This study supports the hypothesis that states with a higher number of firearm-related laws had lower rates of firearm-related suicides among older adults compared with states with fewer firearm laws, building on previous work that has found this same relationship in the general population13. However, there appeared to be a slight increase in non-firearm-related suicides in this age group, which was also seen in a recent study by Ghiani et al (2019), although they observed no impact on state firearm-related laws on non-firearm suicides overall15. Those results differ from those reported here, which found a significant association between the number of gun laws and fewer overall suicides. These results may suggest that, in the presence of gun laws, older adults are finding other means to commit suicide, but not to the same level of lethality. Disrupting access to firearms may only be a useful part—but only a part—of a larger, more comprehensive approach to suicide prevention.
The protective association in firearm-related suicide was stronger when focusing on gun violence restraining orders (GVRO) laws that allow for seizure of weapons from people who exhibit dangerous behavior, revealing a 2.4% reduction in firearm-related suicides among older adults in states with GVRO laws compared to states without these laws. GVRO laws were not associated with non-firearm related suicides, which is as expected given the laws’ purview. We identified two prior studies that have looked at the impact of GVRO laws on the general population. Swanson et al (2017), in their focused study of Connecticut’s GVRO law, found that for every 10 to 20 gun seizures, one suicide was prevented19. Another study by Kivisto and Phalen (2018) similarly found a reduction in suicides among the general population in Connecticut and Indiana, with a 14% and 7.5% reduction in firearm suicides respectively, in the ten years following their enactment20. In this latter study, however, they found that whereas Indiana demonstrated an aggregate decrease in suicide, Connecticut’s estimated reduction in firearm suicides was offset by increased non-firearm suicides. State differences in suicide rate may be contributing to these observed differences, as the overall rate of suicide in Indiana is higher than in Connecticut26. Across the country, overall suicide rates vary up to fourfold, from 6.9 (District of Columbia) to 29.2 (Montana) per 100,000 persons per year26. Differences in regional cultures might also be important to suicidality—such as in the Mountain West, dubbed the “suicide belt,” where there is a pervasive “pick yourself up by the bootstraps” mentality27—as well as the means chosen by individuals to attempt or commit suicide, alongside substance use, job or financial problems and mental health which are other circumstances contributing to suicide. The results from this national study suggest that the net association of GVRO laws is with fewer overall suicides.
A recent study on state-level firearm ownership and suicides found that, among men, higher firearm ownership was associated with an increase in total and firearm suicide rates, and a decrease in non-firearm suicide rate, suggesting that completion of suicides may depend on access to lethal means, such as firearms9. Our data may provide additional evidence for such a trend as we found a negative association between total number of firearm laws and non-firearm related suicide rates among older adults. In fact, for both total firearm laws and for GVRO laws specifically, there was a net negative effect on suicide by any means. This is likely due to the predominant role that the availability of firearms as lethal and commonly used means plays in determining suicide rates in the United States.
Further, although our study does not differentiate between the impact on those with and without dementias—those with dementias being at higher risk of suicide due to symptoms of depression, impulsiveness, aggressiveness—the findings support a protective association of GVRO laws and total firearm laws to firearm-related suicide among the age groups most affected by a spectrum of dementia diagnoses. Future studies could elucidate the particular impact of these policies on dementia patients, including potential unintended consequences (e.g., delaying diagnosis or evaluation of symptoms out of concern for having firearm(s) confiscated; harmful downstream effects of involvement with law enforcement and the criminal justice system).
Importantly, policy level interventions should be complemented by patient-clinician level interventions, which involve increasing physician comfort with talking about guns in the home with patients and their families28. In one qualitative study of stakeholders at Veterans Affairs (VA) regarding mental health and suicide risk at the VA, nearly all patients felt that clinicians should routinely speak about guns with their patients although both patients and clinicians reported limited discussion29. Conversations should not only include discussion about whether or not to remove a weapon entirely, but also how to safeguard the home and environment, or at least having gun access with supervision by another family member or caregiver, with gun access restricted for the patient28. Just as clinicians have incorporated conversations about driving with older adults, it is critical to engage in the difficult conversation about gun ownership and gun safety. For those with dementia, discussions earlier in the course of disease is important as they are less likely to organize a suicide attempt and more likely to be supervised than later in the course of the disease. Conversations about firearm safety at home is also important as people may not be firearm owners themselves but be living in a household with one. At the individual level, clinicians must also be vigilant about addressing mental health symptoms and/or adjustment to cognitive decline. In one study analyzing coroner/medical examiner and law enforcement reports and suicide notes, despondency from cognitive functional decline due to dementia was believed to be a precipitant of suicide among older adults, particularly among those age 85 and older30.
Other efforts can focus on other subgroups of older adults vulnerable to suicide—for example, older men are both more likely to own firearms and more likely to commit suicide using lethal means. In the analysis of Connecticut’s GVRO law, 92% of gun removal subjects were male and 81% were cohabiting or married19. Therefore, even at the policy level, legislative efforts at reducing firearm ownership may reduce suicide rates differently among males and females. In one national study of firearm ownership and suicide rates, there was a strong relationship between firearm ownership and suicides by any means among male, but not female, individuals9, suggesting that policies reducing firearm ownership would reduce both total suicide rates and firearm-related suicides for males, but only firearm-related suicides for females.
There are several potential limitations to this study. First, because this is an associative study, we are unable to establish causality between the firearm legislation and firearm-related suicides. However, given that the study was conducted over a 5-year period, we believe this adds to the robustness of our findings. Second, although we assessed the presence or absence of certain firearm legislation, we were unable to assess the effectiveness of variation amongst these laws or the effectiveness of their enforcement. That is, wide variation in policies across states includes differences in statutory requirement of the number of officers required as co-affiants, number of officers required to present before a judge, and variations in the probation period during which guns can be confiscated or an appeal process can occur19. These differences may also influence enforcement, similar to how high-profile events may increase awareness of GVRO laws and instigation of firearm removal from an individual deemed dangerous to themselves or others. From 1999 when the Connecticut GVRO was first passed to 2006, there were only approximately twenty guns confiscated per year; this saw a dramatic rise in 2006 after high profile mass shooting events19. In this light, we may even be underestimating the impact of these laws as there is often a lag in effect and increased enforcement over time. Further, given the small number of states that have either type of GVRO law currently in place, we could not differentiate between the effects of GVRO laws that grant family members versus only law enforcement officers the authority to remove firearms from an individual. As GVROs gain traction in legislative chambers, these variations will require further study as it will be vital for state and local governments to have a model policy. Lastly, there is potential for firearm-related injury that is not captured in this study but would paint a more holistic picture of firearms’ impact on older adults and GVRO’s impact in mitigating these harms.