Racial/ethnic and age differences in crack use within neighborhoods

Summary overviews of crack use in the U.S. indicate that the lifetime prevalence of crack use for African-Americans is 2-3 times that of White-Americans. However, a recent study that used interview data from the 1988 National Household Survey of Drug Abuse (NHSDA) found no racial/ethnic differences, once neighborhood was held constant. In light of an observed increase in crack use among African-Americans since 1988, we set out to determine whether differences existed between race/ethnicities, using data from the 1990 NHSDA. Employing a post-stratification procedure, we found no overall differences in lifetime prevalence of crack use for African-Americans relative to White-Americans (RO = 1.37, 95% CI = 0.82-2.29) but found that Hispanic-Americans had lower odds than whites (RO = 0.52, 95% CI = 0.28 - 0.96). However, stratification by age group indicated that African-Americans 30-34 years old had significantly higher odds of lifetime crack use than White-Americans in the same age strata when neighborhood is held constant (RO = 2.51, 95% CI = 1.10 - 5.74). Crack use in the year prior to interview was common among African-Americans in this age group who had ever used crack in their lifetime (63% whereas it was unlikely among similarly aged White- or Hispanic-Americans with a lifetime history of crack use (14% and 0% respectively). Generalized Additive Models (GAMs) were used to examine trends in the age-specific prevalence of lifetime crack use between 1988 and 1990 by race/ethnicity. The overall increase in lifetime crack use among African- Americans is explained by an increase in prevalence for African-Americans who are approximately 30 years old-from 5% in 1988 to 8% in 1990. No change was observed for African-Americans of other ages. These findings point to the importance of studying factors specific to older African-Americans (i.e., 30-34 years old) that might have placed them at greater risk than similarly aged individuals of other racial/ethnic groups.