Background: Although increasing data link atypical depression (AD) to the bipolar spectrum, controversies abound about the extent of the overlap. In particular, the Columbia group, which has pioneered in providing data on operational clarity and pharmacological specificity of atypical depressions, has nonetheless consistently avoided studying its discriminatory validity from bipolar II (BP-II). Accordingly, we undertook a full scale validation of such a link in a large clinical sample of BP-II and unipolar (UP) major depressive disorder (MDD). Methods: Consecutive 348 BP-II and 254 MDD outpatients presenting with major depressive episodes (MDE) were interviewed off psychoactive drugs with a modified Structured Clinical Interview for DSM-IV, the structured Family History Screen and the Hypomania Interview Guide. We used the DSM-IV criteria for "atypical features" specifier. Depressive mixed state was defined as >= 3 concurrent hypomanic signs and symptoms during MDE. Bipolar validators were age at onset, high depressive recurrence, depressive mixed state and bipolar family history (types I and II). Univariate and multivariate logistic regression were used to examine associations and control for confounding variables. Results: Frequency of AD was 43.0% in the combined BP-II and MDD sample. AD, versus non-AD, had significantly higher rates of BP-II. AD was significantly associated with all bipolar validators, among which family history was the most robust. A dose-response relationship was found between number of atypical symptoms during MDE and bipolar family history loading. The association between bipolar family history and number of atypical symptoms remained significant after controlling for the confounding effect of BP-II. Bipolar family history was strongly associated with the atypical symptoms of leaden paralysis and hypersomnia. Conclusion: These results confirm a strong link between AD and bipolar validators along psychopathologic and familial grounds. From a practical standpoint, AD is best viewed as a variant of BP-I. Clinicians confronted with MDE patients presenting with atypical features should strongly consider a BP-II diagnosis. In a more hypothetical vein, atypicality or some associated features thereof-might serve as a nosologic bridge between UP and BP-II. (c) 2004 Elsevier B.V. All rights reserved.
