Background: Clinical presentations of depression in bipolar disorder are varied, inconsistent and often confusing. Most previous studies have focused on bipolar I (BP-I). Given that bipolar II (BP-II) is the more common bipolar phenotype, which is often confused with unipolar (UP), the aim of the current analyses is to delineate the symptomalogic differences between BP II vs. UP MDD in a large national sample. Methods: The data derived from the French National EPIDEP study (n = 452 DSM-IV major depressives), subdivided into BP-II (n = 196) and UP (n = 256). The BP II group included major depressives with both spontaneous and antidepressant-associated hypomania based on our finding of similarity in rates of familial bipolarity in the two subgroups. At index presentation, depression was assessed by the clinician (using HAM-D and the Rosenthal Atypical Depression Scale) and by the patient (using the Multi-Visual Analog Scale of Bipolarity, MVAS-BP). Principal component analyses (PCA with varimax rotation) were conducted on HAM-D and MVAS-BP in the total population and separately in BP-II and UP. We performed inter-group comparative tests (UP vs. BP-II) on factorial scores derived from PCAs and correlation tests between these factorial scores. Results: The PCA on "HAM-D +Rosenthal scale" showed the presence of nine major factors: F1-2 "weight changes", F3-4 "sleep disturbances", F5 "sadness-guilt", F6 "retardation-fatigue", F7 "somatic", F8 "diurnal variation" and F9 "insight-delusion". The PCA on MVAS-BP revealed the presence of eight principal components: F1 "psychomotor retardation", F2 "central pain", F3 "somatic", F4 "social contact", F5 "worry", F6 "loss of interest", F7 "guilt" and F8 "anger". Despite uniformity in global intensity of depression, significant differences were observed as follows: higher score on "psychomotor retardation" (p=0.03), "loss of interest" (p=0.057) and "insomnia" (p=0.05) in the UP group, and higher score on "hypersomnia" (p=0.008) in the BP-II group. Correlation analyses between clinician- and self-rating revealed the presence of higher number of significant coefficients in the UP vs. BP-II group (p <= 0.001). Limitation: A three-way comparison between BP-I, BP-II and UP may have yielded somewhat different results. Conclusion: Our data indicate greater psychomotor retardation, stability and uniformity in the clinical picture of strictly defined UP depression. By contrast, bipolar II depression appeared to be characterized, despite the hypersomnic tendency, by psychomotor activation. This would indicate greater mixed features than those observed in UP. Moreover, in BP-II, there was less agreement between clinician vs. self-rating on the presence of various features of depression. Taken together, these findings explain why BP-II depression is missed by clinicians as a genuine depression. (c) 2004 Published by Elsevier B.V.
