Article Text

Download PDFPDF

Firearm carrying and concurrent substance use behaviours in a community-based sample of emerging adults
  1. Robert N Buschmann,
  2. John D Prochaska,
  3. Jacques G Baillargeon,
  4. Jeff R Temple
  1. Department of Preventive Medicine and Community Health, University of Texas Medical Branch, Galveston, Texas, USA
  1. Correspondence to Dr Jeff R Temple, Department of Obstetrics & Gynecology, University of Texas Medical Branch, UTMB Ob/Gyn, Route 0587, 301 University Boulevard, 3.110 John Sealy Annex, Galveston, TX 77550–0587, USA; jetemple{at}


Introduction This paper examines associations between high-risk gun carrying and substance use in emerging adults (ages 18–22). The coexistence of these high-risk behaviours in a general population of emerging adults can have disastrous consequences.

Methods Dating it Safe is an ongoing longitudinal (2010–2016) survey of emerging adults recruited from seven high schools in five south-east Texas-area school districts (current sample n=684). Multiple logistic regression modelling was used to examine the association between past-year use of legal and illegal substances and past-year firearm carrying for a reason other than sport or hunting.

Results 6% of emerging adults carried firearms in the past year, with most (68%) carrying for protection. Use of cocaine, hallucinogens, methamphetamine, ecstasy and prescription medications in the past year, as well as episodic heavy drinking in the past month, was associated with increased risk of carrying a firearm (p<0.05 for all). After controlling for covariates, hallucinogens (OR 2.81, 95% CI 1.00 to 7.81), ecstasy (OR 3.66, 95% CI 1.32 to 10.14) and prescription medications (OR 2.85, 95% CI 1.22 to 6.68) remained associated with firearm carrying. Episodic heavy drinking was associated with firearm carrying, but only for those who had five or more episodes/month (OR 3.61, 95% CI 1.51 to 8.66).

Conclusions In this community-based sample of emerging adults, firearm carrying, mostly for protection, was associated with a variety of past-year substance use behaviours. These findings extend previous research and suggest directions for further exploration of the clustering of high-risk behaviours in emerging adults.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Firearms were involved in over 33 000 deaths in the USA in 2013,1 and are often implicated as a leading cause of death for adolescents and emerging adults. Indeed, firearm violence is a leading cause of death for black males aged 15–34, the second leading cause for white and Hispanic males aged 15–34 and the second leading cause for black females aged 15–24.2 Some limited research indicates that firearm carrying increases the risk of harm to the firearm carrier. For example, firearm carrying in one urban area appeared to increase the risk of being shot in an assault, even among those with a chance to resist.3 In a sample of at-risk emerging adults, those who chose to carry firearms had an increased chance of being a victim of violence, though they believed themselves to be at lesser risk.4 Despite this public health crisis, the complex political environment of the last three decades has resulted in a dearth of firearm research—even basic information about gun access, distribution and carriage, especially among young adults, is not well understood.5

Results from the 2012–2013 Youth Risk Behavior Survey (YRBS) indicated that 5.5% of students in grades 9–12 carried a gun for at least 1 day in the past month, with generally higher rates among white males; this number has remained relatively stable since 1997.6 There is no recent national survey analogous to the YRBS that estimates the number of adults in the USA who recently carried firearms; instead, surveys such as the General Social Survey (GSS) or the Behavioral Risk Factor Surveillance System (BRFSS) ask about firearm ownership. The most recent GSS numbers indicate that 22.4% of adults personally owned a firearm in 2014, a trend that has been stable since 2006 after declining from a high value of 30.5% in 1985.7 Research using BRFSS statistics, the most recent of which are from 2004, found that 38.3% of adults owned a firearm, with considerable variation between states.8

Firearm carrying among emerging adults may pose additional risks, particularly with regard to the potential for dangerous assaults or other violence, when other common behaviours in this age group are considered. For example, adults between ages of 18 and 34 have the highest rate of episodic heavy drinking of any age group.9 In a previous analysis of eight states using the 1996–1997 BRFSS, heavy episodic drinking was associated with firearm carrying, and firearm owners were more likely to binge drink, drink and drive, and have 60 or more drinks per month.10 This same study found firearm carrying for protection—and having confronted another person with a firearm—to be relatively more common among younger individuals (18–34) than older individuals, though again, these data are now almost 20 years old.10 A review of 40 years of studies examining alcohol and firearm violence noted a pattern of associations between heavy drinking and death from firearms via suicide and self-injury, as well as associations between alcohol use and firearm carrying specifically among US adolescents.11 Firearm carrying has also been associated with illicit drug use in adolescents and emerging adults who came to the emergency room for assault-related injuries.12 Taken together, the literature suggests that younger adults might be more likely to carry firearms for reasons that could lead to assaults and injury, and likewise are more likely to engage in risky behaviours related to substance use. However, the relationship between firearm carrying and substance use remains relatively understudied in the general community, particularly in emerging adults. In addition, many studies do not focus specifically on firearm carrying for reasons other than hunting or sport shooting.

Given these gaps in our knowledge, understanding the relationship between substance use and firearm carrying among emerging adults holds particular public health relevance. Thus, we use a large, ethnically diverse, population-based cohort of community-dwelling emerging adults to examine associations between firearm carrying and concurrent substance use. We hypothesise that reported firearm carriage will be associated with the misuse of alcohol and drugs.


We examined data from Wave 5 of Dating it Safe, an ongoing longitudinal cohort study of individuals originally recruited in 2010 from seven geographically and demographically diverse high schools in south-east Texas.13 Of participants completing Wave 5 in 2014 (n=698), those who did not answer key questions on firearm carrying (n=13) or substance abuse (n=1) were excluded from this analysis, resulting in an analytical sample of 684. Participants were ethnically diverse (33% Hispanic, 28% white, 26% African-American, 13% other), were 63% female and had a mean age of 20 (SD=0.8). The study received approval from the last author's institutional review board, and participants were protected by a privacy certificate issued by the National Institute of Justice.


To measure firearm carrying, participants were asked how often in the previous 12 months they had carried a gun outside their home for a reason other than hunting or target shooting. This measure was dichotomised into carriers (one or more times) and non-carriers. Follow-up questions were asked of those who reported past-year gun carriage, including the most important reason for carrying (ie, self-protection, holding it for someone, used it in a crime, to scare someone, to get back at someone, friends carry firearms or to feel important) and the type of gun carried (ie, regular rifle, automatic/semiautomatic rifle, regular shotgun, sawed-off shotgun, revolver or semiautomatic handgun).

Participants reported past-year substance use by responding yes/no to the use of (1) alcohol (more than just a few sips); (2) marijuana; (3) cocaine (powder, crack or freebase); (4) amphetamines (eg, speed, crystal, crank); (5) inhalants (eg, sniffed glue or huffing); (6) hallucinogens (eg, Lysergic acid diethylamide (LSD), Phencyclidine (PCP)); (7) over-the-counter cold or cough medicine with the intent to get high; (8) ecstasy and (9) prescription medications that were not prescribed by a health professional. Participants were asked how many days in the previous month they had participated in binge drinking (five drinks or more for men, four or more for women). This variable was grouped into three categories: none, 1–4 times (average of once a week or less) and 5+ times (average of more than once per week).

Additional covariates included age (years), sex (female/male), race/ethnicity (non-Hispanic white, non-Hispanic black or African-American, Hispanic or other) and highest level of parental education (less than high school, high school graduate, some college, college graduate, missing). The latter variable was determined by whichever parent had the highest completed level of education. If one parent's education was unknown or missing, the non-missing parent's education was used; if both were missing, parental education was classified as missing. Participants were also asked about their current work/school situation, with responses categorised as: in a public or private 4-year university; in trade or technical school or community college; working but not in postsecondary schooling; and neither working nor in postsecondary schooling.

Statistical analysis

Descriptive statistics were calculated for firearm carriers and non-carriers separately. Means and SDs were calculated for continuous variables and cross-tabulations, and percentages were calculated for categorical variables. Differences between carriers and non-carriers were calculated using t-tests for continuous variables and Fisher's exact test statistics for categorical variables. Multivariable logistic regression was used to model the odds of an individual carrying a firearm in the previous year. Separate models were run for each individual reported substance use category and episodic heavy drinking, controlling for age, gender, race/ethnicity, parental education and current work/school situation. All statistical analyses were conducted in 2016 using Stata V.14.1 (StataCorp, College Station, Texas, USA).


Overall, 6% (n=41) of participants reported that they had carried a firearm for a reason other than hunting or sport shooting in the previous year. Firearm carriers were more likely to be male (73% vs 34%, p<0.01) and slightly younger (19.7 vs 20.1, p<0.01) than non-carriers, but did not differ in racial/ethnic background or highest parental education. Firearm carriers were less likely to be enrolled in a public or private 4-year university, and more likely to be in community college, working and not in postsecondary education, and not working and not in postsecondary education (p=0.02). Most firearm carriers (68%) reported that their primary reasons for carrying a gun was to feel safe or for protection. Other reasons included to provide a feeling of importance or power (10%), holding it for a friend (7%), to scare someone, to commit a crime or because most of the persons’ friends carry firearms (all <5%). The most commonly carried type of firearm was a handgun (63%), with fewer carrying shotguns (20%) or rifles (17%).

Generally, firearm carriers were more likely than non-carriers to report past-year substance use; though, this difference did not always rise to statistical significance (table 1). In particular, firearm carriers were significantly more likely than their non-carrying counterparts to have used cocaine, amphetamines, hallucinogens, ecstasy and prescription medications not prescribed to them, as well as more likely to report episodic heavy drinking (p<0.05 for all).

Table 1

Firearm carriers and non-carriers, year 5 of Dating it Safe study, 2014

In logistic regression models, several substance use behaviours remained associated with firearm carrying even after adjusting for demographic and socioeconomic factors (unadjusted and adjusted ORs are shown in table 2). Indeed, relative to non-carriers, emerging adults who carried a firearm in the past year had higher odds of using hallucinogens (OR 2.81, 95% CI 1.00 to 7.81), ecstasy (OR 3.66, 95% CI 1.32 to 10.14) and prescription medications (OR 2.85, 95% CI 1.22 to 6.68). Use of cocaine (OR 2.65, 95% CI 0.95 to 7.42) and alcohol (OR 2.15, 95% CI 0.93 to 4.99) approached significance (p<0.10 for both). Use of methamphetamines was associated with firearm carrying in unadjusted models, but after adjusting for covariates, this association was no longer statistically significant (p=0.20). Finally, past-month episodic heavy drinking was associated with past-year firearm carrying, but only for those who practiced this behaviour frequently (5+ episodes/month; OR 3.61, 95% CI 1.51 to 8.66). Age, sex and current work/school status remained significant predictors of firearm carrying in all multivariable models (results not shown). Younger age and being male were consistently associated with greater odds of firearm carrying, while being enrolled in a 4-year public or private university was associated with lower odds of firearm carrying compared with all other work/school situations. For all fully adjusted models, McFadden's pseudo R-squared values ranged from 0.18 to 0.20 (p<0.001 for all). Fully adjusted models were then compared with models excluding substance use variables for hallucinogens, ecstasy, prescription medications, cocaine use, alcohol use and episodic heavy drinking; increases in pseudo R-squared values for models containing substance use variables ranged from 0.01 (cocaine) to 0.027 (episodic heavy drinking).

Table 2

Association between non-hunting/sport shooting firearm carrying and substance use


In this community-based sample of emerging adults, firearm carriers were significantly more likely to misuse a number of substances, including hallucinogens, ecstasy and prescription medications, and were more likely to have engaged in five or more episodes of heavy drinking in the past month. This is consistent with previous literature showing a link between firearm carrying and excessive alcohol use in national samples of adults10 and in studies of selected groups of adolescents.11 The literature on substance use and firearm carrying is limited, with some studies finding a positive relationship between drug use and firearm carrying in adolescents14–15 and others finding little or no relationship between drug use and general weapon carrying, including firearms.16 Few studies examined the relationship between gun carrying and specific substances; that our study included ethnically diverse emerging adults and filtered out carrying for sport also adds to the literature.

Socioeconomic status (SES) has been associated with weapon carrying in previous studies, though the relationship does not appear straightforward. For example, adolescents whose parents' education stopped with high school graduation have been found to carry weapons at higher rates than adolescents whose parents had either lower or higher education (though this association disappeared in multivariate models).17 In contrast, in a sample of at-risk adolescents and emerging adults, firearm carrying was more likely among those who were in higher SES (less likely to receive public assistance), though again this association disappeared in multivariate models.12 In our study, of the two SES measures, only the emerging adult's current situation was associated with firearm carrying, and this factor remained a significant predictor of firearm carrying in multivariate models, with lower SES associated with higher odds of firearm carrying. Because it is unclear what is driving this difference, further exploration of the role of SES differences between young adult firearm carriers and non-carriers is needed to clarify this relationship.

It is possible that the link between firearm carrying and substance use demonstrates a tendency for risky behaviours to cluster together in emerging adults, a tendency that has repeatedly been demonstrated in adolescents.15 ,18 For example, a recent study showed15 that recent gun possession by adolescents was strongly and consistently associated with alcohol, general tobacco use and selected other drug use over multiple rounds of the YRBS System. Why these risk behaviours tend to cluster together remains an important question. It is possible that any association between substance use and firearm carrying in this population is due to confounding by a so-called ‘third variable’19 that is related to both. For example, negative childhood experiences, such as repeated bullying, may create a perceived need for greater personal protection (and therefore increase firearm carrying) while also increasing the use of substances to cope with those experiences.20 Other candidates might include poor neighbourhood conditions or other social conditions15 or personality traits such as impulsivity. Examination of potential underlying variables will further elucidate the drivers behind this observed clustering of risky behaviours and lead to more effective intervention programming.

While 6% of young adult firearm carriers extrapolated across the USA represents an alarming number, our relatively small sample of firearm carriers (n=41) limits our ability to make conclusive statements. In addition, our firearm carrying questions did not specifically ask how many individuals carried for any reason, including hunting or sport shooting. Instead, the question asked individuals to exclude those reasons in an attempt to separate potentially riskier firearm carrying behaviours from more specific purposes related to common hobbies in Texas (hunting and sport shooting). This exclusion implicitly assumes that carrying for hunting and/or sport shooting is mutually exclusive with carrying for some other reason. In fact, individuals may have carried firearms for multiple reasons including hunting and sport shooting, and it is unclear in this sample how an individual who carried a firearm for both hunting and protection answered our firearm carrying question. Future research may want to more explicitly include hunting and sport shooting as reasons for carrying firearms. As these data are cross-sectional in nature, we cannot infer any causality between substance use and firearm carriage. In general, previous studies have either used larger population-based samples to examine substance use and (any) weapon carrying17 or smaller at-risk samples to study firearm carrying.4 ,12 The strength of our study—that is, using a community-based sample and asking about a specific type of firearm carrying—may also be a limitation, as this approach resulted in relatively low counts of firearm carrying (compared with weapon carrying in general, and compared with high-risk groups). That said, even with these relatively small counts, we did find statistically significant associations between firearm carrying and substance use behaviours. Other limitations include our reliance on self-reports for these relatively sensitive questions, and that past-year firearm carrying and substance use may not have coincided exactly. In other words, any particular individual may not have used substances and carried a firearm at the same time, and vice versa. Finally, this study sample was entirely from a single region within the USA (south-east Texas). In the USA, and Texas in particular, there are laws that allow firearm carrying, both open and concealed, among adults in many public areas. These laws, along with the social acceptability of firearm carrying, may differ significantly from other areas of the USA and other countries. Our results may therefore be limited in their generalisability (though this fact should encourage further research in other regions with differing laws and customs regarding firearms).


Results from this study further add to the literature on the association between firearm carrying and substance use. Additional examination of the drivers behind these associations is warranted, particularly in terms of causality and mediation. Research using large national samples is especially needed. Regardless, our findings are potentially alarming; we found that emerging adults who carried firearms in the last year are more likely to have also misused alcohol and several drugs. Given the potentially devastating impacts of this combination, future public health research on these issues should remain a priority.

What is already known on the subject

  • Rates of general weapon carrying (not limited to firearms) are known among general adolescent populations, and small studies of firearm carrying among specific groups of at-risk adolescents or adults have been published.

  • Substance use, particularly alcohol use, is often linked with risky firearm-related behaviours in adults.

What this study adds

  • Rates of firearm carrying for reasons other than sport or hunting in a general population of young adults in the USA.

  • There are associations between dangerous alcohol-related behaviours and use of other illegal substances, and carrying a firearm for reasons other than sport or hunting (primarily protection).

  • The association between risky firearm carrying and substance use behaviours in a general population of emerging adults is worrisome.

No funding for research on gun violence

The 1996 Dickey Amendment to an appropriations bill prevented funds being used to advocate or promote gun control. Now modified, that provision still deters researchers in this field because of National Rifle Association support. After the San Bernardino shooting Democrats stated, “We dedicate $240 million a year on traffic safety research, more than $233 million a year on food safety and $331 million a year on the effects of tobacco, but almost nothing on firearms that kill 33,000 Americans annually.”

Surviving a hurricane

Many deaths followed the recent hurricanes in the Caribbean and American Southwest. How might such trauma be prevented given the immense power of these events? Obviously, the most effective is to evacuate the danger area when advised to do so. If it is necessary to stay, the advice is to ‘Remain inside a secure shelter and stay away from windows’, and do so with a previously prepared hurricane evacuation kit. Drivers are advised not to use cruise control or drive on flooded streets.



  • Contributors JRT conceptualised the study, supervised the study, and reviewed, revised and edited the manuscript. JDP assisted in study conceptualisation, data interpretation, and drafting and revision of the manuscript. JGB assisted in data interpretation, and revision and editing of the manuscript. RNB conducted the analysis, interpreted the data and drafted the manuscript.

  • Funding This research was supported by Award Number 2012-WG-BX-0005 (PI: Temple) from the National Institute of Justice (NIJ). The content is solely the responsibility of the authors and does not necessarily represent the official views of NIJ.

  • Competing interests None declared.

  • Ethics approval University of Texas Medical Branch Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.