Prospective predictors of mood episodes in bipolar disorder.

Background: : Bipolar disorder (BD) is associated with alterations in mood, personality, cognition and event-related potential (ERP) measures. The relationship between these multidimensional measures of state and subsequent course of the illness is not well understood. Therefore, this study aimed to prospectively identify factors that predicted the course of mood episodes.

Methods: : Sixty-five participants with BD were administered the auditory P300 oddball task, clinical assessment instruments and cognitive tests at baseline, and were subsequently administered the SCID interview once a month by telephone for 12 months.

Results: : Hierarchical regression analyses indicated that the Montgomery-Asberg Depression Rating Scale (MADRS) predicted the number of months spent in a depressed state (p<.001) and in a mixed state (p=.001), while both the MADRS (p<.001) and time to complete Trails A (p=.033) predicted total number of months in a mood episode (across all mood states). Among euthymic patients at entry, Cox regression analyses indicated that higher ratings on both the MADRS (p=.017) and Hypomanic Personality Scale (HPS; p<.001) were associated with both increased likelihood of a mood episode and less time until the onset of a mood episode.

Limitations: : The sample size is relatively small, not all participants completed 12 months, and follow-up assessments were conducted via telephone.

Conclusions: : Our results suggest that affective and cognitive measures, and personality factors, especially the MADRS and HPS, serve as important predictors of the course of mood episodes or relapse in BD patients. These prospective markers of acute mood states may be used to guide treatment decisions.