Two kinds of suicide

Gregg Easterbrook makes an argument about suicide and the selective serotonin reuptake inhibitors (SSRIs — the class of antidepressants including Prozac and Zoloft) that seems plausible, though I’m not qualified to judge the underlying factual claims:

1. The use of SSRIs precipitates suicide in a small but non-trivial number of patients.

2. But widespread SSRI use has coincided with a decline in the overall suicide rate.

3. Untreated depression is a major cause of suicide.

4. Therefore, SSRIs probably prevent many more suicides than they cause.

5. Therefore, scaring people into no using them, or scaring manufacturers into not making them, with a combination of publicity and lawsuits would result in more suicides, not fewer.

Again, I’m not sure that’s right, but it might well be right, and it shows how poorly adapted our existing journalistic and legal mechanisms are to making policy about health care.

But Easterbrook segues from that sensible argument to what seems to me a not-very-sensible assertion:

Suicide should be viewed in public-health terms–as a disease with symptoms and treatment, requiring programs, awareness campaigns, and public outreach.

That assumes that suicide is one bad thing; in fact, it’s at least two things, only one of them bad. (See Update below.)

The bad suicide — the part of suicide that can reasonably be conceived of as a public-health problem — is the impulsive suicide, committed by a person who didn’t want to kill himself last week and who won’t want to kill himself next week if his current attempt is prevented or fails. Victims of that sort of suicide tend to be younger and to be suffering from various diagnosable mental health conditions, especially depression and anxiety. This class of cases ought to be managed with suicide prevention combined with treatment for the underlying disorders. (In addition, there are impulsive suicides stemming from sudden disasters; for these, prevention is still in order but there may be no underlying disorder to treat.)

But some suicides don’t look like that at all. They’re committed by people who, due to intense and prolonged misery, have decided that their lives would be happier if they ended now rather than later. The subjects — I see no reason to call them “victims” — of such acts of voluntary termination tend to be older, have more physical and social problems, and to have fewer psychiatric problems than those who submit to a transient impulse to kill themselves.

These suicides — call them “considered,” in contrast with the bad “impulsive” suicides — don’t look to me like a public health problem, or indeed a problem of any other kind. Suicide is indeed forbidden by most of the world’s religious traditions, but that doesn’t make it a medical problem. (Hume makes what seems to me a convincing case that the arguments offered in support of a supposed divine ordinance against suicide don’t hold water.)

That’s not to say that suicide may not point to real problems. If chronic pain victims commit suicide at high rates, which they do, that’s a convincing critique of our existing mechanisms of pain treatment. But when someone decides, after reflection, that his or her life is more trouble than it’s worth, it shouldn’t be the job of the public health community or the state to overrule that decision.

There are reasons to worry that a relaxation of the suicide taboo and of the laws that support it — in particular the “Kevorkian” laws that criminalize even the provision of the means by which someone else can kill himself — might over time create social pressure for those whose illness and debility make them a burden on others to do away with themselves. The existence of such social pressure would constitute a problem, and one worth attending to through public policy.

But adding considered and impulsive suicides together to make a suicide problem comparable to the problem of automobile accidents seems to me a mistake. So does using the fact that some people kill themselves with guns as an argument in favor of restricting gun ownership, without first deciding how many of those gun suicides were impulsive rather than considered.

Keeping people from doing away with themselves on impulse or under social pressure is a valid goal for public health and public policy. Meddling in people’s considered decisions not to go on living isn’t; there, the role of the state ought to be to get out of the way. The distinction between the two classes of cases probably isn’t a sharp one even conceptually, and it’s certainly hard to draw empirically. But that doesn’t give the public health folks a license to treat decisions as diseases.

Update A reader points out that the distinction between considered and impulsive acts is at least partly misleading, since someone in the grip of depression or anxiety disorder may plan a suicide carefully over a long period. A suicide committed not on impulse, but with impaired decision-making capacity, can be just as tragic and just as preventable as an impulsive act. So the right distinction is between “well-considered” acts, planned over time with relatively unimpaired mental faculties, and “poorly-considered” acts done either rashly or under the burden of mental illness.

Ideally, those judgments would be made looking only at the form of the decision rather than its content. In fact our judgment about whether a suicide is well-considered or not will have to depend to some extent on the substance of the reasons assigned for it.

The same reader suggests that anti-suicide policies are largely harmless, because those truly intent on taking their own lives will likely succeed. I strongly doubt it. Loss of strength and motor control, loss of mobility, and confinement to a nursing facility all make successful suicide much more difficult.