Abstract
The dichotomous distinction between unipolar and bipolar disorders may be challenged by heterogeneity within diagnoses and overlap between different diagnoses. A broad mood disorder category in which patients differ as a result of variation along separate manic and depressive mood dimensions can be proposed. To test this, it is hypothesized that heterogeneity in clinical and other features of subjects selected for unipolar depression can be partly explained by coexisting manic symptoms. A cohort selected for unipolar depressive disorder was followed up for two years at which time co-occurring manic symptoms were assessed, yielding four groups with increasing manic symptomatology: i) pure unipolar depressive disorder (n=1598), ii) unipolar depressive disorder with subthreshold manic symptomatology (n=64), iii) bipolar II disorder (n=39), and iv) bipolar I disorder (n=86). Multivariate logistic regression and analyses of covariance controlled for depression severity were used to investigate whether patients with increasing manic symptomatology could be differentiated from patients with pure depressive disorder. Male gender, a lower age at first episode, a history of suicide attempts and increased aggressive cognitions were independently associated with an increase in manic symptoms. The additional presence of (hypo)mania was associated with greater depression severity and more disability than pure depressive disorder. The groups with manic symptomatology (subthreshold, hypomania and mania) were considerably smaller compared to the pure depression group. The heterogeneity in depressive illness can be partly explained by the coexisting variation along the manic symptom dimension. Co-occurring manic symptoms should be taken into account in depression and a symptom dimensional approach of mood disorders may provide phenotypes that are more informative than current mood disorder categories.