The association of psychiatric diagnosis with weather conditions in a large urban homeless sample.

Assessment of psychiatric disorders encounters unique complexities in homeless populations. Although the use of structured diagnostic instruments has significantly improved research methodology in this area, questions remain about the validity of using cross-sectional diagnostic methods derived from studies of more general populations. In particular, the validity of structured diagnostic instruments in the assessment of schizophrenia, depression, drug use disorder, and antisocial personality disorder (ASPD) in homeless populations has been questioned. The purpose of this study was to examine the association of psychiatric diagnoses with the weather. It was hypothesized that self-report of psychiatric illness may be affected by prevailing weather conditions. Nine hundred homeless subjects randomly sampled from St. Louis shelters, day centers, and unsheltered locations were interviewed over a 1-year period. Official average daily temperature and amount of precipitation on the day of each subject's interview were compared with lifetime and current psychiatric diagnoses ascertained by the Diagnostic Interview Schedule. Similar analyses were performed in general population data from the Epidemiologic Catchment Area study. The study found that among homeless men, inclement weather on the day of interview was associated with lifetime and current diagnoses of major depression, lifetime drug use disorder, lifetime diagnosis of ASPD, and current alcohol use disorder. These findings, however, were not present in homeless women and not reflected in the general population. The results, although limited, suggest that weather may confound cross-sectional, standardized methods of psychiatric diagnosis in homeless men. Weather-related factors among homeless men are associated with ascertainment of both lifetime and current diagnosis of major depression, as well as lifetime drug use disorder and ASPD and current alcohol use disorder. Possible interpretations of these findings are discussed, with implications for intervention strategies for psychiatric disorders in the larger context of homelessness and social problems.