In Defense of Antidepressants

Since I just featured Marcia Angel's piece questioning the effectiveness of psychiatric medications, it seems only fair that I also feature Dr. Peter Kramer's response in the Times, "In Defense of Antidepressants." Kramer can point to a lot of evidence that SSRIs work. He also criticizes the studies relied on by Irving Kirsch, which were done by pharmaceutical companies when seeking to get approval for their drugs. These studies, Kramer says, are often "quick and sloppy," and he complains that they often include people who were not really depressed. The inclusion of people who are not sick, he says, is what generates the high placebo response. This is an interesting point but it strikes me that it tends to support rather than attack the position of the critics; if neither patients nor their doctors can tell who is really depressed in the first place, how well can either evaluate the effectiveness of drugs that are supposed to treat depression?

Kramer prefers narrow studies, many of which have been done to test whether antidepressants are helpful for people with particular conditions: people recovering from surgery, people who are depressed because they have MS or some other neurological conditions, and so on. He cites many that show positive effects. The problem with this is the one cited by many critics of contemporary science: we only know about the studies that are published, and generally only studies with a positive finding are published. Recall that Kirsch had to use the Freedom of Information Act to get his hands on negative studies done by the drug makers. This is a huge problem all across contemporary science, and Kramer ignores it. Besides which, when there are a wide variety of studies with differing results, there is no real validity in citing a bunch of positive ones. The only rigorous method is to do a statistical meta-study, as Kirsch did, and until someone does one of the many small studies cited by Kramer we won't know the real significance of those results.

My biggest complain about Kramer is the way he characterizes the placebo response. As even he notes, there is a big placebo response (or just natural healing) with depression. Kramer says that with mild depression, 1/3 of patients get better without treatment and "over half" with SSRIs. Even with Kramer's figures, then, more people are cured by the placebo response than by antidepressants. But Kramer several times refers to the placebo response as being "all in your head":

One reason the F.D.A. manages to identify useful medicines is that it looks at a range of evidence. It encourages companies to submit “maintenance studies.” In these trials, researchers take patients who are doing well on medication and switch some to dummy pills. If the drugs are acting as placebos, switching should do nothing. In an analysis that looked at maintenance studies for 4,410 patients with a range of severity levels, antidepressants cut the odds of relapse by 70 percent. These results, rarely referenced in the antidepressant-as-placebo literature, hardly suggest that the usefulness of the drugs is all in patients’ heads.

But where else would a response to depression be? In the patients' feet? I can see no basis for distinguishing a "real effect" of drugs from a "psychological effect" or however Kramer would characterize the placebo effect. People get better or they don't. Depression is a mental condition, and it is connected with how people think and feel. That doesn't make it less "real" than heart disease, which also responds strongly to how people think and feel.

I remain agnostic. Many people believe they have been helped by SSRIs, including many of my friends, and many studies show effects beyond placebos. Other studies do not. What this all means remains to be seen. But certainly antidepressants are not the wonder drugs that they once seemed to be, and the people they help may be outnumbered by the people who would do just as well on Dr. Kilmer's Swamp Root.