Ethnic/racial differences in psychiatric disorders: a test of four hypotheses.

OBJECTIVES Ethnic/racial differences in the prevalence of psychiatric disorders have been reported in hospital studies conducted in the United States, but general population studies find little or no evidence of higher rates of disorder in Black communities than White communities. Thus, ethnic/racial differences in the prevalence of psychiatric disorders in treated samples may be due to biased estimates. Aside from sampling bias, 4 hypothetical models of ethnic/racial differences may account for such bias: sociocultural, self-selection, social selection, and clinician bias. The present article reviews and performs limited tests of these 4 models. DESIGN Paranoia is considered a key symptom in biases associated with ethnic/racial differences in psychiatric disorders. Black, Latino, and White community and patient samples are compared, via secondary analysis of data from an epidemiologic study, in terms of their expression of a range of paranoid symptoms. PARTICIPANTS The study sample consisted of 404 community residents, 96 patients with major depression, and 65 patients with nonorganic, nonaffective psychotic disorder (mainly DSM III schizophrenia). MAIN OUTCOME MEASURES Hypothesis-testing involved 2 approaches: 1) standard significance tests of mean differences between groups; and 2) an analysis of the patterns of mean scores across the paranoia scales from the Psychiatric Epidemiology Research Interview by ethnic/racial group and diagnostic category. RESULTS The results of this study favor the sociocultural explanation of bias associated with ethnic/racial differences in psychiatric disorders. CONCLUSION To reduce disparities in ethnic/racial patterns of psychiatric hospitalizations, the biases associated with differences in the sociocultural backgrounds of providers and patients need to be addressed.