Dr. Bratman, co-author of the Natural Pharmacy: Clinical Evaluation of Medicinal Herbs, explained the intricacies of the interpretation of clinical studies of this type. "In many studies of antidepressants, perhaps as many as one-third, the tested drug doesn't do any better than placebo. The cause is probably a combination of the high placebo effect often seen in studies testing antidepressants and the relative coarseness and subjectivity of the type of rating scales that must be used to evaluate severity of depression." Dr. Bratman explained that these rating scales are by their very nature less precise than biomedical tests, such as those that measure cholesterol levels. "In consequence, it is quite easy for a truly effective anti-depressant, such as sertraline, to fail to prove efficacy in a given double-blind, placebo-controlled trial. The problem is not the treatment itself, but in the difficulties of studying such treatments."

Because this study included sertraline, a drug that previously has been shown effective, the inescapable conclusion is that details of the patient group and the methods by which the ratings were determined were such that this study could not discern the effectiveness of a known effective treatment. This also applies to St. John’s wort (which has also been shown effective in many studies) as it does to sertraline.

Bratman took issue with the authors of the study when they used a secondary outcome measure to conclude that sertraline was more effective than placebo and therefore better than St. John’s wort. According to the standard rules of interpreting clinical studies, one should take only the primary outcome measures as meaningful, he explained. On those measures, neither sertraline nor St. John’s wort was effective. Digging into secondary measures is widely accepted as being inappropriate, he added.

Thus, when the authors conclude that St. John’s wort is ineffective for moderate depression and shouldn't be used, noted Bratman, "it would be equally valid to say that sertraline is ineffective for moderate depression, and shouldn't be used," he added. "However, we know that this is not the case; looking at the body of published research as a whole, both sertraline and St. John's wort are effective."

"Herbs should be tested according to a reasonable expectation of their previously documented benefits," said Mark Blumenthal, founder and executive director of the nonprofit American Botanical Council. He referred to10 previous studies on St. John’s wort extract where the herb preparation was compared directly with pharmaceutical antidepressants for treatment of mild to moderate depression. These studies indicated a comparable efficacy with St. John’s wort and the conventional drugs.

He also noted, "In Germany many physicians use St. John’s wort as a first-line remedy for mild to moderate depression; if it doesn’t work, then they can always put the patient on more powerful, pharmaceutical anti-depressants."

Blumenthal added, "It is important for the NIH to continue conducting clinical trials on many popular herbal dietary supplements. This adds to the growing body of scientific information on their safety and efficacy. This process is a constructive contribution to the maturation of the herbal movement. At the same time, however, it is equally important that the results of these studies are accurately interpreted and communicated to the public. Unfortunately, it appears that some aspects of this first NIH-funded study are not being properly characterized." He also noted that the full text of the study is not available prior to this Wednesday, so it is not possible to adequately evaluate all the details.