Richard Friedman, MD, recently wrote an article in the New York Times about our troubled relationship with antidepressants. He opens his article by asking how antidepressants might affect a patient’s psychological development and core identity.
It’s an interesting question, one which some have discussed in the past, but which is worth another attempt. Unfortunately, Friedman doesn’t actually explore this question. What he does explore, however, is the gap between what we know about the medications, and what we don’t know: the paucity of long term efficacy studies, and long term safety studies.
On the one hand, he observes, there are nearly no good studies on efficacy past a two year endpoint. And there is a short commentary on the warning concerning suicidality, in which he opines that while the risk of suicidality is actually quite small, the risk from the disease is greater. Suicide, he writes, is the third leading cause of death in adolescents.
Friedman follows the ambivalent pattern of many psychiatrists who muse on their field. He offers one or two paragraphs about antidepressants’ benefits, followed by another on the risks or the unknowns. And the obligatory words about needing more data, needing better monitoring.
In fact, SSRIs are the most studied class of medications in history. There are more studies, more data, and yes, more monitoring, done for SSRIs than anything else; additionally, there is more debate, both within psychiatry and by the general public, about the effects of these medications-- not just biologically, but socially and existentially.
The problem isn’t, as Friedman characterizes, insufficient information. The problem is the opposite: too much information, and too little context.
The question isn’t, “do they cause suicide?” The question is: can a pill affect complex behavior? If it can cause you to commit suicide, why doesn’t it cause you to bake a cake? How does a pill cause a series of complex behaviors to occur that results in suicide? Or, in the reverse, how does a pill change your mind about wanting to do it?
The question isn’t, “are antidepressants better than placebo for treating depression?” The question is: what does “treating depression” mean? A simple example is that most SSRIs are tested against a scale (e.g. The HAM-D) which has three (out of 17) questions about insomnia, but none about hypersomnia. Is it conceivable that a medication can treat insomnia, and thus be considered artificially good for treating depression? Or better than something that doesn’t treat insomnia? You bet.
Most psychiatrists will agree that depression that results from no clear cause (“I have nothing to be depressed about, yet I am”) is different than depression due to a cause (e.g. divorce.) Surprisingly, however, there are almost no studies differentiating a drug’s efficacy in these two scenarios, whether Zoloft is better for divorce depression than it is for no-reason depression; or if Prozac is better than Zoloft for divorce depression, etc. Everything is lumped together; all the depressions are the same; all the SSRIs have the same efficacy. It is almost never asked whether the uniform efficacy of SSRIs isn’t due to the fact that they are all tested against the same scales; and if new ones were invented (e.g. one with hypersomnia questions) some SSRIs wouldn’t be “better” than others.
Simultaneously, an inordinate amount of energy is spent deciding whether a depression is unipolar or bipolar depression-- and medication selections vary considerably depending on how you decide.
The treatment of depression hasn’t stagnated because of a lack of study; there are easily ten psychiatric journals packed to the covers every month with antidepressant studies and commentary. It has stagnated because we are asking the same questions, and finding the same answers, over and over. Friedman did it, too, in his commentary: he asked an interesting question, but answered the usual one.