Can Sleep Parameters Predict Upcoming Mood Episodes in Bipolar Disorder?

Background Bipolar disorder (BD) is a recurrent disorder, characterised by episodes of (hypo)mania, depression and euthymia with variation in mood, cognition and sleep. Many patients identify changes in sleep before an episode; using daily sleep logs could help identify these changes. Such early warning signs can be a valuable tool for patients and clinicians alike in predicting and preparing for changes in mood. Methods In the BipoSense study, we followed patients with BD, who were in remission at the start of the study, daily for 1 year. Patients reported for each hour if they were awake or asleep through an app and received fortnightly clinical assessments of bipolar symptoms. We used statistical analyses applying person-centred data in multilevel logit models to investigate if sleep patterns could differentiate between the period before an episode (prodromal stage) and euthymia, looking at both mean changes and variability of sleep. Bonferroni-Holm corrections were applied to avoid inflation of type I errors from multiple testing. Results Twenty-nine participants were included (mean age 44.0 years [SD = 11.9], female 55% and BD-I 59%). Waking up later was associated with prodromal depression and was the only significant finding for prodromal mood episodes. Greater variability of sleep duration, total time spent in bed and time waking up were associated with prodromal depression; less variability of time falling asleep and time waking up were linked with prodromal (hypo)mania. Conclusion Using self-assessed sleep changes and especially variability can be potential tools in helping patients identify early warning signs of mood recurrence; however, these analyses were explorative and further investigations are warranted.