Lithium for Acute Mania

Clinical Question

Is lithium a safe and effective therapy for episodes of acute mania?

Evidence-Based Answer

Lithium is more effective at inducing a reduction of at least 50% on the validated Young Mania Rating Scale (YMRS) compared with placebo (number needed to treat [NNT] = 6; 95% CI, 5 to 8; six studies; n = 1,707). (Strength of Recommendation [SOR]: A, based on consistent, good-quality patient-oriented evidence.) Most comparisons of lithium with mood stabilizers and antipsychotics demonstrate little to no difference in effectiveness; however, olanzapine (Zyprexa) is slightly more effective than lithium. When compared with placebo, lithium increases the risk of tremor (number needed to harm [NNH] = 11; 95% CI, 7 to 21; six studies; n = 1,241) and somnolence (NNH = 19; 95% CI, 10 to 50; seven studies; n = 1,351).1 (SOR: A, based on consistent, good-quality patient-oriented evidence.)

Practice Pointers

Bipolar disorder is a common disease in which patients experience some combination of depressed mood, elevated mood (mania), and mixed states. Lithium is effective as a maintenance drug in bipolar disorder, acting as a mood stabilizer and decreasing the risk of suicide.

This Cochrane review evaluated lithium as a first-line therapy for acute mania, and included 36 randomized controlled trials and 4,220 patients.1 Three of the studies included patients younger than 18 years. The other studies included male and female patients of any age with bipolar disorder who fit criteria for a manic episode. The studies compared lithium with placebo, electroconvulsive therapy, and 12 other medications over three to 12 weeks. The study authors used odds ratios to analyze binary efficacy outcomes.

A positive response was substantiated by a reduction of at least 50% in symptoms on the YMRS. Patients who had acute mania treated with lithium vs. placebo were more likely to achieve this effect (NNT = 6; 95% CI, 5 to 8; six studies; n = 1,707). Lithium also allowed patients to achieve remission of acute mania earlier than placebo (NNT = 6; 95% CI, 5 to 8; five studies; n = 1,597). These trials predominantly occurred over three to four weeks, with the longest comparison lasting eight weeks.

For direct drug comparisons, lithium appeared less likely to induce a reduction of at least 50% in symptoms on the YMRS vs. olanzapine (NNT = 6; 95% CI, 3 to 84; two studies; n = 180; low-certainty evidence downgraded from moderate because of the risk of publication bias and low number of studies). Lithium was more effective than topiramate (Topamax) at treating acute mania (NNT = 6; 95% CI, 4 to 10; one study; n = 660; high-certainty evidence; 12-week trial). There was no evidence that lithium was better or worse at treating mania than any of the other studied medications, nor was there enough evidence to draw a conclusion about electroconvulsive therapy.

Compared with placebo, lithium was more likely to cause tremor (NNH = 11; 95% CI, 7 to 21; six studies; n = 1,241; high-certainty evidence) and somnolence (NNH = 19; 95% CI, 10 to 50; seven studies; n = 1,351; high-certainty evidence). Additional reported adverse effects of lithium therapy included nausea, vomiting, polyuria, polydipsia, and anorexia.2 The adverse effect profile of olanzapine is similar, with the addition of extrapyramidal symptoms.

The National Institute for Health and Care Excellence guidelines note that in a secondary care setting, patients with acute mania should be treated with antipsychotics. However, if a maximum dose of an antipsychotic is ineffective, lithium should be considered.3 This Cochrane review suggests that physicians use lithium to treat acute mania because it is similarly effective to other agents and more effective than placebo. Consideration should be given to the potential adverse effects and monitoring required for lithium, as well as any other agent, before administering treatment, keeping in mind the risks and benefits to the individual patient.

Editor's Note: The NNTs, NNHs, and CIs reported in this Cochrane for Clinicians were calculated by the authors based on raw data provided in the original Cochrane review.