Comparison of longitudinal PSQI and actigraphy-assessed sleep from pregnancy to postpartum

Abstract

Despite the importance of sleep for perinatal health, there is limited research examining whether different measurement modalities may yield inconsistent data from pregnancy through postpartum. We aimed to: 1) describe how sleep patterns change across pregnancy and postpartum using self-report PSQI and actigraphy measures and 2) determine the level of correspondence between these two measurement modalities. Pregnant women from the Stress and Health in Pregnancy and Postpartum (SHIPP) study completed visits during the 3rd trimester, 4–6 weeks postpartum, and 4 months, 8 months, and 12 months postpartum. At each study visit, participants completed questionnaires and wore wrist-actigraphy (Actiwatch 2) for one week prior to each visit. Self-reported global sleep quality was measured with the Pittsburgh Sleep Quality Index (PSQI). Actigraphy and self-reported PSQI sleep characteristics were summarized at each of the five study visits. Using generalized linear mixed modeling, we examined if there were differences by sleep measurement (actigraphy vs. PSQI) for the overlapping sleep outcomes: total sleep time, time in bed, sleep latency, and sleep efficiency. Participants (n = 74; 28.9 years ± 4.6) were mostly white (73%), non-Hispanic (96%), married (78.5%) and over 60% had at least one child previously. Average PSQI global score was > 5 (cutoff for poor sleep) at each study visit. Total sleep time, sleep efficiency, and sleep latency measurements were significantly different between self-report and actigraphy throughout pregnancy and postpartum. Actigraphy-assessed sleep may reflect longer total sleep times, shorter sleep latency, and greater sleep efficiency compared to self-reported sleep among pregnant and postpartum women. This may be due to measurement error in actigraphy or recall bias when completing self-reported sleep measures. These factors should be taken into consideration both at the time of study design and when comparing results from different studies to facilitate the highest quality research and clinical decision-making in this population.