Psychiatric Comorbidity, Suicidality, and In-Home Firearm Access Among a Nationally Representative Sample of Adolescents

Importance  Suicide is the second leading cause of death among US adolescents, and in-home firearm access is an independent risk factor for suicide. Given recommendations to limit firearm access by those with mental health risk factors for suicide, we hypothesized that adolescents with such risk factors would be less likely to report in-home firearm access.

Objectives  To estimate the prevalence of self-reported in-home firearm access among US adolescents, to quantify the lifetime prevalence of mental illness and suicidality (ie, suicidal ideation, planning, or attempt) among adolescents living with a firearm in the home, and to compare the prevalence of in-home firearm access between adolescents with and without specific mental health risk factors for suicide.

Design, Setting, and Participants  Cross-sectional analysis of data from the National Comorbidity Survey–Adolescent Supplement, a nationally representative survey of 10 123 US adolescents (age range, 13-18 years) who were interviewed between February 2001 and January 2004 (response rate 82.9%).

Exposures  Risk factors for suicide, including a history of any mental health disorder, suicidality, or any combination of the 2.

Main Outcomes and Measures  Self-reported access to a firearm in the home.

Results  One in three respondents (2778 [29.1%]) of the weighted survey sample reported living in a home with a firearm and responded to a question about firearm access; 1089 (40.9%) of those adolescents reported easy access to and the ability to shoot that firearm. Among adolescents with a firearm in home, those with access were significantly more likely to be older (15.6 vs 15.1 years), male (70.1% vs 50.9%), of non-Hispanic white race/ethnicity (86.6% vs 78.3%), and living in high-income households (40.0% vs 31.8%), and in rural areas (28.1% vs 22.6%) (P < .05 for all). Adolescents with firearm access also had a higher lifetime prevalence of alcohol abuse (10.1% vs 3.8%, P < .001) and drug abuse (11.4% vs 6.9%, P < .01) compared with those without firearm access. In multivariable analyses, adolescents with a history of mental illness without a history of suicidality (prevalence ratio [PR], 1.13; 95% CI, 0.98-1.29) and adolescents with a history of suicidality with or without a history of mental illness (PR, 1.20; 95% CI, 0.96-1.51) were as likely to report in-home firearm access as those without such histories.

Conclusions and Relevance  Adolescents with risk factors for suicide were just as likely to report in-home firearm access as those without such risk factors. Given that firearms are the second most common means of suicide among adolescents, further attention to developing and implementing evidence-based strategies to decrease firearm access in this age group is warranted.

In 2010, suicide was the second leading cause of death among adolescents 13 to 18 years old in the United States (5.38 deaths per 100 000).¹ Approximately 12% of adolescents report a history of suicidal ideation, and 4% have attempted suicide.² Firearms account for almost half of all suicides among this age group^1,3 and are the most lethal manner by which suicide is attempted.³ Approximately 90% of suicide attempts by firearm result in death.^3,4

The presence of a firearm in the home is an important risk factor for suicide.⁵ Individuals living in households with firearms are more likely to make suicide plans involving a firearm,⁶ and in case-control studies^7-13 the presence of a firearm in the home was consistently associated with suicide deaths, especially among adolescents. In areas where firearm ownership rates are high, suicide rates are also high.^14,15

Psychiatric risk factors for adolescent suicide, including substance use disorders, mood disorders (eg, major depression), and a history of suicidal behavior, are well established.^16,17 However, firearm availability may increase the risk of suicide above baseline psychiatric risk.^18,19

Safe firearm storage practices (eg, storing firearms unloaded, locked, and separate from locked ammunition) to decrease firearm access by adolescents are also important. Suicide attempts are often impulsive acts²⁰ and frequently occur after an acute stressor (eg, an interpersonal dispute).²¹ Twenty-four percent of 153 adolescents and young adults who survived an almost lethal suicide attempt reported that less than 5 minutes elapsed from the time when they decided to end their life and when they attempted suicide.²⁰ Therefore, safe storage practices may decrease suicide risk by reducing immediate access to lethal means. In fact, investigators have observed a lower suicide risk in households where safe storage practices exist than in those where firearms are stored unlocked or loaded.^5,12,22,23

In the United States, approximately 1 in 3 adolescents lives in a home with a firearm.^24-26 Results from surveys of adult firearm owners have suggested that a large proportion of those firearms are stored unlocked or loaded,^25,27,28 and a study²⁹ based on 1995 data found that almost 25% of adolescents have access to a firearm in the home.

Given the association between firearm access and the risk of injury and suicide among adolescents, professional organizations such as the American Psychological Association and the American Academy of Pediatrics have long recommended addressing safe storage practices with families, particularly those with at-risk youth in the home.^30-33 If these recommendations are effectively implemented, the prevalence of firearm access should be lower among adolescents at risk of suicide. While the relationship between mental health and in-home firearm access has been studied in adults,³⁴ little information exists regarding firearm access among adolescents with mental health disorders, arguably the group most at risk of self-inflicted firearm injuries. Understanding the prevalence of in-home firearm access among adolescents with and without mental illness is important for informing public policy and designing interventions. Aims of this study were (1) to estimate the prevalence of self-reported in-home firearm access among a 2001 to 2004 nationally representative sample of US adolescents, (2) to quantify the lifetime prevalence of mental illness and suicidality among adolescents living with a firearm in the home, and (3) to compare among adolescents with a firearm in the home the prevalence of in-home firearm access between adolescents with and without specific mental health risk factors for suicide.

The human subjects committees of Harvard Medical School, Boston, Massachusetts, and the University of Michigan, Ann Arbor, approved all original National Comorbidity Survey–Adolescent Supplement (NCS-A) procedures. Participants’ parents or guardians provided written informed consent, and participants provided written assent. This analysis of deidentified data did not require institutional review board approval by the University of Washington Human Subjects Division.

The NCS-A³⁵ is a nationally representative survey of 10 123 adolescents (age range, 13-18 years) who were interviewed between February 2001 and January 2004. The survey used a dual-frame sampling strategy to recruit participants from National Comorbidity Survey Replication³⁶ households (n = 879) and from schools located in counties representative of those in the replication survey (n = 9244). The overall response rate was 82.9%. Poststratification weights were applied to address within-household probability of selection and differences in respondent characteristics compared with those of the US population. Survey weighting and methods have been described in detail elsewhere.³⁵

We abstracted the following sociodemographic variables for each respondent: age, sex, ratio of family income to the federal poverty level, parental education, number of biological parents living in the home, level of urbanicity, and US census region. We also abstracted self-reported race/ethnicity because firearm ownership and storage patterns vary by this characteristic.³⁷ Race/ethnicity classifications initially defined by the NCS-A were categorized as non-Hispanic white, non-Hispanic black, or other.

The NCS-A assessed participants’ mental health histories using a modified version of the World Health Organization Composite International Diagnostic Interview.³⁸ The Composite International Diagnostic Interview was validated among adolescents, and diagnoses are based on the DSM-IV.³⁹ We used diagnostic hierarchy rules to identify specific lifetime and recent (preceding 12 months) mental health disorders, including mood, anxiety, behavioral, and substance use disorders.⁴⁰ We also quantified the total number of lifetime mental health disorders for each participant (0, 1, 2, or more).

The NCS-A assessed any history of suicidal thoughts and behaviors using a modified version of the suicidal behavior module of the Composite International Diagnostic Interview.^38,41 We identified lifetime prevalence of suicidal ideation using the question “[Have you] ever seriously thought about killing yourself?” Of those who reported suicidal ideation, we identified lifetime suicide planning and attempts using the questions “[Have you] made a plan for killing yourself” and “[Have you] tried to kill yourself?” We identified recent suicidal ideation, planning, and attempts using similar questions specific to the preceding 12 months.

We identified the presence of 1 or more functioning firearms in the home using the question “How many guns that are in working condition do you have in your house, including handguns, rifles, and shotguns?” Among those with at least 1 firearm in the home, we assessed firearm access using the question “Could youget [the gun/one of the guns] and shoot it right now if you wanted to? Or is either the gun or the ammunition put away where you can’t get it?” Potential responses were “could get it,” “could not get it,” and “could maybe get it with great effort.” We defined access to a firearm in the home as yes (ie, could get it) or no (ie, could not get it) and categorized those who answered “could maybe get it with great effort” as missing (n = 135).

We estimated the proportion of US adolescents with in-home firearm access using the full, weighted NCS-A sample. We then described sociodemographic characteristics and mental health histories of respondents who reported at least 1 functioning firearm in the home and answered the firearm access question (yes or no). We defined our outcome of interest as self-reported in-home firearm (yes or no) and used t tests and χ² tests to compare differences in continuous and categorical variables, respectively, between adolescents with and without firearm access. The exposures of interest were a lifetime history of any mental health disorder, a lifetime history of suicidality (ie, suicidal ideation, planning, or attempt), or any combination of the 2. We used Poisson regression with robust SE of the variance to obtain prevalence ratio estimates and corresponding CIs, and we compared the prevalence of in-home firearm access between adolescents without suicide risk factors (ie, no lifetime mental health disorders or suicidality) and adolescents with (1) a lifetime history of a mental health disorder without a history of suicidality and (2) a lifetime history of suicidality with or without a history of a mental health disorder. In each model, we included the following covariables based on their a priori relevance: age, sex, race/ethnicity, ratio of family income to the federal poverty level, level of urbanicity, and US census region.

To determine whether the relationship between suicide risk and firearm access differed by age or sex, we analyzed 2 separate multivariable models. These included the following interaction terms: (1) age (continuous variable) and suicide risk category and (2) sex and suicide risk category.

To determine whether the relationship between suicide risk and firearm access was dependent on proximity or severity of mental illness, we performed 2 subgroup analyses. First, we repeated our main analyses using only 12-month histories of mental health disorders and suicidality. Second, we compared the prevalence of in-home firearm access between adolescents with no lifetime history of suicide risk factors and adolescents who reported a suicide attempt in the preceding 12 months. We used STATA statistical software (version 12.0; StataCorp LP) for all analyses.

Overall, 2921 (30.7%) of the weighted NCS-A sample reported 1 or more functioning firearms in the home and 2778 (29.1%) reported 1 or more functioning firearms in the home and responded to the firearm access question. Of those, 1089 (40.9%) reported easy access to and the ability to shoot that firearm (Table 1).

The mean age of the weighted study sample was 15.3 years; 58.7% were male, 81.7% were of non-Hispanic white race/ethnicity, 34.3% lived in major metropolitan areas, and 44.5% lived in the South (Table 1). Compared with adolescents without firearm access, those with access were significantly more likely to be older (15.6 vs 15.1 years), male (70.1% vs 50.9%), of non-Hispanic white race/ethnicity (86.6% vs 78.3%), and living in high-income households (40.0% vs 31.8%) and in rural areas (28.1% vs 22.6%) (P < .05 for all).

Fifty-one percent of the weighted study sample met criteria for at least 1 lifetime mental health disorder, and 28.9% met criteria for at least 2 lifetime disorders (Table 2). Compared with adolescents without firearm access, a greater proportion of those with firearm access met criteria for 1 or more lifetime mental health disorders (53.6% vs 49.2%, P = .04). Compared with adolescents without firearm access, a greater proportion of those with access had a recent history of alcohol abuse (8.2% vs 2.9%), a lifetime history of alcohol abuse (10.1% vs 3.8%), and a lifetime history of drug abuse (11.4% vs 6.9%), and a smaller proportion had a lifetime history of panic disorder (0.7% vs 3.0%) (P < .01 for all).

Thirteen percent of the weighted study sample reported a lifetime history of suicidal ideation, and 4.1% reported a lifetime history of suicide attempt (Table 2). Recent suicidal ideation and attempts were reported by 5.8% and 1.8% of the sample, respectively. No statistically significant differences were found in recent or lifetime suicidal ideation, planning, or attempts between adolescents with and without firearm access.

In unadjusted and multivariable analyses, there were no statistically significant differences in firearm access between adolescents without suicide risk factors (ie, lifetime mental health disorder or suicidality) and either those with a lifetime history of a mental health disorder without suicidality or those with a lifetime history of suicidality with or without a history of mental health disorder (Table 3). In summary, older age, male sex, non-Hispanic white race/ethnicity, and residence in a nonmetropolitan location were independently associated with increased likelihood of reporting in-home firearm access (eTable in the Supplement). No statistically significant interactions were found between age or sex and suicide risk categories in their effect on firearm access.

In subgroup analyses limited to recent (preceding 12 months) mental health disorders and suicidality, there were no statistically significant differences in firearm access between adolescents with and without suicide risk factors. There was no statistically significant difference in firearm access between adolescents without a lifetime history of a mental health disorder or suicidality and adolescents who reported a recent suicide attempt.

In this cross-sectional analysis of a nationally representative sample of US adolescents, almost one-third reported living in a home with a firearm. Of those, approximately 40% reported the ability to access and shoot that firearm. Adolescents living in households with firearms had a significant burden of mental illness, and those with increased risk of suicide were just as likely to report in-home firearm access as those without specific suicide risk factors.

Our estimate of firearm prevalence among households with adolescents is consistent with estimates reported elsewhere.^25,26 In addition, our estimate of adolescent in-home firearm access is similar to estimates based on storage practices reported by adult firearm owners in other national samples.^25,26,42 To our knowledge, the most recent estimate of adolescent-reported firearm access is from a study²⁹ of 1995 data finding that 24.3% of US adolescents reported in-home firearm access, almost double our estimate. Reasons for this discrepancy are unclear. The NCS-A may have identified a more specific subset of adolescents by asking about both immediate access (ie, “right now”) and the ability to shoot the firearm. Regardless, our study confirms that a large proportion of US adolescents have in-home firearm access.

This study makes several important contributions to the firearm access literature. Most studies^25,26,28 have focused on parent or guardian report of storage practices as a proxy for adolescent firearm access (eg, storing a firearm unlocked). However, such reports tell us little about whether adolescents can access the firearm and, perhaps more important, whether adolescents are aware that they can access the firearm. By focusing exclusively on adult storage practices, prior studies^25,26,28 fail to address an important aspect of firearm access: adolescents may be able to access and shoot a firearm even if that firearm is stored locked, unloaded, and separate from locked ammunition.

To our knowledge, this is the first study to comprehensively describe the burden of mental illness among adolescents with firearm access. We found that the prevalence of most mental health disorders and suicidality was similar between adolescents with and without firearm access and, most important, was similar to previously reported estimates of these disorders among the general US adolescent population.^2,40 However, those with access were more likely to endorse a history of alcohol and drug abuse, findings that are particularly concerning given prior research demonstrating greater odds of substance abuse and firearm access among suicide decedents.^43,44 This is consistent with previous studies finding that 10% of US adolescents have access to both firearms and alcohol in the home²⁹ and that parents with substance abuse problems were more likely to report storing a firearm unsafely than parents without such problems.⁴⁵ Given that alcohol and drug–related disorders cluster within families,⁴⁶ these findings may help explain the increased firearm access among adolescents with substance abuse histories.

Because of recommendations to limit firearm access among adolescents with suicide risk factors,^31,33 we hypothesized that such adolescents would be less likely to report firearm access than those without suicide risk factors. However, we observed no difference in firearm access between risk groups even in subgroup analyses limited to adolescents who had recent mental health disorders or those who reported a recent suicide attempt compared with adolescents who had no lifetime risk factors. This suggests that many parents or guardians of adolescents with suicide risk factors and a firearm in the home may not be aware or convinced that (1) their child has access to the firearm,⁴⁷ (2) firearm access is a risk to their child, or (3) limiting firearm access may mitigate the risk of having a firearm in the home. The prevalence estimates of DSM-IV disorders obtained by the NCS-A do not represent formal clinical diagnoses. Therefore, these findings may also represent unrecognized mental health risk factors among US adolescents.

Although the objective of this study was to address firearm access among adolescents at risk of suicide, these findings have several other important implications. First, adolescents and adults living in homes with firearms, particularly loaded or unlocked firearms, are disproportionately at risk of other firearm-related injuries, including unintentional shootings and homicide.^22,48 Second, we estimated that almost 1 in 4 adolescents living in a home with a firearm has a lifetime behavioral disorder such as intermittent explosive disorder or conduct disorder. Several of these illnesses are associated with increased risk of personal injury or violent behavior.^49,50 The implications of firearm access among this population are unclear. Third, these analyses are based on data collected from 2001 to 2004. That we have identified a high prevalence of firearm access among adolescents with suicide risk factors underscores the importance of reassessing this issue among the present US adolescent population.

These findings present an area for intervention by policy makers, health care systems, health care professionals, and parents. Safe storage practices may reduce the likelihood of firearm injuries among adolescents,^12,22,51 are recommended by professional organizations,^31,33 and are a focus of the Institute of Medicine’s recommendations for firearm injury prevention.⁵² The limited literature on this subject demonstrates that most parents are unlikely to comply with health care professionals’ recommendations to remove firearms from their home.^53,54 However, studies^55,56 have shown that health care professional and clinic–based firearm interventions to improve firearm storage practices are acceptable to parents, and a growing body of evidence suggests that brief interventions delivered in clinical settings can decrease adolescent firearm risk behaviors, including alcohol use problems and weapon carrying.^57,58 In fact, community and clinic–based randomized interventions have demonstrated efficacy in improving safe storage practices among firearm owners, including parents with children in the home.^54,59-63

This study has several limitations. First, these analyses are based on data collected from 2001 to 2004 and do not reflect changes in firearm ownership or access since that time. Second, we defined adolescents at risk of suicide using any history of a mental health disorder or suicidality. Most important, some adolescents have completed suicide without identified histories of mental illness.⁷ Third, these estimates are based on adolescent reporting of firearm access and are subject to misclassification bias if adolescents incorrectly believe that they do or do not have access to a firearm. Fourth, parent report of adolescent mental health disorders has been shown to improve diagnostic accuracy in the NCS-A sample, particularly for behavioral diagnoses. Although lack of parent report may have led to underestimates of some disorders, it is unlikely that differential misclassification between adolescents with and without firearm access biased these results.

In summary, a substantial proportion of US adolescents reported in-home firearm access, and those with suicide risk factors were just as likely to report access as adolescents without such risk factors. Given the significant morbidity and mortality associated with firearm injuries among this population, particularly by self-directed violence, further attention to developing and implementing evidence-based strategies to decrease firearm access among adolescents is warranted.

Submitted for Publication: March 20, 2014; final revision received July 21, 2014; accepted July 29, 2014.

Corresponding Author: Joseph A. Simonetti, MD, MPH, Veterans Affairs Puget Sound Health Care System, 1100 Olive Way, Ste 1400, Seattle, WA 98101 (simonja@uw.edu).

Published Online: December 30, 2014. doi:10.1001/jamapsychiatry.2014.1760.

Author Contributions: Dr Simonetti had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: Simonetti, Mackelprang, Rowhani-Rahbar.

Drafting of the manuscript: Simonetti, Mackelprang.

Critical revision of the manuscript for important intellectual content: Rowhani-Rahbar, Zatzick, Rivara.

Conflict of Interest Disclosures: None reported.

Funding/Support: During the preparation of the manuscript, Dr Simonetti received fellowship support by grant T32HP10002 from the Ruth L. Kirschstein National Research Service Award, administered by the Health Resources and Services Administration, and Dr Mackelprang received fellowship support by grant T32HD057822 from The Eunice Kennedy Shriver National Institute of Child Health and Human Development. The National Comorbidity Survey is funded by grants R01 MH/DA46376 and R01 MH49098 from the National Institute of Mental Health, by a supplement to grant R01 MH/DA46376 from the National Institute on Drug Abuse, and by grant 90135190 from the William T. Grant Foundation.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the funding organizations. Dr Rivara is the Editor of JAMA Pediatrics and serves on the editorial board of JAMA. He was not involved in the editorial evaluation or decision to accept this article for publication.

Previous Presentations: Preliminary findings from this study were presented at the Pediatric Academic Societies Annual Meeting; May 4, 2014; Vancouver, British Columbia, Canada; and at the 47th Annual Conference of the American Association of Suicidology; April 10, 2014; Los Angeles, California.