Suicide Attempt Rates and Associations with Discrimination Are Greatest in Early Adulthood for Sexual Minority Adults Across Diverse Racial and Ethnic Groups

Abstract

Purpose: The study purpose was to investigate differences in past 5-year suicide attempts among racially/ethnically diverse sexual minority (SM) and heterosexual adults across the life course and examine the association of discrimination with past 5-year suicide attempts among SM adults.

Methods: Using nationally representative data collected from 2012 to 2013, we assessed differences in age-varying prevalence of past 5-year suicide attempts among Black, Hispanic, and White SM and heterosexual adults (ages 18–60). We tested whether the association between discrimination and suicide attempts differed by race/ethnicity among SM adults. All secondary data analyses were approved by the Institutional Review Boards of The Pennsylvania State University and University of Maryland.

Results: Compared with heterosexual adults of any race/ethnicity, SM adults exhibited elevated suicide attempt rates until the late twenties when prevalence for Black and Hispanic SM adults declined. Disparities persisted into the mid-40s for White SM adults. Among SM adults of all races/ethnicities, the relationship between SM discrimination and suicide attempts was strongest between ages 18 and 25. For SM adults reporting SM discrimination, odds of suicide attempts were 3.6 times higher for White SM adults and 4.5 times higher for Black and Hispanic SM adults, relative to same-race/ethnicity SM adults who did not report SM discrimination. The effect of SM discrimination was robust among Black and Hispanic SM young adults even when accounting for racial/ethnic discrimination.

Conclusions: SM adults of all racial/ethnic groups demonstrated disparities when contrasted with heterosexual adults of any race/ethnicity, although ages characterized by heightened prevalence rates of suicide attempts differed by race/ethnicity. Early adulthood is a critical period for intervention seeking to disrupt the association between SM discrimination and suicide attempts.

Introduction

Sexual minority (SM) adults evidence persistent disparities in suicide attempts relative to their heterosexual peers.^1–4 Indeed, sexual orientation has been identified as an important factor in suicide attempts across the life course⁵; however, research focuses predominantly on adolescence and, to a lesser extent, young adults. Despite a robust body of literature documenting disparities in suicidality and associations between stigma and suicidality,⁶ we know relatively little about suicide attempt rates among SM people of color (POC) and whether suicide attempt rates might systematically vary for SM adults by race/ethnicity.

Developmental research consistently highlights adolescence and young adulthood as the periods with the greatest suicide attempt prevalence.^7,8 The transition to adulthood represents a sensitive period for the onset and progression of anxiety and mood disorders, which may be exacerbated by the stress of increased responsibility, individuation from family, and increasingly intimate social relationships.⁹ This period is uniquely complex for SM young people who often experience anti-SM stigma that ranges from messages to conform their sexuality to outright discrimination and victimization predicated on diverse sexuality.¹⁰ However, it is unclear how SM individuals across life stages respond differently to stigma and discrimination.

Emergent research in adult samples has similarly demonstrated that the link between anti-SM discrimination and recent suicide attempts is stronger at younger ages (18–30 years).¹¹ Yet, these studies have not been attuned to how differences in societal and cultural contexts conferred on the basis of race/ethnicity may contribute to variability in suicide attempts among SM adults over the life course.

Racial/ethnic differences in suicide attempts and the intersection with sexual orientation

In general, higher prevalence of suicide is typical of White populations compared with Black and Hispanic populations.^12,13 However, researchers speculate that SM POC may be at greater risk for poor mental health due to stigma related to multiple minoritized statuses—a phenomenon sometimes called “double jeopardy”,^14,15 yet evidence is mixed. Indeed, as demonstrated in a recent review of adolescent and emerging adult suicide literature, studies inconsistently reported White SM adolescents and emerging adults at greatest risk, SM adolescents and emerging adults of color at greatest risk, or no racial/ethnic differences in suicide outcomes.¹⁶ An intersectional approach to suicide attempt disparities at the intersection of race/ethnicity and sexual orientation would consider not only multiple, co-occurring identity intersections but also how experiences with systems of power (e.g., racism and homophobia), as reflected in exposure to enacted interpersonal discrimination and stigma, are connected to health outcomes.^17–19

Although considerable attention has been given to suicide attempts in SM adolescents, much less research examines SM adults, especially POC. Some studies of SM adults demonstrated that White SM adults were at greatest risk,²⁰ others reported that SM POC exhibited higher risk,²¹ and others found no racial/ethnic differences in risk for suicidality.²² With the exception of these studies—which offer more questions than answers regarding patterns of suicide attempts at the nexus of sexual orientation and race/ethnicity perhaps due, in part, to differences in community samples and national college samples—there remains a paucity of research on racial/ethnic differences in suicide risk within the SM population in middle and older adulthood. In addition, both minority stress theory^23,24 and critical race theory^14,19,25,26 call for a better understanding of social processes that drive disparities between groups. Indeed, it is racist and heterosexist discrimination that underlie disparities rather than characteristics of minoritized groups.^6,27–30

Mixed findings regarding suicide attempts among racially/ethnically diverse SM adults may be partially attributable to age group differences.^11,31–33 That is, findings may reflect confounding age differences both within and across samples that make it difficult to understand how suicide attempts differ by sexual orientation and race/ethnicity. Indeed, both individual development and historical context across the life course dynamically change as individuals age.³¹ In the general population, the prevalence of suicide attempts among adults is greatest in early adulthood³²; however, some evidence suggests the peak may be distributed and elevated further into adulthood for SM adults.^11,33 Yet, there remain few studies that consider how the relationship between sexual orientation and risk for suicide attempts varies by race/ethnicity and across the life course. Likewise, little is known about how the association between discrimination and suicide attempts may differ in various life stages. Such findings are crucial to developing health promotion strategies that reach those most at risk. In the current study, we extend the work of Fish et al.¹¹ to better understand how the risk for suicide attempts varies on the basis of age, sexual orientation, and race/ethnicity using nationally representative data.

Methods

Data source and sample

Data are from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) III. The NESARC protocol was approved by the US Census Bureau and the US Office of Management and Budget. Participants provided verbal consent to participate. All secondary data analyses were approved for ethical study by the Institutional Review Boards of The Pennsylvania State University and University of Maryland.

Collected in 2012–2013, the NESARC (N = 36,309)³⁴ is a nationally representative cross-sectional survey of noninstitutionalized civilian adults aged 18 years and older. Our final, weighted analytical sample (n = 25,550) included non-Hispanic White, non-Hispanic Black, and Hispanic participants. Data were collected by trained interviewers through computer-assisted personal interviews using the Alcohol Use Disorder and Associated Disabilities Interview Schedule-5 (AUDADIS-5).

Measures

Suicide attempt

To calculate any past 5-year suicide attempt, we subtracted participants' reported age at the most recent suicide attempt from age reported at the time of the survey. If the difference was 5 years or less, participants were coded as having a past 5-year suicide attempt = 1. If the difference was >5 years, or if they reported never attempting suicide, they were coded as no past 5-year suicide attempts = 0. The mean difference between current age and age at last suicide attempt for the analytical sample was 13.8 years (standard deviation [SD] = 11.0).

Sexual orientation

Sexual orientation was operationalized using sexual attraction, sexual behavior, and sexual identity. We (inclusively) considered attraction, behavior, and identity because some SM POC may be less likely to disclose their SM identity³⁵ than SM White individuals and may experience conflict between ethnic and sexual identities, which impacts the disclosure process.³⁶ In addition, limited sexual identity response options in secondary data analysis may preclude some SM individuals from self-identifying when better-suited options (e.g., pansexual, queer, downe, same-gender loving, and two-spirit) or open responses are not available.

For sexual attraction, participants' response options were “only attracted to females,” “mostly attracted to females,” “equally attracted to females and males,” “mostly attracted to males,” or “only attracted to males.” Using participant gender/sex (male or female), we recoded attraction to reflect exclusively other-sex attraction and any same-sex attraction. For sexual behavior, participants indicated lifetime sexual partners as “only males,” “only females,” “both males and females,” and “never had sex.” Responses were recoded to capture exclusive other-sex sexual partners or never had sex and any same-sex sexual partners. Participants selected one of the following sexual identities: heterosexual (straight), gay or lesbian, bisexual, or not sure. Participants who reported any same-sex attraction, any same-sex behavior, or a lesbian/gay or bisexual identity were coded as a SM = 1. Participants reporting exclusive other-sex attraction, behavior, and heterosexual identities were coded as heterosexual = 0. Participants who reported their sexual identity as “not sure” were coded as sexual minorities (n = 199), the majority of these participants also reported behavior or attraction that indicated SM status (n = 178).

Race and ethnicity

Participants first indicated whether they were Hispanic/Latino. Then, participants selected one or more categories to describe their race. Response options included White, Black/African American, Asian/Native Hawaiian/Other Pacific Islander, and American Indian/Alaska Native. These two items were used by NESARC staff to create race/ethnicity categories of non-Hispanic White, non-Hispanic Black, non-Hispanic American Indian/Alaska Native, non-Hispanic Asian/Native Hawaiian/Other Pacific Islander, and Hispanic, any race. Due to sample size limitations at the cross-section of race/ethnicity and sexual orientation, we limited our sample to non-Hispanic White, non-Hispanic Black, and Hispanic participants. A descriptive summary of Asian/Pacific Islander and Native American subgroups is included in Supplementary Table S1.

Discrimination

Lifetime experiences of SM and racial/ethnic discrimination were assessed using the six-item Experiences of Discrimination Scale^37,38 among the SM subsample only. Example items included “How often were you made fun of, picked on, shoved, hit, or threatened with harm because you were assumed to be gay, lesbian, or bisexual?” and “How often did you experience discrimination in public, such as on the street, in stores, or in restaurants, because you were assumed to be gay?” with response options ranging from never (0) to very often (5). Identical questions were asked regarding discrimination targeting race/ethnicity/Hispanic heritage. Reponses were coded as any lifetime sexual minority discrimination = 1 or no sexual minority discrimination = 0 and any lifetime racial/ethnic discrimination = 1 or no racial/ethnic discrimination = 0.

Analytic approach

We first calculated weighted frequencies for all variables to describe the full analytical study sample and used weighted logistic regression to calculate the racial/ethnic differences in suicide attempt across sexual orientation, irrespective of age. We then used time-varying effect modeling (TVEM)³⁹ to estimate age-varying prevalence of past 5-year suicide attempts. TVEM allows researchers to estimate developmental differences in health using cross-sectional data in sufficiently powered samples.^11,40 Given that risk for suicide attempt varies between groups and, simultaneously, across the life course, TVEM provides a compelling platform to understand how rates of suicide attempts vary on the basis of race/ethnicity, sexual orientation, and age. We did this by first estimating intercept-only models, which provided the age-varying prevalence of past 5-year suicide attempts across sexual orientation and race/ethnicity. This resulted in six curves that represent the age-varying prevalence of suicide attempts among White heterosexual, Black heterosexual, Hispanic heterosexual, White SM, Black SM, and Hispanic SM adults.

We then aimed to evaluate the contribution of SM discrimination among the SM subsample. We hypothesized that discrimination was a significant predictor of suicide attempts across all racial/ethnic groups. We attempted to evaluate these associations using TVEM. However, due to insufficient sample sizes among Black and Hispanic SM adults, TVEM analyses were underpowered. Thus, we used weighted logistic regression to evaluate the association of discrimination with suicide attempts for White SM adults and Black and Hispanic SM adults. Insufficient power necessitated collapsing Black and Hispanic SM adults into a single subgroup for this portion of analyses. We calculated odds ratios (ORs) for 18- to 25-year-olds, 26- to 35-year-olds, and 36- to 60-year-olds. These age categories were derived from results in Figures 1–3, which indicated that suicide attempt disparities were greatest among 18–25-year-olds, plateaued in the late 20s and early 30s, and minimized by the early 40s. Finally, in the subsample of Black and Hispanic SM adults, we examined the association of both SM discrimination and racial/ethnic discrimination with odds of past 5-year suicide attempts using multivariable, weighted logistic regression stratified by age.

FIG. 1.

Age-varying prevalence of suicide attempts in the past 5 years by sexual orientation among White adults.

FIG. 2.

Age-varying prevalence of suicide attempts in the past 5 years by sexual orientation among Black adults.

FIG. 3.

Age-varying prevalence of suicide attempts in the past 5 years by sexual orientation among Hispanic adults.

Results

Our full weighted analytical sample (n = 25,550) included non-Hispanic Black (n = 3527, 13.8%), non-Hispanic White (n = 17,312, 67.8%), and Hispanic (n = 4711, 18.4%) participants aged 18–60 years (Table 1). The sample was nearly evenly split by gender/sex (50.8% female), with an average age of 39.1 (SD = 12.3). The SM subsample comprised 10.1% (n = 2591) of the full analytical sample.

Table 1.

Sample Demographic Characteristics, Suicide Attempts, and Discrimination

	Total sample, %	Sexual minority (n = 2591, 10.1%)a			Heterosexual (n = 22,959, 89.9%)a
		White (n = 1773), %	Black (n = 378), %	Hispanic (n = 439), %	White (n = 15,539), %	Black (n = 3149), %	Hispanic (n = 4271), %
Age, years
18–25	19.3	22.3	31.0	34.6	17.0	21.5	22.1
26–35	22.3	24.6	24.7	24.9	20.4	22.1	28.2
36–45	22.3	20.3	16.4	20.9	22.0	23.2	24.3
46–55	24.9	22.7	21.0	14.4	27.5	23.8	18.8
56–60	11.2	10.2	7.0	5.3	13.2	9.4	6.6
Sex
Male	49.2	39.2	35.8	39.5	50.7	53.0	51.6
Female	50.8	60.8	64.2	60.5	49.3	47.0	48.4
Lifetime suicide attemptsb
Yes	6.1	15.7	11.2	11.1	5.6	4.0	4.4
No	93.6	84.3	88.7	88.9	94.1	95.5	95.3
Missing	0.4	<0.1	<0.1	<0.1	0.4	0.6	0.3
Past 5-year suicide attemptsb
Yes	1.7	5.9	4.7	4.8	1.4	1.0	1.4
No	98.3	94.0	95.0	95.0	98.1	98.2	98.2
Missing	0.6	0.2	0.4	0.2	0.5	0.8	0.4
Sexual orientation
Identity	38.4	37.5	41.3	39.5	—	—	—
Attraction	85.8	85.2	86.1	87.9	—	—	—
Behavior	67.5	69.1	64.3	63.8	—	—	—
Lifetime sexual minority discriminationc
Yes	22.8	23.3	22.3	22.2	—	—	—
No	77.2	76.7	77.7	77.8	—	—	—
Lifetime racial/ethnic discriminationd
Yes	60.2	—	67.0	51.1	—	—	—
No	39.9	—	33.0	48.9	—	—	—

^aBecause sample sizes represent a weighted sample to achieve national representation of race/ethnicity, the sum of race/ethnicity subsamples within each sexual orientation group may not be exactly equal to the total.

^bChi-square tests revealed significant differences in prevalence of both lifetime suicide attempts [χ² (5) = 390.0] and past 5-year suicide attempts [χ² (5) = 211.7] across the intersections of sexual orientation and race/ethnicity.

^cTotals and analyses for sexual minority status and discrimination did not include heterosexual participants.

^dTotals and analyses for racial/ethnic discrimination did not include White sexual minority and heterosexual participants.

Age-varying prevalence of suicide attempts

The age-varying prevalence of past 5-year suicide attempts comparing heterosexual and SM adults varied by race/ethnicity (Figs. 1–3). Although a greater proportion of SM adults than heterosexual adults reported past 5-year suicide attempts throughout much of adulthood, the patterns varied by race/ethnicity. Across the life course, White SM adults reported prevalence rates similar to Black SM adults and Hispanic SM adults until the mid- to late twenties. At those ages, the prevalence of suicide attempts among Black and Hispanic SM adults began declining and it was not significantly different from heterosexual adults of all races/ethnicities by approximately age 33. In contrast, the prevalence among White SM adults remained significantly elevated compared with White heterosexual adults and all Black and Hispanic participants from the early thirties until approximately age 46. Black and Hispanic SM adults had very similar age-varying patterns of suicide attempts. Meanwhile, racial/ethnic differences among heterosexual adults were not pronounced (Figs. 1–3).

Age-stratified associations of SM discrimination with past 5-year suicide attempts among SM adults, by race/ethnicity

An evaluation of the association of lifetime SM discrimination with past 5-year suicide attempts among White SM adults and Black and Hispanic SM adults revealed age-stratified associations. Across racial/ethnic groups, the relationship between discrimination and suicide attempts was strongest among those aged 18–25 years. White SM adults, aged 18–25 years, who reported discrimination had 3.60 (95% confidence interval [CI]: 1.56–8.29) times the odds of suicide attempts compared with White SM adults, aged 18–25 years, who did not experience discrimination. For Black and Hispanic SM adults aged 18–25 years, the odds of suicide attempts among those who reported discrimination were 4.52 (95% CI: 1.58–12.91) times the odds of those who did not experience discrimination (Table 2). After age 25, the relationship between discrimination and suicide attempts was not significant for any racial/ethnic group. Race/ethnicity did not moderate the association between SM discrimination and odds of past 5-year suicide attempts (results not shown).

Table 2.

Age-Stratified Associations Between Lifetime Sexual Minority Discrimination and Suicide Attempts

	White sexual minority adults, OR (95% CI)	Black and Hispanic sexual minority adults, OR (95% CI)
Age, years
18–25	3.6 (1.56–8.29)	4.52 (1.58–12.91)
26–35	1.07 (0.40–2.82)	2.64 (0.56–12.40)
36–60	0.72 (0.26–1.98)	0.20 (0.02–1.59)

α = 0.05. Bold font indicates statistical significance at p < 0.05.

CI, confidence interval; OR, odds ratio.

Age-stratified associations of SM and racial/ethnic discrimination with past 5-year suicide attempts among Black and Hispanic SM adults

For Black and Hispanic SM adults aged 18–25 years, the odds of suicide attempts remained higher among those who reported SM discrimination (OR = 4.28; 95% CI: 1.46–12.55) when also accounting for racial/ethnic discrimination (Table 3). In contrast, racial/ethnic discrimination was not associated with odds of suicide; however, inclusion of racial/ethnic discrimination in the model attenuated the association of SM discrimination with suicide attempts. After age 25, the relationship between discrimination of any type and suicide attempts was not significant.

Table 3.

Age-Stratified Associations Between Lifetime Discrimination and Suicide Attempts for Black and Hispanic Sexual Minority Adults

	Sexual minority discrimination, OR (95% CI)	Racial/ethnic discrimination, OR (95% CI)
Age, years
18–25	4.28 (1.46–12.55)	1.81 (0.60–5.49)
26–35	2.33 (0.52–10.38)	1.72 (0.38–7.85)
36–60	0.20 (0.03–1.61)	0.81 (0.20–3.27)

α = 0.05. Bold font indicates statistical significance at p < 0.05.

Discussion

Minority stress theory suggests that chronic stressors (e.g., discrimination) unique to SM adults impact mental health outcomes across the life course.^6,11,24,41 We extend previous research supporting the application of minority stress theory to suicidality^3,6,11 by modeling age-varying risk for suicide at the intersection of sexual orientation and race/ethnicity across adulthood in a nationally representative study. Results of this study showed that SM adults, aged 18–25 years, who experienced lifetime SM discrimination had greater odds of past 5-year suicide attempts compared with same-age and -race/ethnicity SM adults who had not experienced SM discrimination. Compared with White, Black, and Hispanic heterosexual adults, SM adults in all three racial/ethnic groups exhibited elevated prevalence of suicide attempts through the early 30s. White SM adults experienced a longer period of elevated risk extending into the mid-40s.

The results are consistent with previous research showing that SM adults exhibit higher prevalence of suicidality than heterosexual adults throughout the life course.^11,42 A strength of the current study is the use of nationally representative data, including adult participants from the general population across the life course, which may help to clarify inconsistencies in suicide attempt rates among racially/ethnically diverse SM adults across community and national college samples.^16,20–22 Age-varying prevalence of past 5-year suicide attempts highlighted critical developmental periods of elevated risk. Risk for past 5-year suicide attempts among SM adults was most elevated in early adulthood across all racial/ethnic groups. These results indicate an especially vulnerable period when SM adults may be experiencing rapid developmental transitions and continued psychological exploration of identity.^43,44

The results of this study add three important findings. First, the period of elevated past 5-year suicide attempts for SM adults relative to heterosexual adults extended well beyond early adulthood into the 30s (Black and Hispanic SM adults) and 40s (White SM adults). Second, elevated prevalence of past 5-year suicide attempts was apparent not only for White SM adults but also for Black and Hispanic SM adults relative to heterosexual adults of all races/ethnicities. Third, SM discrimination appears to be related to past 5-year suicide attempts for White, Black, and Hispanic SM young adults, and for Black and Hispanic SM young adults, this association remains stable even when accounting for racial/ethnic discrimination.

Explaining the elevated suicide attempt risk for White SM adults compared with Black and Hispanic SM adults from the late twenties to mid-40s remains a challenge in public health research. Competing hypotheses seek to explain inconsistent results in health disparities observed between SM POC and both heterosexual POC and White SM adults. Because SM POC are expected to experience psychosocial stressors due to their race/ethnicity, sexuality, and intersections of these two identities, the double jeopardy hypothesis suggests that SM POC experience extra stigma-related stress and, thus, are at greater risk for mental health adversity than White SM adults and heterosexual POC.^14,15 However, some research,⁴⁵ including portions of this study, provides emerging evidence for the minority resilience hypothesis,⁴⁶ which suggests that despite expectations of elevated stress, SM POC may demonstrate unique resilience and no greater risk than White SM adults. Indeed, it appears that even when considering both SM discrimination and racial/ethnic discrimination among Black and Hispanic SM adults, it was SM discrimination alone that drove the association with suicide attempt.

By modeling suicide prevalence with age-varying effects, we uncovered a potential explanation for inconsistent results. Like Hayes et al.,⁴⁷ we found that the double jeopardy hypothesis holds true for SM POC when compared with heterosexual POC, but not necessarily when compared with White SM adults. This divergent evidence for the double jeopardy hypothesis was demonstrated by the elevated risk for suicide among Black and Hispanic SM adults in early adulthood relative to Black and Hispanic heterosexual adults, which contrasted with the elevated prevalence of past 5-year suicide attempt among White SM adults in their 30s and 40s relative to Black and Hispanic SM adults. It is therefore possible that Black and Hispanic SM adults draw on different resources for resilience^45,46 in middle and later adulthood, which are either less readily available or take longer to develop for White SM adults.

Prior studies have demonstrated that ethnic identity can act as a protective resource against behavioral health disparities (e.g., substance use) for racial/ethnic minority adolescents, but not for White adolescents.⁴⁸ Smith and Silva⁴⁹ indicated the need for additional research to examine in what contexts and for what subgroups racial/ethnic identity may be a protective factor for mental health. It is possible that racial/ethnic identity and community resources may be protective against the effects of discrimination and may be accessed differentially across the life course.

Across the life course, the association of discrimination with past 5-year suicide attempts varied such that reporting SM discrimination was associated with suicide risk in young adulthood, but not middle and later adulthood. In addition, it is noteworthy that significant effects of discrimination were of greater magnitude for Black and Hispanic SM adults than for White SM adults. This finding suggests the importance of taking a within-group approach to study SM suicide risk. Although White SM adults may be at risk of suicide attempts across a longer developmental period than Black and Hispanic SM adults, SM discrimination may be especially harmful among Black and Hispanic SM young adults.

The results of this study support SM discrimination as a factor associated with suicide attempts among Black and Hispanic SM young adults; however, the results should not be interpreted to mean that racial/ethnic discrimination has no bearing on mental health or suicidality for SM POC. Instead, the results of this study can provide evidence that the association of SM discrimination with suicide attempt is especially pernicious in early adulthood and it reaches universally across racially/ethnically diverse groups of SM adults. Acknowledging and accounting for both SM discrimination and racial/ethnic discrimination are important parts of applying an intersectional framework to health among SM POC; however, intersectionality calls for research that examines not only separate sources of stigma and oppression but also the unique intersection of multiple forms of oppression.^17–19,50 Future investigation of intersecting and co-occurring forms of heterosexism and racism (e.g., sexual racism in SM communities)⁵¹ may provide a more nuanced understanding of how the intersection of multiple forms of oppression contributes to elevated prevalence of suicide attempts among Black and Hispanic SM young adults compared with Black and Hispanic heterosexual young adults.

Limitations

There are a number of limitations to note. First, the data were restricted to adults aged 18–60 years. We were therefore unable to assess how past 5-year suicide attempts varied for those under age 18. Considering that this is a critical developmental period for the onset of mental health disorders and suicide attempts, future research including both adolescents and adults would benefit from similar analyses. Second, our measures captured suicide attempts in the preceding 5 years. The assessment of past-year suicidality would provide more accurate estimates, but data sparseness necessitated a broader recall window. Third, because data were cross-sectional, we could not disentangle cohort effects from developmental differences. Considering SM-related social change in the past 40 years, there may be differences in risk for suicide attempts on the basis of when different cohorts of SM adults came of age.^52,53 Continued collection of sexual orientation data will help provide data necessary to test potential age–period–cohort effects on SM health disparities in the future.

Although Black and Hispanic SM adults demonstrated similar patterns of past 5-year suicide attempts and of discrimination, we cannot be certain that the association of discrimination with past 5-year suicide attempt is the same for Black and Hispanic SM adults. By collapsing these two groups together in response to power constraints, we were unable to evaluate potential differences in this association between Black and Hispanic SM adults. Finally, we did not explore differences by gender/sex in combination with differences by sexual orientation and race/ethnicity due to sparsity of data within gender/sex-by-race-by-sexual orientation subgroups. Likewise, data on Asian/Pacific Islander and Native American SM participants were insufficient for the analytic strategy we employed. Yet, descriptive results suggest elevated rates of past 5-year suicide attempts among SM adults in both groups, with especially high rates of suicide attempts among SM Native American adults (past 5-year attempts: 8.6%; lifetime: 26.2%). More research with these two subgroups is urgently needed. For now, we are limited by the available data.

Conclusions

Little is known about the suicide attempt disparities between SM and heterosexual adults across racial/ethnic groups and throughout the life course. The results of this study show that although periods of elevated past 5-year suicide attempt rates differed by race/ethnicity, all SM adults demonstrated disparities when contrasted with heterosexual adults of any race/ethnicity from age 18 through the early 30s. Sexual orientation-based discrimination appears to be an especially important factor for past 5-year suicide attempts from the ages of 18 to 25 years, when SM adults experiencing SM discrimination show elevated odds of past 5-year suicide attempts relative to SM adults of the same racial/ethnic group who did not experience SM discrimination. Taken together, these findings point to an extended period of elevated prevalence of past 5-year suicide attempts for SM adults of many races/ethnicities and reemphasize early adulthood as a period of potential vulnerability to discrimination, especially for Black and Hispanic SM adults.

Disclaimer

The NESARC protocol was originally approved by the US Census Bureau and the US Office of Management and Budget. Secondary data analysis for the current article was approved by The Pennsylvania State University and University of Maryland Institutional Review Boards. The views expressed in this article belong to the authors and do not necessarily represent the official views of the National Institutes of Health (NIH), the Centers for Disease Control and Prevention (CDC), the University of Maryland Prevention Research Center, or the Southern Regional Education Board.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

E.K.L. was supported by the National Institute on Drug Abuse (T32DA017629; P50DA039838) and the National Institute of Mental Health (T32MH020031) of the National Institutes of Health (NIH). C.E. was supported by the National Center for Advancing Translational Sciences (KL2TR002015; UL1TR002014) of the NIH. A.B.M. was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P2CHD042849; T32HD007081) and the National Institute of Mental Health (F31MH115608) of the NIH. N.D.W. acknowledges the support of the Southern Regional Education Board. J.N.F. was supported by the University of Maryland Prevention Research Center cooperative agreement number U48DP006382 from the Centers for Disease Control and Prevention (CDC) and the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P2CHD041041) of the NIH. The content of this study does not reflect the official views of the NIH or the CDC.

Supplementary Material

Supplementary Table S1