174:6

J Leng and others and 174:6 763–773 Clinical Study gestational

Physical activity, sedentary behaviors and risk of mellitus: a population-based cross-sectional study in Tianjin, China

Junhong Leng1,2, Gongshu Liu2, Cuiping Zhang2, Shijuan Xin3, Fang Chen2, Baojuan Li2, Huiguang Tian2, Zhijie Yu4, Jaakko Tuomilehto5,6,7,8, Gang Hu9 and Xilin Yang1

1Department of Epidemiology and Biostatistics, School of Public , Tianjin Medical University, Tianjin, China, 2Tianjin Women and Children’s Health Centre, Tianjin, China, 3Tianjin Dong Li Women and Children’s Health Centre, Tianjin, China, 4Population Research Program and Department of Pediatrics, Dalhousie University, Halifax, Nova Scotia, Canada, 5Centre for Vascular Prevention, Correspondence Danube-University Krems, Krems, Austria, 6Diabetes Prevention Unit, National Institute for Health and should be addressed Welfare, Helsinki, Finland, 7Instituto de Investigacion Sanitaria del Hospital Universario LaPaz (IdiPAZ), to X Yang Madrid, Spain, 8Diabetes Research Group, King Abdulaziz University, Jeddah, Saudi Arabia and 9Chronic Email Epidemiology Laboratory, Pennington Biomedical Research Center, Baton Rouge, Louisiana, USA [email protected] or [email protected]

Abstract

Objective: Physical activity in a nonpregnant state or before reduces the risk of and is also associated with reduced risk of gestational diabetes mellitus (GDM). However, it is uncertain whether physical activity during pregnancy reduces the risk of GDM. Design and methods: Using an established universal screening system in Tianjin, China, we prospectively recruited 11 450 pregnant women within the 12th gestational week from 2010 to 2012. These women underwent a 50-g 1-h challenge test (GCT) at 24 – 28 weeks of gestation and a 75-g 2-h oral if GCT glucose European Journal European of Endocrinology ≥7.8 mmol/L. GDM was defined according to the International Association of Study Group’s criteria. Self-reported physical activity in the last month was collected at GCT time using a validated questionnaire. Results: GDM developed in 7.3% (n = 840) of the women. Women with GDM were less likely to be engaged in moderate-to-high physical activity during pregnancy than those without (79.8% vs 81.6%, P = 0.191). Moderate-to- high physical activity during pregnancy was associated with decreased risk of GDM (multivariable odds ratio (OR): 0.81, 95% confidence interval (CI): 0.67 – 0.97). at home for 2 – 4 h per day and >4 h per day were associated with significantly increased risk of GDM (multivariable OR of sitting time for 2 – 4 h vs <2 h: 1.59, 95% CI: 1.18 – 2.15; OR of sitting time for >4 h vs <2 h: 1.73, 95% CI: 1.22 – 2.43). Conclusions: Increased physical activity during pregnancy was associated with reduced GDM risk, whereas sedentary lifestyle was associated with increased GDM risk among Chinese pregnant women.

European Journal of Endocrinology (2016) 174, 763–773

Introduction

The prevalence of gestational diabetes mellitus (GDM) GDM is not only associated with adverse short-term has been increasing worldwide (1), and it comes to be pregnancy outcomes but also associated with long-term one of the most prevalent metabolic disorders during suboptimal health outcomes in both mothers and their pregnancy, affecting 16.8% of all (2). offspring (3, 4, 5).

www.eje-online.org © 2016 European Society of Endocrinology Published by Bioscientifica Ltd. DOI: 10.1530/EJE-15-1103 PrintedinGreat Britain

Downloaded from Bioscientifica.com at 10/01/2021 05:18:44AM via free access

10.1530/EJE-15-1103 Clinical Study J Leng and others Physical activity and 174:6 764 gestational diabetes

Physical inactivity and overnutrition are well- time of screening for GDM in mid-pregnancy and to established risk factors for diabetes, and lifestyle delivery (15). intervention aiming at increasing physical activity From October 2010 to August 2012, 19 669 and healthy diet can reduce diabetes risk among high- pregnant women attended the primary care hospital for risk subjects (6, 7, 8). However, it remains to establish registration of pregnancy and antenatal care within the a causal association between physical inactivity and first 12th gestational week. We sequentially excluded increased risk of GDM. In this connection, a few studies 1080 women who did not perform glucose challenge test have examined the association between physical activity (GCT) and 781 women with a positive GCT who did not during pregnancy and GDM risk (9, 10, 11, 12, 13), and perform the standard oral glucose tolerance test (OGTT). findings from these studies were inconsistent, possibly Of the remaining 17 808 women, 11 450 who agreed due to inadequate study designs, different diagnostic and completed the self-administered questionnaire on criteria for GDM, shortfall of statistical power, inaccurate physical activity and sedentary behaviors at the time measurement of physical activity, and inability to control of GCT were included in the current analysis. Ethics for confounders. is a well-recognized risk factor approval was obtained from the Ethics Committee for for GDM. Of note, the benefits of physical activity for Clinical Research of Tianjin Women and Children’s GDM may be influenced by body composition, although Health Centre (TWCHC), Tianjin, China, and informed few studies have examined whether the effect of physical consent was obtained from all the women. activity on GDM risk varies by obesity and/or before pregnancy. Women tend to be less engaged in Diagnosis of GDM physical activity and more likely to adopt sedentary behaviors during pregnancy compared with during their A two-step screening procedure was used to identify nonpregnant period. It is unknown whether reduced GDM cases. All pregnant women first underwent a 50 g sedentary lifestyle and increased physical activity during 1-h GCT in nonfasting status between 24 and 28 weeks pregnancy are possible measures for prevention of GDM. of gestation at primary care hospitals. The women with Therefore, the current analysis is aimed at examining plasma glucose (PG) ≥7.8 mmol/L were referred to the (i) associations of physical activity and sedentary behaviors, GDM clinic within TWCHC where they underwent a 75 g especially sitting time at home during pregnancy, with 2-h OGTT in the morning after at least 8 h of fasting. GDM risk, and (ii) consistency of associations of physical PGs at fasting, 1 h and 2 h after the glucose load were activity and sedentary behaviors during pregnancy with measured at the Central Laboratory of TWCHC using European Journal European of Endocrinology GDM risk by pre-pregnancy obesity and overweight. an automatic analyzer (TBA-120FR, Toshiba, Japan). The inter-assay and intra-assay coefficients of variation for glucose were <2.59%. GDM was diagnosed according to Subjects and methods the International Association of Diabetes and Pregnancy Study population and settings Study Group (IADPSG) cut-points, that is, any of the following PG values is met: fasting PG (FPG) ≥5.1 mmol/L, Tianjin is a municipality directly under the administration 1-h PG ≥10.0 mmol/L or 2-h PG ≥8.5 mmol/L (16). of the central government of China. It is situated in the eastern part of the North China, covering an area of 11 900 km2 with over 13 million residents. It has six central Assessment of physical activity and sedentary urban districts, one new urban district, four suburban behaviors districts and five counties. There are about 4.3 million Physical activity and sedentary behaviors in 1 month residents living in the six central urban districts. prior to the GCT time was collected by a self-administered Our group established a universal screening system questionnaire. This questionnaire was repeatedly used in for GDM in urban Tianjin in 1999 (14). Using this well- our previous studies (7, 17, 18), and detailed description established screening and management system, we and validation information are available elsewhere (19). conducted a large population-based prospective study To document physical activity of pregnancy women among pregnant women and their offspring in the six more accurately, we further included housework physical central urban districts from 2010 to 2012 with detailed activity. Briefly, physical activity during pregnancy documentation of clinical and biochemical profiles, included occupational, commuting, leisure-time, and longitudinally from their first antenatal care visit to the housework physical activities. Occupational physical

www.eje-online.org

Downloaded from Bioscientifica.com at 10/01/2021 05:18:44AM via free access European Journal of Endocrinology mass index(BMI)wascalculated asweightinkilogram for pregnancyandatthetime ofGCTrespectively. Body and inlightclothingtothe nearest0.1 nearest 0.5 protocol. Heightwasmeasured withoutshoestothe weight, andbloodpressure(BP)usingastandardized at thetimeofGCT. out bypregnantwomenatregistrationforpregnancyor collected usingquestionnairesbycareprovidersorfilled alcohol consumptionbeforeandduringpregnancywere parity, ethnicity, education attainment, smoking and ofdiabetesinthefirst-degree relatives, family history Maternal birth date, date of the last menstrual period, Other assessments categories: <2,2 meal time.Sittingtimeathomewasstratifiedintothree the computer, andothersittingtimesathome,including sitting athome,includingwatchingTV, reading,using activity, weaskedhowmanyhoursweredailyspenton activities was‘moderate-to-high’.Inadditiontophysical women reportedatleastonelevelofthefourtypes of activitieswere‘low’and‘moderate-to-high’when ‘low’ whenwomenreportedthelevelsofallfourtypes categories basedontheabovefourtypesofactivities: during pregnancyweremergedandregroupedintotwo day (moderate),and was classifiedintothreecategories:<1 week, moderate-to-high).Houseworkphysicalactivity engaged inthisactivitymorethan3 example, walking,jogging,swimming,ballgames,and this activitylessthan3 categories: noleisure-timephysicalactivity/engagedin Leisure-time physicalactivitywasclassifiedintotwo high’ whensubjectsreportedmorethan30 subjects reported0 30 day (1 cycling (0 journey: usingmotorizedtransportation/nowalkingor from work,anddailydurationofthiscommutingreturn categories accordingtothetypeoftransportationand Commuting physicalactivitywasdividedintothree or heavymanuallabor, e.g.farmerandtextileworkers). teacher), andhigh(workincludingwalkinglifting including standingandwalking,e.g.,storeassistant moderate(work including sitting,e.g.,secretary), activity wasdividedintothreecategories:low(work Clinical Study min perday( Nurses at primary carehospitalsmeasured height, Nurses atprimary – 29 min/day), walking or cycling 1 min/day), and walking or cycling more than cm. Weights were measuredwithoutshoes – ≥ 4, and 30 – min/day). We renamed‘low’when 29 ≥ 2 min perday, and‘moderate-to- h perweek(<3 ≥ h/day (high).Physicalactivities 4 h/day. J Lengandothers h/day (low),1 h perweek( h/week, low),for kg atregistration min perday. – 29 min per ≥ – 3 h/ 2 h/ physical activity and sedentary behaviors for GDM risk. physical activity and sedentary (CIs)of ratios (ORs)andtheir95%confidenceintervals logisticregressionmodelswereusedtoobtainodds Binary Statistical analysis 10 a calibratedmercury sphygmomanometerafteratleast pregnancy andattheGCT. SittingBPwasmeasuredusing as thedifferenceinweightbetweenatregistrationfor Weight gainfrompre-pregnancytoGCTwascalculated first trimesterofpregnancyissmallandnegligible( used aspre-pregnancyweightbecausegaininthe China ( recommended bytheWorking GrouponObesityin m (18.5 –23.9kg/m weight (<18.5 was categorizedasunderweight (kg) dividedbysquaredbodyheightinmeter(m)and two-tailed testwasconsidered tobestatisticallysignificant. all thestatisticalanalyses.A final analysisandthoseexcluded. and biochemical characteristics of women included in the was alsoperformedtocomparethedistributionsofclinical impacts of exclusion of those subjects. Additional analysis GCT result but were absent from OGTT to check the Secondly, wereincluded455subjectswhohad apositive to checktheconsistencyacrosstwosetsofcriteria. used the1999WHOcriteriafordiagnosisofGDM( that is,BMI use oftheWorld HealthOrganization(WHO)’s criteria, criteria ( statusaccordingtotheChinese by obesity/overweight BMIs wasperformedtochecktheconsistencyofORs P step-down Bonferroniprocedure( activity, behaviors,andGDMrisk.Ryan-Holm sedentary to checkitseffectontheassociationbetweenphysical Fourthly, wefurtheradjustedforpre-pregnancyBMI systolic BPandweightgainfrompre-pregnancytoGCT. in first-degree relatives, parity, education, ethnicity, ofdiabetes confounders, includingheight,familyhistory during pregnancy. Thirdly, wefurtheradjustedforother including habitual smoker and alcohol drinker before or Secondly, we further adjusted for lifestyle confounders, for confounders. First, we adjusted for maternal age. A structuredadjustmentschemewasusedtocontrol gestational diabetes Physical activityand -values and95%CIswhereappropriate. 2 min ofrestatregistrationforpregnancy. ), andobesity( IBM SPSS Statistics 20.0 (IBM SPSS) was used to perform IBM SPSSStatistics20.0(IBM SPSS)wasusedtoperform Two sensitivityanalyseswerealsoperformed.First,we Further subgroupanalysisbydifferentpre-pregnancy 20 20 ). Weight atregistrationforpregnancywas ). Thesubgroupanalysiswasrepeatedbythe ≥ 25.0 kg/m ≥ 28.0 kg/m 2 to define overweight andobesity. todefineoverweight Downloaded fromBioscientifica.com at10/01/202105:18:44AM 2 , vregt (24.0 –27.9kg/ overweight ), P -value of less than 0.05 for a -value oflessthan0.05fora 2 ) accordingtothecriteria 22 ) wasusedtoadjust 174 www.eje-online.org :6 kg/m 2 ), normal 765 21 23 via freeaccess ). ) Clinical Study J Leng and others Physical activity and 174:6 766 gestational diabetes

Results The results were similar if the WHO’s criteria were used to define overweight and obesity. Characteristics of the study population

The mean age of 11 450 women was 28.5 (s.d.: 2.8) Sensitivity analysis years, 96.9% were nulliparous, 19.8% (n = 2263) were overweight, and 6.5% (n = 745) were obese. Of them, Using the 1999 WHO’s criteria for the definition of GDM 81.4% were engaged in moderate-to-high level of physical (n = 732), the protective effects of moderate-to-high activity, and 46.9 and 43.2% of women were sitting at physical activity persisted (adjusted ORs: 0.79, 0.65 – 0.95) home for 2 – 4 h and ≥4 h per day respectively. (Table 4). Also, the graded direct association between A total of 840 women (7.3%) developed GDM. sitting time at home and GDM risk was still significant Women with GDM were more likely to be engaged in (P for trend = 0.040). If we assumed that all the 455 three or more hours of leisure-time physical activity to sit subjects who had a positive GCT result but were absent at home for more hours (Table 1). Occupational physical from a standard OGTT did not have GDM, reinclusion activity, commuting physical activity, and housework of them in the analysis did not change the effect sizes physical activity were similar in women with GDM and of moderate-to-high physical activity and sitting time at those without. In addition, women with older age, shorter home for GDM. However, if we assumed that all those height, higher pre-pregnancy BMI, higher systolic/ women had GDM, reinclusion of them in the analysis diastolic BP at registration for pregnancy, and lower resulted in attenuated ORs of moderate-to-high physical education level, with family in first- activity (still significant) and nonsignificant ORs of sitting degree relatives and habitual use of tobacco and alcohol time at home. before pregnancy, were more likely to develop GDM.

Additional analysis Physical activity and sedentary behaviors for There were no statistical differences in most of the GDM risk measured variables between the women included in the In the multivariable analysis models 3 and 4, moderate- analysis and the women excluded except for maternal to-high physical activity during pregnancy was associated age, family history of diabetes, multipara and alcohol with significantly reduced risk of GDM (OR in model 4: drinking habit (Table 5).

European Journal European of Endocrinology 0.81, 95% CI: 0.67 – 0.97). In all the four multivariable analysis models, sitting time at home for 2 – 4 h per day Discussion and more than 4 h per day significantly increased GDM risks compared with sitting time for less than 2 h per day Our study found that (i) physical activity during in an ordinal manner (ORs and 95% CIs in model 4: 1.59, pregnancy was associated with reduced risk of GDM, 1.18 – 2.15 for 2 – 4 h/day vs <2 h/day; 1.73, 1.22 – 2.43 especially among women who were overweight or obese for ≥4 h/day vs <2 h/day; P for trend: 0.002) (Table 2). before pregnancy, and (ii) sedentary behaviors during Regardless of physical activity or sitting time at home, the pregnancy were also associated with the increased risk of ORs from model 3 to model 4 just changed slightly after GDM, among overweight/obese women as well as among further adjusting pre-pregnancy BMI. women with normal weight. Previous studies reported inconsistent findings regarding the association between physical activity Subgroup analysis of physical activity and sedentary during pregnancy and GDM risk (9, 10, 11, 12, 13). behaviors for GDM risk A small case–control study of 541 pregnant women The OR of moderate-to-high physical activity for GDM found that recreational physical activity during the decreased to nonsignificance among women with first 20 weeks of pregnancy was associated with a 48% BMI <24.0 kg/m2 but increased among women with decrease in GDM risk (OR: 0.52, 95% CI: 0.33 – 0.80) BMI ≥24.0 kg/m2 (Table 3). Similarly, the OR of sitting (10). More recently, a meta-analysis of five large studies time at home for GDM also numerically decreased among reported that physical activity in early pregnancy was women with BMI <24.0 kg/m2 (although still significant) highly protective against GDM (a pooled OR: 0.76, but increased among women with BMI ≥24.0 kg/m2. 95% CI: 0.70 – 0.83) (13). More recent RCTs also generated

www.eje-online.org

Downloaded from Bioscientifica.com at 10/01/2021 05:18:44AM via free access European Journal of Endocrinology physical activity, <3 from work; at least6months; a BMI, bodymassindex;GCT, glucosechallengetest. Variables atGCT Variables atregistration forpregnancy n by theIADPSG’s criteria.Dataarereportedinmean Table 1 Derived fromStudent’s Sitting timeathomeduringpregnancy(h/day) Physical activityduringpregnancy Housework physicalactivityduringpregnancy(h/day) Leisure-time physicalactivityduringpregnancy(h/week) Commuting physicalactivityduringpregnancy(min/day) Occupational physicalactivityduringpregnancy Alcohol drinkinghabit Smoking habit Weight gainfrompre-pregnancytoGCT(kg) Gestational ageatGCT(weeks) Education >12 years Family historyofdiabetesinfirst-degreerelatives Han ethnicity Parity Systolic bloodpressure(mmHg) Gestational ageatregistrationforpregnancy(weeks) Pre-pregnancy BMIgroup(kg/m Pre-pregnancy BMI(kg/m Body height(cm) Age group(years) Age (years) Diastolic bloodpressure(mmHg) Body heightgroup(cm) Clinical Study ≥ 2 –4 <2 Moderate-to-high Low ≥ 1 –2 <1 ≥ <3 ≥ 1 –29 0 High Moderate Low Alcohol drinkerduringpregnancy Alcohol drinkerbeforepregnancy Habitual smokerduringpregnancy Habitual smokerbeforepregnancy ≥ ≥ ≥ <18.5 ≥ ≥ ≥ ≥ <30 <160 4 2 3 30 28 24 18.5 165 160 35 30 − − ≥ − <28 <35 − 1 <165 <24 Clinical andbiochemicalcharacteristicsofsubjectsaccordingtotheoccurrencegestationaldiabetesmellitusdefined f Low wasdefinedwhensubjectssimultaneously reportedthelightlevelofoccupationalphysicalactivity, 0 d Defined ashavingsmokedoneor more cigarettesperdayduringpregnancy; h/week ofleisure-timephysicalactivity, and<1 t -test; b 2 Derived from ) 2 ) f J Lengandothers d c χ 2 testorFisher’s exacttest; ± s.d. or h/day ofhouseworkphysicalactivity; andothersweredefinedas‘moderate-to-high’. e n (%). c Defined ashavingcontinuouslysmoked oneormorecigarettesperdayfor gestational diabetes Physical activityand 10 128 (95.5%) 4554 (42.9%) 4979 (46.9%) 1077 (10.2%) 8656 (81.6%) 1954 (18.4%) 2061 (19.4%) 6021 (56.7%) 2528 (23.8%) 1204 (11.3%) 9406 (88.7%) 9794 (92.3%) 1311 (12.4%) 9293 (87.6%) 3356 (31.6%) 1982 (18.7%) 6931 (65.3%) 1073 (10.1%) 4269 (40.2%) 4603 (43.4%) 2241 (21.1%) 8130 (76.6%) 8918 (84.1%) 1736 (16.4%) 163.2 ±4.7 105.4 ±10.7 566 (5.3%) 250 (2.4%) 319 (3.0%) 328 (3.1%) 621 (5.9%) 239 (2.3%) 847 (8.0%) Non-GDM 24.7 ±2.5 68.3 ±7.7 22.2 ±3.3 28.4 ±2.8 99 (0.9%) 71 (0.7%) 7.5 ±3.4 9.8 ±1.6 10 610 6 (0.1%) e Defined bythetimespentinwalking orcyclingtoand 391 (46.5%) 395 (47.0%) 670 (79.8%) 170 (20.2%) 165 (19.6%) 460 (54.8%) 215 (25.6%) 126 (15.0%) 714 (85.0%) 785 (93.5%) 745 (88.7%) 295 (35.1%) 124 (14.8%) 281 (33.5%) 407 (48.5%) 312 (37.1%) 371 (44.2%) 157 (18.7%) 230 (27.4%) 561 (66.8%) 687 (81.8%) 132 (15.7%) 810 (96.4%) 162.8 ±4.7 108.7 ±11.3 Downloaded fromBioscientifica.com at10/01/202105:18:44AM 95 (11.3%) 54 (6.4%) 38 (4.5%) 17 (2.0%) 10 (1.2%) 37 (4.4%) 31 (3.7%) 28 (3.3%) 49 (5.8%) 24.8 ±1.9 70.8 ±8.1 24.3 ±3.9 29.5 ±3.2 0 (0.0%) 7 (0.8%) 7.5 ±3.8 9.9 ±1.6 GDM 840 – 29 min/day ofcommuting 174 www.eje-online.org :6 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 P 0.628 0.001 0.191 0.454 0.001 0.485 0.460 0.037 0.578 0.025 0.082 0.338 0.104 0.190 0.766 0.203 0.542 0.020 -value b b b a a a a b b b b b b b b b b b b b b a a a a 767 via freeaccess Clinical Study J Leng and others Physical activity and 174:6 768 gestational diabetes

Table 2 The ORs of physical activity and sitting time at home during pregnancy for gestational diabetes mellitus defined by the IADPSG’s criteria.

n (%) OR 95% CI P-value Multivariable analysis 1a Physical activity during pregnancy Low 170 (8.0%) 1.00 Reference Moderate-to-high 670 (7.2%) 0.84 0.70 – 1.01 0.057 Sitting time at home during pregnancy (h/day) 0.001f <2 54 (4.8%) 1.00 Reference 2 – 4 395 (7.4%) 1.62 1.20 – 2.17e 0.001e ≥4 391 (7.9%) 1.77 1.26 – 2.48e <0.001e Multivariable analysis 2b Physical activity during pregnancy Low 170 (8.0%) 1.00 Reference Moderate-to-high 670 (7.2%) 0.84 0.70 – 1.01 0.057 Sitting time at home during pregnancy (h/day) 0.001f <2 54 (4.8%) 1.00 Reference 2 – 4 395 (7.4%) 1.60 1.20 – 2.15e 0.002e ≥4 391 (7.9%) 1.74 1.24 – 2.44e <0.001e Multivariable analysis 3c Physical activity during pregnancy Low 170 (8.0%) 1.00 Reference Moderate-to-high 670 (7.2%) 0.83 0.69 – 0.99 0.040 Sitting time at home during pregnancy (h/day) 0.001f <2 54 (4.8%) 1.00 Reference 2 – 4 395 (7.4%) 1.61 1.20 – 2.17e 0.002e ≥4 391 (7.9%) 1.76 1.25 – 2.48e <0.001e Multivariable analysis 4d Physical activity during pregnancy Low 170 (8.0%) 1.00 Reference Moderate-to-high 670 (7.2%) 0.81 0.67 – 0.97 0.022 Sitting time at home during pregnancy (h/day) 0.002f <2 54 (4.8%) 1.00 Reference 2 – 4 395 (7.4%) 1.59 1.18 – 2.15e 0.002e ≥4 391 (7.9%) 1.73 1.22 – 2.43e <0.001e European Journal European of Endocrinology

aVariables adjusted included maternal age group; bFurther adjusted for habitual smokers before or during pregnancy and alcohol drinkers before or during pregnancy in addition to the variables listed in the footnote ‘a’; cFurther adjusted for maternal height group, family history of diabetes in first-degree relatives, parity≥ 1, education >12 years, Han nationality, systolic BP at registration for pregnancy, and from pre-pregnancy to GCT in addition to the variables listed in the footnote ‘b’; dFurther adjusted for pre-pregnancy BMI group in addition to the variables listed in the footnote ‘c’; eP-values and 95% CIs of ORs were adjusted for multiple comparisons by Ryan-Holm step-down Bonferroni procedure; fP for trend.

conflicting findings. A randomized controlled trial of Diabetes Prevention Study (26) reported that a moderate 855 healthy pregnant women with normal BMI failed to individualized lifestyle intervention on diet, physical confirm that a 12-week standard program during activity, and weight control could reduce the incidence the second half of pregnancy was able to prevent GDM of GDM by 39% in high-risk pregnant women with a (12). The LIMIT multicenter, randomized trial found that history of GDM and/or a pre-pregnancy BMI ≥30 kg/m2. antenatal dietary and lifestyle interventions, including Indeed, different study designs and diagnostic criteria for a combination of dietary, exercise, and behavioral GDM, different definitions of physical activity during strategies, were unable to reduce the risk of GDM and pregnancy, and timing of initiation of the intervention to improve pregnancy outcomes among overweight or may contribute to the inconsistent results. Our findings obese women (24). Similarly, another recent multicenter, support the notion that early lifestyle intervention, randomized controlled trial, the UK Pregnancies Better including physical activity during pregnancy, can reduce Eating and Activity Trial (UPBEAT) study, reported that the risk of GDM in Chinese pregnant women, although a behavioral intervention addressing diet and physical further larger and well-designed RCTs are needed to activity in obese pregnant women was also unable to confirm it. We noticed that women with GDM were more reduce GDM risk (25). However, the Finnish Gestational likely to be engaged in three or more hours of leisure-time

www.eje-online.org

Downloaded from Bioscientifica.com at 10/01/2021 05:18:44AM via free access European Journal of Endocrinology between sedentary behaviorsandtheriskoftype2 between sedentary Although some studies showed a direct association pregnancy andGDMriskhas notbeenwelladdressed. inconsistency. time physicalactivity<3 higher pre-pregnancyBMIamongwomenwithleisure- multipara wasassociatedwithGDM.Itispossiblethat <3 multipara thanthosewithleisure-timephysicalactivity had ahigherpre-pregnancyBMIandlesslikelytobe the womenwithleisure-timephysicalactivity pre-pregnancy BMI and multipara (data was not shown), with leisure-timephysicalactivity with leisure-timephysicalactivity<3 no statisticaldifferencesinmostvariablesbetweenwomen physical activitythanthosewithout.Althoughtherewere for multiplecomparisonsbyRyan-Holmstep-downBonferroniprocedure; a Among womenwithBMI Among womenwithBMI<25 Among womenwithBMI Among womenwithBMI<24 mellitus definedbytheIADPSG’s criteriastratifiedbyoverweightandobesitystatus. Table 3 Variables adjustedinthemultivariableanalysisaresameasfour Sitting timeathomeduringpregnancy(h/day) Physical activityduringpregnancy Sitting timeathomeduringpregnancy(h/day) Physical activityduringpregnancy Sitting timeathomeduringpregnancy(h/day) Physical activityduringpregnancy Sitting timeathomeduringpregnancy(h/day) Physical activityduringpregnancy Clinical Study ≥ 2 <2 Moderate-to-high Low ≥ 2 <2 Moderate-to-high Low ≥ 2 <2 Moderate-to-high Low ≥ 2 <2 Moderate-to-high Low

h/week. Inouranalysis,pre-pregnancyBMIbutnot

– – – – 4 4 4 4 The association between sedentary behaviorsduring The associationbetweensedentary

4 4 4 4

Multivariable analysisofORsphysicalactivityandsittingtimeathomeduringpregnancyforgestationaldiabetes ≥ ≥

h/week mightexplainpartofthe 25 24

kg/m kg/m

kg/m kg/m 2 2 2 2 J Lengandothers a a a a ≥ 3

h/week exceptfor h/week andthose ≥ 3

h/week 417 (5.5%) 104 (5.9%) 205 (5.7%) 200 (5.1%) 186 (14.1%) 195 (13.7%) 322 (12.9%) 151 (15.9%) 151 (14.8%) 253 (14.1%) 348 (5.1%) 240 (6.0%) 244 (5.6%) 83 (16.1%) 66 (18.3%) 24 (9.0%) 30 (3.5%) 87 (5.4%) 17 (9.0%) 37 (3.9%) n (%) c P fortrend. pregnancy, physiological insulinresistancedevelopsto contribute tothedevelopment ofGDM.Duringnormal confounders. after adjustingforotherphysical activitypatternsand pregnant womenweredirectlyassociatedwithGDMrisk behaviors inChinese is thefirsttoreportthatsedentary GDM inChinesepregnantwomen,but,moreimportantly, that physicalactivitywasassociatedwithareducedrisk of risk ( during pregnancyandGDMorabnormalglucosetolerance find anyassociationbetweentelevisionviewingbefore or behaviorsduringpregnancy butfailedto with sedentary Of them, only one investigated the association of GDM behaviorsandGDM risk( between sedentary knowledge, onlythreestudiesexaminedtheassociation diabetes innonpregnantpopulation( gestational diabetes Physical activityand Both insulinresistanceand impaired 29 ). Inthisregard,ourstudynotonlyreconfirmed 0.88 1.00 1.78 1.77 1.00 0.75 1.00 1.71 1.49 1.00 0.88 1.00 2.09 1.96 1.00 0.72 1.00 1.61 1.47 1.00 OR Table 2 ; b P Downloaded fromBioscientifica.com at10/01/202105:18:44AM -values and95%CIsofORswereadjusted 1.05 1.12 1.09 1.01 1.13 1.15 1.07 1.03 Reference Reference Reference Reference Reference Reference Reference Reference 0.70 0.57 0.68 0.52 95% CI

– – – – – – – –

– – – –

3.01 2.79 2.69 2.21 3.89 3.37 2.43 2.11

1.10 0.99 1.12 0.98 b b b b b b b b 174 www.eje-online.org :6 27 β , -cell function -cell function 28 P 29 0.249 0.042 0.034 0.027 0.014 0.016 0.048 0.015 0.014 0.019 0.034 0.287 0.040 0.024 0.027 0.034 -value ), toour , 30 b b b b b b b b c c c c 769 , 31 via freeaccess ). Clinical Study J Leng and others Physical activity and 174:6 770 gestational diabetes

Table 4 Sensitivity analysis of multivariable ORs of physical activity and sitting time at home during pregnancy for gestational diabetes mellitus. Sensitivity analysis 2: Reinclusion of the 455 subjects who had a positive GCT result but did not have a standard OGTT and assuming that all the 455 subjects did not have GDM. Sensitivity analysis 3: Reinclusion of the 455 subjects who had a positive GCT result but did not have a standard OGTT and assuming that all the 455 subjects had GDM.

n (%) OR 95% CI P-value Sensitivity analysis 1a,c Physical activity during pregnancy Low 155 (7.3%) 1.00 Reference Moderate-to-high 577 (6.2%) 0.79 0.65 – 0.95 0.014 Sitting time at home during pregnancy (h/day) 0.040d <2 59 (5.2%) 1.00 Reference 2 – 4 327 (6.1%) 1.16 0.87 – 1.56e 0.313e ≥4 346 (7.0%) 1.37 0.98 – 1.91e 0.074e Sensitivity analysis 2b,c Physical activity during pregnancy Low 170 (7.7%) 1.00 Reference Moderate-to-high 670 (6.9%) 0.81 0.68 – 0.98 0.026 Sitting time at home during pregnancy (h/day) 0.001d <2 54 (4.5%) 1.00 Reference 2 – 4 395 (7.1%) 1.62 1.20 – 2.18e 0.001e ≥4 391 (7.6%) 1.76 1.25 – 2.47e <0.001e Sensitivity analysis 3b,c Physical activity during pregnancy Low 260 (11.7%) 1.00 Reference Moderate-to-high 1035 (10.7%) 0.84 0.73 – 0.98 0.026 Sitting time at home during pregnancy (h/day) 0.287d <2 115 (9.6%) 1.00 Reference 2 – 4 604 (10.8%) 1.14 0.92 – 1.41e 0.240e ≥4 576 (11.2%) 1.19 0.93 – 1.53e 0.236e

N (%), number of cases (percentage of number at risk). aGDM was diagnosed according to the 1999 WHO criteria; bGDM was diagnosed according to the IADPSG cut-points; cVariables adjusted in the multivariable analysis are the same as in the multivariable analysis 4 in Table 2; dP for trend; eP-values and 95% CIs of ORs were adjusted for multiple comparisons by Ryan-Holm step-down Bonferroni procedure.

European Journal European of Endocrinology support the demands of fetal growth and development, Maternal pre-pregnancy BMI is considered to play a role which begins at mid-pregnancy and continues until delivery. in the association between physical activity and GDM risk. As a result, secretion of by pancreatic β-cells increases Dye et al. used the Central New York Regional Perinatal Data by 2 – 2.5 times to compensate for (32). System and reported that physical activity during pregnancy Compared with women with normal glucose tolerance, was associated with a 47% reduction of GDM risk but only those with GDM are unable to secrete enough insulin to among women with BMI >33.0 kg/m2 (11). The enhanced meet the metabolic stress of insulin resistance. Increased protective effect of increased physical activity among physical activity improves peripheral insulin sensitivity and overweight and obese women and the attenuated protective regulates glucose levels in nonpregnant population and is effect among women with normal weight revealed in our also associated with reduced first-phase insulin response in study are more likely to suggest that pre-pregnancy BMI late pregnancy (33). Apart from hormones secreted from may play a moderating role between physical activity and , recent data suggested that adipose tissue-derived GDM risk. Thus, physical activity may be more effective mediators, such as adiponectin, leptin, resistin, tumor on the prevention of GDM among overweight and obese necrosis factor-alpha, and oxidative stress, may be involved women because obesity-related insulin resistance may in the pathogenesis of insulin resistance in GDM (34, 35). play a more important role than impaired β-cell function Physical activity-induced improvements in glycemic control in GDM development among overweight and obese might be due to increased insulin sensitivity, beneficial women. By contrast, for and normal weight changes in adipokines, and reduced oxidative stress and women, impaired β-cell function may be dominant in the antioxidant effects (36). development of GDM, resulting in an attenuated benefit of The prevalence of GDM is increasing in parallel with physical activity for the prevention of GDM. Nevertheless, rising overweight and obesity in pregnant population. our data support the notion that increased sedentary

www.eje-online.org

Downloaded from Bioscientifica.com at 10/01/2021 05:18:44AM via free access European Journal of Endocrinology Body height(cm) Maternal age(years) n presented asmean Table 5 Gestational ageatregistration(weeks) Pre-pregnancy BMI(kg/m Alcohol drinkinghabit Smoking habit Education >12years Family historyofdiabetesinfirst-degreerelatives Han ethnicity Parity Systolic bloodpressureatregistration(mmHg) Diastolic bloodpressureatregistration(mmHg) 6 months; a BMI, bodymassindex. alcohol drinkingarepossible riskfactorsforGDM.Inthis are establishedriskfactors for GDM,andmultipara ( age, familyhistory, multipara,andalcoholdrinking women includedandthose excluded exceptformaternal of clinicalandbiochemicalcharacteristicsweresimilar in questionnaire ormissingothervariables.Thedistributions the analysisduetononresponsephysicalactivity high proportionofwomen(42%)werenotincluded in physical activitylevelsremainlow( Chinese andotherpopulations,especiallyinthosewhose behaviors duringpregnancy, andGDM is causalin that theassociationbetweenphysicalactivity, sedentary randomized controlledtrialswillbewarrantedtoshow and non-Chinesepopulations.Iffurtherconformed, and thefindingsneedtobereplicatedinotherChinese limitations. First, this study was a cross-sectional study, after adjustingforavarietyofconfounders.Thestudyhad for GDMwereavailableandtheresultssignificant of OGTTdata,and(ii)nearlyalltraditionalriskfactors well asreinclusionofwomenwithincompletecollection WHO’s criteriaortheIADPSG’s criteriatodefineGDMas robust because(i)ORswereconstantusingeitherthe1999 associations betweenphysicalactivityandGDMriskwere based studywithalargesamplesize.Secondly, the strength isthatthestudywasaprospectivepopulation- women withnormalpre-pregnancybodyweight. behaviors duringpregnancyincreaseGDMriskamong Derived fromStudent’s al 5 Table Habitual smokerduringpregnancy Habitual smokerbeforepregnancy Alcohol drinkerduringpregnancy Alcohol drinkerbeforepregnancy Clinical Study Our study has strengths and limitations. The first ≥

1 d ). Maternal age and family history ofdiabetes ). Maternalageandfamily history Defined ashavingsmokedoneormorecigarettesperdayduringpregnancy. Clinical andbiochemicalcharacteristicsofsubjectsaccordingtoinclusionorexclusioncriteriainanalysis.Dataare

±

s.d. or t -test; 2 ) b Derived fromChi-squaredtest; n (%). J Lengandothers d c 37 ). Secondly, a c Defined ashavingcontinuouslysmokedoneormorecigarettesperdayforatleast Excluded from theanalysis 6508 (84.1%) 7418 (95.5%) 2243 (28.9%) 163.1 105.6 226 (2.9%) 603 (7.8%) 290 (3.7%) 22.3 28.6 68.5 23 (0.3%) 36 (0.5%) 9.8 7764 was significantly associated with reduced GDM was significantlyassociated withreducedGDM found thathighphysical activityduringpregnancy especially amongmultiparous women. study subjects,andourfindings needfurtherreplications, the study. Seventhly, nullipara accounted for 97% of our activity onGDMmightbehigherthanthosereported in procedure. Consequently, thetrueeffectsizes ofphysical cases may have failed to be identified by the two-step procedure todetectGDMinthepast15years.Some our antenatalcaresystem(alsostudy)usedatwo-step of aone-stepOGTTapproachtoidentifyGDM,whereas to pregnancy. Sixthly, theIADPSG recommendedtheuse was of physicalactivityduringpregnancyontheriskGDM statistical analysis.Itispossiblethatpartofthebenefit and itsconfoundingeffectcouldnotberemovedby Fifthly, pre-pregnancy physical activity was not assessed, so thepotentialself-reportingbiascouldnotbeexcluded. activity wasself-reportedandnotobjectivelymeasured, not inpregnantwomenpopulations.Also,thephysical old andonlyvalidatedinnonpregnantpopulations,but study. Fourthly, thephysicalactivityquestionnairewas excluded. Thirdly, foodhabitwasnotcollectedinthis of existenceunmeasuredconfounderscouldnotbe the highnonrespondingrate.However, thepossibility at least,partiallyremovetheselectionbiasstemmingfrom these variablesinthemultivariableanalyses,whichmight, regard, wehadpaidspecialattentiontoadjustmentforall gestational diabetes Physical activityand

± ± ± ± ± ±

4.7 3.4 3.0 10.6 7.7 1.6 In conclusion, our large population-based study In conclusion,ourlarge population-based study attributable to goodhabitofphysicalactivityprior Included intheanalysis 11 377 (95.6%) 1034 (8.7%) 9974 (83.8%) 3786 (31.8%) 105.7 163.2 378 (3.2%) 119 (1.0%) 374 (3.1%) Downloaded fromBioscientifica.com at10/01/202105:18:44AM 22.4 28.5 68.5 81 (0.7%) 11 905 9.8

± ± ± ± ± ±

3.4 2.8 10.8 7.8 1.6 4.7 174 www.eje-online.org :6 <0.001 <0.001 <0.001 P 0.053 0.294 0.558 0.023 0.943 0.024 0.226 0.202 0.713 0.442 0.157 -value 771 b b b b b b b b a a a a a a via freeaccess Clinical Study J Leng and others Physical activity and 174:6 772 gestational diabetes

risk in Chinese pregnant women, especially among 6 Colberg SR, Sigal RJ, Fernhall B, Regensteiner JG, Blissmer BJ, pre-pregnancy overweight and obese women. The harmful Rubin RR, Chasan-Taber L, Albright AL & Braun B. Exercise and type 2 diabetes: the American College of Sports Medicine and the American effect of sedentary lifestyle was not limited to overweight Diabetes Association: joint statement executive summary. and obese women but extended to women with normal Diabetes Care 2010 33 2692–2696. (doi:10.2337/dc10-1548) body weight before pregnancy. Given the increasing 7 Hu G, Lindstrom J, Valle TT, Eriksson JG, Jousilahti P, Silventoinen K, Qiao Q & Tuomilehto J. Physical activity, , prevalence of GDM in China and other parts of the world, and risk of type 2 diabetes in patients with normal or impaired replications of our findings in other populations of pregnant glucose regulation. Archives of Internal Medicine 2004 164 892–896. women are warranted. If further replicated, randomized (doi:10.1001/archinte.164.8.892) 8 Jeon CY, Lokken RP, Hu FB & van Dam RM. Physical activity of controlled trials are needed to confirm the benefits of moderate intensity and risk of type 2 diabetes: a systematic review. lifestyle interventions for GDM, including increased Diabetes Care 2007 30 744–752. (doi:10.2337/dc06-1842) physical activity and reduced sedentary behaviors. 9 Callaway LK, Colditz PB, Byrne NM, Lingwood BE, Rowlands IJ, Foxcroft K & McIntyre HD. Prevention of gestational diabetes: feasibility issues for an exercise intervention in obese pregnant women. Diabetes Care 2010 33 1457–1459. (doi:10.2337/dc09-2336) Declaration of interest 10 Dempsey JC, Butler CL, Sorensen TK, Lee IM, Thompson ML, The authors declare that there is no conflict of interest that could be Miller RS, Frederick IO & Williams MA. A case-control study of perceived as prejudicing the impartiality of the research reported. maternal recreational physical activity and risk of gestational diabetes mellitus. Diabetes Research and Clinical Practice 2004 66 203–215. (doi:10.1016/j.diabres.2004.03.010) 11 Dye TD, Knox KL, Artal R, Aubry RH & Wojtowycz MA. Physical activity, Funding obesity, and diabetes in pregnancy. American Journal of Epidemiology 1997 This project was supported by BRIDGES (Grant number: LT09-227). 146 961–965. (doi:10.1093/oxfordjournals.aje.a009223) BRIDGES is an International Diabetes Federation program supported by 12 Stafne SN, Salvesen KA, Romundstad PR, Eggebo TM, Carlsen SM & an educational grant from Lilly Diabetes. Morkved S. Regular exercise during pregnancy to prevent gestational diabetes: a randomized controlled trial. & Gynecology 2012 119 29–36. (doi:10.1097/AOG.0b013e3182393f86) Author contribution statement 13 Tobias DK, Zhang C, van Dam RM, Bowers K & Hu FB. Physical X Y, G H and Z Y conceived and designed the study. All authors, except activity before and during pregnancy and risk of gestational Z Y, J T, G H, and X Y, contributed to the collection of the data. J L analyzed diabetes mellitus: a meta-analysis. Diabetes Care 2011 34 223–229. the data and wrote the first draft. All authors gave critical comments and (doi:10.2337/dc10-1368) contributed to the writing of the manuscript. X Y (the corresponding 14 Yang X, Hsu-Hage B, Zhang H, Yu L, Dong L, Li J, Shao P & Zhang C. authors) and J L (the first author) take full responsibility for the work as a Gestational diabetes mellitus in women of single gravidity in whole, including the study design, the access to data, and the decision to Tianjin City, China. Diabetes Care 2002 25 847–851. (doi:10.2337/ submit and publish the manuscript. diacare.25.5.847) 15 Leng J, Shao P, Zhang C, Tian H, Zhang F, Zhang S, Dong L, Li L, Yu Z, Chan JC et al. Prevalence of gestational diabetes mellitus and its risk European Journal European of Endocrinology Acknowledgments factors in Chinese pregnant women: a prospective population-based The authors thank all doctors, nurses and research staff at 65 primary study in Tianjin, China. PLoS ONE 2015 10 e0121029. (doi:10.1371/ care hospitals, six district-level women’s and children’s health centers, journal.pone.0121029) and Tianjin Women and Children’s Health Center, Tianjin, China, for their 16 Metzger BE, Gabbe SG, Persson B, Buchanan TA, Catalano PA, participation in this study. Damm P, Dyer AR, Leiva A, Hod M, Kitzmiler JL et al. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of in pregnancy. Diabetes Care 2010 33 676–682. (doi:10.2337/dc09-1848) References 17 Hu G, Jousilahti P, Antikainen R, Katzmarzyk PT & Tuomilehto J. Joint effects of physical activity, body mass index, waist circumference, and 1 Tutino GE, Tam WH, Yang X, Chan JC, Lao TT & Ma RC. Diabetes waist-to-hip ratio on the risk of heart failure. Circulation 2010 121 and pregnancy: perspectives from Asia. Diabetic Medicine 2014 31 237–244. (doi:10.1161/CIRCULATIONAHA.109.887893) 302–318. (doi:10.1111/dme.12396) 18 Hu G, Qiao Q, Silventoinen K, Eriksson JG, Jousilahti P, Lindstrom J, 2 International Diabetes Federation. Diabetes Atlas, 6th edn, p 44. Valle TT, Nissinen A & Tuomilehto J. Occupational commuting, and Brussels, Belgium: IDF Executive Office, 2013. (available atwww.idf. leisure-time physical activity in relation to risk for Type 2 diabetes org/diabetesatlas). in middle-aged Finnish men and women. Diabetologia 2003 46 3 Kim C, Newton KM & Knopp RH. Gestational diabetes and the 322–329. incidence of type 2 diabetes: a systematic review. Diabetes Care 2002 19 Sallis JF, Haskell WL, Wood PD, Fortmann SP, Rogers T, Blair SN & 25 1862–1868. (doi:10.2337/diacare.25.10.1862) Paffenbarger RS Jr. Physical activity assessment methodology in the 4 Metzger BE, Lowe LP, Dyer AR, Trimble ER, Chaovarindr U, Five-City Project. American Journal of Epidemiology 1985 121 91–106. Coustan DR, Hadden DR, McCance DR, Hod M, McIntyre HD et al. 20 Chen C & Lu FC. The guidelines for prevention and control Hyperglycemia and adverse pregnancy outcomes. New England Journal of overweight and obesity in Chinese adults. Biomedical and of Medicine 2008 358 1991–2002. (doi:10.1056/NEJMoa0707943) Environmental Sciences 2004 17 (Supplement) 1–36. 5 Silverman BL, Metzger BE, Cho NH & Loeb CA. Impaired glucose 21 Fattah C, Farah N, Barry SC, O’Connor N, Stuart B & Turner MJ. tolerance in adolescent offspring of diabetic mothers. Relationship to Maternal weight and body composition in the first trimester of fetal hyperinsulinism. Diabetes Care 1995 18 611–617. (doi:10.2337/ pregnancy. Acta Obstetricia et Gynecologica Scandinavica 2010 89 diacare.18.5.611) 952–955. (doi:10.3109/00016341003801706)

www.eje-online.org

Downloaded from Bioscientifica.com at 10/01/2021 05:18:44AM via free access European Journal of Endocrinology 29 28 27 26 25 24 23 22

Clinical Study

Oken E, Ning Y, Rifas-Shiman SL,Radesky JS,Rich-Edwards JW& Hu FB, Leitzmann MF, Stampfer MJ,Colditz GA,Willett WC & lifestyleandriskofobesitytype2diabetes. Hu FB. Sedentary Koivusalo SB, Rono K,Klemetti MM,Roine RP, Lindstrom J, Poston L, Bell R,Croker H,Flynn AC,Godfrey KM,Goff L, Dodd JM, Turnbull D, McPhee AJ,Deussen AR,Grivell RM, Alberti KG &Zimmet PZ.Definition,diagnosisandclassification Ludbrook J. Multipleinferencesusingconfidenceintervals. 2006 and duringpregnancywithglucosetolerance. Gillman MW. Associationsofphysicalactivityandinactivitybefore Medicine to riskfortype2diabetesmellitusinmen. Rimm EB. Physicalactivityandtelevisionwatchinginrelation Lipids (doi:10.2337/dc15-0511) (RADIEL): arandomizedcontrolledtrial. TheFinnishGestationalDiabetesPreventionStudy intervention: et al Erkkola M, Kaaja RJ,Poyhonen-Alho M,Tiitinen A, Huvinen E 8587(15)00227-2) Diabetes &Endocrinology UPBEAT study):amulticentre,randomisedcontrolledtrial. inobesepregnantwomen(theof abehaviouralintervention Hayes L, Khazaezadeh N,Nelson SM,Oteng-Ntim E g1285) LIMIT randomisedtrial. orobese: Antenatal lifestyleadviceforwomenwhoareoverweight Yelland LN, Crowther CA,Wittert G, Owens JA&Robinson JS. consultation. classification ofdiabetesmellitusprovisionalreportaWHO of diabetesmellitusanditscomplications.Part1:diagnosis (doi:10.1046/j.1440-1681.2000.03223.x) and Physiology and ExperimentalPharmacology . Gestationaldiabetesmellituscanbepreventedbylifestyle 1081200–1207. 2003 2001 38 Diabetic Medicine 103–108. 161 1542–1548. BMJ (doi:10.1007/s11745-003-1038-4) 2015 2014 1998 3 767–777. (doi:10.1001/archinte.161.12.1542) 348 15 J Lengandothers g1285. 539–553. Diabetes Care (doi:10.1016/S2213- 2000 Archives ofInternal Obstetrics &Gynecology (doi:10.1136/bmj. 27 212–215. 2016 et al . Effect Clinical 39 Lancet 24–30.

32 31 30 35 34 33 37 36 Accepted 10March2016 Revised versionreceived16February2016 Received 9November2015 gestational diabetes Physical activityand

Metzger BE, Buchanan TA, Coustan DR,deLeiva A,Dunger DB, Zhang C, Solomon CG,Manson JE&Hu FB.Aprospectivestudy van derPloeg HP, vanPoppel MN,Chey T, Bauman AE&Brown WJ. Falcao-Pires I, Castro-Chaves P, Miranda-Silva D,Lourenco AP&Leite- Coughlan MT, Vervaart PP, Permezel M,Georgiou HM& Gradmark A, Pomeroy J,Renstrom F, Steiginga S,Persson M, Chasan-Taber L, Fortner RT, Gollenberg A,Buonnaccorsi J,Dole N& Golbidi S &Laher I.Potentialmechanismsofexercise ingestational Hadden DR, Hod M,Kitzmiller JL,Kjos SL,Oats JN Medicine to theriskforgestationaldiabetesmellitus. behaviorsinrelationof pregravidphysicalactivityandsedentary Medicine inSport the developmentofgestationaldiabetesmellitus. behaviourin The roleofpre-pregnancyphysicalactivityandsedentary drudis.2012.04.007) of adipokines. Moreira AF. Physiological, pathological and potential therapeutic roles 4004(03)00183-8) diabetes mellitus. Rice GE. Alteredplacentaloxidativestressstatusingestational secretion inpregnantandnon-pregnantwomen. activity, behaviors,andestimatedinsulinsensitivity sedentary Wright A, Bluck L,Domellof M,Kahn SE,Mogren I (Supplement 2)S251–260. Conference onGestationalDiabetesMellitus. and recommendationsoftheFifthInternationalWorkshop- (doi:10.1089/jwh.2009.1416) characteristics. for gestationaldiabetesamongHispanicwomen:designandbaseline Markenson G. Aprospectivecohortstudyofmodifiableriskfactors (doi:10.1155/2013/285948) diabetes. 2006 Journal ofNutritionandMetabolism Journal 2011 Drug Discovery Today Discovery Drug 166 Journal ofWomen’sJournal Health(Larchmt) 11 2011 Placenta 44. 543–548. 14 (doi:10.1186/1471-2393-11-44) 149–152. 2004 (doi:10.2337/dc07-s225) Downloaded fromBioscientifica.com at10/01/202105:18:44AM (doi:10.1001/archinte.166.5.543) 25 (doi:10.1016/j.jsams.2010.09.002) 2012 78–84. 17 2013 (doi:10.1016/S0143- 880–889. 174 Archives ofInternal Diabetes Care www.eje-online.org :6 Journal of Science and ofScienceand Journal 2013 BMC Pregnancy 2010 et al et al (doi:10.1016/j. 285948. . Summary . Summary . Physical 19 2007 117–124. 773

30 via freeaccess