Using an Ordinal Regression to Examine the Association between Housing Insecurity and Escalating Suicidal Behaviors Among Trans-Young Adults: A Structural Determinant Approach

Background: Homelessness is a serious public health issue that disproportionately burdens transgender and gender-diverse people (collectively referred to here as trans). It is known that trans people experience high rates of housing insecurity (homelessness or risk of homelessness) and suicidal behaviors, with emerging evidence of substantial co-occurrence. Gender minority discrimination experienced in families, public settings, and social services contributes to experiences of homelessness and the associated exposure to violence that exacerbates mental health problems, including PTSD, depression, and suicide. Among trans young adults of color, racism in the context of gender minority discrimination adds to risk exposure. These experiences, while housing insecure, often exacerbate prior trauma, family separation, and instability experienced earlier in life. Although conceptual models of gender minority stress and multi-minority stress suggest suicidal ideation and behavior as an outcome of multilevel drivers, there has been little examination of structural determinants such as housing. Thus, there is a need to investigate housing insecurity as a structural determinant of suicidal behavior among trans young adults.

Methods: To examine the association between housing insecurity and suicide risk, this study conducts a secondary analysis of the 2015 USTS. Data analysis uses sequential ordinal regression to explore the relationship between housing insecurity and suicidal behaviors by increasing severity based on Jaccobucci et al. 2020. In contrast to a binomial logistic regression, an ordinal logistic regression assesses how factors contribute to a one-unit increase in a dependent variable with ordered categories. The use of an ordinal regression is a response to the research generating many risk factors for ideation and attempts separately but having less knowledge on what escalates suicide behaviors. An ordinal regression assumes that an individual must pass through less severe states (i.e., ideation) to reach more severe states (i.e., attempts). Results will not only give insight into how housing insecurity escalates suicide risk but also the validity of this assumption compared to other proposed models of suicidal behaviors, such as the fluidity vulnerability model of a suicidal model of suicidal behavior (Bryan et al., 2020).

Results: Housing insecurity was associated with 1.5x (CI 1.42-1.70) the odds of escalating suicidal behaviors. Housing insecurity amplified suicide risk among trans men (OR=1.5, CI=1.24-1.80, z=-3.06) and illicit substance users (OR=1.51, CI=1.23-1.85, z=-2.31) as compared to housing secure trans young adults. Cumulative discrimination, depression, being kicked out by family members, and no family support for trans identity were associated with escalated suicidal behavior among both housing-secure and insecure trans young adults. Sexual orientation, community connections, education, race, and visual non-conformity were not significantly associated with increased odds of escalating suicidal behaviors.

Conclusion: Trans young adults disproportionately experience housing insecurity and suicidal behaviors. Housing insecurity, cumulative discrimination, and family disruption contribute to suicide risk progression among trans young adults. Ongoing analysis is needed to understand better how housing insecurity amplifies suicide risk, particularly among trans men and illicit substance users. Further, there is a need to examine resilience factors and how housing-insecure trans young adults navigate and draw upon structural resources to de-escalate crises.