A prospective study of the impact of comorbid medical disease on bipolar disorder outcomes.

Background: Several studies suggest that medical comorbidity is associated with worse clinical status in bipolar disorder. It is unclear which aspect of medical comorbidity is responsible: simple disease count, risk for future morbidity, or current physical burden.
Methods: We analyzed three years of prospective data from a randomized clinical trial of collaborative care in 306 bipolar veterans. We examined the association of clinical outcome with baseline medical comorbidity defined as: (1) simple active disease count, (2) diseases with risk for future morbidity measured with the Charlson Comorbidity Index, and (3) current physical burden measured with the SF-36 Physical Component Summary score (PCS). Bipolar outcomes were weeks in episode, mean depression score, and change in mental health burden measured by the SF-36 Mental Component Summary score (MCS).
Results: The three medical comorbidity measures were not highly correlated, indicating that each conveyed novel information. Controlling for potential confounders, worse baseline PCS predicted significantly higher mean depression scores (p=0.011) and less improvement in MCS scores (p=0.0099) over three years. Simple disease count and risk for future risk did not predict worse bipolar outcomes.
Limitations: Some potential limitations include not accounting for all confounding factors, selection bias for participants, increased the likelihood of Type I error due to multiple comparisons and having a predominantly male population.
Conclusions: This long-term prospective study extends cross-sectional and retrospective research on the link between medical illness and bipolar outcomes. It is the current experience of burden of physical illness, rather than an unweighted or weighted disease count, that leads to worse bipolar outcomes.