Do Genetic Variants Modify the Effect of Smoking on Risk of Preeclampsia in Pregnancy?

Objective Maternal smoking is associated with as much as a 50% reduced risk of preeclampsia, despite increasing risk of other poor pregnancy outcomes that often co-occur with preeclampsia, such as preterm birth and fetal growth restriction. Researchers have long sought to understand whether this perplexing association is biologically based, or a result of noncausal mechanisms. We examined whether smoking-response genes modify the smoking-preeclampsia association to investigate potential biological explanations. Study Design We conducted a nested case-control study within the Norwegian Mother, Father and Child Birth Cohort (1999-2008) of 2,596 mother-child dyads. We used family-based log-linear Poisson regression to examine modification of the maternal smoking-preeclampsia relationship by maternal and fetal single nucleotide polymorphisms involved in cellular processes related to components of cigarette smoke ( n = 1,915 with minor allele frequency >= 10%). We further investigated the influence of smoking cessation during pregnancy. Results Three polymorphisms showed overall ( p < 0.001) multiplicative interaction between smoking and maternal genotype. For rs3765692 ( TP73 ) and rs10770343 ( PIK3C2G ), protection associated with smoking was reduced with two maternal copies of the risk allele and was stronger in continuers than quitters (interaction p = 0.02 for both loci, based on testing 3-level smoking by 3-level genotype). For rs2278361 ( APAF1 ) the inverse smoking-preeclampsia association was eliminated by the presence of a single risk allele, and again the trend was stronger in continuers than in quitters (interaction p = 0.01). Conclusion Evidence for gene-smoking interaction was limited, but differences by smoking cessation warrant further investigation. We demonstrate the potential utility of expanded dyad methods and gene-environment interaction analyses for outcomes with complex relationships between maternal and fetal genotypes and exposures.