Mai Hougaard Christensen, Ellen Margrethe Mikkelsen, Lauren Anne Wise, Elizabeth Elliott Hatch, Anne Sofie Dam Laursen
We estimated the effects of sleep duration, trouble sleeping, and job time (day, evening, shift, and night work) on fecundability, the per-cycle probability of conception. We analysed data from a preconception cohort study of 10 475 Danish females aged 18-45 years attempting to conceive. On baseline questionnaires, females reported on sleep duration (hours/day), trouble sleeping (never, sometimes, approx. 50% of the time, most of the time), and job time. Pregnancy status was ascertained from follow-up questionnaires completed every 8 weeks for up to 12 months. We used proportional probabilities regression models to estimate fecundability ratios (FRs), adjusted for potential confounders. Compared with 8 h/day of sleep, FRs were 0.94 for ≤6 h/day (95% CI, 0.83-1.08), 0.97 for 7 h/day (95% CI, 0.89-1.06), and 0.96 for ≥9 h/day (95% CI, 0.81-1.13). Compared with no trouble sleeping, FRs were 1.00 (95% CI, 0.94-1.06) for trouble sleeping sometimes, 0.94 (95% CI, 0.86-1.03) for trouble sleeping approx. 50% of the time, and 0.95 (95% CI, 0.82-1.10) for trouble sleeping most of the time. Relative to day work, FRs were 1.12 for shift work (95% CI, 1.04-1.20), 0.77 for night work (95% CI, 0.49-1.22) and 1.10 for evening work (95% CI, 0.91-1.33). Self-reported sleep duration and trouble sleeping were not appreciably associated with fecundability. Shift work was associated with a slightly higher fecundability, while night work was associated with slightly lower fecundability, although associations were imprecise. Potential limitations include misclassification of sleep and residual or unmeasured confounding.
