Suicidality Among Veterans: Implications of Sexual Minority Status

Abstract

Using the California Quality of Life surveys, we examined suicidal ideation and attempts in 129 lesbian, gay, and bisexual (LGB) veterans and in 315 heterosexual veterans in 2008–2009 and 2012–2013. Although there were no significant differences in the past 12-month suicidal ideation and lifetime attempts, LGB veterans had higher odds of lifetime suicidal ideation than heterosexual veterans (adjusted odds ratio = 3.00; 95% confidence interval = 1.38, 6.53). Suicide assessment and prevention efforts in LGB veterans could benefit from a life-course perspective regarding suicide risk.

Suicide prevention is a public health imperative for military veterans,¹ who are estimated to constitute more than 20% of US suicide deaths annually.² Lesbian, gay, and bisexual (LGB or sexual minority) populations also have elevated risks of suicidal ideation and attempt compared with heterosexuals.³ It is unclear, however, if LGB status is significantly associated with suicidal ideation or attempt among veterans, a population particularly vulnerable to suicide risk.⁴ Although 1 previous study found that LGB veterans had a higher prevalence of past-year suicidal ideation than heterosexual veterans did,⁵ little else is known about possible differences in lifetime suicidal ideation or suicide attempts in this subpopulation.

By using a population-based sample, we sought to examine the association between sexual orientation and past-year suicidal ideation, lifetime suicidal ideation, and lifetime suicide attempt among veterans.

METHODS

Data were from the 2008–2009 and 2012–2013 California Quality of Life (Cal-QOL) surveys, which are computer-assisted telephone interviews of probability-based samples of adults aged 18 to 70 years from parent waves of the California Health Interview Surveys (see Cochran et al. for study methodology and sample characteristics of the 2008–2009 wave⁶). Our study focused on respondents who indicated “ever serving on active duty in the Armed Forces of the United States” (n = 444).

Demographic characteristics included gender, age, race/ethnicity, education, and employment status. Because of the restricted sample size, we dichotomized sexual orientation as LGB or heterosexual. Marital status was categorized as partnered (married or cohabiting) versus single (widowed, divorced, separated, never married). Because California has a higher cost of living than the rest of the United States,⁷ we followed previous Cal-QOL analyses⁸ and dichotomized income using 300% of the year-adjusted federal poverty level as defined by the US Department of Health and Human Services. All veterans were asked if they had ever served in a war zone or area of active conflict, which was operationalized as service in a combat zone.⁹

Suicidality was assessed using the World Health Organization Composite International Diagnostic Interview for Suicidality.¹⁰ Analyses focused on dichotomous items of the past 12-month suicidal ideation, lifetime suicidal ideation, and lifetime suicide attempt.

We used a logistic regression model to parsimoniously evaluate demographic characteristic predictors of LGB status.¹¹ We used 3 separate logistic regression models for each suicide-related symptom to investigate the association of LGB status. We adjusted all models for survey cycle, service in a combat zone, and sociodemographic factors related to suicidality (e.g., gender, age, education, income, employment, and relationship status).¹² We weighted all analyses using SAS version 9.2 (SAS Institute, Cary, NC) and SUDAAN version 11.0.1 (RTI International, Research Triangle Park, NC) to account for oversampling of LGB individuals.¹¹

RESULTS

Among veterans, the weighted prevalence of LGB status was 2.46% (95% confidence interval [CI] =  1.77%, 3.42%). Fewer LGB veterans were partnered, and fewer indicated combat exposure than heterosexual veterans (Table 1). In bivariate analyses, LGB veterans had lower prevalence of past 12-month suicidal ideation than heterosexual veterans (4.2% vs 7.4%), although this was not statistically significant. Conversely, 47.0% of LGB veterans indicated lifetime suicidal ideation, which was significantly higher than that among heterosexual veterans (22.1%). Lifetime suicide attempt was elevated among LGB veterans compared with heterosexual veterans, but did not achieve statistical significance (22.6% vs 7.6%; P = .082). In adjusted models, LGB veterans experienced 3 times the odds of lifetime suicidal ideation than heterosexual veterans (95% CI = 1.38, 6.53; Table 2).

TABLE 1—

Demographic Characteristics and Suicidality Among Veterans, by Sexual Minority Status: California Quality of Life Surveys, 2008–2009 and 2012–2013

	Sexual Minority Veterans (n = 129)		Heterosexual Veterans (n = 315)
Variable	No. (%) or Mean	SE	No. (%) or Mean	SE	P
Demographicsa
Age, y	53.1	2.7	52.0	1.0	.881
Gender
Male	108 (78.4)	7.6	282 (89.6)	2.2	.307
Female (Ref)	21 (21.5)		33 (10.4)
Race/ethnicity
Non-Hispanic White	104 (71.4)	7.9	188 (65.9)	3.3	.43
Racial/ethnic minority (Ref)	25 (28.6)		127 (34.1)
Relationship status
Partnered	39 (52.2)*	7.7	235 (76.8)	2.9	.003
Single (Ref)	90 (47.8)		80 (23.2)
Education
< college	51 (44.8)	7.9	191 (59.3)	3.4	.162
≥ college degree (Ref)	78 (55.2)		124 (40.7)
Employment status
Unemployed	6 (1.8)	0.8	15 (5.6)	1.8	.07
Out of workforce	63 (37.9)	7.2	106 (28.7)	2.9	.326
Employed (Ref)	60 (60.3)	7.3	194 (65.7)	3.2
Below 300% FPL	37 (26.5)	3.3	93 (30.8)	6.6	.338
Service in combat zone	25 (18.6)*	7.2	120 (37.6)	3.3	.041
Suicidalityb
Lifetime suicidal ideation	52 (47.0)*	7.9	72 (22.1)	2.8	.007
Lifetime suicide attempt	21 (22.6)	7.7	26 (7.6)	1.7	.082
Past 12 mo suicidal ideation	8 (4.2)	1.8	22 (7.4)	1.8	.224

Note. FPL = federal poverty level (according to The US Department of Health and Human Services). All analyses are weighted; frequencies are unweighted and percentages and SEs are weighted.

^aSingle logistic regression adjusted for survey year and testing association of demographic characteristics with sexual minority status.

^bχ-² test (sexual minority vs heterosexual).

*P < .05.

TABLE 2—

Association of Sexual Minority Status and Other Demographic Characteristics With Suicidality Among Veterans: California Quality of Life Surveys, 2008–2009 and 2012–2013

Variable	Lifetime Suicidal Ideation AOR (95% CI)	Lifetime Suicide Attempt AOR (95% CI)	Past 12 Mo Suicidal Ideation AOR (95% CI)
Age	0.98 (0.95, 1.01)	0.97 (0.92, 1.02)	0.92* (0.87, 0.97)
Female	1.68 (0.68, 4.16)	1.05 (0.21, 5.39)	2.53 (0.42, 15.26)
Non-Hispanic White	1.31 (0.65, 2.65)	2.50 (0.69, 9.03)	3.02 (0.81, 11.24)
Partnered	0.64 (0.32, 1.27)	0.37 (0.13, 1.04)	0.33* (0.12, 0.90)
≤ college degree	1.13 (0.58, 2.22)	1.86 (0.66, 5.25)	1.10 (0.30, 4.05)
Unemployed	1.16 (0.22, 6.23)	1.30 (0.12, 13.87)	2.09 (0.26, 16.99)
Out of workforce	1.61 (0.66, 3.92)	1.28 (0.25, 6.56)	7.18 (1.34, 38.42)
< 300% FPL	2.51* (1.24, 5.09)	3.65* (1.27, 10.48)	6.48* (1.97, 21.28)
Service in combat zone	0.94 (0.48, 1.84)	1.28 (0.46, 3.55)	2.16 (0.70, 6.63)
Sexual minority	3.00* (1.38, 6.53)	3.21 (0.89, 11.59)	0.46 (0.11, 1.90)

Note. AOR = adjusted odds ratio; CI = confidence interval; FPL = federal poverty level (according to The US Department of Health and Human Services). All analyses are weighted and adjusted for survey year. Sample size for all models is 444 respondents.

*P < .05.

DISCUSSION

Unlike a previous study,⁵ we did not observe significant crude differences in past 12-month suicidal ideation by sexual orientation among veterans. However, LGB veterans in our study had higher odds of lifetime suicidal ideation and evidenced a trend for an elevated prevalence of lifetime attempts compared with heterosexuals. In this veteran sample, it was unclear when sexual orientation differences in lifetime suicidality occurred. In general, the literature suggests that adolescence and young adulthood might be specific periods of heightened risk for sexual minority individuals.¹³ These developmental periods might overlap with or be proximal to the typical age of military enlistment. Further research is needed to ascertain when suicidal symptoms occur among LGB veterans with respect to their military service.

In addition to lifetime suicidality, LGB individuals were more likely than heterosexuals to experience anxiety and mood disorders,¹⁴ but it was unclear how characteristics of veteran status might influence mental health and suicidality among LGB individuals. For example, sexual minority veterans had more than twice the odds of keeping firearms in the home than sexual minority nonveterans,¹⁵ and firearm ownership, itself, is linked to risk of suicide.¹⁶ Future research is needed on whether characteristics of veteran status (e.g., firearm ownership,¹⁵ traumatic brain injury,¹⁷ military sexual trauma¹⁸) might play roles in the suicidality of LGB individuals. Furthermore, suicide prevention for veterans should identify the needs of LGB veterans with histories of suicidal symptoms.

We noted several study limitations. First, as a state-based sample, the results might not generalize to the United States. Second, we aggregated sexual minority groups to increase statistical power, obscuring possible heterogeneity among sexual minority identities.¹⁹ Despite aggregation, the sample size was relatively small, which affected the statistical power of our study. Specifically, although we observed elevation in lifetime suicide attempts among LGB veterans, the small sample size might have reduced our ability to detect statistical significance. The Cal-QOL also did not include a measure of lifetime depression, which would have been an important covariate for the lifetime suicidal ideation and attempt outcomes. Finally, self-reported veteran status could introduce misclassification bias.