Dr. Goldberg Replies

To the Editor: Dr. Liebowitz offers an interesting anecdotal observation about the potential importance of diurnal variation in bipolar depression and its possible relationship to mood stabilizer response. Unfortunately, the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) study did not obtain data on this clinical characteristic during depressive episodes. Existing literature on the phenomenology of bipolar depression has not identified diurnal variation as a feature that is more common in bipolar than unipolar depression, in contrast to other constructs such as reversed vegetative signs (1) . However, some investigators have hypothesized that diurnal mood variation may reflect disruptions of circadian rhythms in healthy volunteers (2) , suggesting a possible role in cyclical or highly recurrent mood disorders.

Dr. Liebowitz also raises the hypothesis that short-acting antidepressant preparations may be less prone than longer-acting formulations to induce affective polarity switch. STEP-BD found no differences in treatment-emergent affective switch in bipolar depressed patients who underwent equipoise randomization to adjunctive bupropion sustained-release, paroxetine, or placebo in conjunction with mood stabilizers (3) . Lack of efficacy appears to be an overall greater risk than induction of mania for most depressed bipolar patients who receive adjunctive antidepressants. The STEP-BD study did not evaluate depression outcomes with highly noradrenergic antidepressants, such as venlafaxine, but their apparent higher risk for induction of mania relative to predominantly serotonergic or dopaminergic antidepressants (4) would seem to prompt caution if one chose to expose a bipolar patient to other noradrenergic agents, especially those not studied in bipolar disorder, such as atomoxetine.