Association of adverse pregnancy outcomes with self-reported measures of sleep duration and timing in women who are nulliparous

Study Objectives: To examine the relationship of self-reported sleep during pregnancy with adverse pregnancy outcomes. A secondary objective was to describe the concordance between self-reported and objectively assessed sleep during pregnancy. Methods: In this prospective cohort, women completed a survey of sleep patterns at 6 to 13 weeks' gestation (visit 1) and again at 22 to 29 weeks' gestation (visit 3). Additionally, at 16 to 21 weeks (visit 2), a subgroup completed a week-long sleep diary coincident with an actigraphy recording. Weekly averages of self-reported sleep duration and sleep midpoint were calculated. A priori, sleep duration < 7 hours was defined as "short," and sleep midpoint after 5:00 am was defined as "late." The relationship of these sleep abnormalities with hypertensive disorders of pregnancy (HDP) and gestational diabetes mellitus (GDM) was determined. Results: Of the 10,038 women enrolled, sleep survey data were available for 7, 524 women at visit 1 and 7, 668 women at visit 3. A total of 752 women also provided ≥ 5 days of sleep diary data coincident with actigraphy at visit 2. We did not observe any consistent relationship between self-reported short sleep and HDP or GDM. There was an association between self-reported late sleep midpoint and GDM (visit 1 adjusted odds ratio 1.67, 95% confidence interval 1.17, 2.38; visit 2 adjusted odds ratio 1.73, 95% confidence interval 1.23, 2.43). At visit 2, 77.1% of participants had concordance between their diary and actigraphy for short sleep duration, whereas 94.3% were concordant for sleep midpoint. Conclusions: Self-reported sleep midpoint, which is more accurate than self-reported sleep duration, is associated with the risk of GDM.