A prospective study of substance use and housing stability in a homeless population

This study examined self-report and urine test data about homeless substance use over time, prospectively comparing substance use with attainment of stable housing.

400 homeless people systematically sampled from shelters and streets in St. Louis, Missouri were assessed with structured diagnostic interviews and urine substance testing annually over 2 years. Nearly two-thirds (n = 255) completed all three assessments, constituting the sample for this prospective study.

More than half (55%) of this homeless sample had detectable cocaine use during the study. Most cocaine users continued using during the next 2 years and failed to achieve and maintain stable housing. Cocaine use in the first follow-up year predicted housing patterns over the next 2 years, independent of lifetime diagnosis of cocaine use disorder. Alcohol abuse/dependence in the 2-year follow-up period did not predict housing outcomes.

The course of cocaine use and abuse/dependence, but not continuing alcohol addiction, was associated with subsequent attainment of stable housing, especially cocaine use in the first prospective year. Replication of these findings in other locations to determine generalizability may have implications for designing housing service models.