Exploring Differences in the Lives and Well-Being of Sexual and Gender Minority Adults Experiencing Homelessness Relative to Their Cisgender Heterosexual Counterparts

Abstract Purpose: To better understand the lives and experiences of sexual and gender minority (SGM) adults experiencing homelessness relative to their cisgender heterosexual (non-SGM) counterparts. Methods: A modified time-location sampling strategy was used to reach a diverse sample of individuals with experiences of homelessness. Interviewer or self-administered paper-based surveys were administered to participants on location. Results: SGM and non-SGM participants reported significant differences in the age at which they became homeless, their current housing, and experiences of violence over the past year. SGM participants reported poorer mental health than their non-SGM counterparts. Conclusion: SGM adults may be uniquely impacted by their experiences of homelessness.


Introduction
The disproportionate number of homeless youth who identify as sexual and gender minorities (SGM) is widely cited, with considerable research and resources directed toward rectifying this inequity. [1][2][3][4] Elevated rates of homelessness among SGM youth is often attributed to familial rejection of the youths' sexual or gender identity. 1 What remains unknown is whether homelessness among SGM youth has only recently increased or whether this disparity has gone undocumented over time. Although either is conceivable, societal acceptance of SGM persons has increased over the years, 5 suggesting that SGM youth of earlier generations may have been historically overrepresented among homeless youth. If this is the case, SGM homeless youth of earlier generations may have since transitioned into a disproportionate number of housing unstable SGM adults. Yet, there is much less known about SGM adults experiencing homelessness, with only one known study comparing the causes of homelessness between SGM adults and their cisgender heterosexual (non-SGM) counterparts. 6 Although informative, the study was not designed to assess SGM-specific reasons for homelessness and found no differences between SGM and non-SGM adults. 6 A recent review on SGM adults experiencing homelessness specifically pointed to the lack of information on the pathways to homelessness as a critical gap in the literature. 7 In addition to a lack of understanding about the pathways to homelessness among SGM adults, the current literature tends to focus on HIV or substance use, 8,9 with little information about the ways in which the lived experiences of SGM adults experiencing homelessness may differ from their non-SGM counterparts. The sampling methods used in previous studies are similarly limited with the majority sampling homeless individuals from within SGM communities. [10][11][12][13] Instead, assessing SGM identity among a sample of individuals experiencing homelessness would increase generalizability and allow for comparisons between SGM and non-SGM adults. To partially address this gap, the present study explores differences in causes, experiences, and health of a diverse sample of SGM and non-SGM adults experiencing homelessness.

Methods
The present study was conducted as part of a need assessment for a Federally Qualified Health Center that provides care exclusively to individuals with experiences of homelessness. Most research with individuals experiencing homelessness does not include a representative sample given the sampling strategy challenges with this population. To ensure that our sample included individuals with diverse experiences of homelessness, a modified time-location sampling (TLS) strategy was used in the present study. TLS has been used previously to sample hard-to-reach populations often in the context of infectious disease surveillance. 14,15 TLS involves the identification of a complete set of venues and the days and times at which the population visits these venues. The venues in this study included organizations that serve the homeless community, permanent supportive housing sites that house individuals who have experienced homelessness, and outdoor locations where people experiencing homelessness congregate. A universe of all possible venue, day-time units (VDTs) was established, and VDTs were drawn randomly to make a calendar of visits to venues for data collection. Teams of at least three people visited venues in 4-h blocks to count and approach all individuals at the venue with an invitation to participate in the study. Interviewer or selfadministered paper-based surveys were administered to individuals on location, and participants were provided with a hygiene kit at completion of the program.
Survey questions included sex at birth, 16 gender identity, 16 sexual identity, 16 race/ethnicity, 17 education, 18 employment, 19 felony history, and veteran status. 20 Participants were asked where they slept in the past 12 months, the place they spent their past night, and perceived causes for homelessness. 21 Participants were asked to indicate needed services from a list of local services (e.g., clothing, ID). Consistent with the CDC's Health-Related Quality of Life measure and the mental health item in the Homeless Management Information System, participants rated their mental and physical health on a scale from poor to excellent. 22,23 Procedures were reviewed by a consumer advisory board before implementation. Secondary data analysis was approved by the UTHealth IRB. Differences between SGM and non-SGM participants were explored using chi-square, t-tests, logistic regression, and analysis of covariance.

Results
SGM participants were, on average, a decade (M = 9.70, SD = 1.14) younger (t = 4.87, p < 0.001) than non-SGM participants ( Table 1). In the past 12 months, after controlling for age and race, SGM participants were more likely to spend a night in jail or on the streets ( Table 2). When asked where they slept the night before and where they most frequently slept in the past 30 days, the largest percentage of non-SGM reported that they slept in permanent supportive housing (24.7%, N = 54), whereas the largest percentage of SGM participants reported sleeping on the streets (33.3%, N = 11).
SGM participants reported becoming homeless at a younger age (M = 32.32, SD = 12.58) than their non-SGM counterparts (M = 38.88, SD = 14.98, t = 4.87, p < 0.001). SGM and non-SGM participants did not differ in the amount of time they spent homeless. The most common reason for homelessness reported by participants was that they lost their job or family income (non-SGM: 53.3%, N = 178; SGM = 52.2%, N = 24). SGM participants were more likely to report that they lost housing due to a change in assistance (e.g., temporary assistance for needy families) eligibility (SGM: 28.3%, N = 13, w 2 = 4.33, p < 0.05) or because of their sexuality or gender identity (SGM: 17.4%, N = 8, w 2 = 27.71, p < 0.001).
When asked about current needs, permanent housing was the most common need for SGM and non-SGM participants. SGM participants were more likely to report that they needed assistance getting an ID (w 2 = 4.82, p < 0.05), food stamps (w 2 = 5.74, p < 0.05), food (w 2 = 6.05, p < 0.05), and substance use treatment (w 2 = 6.47, p < 0.05). Needs did not significantly differ between SGM and non-SGM participants after controlling for sociodemographic and living characteristics.
SGM and non-SGM participants did not significantly differ in self-reported physical health. SGM participants reported significantly poorer mental health relative to their non-SGM counterparts (Table 3) in bivariate and multivariate models controlling for sociodemographic and living characteristics (F = 6.73, p < 0.01).

Discussion
This is the first known study to compare the lives, experiences, and needs of SGM and non-SGM adults experiencing homelessness. Consistent with previous research, 6 SGM and non-SGM participants reported few differences in their reasons for homelessness with the majority of participants in both groups reporting that their first experience of homelessness was as an adult. A sizable minority of SGM participants reported that they had to leave previous residences for sexuality or gender-related reasons. One gay-identified cisgender man reported that his homelessness was due to discrimination he experienced when landlords would not include his husband on the lease. Future researchers should continue to explore the complex and subtle ways in which one's sexuality and gender identity may increase their vulnerability for homelessness (e.g., employment discrimination and partner benefits) not captured in the present study.
In the present study, SGM individuals experiencing homelessness were more likely to report recent gun/ knife violence, sexual assault, and elevated service needs, a difference that may be largely attributable to differences in housing. SGM participants were more likely than non-SGM participants to report having spent a night on the streets or jail with the largest percentage of SGM participants reporting that they slept on the streets over the past year. The driving mechanism behind this housing difference remains unknown. It may be that SGM individuals are less likely to seek housing; that those who are in housing/shelters are less likely to identify as SGM; or that people who identify as SGM are not universally welcome in privately funded shelters/housing. 13,24 One participant interviewed in religious housing indicated that she used to identify as a lesbian, but now identifies as heterosexual/straight for religious/moral reasons. Perhaps relatedly, non-SGM and SGM participants significantly differed in their self-reported mental health. Together, these findings suggest that SGM adults experiencing homelessness may benefit from the intentionallyinclusive housing and service strategies designed for SGM youth experiencing homelessness. 13,24 The present study has several limitations, including the sample size, which limited both generalizability and our ability to disaggregate SGM participants to explore the diversity of their experiences. 2 Despite limitations, differences between SGM and non-SGM participants experiencing homelessness may have important policy and practice implications. For instance, adding sexual orientation and gender identity to the Fair Housing Act could increase access to housing and perceived security while housed for SGM individuals. The current lack of housing protection combined with the adverse experiences reported by SGM participants in the present study supports the potential benefit of assessing SGM identity on standardized assessment tools used to determine housing prioritization. Further consideration and development of tailored evidence-based interventions and system-level AOR = Odds ratio adjusted for race and age. ***p < 0.005, ****p < 0.001. Relationship to spending a night in the jail or the streets was significant after adjusting for the False Discovery Rate using the Benjamini-Hochberg procedure.