Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Associations between maternal physical activity in early and late pregnancy and offspring birth size: remote federated individual level meta‐analysis from eight cohort studies

Associations between maternal physical activity in early and late pregnancy and offspring birth... DOI: 10.1111/1471-0528.15476 Epidemiology www.bjog.org Associations between maternal physical activity in early and late pregnancy and offspring birth size: remote federated individual level meta-analysis from eight cohort studies a a b c d e,f g € € S Pastorino, T Bishop, SR Crozier, C Granstrom, K Kordas, LK Kupers, EC O’Brien, h i j,k,l m j m b,n K Polanska, KA Sauder, MH Zafarmand, RC Wilson, C Agyemang, PR Burton, C Cooper, e o h b,n g c j a E Corpeleijn, D Dabelea, WHanke, HM Inskip, FM McAuliffe, SF Olsen, TG Vrijkotte, SBrage, p,q n a a a r, a, A Kennedy, DO’Gorman, P Scherer, K Wijndaele, NJ Wareham, G Desoye, * KK Ong * a b MRC Epidemiology Unit, University of Cambridge, Cambridge, UK MRC Lifecourse Epidemiology Unit (University of Southampton), Southampton General Hospital, Southampton, UK Department of Epidemiology Research, Centre for Fetal Programming, State Serum Institute, Copenhagen, Denmark Epidemiology and Environmental Health, School of Public Health and Health Professions, University at Buffalo, Buffalo, e f NY, USA Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands MRC Integrative Epidemiology Unit, School of Social and Community Medicine, University of Bristol, Bristol, UK Obstetrics & Gynaecology, UCD Perinatal Research Centre, School of Medicine, University College Dublin, National Maternity Hospital, Dublin, Ireland Department of Environmental Epidemiology, Nofer Institute of Occupational Medicine, Lodz, Poland Department of Pediatrics, University of Colorado School of Medicine, Aurora, CO, USA Department of Public Health, Amsterdam Public Health Research Institute, Amsterdam UMC, University of Amsterdam, Amsterdam, the Netherlands Department of Obstetrics & Gynaecology, Amsterdam UMC, University of Amsterdam, Amsterdam, the Netherlands Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Amsterdam Public Health Research Institute, Amsterdam m n UMC, University of Amsterdam, the Netherlands Institute of Health and Society, Newcastle University, Newcastle, UK NIHR Southampton Biomedical Research Centre, University Hospital Southampton NHS Foundation Trust and University of Southampton, Southampton, UK o p Department of Epidemiology, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Denver, CO, USA 3U Diabetes Consortium and School of Health and Human Performance, Dublin City University, Dublin, Ireland School of Biological Sciences, Dublin Institute of Technology, Dublin, Ireland Department of Obstetrics & Gynaecology, Medical University of Graz, Graz, Austria Correspondence: S Pastorino, MRC Epidemiology Unit, Level 3 Institute of Metabolic Science, University of Cambridge School of Clinical Medicine, Cambridge CB2 0SL, UK. Email: silvia.pastorino@mrc-epid.cam.ac.uk Accepted 10 September 2018. Published Online 16 October 2018. Objective Evidence on the impact of leisure time physical activity expenditure were modestly inversely associated with BW, LGA, macrosomia, and ponderal index, without heterogeneity (all: (LTPA) in pregnancy on birth size is inconsistent. We aimed to I = 0%). For each extra hour/week of MVPA, RR for LGA and examine the association between LTPA during early and late macrosomia were 0.97 (95% CI: 0.96, 0.98) and 0.96 (95% CI: 0.94, pregnancy and newborn anthropometric outcomes. 0.98), respectively. Associations were only modestly reduced after Design Individual level meta-analysis, which reduces heterogeneity additional adjustments for maternal BMI and gestational diabetes. across studies. No measure of LTPA was associated with risk for SGA. Setting A consortium of eight population-based studies (seven European and one US) comprising 72 694 participants. Conclusions Physical activity in late, but not early, pregnancy is consistently associated with modestly lower risk of LGA and Methods Generalised linear models with consistent inclusion of macrosomia, but not SGA. confounders (gestational age, sex, parity, maternal age, education, ethnicity, BMI, smoking, and alcohol intake) were used to test Keywords Birth weight, large-for-gestational age, macrosomia, associations between self-reported LTPA at either early (8– physical activity, pregnancy, small-for-gestational age. 18 weeks gestation) or late pregnancy (30+ weeks) and the Tweetable abstract In an individual participant meta-analysis, late outcomes. Results were pooled using random effects meta-analyses. pregnancy moderate to vigorous physical activity modestly Main outcome measures Birth weight, large-for-gestational age reduced birth size outcomes. (LGA), macrosomia, small-for-gestational age (SGA), % body fat, and ponderal index at birth. Linked article This article is commented on by CA Venetis, p.471 in this issue. To view this mini commentary visit https://doi.org/10.1111/ Results Late, but not early, gestation maternal moderate to vigorous 1471-0528.15542. physical activity (MVPA), vigorous activity, and LTPA energy *Equal contributions. ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 459 Royal College of Obstetricians and Gynaecologists. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. [The copyright line for this article was changed on 22 August 2019 after original online publication.] Pastorino et al. Please cite this paper as: Pastorino S, Bishop T, Crozier SR, Granstrom € C, Kordas K, Kupers € LK, O’Brien EC, Polanska K, Sauder KA, Zafarmand MH, Wilson B, Agyemang C, Burton PR, Cooper C, Corpeleijn E, Dabelea D, Hanke W, Inskip HM, McAuliffe FM, Olsen SF, Vrijkotte TG, Brage S, Kennedy A, O’Gorman D, Scherer P, Wijndaele K, Wareham NJ, Desoye G, Ong KK. Associations between maternal physical activity in early and late pregnancy and offspring birth size: remote federated individual level meta-analysis from eight cohort studies. BJOG 2019;126:459–470. meta-analysis to be performed remotely. Compared with a Introduction literature-based meta-analysis, this approach can reduce The prevalence of childhood obesity has increased world- heterogeneity between studies by allowing harmonisation of wide over the last three decades. Babies born with large- exposure and outcome variables, and by allowing the same for-gestational age (LGA), or with macrosomia [birth models to be tested in each study. weight (BW) above 4000 or 4500 g], have higher risks of obesity and raised metabolic disease markers in childhood Methods 2,3 compared with babies with appropriate BW. Physical activity during pregnancy is recommended to enhance the InterConnect is an EU-FP7 funded project that optimises health of the mother-to-be, but has also been explored as the use of existing data by enabling cross-cohort analyses a potential intervention to lower the risk for LGA and within consortia without pooling of individual-level data at 5–10 macrosomia. Physical activity might be especially a central location. For this research question, eight cohorts appealing if it reduced high BW without increasing the risk with data on physical activity in pregnancy and neonatal of small-for-gestational age (SGA) babies. Physical activity outcomes set up a server to allow remote federated analyses during pregnancy might reduce fetal growth by increasing and joined the consortium. The collaborative group com- insulin sensitivity and by modulating glucose regula- prised the following prospective birth cohort studies: the 11,12 tion. Physical activity might also regulate fetoplacental Avon Longitudinal Study of Parents and Children 32,33 growth by altering the rates of oxygen and nutrient sup- (ALSPAC, UK), the Amsterdam Born Children and 13 34 ply. their Development study (ABCD, the Netherlands), the Recent systematic reviews of randomised controlled trials Danish National Birth Cohort (DNBC, Denmark), the on the effect of maternal exercise on birth outcomes report Groningen Expert Center for Kids with Obesity (GECKO)- 14,15 36 modest BW reductions (10–30 g). However, they report Drenthe (the Netherlands), the Healthy Start Study (HSS, wide variation in the types of interventions studied in USA), the Polish Mother and Child Cohort (REPRO_PL, 37 38 terms of form, intensity, and volume of exercise. Systematic Poland), the ROLO study (Ireland), and the Southamp- reviews of observational studies on the association between ton Women’s Survey (SWS, UK). Characteristics of the maternal physical activity during pregnancy with birth participating studies are shown in Table S1. Each partici- 16,17 size report conflicting results: some studies report an pating cohort obtained ethical approval from the corre- 5–10,18,19 20–22 inverse association, some a positive association, sponding local ethics committee (see details at the end). 23–28 and others no significant association. There is also No PPI took place for these analyses. some evidence that the timing of physical activity in preg- We included all live-born singleton full-term births and 18,29 nancy might be important. The heterogeneity among excluded mothers with pre-eclampsia and those with miss- studies limits the ability to pool published results. One ing information for any of the covariates. The percentage meta-analysis reports that ‘high’ physical activity levels of participants with any missing values across cohorts ran- were inversely associated with BW, but conversely ‘moder- ged between 10.2% and 34% for early pregnancy analyses, ate’ physical activity levels were positively associated with and between 12.7% and 43.5% for late pregnancy analyses. BW. The included studies use different definitions of physi- Funding for this study was received from the European cal activity level and there is no standardisation with regard Union Seventh Framework Programme (FP7/2007–2013) to the type and domains of activity or the volume, inten- under grant agreement no. 602068. Core Outcome Set sity, and timing. Most studies did not adjust for any con- (COS), and patient involvement (PPI) is not relevant to founder. this study and hence is not described here. Here, we examined the association between leisure time physical activity (LTPA) during pregnancy and newborn Physical activity during pregnancy anthropometric outcomes across a range of prospective All studies assessed physical activity during pregnancy by cohort studies. Within a consortium created as part of the questionnaire. HSS and SWS used interviewer-administered InterConnect project, we used a federated meta-analysis questionnaires, DNBC used a computer-assisted telephone approach, which allows an individual participant-level interview, and the other studies used self-administered 460 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes questionnaires. Table S2 details the questions in each in early pregnancy up to week 18 of gestation. We applied cohort. We harmonised self-reported data on LTPA during a uniform correction factor to weights measured later than early pregnancy across seven cohorts, and LTPA during late 12 weeks gestation derived by weight gain curves based on pregnancy across five cohorts. The median gestational age repeated maternal weight measures in the ALSPAC study. at which mothers replied to questionnaires was 8–18 weeks There was wide variation in definitions of ethnicity across for early pregnancy, and 30 weeks to 1 day post-delivery cohorts; the ‘other ethnicity’ category includes a variety of for late pregnancy. LTPA was chosen as it is the domain Asian, Hispanic, and other ethnic groups. GDM was most amenable to intervention and therefore more relevant defined using biochemical data at weeks 24–28 in HSS and for public health recommendations; it was also the most ROLO, and by a combination of medical records and self- commonly assessed domain across the eight studies. Inten- reports in the other studies. sity of reported activities was expressed in metabolic equiv- alent of energy expenditure (MET) values according to the Potential confounders and other covariates Compendium of Physical Activity. Four exposure vari- Potential confounders were not harmonised because, in ables were harmonised: (1) duration of LTPA (hours/ federated analysis models involving random-effects meta- week), which included any reported leisure time activity; analysis of the arising study-specific estimates, this would (2) duration of moderate-vigorous LTPA (MVPA) (hours/ not impact the summary effect estimates and P-values. week) including activities with intensity ≥3 MET; (3) dura- However, confounder variables were reasonably comparable tion of vigorous LTPA (VPA) (hours/week) including across studies. Smoking in pregnancy was a dichotomous activities with intensity ≥6 MET; (4) energy expenditure of variable (yes/no) in all studies except DNBC, which deter- LTPA (MET-hours/week) calculated by multiplying dura- mined the number of cigarettes/week. Alcohol intake was tion of LTPA by MET values. Three studies recorded cate- considered as units of alcohol/week in ALSPAC, DNBC, gorical response formats for duration of LTPA (ALSPAC, and SWS; glasses/week in ABCD; and as categorical vari- GECKO, and SWS). These were converted into numerical ables in GECKO (none, <1 glass/week, 1–6 glasses/week, values, where relevant using the mid-point of the stated 7+ glasses/week), HSS (none, once per month or less, twice range (i.e. ‘>7 hours/week’ was converted to 7 hours/week; per month or more), REPRO_PL (yes/no), and ROLO ‘2–6’ to 4; <1 to 0.5; ‘never’ to 0). (yes/no). Educational attainment was considered as a cate- gorical variable in most cohorts (range 2–6 levels) except Outcomes ABCD, which recorded ‘years of education after elementary The following outcome variables were harmonised across school’. Parity (number of previous live births) was self- all studies, based on objective measurements in all studies: reported in all studies, and maternal age at delivery was BW (g), macrosomia (defined as BW >4000 g), LGA (BW calculated from mother’s date of birth and delivery date. for gestational age >90th percentile according to the INTERGROWTH-21st Project ), and SGA (BW for gesta- Statistical analyses tional age <10th percentile according to INTERGROWTH- All analyses were conducted using R within the Data- 21st). Ponderal Index, a measure of leanness (corpulence) SHIELD federated meta-analysis library. In this process, 3 3 [weight/length (kg/m )] at birth was harmonised for six individual participant data from contributing studies are cohorts. Percent (%) body fat at birth was available for held securely on servers at each study location. A com- three cohorts. Of these, one (HSS) assessed newborn body puter within the network sends analytical commands that fat using air displacement plethysmography (PEAPOD), request each local server to undertake an analysis locally while skinfold thickness measurements were available in and return non-identifiable summary statistics. The result HSS, SWS, and in a subset of ROLO (n = 219). Triceps of this process is mathematically equivalent to an individ- and subscapular skinfolds were used to estimate % body fat ual participant meta-analysis with the advantage that data using the algorithm reported by Slaughter et al.: % body remain within the governance structure of each single fat = 1.21 9 (triceps skinfold + subscapular skinfold)  cohort study. 0.008 9 (triceps skinfold + subscapular skinfold)  1.7. To analyse data, we used generalised linear models in each study. Each model was fitted in a federated manner Potential modifiers using the iterative reweighted least squares process. The The following potential modifying variables were harmo- primary models included MVPA duration as exposure and nised across the studies: infant sex, maternal obesity (BMI: each outcome (BW, macrosomia, LGA, SGA, ponderal 2 2 2 ≤20 kg/m , >20–30 kg/m , >30 kg/m ), maternal ethnicity index, % body fat) separately. Moderate to vigorous activ- (white, black, other), and gestational diabetes mellitus (GD: ity was chosen as the primary exposure because it has yes, no). Maternal weight was objectively measured in five higher validity than lower intensity activities; also, the cohorts and self-reported in three cohorts at varying times majority of existing guidelines recommend moderate ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 461 Royal College of Obstetricians and Gynaecologists Pastorino et al. intensity physical activity for pregnant women. The cohorts, macrosomia prevalence ranged between 5.6% in adjusted models included each exposure separately (LTPA HSS and 21.7% in DNBC, and LGA between 8.7% in HSS duration, MVPA duration, VPA duration, LTPA energy and 30.2% in GECKO. SGA prevalence ranged between expenditure) with each outcome (BW, macrosomia, LGA, 0.8% in ROLO and 9.4% in HSS. Median ponderal index SGA, ponderal index, % body fat), and were adjusted for at birth ranged between 20.2 in REPRO_PL and 27.8 in gestational age (except for LGA and SGA), infant sex, par- SWS, and body fat was 10, 11, and 16% in HSS, SWS, and ity, maternal age, smoking, alcohol, maternal education, ROLO, respectively. and ethnicity. Further models were additionally adjusted Reported levels of maternal LTPA during pregnancy var- for maternal early pregnancy BMI. A schematic diagram ied across studies, with DNBC women having the lowest of the analysis plan is shown in Figure S1. All covariates levels in both periods (64% of women reporting no LTPA). were chosen a priori based on literature evidence. To Among the other cohorts, median LTPA duration ranged explore which covariate contributed most to heterogeneity, from 2.0 to 6.5 hours/week for early pregnancy, and 1– we conducted further analyses by including each potential 7 hours/week for late pregnancy. Median MVPA levels ran- confounding variable one at a time. Physical activity is ged from 0 to 4 hours/week for early pregnancy, and 0– likely to exert its effect on birth size by altering maternal 0.8 hours/week for late pregnancy. The proportion of metabolic pathways such as glucose metabolism, and there women reporting any MVPA decreased from the early is evidence of its association with GDM. Therefore, pregnancy in the four studies with data at both time points GDM was added in a subsequent model to explore its (DNBC, 34%; HSS, 72%; REPRO_PL, 20%; SWS, 84%) to possible mediating effect. We explored the possible modi- late pregnancy (DNBC, 25%; HSS, 49%; REPRO_PL, 12%; fying effect of infant sex, maternal obesity, maternal eth- SWS, 78%). The proportion of women reporting any VPA nicity, and GDM by including interaction terms in the was low in most cohorts (range: 6.6–42.5%) and decreased model. These potential effect modifiers were chosen a pri- in late pregnancy (range: 2.9–24.1%). ori. The levels of physical activity and their effects on health differ across ethnic groups. In pregnant women, Physical activity associations in early pregnancy both obesity and GDM might alter physiological charac- Early pregnancy maternal LTPA was not associated with teristics that affect their ability to exercise. All models any measure of offspring birth size (Tables 2, S3, and S4). were conducted separately for early and late pregnancy Heterogeneity across studies was high in unadjusted models physical activity. Early pregnancy physical activity mea- (I = 79–86% for BW, macrosomia, and LGA, Table S1), sures were available for ALSPAC, ABCDS, DNBC, HSS, but was substantially reduced after adjustments for poten- REPRO-PL, ROLO, and SWS. Late pregnancy physical tial confounders (0–54%, Table 2). In sensitivity models, activity measures were available for DNBC, GECKO, HSS, with stepwise inclusion of covariates, ethnicity and mater- REPRO_PL, and SWS. Regression analyses were conducted nal education contributed the most to (positive) confound- for each individual study, and then random-effects meta- ing in some individual studies, with non-white ethnicity analysis was used to combine the effect estimates. A ran- being associated with both lower BW and lower LTPA, and dom effects approach was chosen owing to the reported maternal education being associated with both higher BW heterogeneity between other published studies. Hetero- and higher LTPA (not shown). geneity was assessed using the I statistic. Physical activity associations in late pregnancy Late pregnancy maternal MVPA (Figures 1 and 2, Table 2), Results VPA, and LTPA energy expenditure (Tables 2 andS3) were For early pregnancy physical activity analyses, 72 694 par- inversely associated with all birth size outcomes (except for ticipants from seven studies were included (57 807 across % body fat and SGA) in adjusted models. For each +1 hour/ six studies for ponderal index; 3039 in three studies for % week of MVPA, offspring BW was lower by 6.4 g (95% CI: body fat). For late pregnancy analyses, the available sample 9.1, 3.7; P <0.001) and ponderal index by 0.02 kg/m (95% was 58 820 from five studies (57 172 across four studies CI: 0.03, 0.00; P = 0.02); the relative risks of macrosomia for ponderal index; 2792 in two studies for % body fat). and LGA were lower by 4% (95% CI: 2, 6; P <0.001) and 3% Maternal and infant characteristics are presented in (95% CI: 2, 4; P <0.01), respectively. No association was Table 1. Mean BW ranged between 3356 and 4135 g for found for SGA (OR: 0.99, 95% CI: 0.98, 1.00) and % body male infants, and between 3217 and 3963 g for female fat (0.01, 95% CI: 0.04, 0.02). VPA showed larger associ- infants. ROLO infants had the highest mean BW and high- ations with BW (22 g/hour/week; 95% CI: 31.3, 12.7; est prevalence of macrosomia (51.8%) and LGA (61.7%), P <0.001), ponderal index (0.07 units; 95% CI: 0.13, reflecting their inclusion of only secundigravid women 0.02; P <0.01), macrosomia (lower by 11%, 95% CI: 5, 16; whose first baby had been macrosomic. Among the other P <0.01) and LGA (lower by 11%, 95% CI: 5, 16 P <0.001), 462 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 463 Royal College of Obstetricians and Gynaecologists Table 1. Study population characteristics ALSPAC ABCD DNBC GECKO HSS REPRO_PL ROLO SWS n* 9058 6464 53 671 1335 1054 982 617 1902 Female, n (%) 4428 (48.9) 3244 (50.2) 26 371 (49.2) 638 (41.8) 509 (48.3) 487 (49.6) 309 (50.1) 916 (48.2) Birth weight, g Male, mean (SD) 3551 (479) 3572 (491) 3709 (503) 3708 (505) 3356 (432) 3490 (440) 4135 (481) 3589 (480) Female, mean (SD) 3424 (447) 3435 (456) 3575 (481) 3538 (490) 3217 (420) 3316 (432) 3963 (423) 3445 (458) Macrosomia, n (%) 1158 (12.7) 871 (13.4) 11 681 (21.7) 289 (21.6) 60 (5.6) 84 (8.5) 320 (51.8) 267 (14) LGA, n (%) 1888 (20.8) 1222 (18.9) 15 052 (28) 405 (30.3) 121 (8.7) 183 (18.6) 381 (61.7) 369 (19.4) SGA, n (%) 418 (4.6) 311 (4.8) 1849 (3.4) 59 (6.4) 100 (9.4) 58 (5.9) 5 (0.8) 101 (5.3) Ponderal index**, median (IQR) 26.2 (24.7–27.8) 24.9 (23.5–26.5) 26.9 (24.9–29.2) 20.2 (18.9- 21.6) 27.1 (25.3–29.3) 27.8 (26.3–29.2) % Body fat***, median (IQR) 10 (8–12) 16 (14–18) 11 (10–13) Early pregnancy physical activity, median (IQR) LTPA (hours/week) 4.0 (0.5–5.5) 2.0 (0.5–4.3) 0.0 (0.0–1.0) 3.0 (1.0–5.8) 4.0 (0.0–7.0) 1.7 (1.0–2.3) 6.5 (3.2–11.5) MVPA (hours/week) 4.0 (0.5–5.0) 1.5 (0.0–3.5) 0.0 (0.0–1.0) 1.5 (0.0–3.5) 0.0 (0.0–0.0) 0.3 (0.0–1.0) 1.2 (0.3–3.0) LTPA EE (Met-hours/week) 15.2 (3.0–25.2) 8.1 (1.7–19.3) 0.0 (0.0–6.0) 10.2 (3.1–23.6) 16.5 (0.0–33.0) 4.5 (2.0–7.8) 17.5 (8.7–32.1) Women doing vigorous PA, n (%) 604 (6.6) 1876 (29) 4321 (8.0) 244 (23.1) 84 (8.5) 61 (9.8) 810 (42.5) Late pregnancy physical activity, median (IQR) LTPA (hours/week) 0.0 (0.0–1.0) 1.0 (1.0–1.0) 2.0 (0.5–3.6) 5.0 (0.0–8.0) 7.0 (3.4–12.0) MVPA (hours/week) 0.0 (0.0–1.0) 0.3 (0.0–1.0) 0.0 (0.0–1.5) 0.0 (0.0–0.0) 0.8 (0.1–2.3) Women doing vigorous PA, n (%) 1599 (2.9) 61 (5.7) 77 (8.3) 443 (24.1) LTPA EE (Met-hours/week) 0.0 (0.0–3.0) 1.0 (0.0–4.0) 6.3 (1.5–11.9) 19.8 (0.0–33.0) 16.7 (8.5–31.1) Maternal age (y), mean (SD) 28.5 (4.7) 30.9 (5.1) 30.1 (4.2) 30.8 (4.2) 28 (6.1) 29.0 (4.2) 32.2 (4.1) 30.6 (3.7) Maternal BMI Mean (SD) 22.5 (4.3) 24.0 (4.1) 24.7 (4.1) 24.7 (4.7) 26.7 (6.0) 22.8 (3.6) 26.6 (4.8) 26 (4.8) Overweight, n (%) 1257 (13.8) 1447 (22.3) 14 896 (27.7) 320 (24) 334 (31.6) 162 (16.4) 233 (37.7) 606 (32) Obese, n (%) 586 (6.4) 527 (8.1) 5546 (10.3) 169 (12.6) 225 (21.3) 39 (3.9) 114 (18.4) 323 (17) GDM, n (%) 41 (0.4) 76 (1.1) 380 (10.7) 44 (3.2) 43 (4) 34 (3.4) 12 (2) 19 (1) Ethnicity White 8867 (98) 4490 (69.4) 53 671 (100) 1321 (99) 814 (76.5) 982 (100) 612 (97.5) 1840 (96.8) Black 77 (0.8) 486 (7.6) 0 (0) 0 (0) 162 (15.3) 0 (0) 2 (0.3) 10 (0.5) Other 114 (1.2) 1488 (23) 0 (0) 14 (1) 78 (7.2) 0 (0) 14 (2.2) 52 (2.7) EE, energy expenditure; GDM, gestational diabetes mellitus; LGA, large for gestational age; LTPA, leisure-time physical activity; MVPA, moderate to vigorous leisure time physical activity; SGA, small for gestational age. *Sample size available for late pregnancy physical activity analyses were: DNBC = 53 684, HSS = 1044, REPRO_PL = 919, SWS = 1838. **Sample size available for analyses of Ponderal index for early pregnancy analyses were: ALSPAC = 7118, DNBC = 53 487, HSS = 976, REPRO_PL = 977, ROLO = 523, SWS = 1844; for late pregnancy analyses: DNBC = 53 500, HSS = 968, REPRO_PL = 915, SWS = 1789. ***Sample size available for analyses of % body fat for early pregnancy analyses were: HSS = 988, ROLO = 189, SWS = 1862; for late pregnancy analyses: HSS = 987, SWS = 1805. Pastorino et al. Table 2. Associations between physical activity during pregnancy and offspring birth size BW (grams) Macrosomia LGA Ponderal index SGA RR, 95% CI RR, 95% CI RR, 95% CI Beta, 95% CI Beta, 95% CI 2 2 2 2 2 I I I I I Physical activity Early pregnancy LTPA (hours/week) 0.86 (2.33, 0.61) 0.99 (0.98, 1,01) 0.99 (0.98, 1,00) 0.0 (0.01, 0.01) 0.99 (0.98, 1.01) 23% 51% 46% 0% 0% MVPA (hours/week) 1.38 (3.77, 1.01) 1.00 (0.98, 1,01) 1.00 (0.98, 1,01) 0.00 (0.01, 0.01) 0.99 (0.98, 1.00) 41% 52% 43% 0% 0% VPA (hours/week) 1.38 (3.77, 1.01) 1.00 (0.98, 1,01) 1.00 (0.98, 1,01) 0.00 (0.05, 0.04) 0.99 (0.98, 1.00) 41% 52% 43% 18% 0% LTPAEE (met-hours/week) 0.14 (0.58, 0.30) 1.00 (0.99, 1,00) 0.99 (0.99, 1,00) 0.00 (0.00, 0.00) 0.99 (0.99, 1.00) 49% 53% 38%) 0% 0% Physical activity Late pregnancy LTPA (hours/week) 2.22 (5.54, 1.0) 0.98 (0.96, 1.00) 0.98 (0.97, 0.99) 0.01 (0.02, 0.00) 0.99 (0.97, 1.01) 64% 37% 0% 13% 0% MVPA (hours/week) 6.43 (9.12, 3.74) 0.96 (0.94, 0.98) 0.97 (0.96, 0.98) 0.02 (0.03, 0.00) 1.01 (0.97, 1.03) 0% 0% 0% 0% 0% VPA (hours/week) 22.0 (31.3, 12.7) 0.89 (0.84, 0.95) 0.89 (0.84, 0.94) 0.07 (0.13, 0.02) 1.06 (0.96, 1.17) 0% 0% 0% 0% 0% LTPAEE (met-hours/week) 0.93 (1.43, 0.42) 0.99 (0.99, 0.99) 0.99 (0.99, 0.99) 0.00 (0.01, 0.00) 0.99 (0.99, 1.00) 9% 0% 0% 0% 0% EE, energy expenditure; LGA, large for gestational age; LTPA, leisure time physical activity; MVPA, moderate to vigorous leisure time activity; SGA, small for gestational age; VPA, vigorous leisure time activity. Models are adjusted for gestational age, sex, parity, maternal age, smoking, alcohol, maternal education, and ethnicity. Statistically significant associations are highlighted in bold. and no association with % body fat (0.05; 95% CI: 0.17, with the inconsistent findings of published literature-based 0.06) and SGA (OR: 1.01, 95% CI: 0.96, 1.16). The associa- systematic reviews, heterogeneity between study estimates tions with late pregnancy LTPA were not mediated by GDM was substantially reduced by consistent confounding adjust- and persisted after further adjustment for early pregnancy ment and by harmonisation of exposures and outcomes. maternal BMI (Table S5). The remote federated analysis approach avoided the need No interaction with ethnicity, infant sex, GDM, or to physically pool individual-level data, and hence substan- maternal obesity was found in either pregnancy period for tially reduced the governance burdens and associated time LTPA and birth size (all P-values for interactions >0.05). delays, and avoided barriers due to limitations of consent and research ethics permissions. Another strength is that we were able to analyse the differential association of tim- Discussion ing and intensity of physical activity in pregnancy with off- Main findings spring birth size outcomes. In this large cross-cohort analysis of up to 72 694 individuals, However, there were some limitations in our approach. we found small but consistent inverse associations between Physical activity was self-reported in all included studies, maternal LTPA during late, but not early, pregnancy and off- and only a few of the questionnaires were validated. Phys- spring birth size. Each additional hour/week of MVPA in late ical activity questionnaires are susceptible to measurement pregnancy was associated with 6.4 g lower birth weight and error related to both recall and social desirability with 4% and 3% relative reductions in risk of macrosomia and validity estimated between 0.25 and 0.4. However, they LGA, respectively, without increasing the risk of SGA. are able to rank individuals according to activity levels. Furthermore, validity is higher among women than men Strengths and limitations and for vigorous intensity compared with lighter intensity A major strength of our approach was the planned individ- activities. It remains a challenge to identify thresholds of ual level analysis across several cohort studies. Compared physical activity in terms of health benefits. Contributing 464 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes Study (weight) Birth weight Beta (95% CI) Grams Study (weight) Ponderal Index Beta (95% CI) kg/m Figure 1. Forest plots for late pregnancy moderate to vigorous activity (hours/week) associated with birth weight and ponderal index. Associations were adjusted for gestational age, sex, parity, maternal age, smoking, alcohol, maternal education, and ethnicity. n = 58 820 except for ponderal index (n = 57 172). studies used different questionnaires with varying ways of analyses including such outcomes would be highly infor- assessing LTPA, which made harmonisation challenging. mative. Our use of international INTERGROWTH-21st For example, some listed specific activities (e.g. ‘swim- Project data to define LGA and SGA led to unequal num- ming’, ‘walking’) while others asked only about categories bers for those outcomes and limited the statistical power of activities (i.e. ‘moderate, ‘vigorous’), which included to detect a possible association between VPA and SGA. some activities outside of leisure time. Intensity informa- Although we adjusted for many confounding factors, tion was not available in all questionnaires, which meant residual confounding cannot be ruled out. Limited geo- assumptions had to be made when assigning MET values. graphical and ethnic diversity restricted the power to Differences in average LTPA levels across the studies detect modifying factors. One participating study (DNBC) might therefore reflect differences in methods or real pop- was substantially larger than the other studies and ulation differences. The timing of questionnaire adminis- accounted for more than 70% of the sample size in the tration differed across studies, particularly for early analyses. Whilst the dominance of this study in driving pregnancy LTPA. Unfortunately, data were unavailable on results should be acknowledged, it is noteworthy that, in clinical outcomes associated with LGA and macrosomia adjusted models, heterogeneity was reduced from >70 to (such as shoulder dystocia, 3rd or 4th degree laceration), 0% in most analyses, thus highlighting the consistency or on pregnancies not resulting in live birth. Future across studies and the generalisability of results. ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 465 Royal College of Obstetricians and Gynaecologists Pastorino et al. Study (weight) Macrosomia RR (95% CI) RR Study (weight) LGA RR (95% CI) RR Figure 2. Forest plots for late pregnancy moderate to vigorous activity (hours/week) associated with relative risk of macrosomia and large for gestational age (LGA). Associations were adjusted for gestational age, sex, parity, maternal age, smoking, alcohol, maternal education, and ethnicity. n = 58 820. Interpretations adjustment for confounding reduced heterogeneity The direction of our associations is consistent with some between individual study estimates from I >70% to 0% 5–10,18,19 previous individual studies; however, other studies in several analyses. Furthermore, adjustment for ethnicity 23–28 20– reported null or even directionally opposite results. and maternal education avoided spurious positive associa- 22 17 tions between early pregnancy physical activity and birth A recent meta-analysis reports that a ‘moderate’ level size. We harmonised the intensity of activities by assign- of physical activity was positively associated with BW, ing the same MET values for similar reported activities while a ‘high’ level of physical activity was inversely asso- across studies. Although the diverse nature of the ques- ciated with BW. However, those results were based on a tionnaires used in the individual studies made harmonisa- mixture of adjusted and unadjusted models, and their tion challenging, MVPA was less heterogeneous than reported meta-analysis of only the adjusted models other activity variables, particularly in late pregnancy; this showed null associations for both moderate and high may be because our harmonised MVPA variable was more levels of physical activity. Furthermore, in that meta-ana- robust to underlying methodological differences across lysis, there was substantial heterogeneity, with I values studies. >80%. We demonstrate here that more consistent 466 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes The timing of PA associations with LTPA during late, Disclosure of interests but not early, pregnancy is also consistent with some None declared. Completed disclosure of interests form 18,29,51 51 reported studies. Clapp et al. reported inverse asso- available to view online as supporting information. ciations with newborn adiposity or BW only for late preg- nancy physical activity. Hopkins and Cutfield conjectured Contribution to authorship that high volume exercise only in the first half of preg- GD, KKO, SP contributed to planning the study. SP, KW, nancy increased BW, but if performed throughout preg- SB, AK, DOG, and KKO coordinated harmonisation of all nancy or only in the second half of pregnancy, it reduced variables. TB and PS conducted the federated remote statisti- BW. They suggested that the timing of physical activity cal analyses. SP, KKO, DOG, and SB interpreted the results. caused different fetoplacental adaptations. SP wrote the article. SP, TB, SB, KW, NJW, GD, KKO, SRC, Regarding intensity of LTPA, we found that late preg- CG, KK, LKK, EOB, KP, KAS, MHZ, BW, CA, PRB, CC, EC, nancy MVPA, VPA, and energy expenditure, but not DD, WH, HMI, FM, SFO, and TGV contributed to the anal- duration of LTPA, were inversely associated with offspring ysis plan, the production of the paper, the harmonisation birth size. Some previous studies have assessed the impact algorithms, and the review of the manuscript. of physical activity intensity on offspring birth size, with 21,22,51 some findings consistent with ours, but others Details of ethics approval 27,52–54 reported null results. Different adjustment factors Avon Longitudinal Study of Parents and Children and different definitions, timing, and categories of physi- (ALSPAC): ethical approval for the study was obtained cal activities might lead to inconsistent findings between from ALSPAC Ethics and Law Committee and the Local studies. Although the proportion of women reporting any Research Ethics Committees. Amsterdam Born Children VPA was small, our results suggest that changes in birth and their Development study (ABCD): approval of the size outcomes are dependent on the intensity of LTPA, study was obtained from the Central Committee on with larger effects observed with higher intensity. It is Research Involving Human Subjects in The Netherlands, possible that LTPA intensity needs to reach a certain the medical ethics review committees of the participating threshold before it has an effect on nutrient supply to the hospitals, and the Registration Committee of the Munici- fetus. Alternatively, higher intensity recreational activities pality of Amsterdam. Danish National Birth Cohort may be easier to recall and less prone to measurement 44 (DNBC, Denmark): approved by the Committee on error. Biomedical Research Ethics under case number (KF) 01- Our observed associations remained significant after 471/94. Groningen Expert Center for Kids with Obesity adjustment for maternal BMI, possibly suggesting that the (GECKO )-Drenthe: approved by the Medical Ethics effect of physical activity on birth size is only partially Committee of the University Medical Center Groningen mediated by maternal weight; however, we did not have (UMCG). Healthy Start Study (HSS): approved by the measures of late pregnancy maternal weight gain and Colorado Multiple Institutional Review Board. Polish BMI. Independent of maternal weight, physical activity Mother and Child Cohort (REPRO_PL): approved by the 12,55 increases maternal insulin sensitivity, reduces maternal Ethical Committee of the Nofer Institute of Occupational glucose, and hence might reduce glucose transfer to the Medicine, Łod  z, Poland (Decision No. 7/2007). ROLO fetus. These metabolic changes are more marked at study: approved by the Ethics Committee at the National higher intensities and volumes of exercise and in late Maternity Hospital, June 2006. Southampton Women’s 11,29 pregnancy. Survey (SWS): approved by South Central—Hampshire B Research Ethics Committee. Conclusion Funding In conclusion, LTPA energy expenditure, MVPA, and VPA InterConnect: the research leading to these results received during late, but not early, pregnancy had a small but signif- funding from the European Union Seventh Framework icant and consistent inverse association with offspring birth Programme (FP7/2007–2013) under grant agreement no. size. Larger effects were observed with higher intensity of 602068. Amsterdam Born Children and their Development physical activity. Compared with the inconsistent findings study (ABCD): the ABCD study was supported by grants of reviews of published reports, this remote federated indi- from the Netherlands Organization for Health Research vidual-level analysis substantially reduced heterogeneity and Development (ZonMW) and The Netherlands Heart between individual studies by allowing consistent adjust- Foundation. Genotyping was funded by the BBMRI-NL ment for confounding and careful harmonisation of expo- grant CP2013-50. M.H. Zafarmand was supported by sures and outcomes. ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 467 Royal College of Obstetricians and Gynaecologists Pastorino et al. BBMRI-NL (CP2013-50). T.G.M. Vrijkotte was supported Table S1. Characteristics of the eight contributing cohort by ZonMW (TOP 40–00812–98–11010). Avon Longitudi- studies. nal Study of Parents and Children (ALSPAC): we are Table S2. Questions asked in the participating cohorts grateful to all the families who took part in this study, the used for harmonisation of leisure time physical activity midwives for their help in recruiting them, and the Avon exposure. Longitudinal Study of Parents and Children (ALSPAC) Table S3. Unadjusted associations between physical team, including interviewers, computer and laboratory activity during pregnancy and offspring birth size. technicians, clerical workers, research scientists, volunteers, Table S4. Associations between physical activity during managers, receptionists, and nurses. The UK Medical pregnancy and % body fat. Research Council and Wellcome (grant ref: 102215/2/13/2) Table S5. Associations between late pregnancy physical and the University of Bristol provide core support for activity and offspring birth size with additional adjustments ALSPAC. Danish National Birth Cohort (DNBC, Den- for maternal early pregnancy BMI and GDM. mark): the Danish National Research Foundation has established the Danish Epidemiology of Science Centre that initiated and created the DNBC. The cohort is fur- References thermore a result of a major grant from this foundation. 1 Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, Additional support for the DNBC is obtained from the et al. Global, regional, and national prevalence of overweight and Pharmacy Foundation, the Egmont Foundation, the March obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet of Dimes Birth Defect Foundation, the Augustinus Founda- 2014;384:766–81. tion, and the Health Foundation. Groningen Expert Center 2 Boney CM, Verma A, Tucker R, Vohr BR. Metabolic syndrome in for Kids with Obesity (GECKO)-Drenthe: the GECKO Dren- childhood: association with birth weight, maternal obesity, and the birth cohort was funded by an unrestricted grant of gestational diabetes mellitus. Pediatrics 2005;115:e290–6. Hutchison Whampoa Ltd, Hong Kong, and supported by 3 Gu s, An X, Fang L, Zhang X, Zhang C, Wang J, et al. Risk factors and long-term health consequences of macrosomia: a prospective the University of Groningen, Well Baby Clinic Foundation study in Jiangsu Province, China. J Biomed Res 2012;26: Icare, Noordlease, and Youth Health Care Drenthe. Healthy 235–40. Start Study (HSS) was funded by the following NIH funding 4 Amercian College of Obstetricians and Gynaecolgist. Committee sources: R01DK076645, UL1TR00108. Polish Mother and opinion no. 650: physical activity and exercise during pregnancy and Child Cohort (REPRO_PL) is supported in part by funds the postpartum period. Obstet Gynecol 2015;126:e135–42. 5 Clapp JF III, Capeless EL. Neonatal morphometrics after endurance from National Centre for Research and Development, exercise during pregnancy. Am J Obstet Gynecol 1990;1:1805–11. Poland (grant no. PBZ-MEiN-/8/2/2006; contract no. K140/ 6 Morgan KL, Rahman MA, Hill RA, Zhou SM, Bijsma G, Khanom A, P01/2007/1.3.1.1) and grant PNRF-218-AI-1/07 from Nor- et al. Physical activity and excess weight in pregnancy have way through the Norwegian Financial Mechanism within the independent and unique effects on delivery and perinatal outcomes. Polish–Norwegian Research Fund. ROLO study was funded PLoS One 2014;9:e94532. 7 Juhl M, Olsen J, Andersen PK, Nøhr EA, Andersen AM. Physical by the Health Research Board of Ireland, with additional exercise during pregnancy and fetal growth measures: a study financial support from the National Maternity Hospital within the Danish National Birth Cohort. Am J Obstet Gynecol Medical Fund. Southampton Women’s Survey (SWS) was 2010;202:63.e1–8. supported by grants from the Medical Research Council, 8 Owe KM, Nystad W, Bø K. Association between regular exercise National Institute for Health Research Southampton and excessive newborn birth weight. Obstet Gynecol 2009;114:770–6. Biomedical Research Centre, University of Southampton 9 Currie LM, Woolcott CG, Fell DB, Armson BA, Dodds L. The and University Hospital Southampton National Health association between physical activity and maternal and neonatal Service Foundation Trust, and the European Union’s outcomes: a prospective cohort. Matern Child Health J Seventh Framework Programme (FP7/2007-2013), project 2014;18:1823–30. EarlyNutrition (grant 289346). SB, KW, NW and KO are 10 Alderman BW, Zhao H, Holt VL, Watts DH, Beresford SA. Maternal physical activity in pregnancy and infant size for gestational age. supported by the Medical Research Council (Unit Pro- Ann Epidemiol 1998;8:513–9. gramme numbers: MC_UU_12015/1, MC_UU_12015/2 and 11 Clapp JF III, Capeless EL. The changing glycemic response to exercise MC_UU_12015/3). during pregnancy. Am J Obstet Gynecol 1991;1:1678Y83. 12 van Poppel MN, Oostdam N, Eekhoff ME, Wouters MG, van Mechelen W, Catalano PM. Longitudinal relationship of physical Supporting Information activity with insulin sensitivity in overweight and obese pregnant women. J Clin Endocrinol Metab 2013;98:2929–35. Additional supporting information may be found online in 13 Jackson MR, Gott P, Lye SJ, Ritchie JW, Clapp JF III. The effects of the Supporting Information section at the end of the maternal aerobic exercise on human placental development: article. placental volumetric composition and surface areas. Placenta Figure S1. Schematic representation of the analysis plan. 1995;16:179–91. 468 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes 14 Wiebe HW, Boule NG, Chari R, Davenport MH. The effect of 34 van Eijsden M, Vrijkotte TG, Gemke RJ, van der Wal MF. Cohort supervised prenatal exercise on fetal growth: a meta-analysis. Obstet profile: the Amsterdam Born Children and their Development Gynecol 2015;125:1185–94. (ABCD) study. Int J Epidemiol 2011;40:1176–86. 15 Sanabria-Martınez G, Garcıa-Hermoso A, Poyatos-Leon  R, Gonzalez- 35 Olsen J, Melbye M, Olsen SF, Sørensen TI, Aaby P, Andersen AM, Garcıa A, Sanchez-Lopez  M, Martınez-Vizcaıno V. Effects of et al. The Danish National Birth Cohort – its background, structure exercise-based interventions on neonatal outcomes: a meta-analysis and aim. Scand J Public Health 2001;29:300–7. of randomized controlled trials. Am J Health Promot 2015;30:214–23. 36 L’Abee C, Sauer PJJ, Damen M, Rake JP, Cats H, Stolk RP. Cohort 16 Schlussel € MM, Souza EB, Reichenheim ME, Kac G. Physical activity profile: the GECKO Drenthe study; overweight programming during during pregnancy and maternal-child health outcomes: a systematic early childhood. Int J Epidemiol 2008;37:486–9. literature review. Cad Saude Publica 2008;24(Suppl 4):s531–44. 37 Polanska K, Hanke W, Gromadzinska J, Llgocka D, Gulczynska E, 17 Bisson M, Lavoie-Guenette J, Tremblay A, Marc I. Physical activity Sobala W, et al. Polish mother and child cohort study – defining the volumes during pregnancy: a systematic review and meta-analysis of problem, the aim of the study and methodological assumptions. Int observational studies assessing the association with infant’s birth J Occup Med Environ Health 2009;22:383–91. weight. AJP Rep 2016;6:e170–97. 38 Walsh JM, McGowan CA, Mahony R, Foley ME, McAuliffe FM. Low 18 Harrod CS, Chasan-Taber L, Reynolds RM, Fingerlin TE, Glueck DH, glycaemic index diet in pregnancy to prevent macrosomia (ROLO Brinton JT, et al. Physical activity in pregnancy and neonatal body study): randomised control trial. BMJ 2012;345:e5605. composition: the Healthy Start study. Obstet Gynecol 2014;124:257–64. 39 Inskip HM, Godfrey KM, Robinson SM, Law CM, Barker DJ, Cooper 19 Perkins CC, Pivarnik JM, Paneth N, Stein AD. Physical activity and C. Cohort profile: the Southampton Women’s Survey. Int J fetal growth during pregnancy. Obstet Gynecol 2007;109:81–7. Epidemiol 2006;35:42–8. 20 Leiferman JA, Evenson KR. The effect of regular leisure physical 40 Ainsworth BE, Haskell WL, Herrmann SD, Meckes N, Bassett DR, Tudor- activity on birth outcomes. Matern Child Health J 2003;7:59–64. Locke C, et al. Compendium of physical activities: a second update of 21 Nieuwenhuijsen MJ, Northstone K, Golding J, ALSPAC Study Team. codes and MET values. Med Sci Sports Exerc 2011;43:1575–81. st Swimming and birth weight. Epidemiology 2002;13:725–8. 41 Intergrowth 21 : the international fetal and newborn growth 22 Hatch MC, Shu XO, McLean DE, Levin B, Begg M, Reuss L, et al. consortium [www.intergrowth21.org.uk]. Accessed 03 July 2017. Maternal exercise during pregnancy, physical fitness, and fetal 42 Slaughter MH, Lohman TG, Boileau RA, Horswill CA, Stillman RJ, growth. Am J Epidemiol 1993;137:1105–14. Van Loan MD, et al. Skinfold equations for estimation of body 23 Hegaard HK, Petersson K, Hedegaard M, Ottesen B, Dykes AK, fatness in children and youth. Hum Biol 1988;60:709–23. Henriksen TB, et al. Sports and leisure-time physical activity in 43 DataSHIELD: a software solution for secure bioscience collaboration pregnancy and birth weight: a population-based study. Scand J Med [http://www.datashield.ac.uk]. Accessed 03 July 2017. Sci Sports 2010;20:e96–102. 44 Durante R, Ainsworth BE. The recall of physical activity: using a 24 Duncombe D, Skouteris H, Wertheim EH, Kelly L, Fraser V, Paxton cognitive model of the question-answering process. Med Sci Sports SJ. Vigorous exercise and birth outcomes in a sample of recreational Exerc 1996;28:1282–91. exercisers: a prospective study across pregnancy. Aust N Z J Obstet 45 Evenson KR, Barakat R, Brown WJ. Guidelines for physical activity Gynaecol 2006;46:288–92. during pregnancy: comparisons from around the world. Am J 25 Klebanoff MA, Shiono PH, Carey JC. The effect of physical activity Lifestyle Med 2014;8:102–21. during pregnancy on preterm delivery and birth weight. Am J 46 Aune D, Sen A, Henriksen T, Saugstad OD, Tonstad S. Physical Obstet Gynecol 1990;163:1450–6. activity and the risk of gestational diabetes mellitus: a systematic 26 Magann EF, Evans SF, Weitz B, Newnham J. Antepartum, review and dose-response meta-analysis of epidemiological studies. intrapartum, and neonatal significance of exercise on healthy, low- Eur J Epidemiol 2016;31:967–97. risk pregnant women. Obstet Gynecol 2002;99:466–72. 47 Gill JM, Celis-Morales CA, Ghouri N. Physical activity, ethnicity and 27 Jukic AM, Evenson KR, Daniels JL, Herring AH, Wilcox AJ, Hartmann cardio-metabolic health: does one size fit all? Atherosclerosis KE. A prospective study of the association between vigorous 2014;232:319–33. physical activity during pregnancy and length of gestation and 48 Konstantina D, Triantafyllou A, GrigoriadouI I, Kintiraki E, birthweight. Matern Child Health J 2012;16:1031–44. Triantafyllou GE, Poulios P, et al. Impairments in microvascular 28 Voldner N, Frøslie KF, Bo K, Haakstad L, Hoff C, Godang K, et al. function and skeletal muscle oxygenation in women with gestational Modifiable determinants of fetal macrosomia: role of lifestyle-related diabetes mellitus: links to cardiovascular disease risk factors. factors. Acta Obstet Gynecol Scand 2008;87:423–9. Diabetologia 2017;60:192–201. 29 Hopkins SA, Cutfield WS. Exercise in pregnancy: weighing up the 49 Helmerhorst HJF, Bragen S, Warren J, Besson H, Ekelund U. A long-term impact on the next generation. Exerc Sport Sci Rev systematic review of reliability and objective criterion-related validity 2011;39:120–7. of physical activity questionnaires. Int J Behav Nutr Phys Act 2012; 30 InterConnect: global data for diabetes and obesity research [www. 9:103. interconnect-diabetes.eu]. Accessed 03 July 2017. 50 Wanner M, Probst-Hensch N, Kriemler S, Meier F, Autenrieth C, 31 Jones EM, Sheehan NA, Masca N, Wallace SE, Murtagh MJ, Burton Martin BW. Validation of the long international physical activity PR. DataSHIELD – shared individual-level analysis without sharing the questionnaire: influence of age and language region. Prev Med Rep data: a biostatistical perspective. Nor Epidemiol 2012;21:231–9. 2016;9:250–6. 32 Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson J, 51 Clapp JF, Kim H, Burciu B, Schmidt S, Petry K, Lopez B. Continuing et al. Cohort profile: the ׳children of the 90s׳ – the index offspring regular exercise during pregnancy: effect of exercise volume on of the Avon Longitudinal Study of Parents and Children. Int J fetoplacental growth. Am J Obstet Gynecol 2002;186:142–7. Epidemiol 2013;58:111–27. 52 Mudd LM, Pivarnik J, Holzman CB, Paneth N, Pfeiffer K, Chung H. 33 Fraser A, Macdonald-Wallis C, Tilling K, Boyd A, Golding J, Davey-Smith Leisure-time physical activity in pregnancy and the birth G, et al. Cohort profile: the Avon Longitudinal Study of Parents and weight distribution: where is the effect? J Phys Act Health 2012;9: Children: ALSPAC mothers cohort. Int J Epidemiol 2013;42:97–110. 1168–77. ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 469 Royal College of Obstetricians and Gynaecologists Pastorino et al. 53 Rose NC, Haddow JE, Palomaki GE, Knight GJ. Self-rated physical 55 McMurray RG, Hackney AC, Guion WK, Katz VL. Metabolic and activity level during the second trimester and pregnancy outcome. hormonal responses to low-impact aerobic dance during pregnancy. Obstet Gynecol 1991;78:1078–80. Med Sci Sports Exerc 1996;28:41–6. 54 Sternfeld B, Quesenberry CP Jr, Eskenazi B, Newman LA. Exercise 56 Bonen A, Campagna P, Gilchrist L, Young DC, Beresford P. during pregnancy and pregnancy outcome. Med Sci Sports Exerc Substrate and endocrine responses during exercise at selected stages 1995;27:634–40. of pregnancy. J Appl Physiol 1992;73:134–42. 470 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BJOG An International Journal of Obstetrics & Gynaecology Pubmed Central

Associations between maternal physical activity in early and late pregnancy and offspring birth size: remote federated individual level meta‐analysis from eight cohort studies

Download PDF
12 pages

Loading next page...
 
/lp/pubmed-central/associations-between-maternal-physical-activity-in-early-and-late-3LJiB2Qai4

References (117)

Publisher
Pubmed Central
Copyright
© 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists
ISSN
1470-0328
eISSN
1471-0528
DOI
10.1111/1471-0528.15476
Publisher site
See Article on Publisher Site

Abstract

DOI: 10.1111/1471-0528.15476 Epidemiology www.bjog.org Associations between maternal physical activity in early and late pregnancy and offspring birth size: remote federated individual level meta-analysis from eight cohort studies a a b c d e,f g € € S Pastorino, T Bishop, SR Crozier, C Granstrom, K Kordas, LK Kupers, EC O’Brien, h i j,k,l m j m b,n K Polanska, KA Sauder, MH Zafarmand, RC Wilson, C Agyemang, PR Burton, C Cooper, e o h b,n g c j a E Corpeleijn, D Dabelea, WHanke, HM Inskip, FM McAuliffe, SF Olsen, TG Vrijkotte, SBrage, p,q n a a a r, a, A Kennedy, DO’Gorman, P Scherer, K Wijndaele, NJ Wareham, G Desoye, * KK Ong * a b MRC Epidemiology Unit, University of Cambridge, Cambridge, UK MRC Lifecourse Epidemiology Unit (University of Southampton), Southampton General Hospital, Southampton, UK Department of Epidemiology Research, Centre for Fetal Programming, State Serum Institute, Copenhagen, Denmark Epidemiology and Environmental Health, School of Public Health and Health Professions, University at Buffalo, Buffalo, e f NY, USA Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands MRC Integrative Epidemiology Unit, School of Social and Community Medicine, University of Bristol, Bristol, UK Obstetrics & Gynaecology, UCD Perinatal Research Centre, School of Medicine, University College Dublin, National Maternity Hospital, Dublin, Ireland Department of Environmental Epidemiology, Nofer Institute of Occupational Medicine, Lodz, Poland Department of Pediatrics, University of Colorado School of Medicine, Aurora, CO, USA Department of Public Health, Amsterdam Public Health Research Institute, Amsterdam UMC, University of Amsterdam, Amsterdam, the Netherlands Department of Obstetrics & Gynaecology, Amsterdam UMC, University of Amsterdam, Amsterdam, the Netherlands Department of Clinical Epidemiology, Biostatistics and Bioinformatics, Amsterdam Public Health Research Institute, Amsterdam m n UMC, University of Amsterdam, the Netherlands Institute of Health and Society, Newcastle University, Newcastle, UK NIHR Southampton Biomedical Research Centre, University Hospital Southampton NHS Foundation Trust and University of Southampton, Southampton, UK o p Department of Epidemiology, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Denver, CO, USA 3U Diabetes Consortium and School of Health and Human Performance, Dublin City University, Dublin, Ireland School of Biological Sciences, Dublin Institute of Technology, Dublin, Ireland Department of Obstetrics & Gynaecology, Medical University of Graz, Graz, Austria Correspondence: S Pastorino, MRC Epidemiology Unit, Level 3 Institute of Metabolic Science, University of Cambridge School of Clinical Medicine, Cambridge CB2 0SL, UK. Email: silvia.pastorino@mrc-epid.cam.ac.uk Accepted 10 September 2018. Published Online 16 October 2018. Objective Evidence on the impact of leisure time physical activity expenditure were modestly inversely associated with BW, LGA, macrosomia, and ponderal index, without heterogeneity (all: (LTPA) in pregnancy on birth size is inconsistent. We aimed to I = 0%). For each extra hour/week of MVPA, RR for LGA and examine the association between LTPA during early and late macrosomia were 0.97 (95% CI: 0.96, 0.98) and 0.96 (95% CI: 0.94, pregnancy and newborn anthropometric outcomes. 0.98), respectively. Associations were only modestly reduced after Design Individual level meta-analysis, which reduces heterogeneity additional adjustments for maternal BMI and gestational diabetes. across studies. No measure of LTPA was associated with risk for SGA. Setting A consortium of eight population-based studies (seven European and one US) comprising 72 694 participants. Conclusions Physical activity in late, but not early, pregnancy is consistently associated with modestly lower risk of LGA and Methods Generalised linear models with consistent inclusion of macrosomia, but not SGA. confounders (gestational age, sex, parity, maternal age, education, ethnicity, BMI, smoking, and alcohol intake) were used to test Keywords Birth weight, large-for-gestational age, macrosomia, associations between self-reported LTPA at either early (8– physical activity, pregnancy, small-for-gestational age. 18 weeks gestation) or late pregnancy (30+ weeks) and the Tweetable abstract In an individual participant meta-analysis, late outcomes. Results were pooled using random effects meta-analyses. pregnancy moderate to vigorous physical activity modestly Main outcome measures Birth weight, large-for-gestational age reduced birth size outcomes. (LGA), macrosomia, small-for-gestational age (SGA), % body fat, and ponderal index at birth. Linked article This article is commented on by CA Venetis, p.471 in this issue. To view this mini commentary visit https://doi.org/10.1111/ Results Late, but not early, gestation maternal moderate to vigorous 1471-0528.15542. physical activity (MVPA), vigorous activity, and LTPA energy *Equal contributions. ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 459 Royal College of Obstetricians and Gynaecologists. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. [The copyright line for this article was changed on 22 August 2019 after original online publication.] Pastorino et al. Please cite this paper as: Pastorino S, Bishop T, Crozier SR, Granstrom € C, Kordas K, Kupers € LK, O’Brien EC, Polanska K, Sauder KA, Zafarmand MH, Wilson B, Agyemang C, Burton PR, Cooper C, Corpeleijn E, Dabelea D, Hanke W, Inskip HM, McAuliffe FM, Olsen SF, Vrijkotte TG, Brage S, Kennedy A, O’Gorman D, Scherer P, Wijndaele K, Wareham NJ, Desoye G, Ong KK. Associations between maternal physical activity in early and late pregnancy and offspring birth size: remote federated individual level meta-analysis from eight cohort studies. BJOG 2019;126:459–470. meta-analysis to be performed remotely. Compared with a Introduction literature-based meta-analysis, this approach can reduce The prevalence of childhood obesity has increased world- heterogeneity between studies by allowing harmonisation of wide over the last three decades. Babies born with large- exposure and outcome variables, and by allowing the same for-gestational age (LGA), or with macrosomia [birth models to be tested in each study. weight (BW) above 4000 or 4500 g], have higher risks of obesity and raised metabolic disease markers in childhood Methods 2,3 compared with babies with appropriate BW. Physical activity during pregnancy is recommended to enhance the InterConnect is an EU-FP7 funded project that optimises health of the mother-to-be, but has also been explored as the use of existing data by enabling cross-cohort analyses a potential intervention to lower the risk for LGA and within consortia without pooling of individual-level data at 5–10 macrosomia. Physical activity might be especially a central location. For this research question, eight cohorts appealing if it reduced high BW without increasing the risk with data on physical activity in pregnancy and neonatal of small-for-gestational age (SGA) babies. Physical activity outcomes set up a server to allow remote federated analyses during pregnancy might reduce fetal growth by increasing and joined the consortium. The collaborative group com- insulin sensitivity and by modulating glucose regula- prised the following prospective birth cohort studies: the 11,12 tion. Physical activity might also regulate fetoplacental Avon Longitudinal Study of Parents and Children 32,33 growth by altering the rates of oxygen and nutrient sup- (ALSPAC, UK), the Amsterdam Born Children and 13 34 ply. their Development study (ABCD, the Netherlands), the Recent systematic reviews of randomised controlled trials Danish National Birth Cohort (DNBC, Denmark), the on the effect of maternal exercise on birth outcomes report Groningen Expert Center for Kids with Obesity (GECKO)- 14,15 36 modest BW reductions (10–30 g). However, they report Drenthe (the Netherlands), the Healthy Start Study (HSS, wide variation in the types of interventions studied in USA), the Polish Mother and Child Cohort (REPRO_PL, 37 38 terms of form, intensity, and volume of exercise. Systematic Poland), the ROLO study (Ireland), and the Southamp- reviews of observational studies on the association between ton Women’s Survey (SWS, UK). Characteristics of the maternal physical activity during pregnancy with birth participating studies are shown in Table S1. Each partici- 16,17 size report conflicting results: some studies report an pating cohort obtained ethical approval from the corre- 5–10,18,19 20–22 inverse association, some a positive association, sponding local ethics committee (see details at the end). 23–28 and others no significant association. There is also No PPI took place for these analyses. some evidence that the timing of physical activity in preg- We included all live-born singleton full-term births and 18,29 nancy might be important. The heterogeneity among excluded mothers with pre-eclampsia and those with miss- studies limits the ability to pool published results. One ing information for any of the covariates. The percentage meta-analysis reports that ‘high’ physical activity levels of participants with any missing values across cohorts ran- were inversely associated with BW, but conversely ‘moder- ged between 10.2% and 34% for early pregnancy analyses, ate’ physical activity levels were positively associated with and between 12.7% and 43.5% for late pregnancy analyses. BW. The included studies use different definitions of physi- Funding for this study was received from the European cal activity level and there is no standardisation with regard Union Seventh Framework Programme (FP7/2007–2013) to the type and domains of activity or the volume, inten- under grant agreement no. 602068. Core Outcome Set sity, and timing. Most studies did not adjust for any con- (COS), and patient involvement (PPI) is not relevant to founder. this study and hence is not described here. Here, we examined the association between leisure time physical activity (LTPA) during pregnancy and newborn Physical activity during pregnancy anthropometric outcomes across a range of prospective All studies assessed physical activity during pregnancy by cohort studies. Within a consortium created as part of the questionnaire. HSS and SWS used interviewer-administered InterConnect project, we used a federated meta-analysis questionnaires, DNBC used a computer-assisted telephone approach, which allows an individual participant-level interview, and the other studies used self-administered 460 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes questionnaires. Table S2 details the questions in each in early pregnancy up to week 18 of gestation. We applied cohort. We harmonised self-reported data on LTPA during a uniform correction factor to weights measured later than early pregnancy across seven cohorts, and LTPA during late 12 weeks gestation derived by weight gain curves based on pregnancy across five cohorts. The median gestational age repeated maternal weight measures in the ALSPAC study. at which mothers replied to questionnaires was 8–18 weeks There was wide variation in definitions of ethnicity across for early pregnancy, and 30 weeks to 1 day post-delivery cohorts; the ‘other ethnicity’ category includes a variety of for late pregnancy. LTPA was chosen as it is the domain Asian, Hispanic, and other ethnic groups. GDM was most amenable to intervention and therefore more relevant defined using biochemical data at weeks 24–28 in HSS and for public health recommendations; it was also the most ROLO, and by a combination of medical records and self- commonly assessed domain across the eight studies. Inten- reports in the other studies. sity of reported activities was expressed in metabolic equiv- alent of energy expenditure (MET) values according to the Potential confounders and other covariates Compendium of Physical Activity. Four exposure vari- Potential confounders were not harmonised because, in ables were harmonised: (1) duration of LTPA (hours/ federated analysis models involving random-effects meta- week), which included any reported leisure time activity; analysis of the arising study-specific estimates, this would (2) duration of moderate-vigorous LTPA (MVPA) (hours/ not impact the summary effect estimates and P-values. week) including activities with intensity ≥3 MET; (3) dura- However, confounder variables were reasonably comparable tion of vigorous LTPA (VPA) (hours/week) including across studies. Smoking in pregnancy was a dichotomous activities with intensity ≥6 MET; (4) energy expenditure of variable (yes/no) in all studies except DNBC, which deter- LTPA (MET-hours/week) calculated by multiplying dura- mined the number of cigarettes/week. Alcohol intake was tion of LTPA by MET values. Three studies recorded cate- considered as units of alcohol/week in ALSPAC, DNBC, gorical response formats for duration of LTPA (ALSPAC, and SWS; glasses/week in ABCD; and as categorical vari- GECKO, and SWS). These were converted into numerical ables in GECKO (none, <1 glass/week, 1–6 glasses/week, values, where relevant using the mid-point of the stated 7+ glasses/week), HSS (none, once per month or less, twice range (i.e. ‘>7 hours/week’ was converted to 7 hours/week; per month or more), REPRO_PL (yes/no), and ROLO ‘2–6’ to 4; <1 to 0.5; ‘never’ to 0). (yes/no). Educational attainment was considered as a cate- gorical variable in most cohorts (range 2–6 levels) except Outcomes ABCD, which recorded ‘years of education after elementary The following outcome variables were harmonised across school’. Parity (number of previous live births) was self- all studies, based on objective measurements in all studies: reported in all studies, and maternal age at delivery was BW (g), macrosomia (defined as BW >4000 g), LGA (BW calculated from mother’s date of birth and delivery date. for gestational age >90th percentile according to the INTERGROWTH-21st Project ), and SGA (BW for gesta- Statistical analyses tional age <10th percentile according to INTERGROWTH- All analyses were conducted using R within the Data- 21st). Ponderal Index, a measure of leanness (corpulence) SHIELD federated meta-analysis library. In this process, 3 3 [weight/length (kg/m )] at birth was harmonised for six individual participant data from contributing studies are cohorts. Percent (%) body fat at birth was available for held securely on servers at each study location. A com- three cohorts. Of these, one (HSS) assessed newborn body puter within the network sends analytical commands that fat using air displacement plethysmography (PEAPOD), request each local server to undertake an analysis locally while skinfold thickness measurements were available in and return non-identifiable summary statistics. The result HSS, SWS, and in a subset of ROLO (n = 219). Triceps of this process is mathematically equivalent to an individ- and subscapular skinfolds were used to estimate % body fat ual participant meta-analysis with the advantage that data using the algorithm reported by Slaughter et al.: % body remain within the governance structure of each single fat = 1.21 9 (triceps skinfold + subscapular skinfold)  cohort study. 0.008 9 (triceps skinfold + subscapular skinfold)  1.7. To analyse data, we used generalised linear models in each study. Each model was fitted in a federated manner Potential modifiers using the iterative reweighted least squares process. The The following potential modifying variables were harmo- primary models included MVPA duration as exposure and nised across the studies: infant sex, maternal obesity (BMI: each outcome (BW, macrosomia, LGA, SGA, ponderal 2 2 2 ≤20 kg/m , >20–30 kg/m , >30 kg/m ), maternal ethnicity index, % body fat) separately. Moderate to vigorous activ- (white, black, other), and gestational diabetes mellitus (GD: ity was chosen as the primary exposure because it has yes, no). Maternal weight was objectively measured in five higher validity than lower intensity activities; also, the cohorts and self-reported in three cohorts at varying times majority of existing guidelines recommend moderate ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 461 Royal College of Obstetricians and Gynaecologists Pastorino et al. intensity physical activity for pregnant women. The cohorts, macrosomia prevalence ranged between 5.6% in adjusted models included each exposure separately (LTPA HSS and 21.7% in DNBC, and LGA between 8.7% in HSS duration, MVPA duration, VPA duration, LTPA energy and 30.2% in GECKO. SGA prevalence ranged between expenditure) with each outcome (BW, macrosomia, LGA, 0.8% in ROLO and 9.4% in HSS. Median ponderal index SGA, ponderal index, % body fat), and were adjusted for at birth ranged between 20.2 in REPRO_PL and 27.8 in gestational age (except for LGA and SGA), infant sex, par- SWS, and body fat was 10, 11, and 16% in HSS, SWS, and ity, maternal age, smoking, alcohol, maternal education, ROLO, respectively. and ethnicity. Further models were additionally adjusted Reported levels of maternal LTPA during pregnancy var- for maternal early pregnancy BMI. A schematic diagram ied across studies, with DNBC women having the lowest of the analysis plan is shown in Figure S1. All covariates levels in both periods (64% of women reporting no LTPA). were chosen a priori based on literature evidence. To Among the other cohorts, median LTPA duration ranged explore which covariate contributed most to heterogeneity, from 2.0 to 6.5 hours/week for early pregnancy, and 1– we conducted further analyses by including each potential 7 hours/week for late pregnancy. Median MVPA levels ran- confounding variable one at a time. Physical activity is ged from 0 to 4 hours/week for early pregnancy, and 0– likely to exert its effect on birth size by altering maternal 0.8 hours/week for late pregnancy. The proportion of metabolic pathways such as glucose metabolism, and there women reporting any MVPA decreased from the early is evidence of its association with GDM. Therefore, pregnancy in the four studies with data at both time points GDM was added in a subsequent model to explore its (DNBC, 34%; HSS, 72%; REPRO_PL, 20%; SWS, 84%) to possible mediating effect. We explored the possible modi- late pregnancy (DNBC, 25%; HSS, 49%; REPRO_PL, 12%; fying effect of infant sex, maternal obesity, maternal eth- SWS, 78%). The proportion of women reporting any VPA nicity, and GDM by including interaction terms in the was low in most cohorts (range: 6.6–42.5%) and decreased model. These potential effect modifiers were chosen a pri- in late pregnancy (range: 2.9–24.1%). ori. The levels of physical activity and their effects on health differ across ethnic groups. In pregnant women, Physical activity associations in early pregnancy both obesity and GDM might alter physiological charac- Early pregnancy maternal LTPA was not associated with teristics that affect their ability to exercise. All models any measure of offspring birth size (Tables 2, S3, and S4). were conducted separately for early and late pregnancy Heterogeneity across studies was high in unadjusted models physical activity. Early pregnancy physical activity mea- (I = 79–86% for BW, macrosomia, and LGA, Table S1), sures were available for ALSPAC, ABCDS, DNBC, HSS, but was substantially reduced after adjustments for poten- REPRO-PL, ROLO, and SWS. Late pregnancy physical tial confounders (0–54%, Table 2). In sensitivity models, activity measures were available for DNBC, GECKO, HSS, with stepwise inclusion of covariates, ethnicity and mater- REPRO_PL, and SWS. Regression analyses were conducted nal education contributed the most to (positive) confound- for each individual study, and then random-effects meta- ing in some individual studies, with non-white ethnicity analysis was used to combine the effect estimates. A ran- being associated with both lower BW and lower LTPA, and dom effects approach was chosen owing to the reported maternal education being associated with both higher BW heterogeneity between other published studies. Hetero- and higher LTPA (not shown). geneity was assessed using the I statistic. Physical activity associations in late pregnancy Late pregnancy maternal MVPA (Figures 1 and 2, Table 2), Results VPA, and LTPA energy expenditure (Tables 2 andS3) were For early pregnancy physical activity analyses, 72 694 par- inversely associated with all birth size outcomes (except for ticipants from seven studies were included (57 807 across % body fat and SGA) in adjusted models. For each +1 hour/ six studies for ponderal index; 3039 in three studies for % week of MVPA, offspring BW was lower by 6.4 g (95% CI: body fat). For late pregnancy analyses, the available sample 9.1, 3.7; P <0.001) and ponderal index by 0.02 kg/m (95% was 58 820 from five studies (57 172 across four studies CI: 0.03, 0.00; P = 0.02); the relative risks of macrosomia for ponderal index; 2792 in two studies for % body fat). and LGA were lower by 4% (95% CI: 2, 6; P <0.001) and 3% Maternal and infant characteristics are presented in (95% CI: 2, 4; P <0.01), respectively. No association was Table 1. Mean BW ranged between 3356 and 4135 g for found for SGA (OR: 0.99, 95% CI: 0.98, 1.00) and % body male infants, and between 3217 and 3963 g for female fat (0.01, 95% CI: 0.04, 0.02). VPA showed larger associ- infants. ROLO infants had the highest mean BW and high- ations with BW (22 g/hour/week; 95% CI: 31.3, 12.7; est prevalence of macrosomia (51.8%) and LGA (61.7%), P <0.001), ponderal index (0.07 units; 95% CI: 0.13, reflecting their inclusion of only secundigravid women 0.02; P <0.01), macrosomia (lower by 11%, 95% CI: 5, 16; whose first baby had been macrosomic. Among the other P <0.01) and LGA (lower by 11%, 95% CI: 5, 16 P <0.001), 462 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 463 Royal College of Obstetricians and Gynaecologists Table 1. Study population characteristics ALSPAC ABCD DNBC GECKO HSS REPRO_PL ROLO SWS n* 9058 6464 53 671 1335 1054 982 617 1902 Female, n (%) 4428 (48.9) 3244 (50.2) 26 371 (49.2) 638 (41.8) 509 (48.3) 487 (49.6) 309 (50.1) 916 (48.2) Birth weight, g Male, mean (SD) 3551 (479) 3572 (491) 3709 (503) 3708 (505) 3356 (432) 3490 (440) 4135 (481) 3589 (480) Female, mean (SD) 3424 (447) 3435 (456) 3575 (481) 3538 (490) 3217 (420) 3316 (432) 3963 (423) 3445 (458) Macrosomia, n (%) 1158 (12.7) 871 (13.4) 11 681 (21.7) 289 (21.6) 60 (5.6) 84 (8.5) 320 (51.8) 267 (14) LGA, n (%) 1888 (20.8) 1222 (18.9) 15 052 (28) 405 (30.3) 121 (8.7) 183 (18.6) 381 (61.7) 369 (19.4) SGA, n (%) 418 (4.6) 311 (4.8) 1849 (3.4) 59 (6.4) 100 (9.4) 58 (5.9) 5 (0.8) 101 (5.3) Ponderal index**, median (IQR) 26.2 (24.7–27.8) 24.9 (23.5–26.5) 26.9 (24.9–29.2) 20.2 (18.9- 21.6) 27.1 (25.3–29.3) 27.8 (26.3–29.2) % Body fat***, median (IQR) 10 (8–12) 16 (14–18) 11 (10–13) Early pregnancy physical activity, median (IQR) LTPA (hours/week) 4.0 (0.5–5.5) 2.0 (0.5–4.3) 0.0 (0.0–1.0) 3.0 (1.0–5.8) 4.0 (0.0–7.0) 1.7 (1.0–2.3) 6.5 (3.2–11.5) MVPA (hours/week) 4.0 (0.5–5.0) 1.5 (0.0–3.5) 0.0 (0.0–1.0) 1.5 (0.0–3.5) 0.0 (0.0–0.0) 0.3 (0.0–1.0) 1.2 (0.3–3.0) LTPA EE (Met-hours/week) 15.2 (3.0–25.2) 8.1 (1.7–19.3) 0.0 (0.0–6.0) 10.2 (3.1–23.6) 16.5 (0.0–33.0) 4.5 (2.0–7.8) 17.5 (8.7–32.1) Women doing vigorous PA, n (%) 604 (6.6) 1876 (29) 4321 (8.0) 244 (23.1) 84 (8.5) 61 (9.8) 810 (42.5) Late pregnancy physical activity, median (IQR) LTPA (hours/week) 0.0 (0.0–1.0) 1.0 (1.0–1.0) 2.0 (0.5–3.6) 5.0 (0.0–8.0) 7.0 (3.4–12.0) MVPA (hours/week) 0.0 (0.0–1.0) 0.3 (0.0–1.0) 0.0 (0.0–1.5) 0.0 (0.0–0.0) 0.8 (0.1–2.3) Women doing vigorous PA, n (%) 1599 (2.9) 61 (5.7) 77 (8.3) 443 (24.1) LTPA EE (Met-hours/week) 0.0 (0.0–3.0) 1.0 (0.0–4.0) 6.3 (1.5–11.9) 19.8 (0.0–33.0) 16.7 (8.5–31.1) Maternal age (y), mean (SD) 28.5 (4.7) 30.9 (5.1) 30.1 (4.2) 30.8 (4.2) 28 (6.1) 29.0 (4.2) 32.2 (4.1) 30.6 (3.7) Maternal BMI Mean (SD) 22.5 (4.3) 24.0 (4.1) 24.7 (4.1) 24.7 (4.7) 26.7 (6.0) 22.8 (3.6) 26.6 (4.8) 26 (4.8) Overweight, n (%) 1257 (13.8) 1447 (22.3) 14 896 (27.7) 320 (24) 334 (31.6) 162 (16.4) 233 (37.7) 606 (32) Obese, n (%) 586 (6.4) 527 (8.1) 5546 (10.3) 169 (12.6) 225 (21.3) 39 (3.9) 114 (18.4) 323 (17) GDM, n (%) 41 (0.4) 76 (1.1) 380 (10.7) 44 (3.2) 43 (4) 34 (3.4) 12 (2) 19 (1) Ethnicity White 8867 (98) 4490 (69.4) 53 671 (100) 1321 (99) 814 (76.5) 982 (100) 612 (97.5) 1840 (96.8) Black 77 (0.8) 486 (7.6) 0 (0) 0 (0) 162 (15.3) 0 (0) 2 (0.3) 10 (0.5) Other 114 (1.2) 1488 (23) 0 (0) 14 (1) 78 (7.2) 0 (0) 14 (2.2) 52 (2.7) EE, energy expenditure; GDM, gestational diabetes mellitus; LGA, large for gestational age; LTPA, leisure-time physical activity; MVPA, moderate to vigorous leisure time physical activity; SGA, small for gestational age. *Sample size available for late pregnancy physical activity analyses were: DNBC = 53 684, HSS = 1044, REPRO_PL = 919, SWS = 1838. **Sample size available for analyses of Ponderal index for early pregnancy analyses were: ALSPAC = 7118, DNBC = 53 487, HSS = 976, REPRO_PL = 977, ROLO = 523, SWS = 1844; for late pregnancy analyses: DNBC = 53 500, HSS = 968, REPRO_PL = 915, SWS = 1789. ***Sample size available for analyses of % body fat for early pregnancy analyses were: HSS = 988, ROLO = 189, SWS = 1862; for late pregnancy analyses: HSS = 987, SWS = 1805. Pastorino et al. Table 2. Associations between physical activity during pregnancy and offspring birth size BW (grams) Macrosomia LGA Ponderal index SGA RR, 95% CI RR, 95% CI RR, 95% CI Beta, 95% CI Beta, 95% CI 2 2 2 2 2 I I I I I Physical activity Early pregnancy LTPA (hours/week) 0.86 (2.33, 0.61) 0.99 (0.98, 1,01) 0.99 (0.98, 1,00) 0.0 (0.01, 0.01) 0.99 (0.98, 1.01) 23% 51% 46% 0% 0% MVPA (hours/week) 1.38 (3.77, 1.01) 1.00 (0.98, 1,01) 1.00 (0.98, 1,01) 0.00 (0.01, 0.01) 0.99 (0.98, 1.00) 41% 52% 43% 0% 0% VPA (hours/week) 1.38 (3.77, 1.01) 1.00 (0.98, 1,01) 1.00 (0.98, 1,01) 0.00 (0.05, 0.04) 0.99 (0.98, 1.00) 41% 52% 43% 18% 0% LTPAEE (met-hours/week) 0.14 (0.58, 0.30) 1.00 (0.99, 1,00) 0.99 (0.99, 1,00) 0.00 (0.00, 0.00) 0.99 (0.99, 1.00) 49% 53% 38%) 0% 0% Physical activity Late pregnancy LTPA (hours/week) 2.22 (5.54, 1.0) 0.98 (0.96, 1.00) 0.98 (0.97, 0.99) 0.01 (0.02, 0.00) 0.99 (0.97, 1.01) 64% 37% 0% 13% 0% MVPA (hours/week) 6.43 (9.12, 3.74) 0.96 (0.94, 0.98) 0.97 (0.96, 0.98) 0.02 (0.03, 0.00) 1.01 (0.97, 1.03) 0% 0% 0% 0% 0% VPA (hours/week) 22.0 (31.3, 12.7) 0.89 (0.84, 0.95) 0.89 (0.84, 0.94) 0.07 (0.13, 0.02) 1.06 (0.96, 1.17) 0% 0% 0% 0% 0% LTPAEE (met-hours/week) 0.93 (1.43, 0.42) 0.99 (0.99, 0.99) 0.99 (0.99, 0.99) 0.00 (0.01, 0.00) 0.99 (0.99, 1.00) 9% 0% 0% 0% 0% EE, energy expenditure; LGA, large for gestational age; LTPA, leisure time physical activity; MVPA, moderate to vigorous leisure time activity; SGA, small for gestational age; VPA, vigorous leisure time activity. Models are adjusted for gestational age, sex, parity, maternal age, smoking, alcohol, maternal education, and ethnicity. Statistically significant associations are highlighted in bold. and no association with % body fat (0.05; 95% CI: 0.17, with the inconsistent findings of published literature-based 0.06) and SGA (OR: 1.01, 95% CI: 0.96, 1.16). The associa- systematic reviews, heterogeneity between study estimates tions with late pregnancy LTPA were not mediated by GDM was substantially reduced by consistent confounding adjust- and persisted after further adjustment for early pregnancy ment and by harmonisation of exposures and outcomes. maternal BMI (Table S5). The remote federated analysis approach avoided the need No interaction with ethnicity, infant sex, GDM, or to physically pool individual-level data, and hence substan- maternal obesity was found in either pregnancy period for tially reduced the governance burdens and associated time LTPA and birth size (all P-values for interactions >0.05). delays, and avoided barriers due to limitations of consent and research ethics permissions. Another strength is that we were able to analyse the differential association of tim- Discussion ing and intensity of physical activity in pregnancy with off- Main findings spring birth size outcomes. In this large cross-cohort analysis of up to 72 694 individuals, However, there were some limitations in our approach. we found small but consistent inverse associations between Physical activity was self-reported in all included studies, maternal LTPA during late, but not early, pregnancy and off- and only a few of the questionnaires were validated. Phys- spring birth size. Each additional hour/week of MVPA in late ical activity questionnaires are susceptible to measurement pregnancy was associated with 6.4 g lower birth weight and error related to both recall and social desirability with 4% and 3% relative reductions in risk of macrosomia and validity estimated between 0.25 and 0.4. However, they LGA, respectively, without increasing the risk of SGA. are able to rank individuals according to activity levels. Furthermore, validity is higher among women than men Strengths and limitations and for vigorous intensity compared with lighter intensity A major strength of our approach was the planned individ- activities. It remains a challenge to identify thresholds of ual level analysis across several cohort studies. Compared physical activity in terms of health benefits. Contributing 464 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes Study (weight) Birth weight Beta (95% CI) Grams Study (weight) Ponderal Index Beta (95% CI) kg/m Figure 1. Forest plots for late pregnancy moderate to vigorous activity (hours/week) associated with birth weight and ponderal index. Associations were adjusted for gestational age, sex, parity, maternal age, smoking, alcohol, maternal education, and ethnicity. n = 58 820 except for ponderal index (n = 57 172). studies used different questionnaires with varying ways of analyses including such outcomes would be highly infor- assessing LTPA, which made harmonisation challenging. mative. Our use of international INTERGROWTH-21st For example, some listed specific activities (e.g. ‘swim- Project data to define LGA and SGA led to unequal num- ming’, ‘walking’) while others asked only about categories bers for those outcomes and limited the statistical power of activities (i.e. ‘moderate, ‘vigorous’), which included to detect a possible association between VPA and SGA. some activities outside of leisure time. Intensity informa- Although we adjusted for many confounding factors, tion was not available in all questionnaires, which meant residual confounding cannot be ruled out. Limited geo- assumptions had to be made when assigning MET values. graphical and ethnic diversity restricted the power to Differences in average LTPA levels across the studies detect modifying factors. One participating study (DNBC) might therefore reflect differences in methods or real pop- was substantially larger than the other studies and ulation differences. The timing of questionnaire adminis- accounted for more than 70% of the sample size in the tration differed across studies, particularly for early analyses. Whilst the dominance of this study in driving pregnancy LTPA. Unfortunately, data were unavailable on results should be acknowledged, it is noteworthy that, in clinical outcomes associated with LGA and macrosomia adjusted models, heterogeneity was reduced from >70 to (such as shoulder dystocia, 3rd or 4th degree laceration), 0% in most analyses, thus highlighting the consistency or on pregnancies not resulting in live birth. Future across studies and the generalisability of results. ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 465 Royal College of Obstetricians and Gynaecologists Pastorino et al. Study (weight) Macrosomia RR (95% CI) RR Study (weight) LGA RR (95% CI) RR Figure 2. Forest plots for late pregnancy moderate to vigorous activity (hours/week) associated with relative risk of macrosomia and large for gestational age (LGA). Associations were adjusted for gestational age, sex, parity, maternal age, smoking, alcohol, maternal education, and ethnicity. n = 58 820. Interpretations adjustment for confounding reduced heterogeneity The direction of our associations is consistent with some between individual study estimates from I >70% to 0% 5–10,18,19 previous individual studies; however, other studies in several analyses. Furthermore, adjustment for ethnicity 23–28 20– reported null or even directionally opposite results. and maternal education avoided spurious positive associa- 22 17 tions between early pregnancy physical activity and birth A recent meta-analysis reports that a ‘moderate’ level size. We harmonised the intensity of activities by assign- of physical activity was positively associated with BW, ing the same MET values for similar reported activities while a ‘high’ level of physical activity was inversely asso- across studies. Although the diverse nature of the ques- ciated with BW. However, those results were based on a tionnaires used in the individual studies made harmonisa- mixture of adjusted and unadjusted models, and their tion challenging, MVPA was less heterogeneous than reported meta-analysis of only the adjusted models other activity variables, particularly in late pregnancy; this showed null associations for both moderate and high may be because our harmonised MVPA variable was more levels of physical activity. Furthermore, in that meta-ana- robust to underlying methodological differences across lysis, there was substantial heterogeneity, with I values studies. >80%. We demonstrate here that more consistent 466 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes The timing of PA associations with LTPA during late, Disclosure of interests but not early, pregnancy is also consistent with some None declared. Completed disclosure of interests form 18,29,51 51 reported studies. Clapp et al. reported inverse asso- available to view online as supporting information. ciations with newborn adiposity or BW only for late preg- nancy physical activity. Hopkins and Cutfield conjectured Contribution to authorship that high volume exercise only in the first half of preg- GD, KKO, SP contributed to planning the study. SP, KW, nancy increased BW, but if performed throughout preg- SB, AK, DOG, and KKO coordinated harmonisation of all nancy or only in the second half of pregnancy, it reduced variables. TB and PS conducted the federated remote statisti- BW. They suggested that the timing of physical activity cal analyses. SP, KKO, DOG, and SB interpreted the results. caused different fetoplacental adaptations. SP wrote the article. SP, TB, SB, KW, NJW, GD, KKO, SRC, Regarding intensity of LTPA, we found that late preg- CG, KK, LKK, EOB, KP, KAS, MHZ, BW, CA, PRB, CC, EC, nancy MVPA, VPA, and energy expenditure, but not DD, WH, HMI, FM, SFO, and TGV contributed to the anal- duration of LTPA, were inversely associated with offspring ysis plan, the production of the paper, the harmonisation birth size. Some previous studies have assessed the impact algorithms, and the review of the manuscript. of physical activity intensity on offspring birth size, with 21,22,51 some findings consistent with ours, but others Details of ethics approval 27,52–54 reported null results. Different adjustment factors Avon Longitudinal Study of Parents and Children and different definitions, timing, and categories of physi- (ALSPAC): ethical approval for the study was obtained cal activities might lead to inconsistent findings between from ALSPAC Ethics and Law Committee and the Local studies. Although the proportion of women reporting any Research Ethics Committees. Amsterdam Born Children VPA was small, our results suggest that changes in birth and their Development study (ABCD): approval of the size outcomes are dependent on the intensity of LTPA, study was obtained from the Central Committee on with larger effects observed with higher intensity. It is Research Involving Human Subjects in The Netherlands, possible that LTPA intensity needs to reach a certain the medical ethics review committees of the participating threshold before it has an effect on nutrient supply to the hospitals, and the Registration Committee of the Munici- fetus. Alternatively, higher intensity recreational activities pality of Amsterdam. Danish National Birth Cohort may be easier to recall and less prone to measurement 44 (DNBC, Denmark): approved by the Committee on error. Biomedical Research Ethics under case number (KF) 01- Our observed associations remained significant after 471/94. Groningen Expert Center for Kids with Obesity adjustment for maternal BMI, possibly suggesting that the (GECKO )-Drenthe: approved by the Medical Ethics effect of physical activity on birth size is only partially Committee of the University Medical Center Groningen mediated by maternal weight; however, we did not have (UMCG). Healthy Start Study (HSS): approved by the measures of late pregnancy maternal weight gain and Colorado Multiple Institutional Review Board. Polish BMI. Independent of maternal weight, physical activity Mother and Child Cohort (REPRO_PL): approved by the 12,55 increases maternal insulin sensitivity, reduces maternal Ethical Committee of the Nofer Institute of Occupational glucose, and hence might reduce glucose transfer to the Medicine, Łod  z, Poland (Decision No. 7/2007). ROLO fetus. These metabolic changes are more marked at study: approved by the Ethics Committee at the National higher intensities and volumes of exercise and in late Maternity Hospital, June 2006. Southampton Women’s 11,29 pregnancy. Survey (SWS): approved by South Central—Hampshire B Research Ethics Committee. Conclusion Funding In conclusion, LTPA energy expenditure, MVPA, and VPA InterConnect: the research leading to these results received during late, but not early, pregnancy had a small but signif- funding from the European Union Seventh Framework icant and consistent inverse association with offspring birth Programme (FP7/2007–2013) under grant agreement no. size. Larger effects were observed with higher intensity of 602068. Amsterdam Born Children and their Development physical activity. Compared with the inconsistent findings study (ABCD): the ABCD study was supported by grants of reviews of published reports, this remote federated indi- from the Netherlands Organization for Health Research vidual-level analysis substantially reduced heterogeneity and Development (ZonMW) and The Netherlands Heart between individual studies by allowing consistent adjust- Foundation. Genotyping was funded by the BBMRI-NL ment for confounding and careful harmonisation of expo- grant CP2013-50. M.H. Zafarmand was supported by sures and outcomes. ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 467 Royal College of Obstetricians and Gynaecologists Pastorino et al. BBMRI-NL (CP2013-50). T.G.M. Vrijkotte was supported Table S1. Characteristics of the eight contributing cohort by ZonMW (TOP 40–00812–98–11010). Avon Longitudi- studies. nal Study of Parents and Children (ALSPAC): we are Table S2. Questions asked in the participating cohorts grateful to all the families who took part in this study, the used for harmonisation of leisure time physical activity midwives for their help in recruiting them, and the Avon exposure. Longitudinal Study of Parents and Children (ALSPAC) Table S3. Unadjusted associations between physical team, including interviewers, computer and laboratory activity during pregnancy and offspring birth size. technicians, clerical workers, research scientists, volunteers, Table S4. Associations between physical activity during managers, receptionists, and nurses. The UK Medical pregnancy and % body fat. Research Council and Wellcome (grant ref: 102215/2/13/2) Table S5. Associations between late pregnancy physical and the University of Bristol provide core support for activity and offspring birth size with additional adjustments ALSPAC. Danish National Birth Cohort (DNBC, Den- for maternal early pregnancy BMI and GDM. mark): the Danish National Research Foundation has established the Danish Epidemiology of Science Centre that initiated and created the DNBC. The cohort is fur- References thermore a result of a major grant from this foundation. 1 Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, Additional support for the DNBC is obtained from the et al. Global, regional, and national prevalence of overweight and Pharmacy Foundation, the Egmont Foundation, the March obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet of Dimes Birth Defect Foundation, the Augustinus Founda- 2014;384:766–81. tion, and the Health Foundation. Groningen Expert Center 2 Boney CM, Verma A, Tucker R, Vohr BR. Metabolic syndrome in for Kids with Obesity (GECKO)-Drenthe: the GECKO Dren- childhood: association with birth weight, maternal obesity, and the birth cohort was funded by an unrestricted grant of gestational diabetes mellitus. Pediatrics 2005;115:e290–6. Hutchison Whampoa Ltd, Hong Kong, and supported by 3 Gu s, An X, Fang L, Zhang X, Zhang C, Wang J, et al. Risk factors and long-term health consequences of macrosomia: a prospective the University of Groningen, Well Baby Clinic Foundation study in Jiangsu Province, China. J Biomed Res 2012;26: Icare, Noordlease, and Youth Health Care Drenthe. Healthy 235–40. Start Study (HSS) was funded by the following NIH funding 4 Amercian College of Obstetricians and Gynaecolgist. Committee sources: R01DK076645, UL1TR00108. Polish Mother and opinion no. 650: physical activity and exercise during pregnancy and Child Cohort (REPRO_PL) is supported in part by funds the postpartum period. Obstet Gynecol 2015;126:e135–42. 5 Clapp JF III, Capeless EL. Neonatal morphometrics after endurance from National Centre for Research and Development, exercise during pregnancy. Am J Obstet Gynecol 1990;1:1805–11. Poland (grant no. PBZ-MEiN-/8/2/2006; contract no. K140/ 6 Morgan KL, Rahman MA, Hill RA, Zhou SM, Bijsma G, Khanom A, P01/2007/1.3.1.1) and grant PNRF-218-AI-1/07 from Nor- et al. Physical activity and excess weight in pregnancy have way through the Norwegian Financial Mechanism within the independent and unique effects on delivery and perinatal outcomes. Polish–Norwegian Research Fund. ROLO study was funded PLoS One 2014;9:e94532. 7 Juhl M, Olsen J, Andersen PK, Nøhr EA, Andersen AM. Physical by the Health Research Board of Ireland, with additional exercise during pregnancy and fetal growth measures: a study financial support from the National Maternity Hospital within the Danish National Birth Cohort. Am J Obstet Gynecol Medical Fund. Southampton Women’s Survey (SWS) was 2010;202:63.e1–8. supported by grants from the Medical Research Council, 8 Owe KM, Nystad W, Bø K. Association between regular exercise National Institute for Health Research Southampton and excessive newborn birth weight. Obstet Gynecol 2009;114:770–6. Biomedical Research Centre, University of Southampton 9 Currie LM, Woolcott CG, Fell DB, Armson BA, Dodds L. The and University Hospital Southampton National Health association between physical activity and maternal and neonatal Service Foundation Trust, and the European Union’s outcomes: a prospective cohort. Matern Child Health J Seventh Framework Programme (FP7/2007-2013), project 2014;18:1823–30. EarlyNutrition (grant 289346). SB, KW, NW and KO are 10 Alderman BW, Zhao H, Holt VL, Watts DH, Beresford SA. Maternal physical activity in pregnancy and infant size for gestational age. supported by the Medical Research Council (Unit Pro- Ann Epidemiol 1998;8:513–9. gramme numbers: MC_UU_12015/1, MC_UU_12015/2 and 11 Clapp JF III, Capeless EL. The changing glycemic response to exercise MC_UU_12015/3). during pregnancy. Am J Obstet Gynecol 1991;1:1678Y83. 12 van Poppel MN, Oostdam N, Eekhoff ME, Wouters MG, van Mechelen W, Catalano PM. Longitudinal relationship of physical Supporting Information activity with insulin sensitivity in overweight and obese pregnant women. J Clin Endocrinol Metab 2013;98:2929–35. Additional supporting information may be found online in 13 Jackson MR, Gott P, Lye SJ, Ritchie JW, Clapp JF III. The effects of the Supporting Information section at the end of the maternal aerobic exercise on human placental development: article. placental volumetric composition and surface areas. Placenta Figure S1. Schematic representation of the analysis plan. 1995;16:179–91. 468 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists Pregnancy physical activity and birth size outcomes 14 Wiebe HW, Boule NG, Chari R, Davenport MH. The effect of 34 van Eijsden M, Vrijkotte TG, Gemke RJ, van der Wal MF. Cohort supervised prenatal exercise on fetal growth: a meta-analysis. Obstet profile: the Amsterdam Born Children and their Development Gynecol 2015;125:1185–94. (ABCD) study. Int J Epidemiol 2011;40:1176–86. 15 Sanabria-Martınez G, Garcıa-Hermoso A, Poyatos-Leon  R, Gonzalez- 35 Olsen J, Melbye M, Olsen SF, Sørensen TI, Aaby P, Andersen AM, Garcıa A, Sanchez-Lopez  M, Martınez-Vizcaıno V. Effects of et al. The Danish National Birth Cohort – its background, structure exercise-based interventions on neonatal outcomes: a meta-analysis and aim. Scand J Public Health 2001;29:300–7. of randomized controlled trials. Am J Health Promot 2015;30:214–23. 36 L’Abee C, Sauer PJJ, Damen M, Rake JP, Cats H, Stolk RP. Cohort 16 Schlussel € MM, Souza EB, Reichenheim ME, Kac G. Physical activity profile: the GECKO Drenthe study; overweight programming during during pregnancy and maternal-child health outcomes: a systematic early childhood. Int J Epidemiol 2008;37:486–9. literature review. Cad Saude Publica 2008;24(Suppl 4):s531–44. 37 Polanska K, Hanke W, Gromadzinska J, Llgocka D, Gulczynska E, 17 Bisson M, Lavoie-Guenette J, Tremblay A, Marc I. Physical activity Sobala W, et al. Polish mother and child cohort study – defining the volumes during pregnancy: a systematic review and meta-analysis of problem, the aim of the study and methodological assumptions. Int observational studies assessing the association with infant’s birth J Occup Med Environ Health 2009;22:383–91. weight. AJP Rep 2016;6:e170–97. 38 Walsh JM, McGowan CA, Mahony R, Foley ME, McAuliffe FM. Low 18 Harrod CS, Chasan-Taber L, Reynolds RM, Fingerlin TE, Glueck DH, glycaemic index diet in pregnancy to prevent macrosomia (ROLO Brinton JT, et al. Physical activity in pregnancy and neonatal body study): randomised control trial. BMJ 2012;345:e5605. composition: the Healthy Start study. Obstet Gynecol 2014;124:257–64. 39 Inskip HM, Godfrey KM, Robinson SM, Law CM, Barker DJ, Cooper 19 Perkins CC, Pivarnik JM, Paneth N, Stein AD. Physical activity and C. Cohort profile: the Southampton Women’s Survey. Int J fetal growth during pregnancy. Obstet Gynecol 2007;109:81–7. Epidemiol 2006;35:42–8. 20 Leiferman JA, Evenson KR. The effect of regular leisure physical 40 Ainsworth BE, Haskell WL, Herrmann SD, Meckes N, Bassett DR, Tudor- activity on birth outcomes. Matern Child Health J 2003;7:59–64. Locke C, et al. Compendium of physical activities: a second update of 21 Nieuwenhuijsen MJ, Northstone K, Golding J, ALSPAC Study Team. codes and MET values. Med Sci Sports Exerc 2011;43:1575–81. st Swimming and birth weight. Epidemiology 2002;13:725–8. 41 Intergrowth 21 : the international fetal and newborn growth 22 Hatch MC, Shu XO, McLean DE, Levin B, Begg M, Reuss L, et al. consortium [www.intergrowth21.org.uk]. Accessed 03 July 2017. Maternal exercise during pregnancy, physical fitness, and fetal 42 Slaughter MH, Lohman TG, Boileau RA, Horswill CA, Stillman RJ, growth. Am J Epidemiol 1993;137:1105–14. Van Loan MD, et al. Skinfold equations for estimation of body 23 Hegaard HK, Petersson K, Hedegaard M, Ottesen B, Dykes AK, fatness in children and youth. Hum Biol 1988;60:709–23. Henriksen TB, et al. Sports and leisure-time physical activity in 43 DataSHIELD: a software solution for secure bioscience collaboration pregnancy and birth weight: a population-based study. Scand J Med [http://www.datashield.ac.uk]. Accessed 03 July 2017. Sci Sports 2010;20:e96–102. 44 Durante R, Ainsworth BE. The recall of physical activity: using a 24 Duncombe D, Skouteris H, Wertheim EH, Kelly L, Fraser V, Paxton cognitive model of the question-answering process. Med Sci Sports SJ. Vigorous exercise and birth outcomes in a sample of recreational Exerc 1996;28:1282–91. exercisers: a prospective study across pregnancy. Aust N Z J Obstet 45 Evenson KR, Barakat R, Brown WJ. Guidelines for physical activity Gynaecol 2006;46:288–92. during pregnancy: comparisons from around the world. Am J 25 Klebanoff MA, Shiono PH, Carey JC. The effect of physical activity Lifestyle Med 2014;8:102–21. during pregnancy on preterm delivery and birth weight. Am J 46 Aune D, Sen A, Henriksen T, Saugstad OD, Tonstad S. Physical Obstet Gynecol 1990;163:1450–6. activity and the risk of gestational diabetes mellitus: a systematic 26 Magann EF, Evans SF, Weitz B, Newnham J. Antepartum, review and dose-response meta-analysis of epidemiological studies. intrapartum, and neonatal significance of exercise on healthy, low- Eur J Epidemiol 2016;31:967–97. risk pregnant women. Obstet Gynecol 2002;99:466–72. 47 Gill JM, Celis-Morales CA, Ghouri N. Physical activity, ethnicity and 27 Jukic AM, Evenson KR, Daniels JL, Herring AH, Wilcox AJ, Hartmann cardio-metabolic health: does one size fit all? Atherosclerosis KE. A prospective study of the association between vigorous 2014;232:319–33. physical activity during pregnancy and length of gestation and 48 Konstantina D, Triantafyllou A, GrigoriadouI I, Kintiraki E, birthweight. Matern Child Health J 2012;16:1031–44. Triantafyllou GE, Poulios P, et al. Impairments in microvascular 28 Voldner N, Frøslie KF, Bo K, Haakstad L, Hoff C, Godang K, et al. function and skeletal muscle oxygenation in women with gestational Modifiable determinants of fetal macrosomia: role of lifestyle-related diabetes mellitus: links to cardiovascular disease risk factors. factors. Acta Obstet Gynecol Scand 2008;87:423–9. Diabetologia 2017;60:192–201. 29 Hopkins SA, Cutfield WS. Exercise in pregnancy: weighing up the 49 Helmerhorst HJF, Bragen S, Warren J, Besson H, Ekelund U. A long-term impact on the next generation. Exerc Sport Sci Rev systematic review of reliability and objective criterion-related validity 2011;39:120–7. of physical activity questionnaires. Int J Behav Nutr Phys Act 2012; 30 InterConnect: global data for diabetes and obesity research [www. 9:103. interconnect-diabetes.eu]. Accessed 03 July 2017. 50 Wanner M, Probst-Hensch N, Kriemler S, Meier F, Autenrieth C, 31 Jones EM, Sheehan NA, Masca N, Wallace SE, Murtagh MJ, Burton Martin BW. Validation of the long international physical activity PR. DataSHIELD – shared individual-level analysis without sharing the questionnaire: influence of age and language region. Prev Med Rep data: a biostatistical perspective. Nor Epidemiol 2012;21:231–9. 2016;9:250–6. 32 Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson J, 51 Clapp JF, Kim H, Burciu B, Schmidt S, Petry K, Lopez B. Continuing et al. Cohort profile: the ׳children of the 90s׳ – the index offspring regular exercise during pregnancy: effect of exercise volume on of the Avon Longitudinal Study of Parents and Children. Int J fetoplacental growth. Am J Obstet Gynecol 2002;186:142–7. Epidemiol 2013;58:111–27. 52 Mudd LM, Pivarnik J, Holzman CB, Paneth N, Pfeiffer K, Chung H. 33 Fraser A, Macdonald-Wallis C, Tilling K, Boyd A, Golding J, Davey-Smith Leisure-time physical activity in pregnancy and the birth G, et al. Cohort profile: the Avon Longitudinal Study of Parents and weight distribution: where is the effect? J Phys Act Health 2012;9: Children: ALSPAC mothers cohort. Int J Epidemiol 2013;42:97–110. 1168–77. ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of 469 Royal College of Obstetricians and Gynaecologists Pastorino et al. 53 Rose NC, Haddow JE, Palomaki GE, Knight GJ. Self-rated physical 55 McMurray RG, Hackney AC, Guion WK, Katz VL. Metabolic and activity level during the second trimester and pregnancy outcome. hormonal responses to low-impact aerobic dance during pregnancy. Obstet Gynecol 1991;78:1078–80. Med Sci Sports Exerc 1996;28:41–6. 54 Sternfeld B, Quesenberry CP Jr, Eskenazi B, Newman LA. Exercise 56 Bonen A, Campagna P, Gilchrist L, Young DC, Beresford P. during pregnancy and pregnancy outcome. Med Sci Sports Exerc Substrate and endocrine responses during exercise at selected stages 1995;27:634–40. of pregnancy. J Appl Physiol 1992;73:134–42. 470 ª 2018 The Authors. BJOG: An International Journal of Obstetrics and Gynaecology published by John Wiley & Sons Ltd on behalf of Royal College of Obstetricians and Gynaecologists

Journal

BJOG An International Journal of Obstetrics & Gynaecology – Pubmed Central

Published: Oct 22, 2018

There are no references for this article.