Birth Weight and Development: Bias or Heterogeneity by Polygenic Risk Factors?

Birth weight has long been known to be a proxy for prenatal nutrition as well as social stress during pregnancy and, in this vein, has important associations with life outcomes ranging from infant mortality to cognition to adult stature. However, since birth weight and such developmental outcomes may share a genetic etiology, here we compare estimates of the effect of birth weight on traits with and without controls for genotype. Moreover, a recent literature has suggested that the effects of birth weight on later outcomes may be moderated by genotype. Namely, a handful of studies have assessed whether alleles putatively associated with neural plasticity interact with birth weight in twin and sibling models and have come to conflicting results. This confusion may result from the fact that these researchers have used only candidate gene markers-that is single alleles or a combination of a few alleles. Candidate gene studies, in turn, have failed to replicate and the general consensus in the genetics of behavior literature is that most published associations are false positives. To improve on this prior work, the present study uses well-validated, replicated polygenic scores that combine information across the entire genome by summarizing the effects of over one million individual alleles. Specifically, we use data from the National Longitudinal Study of Adolescent to Adult Health and the UK Biobank to assess whether the effects birth weight on cognitive functioning, educational attainment, height, body mass index, smoking, and depression are biased by or moderated by each phenotype's respective polygenic score. Results do not show any robustly significant interactions in cross-sectional models or in sibling and/or twin fixed effects models. Implications for future research are discussed.