Search

International sharing of cohort data for research is important and challenging. We explored the feasibility of multicohort federated analyses by examining associations between 3 pregnancy exposures (maternal education, exposure to green vegetation, and gestational diabetes) and offspring body mass index (BMI) from infancy to age 17 years. We used data from 18 cohorts (n = 206,180 mother-child pairs) from the EU Child Cohort Network and derived BMI at ages 0–1, 2–3, 4–7, 8–13, and 14–17 years. Associations were estimated using linear regression via 1-stage individual participant data meta-analysis using DataSHIELD. Associations between lower maternal education and higher child BMI emerged from age 4 and increased with age (difference in BMI z score comparing low with high education, at age 2–3 years = 0.03 (95% confidence interval (CI): 0.00, 0.05), at 4–7 years = 0.16 (95% CI: 0.14, 0.17), and at 8–13 years = 0.24 (95% CI: 0.22, 0.26)). Gestational diabetes was positively associated with BMI from age 8 years (BMI z score difference = 0.18, 95% CI: 0.12, 0.25) but not at younger ages; however, associations attenuated towards the null when restricted to cohorts that measured gestational diabetes via universal screening. Exposure to green vegetation was weakly associated with higher BMI up to age 1 year but not at older ages. Opportunities of cross-cohort federated analyses are discussed., International sharing of cohort data for research is important and challenging. We explored the feasibility of multicohort federated analyses by examining associations between 3 pregnancy exposures (maternal education, exposure to green vegetation, and gestational diabetes) and offspring body mass index (BMI) from infancy to age 17 years. We used data from 18 cohorts (n = 206,180 mother-child pairs) from the EU Child Cohort Network and derived BMI at ages 0-1, 2-3, 4-7, 8-13, and 14-17 years. Associations were estimated using linear regression via 1-stage individual participant data meta-analysis using DataSHIELD. Associations between lower maternal education and higher child BMI emerged from age 4 and increased with age (difference in BMI z score comparing low with high education, at age 2-3 years = 0.03 (95% confidence interval (CI): 0.00, 0.05), at 4-7 years = 0.16 (95% CI: 0.14, 0.17), and at 8-13 years = 0.24 (95% CI: 0.22, 0.26)). Gestational diabetes was positively associated with BMI from age 8 years (BMI z score difference = 0.18, 95% CI: 0.12, 0.25) but not at younger ages; however, associations attenuated towards the null when restricted to cohorts that measured gestational diabetes via universal screening. Exposure to green vegetation was weakly associated with higher BMI up to age 1 year but not at older ages. Opportunities of cross-cohort federated analyses are discussed.

Introduction: High prevalence of overweight and obesity already observed in preschool children suggests the involvement of early-life risk factors. Preconception period and pregnancy are crucial windows for the implementation of child obesity prevention interventions with parental lifestyle factors as relevant targets. So far, most studies have evaluated their role separately, with only a few having investigated their potential synergistic effect on childhood obesity. Our objective was to investigate parental lifestyle patterns in the preconception and pregnancy periods and their association with the risk of child overweight after 5 years. Materials and methods: We harmonized and interpreted results from four European mother-offspring cohorts participating in the EndObesity Consortium [EDEN, France; Elfe, France; Lifeways, Ireland; and Generation R, Netherlands] with data available for 1,900, 18,000, 1,100, and 9,500 families, respectively. Lifestyle factors were collected using questionnaires and included parental smoking, body mass index (BMI), gestational weight gain, diet, physical activity, and sedentary behavior. We applied principal component analyses to identify parental lifestyle patterns in preconception and pregnancy. Their association with risk of overweight (including obesity; OW-OB) and BMI z-scores between 5 and 12 years were assessed using cohort-specific multivariable logistic and linear and regression models (adjusted for potential confounders including parental age, education level, employment status, geographic origin, parity, and household income). Results: Among the various lifestyle patterns derived in all cohorts, the two explaining the most variance were characterized by (1) “high parental smoking, low maternal diet quality (and high maternal sedentary behavior in some cohorts)” and, (2) “high parental BMI and low gestational weight gain.” Patterns characterized by high parental BMI, smoking, low diet quality or high sedentary lifestyle b

BACKGROUND: A high prevalence of excess weight in children younger than 5 years suggests the involvement of early-life risk factors. The preconception and pregnancy periods are crucial stages for the implementation of interventions to prevent childhood obesity. Most studies so far have evaluated the effects of early-life factors separately, with only a few investigating the combined effect of parental lifestyle factors. Our objective was to fill the literature gap regarding parental lifestyle factors in the preconception and pregnancy periods and to study their association with the risk of overweight in children after the age of 5 years. METHODS: We harmonised and interpreted data from four European mother-offspring cohorts (EDEN [comprising 1900 families], Elfe [comprising 18 000 families], Lifeways [comprising 1100 families], and Generation R [comprising 9500 families]). Written informed consent was obtained from parents of all involved children. Lifestyle factor data collected through questionnaires comprised parental smoking, BMI, gestational weight gain, diet, physical activity, and sedentary behaviour. We applied principal component analyses to identify multiple lifestyle patterns in preconception and pregnancy. Their association with child BMI z-score and risk of overweight (including obesity, overweight and obesity, as defined by the International Task Force reference) between the ages of 5 and 12 years were assessed using cohort-specific multivariable linear and logistic regression models (adjusted for confounders including parental age, education level, employment status, geographic origin, parity, and household income). FINDINGS: Among the various lifestyle patterns identified in all cohorts, the two that better explained variance were high parental smoking plus low maternal diet quality or high maternal sedentary behaviour, and high parental BMI plus low gestational weight gain. Overall, we observed that patterns characterised by high parental BMI, sm