Bruce Levine’s article in the latest issue of Skeptic Vol. 17, No. 2: “Depression Treatment: What Works and How We Know” (pp. 23–27) paints a bleak view of depression treatment. He destroys all options except for the patient to congratulate himself on his critical thinking skills and to find something he can believe in. I think the reality is more encouraging: treatment can help.
I think Levine goes too far when he says antidepressants don’t outperform placebos. He cites Erick Turner’s study of FDA data showing that negative studies are less likely to be published, thereby skewing the published data; but he leaves out an important part of the story. After incorporating the data from the unpublished studies on 12 different antidepressants, Turner found that every one of the 12 still clearly beat the placebo.
Irving Kirsch looked at a subset of the 12 drugs in Turner’s analysis and came to a different conclusion: that antidepressants were no more effective than placebo. The difference in their conclusions hinges on the interpretation of “effect size.” It’s a bit complicated. I explained it in more detail in an article on ScienceBasedMedicine.org but basically, Kirsch used an arbitrary cutoff of 0.5, and when the studies failed to meet that mark he interpreted that as “antidepressants don’t work.” Turner looked at the actual effect sizes and found a positive effect for every antidepressant drug, with an overall effect size of 0.31. In an elegant metaphor, he explains that the published studies had sold us a glass of juice advertised to contain 0.41 liters, and the actual glass of efficacy (that included the unpublished data) contained only 0.31 liters. Kirsch considers the glass to be empty; Turner correctly concludes that the glass is far from full, but it’s also far from empty. He further points out that patients’ responses are not all-or-none and that partial responses can be meaningful.
Psychiatry bashing is a popular sport. There is much wrong with psychiatry, but it doesn’t deserve all the criticism directed at it. Marcia Angell was critical of psychiatry in her two-part article for The New York Review of Books where she reviewed Kirsch’s book and two others. I have critiqued her article on ScienceBasedMedicine.org, where I go into much more detail than space permits here. I’ll just make a few points. Depression is a complicated subject. The chemical imbalance model of depression is too simplistic. Drugs have side effects. They are often prescribed inappropriately for minor depression and the problems of everyday life. But conversely, many depressed patients suffer unnecessarily because they are untreated or undertreated. Severe depression is a life-threatening illness, and treatment with antidepressant medication has been shown in several studies to significantly reduce the risk of suicide. While psychotherapy is an alternative to drugs, antidepressants may be needed at first to lift the depression enough that patients are able to participate in psychotherapy.
There’s a baby in that depression treatment bathwater; let’s not throw it out just yet!
