Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
334363 | Psychiatry Research | 2007 | 10 Pages |
The study's aim was to find if features often reported to distinguish bipolar and depressive disorders could predict bipolar-II disorder (BP-II). Consecutive major depressive episode (MDE) outpatients, including 284 with BP-II and 196 with major depressive disorder (MDD), were interviewed with the Structured Clinical Interview for DSM-IV, Hypomania Interview Guide, and Family History Screen, in a private practice. The minimum duration of past hypomania was 2 days. Mixed depression was defined as an MDE plus three or more intradepressive, non-euphoric hypomanic symptoms. BP-II predictors were early onset (< 20 years), many recurrences (> 4 MDEs), bipolar family history, mixed depression, and atypical depressions. Bipolar family history had the highest positive predictive value (PPV) (80.8%) but low sample frequency (32.7%); early onset had high PPV (75.2%) and a sample frequency of 37.0%; many recurrences had the highest frequency (70.4%) but the lowest PPV (66.5%). Combinations of three or more predictors had high PPV (79.0%) and a sample frequency of 46.6%. Predictors and combinations of predictors may correctly identify 75% to 80% of BP-II, reducing the misdiagnosis of BP-II as MDD (by prompting careful probing for hypomania history), and improving treatment of depression (as antidepressants alone may worsen BP-II course). As PPV is related to disease prevalence, findings need to be replicated in different settings.