For years, public health researchers have known that there’s an alarming correlation between the availability of firearms and the overall rate of suicide in the United States. Numerous papers have drilled down further, finding that gun-owning households are at a higher risk for suicide. But the issue is still up for debate for some skeptics, who argue that suicide trends can be better attributed to other variables like loneliness or “psychological issues.”
Their argument has been bolstered by a 2004 report on firearms and violence from the National Academies, which concluded that firearms “might be associated with suicide but have no direct effect.” Instead, the study pointed to possible “unmeasured confounders” — or variables — “associated with both access to firearms and the propensity to commit suicide.” In other words, people who own guns might share characteristics, apart from gun ownership itself, that lead to increased rates of suicide. In particular, the study names “isolation,” and the “mistrust of others.”
Now, a new study by a team of researchers from Harvard and Northeastern debunks that notion, putting the spotlight squarely on firearms, which are by far the most likely risk factor responsible for increased rates of suicide in gun-owning households.
Their paper, published this month in Epidemiologic Reviews, a leading public health journal, is a systematic review of four influential studies on firearms suicide. Using what’s known as a bias formula, the researchers set out to determine whether there could be a significant risk factor for suicide that was missed by the earlier studies.
“When you do the math, it is absurd that such a confounder could actually exist,” Matthew Miller, the study’s lead author, tells The Trace. “This study is just one more piece of evidence that the causal connection between living in a home with a gun and being at higher risk of suicide is really strong.”
Because suicides are more common in gun-owning households, any hidden variable that might account for that must also be more prevalent in gun-owning households. And because suicide rates are so highly correlated with gun ownership, for the numbers to work out, an overlooked variable would also have to be an even bigger risk factor than the presence of a firearm.
The rub, according to Miller, is that researchers don’t know of any variable that meets both of those criteria. Psychiatric and substance abuse disorders are potent risk factors for suicide, but they aren’t more prevalent among gun owners. Individuals with mental illnesses or addictions are no more likely to possess firearms than individuals without them, according to existing research. Nor do gun-owning households experience a higher rate of suicide attempts. (What’s true is this: Suicide attempts with a gun are 42 times more effective than suicides by overdose or poisoning.)
Death Threats, Harassment, Intimidation: The Occupational Hazards of Being a Gun Violence Researcher
Miller’s team also could find no evidence that the specific variables suggested in the National Academies’ report — isolation and mistrust of others — might possibly explain away the link between gun access and suicide. No study has ever shown that a mistrust of others presents a risk factor for suicide. Social isolation, on the other hand, has consistently been shown to increase the probability of suicide, but only makes it about two times more likely — not nearly enough to override the correlation between gun ownership and suicide. For that to happen, Miller and his colleagues found, any unmeasured variable would have to increase the risk of suicide by at least five times.
When the National Academies released its report on firearms and violence in 2004, it didn’t intend to dismiss the relationship between gun access and suicide. Little research on the subject had been published at that point, Miller says, so the existence of some unmeasured variable was a legitimate theoretical possibility. Today, however, the relationship has been so strongly established that Miller compares naysayers to tobacco companies in the mid-20th century, shortly after cigarettes had been found to cause lung cancer. In fact, he and his colleagues modeled their paper on studies from the 1950s that proved smoking was causally responsible for lung cancer, not some previously unknown genetic variable.
“With guns, just the same as with cigarette smoking, there’s a hierarchy of risk,” Miller says. “If you smoke half a pack a day, you’re at lower risk of lung cancer than if you smoke four packs a day. Guns stored locked and unloaded are slightly safer, but overall, people who live in homes with guns have three to five times the risk of dying by suicide compared to people who live in homes without guns. It’s a huge, huge risk factor.”