Abstract Introduction Poor sleep has been associated with adverse pregnancy outcomes (APOs), but prior research has been limited to examination of single sleep dimensions and/or assessment of sleep at a single timepoint. The purpose of this study was to examine the trajectories of three sleep health dimensions (duration, efficiency, quality) across pregnancy and their association with APOs. Methods The Pregnancy 24/7 cohort study followed a sample of 500 pregnant women from three sites (Iowa City, Iowa; Pittsburgh, Pennsylvania; Morgantown, West Virginia). Sleep was assessed at each trimester using actigraphy and validated questionnaires; APOs (hypertensive disorders of pregnancy HDP, preterm birth, gestational diabetes, and small-for-gestational age) were abstracted from medical records and adjudicated by an obstetrician at each site. Group-based trajectories of sleep duration and efficiency (based on ~7 nights of total sleep time and sleep efficiency SE, respectively) and quality (single item from the Pittsburgh Sleep Quality Index) were constructed. Binary logistic regression models evaluated associations between the three sleep health dimension trajectories and APOs (composite of all APOs, HDP), with covariate adjustment for site, age, pre-pregnancy body mass index, socioeconomic status, number of children in the home, and minority status. Results Participants with available sleep and APO data (n=478) were, on average, 30.6±4.5 years old; 87.2% were White race, and 22.6% were rural. APOs occurred in 36.8% of participants, with 18.2% specifically developing HDP. Four trajectory groups were identified for duration (representing short, medium, medium-long, and long sleep duration), four groups for efficiency (representing low, low-moderate, high, and very high SE), and two groups for quality (poor, good). None of the sleep health trajectories were significantly associated with the likelihood of either APO or HDP, with efficiency showing the most prominent (albeit nonsignificant) associations (e.g., for APO: relative to low SE, adjusted odds ratios for low-moderate SE, high SE, and very high SE were 0.64 0.32,1.28, 0.66 0.33,1.32, and 0.66 0.31,1.38). Conclusion Our findings failed to observe an association between trajectories of sleep duration, efficiency, or quality with APOs. Future work will explore the potential influence of additional sleep health dimensions and trimester-specific associations with APOs. Support (if any) R01HL153095
Kline et al. (Fri,) studied this question.