Predictors of efficacy in lithium augmentation for treatment-resistant depression.

Background: Lithium augmentation is widely applied for treatment-resistant depression, however, the clinical predictors of its efficacy regarding polarity and bipolarity are unknown.

Methods: We retrospectively examined the predictive value of clinical variables in 79 depressed patients who underwent lithium augmentation after failure to respond to antidepressant monotherapy. Lithium augmentation efficacy was evaluated by Clinical Global Impression Improvement assessment 4 to 8 weeks after initiating lithium administration; subjects with scores of 1 and 2 were defined as responders, and those with scores of 3 to 7 as non-responders. Clinical variables, including demographic and diagnostic variables, psychiatric medication, and clinical variables, were compared between groups. The bipolarity of patients with major depressive disorder as a final diagnosis was evaluated in association with the lithium augmentation efficacy. Data were analyzed using a chi-square test or Fisher's test.

Results: The lithium augmentation efficacy rate was 41% among 79 enrolled patients (14 dropped out, 32 responders, and 33 non-responders). Lithium augmentation was significantly more effective for patients with a final diagnosis of bipolar disorder than with major depressive disorder (p=0.03). Subjects with more than three major depressive episodes showed a significant response to lithium augmentation (p=0.004). The rate of a family history of major depressive disorder/bipolar disorder in a first-degree relative was significantly higher in responders (34%) than in non-responders (7%, p=0.01), consistent with the association between the efficacy of lithium augmentation and bipolarity in major depressive disorder (responders=27% vs. non-responders=3%, p=0.03).

Limitation: The study was retrospective and severity was not analyzed.

Conclusion: Bipolar disorder, frequency of major depressive episodes, and family history of major depressive disorder/bipolar disorder in a first-degree relative were detected as predictors of lithium augmentation efficacy. Among them, family history of major depressive disorder/bipolar disorder in a first-degree relative was the most reliable predictor of lithium augmentation efficacy for bipolar disorder and major depressive disorder.