From Pages to Practice

By Julie Barzilay, MD, MPhil

Published September 14, 2023


A 40-year-old man with bipolar I disorder who is currently on a mood stabilizer is feeling better after a recent depressive episode. During the episode, he was treated with an adjunctive antidepressant. Now that his depression is in remission, you are faced with a choice: Do you continue the antidepressant or taper and discontinue it? Yatham et al. examined this question in a multisite, double-blind, randomized, placebo-controlled trial recently published in NEJM.

The study evaluated 177 patients with bipolar I disorder who were treated with either a mood stabilizer, a second-generation antipsychotic, or both, and recently had remission of a depressive episode. The patients were randomized to continue receiving an adjunctive antidepressant (escitalopram or bupropion XL) for 52 weeks after remission of depression or switch to placebo 8 weeks after remission (6 weeks with a 2-week taper). Most participants were of Asian descent.

The hazard ratio for time to any mood episode (depressive, manic, or mixed) in the 52-week group relative to the 8-week group was 0.68 (95% CI, 0.43–1.10; P=0.12). Does this result mean that stopping the antidepressant after 8 weeks did not make any difference? That is one conclusion, but some nuances make these results difficult to interpret.

First, the primary outcome of any mood episode is heterogeneous: Analysis of the secondary outcome of time to an episode of mania/hypomania or depression indicated that the 52-week group had numerically fewer depressive events, but more manic and mixed events than the 8-week group. These events are not equivalent, and individual patients may be more concerned about preventing one type of event over another. Second, because the secondary analyses were not adjusted for multiplicity, no causal inferences or claims about significance can be made. Another limitation is that the study was stopped before the planned sample size of 216 patients was reached due to slow recruitment during the Covid-19 pandemic, which may have limited the ability to detect a significant difference.

So, what can we take away from this study? For the 40-year-old patient with bipolar I disorder and recently remitted depression, the results suggest no overall benefit from remaining on the antidepressant for a year. Continuing the antidepressant could prolong the time to the next depressive episode, but it may increase the probability of a manic episode. On the other hand, stopping the antidepressant might spare the patient some medication side effects, since numerically more patients in the 52-week group experienced weight gain, fatigue, and memory problems. These nuanced results provide a bit more data to use in thoughtful conversations with patients about treatment options related to antidepressants for maintenance therapy after the remission of depression in bipolar I disorder.

Read the following NEJM Journal Watch summary for more details of this study.

Continuing Adjunctive Antidepressants After Remission in Patients with Bipolar Depression

Peter Roy-Byrne, MD, reviewing Yatham LN et al. N Engl J Med 2023 Aug 3

Antidepressants are prescribed widely for bipolar depression, despite limited evidence of their efficacy for this condition. This multisite trial involved 187 patients with bipolar 1 disorder in whom recent depressive episodes had remitted during treatment with adjunctive antidepressants (i.e., bupropion or escitalopram) added to mood stabilizers. Patients were randomized to continue antidepressants for either 8 or 52 weeks.

Time to next mood episode (either depression or mania/hypomania) was not significantly different in the two groups. Patients in the 52-week group had numerically fewer depressive recurrences but numerically more manic or hypomanic episodes.

Comment: This study suggests that, in patients with bipolar 1 disorder, adjunctive antidepressants can be stopped after resolution of acute episodes. The results add to other data that show no or minimal benefit for standard antidepressant treatment in this population. Interpreting response to treatment in bipolar illness is challenging, because spontaneous mood switches often give the illusion of treatment efficacy. In my practice experience, only a small proportion of patients with bipolar disorder benefit from antidepressants — either acutely or chronically — consistent with multiple guidelines indicating that standard antidepressants are third-line options. In patients with bipolar disorder who are clinically stable on mood-stabilizing agents and standard antidepressants, primary care clinicians could use these data to suggest deprescribing antidepressants.

Browse more From Pages to Practice »

Julie Barzilay, MD, MPhil, is a 2023–2024 NEJM Editorial Fellow. She is currently a pediatrician in the Division of General Pediatrics at Boston Children's Hospital. She completed her pediatrics residency in the Boston Combined Residency Program at Boston Children's Hospital and Boston Medical Center.