BACKGROUND

Mood stabilizers (MS), especially Lithium, are used in augmentation strategies for resistant depression. However the broader bipolar spectrum (depressions with brief [i.e. 2 days] hypomania, cyclothymic and hyperthymic temperaments) has rarely been explored in such strategies. The aim of the current report is to search for predictive factors for response to mood stabilizers when used as augmentation therapy after excluding clear-cut hypomania and focusing on DSM-IV major depressive disorder (UP-MDD), which is best designated as apparently "unipolar".

METHOD

From the total sample of 452 major depressives (MDE) included in the French National study EPIDEP, 256 were classified as UP-MDD after eliminating DSM-IV bipolar II (> or =4 days of hypomania); conservatively, we also excluded MDD with hypomania associated with antidepressants. Lifetime treatment history of UP-MDD revealed that 59 (23.3%) had received at least one MS (lithium, valpromide [French variant of divalproex], and carbamazepine) in the past; from this sub-population, 18 were considered retrospectively as good responders (30%, GR) versus 41 poor responders (70%, PR) to MS augmentation on the basis of the clinical judgment of the treating psychiatrist.

RESULTS

Comparative analyses between patients who received MS and those who did not, revealed the former group as having higher levels on the hypomania checklist and cyclothymic and depressive temperaments. The delay to MS installation was significantly longer in the PR versus GR. The profile of GR could be described as follows: younger current age, higher education; symptom-free interval between major episodes; and fewer prior depressive episodes and hospitalizations; and higher rate of MS prescription. However, no significant differences were obtained from hypomania assessment and affective temperament ratings (cyclothymic, hyperthymic, depressive). During the index (most recent) depressive episode, we obtained a significantly higher rating of "suicidal thoughts" associated with higher levels of "sadness-guilt," psychomotor agitation, and lower "retardation-fatigue" (all from the HAM-D) in the PR group; better and faster response to current treatment (as prospectively assessed) were also observed in the GR. At this time, overall severity of depression was not linked to the quality of response to the MS. LIMITATIONS AND CONCLUSION: Despite its retrospective design, these analyses have important implications in the management of difficult or resistant "unipolar" depression by using MSs as augmentation strategy. Clinicians appeared to have used "subtle" hypomanic and cyclothymic features as a justification for augmentation. However, these features per se were not predictive of response to such augmentation. Instead, the profile of augmentation response to failed antidepressants appears to be an "activated depression" (significantly less retardation and withdrawal and higher agitation associated with greater intensity of painful and guilt-ridden sadness with suicidality), and the significantly higher rate of and earlier prescription of MSs in the course of recurrent MDD. These data suggest that resistant depressives should not stay on antidepressant or antidepressant combination for too long; MS augmentation must be instituted without much delay.