UCLA

PURPOSE: We examined the impact of duration and number of homelessness episodes on health outcomes for unsheltered homeless young adults. METHODS: We analyzed the 2018/2019 Los Angeles County homeless youth demographic surveys. We addressed five summary health outcomes: physical health, mental health, substance use disorder, tri-morbidity, and any condition. Respondents were classified into three homeless trajectory groups: (1) short-term-homeless < 1 year in one episode, (2) episodic-homeless < 1 year and multiple episodes, and (3) long-term-homeless continuously for > 1 year. Weighted bivariate and multivariate logistic regression models tested the relationship between homeless trajectory group and health, with controls for sociodemographic factors and structural exposures. RESULTS: Mental health and substance use were high among unsheltered young adults compared to national rates. Long-term homeless respondents were significantly more likely than short-term to report a mental health condition (53.3% vs. 39.8%, p < .001), substance use disorder (25.5% vs. 18.3%, p < .001), and physical conditions (26.0% vs. 15.6%, p = .008). Episodic respondents were more likely to report a mental health condition (50.5%, p < .001). In multivariate models, long-term respondents had twice the odds of tri-morbidity (odds ratio [OR] = 2.14, p < .05) and any health condition (OR 2.00, p < .01) as short term and significantly higher odds of a physical health condition (OR = 1.64, p < .05). DISCUSSION: Youth with longer durations and more frequent episodes of homelessness have substantially poorer health outcomes. The association of longer duration to poorer health persisted in multivariate models. Longer duration of unsheltered homelessness may drive the onset of physical and mental health problems.