"Adapted from Holzer, Swanson, and Shea (1995).
Hispanics. The corresponding rates for a 1-year prevalence period were 19%, 26%, and 20%, respectively. In both instances the rates are significantly higher for Black than for White respondents. More detailed analyses of lifetime and 6-month prevalence for Mexican Americans have been presented by the Los Angeles ECA group (Burnam et al., 1987; Karno et al., 1987).
In subsequent ECA analyses, Holzer, Swanson, and Shea (1995) presented ethnic comparisons for a 6-month prevalence period, as shown in Table I. These crude rates, weighted to population composition, are higher for Blacks than Whites in every group and higher than Hispanics in all but alcohol abuse or dependence and major depressive episode.
In order to test the ethnic differences, Holzer et al. (1995) created logistic regression models for any disorder that included age, sex, race, and socioeconomic status, as well as their two-way interactions. In the first model using socioeconomic status as a categorical variable, the main effect for race was not significant, although there were significant age-by-race and race-by-socioeconomic status interactions. In a second model, which used socioeconomic status as a linear variable, race and the interactions remained significant, indicating that the ethnic differences are not cancelled by simple demographic controls.
Was this article helpful?