Ambient air pollution and childhood obesity from infancy to late childhood: An individual participant data meta-analysis of 10 European birth cohorts

Abstract

Ambient air pollution may contribute to childhood obesity through various mechanisms. However, few longitudinal studies examined the relationship between pre- and postnatal exposure to air pollution and obesity outcomes in childhood. We aimed to investigate the association between pre- and postnatal exposure to air pollution and body mass index (BMI) and the risk of overweight/obesity throughout childhood in European cohorts. This study included mother-child pairs from 10 European birth cohorts (n = 37111 (prenatal), 33860 (postnatal)). Exposure to nitrogen dioxide (NO₂) and fine particulate matter with aerodynamic diameter < 2.5 µm (PM_2.5) was estimated at the home addresses during pre- and postnatal periods (year prior outcome assessment). BMI z-scores (continuous) and overweight/obesity status (categorical: zBMI≥+2 (<5 years) or ≥+1 (≥5 years) standard deviations) were derived at 0-2, 2-5, 5-9, 9-12 years. Associations between air pollution exposure and zBMI were estimated separately for each pollutant and cohort using linear and logistic longitudinal mixed effects models, followed by a random-effects meta-analysis. The overweight/obesity prevalence ranged from 12.3-40.5 % between cohorts at 0-2 years, 16.7-35.3 % at 2-5 years, 12.5-40.7 % at 5-9 years, and 10.7-43.8 % at 9-12 years. Results showed no robust associations between NO₂ exposure and zBMI or overweight/obesity risk. Exposure to PM_2.5 during pregnancy was associated with 23 % (95%CI 1.05;1.37) higher overweight/obesity risk across childhood, and higher zBMI and overweight/obesity risk at 9-12 years. Heterogeneity between cohorts was considerable (I²:25-89 %), with some cohort-specific associations; e.g., pre- and postnatal exposure to PM_2.5 was associated with lower zBMI across age periods in UK cohorts (ALSPAC and BiB), while postnatal exposure to PM_2.5 and NO₂ was associated with higher zBMI in one Dutch cohort (Generation R). Overall, this large-scale meta-analysis suggests that prenatal PM_2.5 exposure may be associated with adverse childhood obesity outcomes, but provides no evidence to support an effect of postnatal air pollution exposure, although cohort-specific associations were observed.

Keywords: Air pollution; Child; Meta-analysis; Pediatric obesity; Urban.

Graphical abstract

Fig. 1

Map of the 10 included cohorts from the EU Child Cohort Network.

Fig. 2

Forest plot of associations between pre- and postnatal exposure^a to NO₂ and childhood zBMI and overweight/obesity using two-stage meta-analysis. ABCD: Amsterdam Born Children and their Development; ALSPAC: Avon Longitudinal Study of Parents and Children; BiB: Born in Bradford; DNBC: Danish National Birth Cohort; EDEN: Etude des Déterminants pré et postnatals précoces du développement et de la santé de l’Enfant; Gen R: Generation R Study; INMA: Infancia y Medio Ambiente; MoBa: Norwegian Mother, Father and Child Cohort Study; NINFEA: Nascita e INFanzia: gli Effetti dell’Ambiente; RHEA: Rhea Mother & Child Cohort Study; zBMI: Body mass index z-score; OR: Odds ratio; CI: Confidence intervals. Prenatal exposure models adjusted for sex, child age (in months), individual-level SES (maternal education), pre-pregnancy BMI, smoking during pregnancy, and area-level SES. Postnatal exposure models were further adjusted for prenatal air pollution exposure. Further adjustments: INMA, NINFEA and EDEN: for city, in Gen R, ABCD and BiB: for ethnicity. I²: Test for heterogeneity (%). a Values are described as weighted average postnatal air pollution levels in the year before zBMI measurement, in the different age periods.

Fig. 3

Forest plot of associations between pre- and postnatal exposure^a to PM_2.5 and childhood zBMI and overweight/obesity using two-stage meta-analysis. ABCD: Amsterdam Born Children and their Development; ALSPAC: Avon Longitudinal Study of Parents and Children; BiB: Born in Bradford; DNBC: Danish National Birth Cohort; EDEN: Etude des Déterminants pré et postnatals précoces du développement et de la santé de l’Enfant; Gen R: Generation R Study; INMA: Infancia y Medio Ambiente; MoBa: Norwegian Mother, Father and Child Cohort Study; NINFEA: Nascita e INFanzia: gli Effetti dell’Ambiente; RHEA: Rhea Mother & Child Cohort Study; zBMI: Body mass index z-score; OR: Odds ratio; CI: Confidence intervals. Prenatal exposure models adjusted for sex, child age (in months), individual-level SES (maternal education), pre-pregnancy BMI, smoking during pregnancy, and area-level SES. Postnatal exposure models were further adjusted for prenatal air pollution exposure. Further adjustments: In INMA, NINFEA and EDEN: for city, in Gen R, ABCD and BiB: for ethnicity. I²: Test for heterogeneity (%). ^a Values are described as weighted average postnatal air pollution levels in the year before zBMI measurement, in the different age periods.

Fig. 4

Pre- and postnatal air pollution exposure^a and childhood overall zBMI and overweight/obesity using two-stage meta-analysis, stratified by age period. BMI: Body mass index; OR: Odds ratio; CI: Confidence intervals. Prenatal exposure models are adjusted for sex, child age (in months), individual-level SES (maternal education), pre-pregnancy BMI, smoking during pregnancy, and area-level SES. Postnatal exposure models were further adjusted for prenatal air pollution exposure. Further adjustments: In INMA, NINFEA and EDEN: for city, in Gen R, ABCD and BiB: for ethnicity. I²: Test for heterogeneity (%). ^a Values are described as weighted average postnatal air pollution levels in the year before zBMI measurement, in the different age periods.

Fig. 4

Pre- and postnatal air pollution exposure^a and childhood overall zBMI and overweight/obesity using two-stage meta-analysis, stratified by age period. BMI: Body mass index; OR: Odds ratio; CI: Confidence intervals. Prenatal exposure models are adjusted for sex, child age (in months), individual-level SES (maternal education), pre-pregnancy BMI, smoking during pregnancy, and area-level SES. Postnatal exposure models were further adjusted for prenatal air pollution exposure. Further adjustments: In INMA, NINFEA and EDEN: for city, in Gen R, ABCD and BiB: for ethnicity. I²: Test for heterogeneity (%). ^a Values are described as weighted average postnatal air pollution levels in the year before zBMI measurement, in the different age periods.