Efficacy of antidepression drugs challenged.

A controversial new report challenges the view of many psychiatrists that antidepressant drugs act as powerful depression-busters. Standard antidepressants, as well as the new generation of antidepressants such as fluoxetine (Prozac), ease depression only about as well as or slightly better than placebo pills, contend psychologist Roger P. Greenberg of the State University of New York Health Science Center at Syracuse and his colleagues.

"The efficacy of antidepressants and other psychoactive drugs has been inflated because of problems with study designs," asserts SUNY psychiatrist Seymour Fisher, a collaborator on the project. Critics, however, say the new study is itself flawed by design problems.

Since psychological factors generate noticeable improvements among many depressed individuals given placebos, similar factors must also contribute to the comparable effectiveness of antidepressants, Greenberg's group argues in the October JOURNAL OF CONSULTING AND CLINICAL PSYCHOLOGY.

The fatal flaw in most antidepressant studies concerns the inability to keep participants and clinicians truly "blind" to who gets an active drug and who gets a placebo, the researchers hold. Physical reactions and side effects induced by antidepressants often stand in stark contrast to placebo treatment, they note.

More informative, but much rarer studies compare the effectiveness of a newer antidepressant with two control groups, one receiving a standard antidepressant and the other receiving a placebo, according to Greenberg and his associates. With this strategy, it becomes harder to identify recipients of active drugs based on side effects (which occur less often with newer antidepressants), and the experimenters' vested interest in establishing the efficacy of the standard drugs lessens, they maintain.

Greenberg's group used a statistical technique known as meta-analysis to combine the results of 22 independent antidepressant studies that included dual control groups. Participants in the studies suffered from moderate to severe depression. For both standard and newer antidepressants, clinicians reported modest patient improvement beyond that obtained with placebos, while patient ratings revealed roughly equal effectiveness for drugs and placebos.

This finding held steady when the researchers controlled statistically for the ratio of men to women in the sample, drug dosage, age of participants, and severity of depression.

Until larger studies with appropriate controls appear, claims that antidepressant drugs quell depression "need to be conservative," Greenberg's group asserts.

Several studies suggest that cases of severe depression with no clear external cause and featuring disturbances of sleep and endocrine function prove highly responsive to antidepressants, writes psychologist Michael Feinberg of Hahnemann University in Philadelphia in a comment accompanying the new report.

However, such patients did not derive any special benefits from antidepressants in the meta-analysis, responds Fisher.

Psychiatrist David J. Kupfer of the University of Pittsburgh disputes the methods used by Greenberg and his co-workers. Their analysis relied on studies ranging from three to six weeks long, but researchers must track patients for at least eight weeks to gauge accurately the effectiveness of antidepressants, Kupfer argues. Moreover, Greenberg's group chose studies that examined the least effective of the newer antidepressants rather than the widely prescribed Prozac, he adds.

No evidence suggests that longer antidepressant studies would yield different results, Fisher contends. And Prozac proves about as effective as other antidepressants, according to an unpublished meta-analysis by Greenberg and Fisher of more than a dozen fluoxetine studies.