To the Editor.
—The recent article by Lillie-Blanton et al1 and the accompanying Editorial2 underscore the importance of selecting appropriate statistical methods when interpreting prevalence rates showing racial and ethnic differences. Their conclusions are consistent with previous research findings in which adjustments for sociodemographic variables reduced or eliminated racial differences in the symptoms of schizophrenia and depression in personality tests and in intelligence tests.3They also support recent findings of the Epidemiologic Catchment Area study, which, after corrections for age, sex, socioeconomic status, and marital status, found few and unimpressive race-related differences for schizophrenia, affective disorder, obsessive-compulsive disorder, antisocial personality disorder, and drug abuse.4Some of these results have important implications, which may be quite different from those of the raw data. For example, lack of racial differences in schizophrenia and affective disorder is in sharp contrast to unadjusted hospital and clinic data showing racial disparities
Adebimpe VR. Race and Crack Cocaine. JAMA. 1993;270(1):45. doi:10.1001/jama.1993.03510010051021