Journal of Human Hypertension (2012) 26, 405–419 & 2012 Macmillan Publishers Limited All rights reserved 0950-9240/12 www.nature.com/jhh REVIEW Developmental origins of health and disease: experimental and human evidence of fetal programming for

ML de Gusma˜o Correia, AM Volpato, MB A´ guila and CA Mandarim-de-Lacerda Department of Anatomy, Institute of Biology, State University of Rio de Janeiro, Rio de Janeiro, Brazil

The concept of developmental origins of health and rodents are associated with fetal and neonatal program- disease has been defined as the process through which ming for metabolic syndrome. In this paper, recent the environment encountered before birth, or in infancy, experimental models and human epidemiological stu- shapes the long-term control of tissue physiology and dies providing evidence for the fetal programming homeostasis. The evidence for programming derives associated with the development of metabolic syndrome from a large number of experimental and epidemiologi- and related diseases are revisited. cal observations. Several nutritional interventions dur- Journal of Human Hypertension (2012) 26, 405–419; ing diverse phases of pregnancy and lactation in doi:10.1038/jhh.2011.61; published online 23 June 2011

Keywords: fetal programming; developmental origins of disease; metabolic syndrome;

Introduction Metabolic syndrome is an aggregation of risk factors (that is, overweight, abdominal obesity, The concept of developmental origins of health and hypertension, dyslipidaemia and glucose intoler- disease (a.k.a. fetal programming) was first proposed by 1 2 ance) that strongly correlates with cardiovascular David Barker and later defined by Alan Lucas as the disease. Several nutritional interventions during permanent response of an organism to a stimulus or diverse phases of pregnancy and lactation in rodents insult during critical periods of development. The are associated with fetal and neonatal programming concept of programming was later expanded and for metabolic syndrome. In humans, both small and defined as the process through which the environment large for gestational age newborns have been shown encountered before birth, or in infancy, shapes the long- 3 to develop features of metabolic syndrome during term control of tissue physiology and homeostasis. growth and adulthood. Such adaptive responses often cause several diseases in In this paper, recent experimental and epidemiolo- adult life. The evidence for fetal programming comes gical studies providing evidence for the fetal pro- from a large number of experimental and epidemiolo- 4 gramming associated with the development of gical observations. Barraclough demonstrated that metabolic syndrome and related diseases were re- neonatal administration of testosterone to female rats viewed. First, associations between birth weight and causes ovulation failure and sterility, which are not the increase in cardiovascular risk in humans based observed if the hormone is given later in life. 5 on epidemiological observations were discussed. Importantly, Rose et al. firstreportedthattheriskof Second, we revisited evidence linking maternal low- ischaemic heart disease doubled in human subjects protein diet with development of metabolic syndrome whose siblings were still born or died in early child- in rodents. Finally, recent experimental studies hood, suggesting that adverse exposures during preg- showing the effects of maternal overnutrition in the nancy could cause diseases in adulthood. adult offspring were presented.

Cardiovascular risk and fetal program- Correspondence: Professor ML de Gusma˜o Correia, Laboratory of ming in humans: early evidence Morphology, Department of Anatomy, Institute of Biology, State University of Rio de Janeiro, Boulevard 28 de Setembro, 87 fds, Several epidemiological reports have shown asso- 20551-030 Rio de Janeiro, RJ, Brazil. E-mail: [email protected] ciations between prenatal and postnatal exposures Received 13 May 2010; revised 20 February 2011; accepted 9 May and the development of metabolic syndrome in 2011; published online 23 June 2011 adult life. Although genomically imprinted genes Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 406 could potentially determine programming, several accelerate their growth rate if adequate nutrition is epigenetic mechanisms have been associated with provided after birth (that is, catch-up). This growth altered genetic expression and the development of pattern has been associated with an increased risk disease in adult life.6 for coronary heart disease and higher cardiovascular Times of famine caused by human conflicts have mortality.11 In line with these results and using low provided unique opportunities for the study of the birth weight as a surrogate of intra-uterine slow impact of undernourishment during pregnancy on growth, the Hertfordshire and Sheffield cohort of the development of disease in the adult progeny. neonates born between 1911 and 1930 has shown Two events are remarkable in this regard: the siege of that birth weights o2.5 kg were associated with Leningrad and the Dutch famine, which occurred increased coronary artery disease and stroke mor- during World War II. Stanner et al.7 have studied tality, and augmented prevalence of . adult subjects exposed to malnutrition during Indeed, 1 s.d. increase in birth weight reduces all- gestation and afterwards, and those only exposed cause mortality at the age of 75 years old by 0.86% after birth during the siege of Leningrad by the in men and women.12 These results appear to be German army between 1941 and 1944. These independent of social class but are potentiated in subjects were compared with control individuals women.12–15 born concurrently but outside the area of the siege and not exposed to starvation. It was found that obesity-related changes in blood pressure and Metabolic syndrome programming in endothelial dysfunction were more prevalent in small for gestational age humans the exposed groups. Nevertheless, glucose intoler- ance, dyslipidaemia, hypertension, obesity or cardi- Programming for metabolic syndrome in adulthood ovascular diseases were not more prevalent among of infants born small for gestational age could adults exposed to siege’s nutritional conditions. potentially increase cardiovascular risk. However, The Dutch famine occurred in the harsh winter of metabolic syndrome as a consequence of program- 1944–1945 after a food embargo imposed by Ger- ming has been seldom studied. Although low birth many in retaliation to a rail strike. Unlike the siege weight has been inconsistently linked with later of Leningrad, the Dutch famine was short (that is, 5 development of metabolic syndrome,16–18 a much to 7 months) and allowed the study of malnutrition larger number of studies have reported programming around fertilization or later in pregnancy. It has been for its main individual components. For instance, observed that malnutrition during the first and hypertension, the most prevalent component of second trimesters were associated with an increased metabolic syndrome, develops more often in small prevalence of obesity whereas exposure in the third newborns who catch-up between 1 and 5 years of trimester correlated with lower to normal weight in age. Interestingly, low birth weight was also corre- adult life.8,9 In addition, subjects exposed to mal- lated with increased adiposity in adulthood, which nutrition in early pregnancy exhibited atherogenic partly explains the development of systolic hyper- lipid profile, higher plasma fibrinogen, reduced tension in these subjects.19 The relevance of the concentrations of factor VII and higher risk of catch-up growth has been largely confirmed by the coronary heart disease. In contrast, the progeny United States Collaborative Perinatal Project, which exposed to famine in late or mid gestation had studied 455 000 pregnancies in an observational impaired glucose tolerance.10 Therefore, these re- multi-centre cohort between 1959 and 1974. In this sults indicate that malnutrition exposure during study, low birth weight was not associated with diverse periods of pregnancy increases the risk of increased risk of hypertension at the age of 7 years metabolic syndrome differentially in the adult off- old, except in infants who gained weight faster than spring, obesity and coronary heart disease being predicted by growth charts and were at increased more common in subjects exposed to malnutrition risk for higher systolic blood pressure.20 Moreover, during the first half of gestation, whereas insulin in one important clinical trial, small for gestational resistance appears to be more common during late age newborns were either assigned to standard or pregnancy exposure. high-protein diet. Infants in the high-protein diet Of note, the adult phenotypes of the progeny born exhibited faster weight gain, which was associated in Leningrad and Holland are substantially different with elevated diastolic, mean but not systolic blood given that glucose intolerance, dyslipidaemia, obe- pressures.21 The lack of differences in regard to sity or cardiovascular disease was less prevalent in systolic blood pressures in this study might have the first population. One possibility for these been due to the high-protein diet chosen in a differences is that infants born in Leningrad were clinical trial setting. Overall, increased adiposity exposed to malnutrition for a prolonged period of later in life in small for gestational age infants time after the war whereas those born in Holland appears to be a pivotal condition for the develop- recovered from nutrition deficits shortly after the ment of hypertension. liberation of Western Europe and might have Other mechanisms have been proposed for pro- experienced the so-called catch-up growth. Indeed, gramming of hypertension in humans including small babies exposed to intra-uterine malnutrition reduced number of nephrons and glomeruli along

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 407 with smaller kidney volume.22 Indeed, the odds sensitivity because of abnormal mitochondrial me- ratio for reduced creatinine clearance in low and tabolism and altered epigenetic regulation.36 very low birth weight babies reaching early adult- Furthermore, interactions between common genetic hood ranges between 1.65 and 2.4 as compared with variants and low birth weight increasing the risk of normal newborns, indicating increased risk of renal later development of glucose intolerance and type 2 dysfunction and, thus, hypertension in the former diabetes have been reported.37 population.23 Simonetti et al.,24 confirmed smaller Atherogenic lipid profile is another component of kidney size and lower glomerular filtration rate, metabolic syndrome that has been described in which were associated with salt sensitivity of blood adults born small. Specifically, at the age of 60 pressure in children born with low weight. Impor- years old, low birth weight was associated with tantly, catch-up growth, by increasing renal blood higher non-high-density lipoprotein cholesterol and flow, could further impair renal function in this apolipoprotein B whereas slow weight gain at 6 pathological setting.22 Abnormal hypothalamus– months and 2 years of age was correlated with pituitary–adrenal axis, altered autonomic function, increased non-high-density lipoprotein cholesterol increased cardiac sympathetic activation, decreased and lower high-density lipoprotein.38 Importantly, baroreflex sensitivity, endothelial dysfunction and low birth weight without catch-up at 2 years of age, increased metabolic efficiency have all been de- followed by rapid weight gain until the age of 11 scribed in small for gestational age newborns and years old correlated with increased risk for coronary could also contribute for the development of heart disease.39 Altogether, these studies highlight hypertension.22,25 Altogether, these mechanisms that early catch-up growth might be neutral or may partly explain increased blood pressure re- protective in regard to atherogenic lipid profile sponse (rather than resting blood pressure) to and confirmed earlier reports from the Hertfordshire psychosocial stress test in adults born prema- cohort, which showed that low weight at the age of turely.26 1-year old was associated with increased total and Obesity is another important component of the low-density lipoprotein cholesterol, and apolipo- metabolic syndrome that has been associated with protein B plasma concentrations in men.40 Never- fetal programming and the occurrence of catch-up in theless, evidence linking small birth weight and small for gestational age newborns. Indeed, catch-up programming for abnormal lipid profile is contro- growth between birth and 2 years of age predicts versial. Huxley et al.41 published a systematic central obesity and insulin resistance starting at the review of 58 papers involving 68 974 subjects and age of 4 years old.27–29 Interestingly, even non- observed that 1 kg lower birth weight is associated overweight children born small exhibit augmented with 2.0 mg dl–1 increment in total cholesterol in visceral adiposity, increased insulinaemia and lower later life, which might not confer a substantially plasma levels of high molecular weight adiponectin, higher cardiovascular risk in this population. suggesting that programmed endocrine abnormal- ities may determine the development of overweight and obesity.30 These findings are nevertheless con- Metabolic syndrome programming troversial given that, in a large retrospective study, in large for gestational age humans weight at 1-year old but not at birth was associated with body fat mass in adulthood, suggesting that The link between fetal programming for metabolic postnatal factors are stronger predictors for the syndrome in newborns large for gestational age (that development of obesity.31 Also, mother’s weight is, 44 kg) is weaker than that in low birth weight rather than infant’s body mass at birth could explain infants. Also, mechanisms potentially associated later development of obesity.32 with fetal programming of metabolic syndrome in Abnormal glucose metabolism is an important small for gestational age births are much better hallmark of metabolic syndrome. In this regard, the documented than those occurring in large for association of both low and high weight at birth with gestational age newborns. Children’s weight has development of type 2 diabetes is strong. In a meta- long been suspected to be a risk factor for adult analysis comprising 14 studies and 132 180 indivi- obesity but evidence is controversial,42,43 Likely due duals, the odds ratio for the development of type 2 to maternal overweight and reduced prevalence of diabetes ranged between 1.36 and 1.47, for high and smoking, the incidence of large for gestational age low birth weight, respectively.33 A recent systematic births has been increasing over the last decades.44 review confirmed these results even after adjust- Large for gestational infants tend to become obese ments for current body weight and socioeconomic children.45,46 Furthermore, high weight at birth status.34 Notably, the association between low predicted the development of overweight in weight at birth, insulin resistance and type 2 193 056 men at the age of 18 years old in a diabetes appears to be stronger in newborns who registry-based study in Sweden.47 Similarly, high remained thin until the age of 1-year old.35 Program- birth weight and accelerated child growth were also ming mechanisms determining glucose intolerance associated with adult obesity in registry and ques- and type 2 diabetes are obscure but may involve tionnaire-based study in Finland.48 However, in the pancreatic b-cell dysfunction and reduced insulin Muscatine Ponderosity Study, while adult obesity

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 408 often tracked obesity in infancy and adolescence, a onment might explain in part why low-protein diets substantial number of lean children also ended up consistently programme for hypertension,59 im- being obese adults and vice versa.49 Indeed, con- paired pancreas development, altered insulin re- founding factors like socioeconomic status could sponse to the oral glucose tolerance test and hepatic explain the association between child and adult insulin resistance in the offspring.60 obesity.42 It has been suggested that early poor growth The propensity of large for gestational age to results from the development of a thrifty phenotype develop metabolic syndrome is also arguable. In the in the malnourished fetus. Such phenotype would Health Professionals Follow-Up Study, 22 846 North predispose the offspring to develop obesity if over- American men self-reported medical histories and nutrition occurs after birth, enabling the organism to weight at birth. Although large for gestational age maximize energy conservation and storage should it newborns developed obesity more often, hyperten- continue to experience similar periods of nutrient sion and type 2 diabetes, two important components restriction after weaning.61,62 of metabolic syndrome, were only correlated with low weight at birth.50 These results contrast in part with reports showing that both low and high weight Programming of glucose and lipid metabolism at birth are associated with increased risk of type 2 In protein-deprived newborn rats, glucose intoler- diabetes and higher blood pressure in adolescents.51 ance is associated with diminished insulin secretion Metabolic syndrome, notably due to hypertension in response to the oral glucose loading test, which and high plasma triglycerides, was also more persists until adulthood, and with accelerated age- common in children born large for gestational age related islet cell insulin depletion.63 Of note, the in a small study from China.52 young offspring of Sprague–Dawley dams fed with Typically associated with gestational diabetes, 8% protein diet during gestation showed better macrosomia is another important marker for the glucose regulation than the control progeny, but development of obesity in adolescents, although the exhibited faster age-dependent loss of glucose strength of this association decreases after adjust- tolerance and became overtly intolerant 1 year after ments for mother’s .53 Importantly, birth.64,65 In male offspring, glucose intolerance was large for gestational age infants born from mothers related with insulin resistance whereas, in female with gestational diabetes or obesity developed rats, glucose intolerance was associated with insulin metabolic syndrome more often by the age of 11 deficiency, suggesting sex-specific mechanisms.65 years old.54 Taken together, these studies provide Structure and function of the fetal pancreas are evidence for the development of metabolic syn- altered following maternal protein restriction as drome in children and adults born large for gesta- fetuses of dams fed isocaloric diet containing 8% tional age associated or not with gestational protein (as opposed to 20% in the controls) had diabetes. Programming for metabolic syndrome in smaller islets with lower insulin content.57,66 Cel- large for gestational age newborns could be ex- lular replication, notably of b-cells, was also plained in part by the progressive development of reduced by almost 50% in the low-protein offspring, insulin resistance and other endocrine abnormal- because of prolonged cell cycle and higher cyclin ities.55 D1, a marker of G1 phase, but lower NEK2, an indicator of cells in G2 and mitosis.67 These results may reflect cell cycle blockage or lengthening in Maternal low-protein diet programmes for already differentiated b-cells. It may also result from the development of metabolic syndrome permanent programming of cell kinetics that might in rodent’s offspring have been imprinted on precursor cell populations before the beginning of endocrine differentiation. Although several animal species have been used Indeed, the population of endocrine precursor cells experimentally for metabolic programming, we that were immuno-positive for markers such as reviewed only studies with rodents because about nestin, CD34 and c-Kit was decreased during the 80% of publications focused on either mice or rats.56 perinatal period.67 Accelerated apoptosis has also Protein undernutrition during fetal life has been been reported, which could, potentially contribute consistently associated with the development of to reduced endocrine cell mass.67,68 Decreased islet obesity and metabolic syndrome in adult rodents. In cell proliferation and increased apoptosis persisted one classical model described by Snoeck et al.,57 after islets of protein-deprived fetuses were removed pregnant rats were fed either a regular chow from the disturbed metabolic maternal environment (20% protein) or an isocaloric low-protein diet and cultured for 7 days, which suggests program- (8% protein), from the first day of pregnancy until ming of these cells.69,70 delivery. At mid-pregnancy, plasma concentrations Additionally, it has been shown that islet cells of glucose, insulin, prolactin and progesterone were harvested from rat fetuses exposed to protein significantly lower in dams fed with low-protein restriction in utero exhibit altered protein expres- diet as compared with control animals.58 These sion.71 In late pregnancy, gene products involved in abnormalities affecting the pregnant uterine envir- mitochondria respiration, antioxidant defences and

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 409 glucagon metabolism are downregulated in pancrea- protein restriction exhibit increased preference for tic islet cells. Furthermore, reduced numbers of fatty foods, indicating programming of neural net- mitochondria and lower expression of genes in- works involved in feeding behaviour.84 Indeed, volved in mitochondrial respiration have been programming of appetite may stem from remodel- observed in the kidneys and skeletal muscles of ling of hypothalamic structures and from altered adult rats.72 Of note, reduced number of mitochon- expression of genes involved in responses to dria can be found in human leukocytes before the orexigenic hormones.85 Furthermore, rats exposed development of clinical type 2 diabetes.73 Indeed, to in utero undernourishment exhibit greater degree lower number of mitochondria in leukocytes may of obesity if given high-calorie food, indicating correlate with low birth weight and decreased augmented energy conservation in these animals. glucose tolerance in adult humans.74 Studies in animals and humans have also high- lighted the critical role of early nutrition influencing Programming of blood pressure regulation lipid metabolism.75–78 Barker et al.79 have suggested The offspring of undernourished dams shows that nutritionally impaired growth, particularly of elevated systolic blood pressure, markedly in- the liver during late gestation, could result in creased fasting plasma insulin and augmented permanent changes in cholesterol metabolism that leptinaemia.86 Hyperinsulinism and hyperleptinae- persist until adult life. On the other hand, one study mia could contribute to arterial hypertension in using the progeny of rat dams fed 50% food- programmed animals. Whereas insulin resistance restricted diet during pregnancy followed by control reflects impaired insulin actions on glucose and dams nursing has shown changes in triglyceride but carbohydrate metabolism, it is also known that not not cholesterol levels.80 This result suggests that the all insulin-responsive tissues become insulin resis- gestational period during which the dietary inter- tant.87 Thus, secondary hyperinsulinism can have vention is applied may be relevant for programming undesirable effects, such as chronic renal sodium of specific abnormalities in lipid metabolism. Inter- retention and increased sympathetic nervous system estingly, postnatal hypertension and collateral car- activity, whereas resistance in other tissues, such as diovascular adverse remodelling programmed by the endothelium, can impair vasodilation.88 Hyper- perinatal low-protein diet can be improved by leptinaemia has also been shown to increase both maternal fish oil supplement.81 sympathetic activity and blood pressure in mice with diet-induced obesity, despite resistance to the metabolic effects of the hormone.89 Taken together, Programming of hepatic function these observations suggest that insulin resistance, Abnormal function (either due to reduced expres- hyperinsulinism and hyperleptinaemia might be sion or activity) of hepatic enzymes that have key involved in blood pressure abnormalities seen in roles in glucose homeostasis have been reported in the offspring of undernourished dams. the offspring of rat dams fed protein-restricted diets. Thus, it has been proposed that programming of adult offspring glucose intolerance could be caused by altered expression of specific hepatic insulin- Programming of kidney function sensitive enzymes. Indeed, twofold reductions in Renal abnormalities are important hallmarks of fetal glucokinases and twofold increases in phosphoe- programming in rodents and humans, which could nolpyruvate carboxykinase have been observed in potentially cause renal dysfunction and hyperten- the offspring of protein-restricted dams at 11 months sion. Maternal prenatal protein–calorie restriction in of age. The resultant abnormal carbohydrate meta- rats causes renal dysfunction and impaired glomer- bolism has been attributed to reduced in utero ulogenesis in the adult offspring.90 Notably, mater- protein exposure.82 Maternal protein restriction nal protein restriction is associated with altered caused substantial reductions in hepatocyte num- mature to immature glomerular ratio at birth, and ber, steatosis of the liver and arterial hypertension in fewer functional glomeruli at weaning and later in the offspring, which were worsened after exposure life.91 Reduced number of nephrons may result from to high-fat diet. These results highlight the para- suboptimal nephrogenesis during kidney develop- mount importance of intra-uterine conditions and ment and/or loss of nephrons once nephrogenesis postnatal diet quality in the pathogenesis of chronic was complete.92 Importantly, protein restriction liver diseases.83 causes hypertension and renal dysfunction if occur- ring during nephrogenesis in rats.93,94 In addition, the remaining nephrons have thick basal membrane Programming of hypothalamic function and structural alterations in the pedicles of podo- Studies in rodents indicate that fetal undernutrition cytes, which are incapable of regenerative replica- determines adult adiposity. It is unclear whether tion.94 Therefore, loss of podocytes may create areas increases in central adiposity are related to aug- of ‘bare’ glomerular basement membrane, which mented food intake, reduced energy expenditure or represent potential starting points for irreversible both. Interestingly, rats exposed to intra-uterine glomerular injury.95

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 410 Fetal programming seems to be sexually di- gain suggests that hyperphagia is an important factor morphic, given that female rats are relatively to determine the development of metabolic syndrome resistant to hypertension induced by perinatal in the programmed offspring.107,113 Postnatal over- protein restriction. This resistance might be due to feeding induced by restriction of litter size or by the fact that modest maternal protein restriction cross-fostering growth-restricted offspring with nor- does not reduce the number of glomeruli in female mal dams is also associated with increased food rats.96 Interestingly, programming by low-protein intake in adult offspring.114,116 during gestation and/or lactation in rats can be passed transgenerationally,97 apparently in a gender- Programming of pancreatic function specific manner.98 This observation suggests the Insulin-dependent mechanisms appear to be pivotal potential contribution of genetic and/or epigenetic in determining metabolic abnormalities in the mechanisms for the pathophysiology of fetal pro- programmed offspring. Hyperinsulinaemia and hy- gramming in rodents. perglycaemia in mice dams at weaning trigger In humans, the kidneys of growth-retarded still- neonatal insulin responses via milk ingestion. Also, born are remarkably smaller, suggesting a renal increased adiposity in pregnant rats results in mechanism for the development of hypertension in insulin resistance in male and female offspring at adults who were born small for gestational age.99 3 months of age and overt diabetes afterwards, This is in line with epidemiological data showing indicating pancreatic b-cell insufficiency.107,113 that newborns with low birth weight carry an Furthermore, assessment of islet insulin secretory increased risk of mortality from chronic diseases in function in vitro shows reduced glucose-stimulated adulthood, including hypertension and early onset insulin secretion in the offspring of dams fed 20% of end-stage renal disease.93,100 animal fat-rich diet, which indicates blunted insulin secretory capacity. These animals also exhibit enlarged insulin secretory granules suggesting defi- High-fat diet during pregnancy and cient insulin release.117 lactation programmes for the development of metabolic syndrome in rodent’s offspring Programming of hepatic function and the role of lipid abnormalities Fetal programming leading to the development of Abnormal regulation of insulin signalling and in- obesity and related disorders is frequently associated creased levels of triglyceride in the liver have with maternal undernutrition and low birth weight in also been observed in the adult offspring of dams fed animals and humans. Under these conditions, the high-fat diet.118,119 Triglyceride accumulation in the programmed offspring is more likely to develop liver (that is, steatosis) causes hepatic insulin resis- hyperphagia, diabetes and cardiovascular dis- tance.119,120 Fat deposits directly damage the hepato- 75,101,102 eases. At present, however, the worldwide cytes and are followed by inflammation, increased overweight epidemic and maternal obesity during cytokine production and oxidative stress, abnormal pregnancy became serious public health hazards, cellular signalling and activation of stellate cells. which could potentially cause disease in the pro- Importantly, some reports suggest that liver steatosis 103 geny. The impact of fetal programming on the might also be an important cardiovascular risk phenotype of the offspring in adult age has been factor.115 In addition to hyperphagia and endocrine 104,105 assessed through experimental models. Studies abnormalities, other factors could contribute to high- using rodent models of maternal high-fat diet during fat diet-dependent fetal programming. The fatty acid pregnancy often result in the development of meta- composition of the diet appears to be important for the 103 bolic syndrome in the adult offspring. Notably, the fetal programming of metabolic syndrome, notably consumption of palatable processed food with high- excessive maternal saturated fat intake.121,122 Maternal fat (mostly trans fatty acids) and/or high-sugar hypercholesterolaemia associated with saturated fat content causes hyperinsulinism in pregnant rat dams consumption is strongly correlated with altered off- and promotes the development of obesity and spring lipid profile, which is an important component 106–111 diabetes in the offspring. Programmed obesity of the metabolic syndrome. Khan et al.123 reported and/or metabolic syndrome result from complex reduced plasma triglycerides and high-density lipo- interactions between maternal high energy dietary protein cholesterol in the offspring born to mothers 112 and/or maternal fat mass. Several mechanistic fed with saturated fatty acids diet. These effects are pathways could contribute to the development of probably due to altered maternal–fetal cholesterol overt metabolic syndrome or abnormalities in iso- transport across the placenta. lated components of the syndrome in the offspring. Abnormal feeding behaviour, altered endocrine status and changes in pancreatic morphology and function Programming of blood pressure and cardiovascular have been reported.107,113–115 Programmed mice ex- function hibit hyperphagia and reduced energy expenditure. Programmed metabolic syndrome is associated with Notably, augmented energy intake preceding weight several cardiovascular abnormalities such as high

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 411 blood pressure. Interestingly, some cardiovascular 11b-hydroxysteroid dehydrogenase type 2 defi- effects of high-fat diet-dependent fetal programming ciency and/or reduced number of glomeruli.131 can be prevented by the so-called predictive adap- However, the offspring of dams fed with high-fat tive responses. According to this concept, when diet during pregnancy did not present high postnatal diet is similar to dam’s diet, the offspring plasma levels of aldosterone. Alternatively, there is protected from the development of disease in are reports associating low-renin hypertension adulthood.124,125 For instance, when high-fat diet with dysfunction of the renal Na þ ,Kþ ATPase.132 persists during postnatal period, the offspring from Inhibition of Na þ ,Kþ ATPase activity decreases obese mice does not develop endothelial dysfunc- sodium reabsorption and, thereby, natriuresis.133–135 tion.124 In contrast, when fat-rich diet is given only Indeed, the activity of renal Na þ ,Kþ ATPase during pregnancy and lactation, offspring exhibit is significantly reduced in the offspring exposed reduced aortic endothelium-dependent vasodilata- to maternal high-fat diet during pregnancy, which tion and increased aortic stiffness.126 Nonetheless, is strongly associated with hypertension. The some studies have shown that although early mechanisms underpinning the reduced activity exposure to high-fat diet increases blood pressure of renal Na þ ,Kþ ATPase remain speculative; in female offspring, continuing dietary high fat nevertheless there is some evidence that lipids during post weaning period does not prevent including docosahexaenoic acid modulate its abnormal blood pressure.124,125 This may suggest activity.136,137 irreversible effects of maternal diet on mechanisms Taken together, these observations strongly sug- of blood pressure regulation.126 These observations gest that maternal diet alters different pathways that indicate that hypertension in rats is not prevented contribute to the control of blood pressure. The by concordant diets during pregnancy and after consumption of high-fat diets by pregnant rodents weaning, which contradict the predictive adaptive alter hypothalamus–pituitary–adrenal function in response, at least in regard to blood pressure the offspring and lead to abnormal activity of the regulation.124,125 renin–angiotensin–aldosterone system with strong Although the association of endothelial dysfunc- association with defective activity of Na þ K þ tion and hypertension in offspring of fat-fed dams ATPase. Furthermore, defects in the cardiovascular is described in some reports,127 dissociation be- response to nitric oxide are not invariably associated tween blood pressure regulation and endothelial with the development of hypertension in the off- function has also been observed. Khan et al.123 have spring. shown that high-fat diet during rat pregnancy leads to increased blood pressure in the offspring, which is, however, confined to female rats. In Programming of hypothalamic function contrast, both male and female offspring exhibited In addition to the programmed endocrine and abnormal endothelium-dependent vasodilatation. cardiovascular effects of maternal high-fat diet These findings corroborate several reports showing during pregnancy and lactation in rodents, altered that hypertension and endothelial dysfunction are central nervous system plasticity in the offspring not invariably associated in the programmed off- has also been reported.114,138 Exposure to different spring.113 nutrients during critical phases of development There are other potential explanations for the could promote an imbalance between orexigenic development of high blood pressure in programmed neuropeptide Y and appetite suppressor proopio- female rat offspring. First, high-fat diet during melanocortin signalling, which favours increased pregnancy induces deficiency of the cardioprotec- energy intake.139 Neonatal hyperinsulinaemia could tive n-3 polyunsaturated fatty acid, docosahexae- also contribute to altered central nervous system noic acid, in the offspring, which is associated with plasticity because it induces morphological altera- later development of hypertension.128 Second, the tions of hypothalamic nuclei involved in body activation of the hypothalamus–pituitary–adrenal weight regulation and metabolism.140,141 is axis in pregnant rats has been reported to cause another important regulator of energy homeostasis permanent and gender-specific changes in the off- and appetite within the central nervous system. spring’s hypothalamus–pituitary–adrenal axis.129,130 Leptin acts directly in hypothalamic nuclei to There are substantial increases of plasma corticos- reduce neuropeptide Y and activate proopiomela- terone concentration in the offspring as a result of nocortin expression, inhibiting appetite and increas- maternal high-fat diet, which could potentially ing energy expenditure.142 Importantly, maternal increase blood pressure.126 Third, increased blood high-fat diet also programmes hypothalamic leptin pressure observed in female offspring of fat-fed resistance and impairs leptin signalling in the dams could be associated with low renal tissue offspring,138,143 which causes hyperphagia and fa- renin activity.103 In low protein-programmed off- vours energy storage as fat. spring, blunted plasma renin activity has been Maternal nutrition during pregnancy and lacta- associated with hypertension, which could be tion can also modify gene expression in the off- attributed to increases in plasma aldosterone, spring mostly through epigenetic mechanisms such augmented adrenal sensitivity to angiotensin-II, as histone tail modifications, DNA methylation, and

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 412 altered expression of histone- and DNA-modifying Nutritional effects on epigenetic enzymes.144–146 Studies have focused on genes regulation known to be relevant for the development of metabolic syndrome including peroxisomal prolif- The exponential increase in the prevalence of erator-activated receptor and the glucocorticoid obesity has occurred too rapidly to be explained receptor. Changes in the promoters of these genes completely by genetic causes. Increasing evidence and altered mRNA expression in the offspring of suggests that changes in the epigenome caused by high fat-fed dams have been reported.147,148 early environmental cues are associated with in- Exposure to high-fat diet during pregnancy and creased susceptibility to metabolic disease later in lactation predisposes rodent offspring for the devel- life. Conrad Waddington, who is given credit for opment of metabolic syndrome in adulthood. Ab- coining the term, defined epigenetics as ‘the branch normal regulation of insulin signalling and of biology that studies the causal interactions increased levels of triglyceride in the liver has been between genes and their products, which bring the observed in the adult offspring, leading to insulin phenotype into being’.149 Today’s epigenetic re- resistance and liver steatosis. Maternal nutrition search is converging on the study of covalent and also impairs the mechanisms of blood pressure non-covalent modifications of DNA and histone control in the offspring. Furthermore, increased proteins, and the mechanisms by which such intake of saturated fat intake during pregnancy modifications influence overall chromatin struc- and/or lactation augments the expression of orexi- ture.149 Epigenetic mechanisms comprise several genic peptides promoting hyperphagia and obesity levels of interconnected and interdependent codes, in the offspring. Recent findings have shown that including methylation of the cytosine residue in epigenetic mechanisms could help explain the CpG dinucleotides of DNA, covalent modifications molecular basis of metabolic syndrome in the (including methylation, acetylation, phosphoryla- programmed offspring (Figure 1). tion and ubiquitination) of histones, and the gene-

Figure 1 Fetal programming: nutritional interventions during gestation and lactation can programme the metabolic syndrome in the adult offspring. Low-protein and high-fat diets during pregnancy could alter the offspring’s epigenome. The resulting phenotype is characterized by morphological and functional alterations of target-tissues (pancreas, liver, central nervous system, kidneys and cardiovascular system) favouring hyperphagia, obesity, insulin resistance, liver steatosis and altered lipid profile in the adult offspring. CNS, central nervous system; DHA, docosahexaenoic acid; NPY, neuropeptide Y; POMC, proopiomelanocortin.

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 413 regulating and chromatin-organizing activities of transmitter in the hypothalamus. Furthermore, the non-coding RNAs.149,150 Such processes change the expression of mu-opioid receptor and preproenke- binding of transcription activators and repressors to phalin was increased in nucleus accumbens. In specific gene promoters, and/or alter the large-scale association with long-term alterations in gene conformation and function of chromatin itself, expression, maternal high-fat diet can affect gene- which modulates gene expression.151 Epigenetic specific (dopamine reuptake transporter, mu-opioid research partly provides the molecular basis for receptor, and preproenkephalin) promoter DNA experimental and epidemiological observations hypomethylation in the brain of the offspring, showing that the early period of life is critical in altering feeding behaviour.156 determining the susceptibility to chronic non-com- These experimental observations, collectively, municable diseases during adulthood.150 The im- underpin potential molecular bases to the develop- balance between nutrient intake (qualitative and mental origins of obesity in rodents. However, in quantitative) and metabolites during embryonic, humans, one great obstacle for concept proofing is fetal and/or early postnatal development may tissue specific of epigenetic regulation. It is likely irreversibly modify the mammalian epigenome. In that important inter-individual epigenetic variations this way, even a brief period of sub-optimal nutri- occur, specifically in the hypothalamus. Also, tion can alter expression over a long period and whereas initial studies will focus on DNA methyla- compromise normal physiology and metabolism.152 tion because of the relative simplicity and stability Nutritional effects on DNA methylation were of this epigenetic modication, epigenetic dysregula- studied in Agouti mouse by Waterland and Jirtle.153 tion relevant to obesity will almost certainly involve These animals present a yellow coat colour deter- additional interacting epigenetic mechanisms. mined by overexpression of a single gene locus Moreover, current available technology hardly pre- (Avy) and this characteristic can be modified by dicts genomic regions in which epigenetic modifica- nutritional status. When yellow mice dams were tions contribute to the transcriptional regulation of supplemented with folic acid, vitamin B12, choline, specific genes.157 Understanding epigenetic mar- chloride and betaine, the methylation of Avy was kers, such as methylation patterns in specific gene increased in the offspring. This resulted in more promoters linking early nutrition to the develop- mice with an intermediate brown coat colour being ment of disease later in life, could potentially guide born to supplemented mothers. One recent study early-life nutritional interventions aiming at achiev- demonstrated that maternal diet promotes changes ing long-term improvements in human health. in epigenome of the offspring in rodents. Protein restriction during pregnancy induced overexpres- sion of the hepatic glucocorticoid and peroxisome Clinical implications proliferator-activated receptor-a receptors in the The increase in obesity prevalence is a major public offspring. These alterations are accompanied by health issue worldwide. It is clear that life style and epigenetic changes that facilitate transcription of environmental factors can contribute do obesity these receptors. Specifically, changes in glucocorti- development, whereas the susceptibility to obesity coid receptors expression regulation include histone and metabolic diseases may also originate in early modifications, promoter hypomethylation and re- development. Both experimental and epidemiologi- duced expression of methyltransferase 1 (Dnmt1).154 cal observations demonstrate the impact of inade- Also, feeding dams during pregnancy and lactation quate maternal nutrition during gestation and with high-fat diet changed epigenome in primate lactation on the offspring. Understanding the me- fetuses. Impaired lipid metabolism is associated chanisms underlying the predisposition to future with increased expression of several relevant genes disease could lead to preventive interventions through augmented histone H3 acetylation and during early childhood. The new challenge is to decreased histone deacetylase activity, both of identify effective strategies to prevent deleterious which indicate increased transcription.155 These effects of programming, which could be initiated abnormalities in gene expression could potentially during early development in order to reduce the favour the development of the overweight or obesity adverse effects of poor maternal nutrition on the in adulthood. offspring’s health. The novel concepts of fetal In addition, high-fat diet during pregnancy and programming should orient new nutrition guide- lactation has long-term effects on feeding behaviour. lines particularly those addressing feeding habits in Dopamine and endogenous opioids are neurotrans- pregnant women aiming at attenuating obesity and mitters associated with reward networks that affect related metabolic disorders in the adult offspring. animal’s preference for palatable foods. Mice from dams fed with high-fat diet during pregnancy and lactation showed increased preference for sucrose Summary and fat, which is associated with a significant upregulation of dopamine reuptake transporter in This paper reviewed evidence indicating that the ventral tegumental area, nucleus accumbens and maternal nutrition has a pivotal role in the off- prefrontal cortex, and downregulation of this neuro- spring’s lifelong development in both rodents and

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 414 humans. Epidemiologic observations have shown study, a cross sectional study. BMJ 1997; 315(7119): the impact of undernourishment during pregnancy 1342–1348. on the development of disease in the progeny. 8 Ravelli GP, Stein ZA, Susser MW. Obesity in young Programming for metabolic syndrome of infants men after famine exposure in utero and early infancy. born small for gestational age that exhibited catch- N Engl J Med 1976; 295(7): 349–353. 9 Ravelli AC, van Der Meulen JH, Osmond C, Barker DJ, up could have an important role in the higher Bleker OP. Obesity at the age of 50 y in men and cardiovascular risk in these individuals. However, women exposed to famine prenatally. Am J Clin Nutr maternal and gestational overnutrition are more 1999; 70(5): 811–816. prevalent than undernutrition, at least in more 10 Roseboom TJ, van der Meulen JH, Ravelli AC, affluent countries. Although the link between fetal Osmond C, Barker DJ, Bleker OP. Effects of prenatal programming and adult-onset metabolic syndrome exposure to the Dutch famine on adult disease in later in newborns large for gestational age is weaker, it is life: an overview. Twin Res 2001; 4(5): 293–298. possible that heavy newborns are at increased risk 11 Eriksson JG, Forsen T, Tuomilehto J, Winter PD, for the development of obesity and metabolic Osmond C, Barker DJ. Catch-up growth in childhood syndrome. and death from coronary heart disease: longitudinal study. BMJ 1999; 318(7181): 427–431. Reduced protein intake during gestation and 12 Syddall HE, Sayer AA, Simmonds SJ, Osmond C, lactation results in target-organ abnormalities, Cox V, Dennison EM et al. Birth weight, infant which predispose the offspring for metabolic syn- weight gain, and cause-specific mortality: the Hert- drome. The thrifty phenotype is associated with fordshire Cohort Study. Am J Epidemiol 2005; obesity and impaired glucose metabolism. Also, 161(11): 1074–1080. insulin resistance, hyperinsulinism, hyperleptinae- 13 Barker DJ, Winter PD, Osmond C, Margetts B, mia and impaired kidney organogenesis could Simmonds SJ. Weight in infancy and death from contribute to the development of high blood pres- ischaemic heart disease. Lancet 1989; 2(8663): sure in the offspring. Dams’ overnutrition may 577–580. produce similar phenotypes in the adult offspring. 14 Osmond C, Barker DJ, Winter PD, Fall CH, Simmonds SJ. Early growth and death from cardio- The excess of saturated fat in the diet during critical vascular disease in women. BMJ 1993; 307(6918): periods of development causes abnormal regulation 1519–1524. of insulin signalling, increases triglyceridaemia and 15 Martyn CN, Barker DJ, Osmond C. Mothers’ pelvic impairs mechanisms of blood pressure regulation. size, fetal growth, and death from stroke and coronary Furthermore, increased saturated fat intake during heart disease in men in the UK. Lancet 1996; pregnancy and/or lactation augments the expression 348(9037): 1264–1268. of orexigenic peptides promoting hyperphagia and 16 Phillips DI, Jones A, Goulden PA. Birth weight, stress, obesity in the offspring. and the metabolic syndrome in adult life. Ann NY Acad Sci 2006; 1083: 28–36. 17 Silveira VM, Horta BL. [Birth weight and metabolic syndrome in adults: meta-analysis]. Rev Saude Pub- Conflict of interest lica 2008; 42(1): 10–18. 18 Salonen MK, Kajantie E, Osmond C, Forsen T, The authors declare no conflict of interest. Yliharsila H, Paile-Hyvarinen M et al. Childhood growth and future risk of the metabolic syndrome in normal-weight men and women. Diabetes Metab References 2009; 35(2): 143–150. 19 Law CM, Shiell AW, Newsome CA, Syddall HE, 1 Barker DJ. The fetal origins of adult hypertension. Shinebourne EA, Fayers PM et al. Fetal, infant, and J Hypertens Suppl 1992; 10(7): S39–S44. childhood growth and adult blood pressure: a long- 2 Lucas A. Programming by early nutrition in man. Ciba itudinal study from birth to 22 years of age. Circula- Found Symp 1991; 156: 38–50; discussion 50–35. tion 2002; 105(9): 1088–1092. 3 Langley-Evans SC. Developmental programming of 20 Hemachandra AH, Howards PP, Furth SL, Klebanoff health and disease. Proc Nutr Soc 2006; 65(1): MA. Birth weight, postnatal growth, and risk for high 97–105. blood pressure at 7 years of age: results from the 4 Barraclough CA. Production of anovulatory, sterile Collaborative Perinatal Project. Pediatrics 2007; rats by single injections of testosterone propionate. 119(6): e1264–e1270. Endocrinology 1961; 68: 62–67. 21 Singhal A, Cole TJ, Fewtrell M, Kennedy K, Stephen- 5 Rose G. Familial patterns in ischaemic heart disease. son T, Elias-Jones A et al. Promotion of faster weight Br J Prev Soc Med 1964; 18: 75–80. gain in infants born small for gestational age: is there 6 Junien C, Nathanielsz P. Report on the IASO Stock an adverse effect on later blood pressure? Circulation Conference 2006: early and lifelong environmental 2007; 115(2): 213–220. epigenomic programming of metabolic syndrome, 22 Barker DJ, Bagby SP, Hanson MA. Mechanisms of obesity and type II diabetes. Obes Rev 2007; 8(6): disease: in utero programming in the pathogenesis of 487–502. hypertension. Nat Clin Pract Nephrol 2006; 2(12): 7 Stanner SA, Bulmer K, Andres C, Lantseva OE, 700–707. Borodina V, Poteen VV et al. Does malnutrition in 23 Hallan S, Euser AM, Irgens LM, Finken MJ, Holmen J, utero determine diabetes and coronary heart disease Dekker FW. Effect of intrauterine growth restriction in adulthood? Results from the Leningrad siege on kidney function at young adult age: the Nord

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 415 Trondelag Health (HUNT 2) Study. Am J Kidney Dis 39 Barker DJ, Osmond C, Forsen TJ, Kajantie E, Eriksson 2008; 51(1): 10–20. JG. Trajectories of growth among children who have 24 Simonetti GD, Raio L, Surbek D, Nelle M, Frey FJ, coronary events as adults. N Engl J Med 2005; 353(17): Mohaupt MG. Salt sensitivity of children with low 1802–1809. birth weight. Hypertension 2008; 52(4): 625–630. 40 Fall CH, Barker DJ, Osmond C, Winter PD, Clark PM, 25 Jones A, Beda A, Ward AM, Osmond C, Phillips DI, Hales CN. Relation of infant feeding to adult serum Moore VM et al. Size at birth and autonomic function cholesterol concentration and death from ischaemic during psychological stress. Hypertension 2007; heart disease. BMJ 1992; 304(6830): 801–805. 49(3): 548–555. 41 Huxley R, Owen CG, Whincup PH, Cook DG, Colman 26 Feldt K, Raikkonen K, Eriksson JG, Andersson S, S, Collins R. Birth weight and subsequent cholesterol Osmond C, Barker DJ et al. Cardiovascular reactivity levels: exploration of the ‘fetal origins’ hypothesis. to psychological stressors in late adulthood is JAMA 2004; 292(22): 2755–2764. predicted by gestational age at birth. J Hum Hypertens 42 Charney E, Goodman HC, McBride M, Lyon B, Pratt R. 2007; 21(5): 401–410. Childhood antecedents of adult obesity. Do chubby 27 Ong KK, Ahmed ML, Emmett PM, Preece MA, Dunger infants become obese adults? N Engl J Med 1976; DB. Association between postnatal catch-up growth 295(1): 6–9. and obesity in childhood: prospective cohort study. 43 Braddon FE, Rodgers B, Wadsworth ME, Davies JM. BMJ 2000; 320(7240): 967–971. Onset of obesity in a 36 year birth cohort study. Br 28 Ibanez L, Ong K, Dunger DB, de Zegher F. Early Med J (Clin Res Ed) 1986; 293(6542): 299–303. development of adiposity and insulin resistance 44 Surkan PJ, Hsieh CC, Johansson AL, Dickman PW, after catch-up weight gain in small-for-gestational- Cnattingius S. Reasons for increasing trends in large age children. J Clin Endocrinol Metab 2006; 91(6): for gestational age births. Obstet Gynecol 2004; 2153–2158. 104(4): 720–726. 29 Ibanez L, Suarez L, Lopez-Bermejo A, Diaz M, Valls C, 45 Hediger ML, Overpeck MD, McGlynn A, Kuczmarski de Zegher F. Early development of visceral fat excess RJ, Maurer KR, Davis WW. Growth and fatness at after spontaneous catch-up growth in children with three to six years of age of children born small- or low birth weight. J Clin Endocrinol Metab 2008; 93(3): large-for-gestational age. Pediatrics 1999; 104(3): e33. 925–928. 46 Dubois L, Girard M. Early determinants of overweight 30 Ibanez L, Lopez-Bermejo A, Suarez L, Marcos MV, at 4.5 years in a population-based longitudinal study. Diaz M, de Zegher F. Visceral adiposity without Int J Obes (Lond) 2006; 30(4): 610–617. overweight in children born small for gestational age. 47 Rasmussen F, Johansson M. The relation of weight, J Clin Endocrinol Metab 2008; 93(6): 2079–2083. length and ponderal index at birth to body mass index 31 Sayer AA, Syddall HE, Dennison EM, Gilbody HJ, and overweight among 18-year-old males in Sweden. Duggleby SL, Cooper C et al. Birth weight, weight at 1 Eur J Epidemiol 1998; 14(4): 373–380. y of age, and body composition in older men: findings 48 Eriksson J, Forsen T, Tuomilehto J, Osmond C, Barker from the Hertfordshire Cohort Study. Am J Clin Nutr D. Size at birth, childhood growth and obesity in 2004; 80(1): 199–203. adult life. Int J Obes Relat Metab Disord 2001; 25(5): 32 Parsons TJ, Power C, Manor O. Fetal and early life 735–740. growth and body mass index from birth to early 49 Clarke WR, Lauer RM. Does childhood obesity track adulthood in 1958 British cohort: longitudinal study. into adulthood? Crit Rev Food Sci Nutr 1993; 33(4–5): BMJ 2001; 323(7325): 1331–1335. 423–430. 33 Harder T, Rodekamp E, Schellong K, Dudenhausen 50 Curhan GC, Willett WC, Rimm EB, Spiegelman D, JW, Plagemann A. Birth weight and subsequent risk of Ascherio AL, Stampfer MJ. Birth weight and adult type 2 diabetes: a meta-analysis. Am J Epidemiol hypertension, diabetes mellitus, and obesity in US 2007; 165(8): 849–857. men. Circulation 1996; 94(12): 3246–3250. 34 Whincup PH, Kaye SJ, Owen CG, Huxley R, Cook DG, 51 Wei JN, Sung FC, Li CY, Chang CH, Lin RS, Lin CC Anazawa S et al. Birth weight and risk of type 2 et al. Low birth weight and high birth weight infants diabetes: a systematic review. JAMA 2008; 300(24): are both at an increased risk to have type 2 diabetes 2886–2897. among schoolchildren in Taiwan. Diabetes Care 2003; 35 Phillips DI, Goulden P, Syddall HE, Aihie Sayer A, 26(2): 343–348. Dennison EM, Martin H et al. Fetal and infant growth 52 Wang X, Liang L, Junfen FU, Lizhong DU. Metabolic and glucose tolerance in the Hertfordshire Cohort syndrome in obese children born large for gestational Study: a study of men and women born between 1931 age. Indian J Pediatr 2007; 74(6): 561–565. and 1939. Diabetes 2005; 54(Suppl 2): S145–S150. 53 Gillman MW, Rifas-Shiman S, Berkey CS, Field AE, 36 Remacle C, Dumortier O, Bol V, Goosse K, Romanus P, Colditz GA. Maternal gestational diabetes, birth Theys N et al. Intrauterine programming of the weight, and adolescent obesity. Pediatrics 2003; endocrine pancreas. Diabetes Obes Metab 2007; 111(3): e221–e226. 9(Suppl 2): 196–209. 54 Boney CM, Verma A, Tucker R, Vohr BR. Metabolic 37 Pulizzi N, Lyssenko V, Jonsson A, Osmond C, syndrome in childhood: association with birth Laakso M, Kajantie E et al. Interaction between weight, maternal obesity, and gestational diabetes prenatal growth and high-risk genotypes in the mellitus. Pediatrics 2005; 115(3): e290–e296. development of type 2 diabetes. Diabetologia 2009; 55 Giapros V, Evagelidou E, Challa A, Kiortsis D, Drougia 52(5): 825–829. A, Andronikou S. Serum adiponectin and leptin 38 Kajantie E, Barker DJ, Osmond C, Forsen T, Eriksson levels and insulin resistance in children born JG. Growth before 2 years of age and serum lipids 60 large for gestational age are affected by the degree years later: the Helsinki Birth Cohort study. Int J of overweight. Clin Endocrinol (Oxf) 2007; 66(3): Epidemiol 2008; 37(2): 280–289. 353–359.

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 416 56 Vuguin PM. Animal models for small for gestational cause long-term changes in rat liver and muscle age and fetal programming of adult disease. Horm Res mitochondria. J Nutr 2003; 133(10): 3085–3090. 2007; 68(3): 113–123. 73 Lee HK, Song JH, Shin CS, Park DJ, Park KS, Lee KU 57 Snoeck A, Remacle C, Reusens B, Hoet JJ. Effect et al. Decreased mitochondrial DNA content in of a low protein diet during pregnancy on the fetal peripheral blood precedes the development of non- rat endocrine pancreas. Biol Neonate 1990; 57(2): insulin-dependent diabetes mellitus. Diabetes Res 107–118. Clin Pract 1998; 42(3): 161–167. 58 Fernandez-Twinn DS, Ozanne SE, Ekizoglou S, 74 Lee HK. Evidence that the mitochondrial genome is Doherty C, James L, Gusterson B et al. The maternal the thrifty genome. Diabetes Res Clin Pract 1999; endocrine environment in the low-protein model of 45(2–3): 127–135. intra-uterine growth restriction. Br J Nutr 2003; 90(4): 75 Barker DJ, Hales CN, Fall CH, Osmond C, Phipps K, 815–822. Clark PM. Type 2 (non-insulin-dependent) diabetes 59 Langley SC, Jackson AA. Increased systolic blood mellitus, hypertension and hyperlipidaemia (syn- pressure in adult rats induced by fetal exposure to drome X): relation to reduced fetal growth. Diabeto- maternal low protein diets. Clin Sci (Lond) 1994; logia 1993; 36(1): 62–67. 86(2): 217–222; discussion 121. 76 Gluckman PD, Hanson MA. The developmental 60 Nishina H, Green LR, McGarrigle HH, Noakes DE, origins of the metabolic syndrome. Trends Endocrinol Poston L, Hanson MA. Effect of nutritional restriction Metab 2004; 15(4): 183–187. in early pregnancy on isolated femoral artery function 77 Desai M, Hales CN. Role of fetal and infant growth in in mid-gestation fetal sheep. J Physiol 2003; 553(Pt 2): programming metabolism in later life. Biol Rev Camb 637–647. Philos Soc 1997; 72(2): 329–348. 61 Ozanne SE. Metabolic programming in animals. Br 78 Lucas A, Baker BA, Desai M, Hales CN. Nutrition in Med Bull 2001; 60: 143–152. pregnant or lactating rats programs lipid metabolism 62 Hales CN, Barker DJ. The thrifty phenotype hypoth- in the offspring. Br J Nutr 1996; 76(4): 605–612. esis. Br Med Bull 2001; 60: 5–20. 79 Barker DJ, Martyn CN, Osmond C, Hales CN, Fall CH. 63 Dahri S, Snoeck A, Reusens-Billen B, Remacle C, Growth in utero and serum cholesterol concentrations Hoet JJ. Islet function in offspring of mothers on in adult life. BMJ 1993; 307(6918): 1524–1527. low-protein diet during gestation. Diabetes 1991; 80 Desai M, Gayle D, Babu J, Ross MG. Programmed 40(Suppl 2): 115–120. obesity in intrauterine growth-restricted newborns: 64 Petry CJ, Ozanne SE, Wang CL, Hales CN. Early modulation by newborn nutrition. Am J Physiol Regul protein restriction and obesity independently induce Integr Comp Physiol 2005; 288(1): R91–R96. hypertension in 1-year-old rats. Clin Sci (Lond) 1997; 81 Gregorio BM, Souza-Mello V, Mandarim-de-Lacerda 93(2): 147–152. CA, Aguila MB. Maternal fish oil supplementation 65 Shepherd PR, Crowther NJ, Desai M, Hales CN, benefits programmed offspring from rat dams fed low- Ozanne SE. Altered adipocyte properties in the protein diet. Am J Obstet Gynecol 2008; 199(1): 82 offspring of protein malnourished rats. Br J Nutr e81–82 e87. 1997; 78(1): 121–129. 82 Desai M, Byrne CD, Zhang J, Petry CJ, Lucas A, 66 Reusens B, Remacle C. Programming of the endocrine Hales CN. Programming of hepatic insulin-sensitive pancreas by the early nutritional environment. Int J enzymes in offspring of rat dams fed a protein- Biochem Cell Biol 2006; 38(5–6): 913–922. restricted diet. Am J Physiol 1997; 272(5 Pt 1): 67 Petrik J, Reusens B, Arany E, Remacle C, Coelho C, G1083–G1090. Hoet JJ et al. A low protein diet alters the balance of 83 Souza-Mello V, Mandarim-de-Lacerda CA, Aguila islet cell replication and apoptosis in the fetal and MB. Hepatic structural alteration in adult pro- neonatal rat and is associated with a reduced grammed offspring (severe maternal protein restric- pancreatic expression of insulin-like growth factor- tion) is aggravated by post-weaning high-fat diet. Br J II. Endocrinology 1999; 140(10): 4861–4873. Nutr 2007; 98(6): 1159–1169. 68 Boujendar S, Reusens B, Merezak S, Ahn MT, Arany 84 Bellinger L, Langley-Evans SC. Fetal programming of E, Hill D et al. Taurine supplementation to a appetite by exposure to a maternal low-protein diet in low protein diet during foetal and early postnatal the rat. Clin Sci (Lond) 2005; 109(4): 413–420. life restores a normal proliferation and apoptosis 85 Langley-Evans SC, Bellinger L, McMullen S. Animal of rat pancreatic islets. Diabetologia 2002; 45(6): models of programming: early life influences on 856–866. appetite and feeding behaviour. Matern Child Nutr 69 Dahri S, Reusens B, Remacle C, Hoet JJ. Nutritional 2005; 1(3): 142–148. influences on pancreatic development and potential 86 Vickers MH, Breier BH, Cutfield WS, Hofman PL, links with non-insulin-dependent diabetes. Proc Nutr Gluckman PD. Fetal origins of hyperphagia, obesity, Soc 1995; 54(2): 345–356. and hypertension and postnatal amplification by 70 Cherif H, Reusens B, Dahri S, Remacle C. A protein- hypercaloric nutrition. Am J Physiol Endocrinol restricted diet during pregnancy alters in vitro insulin Metab 2000; 279(1): E83–E87. secretion from islets of fetal Wistar rats. J Nutr 2001; 87 Amiel SA, Caprio S, Sherwin RS, Plewe G, Haymond 131(5): 1555–1559. MW, Tamborlane WV. Insulin resistance of puberty: a 71 Sparre T, Reusens B, Cherif H, Larsen MR, Roepstorff defect restricted to peripheral glucose metabolism. P, Fey SJ et al. Intrauterine programming of fetal islet J Clin Endocrinol Metab 1991; 72(2): 277–282. gene expression in rats—effects of maternal protein 88 Steinberg HO, Baron AD. Vascular function, insulin restriction during gestation revealed by proteome resistance and fatty acids. Diabetologia 2002; 45(5): analysis. Diabetologia 2003; 46(11): 1497–1511. 623–634. 72 Park KS, Kim SK, Kim MS, Cho EY, Lee JH, Lee KU 89 Rahmouni K, Morgan DA, Morgan GM, Mark AL, et al. Fetal and early postnatal protein malnutrition Haynes WG. Role of selective leptin resistance in diet-

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 417 induced obesity hypertension. Diabetes 2005; 54(7): 105 Symonds ME, Sebert SP, Hyatt MA, Budge H. 2012–2018. Nutritional programming of the metabolic syndrome. 90 Almeida JR, Mandarim-de-Lacerda CA. Maternal Nat Rev Endocrinol 2009; 5(11): 604–610. gestational protein-calorie restriction decreases the 106 Patel MS, Srinivasan M. Metabolic programming due number of glomeruli and causes glomerular hyper- to alterations in nutrition in the immediate postnatal trophy in adult hypertensive rats. Am J Obstet period. J Nutr 2010; 140(3): 658–661. Gynecol 2005; 192(3): 945–951. 107 Patel MS, Srinivasan M. Metabolic programming due 91 Catta-Preta M, Oliveira DA, Mandarim-de-Lacerda to alterations in nutrition in the immediate postnatal CA, Aguila MB. Adult cardiorenal benefits from period. J Nutr 2010; 140(3): 658–661. postnatal fish oil supplement in rat offspring of low- 108 Plagemann A, Harder T, Rake A, Voits M, Fink H, protein pregnancies. Life Sci 2006; 80(3): 219–229. Rohde W et al. Perinatal elevation of hypothalamic 92 Douglas-Denton RN, McNamara BJ, Hoy WE, Hugh- insulin, acquired malformation of hypothalamic son MD, Bertram JF. Does nephron number matter in galaninergic neurons, and syndrome x-like alterations the development of kidney disease? Ethn Dis 2006; in adulthood of neonatally overfed rats. Brain Res 16(2 Suppl 2): S2-40–S2-45. 1999; 836(1–2): 146–155. 93 Pires KMP, Aguila MB, Mandarim-de-Lacerda CA. 109 Holemans K, Caluwaerts S, Poston L, Van Assche FA. Early renal structure alteration in rat offspring from Diet-induced obesity in the rat: a model for gesta- dams fed low protein diet. Life Sci 2006; 79(22): tional diabetes mellitus. Am J Obstet Gynecol 2004; 2128–2134. 190(3): 858–865. 94 Villar-Martini VC, Carvalho JJ, Neves MF, Aguila MB, 110 Napoli C, de Nigris F, Welch JS, Calara FB, Stuart RO, Mandarim-de-Lacerda CA. Hypertension and kidney Glass CK et al. Maternal hypercholesterolemia during alterations in rat offspring from low protein pregnan- pregnancy promotes early atherogenesis in LDL cies. J Hypertens 2009; 27(Suppl 6): S47–S51. receptor-deficient mice and alters aortic gene expres- 95 Kriz W, Gretz N, Lemley KV. Progression of glomer- sion determined by microarray. Circulation 2002; ular diseases: is the podocyte the culprit? Kidney Int 105(11): 1360–1367. 1998; 54(3): 687–697. 111 Zhang J, Wang C, Terroni PL, Cagampang FR, 96 Woods LL, Ingelfinger JR, Rasch R. Modest maternal Hanson M, Byrne CD. High-unsaturated-fat, protein restriction fails to program adult hypertension high-protein, and low-carbohydrate diet during in female rats. Am J Physiol Regul Integr Comp pregnancy and lactation modulates hepatic lipid Physiol 2005; 289(4): R1131–R1136. metabolism in female adult offspring. Am J 97 Pinheiro AR, Salvucci ID, Aguila MB, Mandarim-de- Physiol Regul Integr Comp Physiol 2005; 288(1): Lacerda CA. Protein restriction during gestation R112–R118. and/or lactation causes adverse transgenerational 112 Bayol SA, Farrington SJ, Stickland NC. A maternal effects on biometry and glucose metabolism in F1 ‘junk food’ diet in pregnancy and lactation promotes and F2 progenies of rats. Clin Sci (Lond) 2008; 114(5): an exacerbated taste for ‘junk food’ and a greater 381–392. propensity for obesity in rat offspring. Br J Nutr 2007; 98 Zambrano E, Martı´nez-Samayoa PM, Bautista CJ, Deas 98(4): 843–851. M, Guillen L, Rodriguez-Gonzalez GL et al. Sex 113 Samuelsson AM, Matthews PA, Argenton M, Christie differences in transgenerational alterations of growth MR, McConnell JM, Jansen EH et al. Diet-induced and metabolism in progeny (F2) of female offspring obesity in female mice leads to offspring hyperpha- (F1) of rats fed a low protein diet during pregnancy gia, adiposity, hypertension, and insulin resistance: a and lactation. J Physiol 2005; 566(Pt 1): 225–236. novel murine model of developmental programming. 99 Kingdom JC, Hayes M, McQueen J, Howatson AG, Hypertension 2008; 51(2): 383–392. Lindop GB. Intrauterine growth restriction is asso- 114 Plagemann A. Perinatal programming and functional ciated with persistent juxtamedullary expression teratogenesis: impact on body weight regulation and of renin in the fetal kidney. Kidney Int 1999; 55(2): obesity. Physiol Behav 2005; 86(5): 661–668. 424–429. 115 Tessari P, Coracina A, Cosma A, Tiengo A. Hepatic 100 Lackland DT, Bendall HE, Osmond C, Egan BM, lipid metabolism and non-alcoholic fatty liver Barker DJ. Low birth weights contribute to high rates disease. Nutr Metab Cardiovasc Dis 2009; 19(4): of early-onset chronic renal failure in the South- 291–302. eastern United States. Arch Intern Med 2000; 160(10): 116 Desai M, Gayle D, Han G, Ross MG. Programmed 1472–1476. hyperphagia due to reduced anorexigenic mechan- 101 Barker DJ. The intrauterine origins of cardiovascular isms in intrauterine growth-restricted offspring. Re- disease. Acta Paediatr Suppl 1993; 82(Suppl 391): prod Sci 2007; 14(4): 329–337. 93–99; discussion 100. 117 Taylor PD, McConnell J, Khan IY, Holemans K, 102 Vickers MH, Breier BH, McCarthy D, Gluckman PD. Lawrence KM, Asare-Anane H et al. Impaired glucose Sedentary behavior during postnatal life is deter- homeostasis and mitochondrial abnormalities in off- mined by the prenatal environment and exacerbated spring of rats fed a fat-rich diet in pregnancy. Am J by postnatal hypercaloric nutrition. Am J Physiol Physiol Regul Integr Comp Physiol 2005; 288(1): Regul Integr Comp Physiol 2003; 285(1): R271–R273. R134–R139. 103 Armitage JA, Taylor PD, Poston L. Experimental 118 Buckley AJ, Keseru B, Briody J, Thompson M, Ozanne models of developmental programming: conse- SE, Thompson CH. Altered body composition and quences of exposure to an energy rich diet during metabolism in the male offspring of high fat-fed rats. development. J Physiol 2005; 565(Pt 1): 3–8. Metabolism 2005; 54(4): 500–507. 104 Taylor PD, Poston L. Developmental programming 119 Gregorio BM, Souza-Mello V, Carvalho JJ, Mandarim- of obesity in mammals. Exp Physiol 2007; 92(2): de-Lacerda CA, Aguila MB. Maternal high-fat intake 287–298. predisposes nonalcoholic fatty liver disease in

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 418 C57BL/6 offspring. Am J Obstet Gynecol 2010; 203(5): states and metabolism. Naturwissenschaften 2003; 495 e491–495 e498. 90(11): 521–523. 120 den Boer M, Voshol PJ, Kuipers F, Havekes LM, 137 Wu BJ, Else PL, Storlien LH, Hulbert AJ. Molecular Romijn JA. Hepatic steatosis: a mediator of the activity of Na(+)/K(+)-ATPase from different sources metabolic syndrome. Lessons from animal models. is related to the packing of membrane lipids. J Exp Arterioscler Thromb Vasc Biol 2004; 24(4): 644–649. Biol 2001; 204(Pt 24): 4271–4280. 121 Siemelink M, Verhoef A, Dormans JA, Span PN, 138 Kirk SL, Samuelsson AM, Argenton M, Dhonye H, Piersma AH. Dietary fatty acid composition during Kalamatianos T, Poston L et al. Maternal obesity pregnancy and lactation in the rat programs growth induced by diet in rats permanently influences and glucose metabolism in the offspring. Diabetologia central processes regulating food intake in offspring. 2002; 45(10): 1397–1403. PLoS One 2009; 4(6): e5870. 122 Korotkova M, Gabrielsson BG, Holmang A, Larsson 139 Franke K, Harder T, Aerts L, Melchior K, Fahrenkrog BM, Hanson LA, Strandvik B. Gender-related long- S, Rodekamp E et al. ‘Programming’ of orexigenic and term effects in adult rats by perinatal dietary ratio of anorexigenic hypothalamic neurons in offspring of n-6/n-3 fatty acids. Am J Physiol Regul Integr Comp treated and untreated diabetic mother rats. Brain Res Physiol 2005; 288(3): R575–R579. 2005; 1031(2): 276–283. 123 Khan IY, Taylor PD, Dekou V, Seed PT, Lakasing L, 140 Bouret SG. Early life origins of obesity: role of Graham D et al. Gender-linked hypertension in hypothalamic programming. J Pediatr Gastroenterol offspring of lard-fed pregnant rats. Hypertension Nutr 2009; 48(Suppl 1): S31–S38. 2003; 41(1): 168–175. 141 Plagemann A, Harder T, Rake A, Janert U, Melchior K, 124 Khan I, Dekou V, Hanson M, Poston L, Taylor P. Rohde W et al. Morphological alterations of hypotha- Predictive adaptive responses to maternal high-fat lamic nuclei due to intrahypothalamic hyperinsulin- diet prevent endothelial dysfunction but not hyper- ism in newborn rats. Int J Dev Neurosci 1999; 17(1): tension in adult rat offspring. Circulation 2004; 37–44. 110(9): 1097–1102. 142 Levin BE, Dunn-Meynell AA, Banks WA. Obesity- 125 Tamaya-Mori N, Uemura K, Iguchi A. Gender prone rats have normal blood-brain barrier transport differences in the dietary lard-induced increase in but defective central leptin signaling before obesity blood pressure in rats. Hypertension 2002; 39(5): onset. Am J Physiol Regul Integr Comp Physiol 2004; 1015–1020. 286(1): R143–R150. 126 Parente LB, Aguila MB, Mandarim-de-Lacerda CA. 143 Ferezou-Viala J, Roy AF, Serougne C, Gripois D, Deleterious effects of high-fat diet on perinatal and Parquet M, Bailleux V et al. Long-term consequences postweaning periods in adult rat offspring. Clin Nutr of maternal high-fat feeding on hypothalamic leptin 2008; 27(4): 623–634. sensitivity and diet-induced obesity in the offspring. 127 Armitage JA, Khan IY, Taylor PD, Nathanielsz PW, Am J Physiol Regul Integr Comp Physiol 2007; 293(3): Poston L. Developmental programming of the R1056–R1062. metabolic syndrome by maternal nutritional imbal- 144 Burdge GC, Hanson MA, Slater-Jefferies JL, Lillycrop ance: how strong is the evidence from experimental KA. Epigenetic regulation of transcription: a mechan- models in mammals? J Physiol 2004; 561(Pt 2): ism for inducing variations in phenotype (fetal 355–377. programming) by differences in nutrition during early 128 Weisinger HS, Armitage JA, Sinclair AJ, Vingrys AJ, life? Br J Nutr 2007; 97(6): 1036–1046. Burns PL, Weisinger RS. Perinatal omega-3 fatty acid 145 Delage B, Dashwood RH. Dietary manipulation of deficiency affects blood pressure later in life. Nat histone structure and function. Annu Rev Nutr 2008; Med 2001; 7(3): 258–259. 28: 347–366. 129 Weinstock M, Matlina E, Maor GI, Rosen H, McEwen 146 Bird A. DNA methylation patterns and epigenetic BS. Prenatal stress selectively alters the reactivity of memory. Genes Dev 2002; 16(1): 6–21. the hypothalamic-pituitary adrenal system in the 147 Plagemann A, Harder T, Brunn M, Harder A, Roepke female rat. Brain Res 1992; 595(2): 195–200. K, Wittrock-Staar M et al. Hypothalamic proopiome- 130 Liu L, Li A, Matthews SG. Maternal glucocorticoid lanocortin promoter methylation becomes altered by treatment programs HPA regulation in adult off- early overfeeding: an epigenetic model of obesity and spring: sex-specific effects. Am J Physiol Endocrinol the metabolic syndrome. J Physiol 2009; 587(Pt 20): Metab 2001; 280(5): E729–E739. 4963–4976. 131 Pratt JH. Low-renin hypertension: more common than 148 Drake AJ, Tang JI, Nyirenda MJ. Mechanisms under- we think? Cardiol Rev 2000; 8(4): 202–206. lying the role of glucocorticoids in the early life 132 Haddy FJ, Pamnani MB. Role of ouabain-like factors programming of adult disease. Clin Sci (Lond) 2007; and Na-K-ATPase inhibitors in hypertension—some 113(5): 219–232. old and recent findings. Clin Exp Hypertens 1998; 149 Goldberg AD, Allis CD, Bernstein E. Epigenetics: 20(5–6): 499–508. a landscape takes shape. Cell 2007; 128(4): 133 Jaitovich A, Bertorello AM. Salt, Na+,K+-ATPase and 635–638. hypertension. Life Sci 2010; 86(3–4): 73–78. 150 Amaral PP, Mattick JS. Noncoding RNA in develop- 134 Iwamoto T, Kita S. Hypertension, Na+/Ca2+ exchan- ment. Mamm Genome 2008; 19(7–8): 454–492. ger, and Na+, K+-ATPase. Kidney Int 2006; 69(12): 151 Gluckman PD, Hanson MA, Buklijas T, Low FM, 2148–2154. Beedle AS. Epigenetic mechanisms that underpin 135 Jaitovich A, Bertorello AM. Salt, Na(+),K(+)-ATPase metabolic and cardiovascular diseases. Nat Rev and hypertension. Life Sci 2010; 86(3–4): 73–78. Endocrinol 2009; 5(7): 401–408. 136 Turner N, Else PL, Hulbert AJ. Docosahexaenoic acid 152 Langley-Evans SC. Nutritional programming of dis- (DHA) content of membranes determines molecular ease: unravelling the mechanism. J Anat 2009; 215(1): activity of the sodium pump: implications for disease 36–51.

Journal of Human Hypertension Fetal programming for metabolic syndrome ML de Gusma˜o Correia et al 419 153 Waterland RA, Jirtle RL. Transposable elements: 155 Aagaard-Tillery KM, Grove K, Bishop J, Ke X, Fu Q, targets for early nutritional effects on epigenetic gene McKnight R et al. Developmental origins of disease regulation. Mol Cell Biol 2003; 23(15): 5293–5300. and determinants of chromatin structure: maternal 154 Lillycrop KA, Slater-Jefferies JL, Hanson MA, Godfrey diet modifies the primate fetal epigenome. J Mol KM, Jackson AA, Burdge GC. Induction of altered Endocrinol 2008; 41(2): 91–102. epigenetic regulation of the hepatic glucocorticoid 156 Vucetic Z, Kimmel J, Totoki K, Hollenbeck E, Reyes receptor in the offspring of rats fed a protein- TM. Maternal high-fat diet alters methylation and restricted diet during pregnancy suggests that gene expression of dopamine and opioid-related reduced DNA methyltransferase-1 expression is genes. Endocrinology 2010; 151(10): 4756–4764. involved in impaired DNA methylation and changes 157 Waterland RA. Epigenetic epidemiology of obesity: in histone modifications. Br J Nutr 2007; 97(6): application of epigenomic technology. Nutr Rev 2008; 1064–1073. 66(Suppl 1): S21–S23.

Journal of Human Hypertension