A 20-Year Longitudinal Study of Homelessness Among 333 Psychiatric Patients in Portugal: From Initial Episodes to Chronic Trajectories

Background: Homelessness is closely related to psychiatric disorders, substance abuse, and premature mortality; however, in-depth analyses of long-term trajectories are lacking. Aims: The objective was to analyze psychiatric diagnoses, clinical patterns and survival outcomes of homelessness trajectories during a 20-year follow-up. Method: A retrospective sample of 333 psychiatric patients experiencing homelessness in Portugal. Cox survival analyses defined the cohorts: Recent (⩽1 year; n = 92), Early-stage (1–5 years; n = 52), Prolonged (5–15 years; n = 93), and Persistent (>15 years; n = 96). Analyses included analysis of variance (ANOVA), Chi-square tests, and logistic regression. Results: Mortality was 31.2%, decreasing after 18 years of homelessness, indicating a survivor effect. Recent cohort and younger patients, and those with alcohol misuse, anxiety, and personality disorders had minimal hospitalization. Early-stage and Prolonged groups had the highest mortality (50%), linked to alcohol-related disorders and organic psychosis, mainly in Prolonged. Persistent cohort, with the longest psychiatric follow-up (18.8 years), more hospitalizations, and higher comorbidity with drug use and intellectual disability, had the lowest mortality (14.6%) and the highest mean age at death (67.3 years). Unspecified psychosis, substance use, and intellectual disability were the key predictors of Persistent group. Alcohol use, psychosis, and personality disorders influenced survival varying by cohort. Conclusions: Patient psychiatric homelessness cohorts have heterogeneous trajectories for clinical, sociodemographic, service use, and survival. Early and mid-stages of homelessness show higher mortality. Findings highlight the Persistent cohort as long-term homelessness survivors. It is stressed that the duration of homelessness is a critical determinant of diagnostic patterns, comorbidities, and survival outcomes, which require differentiated psychiatric and social interventions tailored to cohort-specific risks among people experiencing homelessness.