Skip to main content Back to Top

5/19/2001

Severe Depression Does Not Respond to St. John's Wort, Study Says

Kate Traynor

A study funded by the manufacturer of an antidepressant drug found that St. John’s wort performs no better than placebo in treating major depression.1

By the end of the eight-week study at 11 U.S. medical centers, 27% of the 98 patients with severe depression who were assigned to take St. John’s wort had improved in their scores on a standard depression-rating scale. But 19% of the 102 patients who used a placebo also scored better over the course of the study.

The herb performed better than placebo in one type of analysis: depression scores by all patients who were randomized to a treatment group, regardless of whether they dropped out later or continued for the full eight weeks. In that analysis, 14% of St. John’s wort users, compared with 5% of control patients, did not have symptoms of depression when last assessed. But the study authors asserted that, given these extremely low response rates, the herb might not have "significant antidepressant or antianxiety effects."

According to the researchers, their findings contradict the results of other published studies that touted St. John’s wort as a successful treatment for depression. The research team attributed their discordant findings to the other studies’ "serious methodological flaws" that affected the results. Among the shortcomings cited were nonstandard methods for diagnosing depression, failure to use standardized symptom-rating scales, and too few patients to reliably demonstrate treatment effects.

The study took place between November 1998 and January 2000 and enrolled men and women who had been diagnosed with severe depression of at least four weeks duration. On average, the study participants had had depressive symptoms for more than two years. Nearly two thirds of the study’s enrollees were women.

Initial treatment consisted of three 300-mg tablets of St. John’s wort or three identical placebo tablets each day. After four weeks, patients whose symptoms had improved continued the three-tablet regimen for another four weeks. Patients whose condition had not improved after the first four weeks took a fourth tablet each day for four more weeks.

The researchers used a battery of standardized tests to asses the study participants’ mental state at several points during the trial. Study participants also underwent physical examination, electrocardiographic testing, and blood chemistry analysis.

People in the herbal treatment group were more likely than placebo users to have headaches during treatment. Overall, the adverse events discussed in the article were minor, although one patient in each group withdrew from the study because of an unspecified adverse event. Eighty patients in the herbal group and 87 placebo users completed the study.

The St. John’s wort tablets used in the study were obtained from a German manufacturer and processed using an FDA-approved procedure that masked the herb’s taste and smell. Each tablet lot was assayed for hypericin content, which varied by no more than 5% among the batches. The amount of the purported antidepressive agent hyperforin in the tablets was not reported.

The study, which was funded by Pfizer, manufacturer of the antidepressant sertraline (Zoloft), sparked criticism from trade groups representing herbal product makers. Officials at the American Nutraceutical Association and the American Herbal Products Association questioned the study’s use of St. John’s wort in people diagnosed with severe depression. St. John’s wort, said the officials, is a mood enhancer intended for use by people who have mild to moderate depression.

The research article did state that patients with less severe symptoms of depression might benefit from St. John’s wort therapy but emphasized that well-designed, controlled clinical studies are still needed to address the issue.

  1. Shelton RC, Keller MB, Gelenberg A et al. Effectiveness of St John's wort in major depression: a randomized controlled trial. JAMA. 2001; 285:1978-86.