Mental illness in a representative sample of homeless men in Munich, Germany.

The aim of the study was to reliably assess the 6-month and lifetime prevalence of mental illness according to DSM-III criteria in a representative sample of homeless individuals in the city of Munich, Germany. Because the characteristics of the homeless population were unknown, we first conducted a pre-sampling survey to determine a proportionate allocation of the main interviews in three nested sampling strata. For the pre-sampling we approached 300 males, who appeared to be homeless, for a brief interview; of these, 271 were homeless according to our definition and were allocated to one of three sectors ("shelter", "meal services", "outdoor"). Thereafter, we randomly sampled homeless males in these three strata until the indicated allocations were met. The Diagnostic Interview Schedule (DIS) was used for diagnostic classification according to DSM-III in the main interview. Results from this representative urban sample show that the mean age of the homeless males was 43 years; most were unmarried or divorced, had a relatively low level of school education and a long duration of homelessness. Based on the main interviews with 146 homeless males the following lifetime prevalence rates were obtained: 91.8% for substance use disorder (82.9% alcohol dependence), 41.8% for affective disorders, 22.6% for anxiety disorders and 12.4% for schizophrenia. Of the homeless males in Munich, 94.5% had at least one DIS/DSM-III axis I diagnosis. Six-month prevalence data is also presented. Results are compared with those of a very similar study on homeless individuals in Los Angeles, which also used DIS/DSM-III diagnoses. In comparison with representative community samples in the United States and in Germany, mental illness was much more frequent among homeless individuals in Munich as well as in Los Angeles. Implications for health care planning are discussed.