Prenatal exposure to parental smoking and infertility in sons and daughters: a cohort study.

study questionIs prenatal exposure to parental smoking associated with infertility in sons and daughters? SUMMARY ANSWER: The study does not provide robust evidence that prenatal smoking exposure increases the risk of infertility, defined as clinically diagnosed infertility or ART use, in adult sons and daughters. WHAT IS KNOWN ALREADY: Reproductive function is established early in fetal life and may be adversely affected by prenatal exposure to parental smoking. STUDY DESIGN, SIZE, DURATION: In this cohort study, we used data from the Danish birth cohort, Health Habits for Two, established in 1984-1987. In total, 11 980 (87% of eligible) women participated and gave birth to 11 144 liveborn, singleton sons and daughters. PARTICIPANTS/MATERIALS, SETTING,

methodsAdult sons and daughters born to mothers enrolled in the birth cohort constituted the study population and were aged 35-38 years at follow-up. Around gestational week 36, mothers provided information on parental smoking. The primary exposure of interest was maternal smoking. Paternal smoking and combined parental smoking were also assessed to explore individual and potential synergistic effects. Infertility among offspring was defined as either an ICD-10 infertility diagnoses or use of ART, identified within two nationwide registers and considering their partners' records. Associations were examined using Cox regression models estimating crude and adjusted hazard ratios (HRs) for categories of prenatal smoking, with non-smoking as the reference, accounting for competing risks of infertility. Dose-response associations were explored across categories of maternal and paternal smoking. The proportional hazards assumption was met only when stratifying analyses by age, using cutoffs of ≤26 and >26 years for sons, and ≤27 and >27 years for daughters. In a sensitivity analysis, follow-up was restricted to periods of registered partnership under the assumption that individuals in a relationship are more likely to be identified as infertile and vice versa. In a sub-analysis, analyses were stratified by offspring educational level to explore potential social gradients in being identified as infertile in medical records. MAIN RESULTS AND THE ROLE OF CHANCE: Maternal smoking of >9 cigarettes/day tended to be associated with higher risk of infertility among sons ≤26 years (HR 1.7 (95% CI: 0.8-3.5)) and daughters ≤27 years (HR 1.2 (95% CI: 0.7-1.9)), both indicating suggestive dose-response patterns with higher risk of infertility with increasing maternal smoking levels. No associations for sons >26 years or daughters >27 years were observed. Paternal smoking of >10 cigarettes/day tended to be associated with higher risk of infertility among sons >26 years (HR 1.3 (95% CI: 0.9-1.8)), indicating suggestive dose-response patterns. No higher risk of infertility following paternal smoking was observed among sons ≤26 years or daughters at any age. No synergistic effects of combined parental smoking were observed. When redefining the time at risk using the partnership model, we did not observe higher risks of infertility among maternally exposed sons and daughters. When stratifying the main analyses by the sons' and daughters' educational level, we observed tendencies of higher risks for infertility in some exposure groups among those with longer education. Overall, we found no robust evidence that prenatal exposure to parental smoking is associated with a higher risk of clinically recognized infertility in offspring. LIMITATIONS, REASONS FOR CAUTION: Parental smoking was self-reported by mothers late in pregnancy and might be subject to misclassification, most likely non-differential. Not all infertility cases may be captured in the registers as it requires pregnancy attempt and pursuing medical treatment. Most confidence intervals overlapped the null, indicating that the findings are also compatible with no association. Some analyses, particularly the stratified analyses, were limited by low statistical power. WIDER IMPLICATIONS OF THE

findingsThe observed diminishing effect estimates with increasing age for those exposed to maternal smoking may reflect growing influence from other dominant infertility risk factors. Results from analyses stratified by offspring educational level may be indicative of tendencies of greater awareness of fertility issues and higher health care-seeking among individuals from more advantaged socioeconomic backgrounds. A potential higher, although modest and uncertain, risk of infertility following prenatal smoking exposure cannot be excluded, supporting the current guidelines recommending smoking cessation prior to pregnancy. STUDY FUNDING/COMPETING INTEREST(S): This study was funded by the European Union (ERC, BIOSFER, 101071773); expressed views and opinions are however those of the authors' only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them. Additionally, this study received funding from Aarhus University and was supported by the Research Council of Norway, through its Centre of Excellence funding scheme (No. 262700). All authors have no conflicts or interests to declare. TRIAL REGISTRATION NUMBER: N/A.