Do lithium levels to prevent depressive episodes differ from those to prevent manic/mixed episodes in bipolar disorder? A post-hoc analysis

Introduction: It is unclear, whether in the long-term treatment of bipolar disorder an increase in lithium levels beyond 0.6mmol/l may provide enhanced control of manic/mixed only, or of depressive symptoms/episodes as well. To clarify this issue we performed a post-hoc analysis of the 12-month olanzapine versus lithium maintenance trial (Tohen et al., 2005). Methods: An “inverse-probability-of-treatment“ weighted version of time to event analysis was performed, with two competing events being 1) manic/mixed and 2) pure depressive relapses incorporating lithium dose as time-varying covariate. Lithium levels were divided into three categories: <0.6mmol/l; 0.6–0.8mmol/l; >0.8mmol/l. Results: There was a significant difference between the lowest lithium level group (<0.6mmol/l) versus either of the higher lithium level groups (0.6–0.8mmol/l; >0.8mmol/l) or versus combined group (>0.6mmol/l) in terms of manic/mixed, but not depressive relapse. The relative risk of a manic/mixed relapse for lithium levels at or higher than 0.6mmol/l, compared to those lower than 0.6mmol/l was: HR=0.41, CI 0.22–0.74, p<0.01 (unweighted analysis). Adjustment for possible selection bias due to non-random assignment of lithium dose yielded similar results. Discussion: The results of this analysis suggest that lithium levels >0.6mmol/l provide additional benefits in the prevention of manic/mixed relapse, but not depressive relapse.

This study was supported by Eli Lilly