Last year the US Centers for Disease Control reported an 18% spike in suicide rates among children and adolescents between the years 2003 and 2004. This is a striking increase, particularly when you consider that rates in this group have fallen by 29% since 1990. Yet one swallow does not make a summer and the rise could have represented a single year anomaly.  Not so, according to a research letter published this month in JAMA. Data collected from the National Vital Statistics Systems show a repeat of the excess mortality due to youth suicide in the US in 2005, with 292 deaths in the 10 to 19 year age group. So what could be to blame? Let’s line up the prime suspects…

The Second Amendment: reducing access to guns might have an effect on suicide rates but doesn’t sit squarely with the right to keep and bear arms. Over half of all completed suicides in the United States involve the use of a gun and shooting yourself in the head is unlikely to leave you with the opportunity for a change of heart in the way that an overdose might. So it’s obvious that guns make suicide easier, but that wouldn’t explain the jump in numbers unless they are becoming more available.

The Internet: websites, and in particular social networking sites, have been cited as a contributory factor in a number of suicides. Online ‘suicide pacts’ have frequently been reported in Japan, and more recently in South Wales a cluster of adolescent suicides have been linked to sites such as Bebo and Facebook although the evidence for this seems pretty speculative. The sensational reporting of these stories in the media has been criticised and may in itself be more influential in leading to copycat acts. A reliance, even dependence, on the internet seems increasingly prevalent, but this phenomena isn’t limited to the United States, it’s universal. So far however, there’s little indication that an increase in youth suicide is happening elsewhere.

Falling antidepressant prescriptions: SSRI prescriptions dropped by 20% in 2004 after the FDA put black box warnings on antidepressants linking them with suicidality in young people. Within a year, the suicide rate goes up – cause and effect? Possibly, but bear in mind we’re looking at populations, not individuals and this type of observational data is susceptible to the ecological fallacy. This problem with studies of suicide has cropped up before. In the 19th century, it was noted that suicide rates were higher in Protestant countries than those with largely Catholic populations. The sociologist Emil Durkheim concluded that this reflected the greater social cohesion of Catholic society without considering alternative explanations, such as the coroner’s recording of death or increasing urbanization. Another problem with the black box hypothesis is that there is no evidence of an increase in suicide mortality elsewhere: similar regulatory warnings were issued in the UK in 2003, and prescribing rates fell accordingly. Yet suicide rates continue to decline. I’m not convinced that levels of antidepressant prescribing bear much relation to population suicide rates. We know that many depressed people don’t get them (or take them properly if they do), and many that don’t really need antidepressants do: it’s that ecological fallacy again.

So we have a few suspects but not enough evidence for a conviction. Given the complexities of suicidal behaviour, it may be that seeking proximal causes for a change in population rates will remain little more than guesswork. The switch in the UK from coal gas to natural gas is the only intervention I can recall that had a significant effect. So is there nothing to be done? On the contrary, we know that there is a group of adolescents with an increased risk of suicide, namely those with depression. We also know that while the drugs vs. psychotherapy debate rumbles on about which is the best treatment, only a minority of depressed adolescents get any treatment at all. Finally, in grappling with the question of why people commit suicide we shouldn’t ignore the other important question: how? Suicide is often an opportunistic or impulsive act and as the coal gas story demonstrates, putting hurdles in the way can reduce the odds of an attempt, particularly an attempt that’s likely to be lethal - this might mean locking your gun away if you really insist on having one. If you want to know more about this aspect of suicide take a look at the Means Matter website developed by the Harvard Injury Control Research Center.