Associations of sleep characteristics with outcomes of IVF/ICSI treatment: a prospective cohort study

Are sleep characteristics associated with outcomes of IVF/ICSI treatment?Nocturnal sleep <7 h/night and disturbed sleep are related to impaired oocyte and embryo yield, while longer nocturnal sleep might reduce the chance of a successful pregnancy, and the associations between nocturnal sleep duration and IVF/ICSI outcomes are modified by maternal age and subjective sleep quality.Disturbed sleep and circadian rhythm contribute to impaired fecundity in the general population, but the effects of sleep characteristics on IVF/ICSI outcomes are largely unknown.This study was conducted among 1276 women undergoing IVF/ICSI treatment between December 2018 and September 2019 based on the Tongji Reproductive and Environmental cohort. Owing to the limited number of multiple cycles, we included only the outcomes of their first IVF/ICSI cycle in the current analysis.Data on sleep characteristics were collected via questionnaires on the day of oocyte retrieval. IVF/ICSI outcomes were abstracted from medical records. Quasi-Poisson, quasi-binomial or logistic regression models were used to assess the relations between sleep characteristics and reproductive outcomes after adjusting for relevant confounders. We also performed stratified analyses by subjective sleep quality (good versus poor) and maternal age (≤30 versus >30 years).Compared with women who slept 7 to <8 h/night, those who slept <7 h/night exhibited decreases in the number of retrieved and mature oocytes of 11.5% (95% CI: -21.3%, -0.48%) and 11.9% (95% CI: -22.4%, -0.03%), respectively. A mid-sleep time (MST) earlier than 2:21 a.m. (<2:21 a.m.) or later than 3:00 a.m. (≥3:00 a.m.) and poor subjective sleep quality were inversely associated with the fertilization rate. Women who had trouble falling asleep more than three times per week had a lower number of mature oocytes (-10.5%, 95% CI: -18.6%, -1.6%), normal fertilized oocytes (-14.8%, 95% CI: -23.7%, -4.8%) and good-quality embryos (-15.1%, 95% CI: -25.4%, -3.5%) than those who had no such trouble. In addition, women who slept 9 to <10 h/night had a lower chance of clinical pregnancy compared to women who slept 7 to <8 h/night (odds ratio = 0.65, 95% CI: 0.44, 0.98). In the stratified analyses, the positive associations of nocturnal sleep duration with the number of good-quality embryos and fertilization rate existed only among the women with poor subjective sleep quality (P for interaction = 0.02 and 0.03, respectively). Additionally, we found that the positive associations of nocturnal sleep duration with implantation or clinical pregnancy only existed among women aged over 30 years (P for interaction = 0.04 and 0.01, respectively).Sleep characteristics are self-reported, which may lead to misclassification. MST serves as a proxy of chronotype and may be non-differentially misclassified resulting in an underestimate of the association of interest. Measuring sleep characteristics on the day of oocyte retrieval may lead to bias. Chance findings cannot be excluded because of the limited number of women with <7 h or ≥10 h nocturnal sleep and multiple testing. Our results may be biased by other confounders and may not be generalizable to women of other ethnicities.Unhealthy sleep characteristics, including short nocturnal sleep, inappropriate sleep time, poor subjective sleep quality and having trouble falling asleep, may impair oocyte quantity and its potential to mature or be fertilized. Long nocturnal sleep might reduce the chance of clinical pregnancy among infertile females, especially women younger than 30 years old. Prolonged nocturnal sleep duration may be a potential beneficial behavior for improving IVF/ICSI outcomes for women aged over 30 years and women with poor subjective sleep quality, which requires further investigation.This work was supported by the National Natural Science Foundation of China (81771654) and the National Key R&D Program of China (No. 2018YFC1004201). There are no conflicts of interest to declare.N/A.