Sleep Trajectories Among Pregnant Women and the Impact on Outcomes: A Population-Based Cohort Study

Objectives Sleep problems and deprivation are common during pregnancy, particularly in the third trimester. Previous studies are mostly descriptive or focused on specific clinical groups and late pregnancy. We aimed to identify sleep duration trajectories during the pregnancy period, their associated factors, and impact on pregnancy and birth outcomes. Methods We studied 200 women from a mother-child cohort recruited in 2009-2011 from the French general population. We used semi-parametric models to analyze data collected through questionnaires. Results We detected three sleep duration trajectories during pregnancy: short-decreasing (< 6.5h/night, 10.8% of the sample), medium-decreasing (6.5-8h/night, 57.6%), and long-increasing (> 8h/night, 31.6%) trajectories. Factors associated with the short-decreasing trajectory relative to the medium-decreasing trajectory were older age (odds-ratio/year = 1.13 [95%Confidence-Interval 1.00-1.29]) and working > 28 weeks of gestational age (odds-ratio = 0.30 [0.10-0.90]). Sleep duration during pregnancy in this trajectory group was modified by insomniac symptoms (regression coefficient/trimester = -0.74 [Standard-Error 0.12]) and naps (regression coefficient/trimester = 0.58 [0.25]). Restless legs syndrome was the only factor associated with the long-increasing trajectory and decreased sleep duration (regression coefficient/trimester = -0.88 [0.25]). Assisted delivery (i.e. cesarean section and/or instrumental delivery) and post-partum depression were more frequent among women with the short-decreasing and long-increasing trajectories whereas cesarean section alone was more prevalent among those with the short-decreasing trajectory. Proportion of premature births was higher in the short-decreasing trajectory group. Birth-weight-z-score was lower in the long-increasing trajectory group. Conclusion We identified sleep trajectories among pregnant women with specific risk factors that could affect both pregnancy and birth outcomes. Taking these into consideration could improve both maternal and child health.