When the drug companies started testing their SSRIs, the results were weak. Eventually, the drugs would flunk about half of their tests and pass the others with a gentleman’s C. But it didn’t really matter. FDA approval is binary. Either your drug is approved or it isn’t; its license doesn’t have an asterisk. It’s like the old question about what you call the guy who graduated last in his medical school class. (Doctor)
Not only that, but they had no reason to think that anyone would ever know the truth of their pallid results. No one except regulators, that is. FDA bureaucrats did briefly wonder if it was a good idea to approve one drug–Celexa–based on such weak studies, but they decided that since they’d already approved Prozac, Paxil, and Z0loft based on similarly dismal numbers, they really didn’t have any choice but to keep doing it.
But then Irving Kirsch and some colleagues used the Freedom of Information Act to dig up the studies mouldering in the FDA’s vault, did the analysis, and, in 2002, published the results. Subsequent studies, like the one that came out in January, have expanded on this finding, most recently to show that the drugs work primarily for severely depressed patients–which to me, means mostly that the Hamilton does what it is supposed to do: measure the effects of antidepressants on severely depressed patients. And, no surprise, in the few head-to-head studies, the tricyclics like imipramine, the drugs for which the Hamilton was invented, do better against severe depression than the SSRIs. (And shock treatment does best of all, but that’s a whole different problem.)
Kirsch’s conclusion from his research–that placebo effects explain antidepressant action–is reasonable, but for one thing. It’s not correct. It flies in the face of what we know about the placebo effect (which is, of course, very little, research money for a sugar pill a poor investment for an industry dependent on patents): that placebo effects tend to wear off relatively quickly. But there are millions of people happily taking their SSRIs, people who weren’t severely depressed in the first place, who put up with the trouble and expense and side effects to do so, and who find life without the drugs unpleasant enough not to stop.
A much more reasonable conclusion is that whatever SSRIs are doing for these people (and they are surely doing something–drugs that tweak serotonin, like LSD and Ecstasy and psilocybin are pretty psychoactive, after all) is not showing up on Max Hamilton’s test. Using the Hamilton to find out what the drugs do to the walking wounded is like using a magnet to find feathers. So in this respect the industry is simply reaping the rewards of its own laziness, hoist on their own petard, as the Bard would have said. And the placebo effect, while real, remains something different from the drug effect, whose contours are still poorly understood.
But can this probelm be solved? Can the drug companies come up with an instrument to measure what SSRIs actually do in the people who benefit from them? No doubt. In fact, some of this research has already been done, like a recent study that showed that the drugs affect personality, lowering something called neuroticism and increasing something else called extraversion. What made patients’ depression scores go down, these researchers said, was the fact that with their personalities changed. Their lives got better with the change wrought by the drugs, just as someone suddenly able to play better tennis or sing the high notes might feel happier. Do the drug companies want to be in the business of changing personality? No, they do not. At least not officially. Because that’s what bad drugs–drugs like cannabis and LSD–do. We’re not supposed to change who we are with drugs. That would be cheating.
Pharma might change their tune if they can reverse engineer results like these, use them to create the “disease” constituted by the pre-drug personality. They’ll have to call it something less obvious than Prozac-deficit disorder, but that’s what it will be. And I wouldn’t be surprised if they’re working on it right now.
