Abstract
BACKGROUND: Poor sleep quality is known to precede the onset of mood episodes and to be associated with poor treatment outcomes in bipolar disorder (BD). We sought to identify modifiable factors that correlate with poor sleep quality in BD independent of residual mood symptoms. METHODS: A retrospective analysis was conducted to assess the association between the Pittsburgh Sleep Quality Index and clinical variables of interest in euthymic patients with DSM-IV BD (n = 119) and healthy controls (HC; n = 136) participating in the Prechter Longitudinal Study of Bipolar Disorder. Multivariable linear regression models were constructed to investigate the relationship between sleep quality and demographic and clinical variables in BD and HC participants. A unified model determined independent predictors of sleep quality. RESULTS AND DISCUSSION: Euthymic participants with BD and HC differed in all domains. The best fitting unified multivariable model of poor sleep quality in euthymic participants with BD included rapid cycling (β = .20, p = .03), neuroticism (β = .28, p = 2 × 10(-3)), and stressful life events (β = .20, p = .02). Poor sleep quality often persists during euthymia and can be a target for treatment. Clinicians should remain vigilant for treating subjective sleep complaints independent of residual mood symptoms in those sensitive to poor sleep quality, including individuals with high neuroticism, rapid cycling, and recent stressful life events. Modifiable factors associated with sleep quality should be targeted directly with psychosocial or somatic treatment. Sleep quality may be a useful outcome measure in BD treatment studies.