Abstract
INTRODUCTION: Pregnant individuals who smoke face increased health risks because smoking harms both the mother and their developing offspring.</p>
AIMS AND METHODS: Using 307 417 Europeans from the UK Biobank, we examined whether exposure to maternal smoking during pregnancy (MSP) interacts with genetic risk to predict offspring birth weight (BW) and smoking behaviors. We investigated interactions between MSP and genetic risk at multiple levels: single variant, gene level, and polygenic score. We examined self-reported BW, smoking initiation status (SI), age of smoking initiation, cigarettes per day, and smoking cessation status.</p>
RESULTS: One locus tagged by single-nucleotide polymorphism rs72689499 on chromosome 14 reached significance for interaction with MSP on the multiplicative (log10) scale for BW (p = 5.13 × 10-9). In gene-level testing, three genes on chromosome 1 and one gene on chromosome 14 reached significance for interaction with MSP on both the additive and multiplicative scale for BW. These genes include PTCH2, EIF2B3, PLK3, and TSHR. Single-nucleotide polymorphism and gene-level results were insignificant for all offspring smoking behaviors. We also detected an interaction between polygenic risk for smoking and MSP on SI on both the additive (p = 4.4 × 10-5) and multiplicative (p = 1.0 × 10-5) scale. We found evidence of gene-environment correlation in the polygenic risk analysis using a post hoc t test which showed that MSP-exposed offspring had a higher SI polygenic risk scores than those unexposed to MSP (p = 5.9 × 10-623).</p>
CONCLUSIONS: Our results support the main effect of MSP on BW and show a genetic interaction between MSP and genetic factors influencing BW.</p>
IMPLICATIONS: We detected interactions between maternal smoking and genetic factors to influence birth weight; these interactions were detectable at both the single-nucleotide polymorphism and gene levels. Many of the genes detected to interact with maternal smoking to influence birth weight have other reported associations with height or smoking-related traits. For smoking initiation, we detected a negative interaction between maternal smoking and polygenic risk, as well as evidence of gene-environment correlation.</p>