SAN DIEGO — If a patient with depression comes into the office and says that his antidepressant has stopped working, the drug you gave him probably was never working at all, Dr. Mark Zimmerman said at the annual meeting of the American Psychiatric Association.
That patient probably had a placebo response, said Dr. Zimmerman, director of outpatient psychiatric services at Rhode Island Hospital, Providence.
Dr. Zimmerman said he was interested in why antidepressants seem to “poop out” when patients take them long term, and so he conducted a meta-analysis of continuation studies.
He identified four extension studies—the only type of continuation study that can be analyzed for its placebo effect-related relapse; in these studies, the patients were treated for their acute depression with a selective serotonin reuptake inhibitor for 6–8 weeks, followed by a continuation phase in which patients continued to take their drug for up to an additional year.
Dr. Zimmerman pooled the studies' data and used a method first described in 1993 to estimate the percentage of cases that can be attributed to a loss of placebo response (Am. J. Psychiatry 1993;150:562-5).
Using that formula, he estimated that 84% of the patients who relapsed during the continuation period were most probably patients whose response was a placebo response.
“The bottom line is that, overwhelmingly, relapse in studies occurs in people who are placebo responders,” he said. “It is not due to receptor down-regulation or up-regulation.”
Dr. Zimmerman also noted that continuation studies are not clinical practice, and that in clinical practice placebo response rates are probably higher than the 24%–30% rate described in trials because patients have higher expectations than those enrolled in studies.
“More of our patients are placebo responders than in clinical trials, and perhaps we shouldn't attribute as much of their gain to the particular molecule they are taking,” he said.