Association of Fear and Mistrust with Stress Among Sheltered Homeless Adults and the Moderating Effects of Race and Sex

Geospatial analysis of associations among mental health need, housing need, and involuntary psychiatric hospitalizations of people experiencing homelessness in Los Angeles County

The purpose of this study was to use geospatial indicators of mental health need and homelessness in Los Angeles County Service Planning Areas (SPAs) and a psychiatric sample of adults who were homeless to investigate 1) overlap between SPA level of mental health need and corresponding volume of involuntary psychiatric hospitalizations over time; 2) overlap between SPA level of unsheltered homelessness and corresponding volume of involuntary psychiatric hospitalizations over time; and 3) associations between SPA level of mental health need, SPA level of unsheltered homelessness, and initiation of a mental health conservatorship for grave disability. A sample of 373 adults who were homeless and hospitalized on an involuntary psychiatric hold from 2016 to 2018 were linked to data from the Greater Los Angeles Homeless Count on unsheltered homelessness and from the California Health Interview Survey on need for mental health services and suicidality, using admission zip codes to link variables at the SPA level. Geospatial mapping and bivariate tests were used to examine geographic overlap of SPA mental health need and unsheltered homelessness with volume of involuntary psychiatric admissions over the study period. Multiple logistic regression modeling was used to examine associations of SPA mental health need and unsheltered homelessness with conservatorship initiation. The volume of patients admitted from SPAs with higher levels of mental illness need grew from 2016 to 2018 (Tau = 0.27, P < 0.001; Tau = 0.40, P < 0.001), but there were fewer patients admitted from SPAs with higher levels of unsheltered homelessness over the same years (Tau of -0.33, P < 0.001). Being admitted from SPAs with the highest levels of unsheltered homelessness was associated with higher odds of conservatorship initiation (OR = 1.73, 95% CI = 1.82-16.74). Results suggest a need for targeted mental health and housing services to reach areas of highest need in Los Angeles County.

Guest Support for Outdoor Smoke-Free Policies within a Homeless Shelter

Roughly 70-80% of adults experiencing homelessness smoke cigarettes. Smoke-free living/workplace policies are an empirically-supported tobacco control intervention. However, homeless shelters may be reluctant to implement smoke-free policies due to fears of it discouraging current/potential shelter guests from taking refuge there. The current study was meant to characterize guest support for on-property smoke-free policies within a homeless shelter with an extant indoor tobacco use ban amongst never smokers, former smokers, and current smokers to provide data on this point. Participants comprised a convenience sample of adult guests of a homeless shelter in Texas (N = 394, 28.2% women; 10.2% former; and 75.9% current smokers). Participant sociodemographics, smoking status, behavioral health diagnoses, and support for two versions of an on-property outdoor courtyard smoke-free policy (one partial, one complete) were assessed. Data were collected in two waves in a repeated cross-sectional design. Overall, 64.0% of participants supported a partial, and 32.0% a full smoking ban. Logistic regressions, controlling for wave of data collection, age, sex, and any additional significant predictors from a semi-adjusted model, examined associations between participant characteristics and policy support. Older participants (OR = 1.024, CI_0.95 = 1.005-1.044), non-veterans (OR = 2.523, CI_0.95 = 1.156-5.506), former smokers (OR = 2.730, CI_0.95 = 1.191-6.258), and those without severe mental illness (OR = 1.731, CI_0.95 = 1.061-2.824) had significantly greater odds of supporting a partial smoking ban. Relative to current smokers, never smokers (OR = 3.902, CI_0.95 = 2.133-7.137) and former smokers (OR = 8.257, CI_0.95 = 3.951-17.258) had significantly greater odds of supporting a complete smoking ban. The implementation of smoke-free living/workplace policies in homeless shelters may enjoy more support from guests-specifically, non-smokers-than anticipated by shelter administrators. Aside from reducing ambient smoke exposure for never and former smokers, these policies can help to reduce ubiquitous smoking cues for those who may want to quit, are undergoing a quit attempt, or are trying to maintain abstinence. Interventionists might partner with shelter guests, particularly smokers, to inform the roll-out of such policies for maximal acceptance and adoption.

Perceived Social Support Attenuates the Association between Stress and Health-Related Quality of Life among Adults Experiencing Homelessness

Health-related quality of life (HRQoL) is defined as a multidimensional subjective assessment of one’s physical and mental health. Homelessness is associated with numerous stressors that can reduce HRQoL. Social support is defined as the availability of individuals, or resources provided by individuals, to cope with stress. Interpersonal social support may be important in buffering HRQoL from the negative implications of stress. Here, we examine this association in a marginalized group known for high rates of physical and mental health comorbidities: adults experiencing homelessness. Participants (N = 581; 63.7% men; M_age = 43.6 ± 12.2) were recruited from homeless-serving agencies in Oklahoma City. Social support was measured with the 12-item Interpersonal Support Evaluation List (ISEL). HRQoL was measured by the Behavioral Risk Factor Surveillance System (BRFSS) survey using self-rated health, the number of poor mental and poor physical health days over the preceding 30 days, as well as the number of limited activity days as the result of poor mental and/or physical health. Perceived stress was assessed using the 4-item Perceived Stress Scale (PSS). The potential moderation effect of social support was examined by assessing the interaction term of social support and stress in a series of linear regression analyses, controlling for sex, age, months homeless, race, education, health insurance status, serious mental illness diagnosis, and recruitment agency/site. There was a significant interaction effect of social support and stress on the prediction of days of poor physical health, days of poor mental health, and days of limited activity (p in all cases ≤ 0.05). Results add to a growing literature on the potentially protective benefits of social support for HRQoL, extend them to a large sample of adults experiencing homelessness in the South, and demonstrate the significance of this moderating effect of social support over and above the influence of several prominent sociodemographic and diagnostic variables. Future work should determine if interventions designed to enhance social support can buffer HRQoL from the deleterious effects of stress among this vulnerable population.

Mental Illness and Youth-Onset Homelessness: A Retrospective Study among Adults Experiencing Homelessness

Financial challenges, social and material instability, familial problems, living conditions, structural issues, and mental health problems have been shown to contribute to youth homelessness. Based on the paucity of literature on mental illness as a reason for youth homelessness, the current study retrospectively evaluated the association between the timing of homelessness onset (youth versus adult) and mental illness as a reason for homelessness among homeless adults living in homeless shelters and/or receiving services from homeless-serving agencies. Homeless participants (N = 919; 67.3% men) were recruited within two independent studies from Dallas and Oklahoma. Covariate-adjusted logistic regressions were used to measure associations between homelessness onset and mental illness as a reason for current homelessness, history of specific mental illnesses, the historical presence of severe mental illness, and severe mental illness comorbidity. Overall, 29.5% of the sample reported youth-onset homelessness and 24.4% reported mental illness as the reason for current homelessness. Results indicated that mental illness as a reason for current homelessness (AOR = 1.62, 95% CI = 1.12–2.34), history of specific mental illnesses (Bipolar disorder–AOR = 1.75, 95% CI = 1.24–2.45, and Schizophrenia/schizoaffective disorder–AOR = 1.83, 95% CI = 1.22–2.74), history of severe mental illness (AOR = 1.48, 95% CI = 1.04–2.10), and severe mental illness comorbidities (AOR = 1.30, 95% CI: 1.11–1.52) were each associated with increased odds of youth-onset homelessness. A better understanding of these relationships could inform needs for early interventions and/or better prepare agencies that serve at-risk youth to address precursors to youth homelessness.

Predictors of Overnight and Emergency Treatment among Homeless Adults

High usage of emergency treatment and hospitalization has been reported among homeless individuals. Hence, this study aimed to identify the sociodemographic predictors associated with overnight and emergency hospital treatment among a sample of homeless adults. Participants were recruited from a shelter in Dallas, Texas (N = 354; Mage = 43.7 ± 11.7) and were predominantly uninsured, low-income men from various racial groups. The outcome variables were: (a) stayed overnight for treatment in a hospital; and (b) treated in a hospital emergency room. In logistic regression models, sex emerged as the only predictor of overnight treatment in a hospital (OR = 2.68, 95% CI = 1.61-4.47), and treatment in an emergency room (OR = 2.21, 95% CI = 1.34-3.65), such that women were more likely than men to be treated overnight and use emergency care. Targeted interventions and policies are needed to address homeless women's primary care needs and reduce costlier treatment.