Antisuicidal effect of lithum in bipolar patients: Is there an age-specific effect?

Objective: Our group reported previously a comparable overall antisuicidal effect of lithium and valproate in bipolar patients. We investigated factors associated with higher antisuicidal efficacy of lithium in older individuals. Methods: The age-related antisuicidal effect of lithium and valproate was compared in ninety-four (n = 94) high-risk bipolar suicide attempters who participated in a 2.5-year randomized, double-blind trial. Results: Age significantly moderated the effect of lithium vs. valproate on the risk of suicide event during the study (z = −1.98, p = 0.049). We found that those who were 42 years or older (above the 75th percentile), and on lithium had significantly lower risk of suicidal behavior than older patients on valproate (>42y) or younger (<42 y) patients on either medication (interaction HR = 0.09, 95%CI: 0.01–0.89, z = −2.07, p = 0.039). This difference in risk differences was not explained away by age-related differences in the proportion of participants with bipolar II disorder (Fisher's test p = 0.020) or higher lethality of past suicide attempts in younger participants (Wilcoxon test p = 0.024); neither was there any correlation with age in the longitudinally measured blood lithium levels (t = 1.04, df = 36, p = 0.307) or valproate levels (t = −0.50, df = 41, p = 0.621). Limitations: Besides the fact that this is a secondary analysis, a limitation is that the study is not powered to detect suicide deaths or suicide attempts. Conclusion: Bipolar patients randomized to lithium and older than 42 years had less suicidal behavior compared to same aged patients on valproate or younger patients (<42 y) on either medication. This effect was independent of clinical and sociodemographic characteristics.