A Depressing Debate About Antidepressants

A much publicized study released earlier this year purported to show, based largely on the results of the drug makers' own research, that anti-depressants such as Prozac and Paxil don't work much better than a placebo at ameliorating depression. I don't like the study very much, in part because even the authors' own engineering of the numbers actually shows that the serotonin reuptake inhibiting drugs are in fact effective in cases of major depression. That's something you may not have caught if you saw articles about the study by writers who didn't get past the press release.

One question that studies like this raise for me is whether say more about the drugs or about the criteria for the illness the drugs are supposed to address. Folks who read that antidepressants are not more effective than a placebo pill come away with the idea that these drugs don't do anything. But this is obviously untrue. Anybody who has taken antidepressants knows that they have dramatic effects. They certainly are not always the effects that you might want—they can be bad effects—but that they do something is very clear.

The gap between the results of studies that show little effect from drugs and the experience of people (some of whom swear by antidepressants, some of who detest them) who have used them underlines a more profound split between drugs and the criteria againt which they are tested. The expectation that a drug will have effects that neatly line up with the definitions of the DSM-IV yields the preposterous conclusion that the drugs have little effect when the problem of something like “depression” is so badly defined that it may well be made better or worse by the very same drugs in different people—or in the same people at different times.

Oddly, this situation benefits both critics of drugs such as antidepressants and their developers. Critics can point to studies that show unclear or unpromising results. Drug companies, on the other hand, feel little pressure to explain exactly how the drugs work or what they do, as long as they can come with a study that shows some difference in some condition that fits the DSM-IV standards. With each new antidepressant the endless debate about whether the drugs "work" continues, and we grow barely closer in understanding how they work and no closer at all in defining what we want them to do.