The Antisuicial Effects of Lithium

Bipolar depression is strongly associated with suicide and premature death due to stress-related medical illness and complications of comorbid substance abuse. Because suicidal patients with bipolar depression are excluded from most clinical trials, remarkably little is known about the contributions of mood-altering treatments to reducing mortality rates in these persons. Despite clinical and ethical constraints on research into the therapeutics of suicide, encouraging new information is emerging to show that lithium has a selective effect against suicidal behavior in patients with major affective disorders.

Previous studies of lithium and suicide. We reviewed studies comparing suicidal rates in affectively ill persons treated with lithium. In all studies providing annual suicidal rates with and without lithium treatment, risk was consistently lower with lithium, averaging a seven-fold reduction. Incomplete protection from suicide may reflect limited effectiveness, inappropriate dosing, variable compliance, or the type of illness treated in this broad assortment of patients with severe mood disorders.

The antisuicidal benefit of lithium may represent a distinct action on aggressive behavior, perhaps mediated by serotonergic effects. Alternatively, it may reflect mood-stabilizing effects, particularly against bipolar depression. Our new findings indicate that lithium produces powerful and sustained reductions in depressive phases of both bipolar type I and type II disorders when administered over years of treatment.

Clinicians should not assume that all mood-stabilizers protect equally against both depression and mania or against suicidal behavior. For example, suicidal behavior occurred in a small but significant number of bipolar or schizoaffective patients treated with carbamazepine, but not in those receiving lithium (the anticonvulsant treatment did not follow discontinuation from lithium, a major stressor leading to sharp increases in bipolar morbidity and suicidal behavior).

New study of lithium vs. suicide. These previous findings encouraged additional studies. We examined life-threatening or fatal suicidal acts in over 300 bipolar type I and type II patients before, during, and following long-term lithium treatment at a collaborating mood disorder research center founded by Leonardo Tondo, M.D., of McLean Hospital and the University of Cagliari in Sardinia.

The patients had been ill for over eight years, from onset of illness to the start of lithium maintenance. Lithium treatment lasted over six years, at serum levels averaging 0.6-0.7 mEq/L, reflecting lithium doses consistent with optimal tolerability and patient compliance. Some patients were also followed prospectively for nearly four years after discontinuing lithium, without other maintenance treatments. Treatment discontinuation was monitored and distinguished from interruptions associated with emerging illness. Most discontinuations were clinically indicated for adverse effects or pregnancy, or were based on patients' decisions to stop without consultation, usually after remaining stable for prolonged periods.

Early emergence of suicidal risk. In this population of over 300 patients, life-threatening suicidal acts occurred at a rate of 2.30/100 patient-years (a measure of frequency over cumulative years) before they began on lithium maintenance. Half of all suicide attempts occurred in less than five years from onset of illness, when most subjects had not yet begun regular lithium treatment. Delays in lithium treatment from onset of illness were shortest in men with bipolar type I and longest in type II women, possibly reflecting differences in the social impact of manic versus depressive illness. Most life-threatening suicidal acts occurred before sustained maintenance treatment, suggesting that lithium treatment was protective and encouraging intervention with lithium early in the course of the illness to limit suicidal risk.

Effects of lithium treatment. During maintenance treatment with lithium, the rate of suicides and attempts decreased by nearly seven-fold. These results were strongly supported by formal statistical analysis: by 15 years of follow-up, the computed cumulative annual risk rate was reduced more than eight-fold with lithium treatment. With lithium treatment, most suicidal acts occurred within the first three years, suggesting that greater benefits derive from persistent treatment or earlier risk in more suicide-prone persons.

Effects of lithium discontinuation. Among patients discontinuing lithium, suicidal acts increased 14-fold above rates found during treatment. In the first year off lithium, the rate rose an extraordinary 20-fold. There was a two-fold greater risk after abrupt or rapid (1-14 days) versus more gradual (15–30 days) discontinuation. Although this trend was not statistically significant because of the infrequency of suicidal acts, the documented benefit of slow lithium discontinuation on reducing risk of relapse supports the clinical practice of slow discontinuation.

Risk factors. Concurrent depression or, less commonly, mixed-dysphoric mood, was associated with most suicidal acts and all fatalities; suicidal behavior was rarely associated with mania and no suicides occurred with normal mood. Additional analyses, based on an expanded Sardinian sample, assessed clinical factors associated with suicidal events. Suicidal behavior was associated with depressed or dysphoric-mixed current mood, prior illness with severe or prolonged depression, comorbid substance abuse, previous suicidal acts, and younger age.

Conclusions. These findings demonstrate that lithium maintenance exerts a clinically important and sustained protective effect against suicidal behavior in manic-depressive disorders, a benefit that has not been shown with any other medical treatment. Lithium withdrawal, particularly abruptly, risks a rapid, transient emergence of suicidal behavior. Prolonged delay from onset of bipolar illness to appropriate maintenance lithium treatment exposes many young persons to mortal risks as well as cumulative morbidity, substance abuse, and disability. Finally, the close association of suicidality with depression and dysphoria in bipolar disorders calls for further study to determine safe and effective treatments for these high-risk illnesses.

Additional Reading:

Baldessarini RJ, Tondo L, Suppes T, Faedda GL, Tohen M: Pharmacological treatment of bipolar disorder throughout the life-cycle. In Shulman KI, Tohen M. Kutcher S (eds): Bipolar Disorder Through the Life-Cycle. Wiley & Sons, New York, NY, 1996, pp 299–338

Tondo L, Jamison KR, Baldessarini RJ. Effect of lithium on suicide risk in bipolar disorder patients. Ann NY Acad Sci 1997; 836:339–351

Baldessarini RJ, Tondo L: Effects of discontinuing lithium treatment in bipolar manic-depressive disorders. Clin Drug Investig 1998; in press

Jacobs D (ed): Harvard Medical School Guide to Assessment and Intervention in Suicide. Simon & Shuster, New York, NY, 1998, in press

Tondo L, Baldessarini RJ, Floris G, Silvetti F, Hennen J, Tohen M, Rudas N: Lithium treatment reduces risk of suicidal behavior in bipolar disorder patients. J Clin Psychiatry 1998; in press

Tondo L, Baldessarini RJ, Hennen J, Floris G: Lithium maintenance treatment: Depression and mania in bipolar I and II disorders. Am J Psychiatry 1998; in press

* * * * * * * * * * * *

This article was contributed by Ross J. Baldessarini, M.D., Leonardo Tondo, M.D., and John Hennen, Ph.D., of the Bipolar & Psychotic Disorders Program of McLean Hospital, and the International Consortium for Bipolar Disorder Research. Dr. Baldessarini is also Professor of Psychiatry (Neuroscience) at Harvard Medical School and Director of the Laboratories for Psychiatric Research and the Psychopharmacology Program at McLean Hospital.

Permission of McLean Hospital