Are social supports always protective? A seven-city study on heavy drinking among sexual and gender minority young adults experiencing homelessness

Abstract

Purpose:

Sexual and gender minority (SGM) young adults are disproportionately impacted by homelessness and heavy drinking (i.e., having five or more drinks of alcohol in a row within a couple of hours). Social support, in general, is protective in reducing individuals’ risk of heavy drinking. However, whether and how support from different sources may have different implications on heavy drinking among SGM young adults experiencing homelessness (SGM-YAEH) remains unclear. Informed by the risk amplification and abatement model (RAAM), this study examined the associations between support sources and heavy drinking among SGM-YAEH.

Methods:

A purposive sample of SGM-YAEH (N=425) recruited in homeless service agencies from seven major cities in the U.S. completed a self-administered computerassisted anonymous survey. This survey covered heavy drinking behaviors and social network properties. Logistic regression models were conducted to identify social support sources associated with SGM-YAEH’s heavy drinking.

Results:

Over 40% of SGM-YAEH were involved in heavy drinking in the past 30 days. Receiving support from street-based peers (OR=1.9; 95% CI=1.1, 3.2) and homebased peers (OR=1.7; 95% CI=1.0, 2.8) were each positively associated with SGMYAEH heavy drinking risks.

Conclusion:

This study was not able to identify the protective role social supports may play in reducing SGM-YAEH’s heavy drinking. Furthermore, receiving support from network members was correlated with elevated heavy drinking risks among this population. As heavy drinking prevention programs develop interventions: they should use affirming and trauma approaches to promote protective social ties, as research points to its association in reducing alcohol use disparities among SGM-YAEH.

1. Introduction

Sexual and gender minority (SGM) individuals are overrepresented among the 4.2 million young adults experiencing homelessness (YAEH) in the U.S. [1–4]. While only comprising around 10% of the general population, it is estimated that SGM young adults constitute over 20% of the YAEH population [3, 5]. Furthermore, SGM-YAEH are disproportionately impacted by heavy drinking (i.e., having five or more drinks of alcohol in a row within a couple of hours [3, 6, 7]. Research suggests as high as 42% of SGM-YAEH engage in heavy drinking, compared to 27% of heterosexual and cisgender YAEH, and have a higher prevalence of lifetime alcohol use (87% vs. 77%) [3]. Extensive literature has identified factors associated with SGM-YAEH’s substance use risks. For example, SGM-YAEH are likely to have experienced childhood trauma and street victimization stemming from transphobia and homophobia [8–9]. Such trauma exposure based on their intersecting minoritized identities combined with systemic injustices of inequitable and unstable access to housing, education, and healthcare, exacerbated by cisgenderism and heterosexism [10] may contribute to SGM-YAEH’s higher risk of heavy drinking. Thus, excessive drinking may indeed serve as a mechanism for SGM-YAEH to cope with the trauma, discrimination, and excessive strain experienced while unstably housed [9, 11].

Social support has been documented as a protective factor in reducing risk engagement among YAEH related to substance use [6, 12]. Social support is defined as support sources (e.g., home-based peers, street-based peers, relatives, etc.) providing support via help and advice (emotional support), and money, housing, food, etc. [12–13]. Considering social support is associated with reduced substance use risk behaviors among YAEH, the ability to sustain these supportive social ties while experiencing homelessness may reduce young adults’ risk in heavy drinking specifically [14]. For example, having a trusting adult whom YAEH can count on for support when in need has been found to be protective in relation to reduced substance use among YAEH [6]. Furthermore, YAEH also counts on support from their network members to address their subsistence needs and to access services, including treatment that is critical to their substance use recovery [12]. However, YAEH are likely to be embedded in networks comprising of diverse members (e.g., providers, relatives, home-based peers, peers affiliated with gangs, etc. [15]), which may have different impacts on their heavy drinking behavior.

The Risk Amplification and Abatement Model (RAAM) argues that maintaining network members promoting positive social contact and experiences (e.g., peers who attend school regularly) may attenuate one’s involvement in risk behaviors; while preserving network members promoting negative social contact or experiences (e.g., peers with gang affiliation) may exacerbate one’s engagement in risky behaviors [16]. Additionally, YAEH are at a critical age in which their peer networks tend to have a strong influence [17], and the RAAM posits that YAEH seeking social support may gravitate toward peers with similarities or similar experiences, whether considered negative or positive. Furthermore, if there is limited opportunity to build positive relationships with network members or no positive social contact at all and especially in the context of unstable housing, risk-taking behavior may be compounded [14, 16, 18]. Thus, receiving social support from varied sources may have different implications for YAEH’s heavy drinking behavior. Research focused on the role of different social support sources on SGM-YAEH’s heavy drinking involvement remains scarce.

Former research has illustrated that SGM-YAEH’s social network composition may be distinctly different than their non-SGM peers [15]. A network-based study across four states found that LGB (lesbian, gay, bisexual) youth experiencing homelessness nominated street-based peers, family members other than a parent, friends, and professionals in their overall network at significantly higher rates than their heterosexual counterparts [13]. Furthermore, this study suggests that the social ties relied on for social support differ significantly from those among the heterosexual youth experiencing homelessness, as the LGB youth experiencing homelessness were significantly more likely to report receiving support from street-based peers, friends, or professionals, while significantly less likely to report support from “other” family members, as compared to their heterosexual counterparts [13]. Whereas Barman-Adhikari et al. [12] found that LGBQ (lesbian, gay, bisexual, queer) youth experiencing homelessness were significantly less likely to depend on home-based peers for support as compared to their heterosexual peers.

1.1. Study Aim

Considering SGM-YAEH are disproportionately impacted by heavy drinking and have different supportive sources as compared to their non-SGM peers, highlights the need for research to focus on SGM-YAEH’s sources of social support and their association with heavy drinking. Thus, informed by RAAM [16] and prior research, this study aimed to explore those relationships. A better understanding of this topic will inform future network-based interventions that seek to reduce heavy drinking among SGM-YAEH.

2. Methods

2.1. Sampling and Data Collection Procedure

This study used data from a larger cross-sectional project aimed at understanding YAEH’s health behaviors and service needs. The parent project involved a purposive sample of 1,426 YAEH. YAEH were recruited from service agencies, such as drop-in centers, emergency shelters, and transitional living programs in Los Angeles, Denver, Houston, Phoenix, San Jose, New York, and St. Louis between 2016 and 2017. The eligibility criteria for the parent project were young adults aged 18–26 years old and experiencing homelessness. Experiencing homelessness was defined as one or more of the following: spent the prior night in a shelter; on the streets; in a location not meant for human habitation; or in place where the length of stay may be limited to 30 days or less, including doubled up with others; and housing provided through a temporary voucher. Verbal informed consent was obtained from eligible respondents. Each participant was assigned a unique identification code developed based on the following formula: the first letter of their mother’s first name, the number of older brothers, the number representing the month they were born, and the first letter of their middle name. Such a strategy allowed the research team to maintain respondents’ anonymity considering the sensitivities of the survey questions, while at the same time identifying potential duplications.

Following a standardized protocol, respondents across study cities completed a self-administered anonymous survey on tablets. The first component focused on individual attributes, including respondents’ demographics (e.g., age and race), trauma exposure, and substance use behaviors. The second component focused on respondents’ social network composition. This study adopted the name generator developed by Rice et al. [19] to collect YAEH’s social network data. Respondents were first asked to nominate five network members with whom they were close to and had interacted in the previous three months. Such interactions included both in-person (i.e., face-to-face interaction) and on-line interaction (e.g., through social media and emails). Information regarding the respondent’s relationship with, attributes of, and support received from each nominated person was then collected. Respondents received $20–$25 gift cards to local venders for their participation. IRB approvals were received from each institution involved in the parent study. Detailed methodology can be found elsewhere (masked for review).

For the purposes of this study, only respondents who self-identified as sexual minority (i.e., gay, lesbian, bisexual, questioning, and self-reported categories not listed) and/or gender minority (i.e., anyone who did not identify as cisgender) in the parent project (N=425) were included.

2.2. Measurements

2.2.1. Dependent Variable

The outcome of interest was past 30-day heavy drinking (i.e., having five or more drinks of alcohol in a row within a couple of hours) among SGM-YAEH [3]. The outcome variable was a dichotomous variable (1= yes, 0= no) indicating whether SGM-YAEH had drank heavily in the past 30 days.

2.2.2. Independent Variables

2.2.2a. Social Support

This study’s focal independent variables were SGM-YAEH’s sources of social support. SGM-YAEH’s social support sources were derived combining SGM-YAEH’s responses to the questions whether they had received any support (i.e., money, advice, and service information) from each of the five nominated network members and what was their relationship with the corresponding network members (e.g., relatives, peers who attended school or employment regularly). Specifically, based on RAAM, eight dichotomous variables were derived with 1 depicting SGM-YAEH receiving any types of social support from either: relatives, service providers, home-based peers (i.e., peers met prior to first episode of homelessness experience), street-based peers (i.e., peers met after first episode of homelessness experience), peers attending school/employment regularly, SGM peers, intimate partners, and peers affiliated with gangs [16].

Informed by previous literature on substance use among YAEH [4, 12], this study also included homeless duration, trauma exposure, discrimination experiences, and mental illness diagnosis as independent variables.

2.2.2b. Duration of Homelessness

Duration of homelessness is a dichotomous variable reflecting whether a respondent had experienced at least two years of homelessness in their lifetime (1= yes; 0= no).

2.2.2c. Trauma Exposure

Trauma exposure included childhood trauma and street victimization history. Childhood trauma exposure was measured using the Adverse Childhood Experiences Scale [20]. ACEs include ten dichotomous items with each item reflecting a specific type of adverse childhood experience, such as witnessing domestic violence (1= had been exposed to the specific type of adverse experience growing up). ACEs scores are calculated by summing and the score ranges between 0 and 10. A score of zero indicates no exposure to adverse childhood experiences, while a score of 10 would indicate experiencing all the possible adverse childhood experiences provided as a response option. ACEs have been applied to various populations, including youth experiencing homelessness, with good validity and reliability [21].

Street victimization was a dichotomous variable (1= yes, 0= no) and respondents indicated whether they had been either robbed, or threatened, or assaulted, or witnessed victimization during homelessness.

2.2.2d. Discrimination Experiences

Discrimination experiences were derived from SGM-YAEH’s responses to the Experiences of Everyday Discrimination Scale (EEDS) [22]. EEDS is a five-item, five-point Likert scale that assesses the frequency of experiencing different types of discrimination (e.g., treated with less courtesy and respect than other people) during a year timeframe; response options include never, less than once a year, a few times a year, at least once a week, or always. Consistent with previous reporting [11, 23]. SGM-YAEH’s responses to EEDS were dichotomized with 1= had experienced any types of discrimination at least a few times a year.

2.2.2e. Mental Illness/Diagnosis

Mental illness diagnosis was a dichotomous variable (1= yes, 0= no) reflecting whether a respondent had received a diagnosis of a mental illness from a health provider, including attention deficit – hyperactivity disorder, post-traumatic stress disorder, conduct disorder, major depression, bipolar disorder, schizophrenia or schizoaffective disorder, and anxiety.

2.2.3. Demographic Controls

YAEH’s demographics were included as control variables in this study. These variables covered the seven study cities (nominal), age (continuous), race and ethnicity (nominal; 1= White, 2= Black, 3= Latinx, and 4= Multiracial or Other), gender (nominal; 1= cisgender man, 2= cisgender woman, and 3= gender minority, including people who are transgender or gender nonconforming), and education (dichotomous; 1= at least had high school diploma or GED).

3. Analysis

Less than 3% of the data was missing across all variables. Therefore, listwise-deletion logistic regression models following a stepwise approach [24] were constructed to explore the relationships between different support sources and SGM-YAEH’s heavy drinking. Specifically, bivariate logistic regression models were first conducted to evaluate the relationship between the independent variables and the outcome of interest. Based on the results of the bivariate analyses, independent variables found to be significantly associated with SGM-YAEH’s past 30-day heavy drinking involvement (p <0.05) were included in the final multivariate model. It should be noted that all demographic controls were included in the final multiple logistic regression model regardless of their relationships with the outcome in the bivariate analyses. However, sexual orientation was not included in the multivariate model, as only 29 SGM-YAEH (6.8%) identified as heterosexual. Finally, all necessary diagnostic and assumption testing was performed on the final model with no concerns noted.

4. Results

Over 40% of SGM-YAEH drank heavily in the past 30 days. On average, participants were 20.9 years old (SD=2.1). SGM-YAEH in this study were predominantly racial minorities (91.6%) and had at least a high school diploma or GED (70.8%).

Approximately 57% of the respondents reported having at least one SGM peer supportive tie in their network, making peers who are also SGM constituting the major support source among SGM-YAEH. Receiving support from home-based peers (OR=1.7; 95% CI=1.0, 2.8) or from street-based peers (OR=1.9; 95% CI=1.11, 3.2) were both associated with SGM-YAEH’s increased heavy drinking risk, and these were the only network support sources significantly related to the outcome. Although not the major focus of the study, street victimization (OR=2.3; 95% CI=1.1, 4.6) was also associated with elevated heavy drinking risk. Finally, cisgender women, as compared to cisgender men were at a lower risk of heavy drinking involvement (OR=0.5; 95% CI=0.3, 0.8).

5. Discussion

Consistent with previous literature, SGM-YAEH in this study were at a higher risk of heavy drinking as compared to their non-SGM peers (40% vs. 31%; χ2=10.8) [3, 6, 7]. Potential factors that may be associated with this increased heavy drinking risk may include lack of sexuality and gender affirming care and resources, that are rather rooted in heterosexism and cisgenderism [10], trauma experiences [8–9], along with other inequities (e.g., inequitable access to stable housing and healthcare) faced by the SGM community [10]. The remainder of our discussion will point to the possibilities of these relations in the context of the current study’s findings. Furthermore, not only is the increased heavy drinking risk consistent with prior literature, but our findings also reflect minority stress theory within quantitative and qualitative findings [25–26]. A future direction for research should consider social support sources and these relations using the constructs of minority stress theory.

Heavy drinking involvement may hinder SGM-YAEH’s transitioning from homelessness to housing and may pose as a barrier for them to be sustained in housing programs [27]. Therefore, our finding highlights the imperative need for heavy drinking risk reduction and recovery programs specifically geared toward SGM-YAEH. A systematic review of the literature points out that currently, there are no evidence-based heavy drinking prevention programs specifically tailored for SGM-YAEH’s needs as the review found that most random control trial tested interventions have focused solely on men who have sex with men [28]. However, another systematic review conducted by Dimova et al. [29] that focuses on SGM adults, and their alcohol use may shed some light on potential interventions for adaptation to reduce heavy drinking among SGM-YAEH. Specifically, considering SGM-YAEH’s transiency, brief interventions, such as a tailored feedback intervention [30] and personalized cognitive counseling [31], may be especially suitable for low-barrier drop-in center settings that SGM-YAEH are likely to utilize to meet subsistence needs. Furthermore, with the high prevalence of smart phone access and technology literacy among YAEH [32], substance use reduction programs, such as an online brief expressive writing and self-affirmation intervention [33], and the Screening, Brief Intervention, and Referral to Treatment (SBIRT) may warrant for future adaptation efforts to address heavy drinking risks among this vulnerable and transient population [34].

The ability to foster, manage, and sustain social support ties are critical for individuals experiencing homelessness to meet subsistence needs (e.g., food and shelter), access to services and treatment, and build social capital to exit homelessness [12, 16, 35]. SGM-YAEH in our study demonstrated diverse sources for support. Specifically, peers with shared SGM identities, relatives, and romantic partners are the major sources of social support for SGM-YAEH. Given that peers with similarities tend to connect [18], it is not surprising that SGM-YAEH may rely on network members with whom they have close relationships or share identities and experiences for social support. Previous research suggests that social support may be protective against substance use risks among youth and young adults, including YAEH (6, 12, 36–37]. However, contrasted to previous literature (6, 12, 36–37], this study failed to identify social support sources that may reduce SGM-YAEH’s heavy drinking risks. Furthermore, our study finds that social support, whether received from home-based peers or street-based peers is associated with SGM-YAEH’s elevated heavy drinking risk.

Drinking is a more socially accepted behavior in society, as compared to using illicit substances. In fact, drinking is often incorporated in social events to facilitate interactions, and further enhance marketing campaigns targeting SGM populations [38–39]. It is possible that peers, regardless of home-based or street-based, whom SGM-YAEH rely on for social support are close social ties with whom SGM-YAEH engage in social activities where drinking is encouraged and happening. Further, it should be noted that our findings do not suggest that network-based heavy drinking interventions may not be effective or needed for SGM-YAEH. Peer-led risk reduction interventions have shown positive impacts on promoting YAEH’s behavioral and mental health outcomes, or at the very least connects them with much needed services [40]. Perhaps, the results are indicative of the lack of opportunity to develop positive social support while experiencing homelessness. Therefore, our findings suggest that with the positive association between peer support and heavy drinking, future interventions focused on fostering peer supports with the goal to reduce SGM-YAEH’s heavy drinking risk should be aware of the potential risks among home-based and street-based peer social support and the limited promotion of protective social support.

Compared to the RAAM [16], this study failed to identify the potential protective role receiving support from ties that would presumably promote a protective aspect, such as providers, may have on SGM-YAEH’s heavy drinking risk. In this study, SGM-YAEH tends to not count on providers for social support regularly. Our finding may echo previous literature on SGM-YAEH’s discriminative experiences interacting with the education, health care, justice, and social service system [9–11], which may reduce the protective impacts these ties may have on SGM-YAEH’s heavy drinking behaviors. Moreover, it is possible our study is missing relevant support sources by only including the closest five social network members nominated by SGM-YAEH. Limiting the number of network members YAEH nominate may only identify a few of SGM-YAEH providers in their networks, and thus results in the current insignificant finding. Social interaction and support from providers, although may not be SGM-YAEH’s close network ties, can serve as critical bridges that connect SGM-YAEH with information, resources, and positive norms [41–42]. Therefore, future research focused on network influences on SGM-YAEH’s heavy drinking risk should consider expanding the network size investigated to capture other potential positive supportive ties (e.g., case managers, social workers, health providers, etc.) that may reduce SGM-YAEH’s heavy drinking risks. However, in contrast to the prior possibility, it may be that our findings relate to quality, rather than the quantity of social support ties. Given the traumatic experiences documented among SGM-YAEH, such as not feeling safe and affirmed may prohibit the opportunity to connect to protective social ties, (e.g., providers, etc.), as SGM-YAEH may not perceive or receive these services as protective [42]. More research is needed to understand the nuances of the relationship among different supportive social ties.

Moreover, this study did not identify a protective role that maintaining supportive ties with relatives [6, 12,16] may have on SGM-YAEH’s heavy drinking involvement, as identified in some studies focusing on YAEH. Whereas other literature suggests that a segment of SGM-YAEH may have complex, even conflicted relationships with their relatives, which may contribute to their current unstable housing situation [43]. Therefore, it is possible that although some SGM-YAEH could receive much needed support from relatives, others’ relationships with family members may be contentious, minimizing the protective role of family members’ support.

Consistent with prior literature on heavy drinking among YAEH in general, our study found that SGM-YAEH’s past street victimization was associated with their elevated risk of heavy drinking [44]. It is well-established that SGM-YAEH are especially susceptible to multiple types of traumas, such as street victimization [1, 3, 8, 9, 44]. In fact, over 83% of SGM-YAEH in our study reported to had been exposed to some type of violence while experiencing homelessness, significantly higher than their non-SGM peers (72%; χ2=22.4). Previous literature has suggested that substance use, especially drinking, may be adopted as a “self-medication” strategy for youth and young adults with oppressed experiences or unstable housing status [45] to cope with their environmental stressors, including trauma [46–47]. Future research should focus on how the intersection of homelessness and SGM identity may impact young adults’ use of substances as a coping mechanism. Moreover, heavy drinking risk reduction interventions targeting SGM-YAEH should be cognizant of the high prevalence of trauma incorporating trauma informed components [48], as well as sexuality and gender affirming care [42]. Such strategies would ensure that SGM-YAEH receives trauma related care when addressing their heavy drinking and that they feel safe and comfortable accessing service agencies.

6. Strengths and Limitations

Using data collected from YAEH across seven major cities in the U.S., this study expands previous literature limited in geographical coverage on correlates of heavy drinking among SGM-YAEH. Furthermore, as one of the few studies to investigate the associations between sources of social support and heavy drinking among SGM-YAEH, our findings provide critical implications on future research directions and intervention targets, as discussed in section 5.

Although having numerous strengths, our study has some limitations. Although we were mindful of the timeframes of variables included in our analysis (e.g., lifetime street victimization experiences vs. past 30-day heavy drinking involvement), the cross-sectional nature of the study makes causal inferences impossible. Another limitation of this study is that it does not control for normative impacts from social network members nor capture social network members specific heavy drinking behavior, due to constraints in available data from the parent project. Future studies should include these variables for a more rigorous study. Participants were only asked to nominate five network members, and thus were limited to close network ties. Moreover, we were not able to assess the climate of the service agencies that partnered in this study, such as known affirming organizations and places where the SGM-YAEH feel safe to access. This is an important aspect to explore in future studies as research points to these safe spaces as crucial to the reduction of substance use disparities among SGM-YAEH [42].

Finally, although our dataset could be considered outdated, we still chose to utilize it as there is limited data collecting this breadth of information across seven cities from a difficult to reach subpopulation. This study was able to be conducted as we had enough sample size among the subset of SGM-YAEH. Future studies should continue to reach larger sample sizes across multiple geographical areas of coverage to continue to understand the individual needs of SGM-YAEH and for the generalizability of results.

7. Conclusion

SGM young adults are disproportionally impacted by homelessness and heavy drinking [1–4]. The intersection of sexual and gender minority status and homelessness in heavy drinking is less studied, resulting in the lack of evidence-based interventions targeting SGM-YAEH’s heavy drinking risks. Although supportive ties may play a critical role in SGM-YAEH’s survival during homelessness [12, 16], this study highlights potential harmful impacts supportive ties may have on their heavy drinking risk. Considering the influence network members may have on SGM-YAEH’s health behaviors [6, 12, 16], future studies should build upon the current study to explore social ties that could be protective to SGM-YAEH’s heavy drinking involvement.

Table 1. Descriptive statistics of sexual and gender minority young adults experiencing homelessness (SGM-YAEH; N=425).

	Descriptive statistics
	N (%)	Mean (SD)
Outcome of Interest
Past 30-day heavy drinking involvement	170 (40.5)
Demographic Control Variables
Study cities
Los Angeles	88 (20.7)
Denver	53 (12.5)
Houston	51 (12.0)
Phoenix	85 (20.0)
New York	61 (14.4)
St. Jose	45 (10.6)
St. Louis	42 (9.9)
Age (years)		20.9 (2.1)
Gender Identity
Cisgender man	140 (33.1)
Cisgender woman	184 (43.5)
Gender minority	99 (23.4)
Sexual orientation
Heterosexual	29 (6.8)
Sexual minority	396 (93.2)
Race/Ethnicity
White	78 (18.4)
Black	127 (29.9)
Latinx	69 (16.2)
Multiracial or others	151 (35.5)
Education
High school or GED	301 (70.8)
Independent Variables
Social Support Sources (i.e., received any supports from specific type of network members)
Relatives	204 (48.0)
Service providers	50 (11.8)
Home-based peers	104 (24.5)
Street-based peers	111 (26.1)
Peers attending schools/employment regularly	35 (8.2)
Sexual and gender minority peers	243 (57.2)
Intimate partners	130 (30.6)
Peers with gang affiliation	45 (10.6)
Duration of Homelessness
Homelessness at least 2 years	136 (32.0)
Trauma Exposure
Adverse childhood experiences		5.5 (2.9)
Street victimization experiences	355 (83.5)
Mental Health Status
Mental illness diagnosis	308 (73.7)

Table 2. Multiple logistic regression results of heavy drinking correlates among sexual and gender minority young adults experiencing homelessness (SGM-YAEH; N=425)^a,^b.

	Past 30-Day Heavy Drinking
	OR	95% CI
Demographic Control Variables
Study Cities (ref: Los Angeles)
Denver	0.7	0.3, 1.5
Houston	0.8	0.4, 1.9
Phoenix	0.7	0.4, 1.4
New York	0.7	0.3, 1.4
San Jose	1.0	0.4, 2.2
St. Louis	0.6	0.2, 1.5
Age	1.0	0.9, 1.2
Gender (ref: cisgender man)
Cisgender woman	0.5 **	0.3, 0.8
Gender minority	0.6	0.3, 1.1
Race/Ethnicity (ref: White)
Black	0.7	0.4, 1.4
Latinx	1.4	0.7, 3.1
Multiracial or Others	0.7	0.4, 1.2
Education
High school or GED	0.7	0.4, 1.2
Independent Variables
Social Support Sources
Home-Based peers	1.7 *	1.1, 2.8
Street-Based peers	1.9 *	1.1, 3.2
Sexual and gender minority peers	1.1	0.7, 1.8
Duration of Homelessness
Homelessness at least 2 years	1.0	0.6, 1.7
Trauma Exposure
Adverse childhood experiences	1.0	0.9, 1.1
Street victimization	2.3 *	1.1, 4.6
Mental Health Status
Mental illness diagnosis	0.9	0.5, 1.5

^aBoldface indicates statistical significance

^*p < .05

^**p < .01

^bOnly independent variables found to be significant in bivariate analysis (p<.05) were included in the final model. However, sexual orientation was not included in the multivariate model, as only 29 SGM-YAEH (6.8%) identified as heterosexual. Sociodemographic controls were included in the multivariate model.

Highlights.

• For sexual and gender minority young adults experiencing homelessness (SGM-YAEH), receiving social support from street-based and home-based peers is significantly associated with heavy drinking

• Over 40% of SGM-YAEH had drank heavily in the past 30 days

• Further research is needed to explore which supportive ties are protective of SGM-YAEH’s heavy drinking

• Heavy drinking reduction programs should be tailored to SGM-YAEH distinctive needs, including affirmative and trauma informed approaches