Clinical Summary
Clinical Summary: Association of Lithium Use With Suicide and Self-Harm Among Chinese Adolescents and Adults With Bipolar Disorder: A Nationwide Multicenter Retrospective Cohort Study
Patients with bipolar disorder face a high burden of suicidal ideation, suicide attempts, and self-harm, yet clinicians often have less evidence about preventing early suicidal thoughts than about preventing attempts or death. This study addresses a practical treatment question in a large Chinese cohort: whether sustained lithium use is associated with lower risk across the suicidal spectrum, especially suicidal ideation.
Design
a multicenter retrospective cohort design spanning a 52-week observation period across 15 medical centers in China
N
585
Population
adolescents and adults with BD
Duration
a 52-week observation period
Key Findings
- Past-year suicidal ideation was less common with lithium than without lithium (53.1% vs 65.4%, P = .003), and lithium remained associated with lower odds after adjustment (adjusted odds ratio [aOR] = 0.57, 95% CI = 0.38–0.86, P = .008).
- Lithium was associated with lower rates of suicidal behaviors overall (23.8% vs 36.3%, P = .001), with adjusted analysis showing reduced odds as well (aOR = 0.60, 95% CI = 0.39–0.90, P = .02).
- Lithium users reported lower levels of earlier suicidal cognition, including wish to be dead (50.0% vs 61.7%, P = .006; aOR = 0.59, 95% CI = 0.39–0.88, P = .01) and nonspecific active suicidal thoughts (41.0% vs 52.5%, P = .007; aOR = 0.65, 95% CI = 0.44–0.96, P = .03).
- Time-to-event analysis for suicide attempt did not reach statistical significance, although the lithium group had a numerically lower cumulative incidence (adjusted HR = 0.74, 95% CI = 0.53–1.04, P = .08; log-rank P = .08).
- Lithium was not associated with less nonsuicidal self-injury (P = .167), but hostility scores were lower in the lithium group (mean [SD]: 21.42 [7.27] vs 23.03 [7.73], P = .01).
Clinical Bottom Line
In bipolar disorder, sustained lithium treatment was associated with markedly lower suicidal ideation and suicidal behaviors over 1 year, with the clearest signal for ideation. Lithium should remain a core maintenance option when suicide prevention is a treatment priority.
Practice Implications
- When treating bipolar disorder with elevated suicide risk, consider lithium not only for mood stabilization but also for reducing suicidal ideation, with adjusted odds of ideation lower at aOR = 0.57 (95% CI = 0.38–0.86, P = .008).
- Assess suicidal ideation in detail, including wish to be dead and nonspecific active suicidal thoughts, because lithium’s association was evident at these earlier stages of suicidality (50.0% vs 61.7%, P = .006; 41.0% vs 52.5%, P = .007).
- Do not assume lower suicidality means lower nonsuicidal self-injury; NSSI did not differ significantly (P = .167), so self-harm assessment should distinguish suicidal behaviors from NSSI.
- Patients with comorbid anxiety, rapid cycling, or predominant depressive episode type may need closer follow-up, as exploratory subgroup analyses found less robust protective associations in these clinical profiles.
