Maternal caffeine consumption during pregnancy and offspring cord blood DNA methylation: a meta-analysis of epigenome-wide association studies

Abstract

AbstractBackgroundWomen are advised to limit caffeine consumption during pregnancy. To uncover the potential epigenetic effects of intrauterine caffeine exposure, we investigated the association of maternal caffeine consumption during pregnancy with genome-wide DNA methylation in cord blood.MethodsWe meta-analysed results from epigenome-wide methylation studies at individual probes and in differentially methylated region (DMR) analysis across 6 European pregnancy and birth cohorts (ALSPAC, BiB, MoBa, Generation R, INMA, EDEN; total n = 3742). Methylation was assessed with lllumina Infinium 450k or EPIC arrays. Maternal caffeine consumption (mg/day) from coffee, tea and cola was derived from questionnaires between weeks 12 - 22 of pregnancy. We investigated associations of methylation with overall and beverage-specific caffeine intake in models adjusted for maternal education, age, BMI, smoking during pregnancy, parity, cord-blood cell proportions and 20 surrogate variables.ResultsOne CpG site (cg19370043, nearest genePRRX1) was associated with maternal caffeine consumption after FDR adjustment for multiple testing and one CpG sites (cg14591243, nearest geneSTAG1) was associated with maternal cola consumption. We found evidence for 12-22 DMRs for each of the caffeine models but little overlap between DMRs identified for specific caffeinated beverages.ConclusionsIn models adjusted for maternal smoking and other potential confounders, we found little evidence to support an intrauterine effect of caffeine on offspring DNA methylation. Comparing associations across different sources of caffeine provided no evidence for caffeine being the causal agent. It is possible that our study did not have adequate statistical power to detect very small associations between maternal caffeine and offspring DNA methylation.Key messages- This large-scale meta-analysis of epigenomewide association studies across six European cohorts does not support an intrauterine effect of caffeine on offspring cord blood DNA methylation.- Lack of overlap between associations with different caffeinated drinks suggest that any (weak) associations were driven by diverse confounding structures of different caffeinated drinks, rather than caffeine per se.- More research is needed to understand the biological mechanisms driving potential effects of caffeine on offspring health.