Guns and Suicide

In their article published concurrently with this editorial, Reisch et al. (1) report that legislation enacted in 2003 in Switzerland that halved the size of the citizen army (in turn significantly reducing the number of service weapons available in households) was associated with a substantial reduction in the rate of suicide by firearm in the most directly affected group, men 18–43 years of age. The overall suicide rate in this group also declined. No such effect was observed in two control groups, men 44–53 years old and women 18–44 years old. Equally interesting was the minimal evidence for method substitution. The firearm suicide rate did not attenuate over the period 2003–2008 except for a small increase in railway suicides that was also seen in the control groups. To the extent that this rise in railway suicides may represent method substitution, it attenuated the decrease in suicide rate by only 22% over a period of 5 years. The striking impact of abruptly reducing the size of the Swiss army on suicide rates may be explained by the fact that at that time, about half of all homes in Switzerland were thought to have a firearm, and firearms were used in 39% of all suicides in men 18–43 years of age.

The findings are consistent with other studies of means restriction on such a scale, such as the detoxification of domestic gas (2). For example, in England and Wales, the gradual detoxification of domestic gas by lowering its carbon monoxide level probably explains why gas as a suicide method was used in 49.8% of suicides in 1958 but only 0.2% in 1977. This decline was also associated with a decline in overall suicide rates. Many years later, there was evidence of method substitution as rates of suicide by use of car exhaust fumes rose. It may be too early to know what the form and extent of method substitution will be in Switzerland, or how quickly it will become manifest, although railway suicides are rising within 5 years of halving the size of the army.

Another recent striking example of gun restriction was observed in New Zealand, where legislation introduced in 1992 in response to a lone gunman killing 13 people with a military-type assault weapon in 1990 required stricter licensing for gun ownership and locked storage for guns and ammunition. Police confirmed compliance by home visits. The result was a decline in suicide by firearm, and the decline was more pronounced for youths, perhaps because they generally do not have the key to the gun closet (3). This trend has continued until the most current period for which data exist: whereas prior to the legislation guns accounted for about 18% of all suicides in New Zealand, in 2010 they accounted for 8%. There was a transient reduction in the overall youth suicide rate (1993–1996). Strict gun control can help reduce firearm suicides, but it is less likely to affect national rates in countries where guns are used less commonly in suicides to begin with.

What lessons can we draw for the United States from these data? The Centers for Disease Control and Prevention reports that in 2010, 31,672 persons died from firearm injuries in the United States, accounting for 17.5% of all injury deaths that year. The main types of firearm-related deaths were suicide (61.2%) and homicide (35.0%). The majority of suicides in the United States are by firearm (58%). The age-adjusted death rate for firearm suicide reversed a multiyear trend by increasing 3.4% in 2010 from 2009, while the death rate for firearm homicide decreased 5.3%. Males are the main victims of death by gunshot (86%), and the gun homicide death rate for African American men 18–29 years of age is 25 times that of white men in that age group (4).

According to estimates from the Bureau of Alcohol, Tobacco, Firearms, and Explosives on gun ownership in the United States, some 34 million individuals own about 195 million guns, and an estimated 35% of all households have at least one gun. Although precise gun ownership rates are unknown, two proxy measures—sales of the magazine Guns & Ammo and the fraction of suicides that involve use of firearms—suggest that gun ownership in the United States has been declining since 1973 (5).

It is of note that the increase in suicides in midlife over the past decade or so involves more hanging/suffocation in men and more poisoning in women (6). One of the arguments made in favor of firearms is that they increase safety because they are a means for self-defense. However, Kellerman and Reay (7) reported that for every case of a firearm being used in a self-defense homicide, there were 37 firearm suicides, and the California Department of Justice (1989) reported that only 17% of people murdered in a private home were killed by a stranger. Guns appear to be a much bigger threat to those they are intended to protect than to potential external assailants. People living in a household with a firearm have almost five times the risk of dying by suicide than those in gun-free homes (8). Homes of people who have died by suicide have twice the gun ownership rate of nonfatal suicide attempters (9). Not surprisingly, regional firearm suicide rates correlate with regional gun ownership rates in the United States (10).

An indication that the Swiss experience may have particular relevance to the United States is the report (11) that the highest suicide rate in U.S. veterans is in the 18- to 34-year age group and that firearm suicides in male and female veterans are, respectively, 1.3 and 1.5 times the rates in the demographically adjusted general population. Given the high rates of suicide in U.S. veterans, and the growing numbers of veterans, the problem of firearm suicide deserves even more urgent attention in this at-risk group.

If the effect seen in Switzerland, where the rate of firearm suicide in 18- to 43-year-old men decreased from 9.9 per 100,000 per year to 7.26 per 100,000 per year (a 27% decrease), were to apply to the United States, where in 2010 the rate of firearm suicides was 11.06 per 100,000 in this demographic group (6,045 suicides, in a population of 54,639,456), the rate would decrease to 8.07 per 100,000 per year, saving 1,636 lives per year on average.

The very real possibility of saving the lives of so many individuals in the United States renders pointless the debate as to whether access to firearms should be reduced; instead, the vexing question for Congress is how to reduce access to guns. The benefits for the country in terms of lives saved are potentially enormous.