Purpose:

pmc

9102136

1030

J Adolesc Health

The Journal of adolescent health : official publication of the Society for Adolescent Medicine

1054-139X1879-1972

40838905

12743606

10.1016/j.jadohealth.2025.06.025

HHSPA2124312

Article

Exposure to Suicidal Behavior by Sexual and Gender Minority Status among Adolescents in the Southeastern United States

Clark

Kirsty A.

Ph.D.abO’Connor

Kelly E.

Ph.D.cPark

Elisa

dSmith

Phillip

Ph.D.eMehari

Krista

Ph.D.b

aDepartment of Medicine, Health, and Society, Vanderbilt University, Nashville, TennesseebDepartment of Psychology and Human Development, Vanderbilt University, Nashville, TennesseecDepartments of Surgery and Psychology, Virginia Commonwealth University, Richmond, VirginiadDepartment of Social and Behavioral Sciences, Yale University, New Haven, ConnecticuteDepartment of Psychology, University of South Alabama, Mobile, Alabama

Corresponding Author: Kirsty A. Clark, PhD, MPH, Department of Medicine, Health, and Society, Vanderbilt University, 2301 Vanderbilt Pl, PMB 35-1665, Nashville, TN, 37235-1665, kirsty.clark@vanderbilt.edu

13122025

102025

2182025

01102026

774659666Purpose:

To investigate exposure to others’ suicidal behavior in explaining the heightened risk for suicidal thoughts and behaviors (STBs) among sexual and gender minority (SGM) adolescents.

Methods:

From December 2022 to May 2023, this cross-sectional survey study recruited 1,128 students from four public high schools (grades 9–12) in one historically segregated county in the southeastern U.S. We estimated exposure to others’ suicidal behavior and one’s own STBs by SGM status.

Results:

SGM adolescents reported significantly higher levels of STBs and more prevalent exposure to friends’ and family members’ suicide attempts and deaths (all p < .01), with 50.5% of SGM adolescents reporting exposure to friends’ suicide attempts. In multivariate mediation models adjusted for confounders, exposure to friends’ and family members’ suicide attempts accounted for 7%–12% of the disparities in STBs between SGM and cisgender/heterosexual adolescents.

Discussion:

Exposure to others’ suicidal behavior is prevalent among SGM adolescents and associated with the group’s increased risk of STBs. Suicide prevention policies and programs targeting SGM adolescents should enhance school-based screening, increase access to SGM-affirmative mental health resources, and leverage peer-engaged interventions.

exposure to violenceself-injurious behaviorsexual orientationgender identityadolescence

Exposure to others’ suicidal behavior (i.e., suicide attempt and suicide death) is a risk factor for one’s own suicidal thoughts and behaviors (STBs) [1–3] especially during the sensitive developmental period of adolescence when peer relationships and social processes are particularly salient. Population-based research from the United States (US) shows that the prevalence of exposure to others’ suicidal behavior sharply increases during the adolescent period, with 13% of 13–15-year-olds and 21% of 16–17-year-olds reporting such exposure [4]. A cross-sectional study of approximately 4,000 14–15-year-olds found that over 40% of adolescents reported exposure to others’ suicide attempts, which was then linked to adolescents’ own increased depressive symptoms [5].

The link between exposure to suicidal behavior and STBs might be particularly relevant among sexual and gender minority (SGM) adolescents, who face sharply elevated rates of STBs when compared to their cisgender/heterosexual (cis/het) adolescent peers [6,7]. Previous studies attempting to explain this disparity have primarily focused on minority stress theory [8], highlighting that SGM youths’ greater exposure to discrimination on the basis of minority identity confers increased risk for STBs [9,10]. Given the high frequency of STBs among SGM adolescents due to their persistent exposure to minority stress [9,10], exposure to suicidal behavior may be more frequent within SGM peer networks, a hypothesis supported in both survey-based [5] and qualitative research [11].

Research conducted without attention to SGM identity has shown that repeated exposure to suicide within one’s social network can shape youths’ cultural scripts and social norms surrounding suicide, leading some to internalize the notion of suicide as a viable or expected option [12]. Some researchers posit that SGM adolescents in particular may encounter cultural scripts whereby suicidal behavior is perceived as expected or even as a “rite of passage” in response to chronic minority stress exposure, a notion can then be internalized within communities of SGM young people [13,14]. Consistent with this perspective, survey research has found that SGM adolescents and adults report more accepting and empathic attitudes towards suicidal behavior compared to their cis/het peers, perhaps reflecting a normalization of STBs within SGM social networks [13,15].

Previous research has shown that SGM adults’ exposure to others’ suicidal behavior is associated with their own depressive symptoms and psychiatric distress [16,17]. However, research involving adolescents is limited and no studies have investigated associations between such exposure and heightened risk for STBs among SGM adolescents. Research has also found that Black individuals are exposed to significantly more deaths of family members throughout their life course compared to non-Hispanic white individuals, likely contributing to increased health risks stratified by race, but the causal pathways for such disparities are not yet well understood [18]. Given the increasing rate of suicide among Black youth in the US [19,20] coupled with calls for more intersectional lenses applied to public health research, there is an urgent need for research focusing on risk factors for STBs in racially diverse samples of SGM adolescents. To address these gaps, we analyzed data from a large sample of racially diverse adolescents residing in one historically segregated county in the southeastern U.S. We had two primary hypotheses: (1) SGM adolescents, compared to cis/het adolescents, would report a greater level of STBs as well as greater exposure to suicidal behavior among friends but not family members; (2) exposure to friends’ suicidal behavior would partially explain the disparity in STBs between SGM and cis/het adolescents.

MethodsParticipants

A total of 1,126 students participated in the study and provided usable data. The sample consisted of primarily male (45%) or female (55%) adolescents (based on sex assigned at birth) between the ages of 14 and 21 years (M=16.2, SD=1.15). Most adolescents identified as African American/Black (51%) or White (35%) as their sole racial identity. In terms of ethnicity, 5% of adolescents identified as Latine. Inclusion criteria for youth included being at least 14 years old, being enrolled in one of the classes in participating schools that were selected to participate, having written consent from parents or guardians, being present at least one of the days of the surveys, and providing active assent. Public schools in one county were recruited to reflect the county’s ethnic diversity and range of urbanicity. The final sample included two predominantly Black schools (one urban, one suburban), one ethnically diverse urban school, and one predominantly White suburban school.

Procedures

This cross-sectional, observational study took place in one county in the southeastern U.S. Four high schools (ninth through twelfth grades) across two public school districts were recruited to participate in a study focused on gun-related injury prevention. Research team members provided study information and consent forms to students in the schools, and researchers returned to each school to collect consent forms. Adolescents received $10 for returning a consent form signed by a parent or legal guardian, regardless of whether consent to participate was given. Of the students enrolled in the schools, 2,587 were randomly selected for recruitment, 58% of whom returned the forms with parent or legal guardian written consent. Of those, 76% were present and available at school on one of the survey days. This resulted in a total participation rate of 44%. Surveys were administered in December 2022 in one school and in February, March, and May 2023 in the remaining three schools, respectively. Surveys were administered through Qualtrics surveys in classrooms; trained research team members were present in each classroom and were available to answer questions. Participants completed the survey in approximately 30 minutes on average, and they received $25 for completing any part of the survey. This study was approved by the university’s Institutional Review Board and by school district leaders in both of the participating school districts. All surveys were fully anonymous due to the study’s sensitive questions which prevented response to any student’s individual suicidality. Investigators shared school-level suicidality data with each school’s administration and school counselors.

MeasuresSuicidal Thoughts and Behaviors.

Adolescents’ STBs were measured by the Paykel Suicide Scale and the Suicide Capacity Scale. The Paykel Suicide Scale [21] is a 5-item measure designed to assess past-year suicidal ideation and attempts (e.g., “Has there been a time that you felt life was not worth living?” and “Have you made an attempt on your life?” Cronbach α=.85). The Suicide Capacity Scale (SCS-3) [22] is a 6-item scale designed to assess dispositional capacity (i.e., a disposition to have low fear of pain or death), acquired capacity (i.e., decrease in fear of pain or death over time), and Practical Capacity (i.e., access to and knowledge of suicide methods). Items are rated on a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree). The total scale score of the Suicide Capacity Scale had a Cronbach α of .73.

Family and Friend Suicidal Behavior was assessed with four items developed for this study. Participants answered two yes-or-no questions about friends and two about family members: “Have any of your [friends/family members] attempted suicide?” and “Have any of your [friends/family members] died by suicide?”

Exposure to Racism and Discrimination was assessed by an adapted version of the Adolescent Discrimination Distress Index [23], a 13-item measure designed to assess the frequency of experiences racial-ethnic discrimination among adolescents in the past 30 days. The measure was designed to assess peer discrimination (e.g., “Were you called insulting names by other kids because of your race/ethnicity?;” Cronbach α=.79); school discrimination (e.g., “Were you treated unfairly by teachers because of your race/ethnicity?;” Cronbach α=.91); and institutional discrimination (e.g., “Were you treated unfairly by a store clerk or security guard because of your race/ethnicity?;” Cronbach α=.86). Participants responded using a 6-point scale ranging from 1 (Never) to 6 (20 or more times). We used this variable as a covariate to account for the confounding effects of exposure to structural racism (rather than biological race) on the associations of interest [24].

Demographics and SGM Status.

Age was recorded in whole years. Participants reported their grade in school. Participants reported on race by indicating all that applied of the following options: White or Caucasian American; Black or African American; Asian or Asian American; American Indian or Alaska Native; Native Hawaiian or Other Pacific Islander; Middle Eastern or North African; or other. Hispanic or Latino/Latina ethnicity was assessed in a separate item.

SGM status was assessed using three questions adapted from the Fenway Institute guide to collecting sexual orientation and gender identity [25]. To assess sex assigned at birth, participants indicated whether they were born male, female, or intersex in response to the question, “When you were born, what did they say your sex was?” Sexual orientation was assessed with the following question: “Do you think of yourself as…” with possible responses including lesbian or gay; straight or heterosexual (not gay or lesbian); bisexual; queer; pansexual; and an open-ended response option. Gender identity was assessed with the following question: “What is your current gender identity” with possible response options including woman/girl; man/boy; nonbinary, genderqueer, or not exclusively female or male; transgender woman/girl; transgender man/boy; and an open-ended response option. For sexual orientation and gender identity items, trolling responses to open-ended responses (e.g., “attack helicopter”, “animals”) were coded as missing data. Adolescents who indicated having either a minority (non-heterosexual) sexual orientation or a minority (non-cisgender) gender identity were coded as SGM (=1) versus cis/het (=0).

Data Analysis Plan

Analyses were conducted in Mplus Version 8.11. Chi-square analyses and independent samples t-tests were conducted to examine differences between SGM and cis/het adolescents in demographic characteristics and variables of interest. Pearson correlations among study variables were also examined. All subsequent analyses were stratified by school and used maximum likelihood estimation with robust standard errors to handle missing data and account for non-normality of outcome variables [26]. Additionally, continuous variables were standardized within each model. To test the study’s hypotheses, a model-building approach was used in which simple mediation models were tested prior to testing a multivariate mediation model that included all four mediator variables in the same model. Age, sex, and peer, school, and institutional racial/ethnic discrimination were included in all mediation models as covariates.

Simple mediation models specified SGM status as the independent variable, one item representing exposure to friend or family suicide attempt or suicide as the mediator, and both the Suicide Capacity Scale and Paykel Suicide Scale as dependent variables assessing suicide capability and STBs, respectively. Next, a multiple mediation model that included all significant mediators from the previous step was tested. Given the use of dichotomous mediators, direct and indirect effects were calculated using the approach described by MacKinnon and Dwyer [27], which adjusts the regression coefficients to be on the same scale and thus enables the standard decomposition of the total effect into direct and indirect effects [28]. The calculated adjusted effects were then examined using bias corrected bootstrap estimates as implemented in Mplus [29]. Proportion mediated (i.e., proportion of the total effect that is attributed to the indirect path through each mediator) was calculated using the model constraint command in Mplus [30].

ResultsDescriptive Statistics

The results of chi-square analyses and independent samples t-tests comparing SGM (19%; n = 216) and cis/het adolescents (81%; n = 912) on demographic characteristics and study variables are reported in Table 1. Relative to cis/het adolescents, a greater proportion of SGM adolescents endorsed female sex at birth (SGM: 72.5%, cis/het: 50.6%; X² = 33.1, p <.001) and Hispanic/Latine ethnicity (SGM: 10.2%, cis/het: 3.6%; X² = 16.4, p <.001) and were slightly older on average (t = −2.9, p =.004). Approximately 16.4% of SGM adolescents reported a non-binary or transgender gender identity. SGM adolescents most commonly identified their sexual orientation as bisexual (41.8%).

A substantially greater proportion of SGM adolescents reported exposure to family and friend suicide attempts and deaths compared to cis/het adolescents (all ps <.01). Notably, among SGM adolescents, more than half had been exposed to a friend’s suicide attempt (50.2%) and over one-third had been exposed to a family member’s suicide attempt (36.3%). SGM adolescents also reported higher levels of suicide capability (t = −4.0, p <.001) and more suicidal ideation (t = −6.3, p <.001) and attempts (t = −6.9, p <.001) on the Paykel Suicide Scale (PSS). Thus, for all subsequent mediation analyses, the PSS total score was used as an overall measure of STBs.

Relative to cis/het adolescents, SGM adolescents reported higher levels of peer racial/ethnic discrimination (t = −2.5, p = .014), school racial/ethnic discrimination (t = −3.2, p =.002), and institutional racial/ethnic discrimination (t = −3.1, p =.002). There were no significant differences between SGM and cis/het adolescents in their grade level, school, or racial/ethnic identity.

Bivariate correlations are reported in Table 2. Exposure to friends’ suicide attempts was positively associated with exposure to family members’ suicide attempts (r = .38). Additionally, exposure to friend or family member suicide attempt was positively associated with exposure to friend or family member death by suicide (rs range from .21 to .40). Suicide capability and STBs were positively correlated (r = .28). Peer, school, and institutional racial/ethnic discrimination were strongly and positively correlated (rs range from .74 to .83).

Simple Mediation Models

Each simple mediation model included SGM status as the independent variable, one variable measuring exposure to suicidal behavior as the mediator (i.e., friend suicide attempt, friend suicide death, family member suicide attempt, family member suicide death), both suicide capability and STBs as dependent variables, and age, sex, and peer, school, and institutional racial/ethnic discrimination variables as covariates. Suicide capability and STBs, as measured by the SCS and PSS scores, were allowed to covary. The results are reported in Table 3. SGM status was indirectly associated with higher suicide capability and STBs via friends’ suicide attempts (both βs = .03, ps <.001) and via family members’ suicide attempts (βs = .02 and .01, respectively, both ps <.05). None of the indirect effects of SGM status on suicide capability or STBs via exposure to friend or family member death by suicide were statistically significant. Thus, exposure to friend and family member death by suicide variables were not included in the multivariate mediation model, described next.

Multiple Mediation Model

The multiple mediation model tested SGM status as the independent variable, exposure to friend and family member suicide attempt as mediators, and suicide capability and STBs as dependent variables, with age, sex, and peer, school, and institutional racial/ethnic discrimination as covariates (see Table 4). The mediators were allowed to covary, as were the dependent variables. SGM status was associated with exposure to friends’ suicide attempts (β = .04, p <.001) and exposure to friends’ suicide attempts was then associated with higher suicide capability (β = .79, p <.001) and STBs (β = .68, p <.001). Regarding the indirect effect, SGM status was associated with higher suicide capability and STBs scores via exposure to friends’ suicide attempts (both βs = .03, ps <.001). The indirect effect of SGM status on suicide capability and STBs via friends’ suicide attempts accounted for 12% and 7% of the total effect, respectively.

SGM status was associated with exposure to family members’ suicide attempts (β = .02, p =.011) and exposure to family members’ suicide attempts was then associated with higher suicide capability (β = .82, p <.001) and STBs (β = .43, p =.021). Regarding the indirect effect, SGM status was associated with higher suicide capability (β = .02, p = .015) but not STBs (β = .01, p =.092) via exposure to family members’ suicide attempts. The indirect effect of SGM status on suicide capability via family members’ suicide attempts accounted for 7% of the total effect, respectively.

Discussion

In this racially diverse high school-based sample of adolescents from the southeastern U.S., SGM adolescents reported greater STBs and were more frequently exposed to others’ suicidal behavior, including friends’ and family members’ suicide attempts and deaths, with the highest proportion observed in exposure to friends’ suicide attempts. As hypothesized, the increased exposure to friends’ suicide attempts partially accounted for the elevated levels of suicide capability and STBs exhibited among SGM adolescents. Unexpectedly, however, greater exposure to family members’ suicide attempts also partially explained higher suicide capability reported among SGM adolescents. We discuss the implications of these findings for theory, research, and public health interventions.

Our findings suggest that existing models of suicide risk among SGM adolescents should be expanded to include social network factors, particularly exposure to others’ suicide attempts. Indeed, only the indirect effects via exposure to others’ suicide attempts were statistically significant, whereas those via exposure to suicide deaths were not. One plausible explanation for this pattern is a difference in statistical power, given that exposure to suicide attempts was more common in our sample than exposure to suicide deaths. Alternatively, it may reflect qualitative differences in how adolescents experience and internalize these exposures. Suicide attempts occur more frequently and therefore may be witnessed or discussed more often, offering multiple opportunities for interpersonal reinforcement over time. Future research should investigate the distinct effects of repeated exposure to suicide attempts versus suicide deaths, with particular attention to how these processes unfold within SGM adolescents’ social networks.

Prior studies have demonstrated that mental disorders including STBs can spread through a contagion process within adolescent peer networks [31], which may be exacerbated when suicide is normalized or even romanticized. For example, media depictions of suicide, such as in the television series 13 Reasons Why [32], and the suicides of popular celebrities like Robin Williams and Anthony Bourdain [33], have been linked to increased suicide risk at the population level. The phenomenon of suicide contagion has also been documented in “suicide clusters” among adolescent populations, whereby a series of suicides occurs within a particular community or school [12,34]. A small but growing body of research, drawing from qualitative interviews with SGM suicide attempt survivors [11], administrative suicide death records [35], and survey data from LGB adolescents [13], suggests that suicidal behavior may be particularly frequent, normalized, and even accepted within SGM peer networks, especially in the context of persistent minority stress. In one qualitative study, survivors of near-fatal suicide attempts described how chronic exposure to minority stress contributed to viewing suicide as a viable option; feeling effectively erased by society, the act of ending one’s life no longer felt like a significant leap [11]. These experiences were often shared within peer networks of other young SGM individuals struggling with suicidal thoughts. These findings, in conjunction with the results of the current study, underscore the critical need to investigate suicide contagion dynamics specifically within SGM adolescent peer networks.

This study’s finding that SGM adolescents experience greater exposure to family members’ suicide attempts and deaths was unexpected. However, this finding aligns with recent research indicating that SGM people report elevated exposure to adverse childhood experiences, including living in the home with someone with depression, mental illness or suicidality [36]. While the reasons for this heightened exposure to household mental illness among SGM youth remain unknown, this may help to explain the increased exposure to family member suicide attempts and deaths, as well as the heightened risk of STBs, identified in the current study, an important avenue for future research. Considering the sample of racially diverse SGM adolescents in the current study, with more than half identifying as Black, this finding also aligns with previous research demonstrating that Black individuals experience heightened exposure to family members’ deaths throughout their lives, including during childhood [18].

This study had several limitations. First, it was conducted in a single county in the southeastern U.S., which may limit the generalizability of the findings to other regions. Future population-representative research could help to clarify these results at the population level. Second, the cross-sectional design is not ideal for mediation analyses and limits causal inference, underscoring the need for future prospective research that assesses SGM identity at one time point, exposure to others’ suicidal behavior at a subsequent time point, and adolescents’ own STBs at a third time point. Third, while the measures included assessments of exposure to others’ suicide attempts and deaths, they did not explore other characteristics of adolescents’ social networks, such as the network size and density, network members’ sexual orientations or gender identities, the severity of others’ suicidal behavior, and whether the exposure occurred in-person or online. Such factors may serve as important moderators of the observed associations, an important avenue for future research. Finally, while we incorporated exposure to racial discrimination into analyses as a covariate, we did not study potential synergistic effects of racial discrimination exposure on the observed associations. Examining the intersectional experiences of racial discrimination and SGM identity on exposure to suicidal behavior and STBs represents an important next step for research in this area.

Adolescent Health Implications

Peer-engaged interventions may be one strategy to address and disrupt the potential for suicide contagion within SGM adolescent peer networks. Sources of Strength, a school-based peer leader program, trains adolescents from diverse social circles, under adult mentorship, to shift norms within their peer groups by engaging with and connecting suicidal friends to trusted adults and coping resources [37]. Findings from the current study suggest that adapting interventions like Sources of Strength for SGM adolescents – perhaps by incorporating SGM-tailored messaging, activities focused on disrupting the normalization of suicide among SGM peers, and strategies for engaging affirming and supportive adults – could be a promising next step. Additionally, Gender Sexuality Alliances (GSAs) and their student leaders may be a promising peer-led support structure [38], particularly for reaching SGM students who are already engaged in these groups. Engaging GSAs in school-based suicide prevention interventions like Sources of Strength could be a practical way to leverage SGM youth’s existing trusted peer networks to promote mental health and reduce risk. Such adaptations could equip SGM adolescents with the skills to intervene effectively and connect suicidal peers to appropriate mental health resources. Yet, the current sociopolitical climate, characterized by increasing restrictions on LGBTQ+ rights and visibility poses significant challenges to the availability and implementation of SGM-affirmative programming, particularly within schools [39,40]. Advocating for policies that protect and promote access to inclusive mental health support for SGM adolescents, including school-based programs, is essential.

Conclusion

This study provides rigorous evidence highlighting the heightened exposure to suicidal behavior among SGM adolescents and its contribution to their increased risk for STBs. Future research should integrate exposure to others’ suicidal behavior into existing models of suicide risk while exploring the dynamics of suicide contagion within SGM peer networks and families. Addressing these factors through SGM-affirmative, school-based interventions and policy advocacy is an essential step forward in reducing suicide risk among SGM adolescents.

References1.

Abrutyn

, Mueller

. Are suicidal behaviors contagious in adolescence? Using longitudinal data to examine suicide suggestion. Am Sociol Rev. 2014;79(2):211–227. doi:10.1177/0003122413519445

26069341

Maple

, Cerel

, Sanford

, Is exposure to suicide beyond kin associated with risk for suicidal behavior? A systematic review of the evidence. Suicide Life Threat Behav. 2017;47(4):461–474. doi:10.1111/sltb.12308

27786372

Hill

NTM

, Robinson

, Pirkis

, Association of suicidal behavior with exposure to suicide and suicide attempt: A systematic review and multilevel meta-analysis. PLOS Medicine. 2020;17(3):e1003074. doi:10.1371/journal.pmed.1003074

32231381

Mitchell

, Turner

, Jones

. Youth exposure to suicide attempts: Relative impact on personal trauma symptoms. Am J Prev Med. 2019;56(1):109–115. doi:10.1016/j.amepre.2018.09.008

30573139

Mitchell

, Banyard

, Ybarra

. Overlap between exposure to suicidal behavior and indirect interpersonal violence: Evidence for a more integrated approach to violence research. J Interpers Violence. 2023;38(15–16):9189–9214. doi:10.1177/08862605231163238

37070812

Marshal

, Dietz

, Friedman

, Suicidality and depression disparities between sexual minority and heterosexual youth: A meta-analytic review. J Adolesc Health. 2011;49(2):115–123. doi:10.1016/j.jadohealth.2011.02.005

21783042

Thoma

, Salk

, Choukas-Bradley

, Suicidality disparities between transgender and cisgender adolescents. Pediatrics. 2019;144(5):e20191183. doi:10.1542/peds.2019-1183

31611339

Meyer

. Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychol Bull. 2003;129:674–697. doi:10.1037/0033-2909.129.5.674

12956539

de Lange

, Baams

, van Bergen

, Minority stress and suicidal ideation and suicide attempts among LGBT adolescents and young adults: A meta-analysis. LGBT Health. 2022;9(4):222–237. doi:10.1089/lgbt.2021.0106

35319281

10.

Pellicane

, Ciesla

. Associations between minority stress, depression, and suicidal ideation and attempts in transgender and gender diverse (TGD) individuals: Systematic review and meta-analysis. Clin Psychol Rev. 2022;91:102113. doi:10.1016/j.cpr.2021.102113

34973649

11.

Clark

, Salway

, McConocha

, How do sexual and gender minority people acquire the capability for suicide? Voices from survivors of near-fatal suicide attempts. SSM - Qualitative Research in Health. 2022;2:100044. doi:10.1016/j.ssmqr.2022.100044

35845718

12.

Abrutyn

, Mueller

, Osborne

. Rekeying cultural scripts for youth suicide: How social networks facilitate suicide diffusion and suicide clusters following exposure to suicide. Society and Mental Health. 2020;10(2):112–135. doi:10.1177/2156869319834063

13.

Canetto

, Antonelli

, Ciccotti

Suicidal as normal – A lesbian, gay, and bisexual youth script?

Crisis. 2021;42(4):292–300. doi:10.1027/0227-5910/a000730

33151088

14.

Savin-Williams

. Suicide attempts among sexual-minority youths: Population and measurement issues. J Consult Clin Psychol. 2001;69(6):983–991. doi:10.1037/0022-006X.69.6.983

11777125

15.

Blosnich

, Lytle

, Coulter

RWS

, Suicide acceptability and sexual orientation: Results from the General Social Survey 2008–2014. Arch Suicide Res. 2018;22(4):542–554. doi:10.1080/13811118.2017.1377132

16.

Clark

, Sexton

, McKay

. Association of sexual orientation with exposure to suicide and related emotional distress among US adults. Arch Suicide Res. 2023;27(4):1363–1372. doi:10.1080/13811118.2022.2127386

36165026

17.

Tucker

, Pardue-Bourgeois

, Snow

, The relationship between suicide-related exposure and personal history of suicidal behavior in transgender and gender-diverse veterans. LGBT Health. 2019;6(7):335–341. doi:10.1089/lgbt.2019.0120

31618166

18.

Umberson

Black deaths matter: Race, relationship loss, and effects on survivors. J Health Soc Behav. 2017;58(4):405–420. doi:10.1177/0022146517739317

29172766

19.

Meza

, Patel

, Bath

. Black youth suicide crisis: Prevalence rates, review of risk and protective factors, and current evidence-based practices. FOC. 2022;20(2):197–203. doi:10.1176/appi.focus.20210034

20.

Sheftall

, Vakil

, Ruch

, Black youth suicide: Investigation of current trends and precipitating circumstances. J Am Acad Child Adolesc Psychiatry. 2022;61(5):662–675. doi:10.1016/j.jaac.2021.08.021

34509592

21.

Paykel

, Myers

, Lindenthal

, Suicidal feelings in the general population: A prevalence study. Br J Psychiatry. 1974;124(582):460–469. doi:10.1192/bjp.124.5.460

4836376

22.

Klonsky

, May

. The Three-Step Theory (3ST): A new theory of suicide rooted in the “ideation-to-action” framework. Int J Cogn Ther. 2015;8(2):114–129. doi:10.1521/ijct.2015.8.2.114

23.

Fisher

, Wallace

, Fenton

. Discrimination distress during adolescence. J Youth Adolesc. 2000;29(6):679–695. doi:10.1023/A:1026455906512

24.

Paradies

, Ben

, Denson

, Racism as a determinant of health: A systematic review and meta-analysis. PLoS One. 2015;10(9):e0138511. doi:10.1371/journal.pone.0138511

26398658

25.

National LGBTQIA+ Health Education Center: A Program of the Fenway Institute. Ready, Set, Go! A guide for collecting data on sexual orientation and gender Identity; 2022. Accessed August 13, 2024. https://www.lgbtqiahealtheducation.org/wp-content/uploads/2022/05/TFIE-64_Updates2022_ReadySetGo_10_18_22.pdf

26.

Little

. Longitudinal structural equation modeling. Guilford Press; 2013.

27.

Mackinnon

, Dwyer

. Estimating mediated effects in prevention studies. Eval Rev. 1993;17(2):144–158. doi:10.1177/0193841X9301700202

28.

Stride

, Gardner

, Catley

, Mplus code for mediation, moderation, and moderated mediation models. Published online 2015. http://www.offbeat.group.shef.ac.uk/FIO/mplusmedmod.htm

29.

MacKinnon

Introduction to statistical mediation analysis. 2nd ed. Routledge; 2017. doi:10.4324/9780203809556

30.

Preacher

, Kelley

. Effect size measures for mediation models: Quantitative strategies for communicating indirect effects. Psychol Methods. 2011;16(2):93–115. doi:10.1037/a0022658

21500915

31.

Alho

, Gutvilig

, Niemi

, Transmission of mental disorders in adolescent peer networks. JAMA Psychiatry. 2024;81(9):882–888. doi:10.1001/jamapsychiatry.2024.1126

38776092

32.

Bridge

, Greenhouse

, Ruch

, Association between the release of Netflix’s 13 Reasons Why and suicide rates in the United States: An interrupted time series analysis. J Am Acad Child Adolesc Psychiatry. 2020;59(2):236–243. doi:10.1016/j.jaac.2019.04.020

31042568

33.

Shaman

, Kandula

, Pei

, Quantifying suicide contagion at population scale. Sci Adv. 2024;10(31):eadq4074. doi:10.1126/sciadv.adq4074

39083618

34.

Hawton

, Hill

NTM

, Gould

, Clustering of suicides in children and adolescents. Lancet Child Adolesc Health. 2020;4(1):58–67. doi:10.1016/S2352-4642(19)30335-9

31606323

35.

Clark

, Blosnich

. Sexual orientation and disclosure of suicidal thoughts before suicide mortality. Am J Prev Med. 2023;65(6):953–963. doi:10.1016/j.amepre.2023.07.002

37429387

36.

Tran

, Henkhaus

, Gonzales

. Adverse childhood experiences and mental distress among US adults by sexual orientation. JAMA Psychiatry. 2022;79(4):377–379. doi:10.1001/jamapsychiatry.2022.0001

35195677

37.

Wyman

, Brown

, LoMurray

, An outcome evaluation of the Sources of Strength suicide prevention program delivered by adolescent peer leaders in high schools. Am J Public Health. 2010;100(9):1653–1661. doi:10.2105/AJPH.2009.190025

20634440

38.

Marraccini

, Ingram

, Naser

, The roles of school in supporting LGBTQ+ youth: A systematic review and ecological framework for understanding risk for suicide-related thoughts and behaviors. J Sch Psychol. 2022;91:27–49. doi: 10.1016/j.jsp.2021.11.006

35190078

39.

Goldbach

, Rhoades

, Mamey

, Reducing behavioral health symptoms by addressing minority stressors in LGBTQ adolescents: a randomized controlled trial of Proud & Empowered. BMC Public Health. 2021;21(1):2315. doi:10.1186/s12889-021-12357-5

34949171

40.

McQuillan

, Lebovitz

, Harbin

. The disruptive power of policy erasure: How state legislators and school boards fail to take up trans-affirming policies while leaning into anti-LGBTQ+ policies. Educational Policy. 2024;38(3):642–699. doi:10.1177/08959048231218206

Table 1.

Comparison of Study Variables by Sexual and Gender Minority (SGM) Status^a.

	SGM (N = 215)			Cis/Het (N = 911)
	N	n	%	N	n	%	χ²

Sex at birth	211			856			33.1***
Male		58	27.5		423	49.4
Female		153	72.5		433	50.6
Gender	207			837			185.3***
Woman or girl		122	58.9		425	50.8
Man or boy		43	20.8		412	49.2
Nonbinary/Genderqueer		20	9.7		0	0.0
Transgender woman/girl		6	2.9		0	0.0
Transgender man/boy		8	3.9		0	0.0
Other		8	3.9		0	0.0
Sexual Orientation	208			782			907.2***
Heterosexual		11	5.3		782	100.0
Lesbian or gay		46	22.1		0	0.0
Bisexual		87	41.8		0	0.0
Queer		11	5.3		0	0.0
Pansexual		24	11.5		0	0.0
Other		29	13.9		0	0.0
Grade	213			859			6.8
Grade 9		36	16.9		188	21.9
Grade 10		64	30.0		278	32.4
Grade 11		62	29.1		249	29.0
Grade 12		51	23.9		144	16.8
School	195			821			8.4
School 1		45	23.1		149	18.1
School 2		42	21.5		164	20.0
School 3		39	20.0		134	16.3
School 4		69	35.4		374	45.6
Race	204			850			6.5
African American/Black		109	53.4		433	50.9
White		63	30.9		302	35.5
Asian or Asian American		5	2.5		11	1.3
American Indian or Alaska Native		3	1.5		15	1.8
Native Hawaiian or Other Pacific Islander		0	0.0		4	0.5
Middle Eastern or North African		1	0.5		2	0.2
Other only		2	1.0		13	1.5
Multiracial		21	10.3		70	8.2
Hispanic/Latine ethnicity	215	22	10.2	911	33	3.6	16.4***
Exposure to friend suicide attempt	215	108	50.2	908	251	27.6	41.9***
Exposure to friend death by suicide	214	27	12.6	909	64	7.0	7.1**
Exposure to family member suicide attempt	215	78	36.3	909	201	22.1	19.6***
Exposure to family member death by suicide	215	40	18.6	909	95	10.5	12.0***

	N	M	SD	N	M	SD	t

Age (in years)	214	16.4	1.2	846	16.1	1.1	−2.5*
Paykel Suicide Scale – Suicidal Ideation subscale^b	210	1.5	1.3	860	0.8	1.2	−6.3***
Paykel Suicide Scale – Suicide Attempt subscale^c	210	0.7	0.8	863	0.3	0.6	−6.9***
Paykel Suicide Scale – Total STBs score^d	210	2.1	1.9	858	1.1	1.6	−7.2***
Suicide Capability Scale^e	209	3.2	0.8	867	3.0	0.8	−4.0***
Peer Racial/Ethnic Discrimination^f	209	1.7	1.0	867	1.5	0.9	−2.5*
School Racial/Ethnic Discrimination^f	209	1.5	0.9	866	1.3	0.7	−3.2**
Institutional Racial/Ethnic Discrimination^f	209	1.6	1.0	867	1.4	0.8	−3.1**

Note. Cis/het=Cisgender/heterosexual youth.

SGM status was dummy-coded to reflect 0=cis/het and 1=SGM.

Possible scores on this measure range from 0 to 3.

Possible scores on this measure range from 0 to 2.

Possible scores on this measure range from 0 to 5.

Possible scores on this measure range from 1 to 5.

Possible scores on this measure range from 1 to 6.

p < .05,

p < .01,

***

p < .001.

Table 2.

Correlations, Means (M), and Standard Deviations (SD) of Study Variables.

	1	2	3	4	5	6	7	8	9	10	11	12

1. SGM status (1=SGM; 0=cis/het)	–
2. Exposure to friend suicide attempt	.20^†	–
3. Exposure to friend death by suicide	.09**	.29^†	–
4. Exposure to family member suicide attempt	.13^†	.38^†	.25^†	–
5. Exposure to family member death by suicide	.11^†	.21^†	.22^†	.40^†	–
6. Suicide capability (SCS)	.12^†	.20^†	.12^†	.21^†	.10^†	–
7. Suicidal thoughts and behaviors (PSS)	.23^†	.28^†	.20^†	.21^†	.13^†	.28^†	–
8. Age	.06**	−.05	.04	−.01	−.03	−.03	.01	–
9. Sex at birth ((1=male; 0=female)	−.17^†	−.29^†	−.09*	−.22^†	−.13^†	.03	−.26^†	.01	–
10. Peer racial/ethnic discrimination	.07**	.07*	.19^†	.10**	.09^†	.08**	.22^†	.04	.04	–
11. School racial/ethnic discrimination	.10^†	.02	.14^†	.04*	.07**	.02	.21^†	.09^†	.03	.79^†	–
12. Institutional racial/ethnic discrimination	.11^†	.04	.14^†	.07**	.08**	.04	.20^†	.11^†	.04	.74^†	.83^†	–
M	.20	.33	.08	.24	.12	3.03	1.34	16.14	.46	1.49	1.33	1.38
SD	.40	.47	.27	.43	.32	.82	1.72	1.13	.50	.90	.74	.80

p < .05,

p < .01,

†

p < .001.

SCS = Suicide Capability Scale (assesses capability for suicide); PSS = Paykel Suicide Scale

Table 3.

Simple Mediation Models.

Effect	Type	β	S.E.	95% CI (LL, UL)

Friend suicide attempt as the mediator
SGM → Friend suicide attempt	Direct	.04***	.01	(.03, .06)
Friend suicide attempt → SCS	Direct	.84***	.08	(.70, .95)
SGM → SCS	Indirect	.03***	.01	(.02, .05)
Friend suicide attempt → PSS	Direct	.72***	.09	(.55, .85)
SGM → PSS	Indirect	.03***	.01	(.02, .05)
Friend death by suicide as the mediator
SGM → Friend death by suicide	Direct	.01	.01	(.00, .02)
Friend death by suicide → SCS	Direct	.58**	.18	(.25, .83)
SGM → SCS	Indirect	.01	.00	(.00, .01)
Friend death by suicide → PSS	Direct	.58***	.13	(.34, .77)
SGM → PSS	Indirect	.01	.00	(.00, .01)
Family suicide attempt as the mediator
SGM → Family suicide attempt	Direct	.02**	.01	(.01, .03)
Family suicide attempt → SCS	Direct	.83***	.07	(.69, .94)
SGM → SCS	Indirect	.02*	.01	(.01, .03)
Family suicide attempt → PSS	Direct	.56***	.13	(.34, .75)
SGM → PSS	Indirect	.01*	.01	(.01, .02)
Family death by suicide as the mediator
SGM → Family death by suicide	Direct	.01*	.01	(.00, .03)
Family death by suicide → SCS	Direct	.59***	.16	(.28, .81)
SGM → SCS	Indirect	.01	.01	(.00, .02)
Family death by suicide → PSS	Direct	.34	.18	(.04, .61)
SGM → PSS	Indirect	.01	.00	(.00, .01)

Note. CI= Confidence Interval; LL=Lower Limit; UL=Upper Limit; SCS = Suicide Capability Scale (assesses capability for suicide); PSS = Paykel Suicide Scale (assesses suicidal thoughts and behaviors).

p < .05,

p < .01,

***

p < .001.

Table 4.

Multivariate Mediation Model.

		M1: Friend suicide attempt			M2: Family suicide attempt
Effect	Type	β	S.E.	95% CI	β	S.E.	95% CI


SGM → M	Direct	.04***	.01	(.03, .06)	.02*	.01	(.01, .04)
M → SCS	Direct	.79***	.11	(.57, .93)	.82***	.11	(.63, .95)
SGM → SCS	Indirect	.03***	.01	(.02, .05)	.02*	.01	(.01, .03)
M → PSS	Direct	.68***	.11	(.49, .84)	.43*	.21	(.09, .70)
SGM → PSS	Indirect	.03***	.01	(.02, .04)	.01	.01	(.00, .02)

Note. SCS = Suicide Capability Scale (assesses capability for suicide); PSS = Paykel Suicide Scale (assesses suicidal thoughts and behaviors).

p < .05,

p < .01,

***

p < .001.

Implications and Contribution

Sexual and gender minority (SGM) adolescents are frequently exposed to others’ suicide attempts and deaths within their social networks. This exposure partially explains SGM adolescents’ increased risk of suicidal thoughts and behaviors. Findings underscore the importance of targeted suicide prevention efforts, including peer-engaged interventions, to reduce disparities in adolescent suicidality.