Lithium Shows Promising but Statistically Unclear Anti-Suicide Effects in New Meta-Analysis
A 2025 meta-analysis of 15 RCTs finds lithium reduces suicide attempts and completed suicide, though results fall short of statistical significance.
Summary
A 2025 systematic review and meta-analysis examined lithium's effect on suicidality across 15 randomized controlled trials involving over 3,000 participants. Lithium was associated with a 27% reduction in suicide attempts and a 39% reduction in completed suicide compared to placebo, but neither result reached statistical significance. Researchers attribute this to methodological limitations — including small sample sizes, diagnostic heterogeneity, subtherapeutic lithium levels, and inconsistent suicidality measurement tools — rather than a true absence of effect. The findings align with robust observational data supporting lithium's anti-suicide properties, suggesting the RCT evidence base is underpowered rather than contradictory. Lithium remains a first-line treatment for bipolar disorder with a plausible and clinically meaningful role in suicide prevention.
Detailed Summary
Suicide prevention remains one of psychiatry's most urgent challenges, and lithium has long been regarded as one of the few medications with potential anti-suicide properties. Despite strong observational evidence, the picture from randomized controlled trials has been less definitive. This updated meta-analysis sought to clarify lithium's role by pooling the best available RCT data.
Researchers followed Cochrane and PRISMA protocols, searching major databases from January 2013 to July 2024 and supplementing with earlier manual references. From 1,793 screened articles, 15 studies met eligibility criteria — eight placebo-controlled (n=1,698) and seven open-label (n=1,338) — covering outcomes including suicidal ideation, suicide attempts, and completed suicide.
The pooled results showed lithium was associated with fewer suicide attempts (OR=0.73) and fewer completed suicides (OR=0.61) versus placebo or comparators. However, neither finding reached conventional statistical significance, with wide confidence intervals spanning 1.0. Heterogeneity in how suicidal ideation was measured across studies prevented any quantitative pooling of that outcome.
The authors argue the lack of significance likely reflects type II error — insufficient statistical power due to small samples, inconsistent diagnostic criteria, variable lithium dosing, poor treatment adherence, and differing baseline suicidality levels — rather than a true null effect. These methodological gaps reduce the sensitivity of the studies to detect real differences.
Clinically, the directional consistency of the RCT data with large observational studies strengthens confidence in lithium's anti-suicide effect. The review underscores the need for larger, better-designed trials with standardized suicidality measures, adequate lithium levels, and homogeneous patient populations to produce definitive evidence on one of psychiatry's most important clinical questions.
Key Findings
- Lithium reduced suicide attempts by 27% (OR=0.73) vs. placebo, but did not reach statistical significance.
- Completed suicide occurred in 4 lithium vs. 13 placebo patients across trials (OR=0.61), also nonsignificant.
- 15 RCTs with over 3,000 participants were included; suicidal ideation could not be meta-analyzed due to measure heterogeneity.
- Authors attribute nonsignificant results to type II error from small samples and methodological inconsistencies, not true lack of effect.
- Findings directionally align with observational studies consistently showing lithium reduces suicide risk in bipolar disorder.
Methodology
Systematic review and meta-analysis of 15 RCTs following Cochrane and PRISMA guidelines, searching OVID databases (Embase, MedLine, PsychINFO) from January 2013 to July 2024, supplemented by manual reference searching. Two independent reviewers extracted data and assessed study quality. Both placebo-controlled and open-label trials were included.
Study Limitations
No individual trial was adequately powered to detect significant differences in rare events like suicide. Substantial heterogeneity in diagnostic criteria, lithium dosing, treatment duration, and suicidality measurement tools limits pooling and generalizability. Several included studies used subtherapeutic lithium levels, potentially diluting any detectable treatment effect.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.