Lithium Better Than Divalproex In Keeping Suicides Down

Patients with bipolar disorder who take divalproex are 2.7 times more likely to die by suicide than bipolar patients taking lithium, a new study asserts. They are also 1.7 times more likely to attempt suicide in general and 1.8 times more likely than patients taking lithium to make a serious enough attempt at suicide to require hospitalization.

The new study, led by Frederick Goodwin, M.D., a research professor of psychiatry and behavioral sciences at the George Washington University School of Medicine and Health Sciences, was published in the September 17 Journal of the American Medical Association.

The study, funded by Solvay Pharmaceuticals, maker of the Lithobid brand of extended-release lithium, is believed to be the first to compare suicide risk of patients taking divalproex with that of patients taking lithium.

Divalproex (Depakote), an anticonvulsant with antimanic efficacy, is the leading medication used as a mood stabilizer in patients with bipolar disorder in the United States with roughly 10 times the sales volume of lithium, according to industry analysts at IMS Health. Goodwin, long a proponent of lithium’s virtues (it is the second-most-prescribed mood stabilizer in the United States) pioneered research into that medication’s ability to reduce the incidence of suicide in patients with bipolar disorder.

In the new study, Goodwin and his colleagues conducted a retrospective cohort study at two large integrated health plans on the West Coast: Kaiser Permanente in the San Francisco Bay and Sacramento areas and the Group Health Cooperative (GHC) in western Washington state. Those plans provide medical coverage to more than 3.4 million persons.

The investigators looked for any person in either plan who, between January 1, 1994, and December 31, 2001, had at least one outpatient diagnosis of bipolar disorder (type I or II) made on or after the patients’ 14th birthday. To be included in the study, patients had to have at least one prescription for either lithium, divalproex, or carbamazepine, which is the third-best-selling mood stabilizer in the United States, filled at a Kaiser or GHC pharmacy during the study period. Each patient included in the study cohort was followed from the time the first prescription was filled until either death, disenrollment from the health plan, or the end of the study on December 31, 2001.

Three Outcomes Tracked

Three potential outcomes were tracked: completed suicides, suicide attempts resulting in an emergency room visit or hospitalization, and suicide attempts resulting in assessment and discharge.

The researchers identified nearly 21,000 patients in the two health plans treated for bipolar disorder with one of the three mood stabilizers during the study period. In that group, there were 53 suicides, 338 attempts resulting in hospitalization, and 642 attempts identified in the emergency room and discharged. Suicide attempts resulting in hospitalization occurred 6.2 times more frequently than completed suicides.

After adjusting for a number of potentially confounding variables—age, sex, health plan, year of diagnosis, comorbid medical conditions, and other psychotropic medication use—the risk of each of the three outcomes was significantly greater during exposure to either divalproex or carbamazepine, compared with lithium. Most comparisons of carbamazepine and lithium, however, while qualitatively similar to those of divalproex and lithium, were not statistically significant due to the relatively small number of patients taking carbamazepine. Only one result was statistically significant: Patients taking carbamazepine were 2.9 times more likely to attempt suicide resulting in hospitalization than those taking lithium.

Prescribing Patterns Shift

Interestingly, the eight-year study captured a significant shift in prescribing habits in the treatment of bipolar patients. In 1994 the ratio of prescriptions for lithium to divalproex was roughly 6 to 1, while at the end of the study in 2001, the ratio was closer to 1 to 2. (That trend has continued and is now nearer to 1 to 3 today, according to IMS Health data.)

Regardless of the shifting habits in prescribing, rates for suicide attempts of patients receiving lithium were remarkably stable during the period. Rates for suicide death were less consistent, however; Goodwin believes that may be due to the relatively small number of deaths over the eight-year period.

In addition, a history of any medication switch—whether from lithium to divalproex, or the other way around—was associated with a higher risk of suicide attempt. Many patients were taking more than one mood stabilizer at a time, either as a combination therapy or while being switched from one medication to another. This could confound the results.

Goodwin stressed, however, that “only a randomized trial can completely exclude the possibility of confounding or bias, but large observational studies such as this one may be the only realistic option for studying relatively rare outcomes such as suicide death.”

In an accompanying editorial, Ross Baldessarini, M.D., a professor of psychiatry at Harvard Medical School and McLean Hospital, along with Leonardo Tondo, M.D., a lecturer in psychiatry at Harvard and McLean, wrote that bipolar disorder, because of its episodic nature, high depressive morbidity, comorbidity with substance abuse and anxiety disorders, and consequently high mortality rate, is “a major unresolved public health challenge.”

Suicide death in patients with bipolar disorder “represents an extraordinarily high risk” with an estimated rate of 0.4 percent a year, compared with the international background rate of 0.017 percent a year.

The two noted that “no currently available treatment provides full protection from recurrences of manic, mixed, manic-depressive, major depressive, or highly prevalent milder depressive states.”

In particular, Baldessarini and Tondo said, Goodwin’s study “not only has merits in its methods and findings, but also represents a significant contribution to a newly emerging interest in suicide as the major unresolved medical problem that it is.”

An abstract of “Suicide Risk in Bipolar Disorder During Treatment With Lithium and Divalproex” is posted on the Web at http://jama.ama-assn.org/cgi/content/abstract/290/11/1467. ▪

JAMA 2003 290 1467