Research: Is poor sleep a risk factor for postnatal depression?

My former PhD student (Dr Lauren Kita) explored the extent that poor sleep may be a risk factor for postnatal depression. This is a summary of her work, conducted under my supervision.

Background: Research has suggested that a bi-directional relationship exists between sleep disruption and depression. Not only is poor sleep a commonly reported symptom in those with depression, some aspects of sleep have also been shown to predict the onset of depression. Despite sleep problems being a commonly reported occurrence throughout the perinatal period, the field of perinatal sleep research remains in its relative infancy. However, recent studies suggest that sleep disturbances during this time may increase the risk of developing postpartum depression (PPD).  Currently, research in this area is limited by studies that have failed to control for depressive symptoms at baseline, relied upon subjective, often retrospective, measures of sleep, and have only measured symptoms of postpartum depression in the early postpartum period. Few studies have used polysomnography (PSG), considered the ‘gold standard’ of sleep, and no studies to date have specifically compared the relationship between subjective and objective sleep.

What we did: A longitudinal study was carried out to examine the relationship between sleep and postpartum depression. This study was divided into three phases. Phase 1 involved comparing the sleep of pregnant, third trimester women (N=29) to non-pregnant controls (N=24) using PSG. We found that pregnant women experienced poorer sleep both subjectively and objectively. This included less REM sleep and more Stage 1 and 2 sleep in the pregnant group.

Phase 2 examined longitudinal changes in sleep quality over three time-points: the third trimester of pregnancy, the first postpartum week, and 12 weeks postpartum. Compared to late pregnancy and 12 weeks postpartum, women reported the most dramatic changes to their sleep and most fatigue during the first postpartum week. Despite attempts to compensate for lack of sleep through daytime napping, women experienced over one and a half hours less sleep compared to late pregnancy and reported significantly poorer sleep quality. This highlights the scope of sleep changes that occur across a short-time interval. The final phase of the study examined whether these sleep changes precipitated changes in depressive symptoms.

Phase 3 investigated the longitudinal relationship between sleep and postpartum depressive symptoms. The key factors that were predictive of early PPD symptoms were: increased amount of sleep and difficulty falling asleep during the last trimester of pregnancy. This relationship was significant while controlling for initial depressive symptoms, suggesting that these specific aspects of sleep are important predictors of PPD. Additionally, depressive symptoms during the last trimester of pregnancy was a stronger predictor of depression at 12 weeks postpartum than symptoms arising in the early postpartum period.

The findings of this study highlight the need to educate women and health-care providers about pregnancy-related sleep changes. If women and understand the extent and importance of these changes, particularly in the period between late pregnancy and the early postpartum period, then they may be able to better prepare. The finding that increased total sleep during late pregnancy predicted increased symptoms of PPD may be due to several reasons that require further investigation. It may be that women who slept longer towards the end of pregnancy experienced a more drastic decrease in total amount of sleep from pre to postpartum. Such changes are likely to be associated with changes in the circadian system which may give rise to hormonal changes. Given that difficulty falling asleep was a key factor related to PPD, this suggests that more research is needed in order to investigate perinatal insomnia.