a prospective [38] P. S. Shah and A. Ohlsson, “Effects of prenatal multimi- cohort study for the investigation of drug use in pregnancy,” cronutrient supplementation on pregnancy outcomes: a meta- International Journal of Clinical Pharmacology and Therapeu- analysis,” Canadian Medical Association Journal, vol. 180, no. tics, vol. 35, no. 12, pp. 572–576, 1997. 12, pp. E99–E108, 2009. [22] H. Malm, J. Martikainen, T. Klaukka, and P. J. Neuvonen, [39] B. A. Haider and Z. A. Bhutta, “Multiple-micronutrient “Prescription of hazardous drugs during pregnancy,” Drug supplementation for women during pregnancy,” Cochrane Safety, vol. 27, no. 12, pp. 899–908, 2004. Database of Systematic Reviews, no. 4, Article ID CD004905, [23] S. S. Mengue, E. P. Schenkel, B. B. Duncan, and M. I. Schmidt, 2006. “Drug use by pregnant women in six Brazilian cities,” Revista [40] C. Damase-Michel, M. Lapeyre-Mestre, C. Moly, A. Fournie,´ de Saude Publica, vol. 35, no. 5, pp. 415–420, 2001. and J. L. Montastruc, “Drug use during pregnancy: survey in [24]A.J.D.Barros,I.D.S.dosSantos,C.G.Victoraetal.,“The 250 women consulting at a university hospital center,” Journal 2004 Pelotas birth cohort: methods and description,” Revista de Gynecologie Obstetrique et Biologie de la Reproduction, vol. de Saude Publica, vol. 40, no. 3, pp. 402–413, 2006. 29, no. 1, pp. 77–85, 2000. ff [25] R. Klasco, “DRUGDEX System,” in Thomson MICROMEDEX, [41] J. P. Pena-Rosas˜ and F. E. Viteri, “E ects and safety of 2010. preventive oral iron or iron+folic acid supplementation for Journal of Pregnancy 11

women during pregnancy,” Cochrane Database of Systematic Reviews, no. 4, Article ID CD004736, 2009. [42] K. B. Feibus, “FDA’s proposed rule for pregnancy and lactation labeling: improving maternal child health through well- informed medicine use,” Journal of Medical Toxicology, vol. 4, no. 4, pp. 284–288, 2008. [43] L. B. Holmes, “Human teratogens: update 2010,” Birth Defects Research Part A, vol. 91, no. 1, pp. 1–7, 2011. Hindawi Publishing Corporation Journal of Pregnancy Volume 2012, Article ID 945196, 9 pages doi:10.1155/2012/945196

Research Article Effects of Tobacco Smoking in Pregnancy on Offspring Intelligence at the Age of 5

Hanne-Lise Falgreen Eriksen,1, 2 Ulrik Schiøler Kesmodel,1, 3 Theresa Wimberley,4, 5 Mette Underbjerg,1, 6 Tina Røndrup Kilburn,1, 7 and Erik Lykke Mortensen8

1 Section of Epidemiology, Department of Public Health, Aarhus University, Bartholins All´e 2 Aarhus C, 8000 Aarhus C, Denmark 2 Department of Clinical Medicine, Center of Functionally Integrative Neuroscience, Aarhus University, 8000 Aarhus C, Denmark 3 Department of Obstetrics and Gynaecology, Aarhus University Hospital, 8200 Aarhus N, Denmark 4 Section of Biostatistics, Department of Public Health, Aarhus University, 8000 Aarhus C, Denmark 5 Department of Economics and Business, National Centre for Register-Based Research, Aarhus University, 8000 Aarhus C, Denmark 6 Children’s Neurocenter Vejlefjord Rehabilitation Center, 7140 Stouby, Denmark 7 Center of Child and Adolescent Psychiatry, Central Denmark Region, Section C, 8240 Risskov, Denmark 8 Institute of Public Health and Center for Healthy Ageing, University of Copenhagen, 1353 Copenhagen K, Denmark

Correspondence should be addressed to Hanne-Lise Falgreen Eriksen, [email protected]

Received 7 October 2012; Accepted 22 November 2012

Academic Editor: Riitta Luoto

Copyright © 2012 Hanne-Lise Falgreen Eriksen et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The aim of the study was to examine the effects of tobacco smoking in pregnancy on children’s IQ at the age of 5. A prospective follow-up study was conducted on 1,782 women, and their offspring were sampled from the Danish National Birth Cohort. At 5 years of age, the children were tested with the Wechsler Preschool and Primary Scale of Intelligence-Revised. Parental education, maternal IQ, maternal alcohol consumption in pregnancy, the sex and age of the child, and tester were considered core confounders, but the full model also controlled for prenatal paternal smoking, maternal age and Bodymass Mass Index, parity, family/home environment, postnatal parental smoking, breast feeding, the child’s health status, and indicators for hearing and vision impairments. Unadjusted analyses showed a statistically significant decrement of 4 points on full-scale IQ (FSIQ) associated with smoking 10+ cigarettes per day compared to nonsmoking. After adjustment for potential confounders, no significant effects of prenatal exposure to tobacco smoking were found. Considering the indisputable teratogenic effects of tobacco smoking, these findings should be interpreted with caution. Still, the results may indicate that previous studies that failed to control for important confounders, particularly maternal intelligence, may be subject to substantial residual confounding.

1. Introduction associated with prenatal exposure to tobacco smoke includes preterm delivery, pre- and postnatal growth restriction [4, 5], Thenegativeeﬀects on health of active as well as passive congenital malformations, [6] stillbirth, [7] and increased exposure to tobacco smoking have long been known. A risk of Sudden Infant Death Syndrome [8, 9]. special case of exposure is that of the developing fetus Further, tobacco smoking may act as a neuroterato- when a pregnant woman is smoking, exposing the fetus gen through various mechanisms [10]. Nicotine and its to the adverse eﬀects of the numerous toxins contained in metabolite cotinine may alter the function of several neu- tobacco, such as vasoconstriction and hypoxia [1]. Blood rotransmitter systems, primarily acetylcholine, serotonin, concentrations of cotinine, a nicotine metabolite, in exposed and catecholamines [3]. Functionally, prenatal exposure newborns indicate that the fetus is exposed to equal—or even to tobacco smoking has been associated with cognitive higher—levels of nicotine than the smoking mother [2, 3]. impairments, particularly in attention and linguistic skills The list of known adverse short- and long-term sequelae presumably related to compromised auditory processing 2 Journal of Pregnancy

[11–15]. Studies of the effects on general cognitive abilities of the test session, we used three verbal (Arithmetic, Infor- or intelligence have provided mixed results. Hence, while mation, and Vocabulary) and three performance subtests some studies reported lower IQ scores in exposed individuals (Block Design, Geometric Design, and Object Assembly). [16–20], others reported significant associations to disappear This set of subtests was selected taking into consideration (a) after adjustment for confounders [21–24], and yet others correlation with FSIQ and (b) variety in the composition of found no association [25]. The overall conclusion regarding the test battery which should make it possible to derive verbal the effects of smoking in pregnancy on offspring intelligence and performance IQ in addition to FSIQ. thus remains ambiguous, and it is widely debated whether Standard procedures [30]wereusedtoprorateVerbal previously reported associations reflect causal relations or IQ (VIQ), Performance IQ (PIQ), and Full-Scale IQ (FSIQ) rather methodological shortcomings, such as residual con- from this shortened form of the test. founding [23, 26]. No Danish WPPSI-R norms were available at the time The purpose of this study was to examine the effects of of the study, and consequently Swedish norms were used to prenatal exposure to tobacco smoking on psychometric intel- derive scaled scores and IQs [31]. Because Swedish norms ligence (IQ) in a large sample of 5-year-old children while were used, the theoretical IQ distribution of a mean of taking into account a wide range of important confounders, 100 and a standard deviation (SD) of 15 cannot necessarily including parental education, maternal intelligence, and be expected in this sample, and IQ-scores may not be alcohol consumption in pregnancy, and postnatal smoke uncorrelated with age. This, however, will not affect internal exposure. comparisons made within the sample with respect to effects of smoke exposure. 2. Materials and Methods Testing took place in one of the four major cities of Denmark (Copenhagen, Odense, Aalborg, and Aarhus). Test 2.1. Study Sample. The study was a part of the Lifestyle procedures were standardised in detail and carried out by During Pregnancy Study (LDPS) [27], a prospective follow- 10 trained psychologists. Tester differences were taken into up study of the effects of various maternal lifestyle factors account by the inclusion of a categorical variable for tester in in pregnancy, primarily intake of alcohol, on motor and the statistical analyses. cognitive outcomes at the age of 5 years. The LDPS is based on a subsample from the Danish National Birth Cohort 2.4. Covariates. The following information was obtained by [28], a large cohort study with information on 101,042 the prenatal telephone interview and subsequently coded women and their children, collected by two prenatal and two as shown in parenthesis: maternal alcohol consumption postnatal telephone interviews. during pregnancy (yes/no), parity (0, 1, 2+), maternal pre- The data collection of the LDPS took place from Septem- pregnancy BMI (weight in kg/(height in m)2), and prenatal ber 2003 to June 2008 the period during which 3,478 mothers paternal smoking (yes/no). and their children were invited to a followup when the child A questionnaire administered at the followup provided was from 60 to 64 months of age. Of these, 1,782 (51%) information on the following variables: maternal marital sta- participated in a comprehensive three-hour assessment of the tus (single at either the prenatal interview or followup/with child’s cognitive ability, including tests of global and specific partner at both times), parental education in years (averaged functions. for both parents if paternal information was available, Exclusion criteria were multiple pregnancies, inability otherwise maternal only), postnatal parental smoking (one to speak Danish, impaired hearing or vision likely to or both parents smoked/both were nonsmokers), an index compromise the ability to perform the cognitive tests, and of the quality of postnatal home environment (dichotomised congenital disabilities likely to imply mental retardation (e.g., as normal or suboptimal in the presence of two or more Down’s syndrome, infantile autism). of the following adverse conditions: living with only one biological parent, changes in primary care givers, daycare for 2.2. Exposure Measure. Data on maternal smoking habits in more than 8 hours/day before age 3, 14+ days of separation pregnancy was obtained by the first prenatal interview in the from parents, irregular breakfast, maternal depression, and DNBC carried out at a median of 17 gestational weeks (range maternal/paternal alcohol intake above the official recom- 7–39). The women were asked about their daily and weekly mendations from the Danish National Board of Health at the number of smoked cigarettes, and based on this information time of the data collection), an index of the child’s health sta- the women were categorised in three exposure levels (0, 1– tus (dichotomised as normal or suboptimal in the presence 9, and 10+ cigarettes per day). The interview also comprised of any handicaps, illness, diseases and/or medication with information on smoked pipes, cheroots, and cigars, but none potential influence on test performance), and breast feeding of the participants reported smoking any of these types. (≤1 month, >1 month). To exclude potential undetected impairments, hearing and vision abilities (impaired/normal) 2.3. Outcome Measure. Intelligence was assessed with the of the child were assessed at the follow-up examination, Wechsler Primary and Preschool Scales of Intelligence- as was maternal IQ; two verbal subtests (information and Revised (WPPSI-R) [29, 30], which is one of the most widely vocabulary) from the Wechsler Adult Intelligence Scale [32] used, standardised measures of intelligence for children of (WAIS) were used to assess verbal IQ, and the Raven 3 to 7 years. The full WPPSI-R comprises five verbal and Standard Progressive Matrices [33]providednonverbalIQ. five performance (nonverbal) subtests. To reduce the length The raw scores of each test were standardised based on Journal of Pregnancy 3 the results from the full sample and weighted equally in a shorter education and lower IQs than nonsmokers. There combined score that was restandardised to an IQ scale with a were significantly higher proportions of single mothers and mean of 100 and an SD of 15. suboptimal home conditions in the two smoking categories. Maternal age was obtained from the Danish Civil Reg- Children of smokers had lower birth weights than children istration System as was the sex and age of the child. Birth of nonsmokers and were less likely to have been breast-fed. weight (grams) and gestational age (days) were obtained It should be noted that these differences were unweighted from the Danish Medical Birth Registry. for the stratified sample and thus not representative for the background population. There were slightly more binge 2.5. Data Analysis. All statistical analyses were conducted drinkers (66.3% versus 61.8%) and slightly fewer smokers with Stata 11 (StataCorp LP, College Station, TX, USA). (31.6% versus 35.2%) among participants compared to non- In the LDPS, the higher alcohol categories were over- participants (data not shown), but otherwise no substantial sampled, and consequently all analyses were weighted by or significant differences were observed. sampling probabilities. All statistical tests were two-sided Pairwise, weighted correlations between all core and and declared significant at the 5% level. All estimates are potential confounders (complete case) showed significant accompanied by 95% confidence intervals. The extent of correlations between maternal IQ and parental education missing values on individual variables ranged from 2 (0.1%) (r = 0.47), maternal age and parental education (r = 0.21), on hearing to 59 (3.3%) on prenatal paternal smoking, maternal age and parity (r = 0.42), paternal smoking at the with 36 (2.0%) missing values on maternal BMI and 8 time of interview and postnatal parental smoking (r = 0.42), (0.4%) missing values on full-scale IQ. For the remaining parental education and maternal BMI (r =−0.24), and variables, the extent of missing values was below 0.8 percent. single mother and home environment (r = 0.64). All other Missing values were imputed based on a model in which coefficients were lower than 0.2 and most were close to zero. variables were modelled from other variables considered predictive. All conclusions were maintained when complete 3.1. WPPSI-R. The unadjusted analyses showed a statisti- case analysis was conducted (N = 1,702–1,774). We report cally significant effect of smoking 10+ cigarettes per day, the results from the imputed analyses. All imputations were with a decrement of 3.7 FSIQ points (95% CI: −6.1, −1.2) implemented with the -ice- add-on command, and the built- compared to nonsmokers (Table 2). On the subscales, the in -mi estimate-command of Stata 11. crude effect was larger on PIQ (mean diff. = −4.0, 95% Associations between smoking exposure categories (0, 1– CI: −7.1, −0.9) than VIQ (mean diff. = −2.5, 95% CI = 9, 10+) and the continuous FSIQ, VIQ, and PIQ outcome −4.7, −0.4). In the model including the core confounders scores were estimated using multiple linear regression. (i.e., maternal IQ, parental education, maternal prenatal Parental education, maternal IQ, and maternal alcohol alcohol consumption, the sex and age of the child, and consumption during pregnancy plus the child’s age at testing, tester), the effect estimates were substantially reduced and the child’s sex, and tester were considered core confounders not statistically significant. This pattern was maintained included as covariates in a separate model. The final model, when additional adjustment was made for the potential in addition, included all potential confounders. Birth weight confounders. and gestational age were considered potential mediators of The logistic regression analyses of the IQ outcomes the effects of smoking exposure and not included in these dichotomised at the sample mean minus one SD showed main analyses. neither statistically significant nor systematic differences Additionally, we analysed the three IQ outcomes between the two exposure categories and the reference dichotomised, using the sample mean minus one SD for group, except for an increased risk of low VIQ in the the relevant IQ score (FSIQ, VIQ, or PIQ) as cutoff score 1–9 category compared to the reference group that was for subnormal test performance. Because logistic regressions marginally significant in the unadjusted analysis (OR = 1.78, were used in these analyses, we report odds ratios, with the 95% CI: 1.03, 3.09) (Table 3). After adjustment, this effect category of IQ above the cut-off as the reference group. was slightly reduced and not statistically significant (OR = In supplementary analyses, we analysed raw scores of 1.65, 95% CI: 0.85, 3.22). each individual WPPSI-R subtest with linear regression In the unadjusted analyses of outcomes on the sub- models adjusting for core and all confounders. Potential scales, smoking 10+ cigarettes per day was associated with interactions with smoking exposure were assessed for sex, marginally significant, lower scores on the information parental education, and maternal alcohol consumption in (mean diff. = −0.6, 95% CI = −1.1, 0.0) and the arith- pregnancy. metic subtests (mean diff. = −0.6, 95% CI: −1.2, 0.0) Pairwise correlations between all core and potential (Table 4). Both exposure categories were associated with confounders were tested. For all continuous covariates, significantly poorer performance on the geometric design potential quadratic associations with the IQ outcomes were subtest compared to the reference group, with unadjusted tested. No significant nonlinear associations were found. effect estimates of −2.0 (95% CI: −3.8, −0.1) and −2.6 (95% CI: −4.7, −0.5) for the 1–9 and the 10+ categories, 3. Results respectively. Adjustment for core and all confounders did not change the effect estimates, and the effect of 1–9 Table 1 presents sample characteristics. Women, who cigarettes/day was still marginally significant (mean dif- smoked during pregnancy tended to be younger, have ference = −2.0, 95% CI: −3.8. −0.1) while the effect of 4 Journal of Pregnancy b P 0.800 Total a Average number of cigarettes per day 0 1–9 10+ 1: Sample characteristics across levels of maternal cigarette smoking in pregnancy. Table ) 22.7 (19.6/29.0) 22.5 (19.7/28.4) 22.3 (19.2/28.4) 22.6 (19.6/28.7) 0.810 2 1 month (%) 12.7 15.8 21.1 14.3 0.004 ≤ 1 (%)2+ (%) 32.3 17.1 33.7 14.8 29.2 18.9 32.1 17.0 0 (%) 50.6 51.5 51.9 50.9 values for unweighted data. P Range 10–25 cigarettes/day. Maternal Body Mass Index (kg/m Number of participantsTiming of interview (gestational week)Median number of cigarettesMaternal age (years); mean (SD)Parity 17.0 (13.0/24.0) 31.0 16.0 (4.2) (12.0/24.0) 1276 0 17.0 (13.0/24.0) 30.2 (4.9) 263 17.0 (13.0/24.0) 5 30.5 (4.8) 0.608 243 13 30.8 (4.4) 0.019 1,782 0 Condition/medicine (%)Impaired hearing (%)Impaired vision (%) 2.7 4.6 2.4 3.4 4.2 2.3 5.3 7.0 7.8 3.2 4.8 3.1 0.104 0.229 0.000 Single mother (%)Parental education (years)Suboptimal home environment (%)Maternal IQ; mean (SD)Maternal alcohol consumption in pregnancy (%)Paternal prenatal smoking (%)Parental postnatal smoking (%)Child’s sex (male, %)Age at testing (years)Birth weight (grams)Gestational age (days) 52.8Breast feeding 14.6 13.0 (11.0/16.5) 101.6 (14.8) 9.7 23.4 12.5 16.4 (10.5/15.5) 50.8 25.0 5.2 282.0 (5.1/5.3) 53.4 (268.0/293.0) 3649.0 96.6 (513.1) (14.8) 12.0 (10.5/14.5) 15.9 57.1 282.0 (267.0/292.0) 68.6 13.0 51.0 (11.0/16.0) 3546.4 (492.9) 5.2 (5.1/5.3) 95.2 (14.7) 280.0 37.3 (267.0/294.0) 48.1 0.000 3424.1 (508.8) 281.0 (267.0/293.0) 22.8 5.2 64.9 (5.1/5.3) 76.3 100.0 (15.0) 52.3 0.229 19.2 3603.2 (515.5) 47.7 5.2 0.000 (5.1/5.3) 12.4 0.760 33.8 0.000 32.3 0.551 51.8 0.000 0.000 0.000 0.125 Data are presented asa medians (10/90 percentiles), unless otherwise specified. b Journal of Pregnancy 5

Table 2: Maternal smoking in pregnancy and WPPSI-Ra performance.

Adjusted for potential Crude Adjusted for core confoundersb Average number of cigarettes/day confoundersc Mean difference Mean score Mean difference (95% CI) Mean difference (95% CI) (95% CI) Full-scale IQ 0 106.1 Reference Reference Reference 1–9 103.3 −2.8 (−5.5, −0.1) −1.0 (−3.4, 1.4) −0.9 (−3.6, 1.8) 10+ 102.5 −3.6 (−6.1, −1.2) −1.0 (−3.3, 1.4) −0.3 (−3.1, 2.5) Verbal IQ 0 105.3 Reference Reference Reference 1–9 103.2 −2.1 (−4.4, 0.2) −0.5 (−2.6, 1.6) −0.6 (−2.8, 1.7) 10+ 102.8 −2.5 (−4.7, −0.4) −0.1 (−2.2, 2.0) 0.2 (−2.1, 2.8) Performance IQ 0 105.6 Reference Reference Reference 1–9 102.6 −3.0 (−6.4, 0.5) −1.4 (−4.5, 1.7) −1.0 (−4.7, 2.7) 10+ 101.5 −4.1 (−7.1, −0.9) −1.7 (−4.6, 1.2) −1.0 (−4.6, 2.6) a Wechsler Preschool and Primary Scale of Intelligence-Revised. bParental education, maternal IQ, prenatal maternal alcohol consumption, the child’s sex, age at testing, and tester. cParental education, maternal IQ, prenatal maternal alcohol consumption, the child’s sex, age at testing, and tester, maternal age, parity, maternal marital status, prenatal paternal smoking, postnatal parental smoking, breast feeding, maternal prepregnancy BMI, the child’s sex, age at testing, health status, family/home environment.

Table 3: Maternal smoking in pregnancy and the risk of low IQ.

Crude Adjusted for core confoundersa Adjusted for potential confoundersb Average number of cigarettes/day OR (95% CI) OR (95% CI) OR (95% CI) Full-scale IQ 0 1.00 1.00 1.00 1–9 1.60 (0.94, 2.70) 1.39 (0.78, 2.48) 1.52 (0.78, 2.99) 10+ 1.31 (0.74, 2.31) 1.02 (0.56, 1.85) 1.06 (0.55, 2.19) Verbal IQ 0 1.00 1.00 1.00 1–9 1.78 (1.03, 3.09) 1.49 (0.81, 2.73) 1.65 (0.85, 3.22) 10+ 1.19 (0.64, 2.19) 0.90 (0.48, 1.69) 0.94 (0.50, 1.75) Performance IQ 0 1.00 1.00 1.00 1–9 1.19 (0.69, 2.05) 1.08 (0.59, 1.95) 1.04 (0.54, 2.03) 10+ 1.65 (0.97, 2.79) 1.41 (0.81, 2.48) 1.31 (0.65, 2.62) a Parental education, maternal IQ, prenatal maternal alcohol consumption, the child’s sex, age at testing, and tester. bParental education, maternal IQ, prenatal maternal alcohol consumption, the child’s sex, age at testing, and tester, maternal age, parity, maternal marital status, prenatal paternal smoking, postnatal parental smoking, breast feeding, maternal prepregnancy BMI, the child’s sex, age at testing, health status, family/home environment.

10+ cigarettes per day approached statistical significance exposure and sex, parental education, or prenatal alcohol (mean difference = −2.0, 95% CI: −4.3, 0.2). Adjustment exposure. for birth weight and gestational age did not alter this result. There were no significant associations between exposure 4. Discussion status and the remaining subtests at any level of analysis. This study confirmed previous consistent findings that Additional adjustment for gestational weeks of interview smoking in pregnancy covaries with a range of social and did not alter any of these conclusions. The supplementary family characteristics, including maternal education [34], analyses showed no significant interactions between smoking socioeconomic status [35], maternal age, and marital status 6 Journal of Pregnancy

Table 4: Maternal smoking in pregnancy and WPPSI-Ra subtest raw scores.

Adjusted for potential Unadjusted Adjusted for core confoundersb Avgerage number of cigarettes/day confoundersc Mean difference Mean score Mean difference (95% CI) Mean difference (95% CI) (95% CI) Information 0 19.3 Reference Reference Reference 1–9 19.0 −0.3 (−0.9, 0.3) −0.1 (−0.6, 0.5) −0.1 (−0.7, 0.5) 10+ 18.7 −0.6 (−1.1, 0.0) −0.2 (−0.7, 0.4) −0.1 (−0.7, 0.5) Vocabular y 0 21.1 Reference Reference Reference 1–9 20.6 −0.5 (−1.6, 0.6) 0.0 (−1.0, 1.0) 0.0 (−1.1, 1.0) 10+ 20.8 −0.3 (−1.3, 0.6) 0.6 (−0.3, 1.5) 0.7 (−0.3, 1.8) Arithmetic 0 14.9 Reference Reference Reference 1–9 14.3 −0.6 (−1.1, 0.2) −0.3 (−0.9, 0.2) −0.3 (−0.9, 0.3) 10+ 14.2 −0.7 (−1.2, 0.0) −0.1 (−0.8, 0.5) −0.1 (−0.8, 0.6) Object assembly 0 23.6 Reference Reference Reference 1–9 23.3 −0.3 (−1.2, 0.7) 0.0 (−0.9, 0.8) 0.2 (−0.8, 0.1) 10+ 23.1 −0.5 (−1.3, 0.3) 0.1 (−0.8, 0.9) 0.2 (−0.7, 0.1) Block design 0 24.3 Reference Reference Reference 1–9 23.9 −0.4 (−1.7, 0.8) −0.2 (−1.4, 1.0) 0.0 (−1.4, 1.4) 10+ 23.4 −0.9 (−2.3, 0.5) −0.2 (−1.5, 1.1) 0.1 (−1.5, 1.5) Geometric design 0 37.7 Reference Reference Reference 1–9 35.8 −1.9 (−3.8, −0.1) −1.6 (−3.3, 0.0) −2.0 (−3.8, −0.1) 10+ 35.1 −2.6 (−4.7, −0.5) −1.9 (−4.0, 0.1) −2.0 (−4.3, 0.2) a WPPSI-R Wechsler Preschool and Primary Scale of Intelligence-Revised. bParental education, maternal IQ, maternal prenatal alcohol consumption, the child’s sex, age at testing, and tester. cParental education, maternal IQ, prenatal maternal alcohol consumption, the child’s sex, age at testing, and tester, maternal age, parity, maternal marital status, prenatal paternal smoking, postnatal parental smoking, breast feeding, maternal prepregnancy BMI, the child’s sex, age at testing, health status, family/home environment.

[34, 36, 37], while, in this sample, we did not observe measured by a military draft board test) of 14,722 pairs the commonly reported association between smoking and of full siblings, only one of which had been exposed to alcohol consumption [38, 39]. smoking in utero. There were no differences between exposed We found no evidence of an effect of smoking exposure and unexposed siblings but an increased risk of low test per se on offspring intelligence after adjustment for con- performance for both if the mother had smoked only founders. Thus, significant effects of smoking 10+ cigarettes during her first pregnancy and no difference compared to per day in pregnancy on the three IQ scales disappeared nonexposed controls for either sibling if she had smoked when adjustment was made for parental education, maternal only during her second. These results support no effect IQ, and prenatal maternal alcohol consumption. Adjustment of smoking per se but rather of maternal and familial for additional covariates did not change this conclusion. A characteristics although it should be noted that the study similar pattern applied to scores on some subtests whereas, only used a rather crude, dichotomous outcome measure. in the analyses of dichotomised IQ, no significant differences By contrast, other studies have reported negative effects were found. of maternal smoking in pregnancy on child IQ [16–20, 45– The overall results of the present study are thus in 47]. A series of followups conducted in the Ottawa Prenatal line with previous studies in which statistical adjustment of Prospective Study (OPPS) reported significant effects on potentially confounding factors eliminated an apparent effect Full-Scale and/or Verbal IQ at ages 3–4, 9–12, and 13–16 of smoking exposure on IQ [21–24, 26, 40–43]. Lundberg [20, 45, 48]. At the 13–16 year follow-up, prenatal exposure et al. [44] addressed the causal effect of prenatal smoking to 16 mg nicotine/day or more per day was associated with exposure by comparing the intellectual performance (as an adjusted decrement of 8 FSIQ points on the Wechsler Journal of Pregnancy 7

Intelligence Scale for Children. The sample sizes in the OPPS, than the prevalence of smoking among Danish pregnant however, are generally small, and the 13–16 year follow-up women reported elsewhere [54]—probably reflecting the included a total of 145 individuals with only 36 individuals oversampling of women with a high alcohol intake—only in the 16+ category. three women reported smoking more than 20 cigarettes The divergent results may arise from methodological per day at the time of interview. Because of this sampling differences, one of the most important being confounder design, the sample was not representative for the background adjustment. As indicated by the results of the present study, population. This was accounted for statistically by weighting socioeconomic position, often measured by education or the analyses by the sampling probabilities. income, and maternal IQ seem to be particularly important, Sample selection bias due to sample attrition may be and the lack of control for each may produce spurious effects present in studies of this type. A comparison of partici- of smoking exposure on outcomes such as IQ [26, 40]. pants and nonparticipants did not indicate any substantial Adjustment for education alone has been reported to differences on the available measures. Selection bias on attenuate the association between smoking and outcomes variables on which information on nonparticipants was on IQ by 30%–40% [49]. While most studies did control unavailable, however, cannot be excluded. Another potential socioeconomic factors to some extent, only few studies have source of bias are the exclusion criteria, which arguably taken maternal IQ into account. Of six studies controlling may exclude children of less-advantaged families, potentially maternal IQ [17, 22, 26, 40, 42, 50], two studies reported more vulnerable to the harmful effects of smoking exposure. significant effects of smoking [17, 50]; one of these, however, In this study only a total of 24 children were excluded; 22 who only at age 10 but not at age 5 and in adolescence, and both were twins, and two who were diagnosed with Asperger’s studies only measured verbal IQ using the Peabody Picture Syndrome. However, to our knowledge Asperger’s Syndrome Vocabulary Test. In the present study, separate adjustments has not been associated with tobacco smoking. In fact, it may for either parental education or maternal IQ reduced the be that the isolated effect of smoking exposure per se is more unadjusted effect estimates by 50–60% and resulted in accurately assessed in a sample with fewer competing risks. the association being statistically nonsignificant. Reversely, Further, because smoking, particularly in pregnancy, is removing both confounders from the fully adjusted model associated with social stigmatisation, some degree of under- resulted in statistically significant effects that were aug- reporting may well be present. The resulting misclassification mented by approximately 60%. could potentially bias estimates toward null effects. The present study was based on a large sample size and The age of the children in the study sample implies spe- controlled for important confounders which not all have cific methodological issues. Test reliability and in particular been included simultaneously in most previous studies. In stability are relatively low in children at age from 4 to 5 addition to the already mentioned confounders, we were able [55]. For the WPPSI-R, however, reliability coefficients for to control for paternal prenatal smoking. This variable both the present age group for the IQs are very high (0.90–0.96), accounts for some of the variance related to unmeasured and yet lower for the individual subtests (0.49–0.80) [31]. potentially confounding paternal and familial factors and The fact that controlling maternal education and IQ provides a measure of maternal passive smoking during preg- dilute or deplete the association between maternal smoking nancy. Validations of self-reported smoking in pregnancy by and child IQ does not per se preclude causality between measures of cotinine levels suggest that self-reports may lead exposure and outcome [26], and precautions should be to misclassification, not because women report their own taken in interpreting negative results as evidence against smoking unreliably, but because of passive smoking [51, 52]. harmful effects of smoking in pregnancy on the cognitive Many previous studies have been characterised by small development of the child. sample sizes resulting in low statistical power. Still, even when including a large number of observations, the risk may still be present of insufficient statistical power to detect 5. Conclusions potential effects if these are subtle and if many covariates The adverse effects of smoking on pregnancy outcomes are are included. In this study, the close-to-zero effect estimates indisputable, and animal studies have provided basis for and the reasonably narrow confidence intervals support assuming that nicotine and/or other components of tobacco the validity of the findings of no-association and speak and tobacco smoke may affect human brain development in against a type II error. The effects of maternal smoking a harmful manner [56]. This study, however, did not show on the geometric design subtest could indicate effects on any significant effects on intelligence at age 5 of maternal more specific and sensitive cognitive measures. Although smoking in pregnancy when adjusting statistically for a this finding is a natural aim for further investigation, an number of important confounders not included in many effect confined to a single subtest is of minor relevance in previous studies. this context, the focus being on general intelligence as an outcome. Some limitations of the study should be noted. Well- Acknowledgments educated women are likely to be overrepresented in the study population [53], and smoking exposure may therefore be of This study was supported by the Centers for Disease Control restricted range with respect to heavy exposure. Although and Prevention (CDC), Atlanta, GA, USA. The Danish the proportion of smokers in this sample (28%) is higher National Research Foundation has established the Danish 8 Journal of Pregnancy

Epidemiology Science Centre that initiated and created the [14] P. A. Fried, C. M. O’Connell, and B. Watkinson, “60- and 72- Danish National Birth Cohort. The cohort is furthermore month follow-up of children prenatally exposed to marijuana, a result of a major grant from this Foundation. Additional cigarettes, and alcohol: cognitive and language assessment,” support for the Danish National Birth Cohort is obtained Journal of Developmental and Behavioral Pediatrics, vol. 13, no. from the Pharmacy Foundation, the Egmont Foundation, 6, pp. 383–391, 1992. the March of Dimes Birth Defects Foundation, the Augusti- [15] J. S. McCartney, P. A. Fried, and B. Watkinson, “Central auditory processing in school-age children prenatally exposed nus Foundation, and the Health Foundation. The authors to cigarette smoke,” Neurotoxicology and Teratology, vol. 16, would like to thank all the participants for their time, as no. 3, pp. 269–276, 1994. well as Dr. Poul Thorsen for his contribution to initiating the [16] J. Makin, P. A. Fried, and B. Watkinson, “A comparison Lifestyle During Pregnancy Study. of active and passive smoking during pregnancy: long-term effects,” Neurotoxicology and Teratology,vol.13,no.1,pp.5– References 12, 1991. [17] D. L. Olds, C. R. Henderson, and R. Tatelbaum, “Intellectual [1]J.M.Rogers,“Tobaccoandpregnancy,”Reproductive Toxicol- impairment in children of women who smoke cigarettes ogy, vol. 28, no. 2, pp. 152–160, 2009. during pregnancy,” Pediatrics, vol. 93, no. 2, pp. 221–227, [2] I. Berlin, C. Heilbronner, S. Georgieu, C. Meier, and O. 1994. Spreux-Varoquaux, “Newborns’ cord blood plasma cotinine [18] C. D. Drews, C. C. Murphy, M. Yeargin-Allsopp, and P. concentrations are similar to that of their delivering smoking Decoufle,´ “The relationship between idiopathic mental retar- mothers,” Drug and Alcohol Dependence, vol. 107, no. 2-3, pp. dation and maternal smoking during pregnancy,” Pediatrics, 250–252, 2010. vol. 97, no. 4, pp. 547–553, 1996. [3]J.Blood-SiegfriedandE.K.Rende,“Thelong-termeffects [19] E. L. Mortensen, K. F. Michaelsen, S. A. Sanders, and J. of prenatal nicotine exposure on neurologic development,” M. Reinisch, “A dose-response relationship between maternal Journal of Midwifery and Women’s Health,vol.55,no.2,pp. smoking during late pregnancy and adult intelligence in male 143–152, 2003. offspring,” Paediatric and Perinatal Epidemiology, vol. 19, no. [4] J. Leonardi-Bee, A. Smyth, J. Britton, and T. Coleman, 1, pp. 4–11, 2005. “Environmental tobacco smoke and fetal health: systematic [20] P. A. Fried, B. Watkinson, and R. Gray, “Differential effects review and meta-analysis,” Archives of Disease in Childhood, on cognitive functioning in 13- to 16-year-olds prenatally vol. 93, no. 5, pp. F351–F361, 2008. exposed to cigarettes and marihuana,” Neurotoxicology and [5] S. E. Vielwerth, R. B. Jensen, T. Larsen, and G. Greisen, “The Teratology, vol. 25, no. 4, pp. 427–436, 2003. impact of maternal smoking on fetal and infant growth,” Early [21] D. M. Fergusson and M. Lloyd, “Smoking during pregnancy Human Development, vol. 83, no. 8, pp. 491–495, 2007. and its effects on child cognitive ability from the ages of 8 to [6] M. Shi, G. L. Wehby, and J. C. Murray, “Review on genetic 12 years,” Paediatric and Perinatal Epidemiology, vol. 5, no. 2, variants and maternal smoking in the etiology of oral clefts pp. 189–200, 1991. and other birth defects,” Birth Defects Research C, vol. 84, no. [22] N. Breslau, N. Paneth, V. C. Lucia, and R. Paneth-Pollak, 1, pp. 16–29, 2008. “Maternal smoking during pregnancy and offspring IQ,” [7] K. Wisborg, U. Kesmodel, T. B. Henriksen, S. F. Olsen, and N. International Journal of Epidemiology, vol. 34, no. 5, pp. 1047– J. Secher, “Exposure to tobacco smoke in utero and the risk of 1053, 2005. stillbirth and death in the first year of life,” American Journal [23] S. E. Gilman, H. Gardener, and S. L. Buka, “Maternal of Epidemiology, vol. 154, no. 4, pp. 322–327, 2001. smoking during pregnancy and children’s cognitive and [8] J. Nicholl and A. O’Cathain, “Antenatal smoking, postnatal physical development: a causal risk factor?” American Journal passive smoking, and the Sudden Infant Death Syndrome,” in of Epidemiology, vol. 168, no. 5, pp. 522–531, 2008. Effects of Smoking on the Fetus, Neonate, and Child,D.Poswillo [24] C. MacArthur, E. George Knox, and R. J. Lancashire, “Effects and E. Alberman, Eds., pp. 138–149, Oxford University Press, at age nine of maternal smoking in pregnancy: experimental Oxford, UK, 1992. and observational findings,” British Journal of Obstetrics and [9] E. A. Mitchell and J. Milerad, “Smoking and the sudden infant Gynaecology, vol. 108, no. 1, pp. 67–73, 2001. death syndrome,” Reviews on Environmental Health, vol. 21, [25] B. Eskenazi and L. S. Trupin, “Passive and active maternal no. 2, pp. 81–103, 2006. smoking during pregnancy, as measured by serum cotinine, [10] A. K. Shea and M. Steiner, “Cigarette smoking during and postnatal smoke exposure. II. Effects on neurodevelop- pregnancy,” Nicotine and Tobacco Research, vol. 10, no. 2, pp. ment at age 5 years,” American Journal of Epidemiology, vol. 267–278, 2008. 142, supplement 9, pp. S19–S29, 1995. [11] S. Milberger, J. Biederman, S. V. Faraone, L. Chen, and J. [26] G. D. Batty, G. Der, and I. J. Deary, “Effect of maternal Jones, “Is maternal smoking during pregnancy a risk factor for smoking during pregnancy on offspring’s cognitive ability: attention deficit hyperactivity disorder in children?” American empirical evidence for complete confounding in the US Journal of Psychiatry, vol. 153, no. 9, pp. 1138–1142, 1996. national longitudinal survey of youth,” Pediatrics, vol. 118, no. [12] S. Milberger, J. Biederman, S. V. Faraone, and J. Jones, “Fur- 3, pp. 943–950, 2006. ther evidence of an association between maternal smoking [27] U. S. Kesmodel, M. Underbjerg, T. R. Kilburn et al., “Lifestyle during pregnancy and attention deficit hyperactivity disorder: during pregnancy: neurodevelopmental effects at 5 years of findings from a high-risk sample of siblings,” Journal of age. The design and implementation of a prospective follow- Clinical Child Psychology, vol. 27, no. 3, pp. 352–358, 1998. up study,” Scandinavian Journal of Public Health,vol.38,no.2, [13] A. Rodriguez and G. Bohlin, “Are maternal smoking and pp. 208–219, 2010. stress during pregnancy related to ADHD symptoms in [28] J. Olsen, M. Melbye, S. F. Olsen et al., “The Danish National children?” Journal of Child Psychology and Psychiatry and Birth Cohort—its background, structure and aim,” Scandina- Allied Disciplines, vol. 46, no. 3, pp. 246–254, 2005. vian Journal of Public Health, vol. 29, no. 4, pp. 300–307, 2001. Journal of Pregnancy 9

[29] D. Wechsler, Manual For the Wechsler Preschool and Primary and marijuana,” Neurotoxicology and Teratology,vol.19,no.3, Scale of Intelligence—Revised, The Psychological Corporation, pp. 171–183, 1997. Kent, UK, 1990. [47] J. Julvez, N. Ribas-Fito,´ M. Torrent, M. Forns, R. Garcia- [30] D. Wechsler, Manual For the Wechsler Preschool and Pri- Esteban, and J. Sunyer, “Maternal smoking habits and cogni- mary Scale of Intelligence—Revised, Dansk Psykologisk Forlag, tive development of children at age 4 years in a population- Copenhagen, Denmark, 2006, Danish Version. based birth cohort,” International Journal of Epidemiology, vol. [31] D. Wechsler, Manual For the Wechsler Preschool and Primary 36, no. 4, pp. 825–832, 2007. Scale of Intelligence—Revised, Psykologiforlaget¨ AB, Stock- [48] P. A. Fried, B. Watkinson, and R. Gray, “Differential effects on holm, Sweden, 1999, Swedish Version. cognitive functioning in 9- to 12-year olds prenatally exposed [32] D. Wechsler, Manual For the Wechsler Adult Intelligence Scale, to cigarettes and marihuana,” Neurotoxicology and Teratology, The Psychological Corporation, New York, NY, USA, 1955. vol. 20, no. 3, pp. 293–306, 1998. [33] J. C. Raven, Standard Progressive Matrices, Oxford Psycholo- [49] J. M. Braun, J. L. Daniels, A. Kalkbrenner, J. Zimmerman, and gists Press, Oxford, UK, 1958. J. S. Nicholas, “The effect of maternal smoking during preg- [34] C. R. Gale, W. Johnson, I. J. Deary, I. Schoon, and G. D. Batty, nancy on intellectual disabilities among 8-year-old children,” “Intelligence in girls and their subsequent smoking behaviour Paediatric and Perinatal Epidemiology, vol. 23, no. 5, pp. 482– as mothers: the 1958 National Child Development Study 491, 2009. and the 1970 British Cohort Study,” International Journal of [50] K. E. Bauman, R. L. Flewelling, and J. LaPrelle, “Parental Epidemiology, vol. 38, no. 1, pp. 173–181, 2009. cigarette smoking and cognitive performance of children,” [35] P. Rantakallio, “A follow-up study up to the age of 14 of Health Psychology, vol. 10, no. 4, pp. 282–288, 1991. children whose mothers smoked during pregnancy,” Acta [51] L. George, F. Granath, A. L. Johansson, and S. Cnattingius, Paediatrica Scandinavica, vol. 72, no. 5, pp. 747–753, 1983. “Self-reported nicotine exposure and plasma levels of cotinine [36]M.D.Cornelius,D.Geva,N.L.Day,J.R.Cornelius,andP. in early and late pregnancy,” Acta Obstetricia et Gynecologica M. Taylor, “Patterns and covariates of tobacco use in a recent Scandinavica, vol. 85, no. 11, pp. 1331–1337, 2006. sample of pregnant teenagers,” Journal of Adolescent Health, [52] G. K. Swamy, K. L. B. Reddick, R. J. N. Brouwer, K. I. Pollak, vol. 15, no. 7, pp. 528–535, 1994. and E. R. Myers, “Smoking prevalence in early pregnancy: [37] H. Graham, S. S. Hawkins, and C. Law, “Lifecourse influences comparison of self-report and anonymous urine cotinine on women’s smoking before, during and after pregnancy,” testing,” Journal of Maternal-Fetal and Neonatal Medicine, vol. Social Science and Medicine, vol. 70, no. 4, pp. 582–587, 2010. 24, no. 1, pp. 86–90, 2010. [38] M. M. Morales-Suarez-Varela,´ C. Bille, K. Christensen, and [53] E. A. Nohr, M. Frydenberg, T. B. Henriksen, and J. Olsen, J. Olsen, “Smoking habits, nicotine use, and congenital “Does low participation in cohort studies induce bias?” malformations,” Obstetrics and Gynecology, vol. 107, no. 1, pp. Epidemiology, vol. 17, no. 4, pp. 413–418, 2006. 51–57, 2006. [54] K. Egebjerg Jensen, A. Jensen, B. Nøhr, and S. Kruger Kjær, [39] E. J. Waterson and I. M. Murray-Lyon, “Drinking and smoking “Do pregnant women still smoke? A study of smoking patterns patterns amongst women attending an antenatal clinic—II. among 261,029 primiparous women in Denmark 1997–2005,” During pregnancy,” Alcohol and Alcoholism,vol.24,no.2,pp. Acta Obstetricia et Gynecologica Scandinavica, vol. 87, no. 7, 163–173, 1989. pp. 760–767, 2008. [40] P. A. Baghurst, S. L. Tong, A. Woodward, and A. J. McMichael, [55] J. M. Schuerger and A. C. Witt, “The temporal stability of “Effects of maternal smoking upon neuropsychological devel- individually tested intelligence,” Journal of Clinical Psychology, opment in early childhood: importance of taking account of vol. 45, pp. 294–302, 1989. social and environmental factors,” Paediatric and Perinatal [56] J. B. Dwyer, R. S. Broide, and F. M. Leslie, “Nicotine and brain Epidemiology, vol. 6, no. 4, pp. 403–415, 1992. development,” Birth Defects Research C, vol. 84, no. 1, pp. 30– [41] R. McGee and W. R. Stanton, “Smoking in pregnancy and 44, 2008. child development to age 9 years,” Journal of Paediatrics and Child Health, vol. 30, no. 3, pp. 263–268, 1994. [42] N. Trasti, T. Vik, G. Jacobsen, and L. S. Bakketeig, “Smoking in pregnancy and children’s mental and motor development at age 1 and 5 years,” Early Human Development,vol.55,no.2, pp. 137–147, 1999. [43] R. Alati, J. Macleod, M. Hickman et al., “Intrauterine exposure to alcohol and tobacco use and childhood IQ: findings from aparental-offspring comparison within the avon longitudinal study of parents and children,” Pediatric Research, vol. 64, no. 6, pp. 659–666, 2008. [44] F. Lundberg, S. Cnattingius, B. D’Onofrio et al., “Maternal smoking during pregnancy and intellectual performance in young adult Swedish male offspring,” Paediatric and Perinatal Epidemiology, vol. 24, no. 1, pp. 79–87, 2010. [45] P. A. Fried and B. Watkinson, “36- and 48-month neu- robehavioral follow-up of children prenatally exposed to marijuana, cigarettes, and alcohol,” Journal of Developmental and Behavioral Pediatrics, vol. 11, no. 2, pp. 49–58, 1990. [46] P. A. Fried, B. Watkinson, and L. S. Siegel, “Reading and language in 9- to 12-year olds prenatally exposed to cigarettes Hindawi Publishing Corporation Journal of Pregnancy Volume 2012, Article ID 470247, 33 pages doi:10.1155/2012/470247

Review Article Preventing Long-Term Risk of Obesity for Two Generations: Prenatal Physical Activity Is Part of the Puzzle

Stephanie-May Ruchat1 and Michelle F. Mottola1, 2, 3

1 R. Samuel McLaughlin Foundation Exercise and Pregnancy Laboratory, School of Kinesiology, The University of Western Ontario, London, ON, Canada N6A 3K7 2 Department of Anatomy and Cell Biology, The University of Western Ontario, London, ON, Canada N6A 3K7 3 Children’s Health Research Institute, The University of Western Ontario, London, ON, Canada N6A 3K7

Correspondence should be addressed to Stephanie-May Ruchat, steph [email protected]

Received 22 August 2012; Accepted 4 October 2012

Academic Editor: Riitta Luoto

Copyright © 2012 S.-M. Ruchat and M. F. Mottola. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. The period surrounding pregnancy has been identified as a risk period for overweight/obesity in both mother and child because of excessive gestational weight gain (GWG). The promotion of a healthy GWG is therefore of paramount importance in the context of the prevention of obesity in the current and next generations. Objective. To provide a comprehensive overview of the effect of prenatal physical activity interventions, alone or in combination with nutritional counselling, on GWG and to address whether preventing excessive GWG decreases the incidence of infant high birth weight and/or postpartum weight retention. Method. A search of the PubMed database was conducted to identify all relevant studies. Nineteen studies were included in this review: 13 interventions combining physical activity, nutrition, and GWG counselling and 6 interventions including physical activity alone. Results. Prenatal lifestyle interventions promoting healthy eating and physical activity habits appear to be the most effective approach to prevent excessive GWG. Achievement of appropriate GWG may also decrease the incidence of high infant birth weight and postpartum weight retention. Conclusion. Healthy eating habits during pregnancy, combined with an active lifestyle, may be important elements in the prevention of long-term risk of obesity for two generations.

1. Introduction outcomes [5]. Large weight gain has been linked to postpartum weight retention, which in turn has been associated There has been a significant increase over the past few with long-term risk of maternal obesity [6–11]. Excessive decades in the prevalence of maternal overweight (body mass GWG has also been associated with adverse infant outcomes, 2 2 index, BMI ≥ 25 kg/m ) and obesity (BMI ≥ 30 kg/m ). In such as an increased risk of being heavier and fatter at birth the United States, data from the “Pregnancy Risk Assessment [8, 12–15], which increases the infant’s risk of becoming Monitoring System” in nine states indicated that pre-pre- overweight/obese later in life [8, 13, 16, 17]. These findings gnancy obesity increased from 13% to 22% between 1993 suggest that influences occurring very early in life are and 2002 [1]. Worldwide population estimates of pre-pre- contributing to the obesity onset. The promotion of healthy gnancy overweight is approximately 34% [2, 3] and that of weight gain during pregnancy is therefore of paramount pre-pregnancy obesity is 25% [4]. Women who enter pre- importance to reduce the risk of long-term obesity and gnancy either normal weight or overweight/obese are at an associated comorbidities in both the mother and infant. increased risk of developing obesity or increasing BMI categories because of excessive gestational weight gain (GWG). 1.2. Gestational Weight Gain Guidelines. To optimize both maternal and infant outcomes, recommendations regarding 1.1. Excessive Gestational Weight Gain: A Link with Obesity healthy weight gain have been developed. The Institute Risk in Mother and Child. Excessive GWG has been associ- of Medicine (IOM) published in 2009 a revision of the ated with short- and long-term adverse maternal and infant 1990 guidelines for how much weight a woman should 2 Journal of Pregnancy

Table 1: Gestational weight gain recommendations based on the 1990 and 2009 Institute of Medicine guidelines [5, 79].

IOM 1990 IOM 2009 Pre-pregnancy BMI Recommendedb range of Pre-pregnancy BMI Meana rate of weight gain in the Recommendedb range category total weight gain category 2nd and 3rd trimester of total weight gain BMI < 19.8 kg/m2 BMI < 18.5 kg/m2 12.5–18.0 kg 0.5 kg/week 12.5–18.0 kg (low) underweight BMI 19.8–26.0 kg/m2 BMI 18.5–24.9 kg/m2 11.5–16.0 kg 0.4 kg/week 11.5–16 kg (normal) normal weight BMI 26.1–29.0 kg/m2 BMI 25.0–29.9 kg/m2 7.0–11.5 kg 0.3 kg/week 7.0–11.5 kg (high) overweight BMI > 29.0 kg/m2 BMI ≥ 30c kg/m2 at least 6.0 kg 0.2 kg/week 5.0–9.0 kg (obese) obese a Rounded values. bCalculations assume a total of 0.5–2.0 kg weight gain in the first trimester. cA narrower range of weight gain may be advised for women with a pre-pregnancy BMI of 35.0 kg/m2 or greater. Individualized advice is recommended for these women. gain during pregnancy [5](Table 1). Using the 1990 IOM per week, starting with 15 minutes of aerobic activity at a recommendations, it has been reported that up to 40% of specific target heart rate intensity based on age and increase normal weight women and 60% of overweight/obese women time slowly to a maximum of 30 minutes per exercise session. gained excessive weight during pregnancy [11, 13, 18]. Allaerobicactivityshouldbeprecededby10to15minutes However, a study that quantified how the 2009 revisions of of warm-up and followed by 10 to 15 minutes of cool-down. the 1990 IOM GWG guidelines changed BMI-specific GWG Appropriate exercise intensity may be monitored by using adherence categories showed that 17.1% of 1990 appropriate target heart-rate zones, the Borg-scale (rating of perceived weight gaining women would now be classified as excessive exertion, RPE), or the “talk test” [23]. Heart-rate zones gainers, given the new guidelines [19]. Health care providers that are provided in the guidelines correspond to moderate- must therefore determine the women’s pre-pregnancy BMI intensity exercise (i.e., 60–80% of maximal aerobic capacity, and provide BMI-specific GWG recommendations in accor- VO2max)[23]. dance with current guidelines, before pregnancy if possible. The recent opinion statement from the SOGC [34]on Further, the importance of healthy weight gain and healthy obesity during pregnancy strongly suggests that regular exer- lifestyle habits to optimize pregnancy outcomes should also cise during pregnancy may help to reduce the risk of medi- be emphasized throughout pregnancy. cal complications associated with maternal obesity. Over- weight and obese women can participate in exercise, if they 1.3. Physical Activity during Pregnancy. Research over the have no contraindications to being physically active. They past 25 years has demonstrated that healthy women with should therefore be medically pre-screened and consult their low-risk pregnancies can safely participate in physical activ- health care provider before engaging in an exercise program. ities without affecting fetal growth and development [20– Target heart rate zones developed for normal weight pregn- 23]. In fact, maternal physical activity has been identified ant women [23]maybetoodifficult for overweight and as an important component of a healthy pregnancy and is obese women to obtain. Target heart rate zones for these beneficial to the mother and fetus [24, 25]. A sedentary women were thus developed and validated at a much lower lifestyle, on the other hand, may put the mother and fetus intensity but high enough to achieve aerobic benefits [35]. at risk for diseases through altered maternal pregnancy The American College of Sports Medicine (ACSM) suggested adaptations, such as excessive GWG, gestational diabetes, that previously sedentary overweight and obese pregnant or gestational hypertension [24, 25]. Observational studies women should initiate an aerobic exercise program at an showed that physical activity during pregnancy reduces the intensity equivalent to 20% to 39% of reserve aerobic risk of glucose intolerance, gestational diabetes [26–29], and capacity (VO2reserve)[36]. The developed and validated target excessive GWG [30, 31]. heart rate zones based on age are 102 to 124 beats per minute Active promotion of physical activity for pregnant (bpm)foroverweightandobesewomen20to29yearsofage women is strongly recommended in the absence of either and 101 to 120 bpm for those aged 30 to 39 years [35]. medical or obstetric contraindications by professional orga- Although maternal physical activity has clear health nizations [23, 32, 33]. The joint Society of Obstetricians and benefits on pregnancy outcomes and is strongly recom- Gynaecologists of Canada (SOGC) and Canadian Society mended by professional organizations, a large proportion for Exercise Physiology (CSEP) Clinical Practice Guidelines of women do not meet the recommendations for physical provide detailed recommendations about the frequency, activity during pregnancy, based on information collected by intensity, time, and type of exercise, following the FITT interviews or questionnaires [37–40]. Similarly, studies using principle for exercise prescription [23]. Low risk medically pedometers to assess the level of physical activity during pre-screened pregnant women should exercise 3 to 4 times pregnancy found that many women have sedentary to low Journal of Pregnancy 3 physical activity levels during pregnancy and that the levels 2. Results of activity declined as their pregnancy progressed [41, 42]. Previous studies conducted in our laboratory showed that Using the above mentioned keywords for searches of the overweight and obese pregnant women are mostly inactive, PubMed database, 256 studies were detected. After applying accumulating 6, 100 ± 1, 700 steps per day [43, 44]. However, the inclusion/exclusion criteria, 16 studies remained. An when a 40-min structured walk (4, 300 ± 680 steps per 40- additional 3 studies were identified from the literature min session) was added to their usual daily activities, these obtained. Among the intervention studies that combined women were taking 10,000 steps or more [43, 44]. As walking physical activity, nutrition and recommendations regarding is the most reported activity during pregnancy [37–39, 45], healthy GWG (Table 2), 9/13 (69%) were successful at providing step-count goals may encourage pregnant women decreasing mean GWG and/or preventing excessive GWG to be more active. The recommendation would be to add based on the 1990 (Table 2(a)) or 2009 (Table 2(b)) IOM 3,000–4,000 steps in a 30–40-min walk (moderate-intensity) guidelines [44, 55–62] and 4/13 (31%) were unsuccessful to the usual number of daily steps. [63–66]. Among intervention studies that included a physical Taken together, these findings showed that a large pro- activity component alone (Table 3), 2/6 (33%) were only par- portion of women are physically inactive during pregnancy. tially successful at reducing mean GWG and/or preventing This may be contributing to excessive GWG, high infant excessive GWG based on the IOM 1990 (Table 3(a)) or 2009 weight at birth, and long-term obesity risk in both the (Table 3(b)) guidelines [67, 68] and 4/6 (67%) were unsuc- mother and child. Pregnancy may be a teachable time cessful [69–72]. when women are motivated to adopt healthy behaviors and may therefore be the perfect opportunity to introduce life- 2.1. Effect of Prenatal Lifestyle Interventions on style modification strategies promoting physical activity and Gestational Weight Gain healthy eating habits in order to achieve a healthy weight gain and pregnancy outcomes. 2.1.1. Successful Interventions at Reducing Mean Gestational Several systematic reviews and meta-analyses assessing Weight Gain. Asbee at al. [55](Table 2(a)) examined the the effect of prenatal lifestyle interventions on GWG have effect of a prenatal intervention on GWG in women of been published in 2010–2012 [46–54] and their conclusions various pre-pregnancy BMI categories (normal weight, over- were that prenatal lifestyle interventions had a modest effect weight, and obese women). This randomised controlled trial on the decrease in GWG. However, these papers included (RCT) included detailed recommendations regarding dietary only randomised control trials, considered mean GWG as intake, physical activity, and appropriate GWG based on the the main outcome, discussed only briefly the adherence to women’s pre-pregnancy BMI. The authors found lower GWG the IOM GWG guidelines and if so, did not differentiate in the intervention group (16.2 ± 7.0 kg) compared to the findings between the 1990 and 2009 IOM recommendations. control group (13.0±5.7, P = 0.01), regardless of BMI status. Furthermore, few of these reviews discuss the effect that Although the participants received specific recommenda- prenatal interventions may have on the prevention of infant tions about dietary intake and physical activity, compliance high birth weight and none on the prevention of maternal to these recommendations was not reported. In addition, weight retention after delivery, which we believe to be no information was given regarding physical activity levels extremely important in the context of the prevention of and eating habits of the control group. Two other success- obesity in the current and next generations. ful interventions included only women who were overweight The objective of this integrative review was therefore to and/or obese prior to pregnancy. Shirazian et al. [57] provide a comprehensive overview of prenatal physical activ- (Table 2(b)) conducted a prenatal intervention that included ity interventions, alone or in combination with a nutritional six seminars and one-on-one counseling sessions (n ≥ intervention, whose main objective was to decrease mean 5) aimed at promoting healthy eating and encouraging GWG and/or prevent excessive GWG and to raise method- walking (food diaries and pedometers were provided to the ological issues and relevant points to discuss in order to high- participants) and educating the women on obesity during light the effective/missing aspects of the reviewed interven- pregnancy. The intervention group gained less weight than tions. Further, our objective was also to address important the control group (8.1 ± 7.4kgversus15.4 ± 7.5kg,resp.; issues, such as whether preventing excessive GWG decreases P = 0.003). Unfortunately, eating and physical activity the incidence of infant high birth weight, or prevents post- habits of the women during the intervention were not partum weight retention. Searches of the PubMed database reported, although the participants received tools (diaries were performed using the following keywords: “exercise” and pedometers) to monitor their daily dietary intake and OR “physical activity” OR “intervention” OR “lifestyle” physical activity levels. Further, physical activity levels and AND “pregnant” AND “weight gain”. Relevant references eating habits of the control group were not reported. Fin- in the literature obtained were further reviewed to identify ally, the study of Nascimento et al. [68](Table 3(b)) included additional studies on this topic. All types of intervention weekly supervised aerobic dance classes of moderate- (randomised controlled trials, historical cohort, single-arm intensity, combined with recommendations about weekly study) published in English were considered. Studies includ- physical activity level and healthy GWG. The authors ing gestational diabetic women were excluded. This review reported lower mean GWG in the intervention group (10.0± includes references published between January 2000 and July 1.7 kg) compared with the control group (16.4 ± 3.9kg, 2012. P = 0.001) but only in overweight women (17.5% of 4 Journal of Pregnancy NS value NS NS P 686 ± 3(3%) 37 (35%) 4,000 g 2,500 g = > < birth weight. = n 3,686 n 523 ± 2(2%) 31 (30%) 4,000 g 2,500 g = > < = n 3,741 n 0.05 value Control Intervention > P 6.7 ± 6.1 At 6 weeks PP 8.5 ± 7.4 At 6 weeks PP vity on gestational weight gain, maternal weight retention, and infant NS value Control Intervention NS 0.29 P 29: 29: 6.4 0.27 0.24 > ≤ ± ± ± At delivery 12.0 0.62 Weekly GWG BMI BMI 0.44 Adherence to 1990 IOM NA 29: 29: 8.3 0.32 0.30 > ≤ ± ± ± 0.63 0.44 Weekly GWG BMI BMI Adherence to 1990 IOM NA At delivery 13.2 (a) Prenatal intervention studies using the 1990 IOM gestational weight gain recommendations Community-based intervention for specific population (Aboriginal Cree) Based on the social learning theory Nutrition Pamphlets about nutritional choices, supermarket tours and cooking demonstrations, individual counseling Physical activity Exercise/walking groups Other Advice to stay within GWG recommendations Delivered by dieticians and health workers Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care ects of prenatal intervention studies combining nutrition and physical acti ff 2: E 5.86 6.45 6.9 6.29 6.85 7.1 ± ± ± ± ± ± Table 219 107 23.8 29.6 73% nulliparous 18.5 112 24.3 30.8 31% nulliparous 17.1 ] ======64 I I C Reference, sample characteristics Gray-Donald et al. [ Design: Time-series control trial N SES NA Prepreg BMI C Parity C Duration of study until delivery Country Canada Ethnicity Cree women Age C Gestational age at recruitment C I I I Journal of Pregnancy 5 NS NS value NS NS NS NS P 0 = 1(3%) 1(4%) n 4 (13%) 2,500 g 4,000 g 4,000 g 3,133 3,283 2,500 g = = < > > = < n n n 0 0 = = 2(9%) n n 3 (10%) 3,226 3,349 2,500 g 2,500 g 4,000 g 4, 000 g = = < < > > n n NS NS value Control Intervention P 5.4 5.6 7wPP 7wPP ± ± ± ± 3.6 4.4 At 8 At 8 4.5 7.0 7wPP 7wPP ± ± ± ± 0.3 6.2 At 8 At 8 0.05 value Control Intervention NS NS 0.09 < P (a) Continued 7.1 NS NA NA7.1 — NA7.2 NA — ± ± ± At 36–40 weeks 15.4 At 36–40 weeks 13.6 Adherence to 1990 IOM 67% did not exceed Adherence to 1990 IOM 41% did not exceed At 36–40 weeks 14.5 5.4 4.8 6.2 ± ± ± At 36–40 weeks 16.4 At 36–40 weeks 10.1 Adherence to 1990 IOM 42% did not exceed Adherence to 1990 IOM 68% did not exceed At 36–40 weeks 13.8 ff Nutrition Bi-weekly newsletters on healthy eating Physical activity Bi-weekly newsletters on healthy exercise habits, increasing walking and developing a more active lifestyle Other Written and oral info about appropriate GWG, personalized graphs showing GWG Delivered by trained sta Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care ] 58 2.0 7.2 1.9 6.0 ± ± ± ± 4.8 3.1 ± ± 110 53 31 22.5 22 34.1 57 30 22.8 27 31.4 ======I I I Parity 47% nulliparous SES low income Gestational age at recruitment 14.5 Duration of study NA C Reference, sample characteristics Polley et al. [ Normal weight group C Overweight group C Design: RCT N Country United States Ethnicity 39% black 61% white, Age 25.5 Prepreg BMI C Prepreg BMI C I I 6 Journal of Pregnancy NS value NS NS P 2(1%) 32 (18%) 2,500 g 4,000 g 3,610 (2,475– 5,345) = < > = n n 3(1%) 83 (22%) 4,000 g 2,500 g 3,621 (2,060– 5,486) = > < = n n value Control Intervention 0.05 0.04 0.14 > P 4.8 ± SES 2.27 kg OW 25% ≥ At 1 year 0.6 NW 37% Weight retention Sub-group, Low 5.6 ± SES 2.27 kg OW 55% ≥ At 1 year 1.3 NW 40% Weight retention Sub-group, Low value Control Intervention 0.04 0.09 0.30 0.05 P (a) Continued 4.5 ± At delivery 14.1 Adherence to 1990 IOM 59% did not exceed Sub-group, Low SES Adherence to 1990 IOM NW 71% OW 56% did not exceed 4.7 ± Adherence to 1990 IOM 55% did not exceed Sub-group, Low SES Adherence to 1990 IOM NW 55% OW 28% did not exceed At delivery 14.8 Nutrition Tips and newsletters about healthy eating, “health checkbook” to help self-monitoring diet Physical activity Tips and newsletters about PA during pregnancy Other Guidance about self-monitoring of GWG, newsletters about GWG Delivered by healthcare providers Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care ] 61 560 381 94% 20–40 yr 41% 43% low 179 94% 20–40 yr 37% low 41% nulliparous ======27 weeks I I nulliparous I income I C SES C Reference, sample characteristics Olson et al. [ Design Historical cohort N Country United States Ethnicity 96% white Age: C Parity C income Prepreg BMI 75% normal weight in early pregnancy Gestational age at recruitment < Duration of study NA Journal of Pregnancy 7 value 0.006 P 0 = n 4,000 g ≥ 8 (15%) 4,000 g = ≥ n value Control Intervention P NA NA — value Control Intervention 0.77 0.053 P (a) Continued 5.4 ± At 36-37 weeks 14.6 Adherence to 1990 IOM 54% did not exceed 4.1 ± Adherence to 1990 IOM 70% did not exceed At 36-37 weeks 14.3 Nutrition 4 individual counseling sessions on healthy diet Physical activity 5 counseling sessions to increase LTPA and design an individual activity plan. Possibility to join group exercise sessions once a week for 45–60 min Other Information on GWG recommendations Delivered by nurses Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care ] 63 4.1 2.1 4.5 3.9 ± ± ± ± 105 56 28.8 36% 22.3 27.6 57% 23.7 49 ======Design non RCT N I Basic/secondary education I Country Finland Ethnicity Finish Age C C Reference, sample characteristics Kinnunen et al. [ Parity 100% nulliparous SES C Gestational age at recruitment 8-9 weeks Duration of study until 37 weeks of gestation Basic/secondary education Prepreg BMI C I I 8 Journal of Pregnancy value P value Control Intervention P NA NA — NA NA — value Control Intervention 0.21 0.01 P (a) Continued 5.7 ± At delivery 13.0 Adherence to 1990 IOM 61% did not exceed 7.0 ± Adherence to 1990 IOM 49% did not exceed At delivery 16.2 Nutrition Nutrition plan containing 40% carbohydrate 30% protein 30% lipid Physical activity Recommendations Frequency 3–5/week Intensity moderate-intensity exercise Other Information about GWG guidelines Delivered by dieticians Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care ] 55 5.0 5.1 3.6 6.0 6.0 3.6 ± ± ± ± ± ± 100 43 55% Hispanic 26.4 44% 25.6 13.6 65% high school 57 58% Hispanic 26.7 68% high school 25.5 13.7 46% nulliparous < < ======I I nulliparous I C or I SES C Reference, sample characteristics Asbee et al. [ Design RCT N Parity C Gestational age at recruitment C Duration of study until delivery Prepreg BMI C Country United States Ethnicity mixed C Age C or I I I Journal of Pregnancy 9 value NS NS NS 0.07 0.05 P 600 ± 2,500 g 4,000– OB 0% 4,500 g OB 26% OW 0% < OW 3.2% 3,564 500 ± 2,500 g 4,000– 4,500 g OB 3% OW 4% < OW 18% OB 13.3% 3,590 — value Control Intervention P 5.6 ± 2.2 At 8 w PP body weight of pre-pregnancy 53% within 2.0 kg NA — value Control Intervention — P (a) Continued ∗ 4.1 ± GWG during 6.8 Adherence to 1990 IOM 80% ∗ the intervention NA 25– Nutrition Individualized nutrition plan with total energy intake of approx 2000 kcal/d 40%–55% of total energy intake from CHO Physical activity Supervised exercise program Frequency 3-4/week Intensity 30% heart rate reserve Duration 40 min Type walking Delivered by dieticians and kinesiologists Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care ] 44 3.5 6.3 4.0 6.2 ± ± ± ± 325 260 31.9 38% 33.4 65 32.4 32.1 38% nulliparous ======50% aboriginal I nulliparous I C SES na Prepreg BMI C Reference, sample characteristics Mottola et al. [ Design Case (historical cohort)-control matched design N Parity C Gestational age at recruitment 16–20 weeks Duration of study until 34–36 weeks of gestation Country Canada Ethnicity > Age: C I I 10 Journal of Pregnancy NS value NS P 398 ± 5 (14%) 4,000 g = > n 3,585 425 ± 3(7%) 4,000 g = > n 3,419 value Control Intervention P NA NA — NS value Control Intervention NS P (a) Continued 5.6 ± At delivery 10.9 Adherence to 1990 IOM 54% did not exceed 6.9 ± At delivery 10.6 Adherence to 1990 IOM 52% did not exceed Passive Intervention: Received a brochure designed for the study providing advice about nutrition, physical activity and tips to limit GWG Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care ] 65 4.0 2.6 4.4 2.4 ± ± ± ± 37 28.7 41% na 10.2 43 29.4 40% na 10.2 ======29 Age C nulliparous SES C Gestational age at recruitment C PI nulliparous PI PI Reference, sample characteristics Guelinckx et al. [ Design RCT Country Belgium Ethnicity Caucasians Prepreg BMI > Duration of study until 32 weeks of gestation C Parity C PI PI Journal of Pregnancy 11 NS NS value NS NS NS NS NS NS P 4 6 5 14 468 459 650 = = = = ± ± ± OW NW 5 (12%) 4,000 g 4,000 g 2,500 g (4%) (7%) (6%) (17%) OW NW NW = > > < OW n 3,367 3,430 3,492 5 3 4 14 425 467 629 = = = = ± ± ± 3(7%) OW NW 4,000 g 4,000 g (5%) (3%) (5%) 2,500 g (16%) = OW NW NW > > OW < n 3,419 3,271 3,442 = = value Control Intervention P 2.1 OR 0.005 P 4.7 5.9 ± ± OW 26% NW 36% at or below At 6 months body weight OW 3.7 NW 2.1 pre-pregnancy 3.5 6.2 ± ± NA NA — OW 17% NW 21% at or below At 6 months body weight NW 3.3 OW 4.3 pre-pregnancy NS NS value Control Intervention 0.05 0.05 NS 0.003 > > P (a) Continued 4.4 6.9 ± ± 7.6 ± At delivery 9.8 At final visit NW 15.3 OW 14.7 Adherence to 1990 IOM 59% did not exceed Adherence to 1990 IOM NW 60% OW 33% did not exceed 4.6 7.5 ± ± 6.9 ± OW 15.1 Adherence to 1990 IOM 52% did not exceed Adherence to 1990 IOM NW 48% OW 39% did not exceed At delivery 10.6 At final visit NW 16.2 Active Intervention: Received the same brochure plus received 3 counseling group-sessions given by a dietician Delivered by dieticians One face-to-face visitatstudyentry to discuss Nutrition Recommendations appropriate calorie goals (20 kcal/kg). Physical activity Recommendations appropriate PA (30 min walking most days of the week). Other Recommendations appropriate GWG. Food records, pedometers and body weight scale were provided, postcard about healthy eating and exercise habits Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care Standard prenatal care ] 59 3.6 4.4 2.4 5.2 5.9 2.8 5.2 5.6 ± ± ± ± ± ± ± ± 42 28.0 48% na 9.3 401 43 29.4 40% na 10.2 200 68% Caucasian 28.8 77% nulliparous 19% 26.5 201 69% Caucasian 28.6 16% 26.3 76% nulliparous ======AI nulliparous AI AI I I I unemployed I Parity C C SES C Design RCT N Reference, sample characteristics C Phelan et al. [ Age C nulliparous SES C Gestational age at recruitment C Parity C Country United States Ethnicity C Age C unemployed Prepreg BMI C AI AI I I 12 Journal of Pregnancy value value P P 681 0.89 ± 571 3,688 ± 3,679 value Control Intervention 0.001 value Control Intervention < P P 5.88 ± 2.15 − At 10–12 w PP 5.34 ± 0.75 At 10–12 w PP value Control Intervention value Control Intervention 0.001 0.003 P < P (a) Continued 5.5 ± 7.0 kg (study At delivery 8.7 < goal) 36% 5.8 ± 7.0 kg < (study goal) 21% At delivery 11.3 (b) Prenatal intervention studies using the 2009 IOM gestational weight gain recommendations were mailed weekly. 3 brief phone calls from the dietician Personalized graph of GWG Delivered by dieticians Motivational interview/talk in early pregnancy Nutrition Information about nutrition during pregnancy Physical activity Possibility to attend aqua aerobic classes (1-2/week) Other 30 min-visit/week for motivational talks, weight control, Additional written information about eating and food intake if needed Delivered by trained midwives Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Control Intervention Control Intervention Standard prenatal care ] 56 4.9 4.5 ± ± 30 ≥ 368 208 30.2 48% 26% unskilled 160 29.7 20% unskilled 42% nulliparous ======I nulliparous I workers I Gestational age at recruitment 10–12 weeks Duration of study NA C SES C Reference, sample characteristics Gestational age at recruitment 10–16 weeks Duration of study NA Reference, Sample characteristics Claesson et al. [ Design non RCT N Parity C Country Sweden Ethnicity Swedish Age C workers Prepreg BMI BMI I Journal of Pregnancy 13 value P 603 0.40 ± 548 3,458 ± value Control Intervention P NA NA — 3,610 value Control Intervention 0.16 0.003 P (b) Continued ∗ 7.4 ± 6.8 kg Time of last Last body weight 8.1 ≤ ∗ body weight measurement unclear (study goal) 38% did not exceed ∗ 7.5 ± 6.8 kg Time of last Last body weight 15.4 ≤ ∗ body weight measurement unclear (study goal) 15% did not exceed 7kg < Delivered by study coordinators 6 seminars and at least 5 one-on-one counseling sessions or phone calls Nutrition Written material, seminars and counseling sessions to promote healthy eating, facilitate calories counting, kept food diary Physical activity Written material, seminars and counseling sessions to encourage walking, received pedometer to monitor Other Written material, seminars and counseling sessions to educate the women on obesity and pregnancy and healthy living; goal: GWG Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care ] 57 5.6 5.3 5.1 5.2 ± ± ± ± 28 20 80% Latina 24.4 50% nulliparous 34.2 21 67% Latina 29.0 36.2 19% nulliparous ======12 weeks = = = = = I I I C SES NA Prepreg BMI C Reference, Sample characteristics Shirazian et al. [ Design Historical cohort N ≤ Parity C Gestational age at recruitment Country United States Ethnicity C Age C Duration of study until delivery I I 14 Journal of Pregnancy value P 453 — ± value Control Intervention P NA NA — NA 3,484 — value Control Intervention — P (b) Continued ∗ 6.4 ± 6.0 kg Time of last Last body weight 6.9 ≤ ∗ body weight measurement unclear (goal of study) 56% NA Meeting a midwife every 15 days and 2 group support sessions. Nutrition Initial individual consultation, instructed to eat after “the plate model” and to keep an all-inclusive food intake diary. Physical activity Supervised exercise program Frequency 1/week Intensity NA Duration 40 min Type Water aerobics Askedtoexercise 30 min/day PA on other days Delivered by midwives and dietician Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention ] 66 1 (0–3) = 3.2 4.4 ± ± 25 = Parity median Gestational age at recruitment 8–12 weeks Duration of study NA Country Sweden Ethnicity Swedish Age 31.7 Reference, Sample characteristics Lindholm et al. [ Design single-arm study N SES na Prepreg BMI 35.4 Journal of Pregnancy 15 value 0.41 0.73 P 509 ± LGA 12 (12%) = 3,490 n 530 ± LGA 15 (17%) = 3,516 n value Control Intervention P NA NA — value Control Intervention 0.28 0.008 P (b) Continued 6.0 ± At delivery 14.1 Adherence to 2009 IOM 65% did not exceed 5.9 ± Adherence to 2009 IOM 45% did not exceed At delivery 15.2 Nutrition Personalized dietary counseling Physical activity Supervised exercise session Frequency 1/week Intensity mild-to-moderate Duration 30–45 min Type aerobic and strength exercise Women wereto asked perform several sessions of exercise sessions at home. VHS and DVD to assist with home-based exercises Delivered by dietitians and licensed fitness trainers Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care 29,628 23,792 ] ± 5.9 5.1 ± 5.2 5.4 60 ± ± ± ± 190 88 25% aboriginal 28.7 na 48,602 25.7 102 17% aboriginal 30.1 na 50,833 24.9 ======26 weeks I I I (high) I C Reference, Sample characteristics Hui et al. [ Design RCT N Parity C Gestational age at recruitment < Duration of study until 36 weeks of gestation Country Canada Ethnicity C Age C SES C (high) Prepreg BMI C I I 16 Journal of Pregnancy value 0.07 0.039 P 3,742 40 (32%) 4,000 g (3,464– 4,070) > = n 3,593 39 (25%) 4,000 g (3,335– 3,930) > = n value Control Intervention P NA NA — value Control Intervention 0.01 0.058 P (b) Continued : intervention group; SES: socioeconomic status; prepreg BMI: body mass index prior to pregnancy; IOM: Institute of I At 35 weeks gestation 7.0 (4.7–10.6) Adherence to 2009 IOM 65% did not exceed : control group; C At 35 weeks gestation 8.6 (5.7–11.5) Adherence to 2009 IOM 53% did not exceed :totalnumber; N 60 min Nutrition Personalized dietary counseling 4 times during the intervention Physical activity free membership to gym Supervised exercise session 1x week Frequency 1/week Intensity moderate Duration Type aerobic, elastic band, balance exercise Other group sessions 4–6 times. Women were encouraged to be active every day for 30–60 min Pedometers were provided Delivered by dieticians and physiotherapists Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Access to awebsite with nutrition and PA advices ] 62 12 yrs 12 yrs ≥ ≥ 304 154 29.0 (27–32) 55% nulliparous 65% 33.3 (31.7–36.9) 150 29.0 (26–31) 74% 33.4 (30.7–36.5) 53% nulliparous ======I I I I I C SES C Reference, Sample characteristics Vinter et al. [ Design RCT N Country Denmark Ethnicity: Danish Age: C Parity C Gestational age at recruitment 10–14 weeks Duration of study until delivery Prepreg BMI: C Abbreviations: RCT: randomized controlled trial; Medicine; NA: non available; NS: non significant, NW: normal weight; OW: overweight; OB: obese. Journal of Pregnancy 17 value P ight. value Control Intervention P NA NA — NA NA — value Control Intervention NS NS P 5.9 ± Walkers At delivery 15.4 Adherence to 1990 IOM 0% did not exceed 6.8 ± Stretchers At delivery 15.9 Adherence to 1990 IOM 11% did not exceed (a) Prenatal intervention studies using the 1990 IOM gestational weight gain recommendations Physical activity: Supervised and home-based exercise program Stretching: Frequency 5/week Intensity — Duration 40 min Type Videotape Walking: Frequency 5/week Intensity 55–69% of HR max Duration 40 min Delivered by exercise specialists Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) ects of prenatal intervention studies including physical activity alone on gestational weight gain, maternal weight retention, and infant birth we ff Control Intervention Control Intervention — 3: E 2 Table 75,000 26 75,000 > = ≥ > ] 5.0 26 kg/m W 2 ≥ 64 87% 61% 71 ± 124 60 82% Caucasians 43% 98% ======14 weeks W W kg/m Parity NA SES S S Reference, Sample characteristics Yeo [ 97% Gestational age at recruitment < Duration of study until end of pregnancy Design RT Stretching (S) versus walking (W) N Country United States Ethnicity S Caucasians Age 31.0 (high) Prepreg BMI S (high) W 18 Journal of Pregnancy value 0.44 0.54 P 563 ± 3 (33%) 2,500 g = < n 3,222 656 ± 2 (37%) 2,500 g = < n 3,313 value Control Intervention P value Control Intervention 0.38 NA NA — P 2.1 ± At delivery 14.3 Adherence to IOM NA 1.6 ± At delivery 15.1 Adherence to IOM NA (b) Prenatal intervention studies using the 2009 IOM gestational weight gain recommendations Physical activity: Supervised exercise program Frequency 3/week Intensity 70% pred HR max Duration 50 min Type Water aerobic Delivered by trained instructors Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Sedentary during pregnancy ] 70 3.8 4.5 ± ± 71 37 23.4 24 24.1 = = = = = I C Reference, Sample characteristics Cavalcante et al. [ Design RCT N Gestational age at recruitment 18–20 weeks Duration of study until end of pregnancy Country Brazil Ethnicity NA Age NA Parity NA SES Low SES Prepreg BMI C I Journal of Pregnancy 19 0.1 0.1 0.1 value > > > P 411 ± 1(1%) 4(6%) 4,000 g 2,500 g = = > < n n 3,165 477 ± 4(6%) 7 (10%) 4,000 g 2,500 g = = > < n n 3,307 value Control Intervention P value Control Intervention 0.1 NA NA — > P (b) Continued 3.7 ± At delivery 11.5 Adherence to 2009 IOM NA 3.4 ± At delivery 12.4 Adherence to 2009 IOM NA y Physical activity: Supervised exercise program Frequency 3/week Intensit 60% HR max Duration 35–40 min Type Muscle strengthening and toning program Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Continue usual daily activities ] 69 3.7 0.5 2.9 0.5 ± ± ± ± 160 80 29.5 57% nulliparous 13% high school 23.4 80 30.4 72% nulliparous 26% high school 24.3 ======I I I C Reference, Sample characteristics Barakat et al. [ Design RCT N Parity C Gestational age at recruitment 12-13 weeks Duration of study until end of pregnancy Country Spain Ethnicity: NA Age: C SES C Prepreg BMI: C I I 20 Journal of Pregnancy 0.05 value > P 465 ± 411 3,404 ± value Control Intervention P value Control Intervention 0.05 NA NA — 3,465 > P (b) Continued 3.2 ± At delivery 12.5 Adherence to 2009 IOM NA 3.1 ± Adherence to 2009 IOM NA At delivery 13.8 y 70% HR max Physical activity: Supervised exercise program Frequency 3/week Intensit ≤ Duration 35–45 min Type aerobic class (2/week) aquatic activities (1/week) Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Continue usual daily activities ] 72 2.9 high 2.8 high 3 > ± 4 > ± ± ± 83 43 31 49% nulliparous 14% 23.0 40 32 65% nulliparous 83% 22.7 ======I I Parity C C Reference, Sample characteristics Barakat et al. [ Design RCT N Gestational age at recruitment 6–9 weeks Duration of study until end of pregnancy I school Prepreg BMI: C Country Spain Ethnicity NA Age C SES C I school I Journal of Pregnancy 21 value P NA NA — value Control Intervention 0.93 0.001 P 1.6 1.7 3.9 ± ± ± we PP 0.8 3.3 At 7.1 if ex sessions 2/week 1.5 w 4.1 ± ± PP 3.3 At 8.1 value Control Intervention 0.31 0.01 0.59 0.006 P (b) Continued 24) = n 4.0 2.0 ± ± At 36–38 weeks 13.0 11.0 If attendance to all ex sessions (2/week, Adherence to 2009 IOM 67% if ex 2/week 0% did not exceed 4.0 ± Adherence to 2009 IOM 62% did not exceed At 36–38 weeks 13.8 2/week Physical activity: Supervised exercise program Frequency ≥ Intensity 12–14 RPE Duration 60 min Type Aerobic dance Women wereto asked exercise 30 min/day on other days Delivered by certified aerobic instructors Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Continue usual daily activities 4.4 4.7 4.3 3.7 3.8 4.1 ± ± ± ± ± ± 105 53 30.3 85% 23.9 18.0 52 31.2 85% 23.8 17.3 ] ======67 I college/university I C Reference, Sample characteristics Haakstad and Bø [ Design RCT N Parity 100% nulliparous SES C Gestational age at recruitment C Duration of study 12 weeks Country Norway Ethnicity NA Age C college/university Prepreg BMI C I I I 22 Journal of Pregnancy NS value NS 0.79 P 700 ± 2(6%) 8 (24%) SGA LGA = = n n 3,267 591 ± 1(3%) 8 (24%) SGA LGA = = n n 3,228 value Control Intervention P NA NA — value Control Intervention 0.54 0.76 0.001 0.43 P (b) Continued 5.6 1.7 5.0 ± ± : intervention group; SES: socioeconomic status; prepreg BMI: body mass index prior to pregnancy; IOM: Institute of I ± 10.4 10.0 OB OW Adherence to 2009 IOM 52% did not exceed At final visit 10.3 = = 7.6 3.9 : control group; 7.4 ± C ± ± 10.9 16.4 At final visit 11.5 OB OW = = Adherence to 2009 IOM 43% did not exceed :totalnumber; N 140 < Physical activity: Supervised exercise program Frequency 1/week Intensity HR Duration 40 min Type Aerobic dance Women wereto asked exercise 5/week at home Other: women were counseled on recommended GWG Delivered by trained physical therapists Treatment Gestational weight gain (GWG) (kg) Maternal weight retention postpartum (PP) (kg) Infant birth weight (g) Control Intervention Control Intervention Standard prenatal care 5.9 6.9 3.7 12 yr of 6.8 6.6 4.2 12 yr of > ± ± ± > ± ± ± 82 42 30.9 24% nulliparous 38% 36.4 17.8 29.7 30% nulliparous 43% 34.8 17.6 ] 40 ======68 I I Country Brazil Ethnicity NA Age C C Reference, Sample characteristics Nascimento et al. [ Design RCT N Parity C Gestational age at recruitment C Duration of study until end of pregnancy school Prepreg BMI C SES C I I I school I Abbreviations: RCT: randomized controlled trial; Medicine; NA: non available; NS: non significant, NW: normal weight; OW: overweight; OB: obese. Journal of Pregnancy 23 the sample) [68]. Sixty-three percent of the women were Studies Using the 2009 IOM Guidelines. Hui et al. [60] compliant with the exercise intervention and accumulated a (Table 2(b)) conducted an intervention that combined per- mean of 80 ± 49 minutes of walking every week and 57 ± 22 sonalized dietary counseling, weekly supervised exercise minutes of exercise from the study protocol. However, no sessions and home-based exercises (VHS and DVD). Women information was provided regarding physical activity levels of different pre-pregnancy BMI categories were included in in overweight and obese women separately. Moreover, the the study. Results showed that 65% of the women who fol- authors did not report physical activity levels for the control lowed the intervention did not gain above the 2009 IOM group. guidelines compared with only 45% of those in the control Taken together, although these 3 studies were successful group (P = 0.008). Only 56% of the women completed at reducing mean GWG, none of them produced statisti- a 3-day food intake record at baseline and 2 months after cally significant differences in adherence to IOM recom- enrolment and the results showed lower energy and fat intake mendations, based on the 1990 [55] or 2009 [57, 68] guide- in the intervention group compared to the control group in lines. response to the intervention. Further, 94% of the women completed physical activity questionnaires and the results 2.1.2. Successful Interventions at Preventing Excessive showed that women in the intervention group increased their Gestational Weight Gain Based on the IOM Guidelines physical activity levels and were active more than twice/week for more than 20 minutes/session 2 months after enrolment. Studies Using the 1990 IOM Guidelines. Polley et al. No changes in physical activity levels were found in the [58](Table 2(a)) conducted a behavioral intervention with control group. All women in the intervention group self- stepped care (increasing one-on-one care if needed) in low- reported home exercise for 3–5 times/week during the study. income women of various pre-pregnancy BMI categories Taken together, although these 4 studies were successful (normal weight, overweight, and obese women). The inter- at preventing excessive GWG based on the 1990 [58, 59, 61] vention group received education about weight gain, healthy or 2009 [60] IOM guidelines, they did not show a significant eating and exercise, and individual graphs of their weight reductioninmeanGWG. gain. The intervention was successful only among normal weight women (55% of the sample), with more women not 2.1.3. Successful Interventions at Reducing Mean Gestational exceeding the IOM guidelines in the intervention versus con- Weight Gain and Preventing Excessive Gestational Weight Gain trol group (67% versus 42%, resp.; P<0.05) [58]. Changes Based on the IOM Guidelines. Only 4/19 (21%) intervention in dietary intake and exercise expenditure were assessed by studies were successful at both reducing mean GWG and questionnaires pre- and post-intervention and showed that preventing excessive GWG based on the IOM guidelines normal weight women in the intervention group decreased [44, 56, 62, 67]. their fat intake whereas no changes were found in normal weight women in the control group. Regarding physical Studies Using the 1990 IOM Guidelines. Mottola et al. [44] activity levels, no changes were found in any of the groups (Table 2(a)) used a Nutrition and Exercise Lifestyle Inter- [58]. A larger study, based on a clinical component and a vention Program (NELIP) for overweight and obese preg- by-mail education program targeting healthy eating, physical nant women. This single-arm intervention included a per- activity and GWG, with an emphasis on self-monitoring, was sonalized meal plan combined with a supervised walking successful but only in women of low income (41% of the program of mild-intensity 3 to 4 times/week. This inter- sample) [61](Table 2(a)). Among women of normal weight, vention was highly successful, with 80% of the women not 71% of those who participated in the intervention did not exceeding recommended GWG while on NELIP. Mean GWG exceed the IOM guidelines compared to only 55% (P = 0.05) during the intervention was 6.8 ± 4.1kg(0.38 ± 0.2kg/week) of the historical cohort. Among the overweight women, 56% [44]. During the intervention, the women walked 2.84 ± 0.87 of those in the intervention group did not gain excessively times per week and dietary intake assessments showed a compared to only 28% (P = 0.04) of the historical cohort. significant decrease in daily intake of total energy (from Mean GWG, as well as nutrition and physical activity habits 2,228.0±474.6 to 1,900.2±343 kcal) and carbohydrate (from of these low-income women were not reported. Finally, 318.5±155.1 to 259.1±93.9 g) and an increase in the percent Phelan et al. [59](Table 2(a)) examined the effect of a of daily energy from protein (16.9%±2.4% to 18.4%±2.3%). low-intensity, partially mail-based behavioral intervention targeting dietary intake, physical activity, and weight gain Studies Using the 2009 IOM Guidelines. Claesson et al. [56] monitoring. The authors found that their intervention was (Table 3(b)) examined the effect of an intervention that successful but only in women of normal weight prior to included only women who were obese prior to pregnancy. pregnancy, with 60% of the women who participated in The cornerstone in the intervention was a motivational inter- the intervention not gaining excessively compared with view/talk in early pregnancy, focusing on behavior change. only 48% of those receiving standard prenatal care (P = During the intervention, the women were invited to a 30- 0.003). Although the participants received food records and minute individual session every week for motivational talks pedometers, nutrition and physical activity habits of the and weight gain control. Further, the women were invited to women during the intervention were not reported. Simi- attend aqua aerobic classes 1-2 times/week. Women in the larly, information regarding lifestyle habits of the control intervention group gained less weight than those in the con- group was not reported. trol group (8.7 ± 5.5kgversus11.3 ± 5.8, resp., P<0.001) 24 Journal of Pregnancy and were more likely to gain less than 7.0 kg (study goal) women of different pre-pregnancy BMI categories (mean (36% versus 21%, resp., P = 0.003) [56]. Behavior changes BMI of 30.0 ± 6.5kg/m2), had no effect on total maternal in response to the intervention were not reported, nor atten- weight gain and did not prevent excessive GWG based on dance for the aqua aerobic classes. Similarly, no informa- IOM guidelines. Eating and physical activity habits did not tion was given regarding the lifestyle habits of the control improve. In fact, self-reported sedentary behaviors were sig- group. The intervention of Vinter et al. [62](Table 2(b)) nificantly higher in the intervention group (61%) compared included personalized dietary counseling provided on 4 with the control group (23%) [64]. Guelinckx et al. [65] occasions during the study, supervised exercise sessions (Table 2(a)) examined the effect of an intervention study that offered once per week, with free access to a fitness center for was based on a brochure (passive intervention) or on active the duration of the study, and 4–6 group sessions given to education (active intervention) about GWG, dietary habits improve participants’ integration of physical activities in pre- and physical activity in women who were obese prior to gnancy and daily life. Mean weight gain was 7.0 kg (4.7– pregnancy. Although improvements in dietary habits were 10.6 kg) in the intervention group and 8.6 kg (5.7–11.5 kg) in observed in both intervention groups, both interventions the control group (P = 0.01) [62]. Although more women in were unsuccessful at reducing mean GWG and preventing the intervention group versus control group did not exceed excessive GWG. Physical activity levels were similar in all the 2009 IOM recommendations, the difference was statisti- three groups and decreased significantly throughout preg- cally not significant (65% versus 53%, resp., P = 0.058). nancy. Finally, Kinnunen et al. [63](Table 2(a)) examined Compliance with the nutrition counseling sessions was very the effect of an intervention that included 4 individual good, with 92% of the women having completed all 4 nutri- counseling sessions on healthy diet and 5 counseling sessions tion counseling sessions. Eighty-five percent of the women in to help increasing leisure-time physical activity (LTPA), and the intervention group and 21% of those in the control group designing an individual activity plan to achieve a healthy reported that their participation in the study resulted in more GWG. In addition, the participants had the opportunity to healthy eating habits. However, no information regarding join group exercise sessions once a week for 45–60 minutes. changes in dietary intake of the women was provided by the The women, who were of various pre-pregnancy BMI cate- authors. Attendance for physical activity was more difficult gories (75% were of normal weight) exhibited similar GWG to maintain, with only 56% of the women having attended (intervention group, 14.6±5.4 kg versus control group, 14.3± the aerobic classes for at least half of the lessons. Seventy- 4.1 kg, resp.). Surprisingly, the authors reported a trend in a eight percent of the women in the intervention group versus higher number of women not exceeding the IOM guidelines 65% of those in the control group engaged in leisure time in the control group compared to the intervention group physical activities during pregnancy. However, the frequency, (70% versus 54%, P = 0.053). However, the intervention duration, and intensity of these activities were not reported. improved dietary habits (increase in the intake of vegetables, Finally, Haakstad et Bø [67](Table 2(a)) investigated the fruits, and berries) but had no effect on LTPA levels. effect of a 12-week exercise program in women of different pre-pregnancy BMI categories (75% were of normal weight). Studies Using the 2009 IOM Guidelines. Lindholm et al. [66] The program consisted of 60 minutes of supervised aerobic (Table 2(b)) conducted a single-arm intervention for obese dance performed at least twice/week. The women were asked women who were invited to an initial individual consultation to exercise 30 minutes/day on the other days of the week. with a dietician and then to meet with a midwife once The intervention was found to be effective only in women every two weeks and to attend 2 group-support sessions. who attended 2 dance classes per week (27% of the exercis- In addition, the intervention included water aerobic classes ing group), with a mean weight gain of 11.0 ± 2.0 kg in the once per week and the women were asked to exercise exercising women and 13.0 ± 4.0 kg in the control women 30 minutes/day on other weekdays. Excessive GWG (i.e., (P = 0.01). All exercising women gained within the recom- >6.0 kg) was prevented in only 56% of the women, with mendations compared to 62% of the control women (P = ameanGWGof6.9 ± 6.4 kg. No information was given 0.006). regarding improvements in eating and physical activity habits of the women in response to the intervention. Further, 2.1.4. Unsuccessful Interventions at Reducing Mean Gesta- attendance to the water aerobic classes was not reported. tional Weight Gain and Preventing Excessive Gestational Finally, 3 studies that included only an exercise intervention Weight Gain Based on the IOM Guidelines. Alargenumber were found to be unsuccessful. Barakat et al. [69, 72] of intervention studies were unfortunately not successful at (Table 3(b)) published two studies examining the effect of reducing mean GWG and preventing excessive GWG based a supervised exercise program performed 3 times/week for on the IOM guidelines. 35–40 minutes/session, one focusing on light resistance and toning exercises and the other on aerobic exercises (1/week) Studies Using the 1990 IOM Guidelines. Gray-Donald et al. and aquatic activities (2/week). Both interventions included [64](Table 2(a)) developed a community-based intervention women of different pre-pregnancy BMI categories (mean for the Aboriginal Cree population, based on the social learn- BMI in the normal range). Although adherence to the ing theory. Components of this intervention were pamphlets exercise program was >85%, similar GWG was found in the about nutritional choices, supermarket tours and cook- intervention and control groups. No information was given ing demonstrations, exercise and walking group sessions, regarding whether the intervention was successful at prevent- and individual counseling. The intervention, that included ing excessive GWG based on the 2009 IOM guidelines. Journal of Pregnancy 25

Similarly, a supervised water aerobic exercise program of becoming pregnant. Asking them to make behavioral moderate intensity performed 3 times/week for 50 min- changes while participating in a prenatal lifestyle interven- utes/session had no effect on GWG in sedentary pregnant tion may be challenging. Moreover, they may have body women of different pre-pregnancy BMI categories [70] weight issues and low self-esteem. Because of these factors, (Table 3). The women participated in a mean of 25 exercise overweight and obese pregnant women may need more sup- sessions during this 15-week intervention and GWG was port and encouragement during an intervention to be able to similar in the intervention and control groups (14.3 ± 2.1kg overcome the barriers they may face in order to improve their and 15.1 ± 1.6 kg, resp.). Adherence to IOM guidelines was lifestyle habits and be successful in limiting their weight gain. not reported by the authors. Finally, Yeo [71](Table 3(a)) Accordingly, intervention studies that included overweight compared the effect of a supervised walking program to a and obese women and were successful at reducing mean stretching program on preeclampsia risk factors, including GWG and preventing excessive GWG were based on frequent excessive GWG, in women of different BMI categories prior individualized nutrition counseling sessions and discussions to pregnancy (81% were obese). The intervention consisted regarding healthy weight gain throughout pregnancy, com- of 40 minutes of stretching or moderate-intensity walking, 5 bined with supervised exercise sessions [44, 56, 62]. On times/week. Mean GWG was similar in both groups (15.4 ± the other hand, intervention studies based only on written 5.9 kg in the women who walked and 15.9 ± 6.8 kg in the or oral recommendations regarding healthy lifestyle habits stretching women), although the walkers tracked more daily during pregnancy and of less frequent interactions were steps compared to the stretching group, with an average of unsuccessful at reducing mean GWG or preventing excessive 7, 790 ± 3, 890 and 5, 355 ± 3, 044 steps per day, respectively GWG in overweight and obese women [59, 65]. The way (P = 0.0002). Even if the step difference between the groups of approaching behavioral change is also of paramount was statistically significant, a difference of approximately importance, especially in overweight and obese women. Only 2,500 steps per day may not have been enough to impact few prenatal lifestyle interventions included any theoretical on GWG. Noteworthy, 2,500 steps is equivalent to approxi- background as a basis for behavioral change. However, we mately 25 min of moderate-intensity walking, implying that may speculate that interventions based on frequent interac- the women were not compliant with the exercise program tions with the participants and discussions regarding healthy recommending 40 min of moderate-intensity walking. weight gain included behavioral change objectives, which may have contributed to the success of these interventions. 2.1.5. Summary and Points to Discuss. Based on the studies The background of the providers of counseling may also have presented above, it appears that prenatal lifestyle interven- influenced the success of the intervention. Counseling pro- tions promoting healthy eating and physical activity habits vided by the same team of healthcare professionals through- were more successful at decreasing mean GWG and/or out the intervention (nurse, midwife, dietician, kinesiologist, preventing excessive GWG compared to those including an or psychologist) may have had a stronger impact on helping exercise component alone (69% versus 33%). This suggests the women to improve their lifestyle habits and be successful that the same principle may apply to the prevention of in limiting their weight gain, emphasizing a team approach. excessive GWG as in weight loss: the major component is in Another factor that may have contributed to the mixed dietary changes and the role of physical activity is to support results of prenatal lifestyle interventions may be the use of and maintain these achievements [73]. the 1990 versus 2009 IOM GWG guidelines, although we However, there are still a large number of prenatal cannot draw conclusions as to whether studies using the 1990 lifestyle interventions that were unsuccessful and it is impor- or 2009 IOM guidelines were more likely to be successful tant to discuss the factors that may explain their ineffective- at preventing excessive GWG. However, with the increased ness in order to design future successful prenatal lifestyle prevalence of maternal obesity and the new GWG guidelines interventions. These factors may be numerous and include that are more restrictive, especially for obese women, we may characteristics of the women (parity, socioeconomic status, expected that upcoming studies will have to provide more maternal pre-pregnancy BMI, pre-pregnancy physical fitness interactive and frequent counseling sessions to be successful levels, pre-pregnancy lifestyle habits), the design of the at preventing excessive GWG in obese women. intervention, the gestational age at which the intervention Finally, an important methodological issue that was was initiated, the components of the intervention (nutrition, identified in several prenatal lifestyle interventions presented physical activity, behavioral change), the frequency and above is that although they all included a physical activity intensity of physical activity, the type of intervention (i.e., component, combined or not with nutritional advice, they phone-based, mail-based, or supervised intervention), the did not assess dietary intake and physical activity levels of frequency of the interaction with the women, the contexts the participants pre- and post-intervention. Similarly, some or providers of counseling, and, of course, compliance with interventions provided food diaries and pedometers to the the intervention. participants in order to encourage them to monitor their There is good evidence to suggest that pre-pregnancy daily dietary intake and physical activity levels but the data BMI is an important factor influencing the success of an collected using these tools were not reported by the authors. intervention and that it should be taken into considera- This is important missing information because researchers tion when designing a prenatal lifestyle intervention. For need to identify from previous successful interventions what example, overweight and obese pregnant women likely had contributed to their success in order to design future success- unhealthy eating habits and a sedentary lifestyle before ful prenatal lifestyle interventions. Consequently, it is highly 26 Journal of Pregnancy important that researchers document the important/effective recommendations, two studies did not report infant mea- aspects of their intervention by examining whether the surements at birth [55, 67] and seven reported no differences participants were compliant with the recommendations in infant birth weight or in the incidence of low (<2,500 g) or given during the intervention and whether they had modified high (>4,000 g) birth weight between the intervention and eating and/or physical activity habits. Further, it is important control groups [56–61, 68]. Only two studies showed a sig- to examine eating and/or physical activity habits of the nificant effect of limiting GWG on infant weight at birth. control group. The few authors who reported data regarding Mottola et al. [44](Table 2(a)) reported that in the over- the impact of their intervention on changes in nutrition weight women, a lower percentage of babies born weighing and physical activity habits of the participants have found between 4,000 g and 4,500 g was found in the intervention that decreasing fat intake [58, 60], decreasing carbohydrate group compared to the historical control cohort (3.2% versus intake while increasing protein consumption [44], and 18%, resp.; P = 0.048). However, similar overall mean birth increasing physical activity levels to meet physical activity weight was found between the intervention group (3, 590 ± recommendations during pregnancy [44, 60, 67, 68]helped 500 g) and the historical control cohort (3, 560 ± 600 g; P> to achieve a healthy GWG. 0.05). Also, no babies born to the NELIP women weighed However, compliance is always difficult to interpret and less than 2,500 g, whereas 3.5% of the babies born to the the authors should report as much information regarding historical control cohort were in this weight category. Con- compliance as possible. For example, if compliance with versely, Vinter et al. [62](Table 2(b)) surprisingly found physical activity recommendations is defined as being active higher birth weights in the intervention group (3,742 g) com- 3 times per week, the authors should not only report the pared to the control group (3,593 g; P = 0.039), with a trend mean number of weekly exercise sessions performed by in higher frequency of high birth weight babies (>4,000 g) in the women but also the percentage of women having been the intervention group compared to the control group (32% active 3 times per week. In the case where there is few but versus 25%; P = 0.07). exceptionally well-adhered women who were active 6 times per week, these women will pull the group average up. The 2.2.1. Summary and Points to Discuss. Based on these prena- reader will then consider that the intervention was effective tal lifestyle intervention studies, conclusions suggesting that because all women were compliant with physical activity decreasing GWG or preventing excessive GWG may be linked recommendations. In fact, the few but exceptionally well- to lower incidence of high birth weight cannot be drawn, adhered women were mostly responsible for the positive with the exception of one study [44]. Noteworthy, the main results. However, compliance with the physical activity outcome of the prenatal lifestyle intervention studies pre- component of a prenatal lifestyle intervention was found to sented above was preventing excessive GWG, with secondary be a major problem. Factors such as concerns for the safety of outcomes being preventing high infant birth weight and the baby, physical limitations, and lack of energy, motivation, pregnancy complications. Thus, it is likely that these studies or resources may contribute to the low compliance with the were underpowered to find a significant effect on these out- physical activity program/recommendations. The inclusion comes. Future larger and well-controlled intervention studies of supervised exercise sessions may help to favor compliance are therefore warranted to examine the effect of the promo- with the physical activity component of the intervention and tion of healthy lifestyle habits and adequate GWG on new- achievement of healthy weight gain. It would allow frequent born weight. interactions with the women during which physical activity- Moreover, future studies should assess body composition related behavior change objectives may be provided and of the neonates. First, even if infant weight at birth is similar emphasis put on the safety and health benefit of physical between the intervention and control groups, differences activity during pregnancy. in body fat distribution might exist and have a long-term Taken together, differences in the characteristics of the impact on the metabolic health of the infant. Second, if a women, the nutritional, physical activity, and behavioral higher birth weight is observed, it is highly important to change approaches used make the comparison of different evaluate whether this is due to higher lean mass or fat mass. interventions difficult and the identification of the most Vinter et al. [62], who found a higher infant weight at birth effective ways to prevent GWG challenging. Further research in the intervention group compared to the control group, is needed to identify the effectiveaspectsofprenatallifestyle mentioned the absence of information regarding body interventions, especially for overweight and obese women, composition of the neonates as a limitation of their study. in order to support the current evidence suggesting that They also suggested that their findings may be explained prenatal lifestyle interventions may play an important role by the stimulation of placenta development in response in the prevention of excessive GWG. to maternal exercise, which has been previously reported [74]. Another explanation may be the time-specific effect of 2.2. Does Limiting Gestational Weight Gain Decrease the exercise on fetal growth, as discussed in a recent review paper Incidence of High Infant Birth Weight? Although the preven- [75]. Beginning an exercise program in early pregnancy tion of excessive GWG is important to decrease the inci- (i.e., first trimester) has been reported to have a stimulatory dence of high infant birth weight, 4/19 (21%) of the inter- effect on placenta growth and function, which may increase vention studies did not report infant birth weight. Among infant birth weight [75]. Although this may be a beneficial the intervention studies that were successful in decreasing adaptation in lean and physically active women, it may have mean GWG or preventing excessive GWG based on the IOM a less desirable effect of promoting excess fetal growth in Journal of Pregnancy 27 overweight/obese women. Interestingly, the intervention of retention of −2.15 ± 5.88 kg and 0.75 ± 5.34 kg at 10–12 Vinter et al. [62] started early in pregnancy (10–14 weeks) weeks postpartum in obese women who followed the inter- and the data support this time-specific effect of exercise on vention and in those who received standard prenatal care, fetal growth. Initiating an exercise program for overweight/ respectively (P<0.001). In addition, Mottola et al. obese women during the second trimester of pregnancy [44](Table 2(a)) showed that overweight/obese women who may be more beneficial in preventing excess fetal growth, as participated in the NELIP retained a mean of 2.2 ± 5.6kg shown by the study of Mottola et al. [44]. at 2 months postpartum, with 53% of the women being The timing of excessive GWG during pregnancy also within 2.0 kg of pre-pregnancy body mass. Phelan et al. [59] deserves more attention. Indeed, excessive GWG that occurs (Table 2(a)) reported that normal weight and overweight/ early in pregnancy may not have the same influence on fetal obese women prior to pregnancy who received the interven- growth and infant weight at birth compared to excessive tion were more likely to be at, or below, their preconception GWG occurring later in pregnancy. For example, excessive weight at 6 months postpartum than women who received GWG, especially during the first trimester of pregnancy, has standard prenatal care (OR: 2.1; 95% CI: 1.3, 3.5; P = been found to be associated with an increased risk of gesta- 0.005). Finally, Olson et al. [61] found that a significantly tional diabetes [76, 77], suggesting that timing of excessive smaller proportion of intervention group women retained GWG may be related to different patterns of metabolic 2.27 kg or more at 1-year postpartum in the low-income and change affecting glucose metabolism regulation as well as overweight BMI group (25% versus 55%; P = 0.04). fetal growth. Finally, it is important to consider that even if prenatal 2.3.1. Summary and Points to Discuss. Taken together, these lifestyle interventions were successful at limiting GWG or findings support the importance of promoting a healthy life- preventing excessive GWG but had no effect on infant style during pregnancy to reduce postpartum weight reten- outcome at birth, there still may be a beneficial effect on tion. However, because only 32% of prenatal lifestyle inter- the long-term health of the infant. As suggested by the fetal vention studies aiming at preventing excessive GWG fol- programming (Barker’s) hypothesis [78], the prenatal period lowed the women into the postpartum period, future studies is a unique physiological window in which maternal and should include follow-up in their design. Further, studies are fetal adaptations can have major consequences for the long- also needed to examine whether the promotion of healthy term health and well-being of the offspring. More research lifestyle habits during pregnancy has a long-lasting effect in is needed to examine the effect of prenatal lifestyle interven- both mother and offspring, beyond the postpartum period. tions on the infant weight trajectory during childhood and adulthood. Moreover, although epidemiological studies have 3. Conclusion suggested that excessive GWG is associated with an increased risk for the infant of becoming overweight/obese later in life Obesity is reaching epidemic proportions in today’s society [8, 13, 16, 17], the molecular mechanisms through which and women of childbearing age are at an increased risk the exposure to excessive GWG translates into the devel- for developing this disease because of excessive weight gain opment of obesity among offspring is currently unknown. during pregnancy and weight retention after birth. Further, Epigenetics (i.e., changes in gene expression) has been offspring of women who gained excessively during pregnancy suggested as a very likely mechanism. Genes and metabolic are at an increased risk of becoming overweight and obese pathways showing epigenetic dysregulation in response to in the long term. Promoting a healthy lifestyle during preg- the exposure to excessive GWG are likely involved in obesity nancy and providing recommendations regarding healthy development among children. The identification of these weight gain, especially in overweight and obese women, genes and metabolic pathways is of high interest and may become therefore increasingly important in the context of the have the potential in the long term to foster the development prevention of obesity. Several prenatal lifestyle interventions of improved prevention programs and treatments. have been published but due to several issues discussed in this review, it is difficult to make any conclusions about the most 2.3. Does Limiting of Excessive Gestational Weight Gain effective way to prevent excessive GWG, reduce the incidence Prevent Postpartum Weight Retention? Finally, one of the of small or large babies and prevent weight retention. How- objectives of preventing excessive GWG is to optimize ever, prenatal interventions combining a nutritional coun- maternal outcomes, such as preventing maternal weight seling, supervised physical activity sessions, and a behavioral retention. However, only 7/19 (37%) of the prenatal lifestyle change approach might be the most successful (Table 4). intervention studies followed the women after delivery [44, Noteworthy, this review is restricted to studies published 56, 58, 59, 61, 64, 67]. Gray-Donald [64]andPolleyetal. in English, which included mainly Caucasian women. It is [58](Table 2(a)) found similar weight retention at the first possible that the most effective way to prevent excessive GWG postnatal visit in women who followed the intervention and is different among women of other ethnicities. in those who were in the control group. Haakstad and Bø Although the role of prenatal physical activity to prevent [67](Table 3(b)) reported lower weight retention at 8–12 excessive GWG has been recently questioned, the prevailing weeks postpartum in women who were compliant with the literature indicates that women who were compliant with the prenatal exercise program compared to those in the control physical activity program and met the physical activity rec- group (0.8±1.7kgversus3.3±4.1kg,resp.;P = 0.001). Simi- ommendations were more likely to achieve appropriate GWG larly, Claesson et al. [56](Table 2(b)) found a mean weight and return to their pre-pregnancy body mass after delivery. 28 Journal of Pregnancy 4,000 g 2,500 g 4,000 g < successful 4,000– 4,500 g successful > unsuccessful ≥ successful Mean birth weight NA Mean birth weight unsuccessful Mean birth weight unsuccessful Mean birth weight NA Mean birth weight unsuccessful Mean birth weight unsuccessful Changes in infant birth weight/prevention of large and small infant at birth Mean birth weight unsuccessful 7 weeks ± Mean weight retention NA Mean weight retention at 6 weeks unsuccessful Mean weight retention at 8 Mean weight retention at 1 year Partially successful (only in low SES OW women) Mean weight retention NA Mean weight retention at 8 weeks successful Mean weight retention NA Maternal weight retention postpartum unsuccessful Adherence to 1990 IOM unsuccessful Adherence to 1990 IOM NA Adherence to 1990 IOM Partially successful (only in NW women) Adherence to 1990 IOM Partially successful (only in low SES women) Adherence to 1990 IOM unsuccessful Adherence to 1990 IOM successful Adherence to 1990 IOM unsuccessful Gestational weight gain (GWG) Mean GWG unsuccessful Mean GWG unsuccessful Mean GWG unsuccessful Mean GWG unsuccessful Mean GWG successful Mean GWG successful Mean GWG unsuccessful ects of prenatal lifestyle intervention studies with regard to gestational weight gain, maternal weight retention, and infant birth weight. ff Nutrition (counseling) Physical activity (supervised) Guidance about self-monitoring of GWG Nutrition (recommendations) Physical activity (supervised) Advice to stay within GWG recommendations Nutrition (newsletters) Physical activity (newsletters) Info about appropriate GWG Nutrition (newsletters) Physical activity (newsletters) Guidance about self-monitoring of GWG Nutrition (plan) Physical activity (recommendations) Information about GWG guidelines Nutrition (plan) Physical activity (supervised) Nutrition (brochure/counseling) Physical activity (brochure/counseling) Tips to limit GWG ] 64 ] ] 4: Summary of the e ] 65 63 ] ] ] 44 58 61 55 Table ReferenceGray-Donald et al. [ Intervention Polley et al. [ Olson et al. [ Kinnunen et al. [ Asbee et al. [ Mottola et al. [ Guelinckx et al. [ Journal of Pregnancy 29 2,500 g 4,000 g 2,500 g 4,000 g < unsuccessful LGA (not defined) unsuccessful > unsuccessful < unsuccessful > unsuccessful infant at birth Changes in infant birth Mean birth weight unsuccessful Mean birth weight unsuccessful Mean birth weight unsuccessful Mean birth weight NA weight/prevention of large and small Mean birth weight unsuccessful Mean birth weight NA Mean birth weight unsuccessful Mean birth weight unsuccessful pre-pregnancy body weight at Mean weight retention at 10–12 weeks successful Mean weight retention NA Mean weight retention NA Mean weight retention NA Maternal weight retention postpartum Mean weight retention NA Mean weight retention NA ≤ Mean weight retention NA 6months successful 0.058) = P 4: Continued. 6.8 kg (study goal) 6.0 kg (goal of study) 7.0 kg (study goal) < successful Adherence to IOM NA Adherence to 1990 IOM Partially successful (only in NW women) Adherence to 2009 IOM successful Adherence to 2009 IOM successful ( Adherence to 1990 IOM unsuccessful ≤ ≤ unsuccessful unsuccessful Table Gestational weight gain (GWG) Mean GWG successful Mean GWG successful Mean GWG unsuccessful Mean GWG unsuccessful Mean GWG Mean GWG unsuccessful Mean GWG successful Mean GWG unsuccessful Nutrition (recommendations) Physical activity (supervised) Motivational interview Nutrition (counseling) Physical activity (counseling) GWG goal Nutrition (recommendations) Physical activity (recommendations) Recommendations about appropriate GWG Nutrition (counseling) Physical activity (supervised) Nutrition (counseling) Physical activity (supervised) Nutrition (counseling) Physical activity (supervised) ] Physical activity (supervised) ] ] ] 70 66 57 ] 56 ] 59 62 ] 60 ] Physical activity (supervised) 71 ReferencePhelan et Intervention al. [ Claesson et al. [ Shirazian et al. [ Lindholm et al. [ Hui et al. [ Vinter et al. [ Yeo [ Cavalcante et al. [ 30 Journal of Pregnancy 2,500 g 4,000 < unsuccessful > unsuccessful SGA (not defined) unsuccessful LGA (not defined) unsuccessful infant at birth Changes in infant birth Mean birth weight unsuccessful Mean birth weight unsuccessful Mean birth weight NA Mean birth weight unsuccessful weight/prevention of large and small Mean weight retention NA Mean weight retention NA Mean weight retention NA Mean weight retention at 8–12 weeks Partially successful (only if attendance to 2 ex sessions/week) Maternal weight retention postpartum 4: Continued. Adherence to 2009 IOM NA Adherence to 2009 IOM NA Adherence to 2009 IOM unsuccessful Adherence to 2009 IOM Partially successful (only if attendance to 2 ex sessions/week) Table Gestational weight gain (GWG) Mean GWG unsuccessful Mean GWG unsuccessful Mean GWG Partially successful (in OW women) Mean GWG Partially successful (only if attendance to 2 ex sessions/week) ] Physical activity (supervised) ] Physical activity (supervised) 68 67 ] Physical activity (supervised) ] Physical activity (supervised) 69 72 ReferenceBarakat et Intervention al. [ Barakat et al. [ Haakstad and Bø [ Nascimento et al. [ Abbreviations: IOM: Institute of Medicine; NA: non available: SES: socioeconomic status; NW: normal weight; OW: overweight. Journal of Pregnancy 31

Consequently, the role of physical activity during pregnancy [10] E. P. Gunderson, B. Abrams, and S. Selvin, “The relative per se is not the issue. Instead, it is the designs of the physical importance of gestational gain and maternal characteristics activity approaches that may not have been adequate for pre- associated with the risk of becoming overweight after preg- venting excessive GWG and/or protocols that were not struc- nancy,” International Journal of Obesity, vol. 24, no. 12, pp. tured to achieve desired compliance. Only few interventions 1660–1668, 2000. included any theoretical background as a basis for behavioral [11] H. Lowell and D. C. Miller, “Weight gain during pregnancy: adherence to Health Canada’s guidelines.,” Health Reports/Sta- change and such approach may help to increase compliance tistics Canada, vol. 21, no. 2, pp. 31–36, 2010. with physical activity program/recommendations. Rigorous [12] D. S. Ludwig and J. Currie, “The association between preg- scientific research, designed to increase compliance of the nancy weight gain and birthweight: a within-family compari- women, is still required to make an informed decision about son,” The Lancet, vol. 376, no. 9745, pp. 984–990, 2010. the role of physical activity in the prevention of excessive [13] S. R. Crozier, H. M. Inskip, K. M. Godfrey et al., “Weight GWG. In addition, research is needed to examine the long- gain in pregnancy and childhood body composition: findings lasting effect of the prevention of excessive GWG on the risk from the Southampton Women’s Survey,” American Journal of of obesity for the mother and offspring. Finally, investigating Clinical Nutrition, vol. 91, no. 6, pp. 1745–1751, 2010. the molecular mechanisms through which the exposure to [14] P. M. Dietz, W. M. Callaghan, and A. J. Sharma, “High preg- excessive GWG translates into the development of obesity nancy weight gain and risk of excessive fetal growth,” American among offspring (i.e., epigenetic factors) may help to better Journal of Obstetrics and Gynecology, vol. 201, no. 1, pp. 51.e1– understand the aetiology of obesity and to develop more 51.e6, 2009. ffi [15] N. E. Stotland, Y. W. Cheng, L. M. Hopkins, and A. B. e cient prevention intervention and treatment. Caughey, “Gestational weight gain and adverse neonatal outcome among term infants,” Obstetrics and Gynecology, vol. Acknowledgments 108, no. 3 I, pp. 635–643, 2006. [16] L. Schack-Nielsen, K. F. Michaelsen, M. Gamborg, E. L. S. M. Ruchat was supported by a postdoctoral fellowship Mortensen, and T. I. A. Sørensen, “Gestational weight gain in from the Canadian Diabetes Association. The authors have relation to offspring body mass index and obesity from infancy no conflict of interests that is directly relevant to the content through adulthood,” International Journal of Obesity, vol. 34, of this paper. no. 1, pp. 67–74, 2010. [17]E.Oken,E.M.Taveras,K.P.Kleinman,J.W.Rich-Edwards, References and M. W. Gillman, “Gestational weight gain and child adiposity at age 3 years,” American Journal of Obstetrics and Gyne- [1] S. Y. Kim, P. M. Dietz, L. England, B. Morrow, and W. M. cology, vol. 196, no. 4, pp. 322.e1–322.e8, 2007. Callaghan, “Trends in pre-pregnancy obesity in nine states, [18] S. Y. Chu, W. M. Callaghan, C. L. Bish, and D. D’Angelo, “Gest- 1993–2003,” Obesity, vol. 15, no. 4, pp. 986–993, 2007. ational weight gain by body mass index among US women [2] L. K. Callaway, J. B. Prins, A. M. Chang, and H. D. McIntyre, delivering live births, 2004-2005: fueling future obesity,” “The prevalence and impact of overweight and obesity in an American Journal of Obstetrics and Gynecology, vol. 200, no. Australian obstetrics population,” Medical Journal of Australia, 3, pp. 271.e1–271.e7, 2009. vol. 184, no. 2, pp. 56–59, 2006. [19] T. A. M. Simas, X. Liao, A. Garrison, G. M. T. Sullivan, A. [3] D. Y. LaCoursiere, L. Bloebaum, J. D. Duncan, and M. W. E. Howard, and J. R. Hardy, “Impact of updated institute of Varner, “Population-based trends and correlates of maternal medicine guidelines on prepregnancy body mass index cate- overweight and obesity, Utah 1991–2001,” American Journal of gorization, gestational weight gain recommendations, and Obstetrics and Gynecology, vol. 192, no. 3, pp. 832–839, 2005. needed counseling,” Journal of Women’s Health,vol.20,no.6, [4] S. Y. Chu, S. Y. Kim, and C. L. Bish, “Prepregnancy obesity pre- pp. 837–844, 2011. valence in the United States, 2004-2005,” Maternal and Child [20] L. A. Wolfe, P. J. Ohtake, M. F. Mottola, and M. J. McGrath, Health Journal, vol. 13, no. 5, pp. 614–620, 2009. “Physiological interactions between pregnancy and aerobic [5] IOM, Weight Gain during Pregnancy: Reexamining the Guide- exercise.,” Exercise and Sport Sciences Reviews, vol. 17, pp. 295– lines, The National Academies Press, Washington, DC, USA, 351, 1989. 2009. [21] F. K. Lotgering, R. D. Gilbert, and L. D. Longo, “Maternal [6]A.M.Siega-Riz,A.H.Herring,K.Carrier,K.R.Evenson,N. and fetal responses to exercise during pregnancy,” Physiological Dole, and A. Deierlein, “Sociodemographic, perinatal, behav- Reviews, vol. 65, no. 1, pp. 1–36, 1985. ioral, and psychosocial predictors of weight retention at 3 and [22] L. A. Wolfe, I. K. Brenner, and M. F. Mottola, “Maternal exer- 12 months postpartum,” Obesity, vol. 18, no. 10, pp. 1996– cise, fetal well-being and pregnancy outcome.,” Exercise and 2003, 2010. Sport Sciences Reviews, vol. 22, pp. 145–194, 1994. [7] A. M. Siega-Riz, K. R. Evenson, and N. Dole, “Pregnancy- [23] G. A. L. Davies, L. A. Wolfe, M. F. Mottola, and C. MacKinnon, related weight gain—a link to obesity?” Nutrition Reviews, vol. “Joint SOGC/CSEP clinical practice guideline: exercise in 62, no. 7, pp. S105–S111, 2004. pregnancy and the postpartum period,” Canadian Journal of [8]E.Oken,K.P.Kleinman,M.B.Belfort,J.K.Hammitt,and Applied Physiology, vol. 28, no. 3, pp. 330–341, 2003. M. W. Gillman, “Associations of gestational weight gain with [24] M. F. Mottola, “The role of exercise in the prevention and short- and longer-term maternal and child health outcomes,” treatment of gestational diabetes mellitus,” Current Diabetes American Journal of Epidemiology, vol. 170, no. 2, pp. 173–180, Reports, vol. 8, no. 4, pp. 299–304, 2008. 2009. [25] T. L. Weissgerber, L. A. Wolfe, G. A. Davies, and M. F. Mottola, [9] B. L. Rooney and C. W. Schauberger, “Excess pregnancy weight “Exercise in the prevention and treatment of maternal- gain and long-term obesity: one decade later,” Obstetrics and fetal disease: a review of the literature,” Applied Physiology, Gynecology, vol. 100, no. 2, pp. 245–252, 2002. Nutrition, and Metabolism, vol. 31, no. 6, pp. 661–674, 2006. 32 Journal of Pregnancy

[26]T.D.Dye,K.L.Knox,R.Artal,R.H.Aubry,andM.A. in women with gestational diabetes mellitus: a pilot study,” Wojtowycz, “Physical activity, obesity, and diabetes in preg- Applied Physiology, Nutrition and Metabolism,vol.33,no.3, nancy,” American Journal of Epidemiology, vol. 146, no. 11, pp. pp. 511–517, 2008. 961–965, 1997. [44] M. F. Mottola, I. Giroux, R. Gratton et al., “Nutrition and [27]A.L.Gollenberg,P.Pekow,E.R.Bertone-Johnson,P.S.Freed- exercise prevent excess weight gain in overweight pregnant son, G. Markenson, and L. Chasan-Taber, “Sedentary behav- women,” Medicine and Science in Sports and Exercise, vol. 42, iors and abnormal glucose tolerance among pregnant latina no. 2, pp. 265–272, 2010. women,” Medicine and Science in Sports and Exercise, vol. 42, [45] M. F. Mottola and M. K. Campbell, “Activity patterns during no. 6, pp. 1079–1085, 2010. pregnancy,” Canadian Journal of Applied Physiology, vol. 28, [28] J. C. Dempsey, C. L. Butler, T. K. Sorensen et al., “A case- no. 4, pp. 642–653, 2003. control study of maternal recreational physical activity and [46] J. M. Dodd, R. M. Grivell, C. A. Crowther, and J. S. Robinson, risk of gestational diabetes mellitus,” Diabetes Research and “Antenatal interventions for overweight or obese pregnant Clinical Practice, vol. 66, no. 2, pp. 203–215, 2004. women: a systematic review of randomised trials,” British [29] J. Liu, J. N. Laditka, E. J. Mayer-Davis, and R. R. Pate, “Does Journal of Obstetrics and Gynaecology, vol. 117, no. 11, pp. physical activity during pregnancy reduce the risk of gesta- 1316–1326, 2010. tional diabetes among previously inactive women?” Birth, vol. [47] B. Gardner, J. Wardle, L. Poston, and H. Croker, “Changing 35, no. 3, pp. 188–195, 2008. diet and physical activity to reduce gestational weight gain: a [30] A. M. Stuebe, E. Oken, and M. W. Gillman, “Associations of meta-analysis,” Obesity Reviews, vol. 12, no. 7, pp. e602–e620, diet and physical activity during pregnancy with risk for exces- 2011. sive gestational weight gain,” American Journal of Obstetrics [48] A. K. Ronnberg and K. Nilsson, “Interventions during pre- and Gynecology, vol. 201, no. 1, pp. 58.e1–58.e8, 2009. gnancy to reduce excessive gestational weight gain: a sys- [31] C. M. Olson and M. S. Strawderman, “Modifiable behavioral tematic review assessing current clinical evidence using the factors in a biopsychosocial model predict inadequate and Grading of Recommendations, Assessment, Development and excessive gestational weight gain,” Journal of the American Die- Evaluation (GRADE) system,” British Journal of Obstetrics and tetic Association, vol. 103, no. 1, pp. 48–54, 2003. Gynaecology, vol. 117, no. 11, pp. 1327–1334, 2010. [32] ACOG, “Exercise during pregnancy and the postpartum [49] H. Skouteris, L. Hartley-Clark, M. McCabe et al., “Preventing period,” Obstetrics and Gynecology, vol. 99, pp. 171–173, 2002. excessive gestational weight gain: a systematic review of [33] RCOG, “Exercise in pregnancy,” RCOG Statement No. 4., interventions,” Obesity Reviews, vol. 11, no. 11, pp. 757–768, 2006. 2010. [50] I. Streuling, A. Beyerlein, E. Rosenfeld, H. Hofmann, T. Schulz, [34] G.A. Davies, “Obesity in pregnancy,” Journal of Obstetrics and and R. Von Kries, “Physical activity and gestational weight Gynaecology Canada, vol. 32, no. 2, pp. 165–173, 2010. gain: a meta-analysis of intervention trials,” British Journal of [35] M. H. Davenport, S. Charlesworth, D. Vanderspank, M. M. Obstetrics and Gynaecology, vol. 118, no. 3, pp. 278–284, 2011. Sopper, and M. F. Mottola, “Development and validation of [51] I. Streuling, A. Beyerlein, and R. Von Kries, “Can gestational exercise target heart rate zones for overweight and obese preg- weight gain be modified by increasing physical activity and nant women,” Applied Physiology, Nutrition and Metabolism, diet counseling? A meta-analysis of interventional trials,” vol. 33, no. 5, pp. 984–989, 2008. American Journal of Clinical Nutrition, vol. 92, no. 4, pp. 678– [36] ACSM, Guidelines for Exercise Testing and Exercise Prescription, 687, 2010. Lippincott Williams & Wilkins, Philadelphia, Pa, USA, 7th [52] F. Campbell, M. Johnson, J. Messina, L. Guillaume, and E. edition, 2005. Goyder, “Behavioural interventions for weight management in [37]A.M.Petersen,T.L.Leet,andR.C.Brownson,“Correlatesof pregnancy: a systematic review of quantitative and qualitative physical activity among pregnant women in the United States,” data,” BMC Public Health, vol. 11, article 491, 2011. Medicine and Science in Sports and Exercise, vol. 37, no. 10, pp. [53] Z. Sui, R. M. Grivell, and J. M. Dodd, “Antenatal exercise to 1748–1753, 2005. improve outcomes in overweight or obese women: a system- [38] K. R. Evenson, D. A. Savitz, and S. L. Huston, “Leisure-time atic review,” Acta Obstetricia et Gynecologica Scandinavica, vol. physical activity among pregnant women in the US,” Paedia- 91, no. 5, pp. 538–545, 2012. tric and Perinatal Epidemiology, vol. 18, no. 6, pp. 400–407, [54] S. Thangaratinam, E. Rogozinska,´ K. Jolly et al., “Effects of 2004. interventions in pregnancy on maternal weight and obstetric [39] K. R. Evenson and F. Wen, “National trends in self-reported outcomes: meta-analysis of randomised evidence,” British physical activity and sedentary behaviors among pregnant Medical Journal, vol. 344, Article ID e2088, 2012. women: NHANES 1999–2006,” Preventive Medicine, vol. 50, [55] S. M. Asbee, T. R. Jenkins, J. R. Butler, J. White, M. Elliot, no. 3, pp. 123–128, 2010. and A. Rutledge, “Preventing excessive weight gain during pre- [40] K. M. Borodulin, K. R. Evenson, F. Wen, A. H. Herring, and gnancy through dietary and lifestyle counseling: a randomized A. M. Benson, “Physical activity patterns during pregnancy,” controlled trial,” Obstetrics and Gynecology, vol. 113, no. 2, pp. Medicine and Science in Sports and Exercise, vol. 40, no. 11, pp. 305–312, 2009. 1901–1908, 2008. [56] I. M. Claesson, G. Sydsjo,¨ J. Brynhildsen et al., “Weight gain [41] G. Lindseth and P. Vari, “Measuring physical activity during restriction for obese pregnant women: a case-control interven- pregnancy,” Western Journal of Nursing Research, vol. 27, no. 6, tion study,” British Journal of Obstetrics and Gynaecology, vol. pp. 722–734, 2005. 115, no. 1, pp. 44–50, 2008. [42] D. S. Downs, G. C. LeMasurier, and J. M. DiNallo, “Baby steps: [57] T. Shirazian, S. Monteith, F. Friedman, and A. Rebarber, “Life- pedometer-determined and self-reported leisure-time exercise style modification program decreases pregnancy weight gain behaviors of pregnant women,” JournalofPhysicalActivityand in obese women,” American Journal of Perinatology, vol. 27, no. Health, vol. 6, no. 1, pp. 63–72, 2009. 5, pp. 411–414, 2010. [43] M. H. Davenport, M. F. Mottola, R. McManus, and R. Gratton, [58] B. A. Polley, R. R. Wing, and C. J. Sims, “Randomized control- “A walking intervention improves capillary glucose control led trial to prevent excessive weight gain in pregnant women,” Journal of Pregnancy 33

International Journal of Obesity, vol. 26, no. 11, pp. 1494–1502, at 24–28 weeks: a randomised controlled trial,” British Journal 2002. of Sports Medicine, vol. 46, no. 9, pp. 656–661, 2012. [59] S. Phelan, M. G. Phipps, B. Abrams, F. Darroch, A. Schaffner, [73] S. Phelan, K. Jankovitz, T. Hagobian, and B. Abrams, “Reduc- and R. R. Wing, “Randomized trial of a behavioral intervening excessive gestational weight gain: lessons from the weight tion to prevent excessive gestational weight gain: the Fit for control literature and avenues for future research,” Womens delivery study,” American Journal of Clinical Nutrition, vol. 93, Health, vol. 7, no. 6, pp. 641–661, 2011. no. 4, pp. 772–779, 2011. [74] J. F. Clapp III, “The effects of maternal exercise on fetal oxy- [60] A. Hui, L. Back, S. Ludwig et al., “Lifestyle intervention on genation and feto-placental growth,” European Journal of Ob- diet and exercise reduced excessive gestational weight gain in stetrics & Gynecology and Reproductive Biology, vol. 110, no. 1, pregnant women under a randomised controlled trial,” British supplement, pp. S80–S85, 2003. Journal of Obstetrics and Gynaecology, vol. 119, no. 1, pp. 70– [75] S. A. Hopkins and W. S. Cutfield, “Exercise in pregnancy: 77, 2012. weighing up the long-term impact on the next generation,” [61] C. M. Olson, M. S. Strawderman, and R. G. Reed, “Efficacy of Exercise and Sport Sciences Reviews, vol. 39, no. 3, pp. 120–127, an intervention to prevent excessive gestational weight gain,” 2011. American Journal of Obstetrics and Gynecology, vol. 191, no. 2, [76] M. M. Hedderson, E. P. Gunderson, and A. Ferrara, “Gesta- pp. 530–536, 2004. tional weight gain and risk of gestational diabetes mellitus,” [62] C. A. Vinter, D. M. Jensen, P. Ovesen, H. Beck-Nielsen, and Obstetrics and Gynecology, vol. 115, no. 3, pp. 597–604, 2010. J. S. Jørgensen, “The LiP (Lifestyle in Pregnancy) Study: a [77] A. S. Morisset, A. Tchernof, M. C. Dube,´ J. Veillette, S. J. randomized controlled trial of lifestyle intervention in 360 Weisnagel, and J. Robitaille, “Weight gain measures in women obese pregnant women,” Diabetes Care, vol. 34, no. 12, pp. with gestational diabetes mellitus,” Journal of Women’s Health, 2502–2507, 2011. vol. 20, no. 3, pp. 375–380, 2011. [63] T. I. Kinnunen, M. Pasanen, M. Aittasalo et al., “Preventing [78] C. N. Hales and D. J. P. Barker, “Type 2 (non-insulin-depend- excessive weight gain during pregnancy—a controlled trial in ent) diabetes mellitus: the thrifty phenotype hypothesis,” Dia- primary health care,” European Journal of Clinical Nutrition, betologia, vol. 35, no. 7, pp. 595–601, 1992. vol. 61, no. 7, pp. 884–891, 2007. [79] IOM, Nutrition During Pregnancy—Part I, Weight Gain,Na- [64] K. Gray-Donald, E. Robinson, A. Collier, K. David, L. Renaud, tional Academy Press, Washington, DC, USA, 1990. and S. Rodrigues, “Intervening to reduce weight gain in pregnancy and gestational diabetes mellitus in Cree communities: an evaluation,” Canadian Medical Association Journal, vol. 163, no. 10, pp. 1247–1251, 2000. [65] I. Guelinckx, R. Devlieger, P. Mullie, and G. Vansant, “Effect of lifestyle intervention on dietary habits, physical activity, and gestational weight gain in obese pregnant women: a randomized controlled trial,” American Journal of Clinical Nutrition, vol. 91, no. 2, pp. 373–380, 2010. [66] E. S. Lindholm, M. Norman, C. P. Kilander, and D. Altman, “Weight control program for obese pregnant women,” Acta Obstetricia et Gynecologica Scandinavica, vol. 89, no. 6, pp. 840–843, 2010. [67] L. A. H. Haakstad and K. Bø, “Effect of regular exercise on prevention of excessive weight gain in pregnancy: a randomised controlled trial,” European Journal of Contraception and Repro- ductive Health Care, vol. 16, no. 2, pp. 116–125, 2011. [68] S. L. Nascimento, F. G. Surita, M. A.ˆ Parpinelli, S. Siani, and J. L. Pinto e Silva, “The effect of an antenatal physical exercise programme on maternal/perinatal outcomes and quality of life in overweight and obese pregnant women: a randomised clinical trial,” British Journal of Obstetrics and Gynaecology, vol. 118, no. 12, pp. 1455–1463, 2011. [69] R. Barakat, A. Lucia, and J. R. Ruiz, “Resistance exercise training during pregnancy and newborn’s birth size: a randomised controlled trial,” International Journal of Obesity, vol. 33, no. 9, pp. 1048–1057, 2009. [70]S.R.Cavalcante,J.G.Cecatti,R.I.Pereira,E.P.Baciuk, A. L. Bernardo, and C. Silveira, “Water aerobics II: maternal body composition and perinatal outcomes after a program for low risk pregnant women,” Reproductive Health,vol.6,no.1, article 1, 2009. [71] S. Yeo, “Adherence to walking or stretching, and risk of preeclampsia in sedentary pregnant women,” Research in Nursing and Health, vol. 32, no. 4, pp. 379–390, 2009. [72] R. Barakat, Y. Cordero, J. Coteron, M. Luaces, and R. Montejo, “Exercise during pregnancy improves maternal glucose screen Hindawi Publishing Corporation Journal of Pregnancy Volume 2012, Article ID 812094, 8 pages doi:10.1155/2012/812094

Research Article Differential Effects of Chronic Pulsatile versus Chronic Constant Maternal Hyperglycemia on Fetal Pancreatic β-Cells

Mackenzie S. Frost,1 Aqib H. Zehri,2 Sean W. Limesand,2 William W. Hay Jr.,3 and Paul J. Rozance3

1 Department of Pediatrics, Drexel University College of Medicine, Philadelphia, PA, USA 2 Department of Animal Sciences, University of Arizona, Tucson, AZ 85719, USA 3 Department of Pediatrics, UCD Perinatal Research Center, University of Colorado School of Medicine, 13243 East 23rd Avenue, MS F441, Aurora, CO 80045, USA

Correspondence should be addressed to Paul J. Rozance, [email protected]

Received 6 August 2012; Accepted 2 October 2012

Academic Editor: Leena Hilakivi-Clarke

Copyright © 2012 Mackenzie S. Frost et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Constant maternal hyperglycemia limits, while pulsatile maternal hyperglycemia may enhance, fetal glucose-stimulated insulin secretion (GSIS) in sheep. However, the impact of such different patterns of hyperglycemia on the development of the fetal β-cell is unknown. We measured the impact of one week of chronic constant hyperglycemia (CHG, n = 6) versus pulsatile hyperglycemia (PHG, n = 5) versus controls (n = 7) on the percentage of the fetal pancreas staining for insulin (β-cell area), mitotic and apoptotic indices and size of fetal β-cells, and fetal insulin secretion in sheep. Baseline insulin concentrations were higher in CHG fetuses (P<0.05) compared to controls and PHG. GSIS was lower in the CHG group (P<0.005) compared to controls and PHG. PHG β- cell area was increased 50% (P<0.05) compared to controls and CHG. CHG β-cell apoptosis was increased over 400% (P<0.05) compared to controls and PHG. These results indicate that late gestation constant maternal hyperglycemia leads to significant β- cell toxicity (increased apoptosis and decreased GSIS). Furthermore, pulsatile maternal hyperglycemia increases pancreatic β-cell area but did not increase GSIS, indicating decreased β-cell responsiveness. These findings demonstrate differential effects that the pattern of maternal hyperglycemia has on fetal pancreatic β-cell development, which might contribute to later life limitation in insulin secretion.

1. Introduction Previous studies in pregnant sheep have tested the impact of chronic hyperglycemia on fetal β-cell function, using con- Fetuses exposed to chronic hyperglycemia secondary to mat- trolled experimental manipulation of maternal and fetal glu- ernal diabetes are prone to develop β-cell hyperplasia and cose concentrations and in vivo measurement of fetal insulin increased insulin secretion that underlie increased postnatal secretion to begin to determine mechanisms responsible for risk of exaggerated glucose-stimulated insulin secretion the impact of different patterns of maternal and fetal glucose (GSIS) and hyperinsulinemic hypoglycemia [1–6]. In con- concentration, as occur in diabetic pregnancies, on fetal GSIS trast, severe maternal diabetes during pregnancy that is [7–11]. Such studies demonstrated that constant maternal complicated by poor glycemic control and vasculopathy and hyperglycemia produced an initial increase in fetal glucose is further associated with intrauterine growth restriction and insulin concentrations [9], but over eight to ten days of (IUGR) results in decreased fetal pancreatic β-cell area and constant high maternal and fetal glucose concentrations the GSIS [3]. In both cases, not only are there short-term compli- fetal insulin concentrations returned to normal [8, 9]. Fur- cations for these offspring, but also long-term consequences. thermore, after ten days of constant maternal hyperglycemia, These long-term consequences for the offspring include an fetal GSIS and arginine-stimulated insulin secretion (ASIS) increased risk of developing adult onset diabetes [3]. were decreased compared to normal fetuses [8, 9]. 2 Journal of Pregnancy

More variable results were obtained when chronic hyper- arterial plasma glucose concentration measurements (con- glycemia was limited to three one-hour pulses each day stant hyperglycemia, CHG, n = 6); (3) basal D70 infusion (pulsatile hyperglycemia, akin to meal-associated hyper- at a rate adjusted to increase maternal plasma glucose 20% glycemia); in these studies over a similar period of 10–11 with additional 60 minute infusions of D70 three times a days, fetal GSIS increased [9]. However, a subsequent study day (8 AM, 2 PM, and 8 PM) at a rate targeted to increase demonstrated that changes in fetal GSIS following chronic maternal arterial plasma glucose concentrations 80% higher pulsatile maternal hyperglycemia depended on the degree than controls (pulsatile hyperglycemia, PHG, n = 5) [9]. and duration of the increased glucose concentrations [12]. These treatments were maintained for seven days. Maternal Despite such striking differences in fetal GSIS in response glucose concentrations were checked at least twice daily, with to different patterns of maternal and fetal glucose con- additional measurements in the PHG group. On day seven of centrations over extended periods, these studies did not the infusion period maternal and fetal arterial blood acid- provide insight into mechanisms responsible for such unique base balance, oxygen (PaO2,SaO2, and arterial blood O2 changes in fetal pancreatic function. Specifically, there has content), and carbon dioxide (PaCO2)weremeasured.At not been any morphological evaluation of the fetal pancreas baseline and throughout the infusion fetal arterial plasma in response to such different hyperglycemic patterns. It is glucose and insulin concentrations were measured. not known, for example, whether chronic constant maternal hyperglycemia, even confined to the latter portion of gesta- 2.3. In Vivo Fetal Insulin Secretion. On day seven of the β tion, will lead to or directly produce other signs of fetal - maternal infusions GSIS was measured with a square wave β cell toxicity besides decreased GSIS, such as increased -cell fetal hyperglycemic clamp followed by an arginine bolus β apoptosis or decreased pancreatic -cell area. It also is not to measure glucose potentiated ASIS [13]. Baseline arterial known how the variability in fetal GSIS following chronic blood samples for fetal arterial plasma glucose and insulin β pulsatile maternal hyperglycemia relates to fetal -cell area, concentrations were drawn at −25, −15, and −5minutes apoptosis, mitosis, or size. (relative to initiation of the glucose clamp at minute 0). Fetal Therefore, we undertook the studies reported herein to arterial blood also was sampled at −25 and −15 for acid- test whether chronic constant fetal hyperglycemia for seven base balance, PaO ,SaO, arterial blood O content, and β 2 2 2 days would decrease fetal pancreatic -cell area in association PaCO measurements. The fetal hyperglycemic clamp was β 2 with decreased -cell mitosis and increased apoptosis. We performed with a direct fetal glucose infusion adjusted to also tested whether chronic pulsatile maternal hyperglycemia double fetal arterial plasma glucose concentrations [9, 13]. β (limited to seven days) would result in increased -cell area At 125 minutes an infusion of arginine (261 mg in 5 mL) β due to increased -cell mitosis and decreased apoptosis. was administered over four minutes to measure ASIS. Blood samples were drawn at 5, 10, 20, 30, 60, 90, 120, 130, 135, 145, and 155 minutes to measure fetal arterial plasma glucose 2. Materials and Methods and insulin concentrations. Blood was sampled at minute 60, 90, and 120 for acid-base balance, PaO2,SaO2, arterial blood 2.1. Animal Preparation. Studies were conducted in preg- O2 content, and PaCO2 measurements. After minute 155 the nant Columbia-Rambouillet sheep carrying single fetuses fetal glucose infusion was stopped and the maternal glucose in compliance with the Institutional Animal Care and Use or saline infusions continued overnight to allow the fetus to Committee, University of Colorado Denver at the Perinatal return to pre-GSIS study conditions. Research Center in Aurora, CO. This laboratory is accredited by the National Institutes of Health, the United States Department of Agriculture, and the American Association 2.4. Biochemical Analysis. Wholebloodwascollectedin EDTA-coated syringes and immediately centrifuged for Accreditation of Laboratory Animal Care. Sheep under- ◦ went surgery at 118–122 days gestation (term = 148 days). (14,000 g) for 3 min at 4 C. Plasma was removed and Maternal catheters were placed into the femoral artery and the glucose and lactate concentrations were immediately vein via a groin incision. Fetal catheters were placed into the determined using the YSI model 2700 select biochemistry abdominal aorta and inferior vena cava via pedal incisions analyzer (Yellow Springs Instruments, Yellow Springs, OH) − ◦ as previously described [13]. Fetal and maternal catheters [13]. The remainder of the plasma was stored at 70 Cfor were flushed daily with 1.5 mL (fetus) or 3 mL (maternal) insulin measurements which was by ELISA (Alpco; inter- heparinized saline solution (50 U/mL heparin in 0.9% NaCl and intra-assay CV’s: 2.9 and 5.6%) [13]. For O2,CO2,pH, and hematocrit concentrations whole blood was collected in in H20). Animals were allowed a minimum of three recovery days prior to initiation of experimental infusions. heparinized syringes, and concentrations were immediately determined using an ABL 520 analyzer (Radiometer, Copenhagen, Denmark). Oxygen content of the blood was 2.2. Experimental Protocol. Animals were randomly assigned calculated by the ABL 520 analyzer [13]. to one of three study groups and managed as previously described: (1) euglycemia with saline infusion (control, 2.5. Organ Isolation. Necropsies and organ isolation were n = 7); (2) continuous 70% dextrose (w/v, D70) infusion performed the day after measurement of in vivo insulin adjusted to increase maternal arterial plasma glucose con- secretion as previously described [13]. The splenic portion centrations80%basedonanaverageoftwicedailymaternal of the pancreas was fixed overnight in 4% Paraformaldehyde Journal of Pregnancy 3

(w/v) in Phosphate-buffered Saline (PBS), and then trans- 3. Results ferred to 70% ethanol (v/v) until it was paraffinembedded. The method by which we obtained the pancreas precluded 3.1. Maternal Parameters during Treatment. The infusion . ± . measurement of pancreatic weight. rate of D70 in the CHG group started at 13 2 1 1gm/hr and increased significantly over the seven day study period (P = 0.0003). The maximum rate was on day five, 19.3 ± . 2.6. Fetal Pancreatic Histology. Paraffinembeddedtissue 1 2 gm/hr, at which point the rate decreased slightly to a final . ± . sections (5 μm) were cut at 100 μm intervals from the splenic rate of 17 9 2 2 gm/hr. The chronic infusion rate in the . ± portion of the pancreas. β-cell area, apoptosis, and mitosis PHG group did not change over time and averaged 1 1 . were determined as previously described with slight modifi- 0 1 gm/hr. The rate of the one hour bolus infusions also did . ± . cation in the dewaxing and antigen retrieval steps for sections not change over time and averaged 14 7 1 4 gm/hr. Maternal evaluated for β-cell area only [14–16]. The modifications arterial plasma glucose concentrations were significantly were that dewaxing was performed by washing slides twice increased in the CHG and PHG groups compared to P< . in EZ Dewax tissue deparaffinization solution (Biogenex) controls throughout the infusion ( 0 0001, Figure 1(a)). for five minutes. Slides were then placed in water to rinse Furthermore, arterial plasma glucose concentrations in the for five minutes followed by washing in Supersensitive Wash CHG group were significantly increased compared to the P< . Buffer (Biogenex) for 5 minutes. Antigen unmasking was PHG group ( 0 0001, Figure 1(a)). To produce “pulsatile” performed by placing the slides in Antigen Unmasking hyperglycemia and model meal associated hyperglycemia Solution (Vector) and heating to 95◦Cfor20minutes. in pregnant women, the PHG group received a 60 minute This was followed by blocking with 1.5% normal donkey dextrose infusion that increased their arterial plasma glucose serum in PBS (v/v) for 30 min. Mature endocrine hormone+ concentrations to an average of 114 ± 3 mg/dL three times a cells were identified with the following primary antibodies day. There was a slight, but statistically significant, increase in diluted in blocking buffer: guinea pig anti-porcine insulin arterial pH in both the CHG and PHG groups compared to P ≤ . (Dako, Carpinteria CA, 1 : 500) or mouse anti-human insulin controls ( 0 028, Supplementary Table available online at (Abcam, Cambridge UK, 1 : 1000), mouse monoclonal anti- doi:10.1155/2012/812094), but no other changes were noted human glucagon, (Sigma-Aldrich, St. Louis MO, 1 : 500), in maternal acid-base balance, O2 values (PaO2,SaO2,and rabbit anti-human somatostatin, (Dako, 1 : 500), rabbit anti- arterial blood O2 content), or PaCO2. human pancreatic polypeptide, (Dako, 1 : 500). Sections were incubated at 4◦C overnight and immunocomplexes detected ffi 3.2. Fetal Parameters during Treatment. Fetal arterial plasma the next day with a nigy purified secondary antiserum glucose concentrations were significantly increased in the conjugated to Rhodamine Red (Jackson ImmunoResearch CHG group compared to both control and PHG fetuses Laboratories, West Grove PA), 7-amino-4methylcoumarin- throughout the maternal dextrose infusion period (P< 3-acteic acid (AMCA, Jackson ImmunoResearch Laborato- 0.0001, Figure 1(b)). There were no differences in fetal arteries), and AlexaFluor 488 (Molecular Probes, Eugene OR) rial plasma glucose concentrations between PHG and control [14]. β-cell mitosis and apoptosis were determined as the + groups. Fetal arterial plasma insulin concentrations did not percentage of insulin cells which were also positive for change in the control and PHG groups (Figure 1(c)). In the phosphorylated Histone H3 or terminal deoxynucleotidyl CHG group insulin concentrations increased on day two and transferate (TdT)-mediated dUTP nick translation end remained increased throughout the infusion compared to labeling (TUNEL), respectively, as previously described [14]. day one (baseline), (P ≤ 0.028, Figure 1(c)). There were β + -cell size was determined by dividing the insulin area small but statistically significant changes in the fetal blood by number of nuclei within that area as determined with pH, PaCO2, and lactate concentrations (Supplementary 4 ,6 diamidino-2-phenylindole (DAPI, Vector Laboratories, Table). Burlingame CA) [14].

3.3. Fetal Insulin Secretion. Fetal insulin secretion was mea- 2.7. Statistical Analysis. Statistical analysis was performed sured with a square-wave fetal hyperglycemic clamp on day using SAS version 9.1 or GraphPad Prism 4.0 for Windows. seven. By design, fetal glucose concentrations were doubled Results are expressed as mean ± SEM. A mixed models in each group. This led to greater absolute increases in ANOVA with a random animal term to account for repeated fetal glucose concentrations in the CHG group compared measurements made within an animal was performed to to the other groups (P<0.0001, Figure 2(a)). For all determine effects of treatment group (control, CHG, or groups, fetal arterial plasma glucose concentrations were PHG), time (days of treatment or minutes of hyperglycemic significantly greater than baseline beginning at minute five glucose clamp), and treatment-time interactions for all and lasting throughout the hyperglycemic clamp period in vivo measurements except for those which were only (P ≤ 0.005, Supplementary Figure). Fetal arterial plasma measured at the end of the glucose clamp infusions. Measure- insulin concentrations increased in all three groups during ments made only once, including fetal weight, fetal length, the hyperglycemic clamps (Figure 2(b)). However, insulin and pancreatic β-cell area, size, mitosis, and apoptosis, were concentrations in the CHG group did not become statis- compared using a one way ANOVA or the Kruskal-Wallis tically greater than baseline concentrations until minute test. 90 of the hyperglycemic clamp (Supplementary Figure), 4 Journal of Pregnancy

∗∗ 70 ∗ 150 60 ## 50

100 40

50 20 Fetal glucose (mg/dL) glucose Fetal Maternal glucose (mg/dL) glucose Maternal 10

0 0 123456 123456 Day Day (a) (b) 0.8

0.7 # ∗ 0.6 ∗ 0.5

0.4

0.3

Fetal insulin (ng/mL) Fetal 0.2

0.1

0 123456 Day (c)

Figure 1: Maternal and fetal glucose and fetal insulin concentrations. Maternal (a) and fetal (b) arterial plasma glucose and fetal arterial plasma insulin (c) concentrations were measured throughout the infusion in CHG (n = 6, ), PHG (n = 5, ), and control animals (n = 7, ). ∗∗ indicates a significant difference between both CHG and PHG from controls for maternal glucose concentrations, P<0.0001. ## indicates a significant difference between CHG and PHG animals for maternal glucose concentrations, <0.0001. ∗ indicates a significant difference between CHG and both PHG and control groups for fetal arterial glucose and insulin concentrations, P = 0.026. # indicates a significant increase in CHG fetuses only compared to their baseline (Day 1) for fetal arterial insulin concentrations, P = 0.028.

despite significantly higher plasma glucose concentrations CHG, 3.51 ± 1.94 ng/mL PHG). Although fetal acid-base during the hyperglycemic clamp and a significantly greater balance, blood oxygen values (PaO2,SaO2, and arterial blood increase in glucose concentrations between the basal and O2 content), and PaCO2 changed in all groups during the hyperglycemic clamp period. Basal insulin concentrations in hyperglycemic clamp, differences among the groups were the CHG fetuses were higher than in the PHG and Control minimal (Supplementary Table). groups. Insulin secretion in the CHG group, determined as the difference between the mean basal and clamp period insulin concentrations, was less than in the other groups 3.4. Fetal Measurements, Organ Weights, and Histology of during the hyperglycemic clamp (P = 0.0002, Figure 2(b)). the Fetal Pancreas. Fetal measurements and organ weights Nevertheless, mean insulin concentrations during the hyper- were not different among the groups (Table 1). There was no glycemic clamps were not different among the three groups. difference in the β-cell area between control and CHG fetal Insulin secretion in PHG fetuses was not different from pancreases, but there was a significant increase in the β-cell control fetuses. Arginine-stimulated insulin secretion was area of the PHG pancreases (P = 0.021, Figure 3(a)). β-cell not statistically different among the three groups. All groups mitosis was not different among the groups (Figure 3(b)), had maximum insulin concentrations at five minutes post- but apoptosis was significantly increased in the CHG group arginine bolus (1.68±0.24 ng/mL control, 2.61±0.59 ng/mL (P = 0.021, Figure 3(c)). β-cell size (57.4 ± 3.5 μm2 control, Journal of Pregnancy 5

125 • 1 ∗ ∗∗ # 100 ∗∗ 0.75 ∗

75 0.5

50 ∗ ∗ Insulin (ng/mL) Insulin Glucose (mg/dL) Glucose 0.25 25

0 0 Control CHG PHG Control CHG PHG Basal hyperglycemic clamp Basal hyperglycemic clamp (a) (b)

Figure 2: Fetal glucose-stimulated insulin secretion. A square-wave fetal hyperglycemic clamp was used to test insulin secretion. The average basal (minute −25, −15, −5) and hyperglycemic clamp (minute 60, 90, 120) glucose (a) and insulin concentrations (b) in control (n = 7, black bars), CHG (n = 6, white bars), and PHG (n = 5, gray bars) fetuses are shown. ∗, ∗∗ indicate a significant difference between basal and hyperglycemic clamp concentrations within a treatment group, P = 0.029, P<0.0001, respectively. • indicates a significant increase in the incremental change of glucose concentrations in CHG compared to both PHG and controls, P<0.0001. # indicates a significant decrease in the incremental change of insulin concentrations in CHG compared to both PHG and controls, P = 0.0002.

59.3 ± 4.4 μm2 CHG, 53.4 ± 0.8 μm2 PHG) was not different with decreased GSIS seen in more recent studies of chronic among groups. pulsatile maternal hyperglycemia of a longer duration [12]. These unique observations, therefore, demonstrate cellular 4. Discussion morphological and developmental changes that precede different patterns of insulin secretion that can be produced The pattern of chronic maternal and thus fetal hyperglycemia by pulsatile versus constant hyperglycemia, providing new is fundamental for determining the regulation of fetal β-cell insight into the pathogenesis of abnormal patterns of fetal development and function throughout the second half of insulin secretion that can occur in type 1 and type 2 human gestation during which such development normally matures. diabetic pregnancies. In order to determine the differential effects of chronic In our studies, chronic constant maternal hyperglycemia constant versus pulsatile hyperglycemia, aimed to mimic pat- (CHG group) for one week increased basal fetal insulin con- terns of glucose concentration common to pregnant women centrations. These data are consistent with previous studies with diabetes, on the morphological and functional devel- which show that over a one week period, fetal insulin concen- opment of fetal pancreatic β-cells, we infused dextrose to trations initially increase with constant fetal hyperglycemia produce different patterns of maternal hyperglycemia over [7, 9, 11]. This is followed by a variable but progressive one week in late gestation in pregnant sheep. There are decline in insulin concentrations. In some fetuses insulin several important novel findings of this study. First, chronic concentrations returned to baseline by day six to seven and constant maternal hyperglycemia decreased fetal GSIS and for others this return to baseline required several more days increased fetal β-cell apoptosis, even though fetal β-cell [9]. More prolonged constant fetal hyperglycemia ultimately size, the degree of fetal β-cell mitosis, and the proportional results in a decline in basal fetal insulin concentrations to fetal β-cell area within the pancreas were not decreased. even lower values than those seen in normal fetuses [8, 9]. These unique and important observations indicate that Also consistent with previous studies, we found that apoptosis and decreased GSIS are early manifestations of chronic constant maternal hyperglycemia for seven days fetal β-cell glucotoxicity and that decreased fetal β-cell decreased fetal GSIS. We have extended these observations by glucose responsiveness is independent of decreased β-cell demonstrating that the pancreatic β-cell area and individual area and precedes the decrease in basal insulin secretion β-cell size are unchanged. These results show for the first previously observed following ten days of chronic constant time that decreased β-cell glucose responsiveness is a mech- maternal hyperglycemia [8, 9]. Second, chronic pulsatile anism for decreased fetal GSIS following chronic constant hyperglycemia increased the fetal pancreatic β-cell area, maternal hyperglycemia. We also have shown increased β- indicating that this cellular change might underlie increased cell apoptosis, another indicator of fetal β-cell glucotoxicity. GSIS that has been seen in other studies of chronic pulsatile Future experiments will determine if a longer exposure to maternal hyperglycemia [9]. However, because we did not CHG also reduces β-cell area, which would be expected find increased GSIS in the PHG fetuses despite the increase given increased β-cell apoptosis in this model, and would in β-cell area, the current results demonstrate a degree of help explain decreased basal insulin concentrations following β-cell dysfunction in this PHG group that is consistent longer durations of CHG [8, 9]. 6 Journal of Pregnancy

Table 1:FetalAge,Weights,andLength.

Constant Pulsatile Control Hyperglycemia Hyperglycemia Gestational Age (days) 134.7 ± 0.7 135.0 ± 0.6 133.6 ± 0.9 Fetal Weight (gm) 3593 ± 138 3601 ± 290 3678 ± 223 CrownRumpLength(cm) 48.8 ± 1.048.6 ± 0.850.1 ± 2.0 Liver (gm) 108.5 ± 5.2 117.9 ± 14.6 123.8 ± 6.6 Heart (gm) 29.8 ± 1.630.4 ± 2.430.7 ± 1.8 Lung (gm) 116.3 ± 4.8 100.0 ± 9.6 122.3 ± 10.3 Kidney (gm) 20.6 ± 1.525.5 ± 2.224.0 ± 1.1 Spleen (gm) 8.2 ± 0.67.4 ± 1.611.8 ± 2.2 Brain (gm) 46.4 ± 0.843.2 ± 0.844.2 ± 0.9 Carcass (gm) 2896 ± 92 2841 ± 245 2873 ± 167 Sex (% Female) 43 50 40

8 ∗ 1 7

6 0.75

4 0.5 -cell area (%) area -cell

β 3 -cell mitosis (%) mitosis -cell β 2 0.25

0 0 Control CHG PHG Control CHG PHG (a) (b) 3

∗

1 -cell apoptosis (%) apoptosis -cell β

0 Control CHG PHG (c)

Figure 3: Fetal pancreatic β-cell area, mitosis, and apoptosis. Fetal pancreatic β-cell area (a), mitosis (b), and apoptosis (c) were measured in control (n = 7, black bars), CHG (n = 6, white bars), and PHG (n = 5, gray bars) fetal pancreases. ∗ indicates a signiﬁcant diﬀerence, P = 0.021.

An alternative explanation for decreased GSIS in the normal fetuses [15]. However, chronic constant fetal hyper- CHG fetuses is that there may be a plateau insulin concentra- glycemia for longer than eight days results in an unequivocal tion, representing a plateau in fetal insulin secretion and/or decrease in GSIS and ASIS, not explainable by plateau a balance between insulin secretion and insulin clearance, glucose-stimulated insulin concentrations [8, 9]. These which cannot be exceeded regardless of further increases results from longer duration experiments, as well as the in fetal glucose concentrations, as recently demonstrated in increased β-cell apoptosis demonstrated in the current study, Journal of Pregnancy 7 strongly support toxicity of the fetal β-cell following constant allow for programming of unique β-cell fates that might maternal hyperglycemia of only seven days duration. produce later life (even lifelong) limitations in function. In the PHG fetuses we found an increase in the pancreatic β -cell area but not increased GSIS, as was previously Acknowledgments observed after ten days [9]. Besides differences in the duration of chronic hyperglycemia between the current and M. S. Frost was supported by NIH Grant T32HD07186. S. previous study, there were differences in the fetal glucose W. Limesand was supported by NIH Grant R01DK084842. concentrations between studies. In the current study glucose P. J. Rozance was supported by NIH Grants R01DK088139 concentrations were increased by approximately 15%, which and K08HD060688, a Junior Faculty Research Development was not statistically significant, despite a significant 20% Award from the Center for Women’s Health Research, increase in maternal plasma arterial glucose concentrations. University of Colorado Denver, and a Pilot and Feasibility In the previous study fetal plasma arterial glucose concen- Award from the Diabetes and Endocrinology Research trations were increased approximately 20–25% [9]. Fetal Center, University of Colorado Denver (P30DK057516, PI: insulin concentrations also were higher in the previous Hutton). Histological support was provided by the Diabetes study [9]. Perhaps a more important difference is the and Endocrinology Research Center, University of Colorado magnitude of the pulsatile hyperglycemia, which in the Denver (P30DK057516, PI: Hutton). The content is solely the current study was 77% greater than controls in contrast to responsibility of the authors and does not necessarily repre- a 60% greater concentration in the previous study [9]. While sent the official views of the National Institutes of Health. such differences are small, a recent study tested the impact of the magnitude of the hyperglycemic pulse on GSIS and References found that higher pulsatile glucose concentrations resulted in inhibition of fetal GSIS [12]. Normal GSIS in the current [1] R. L. Naeye, “Infants of diabetic mothers: a quantitative, mor- PHG fetuses, despite their increased β-cell area, indicates a phologic study,” Pediatrics, vol. 35, pp. 980–988, 1965. degree of β-cell dysfunction, although one not as severe as in [2] M. Osler and J. Pedersen, “The body composition of newborn the CHG fetuses. infants of diabetic mothers,” Pediatrics, vol. 26, pp. 985–992, Despite an increase in β-cell area in the PHG fetuses, we 1960. did not observe changes in β-cell size, mitosis, or apoptosis. [3] F. A. Van Assche, K. Holemans, and L. Aerts, “Long-term con- It is possible that increased mitosis or decreased apoptosis sequences for offspring of diabetes during pregnancy,” British may have been present before we obtained the pancreas as Medical Bulletin, vol. 60, pp. 173–182, 2001. possible explanations for these observations. Furthermore, [4] P. J. Rozance and W. W. Hay, “Hypoglycemia in newborn the variability in rates of β-cell mitosis may have precluded infants: features associated with adverse outcomes,” Biology of demonstrating a statistically significant higher rate in the the Neonate, vol. 90, no. 2, pp. 74–86, 2006. PHG fetuses despite a 40% higher mean rate compared to [5] L. Aerts and F. A. Van Assche, “Rat foetal endocrine pancreas control fetuses. We also did not measure the rate of β-cell in experimental diabetes,” Journal of Endocrinology, vol. 73, no. 2, pp. 339–346, 1977. neogenesis, which if increased, despite being very low near β [6] R. Schwartz, P. A. Gruppuso, K. Petzold, D. Brambilla, V. term, would tend to increase the -cell population [14, 16]. Hillesmaa, and K. A. Teramo, “Hyperinsulinemia and macro- Future studies will be required to obtain pancreatic tissue somia in the fetus of the diabetic mother,” Diabetes Care, vol. from PHG animals earlier in the hyperglycemic infusion 17, no. 7, pp. 640–648, 1994. period to determine the relative contributions of decreased [7] J. M. Bassett and D. Madill, “Influence of prolonged glucose apoptosis, increased mitosis, or neogenesis, and to determine infusions on plasma insulin and growth hormone concentra- how these processes directly and mechanistically affect fetal tions of foetal lambs,” Journal of Endocrinology, vol. 62, no. 2, β-cell insulin secretion, at baseline and in response to glucose pp. 299–309, 1974. stimulation and stimulation by other secretagogues. [8]T.D.Carver,S.M.Anderson,P.A.Aldoretta,A.L.Esler,and W. W. Hay Jr., “Glucose suppression of insulin secretion in 5. Conclusions chronically hyperglycemic fetal sheep,” Pediatric Research, vol. 38, no. 5, pp. 754–762, 1995. In summary, chronic constant hyperglycemia for seven [9]T.D.Carver,S.M.Anderson,P.W.Aldoretta,andW.W. ff days decreased GSIS and increased fetal β-cell apoptosis, Hay Jr., “E ect of low-level basal plus marked “pulsatile” though the proportional fetal β-cell area within the pancreas hyperglycemia on insulin secretion in fetal sheep,” American Journal of Physiology, vol. 271, no. 5, pp. E865–E871, 1996. was preserved. In contrast, chronic pulsatile hyperglycemia increased the proportional fetal β-cell area within the [10] J. P. Lips, H. W. Jongsma, J. Crevels, and T. K. A. B. Eskes, “Chronic hyperglycemia and insulin concentrations in fetal pancreas. Based on these results, as well as previous studies ff lambs,” American Journal of Obstetrics and Gynecology, vol. [8, 9, 12], we speculate that di erences found for human 159, no. 1, pp. 247–251, 1988. newborn GSIS following a diabetic pregnancy are due to [11] A. F. Philipps, P. J. Porte, and S. Stabinsky, “Effects of chronic ff subtle di erences in fetal glucose concentrations [17–20]. fetal hyperglycemia upon oxygen consumption in the ovine Furthermore, these studies show differential effects that the uterus and conceptus,” Journal of Clinical Investigation, vol. 74, pattern of maternal and fetal hyperglycemia has on fetal β- no. 1, pp. 279–286, 1984. cells, which are plastic in their developmental and functional [12] A. S. Green, X. Chen, A. R. Macko et al., “Chronic pul- capacities even near the end of gestation. Such changes could satile hyperglycemia reduces insulin secretion and increases 8 Journal of Pregnancy

accumulation of reactive oxygen species in fetal sheep islets,” Journal of Endocrinology, vol. 212, no. 3, pp. 327–342, 2012. [13] P. J. Rozance, S. W. Limesand, J. S. Barry, L. D. Brown, and W. W. Hay Jr., “Glucose replacement to euglycemia causes hypoxia, acidosis, and decreased insulin secretion in fetal sheep with intrauterine growth restriction,” Pediatric Research, vol. 65, no. 1, pp. 72–78, 2009. [14] S. W. Limesand, J. Jensen, J. C. Hutton, and W. W. Hay Jr., “Diminished β-cell replication contributes to reduced β-cell mass in fetal sheep with intrauterine growth restriction,” Ame- rican Journal of Physiology, vol. 288, no. 5, pp. R1297–R1305, 2005. [15] A. S. Green, A. R. Macko, P.J. Rozance et al., “Characterization of glucose-insulin responsiveness and impact of fetal number and sex diﬀerence on insulin response in the sheep fetus,” American Journal of Physiology, vol. 300, no. 5, pp. E817–E823, 2011. [16] L. Cole, M. Anderson, P. B. Antin, and S. W. Limesand, “One process for pancreatic β-cell coalescence into islets involves an epithelial-mesenchymal transition,” Journal of Endocrinology, vol. 203, no. 1, pp. 19–31, 2009. [17] T. E. Isles, M. Dickson, and J. W. Farquhar, “Glucose tolerance and plasma insulin in newborn infants of normal and diabetic mothers,” Pediatric Research, vol. 2, no. 3, pp. 198–208, 1968. [18] K. C. King, A. Oliven, and S. C. Kalhan, “Functional enteroin- sular axis in full-term newborn infants,” Pediatric Research, vol. 25, no. 5, pp. 490–495, 1989. [19] S. S. Obenshain, P. A. Adam, K. C. King et al., “Human fetal insulin response to sustained maternal hyperglycemia,” The New England Journal of Medicine, vol. 283, no. 11, pp. 566– 570, 1970. [20]R.S.Pildes,R.J.Hart,R.Warrner,andM.Cornblath,“Plasma insulin response during oral glucose tolerance tests in newborns of normal and gestational diabetic mothers,” Pediatrics, vol. 44, no. 1, pp. 76–83, 1969.