Sex-Specific Effects of Birth Weight on Longitudinal Behavioral Outcomes: A Mendelian Randomization Approach Using Polygenic Scores

Abstract

Background: It is unclear whether sex differences in behavior arising from birth weight (BW) are genuine because of the cross-sectional nature and potential confounding in previous studies. We aimed to test whether sex differences associated with BW phenotype were reproducible using a Mendelian randomization approach, i.e., association between polygenic score (PGS) for BW and behavior outcomes across childhood and adolescence.

Methods: Using data from the Raine Study, we had 1484 genotyped participants with a total of 6446 Child Behavior Checklist assessments from ages 5 to 17 years. We used BW-PGSs in linear mixed-effect models to predict parentally assessed attention, aggression, and social problems scales; we also derived estimates and significance for a sex-by-genotype interaction. We used a Bonferroni-corrected significance threshold and tested robustness of the results with teacher assessments of behavior and a second PGS.

Results: We found a sex-by-genotype interaction with lower BW-PGSs associated with increased aggression in males compared with females. These findings were consistent across various analyses, including teacher assessments. Surprisingly, a lower BW-PGS showed protective effects in females, while a lower BW phenotype had detrimental effects in males with evidence of a genotype-phenotype mismatch increasing aggression problems in males only.

Conclusions: This study underscores the genuine nature of behavioral sex differences arising from low BW and highlights the sex-dependent and diverging effects of environmental and genetic BW determinants.

Keywords: Aggression; Birth weight; Child Behavior Checklist; Children and adolescents; Developmental programming; Gene-environment interactions.

Figure 1

Flow diagram of cohort selection. CBCL, Child Behavior Checklist.

Figure 2

Directed acyclic graph and birth weight polygenic score (BW-PGS). Conceptual framework for the primary analysis demonstrated in a directed acyclic graph (A). Density plot and histogram of the distribution of our BW-PGS in males and females (B) and the effect of the PGS on measured birth weight in males and females (C). CBCL, Child Behavior Checklist.

Figure 3

Graphical illustration of the sex interaction on behavioral effects of birth weight genetics. The linear relationship between the normalized polygenic score and attention (A–C), aggression (D–F), and social problems (G–I) is shown for ages 5, 10, and 17 years in males and females. Note that the y-axis changes depending on the Child Behavior Checklist (CBCL) syndrome and the age at assessment.

Figure 4

Birth weight (BW) phenotype-genotype interactions for aggression. In phenotypically smaller males (first BW quintile), a higher BW genotype increases Child Behavior Checklist (CBCL) aggression problem raw scores, whereas in phenotypically larger males (fifth BW quintile), a lower BW genotype increases aggression scores. This changing effect was statistically significant in males (A–C) but not in females (D–F). Because of repeated measures, no confidence interval is given. PGS, polygenic score.