Probing the meaning of racial/ethnic group comparisons in crack cocaine smoking

JAMA. 1993 Feb 24;269(8):993-7.


Objective: To probe the meaning of reported racial and ethnic group differences in the prevalence of crack cocaine smoking and to estimate the degree to which crack cocaine smoking is associated with personal factors specific to race/ethnicity.

Design: Through reanalysis of data from the 1988 National Household Survey of Drug Abuse (NHSDA), we compared racial/ethnic group differences in crack cocaine smoking. To hold constant social and environmental risk factors that might potentially confound racial comparisons, we used an epidemiologic strategy that involves poststratification of respondents into neighborhood risk sets. A conditional logistic regression model was used to estimate the relative odds of crack use by race/ethnicity.

Patients or other participants: The 1988 NHSDA interviewed 8814 individuals residing within households in the United States. Subjects were selected using a multistage area probability sampling of all residents aged 12 years and older.

Results: Once respondents were grouped into neighborhood clusters, the relative odds (RO) of crack use did not differ significantly for African Americans (RO, 0.85; 95% confidence interval [Cl], 0.37 to 1.93) or for Hispanic Americans (RO, 0.88; 95% Cl, 0.47 to 1.67) compared with white Americans.

Conclusion: Findings of race-associated differences are often presented as if a person's race has intrinsic explanatory power. This analysis provides evidence that, given similar social and environmental conditions, crack use does not strongly depend on race-specific (eg, biologic) personal factors. Although the study finding does not refute the previous analysis, it provides evidence that prevalence estimates unadjusted for social environmental risk factors may lead to misunderstanding about the role of race or ethnicity in the epidemiology of crack use. Future research should seek to identify which characteristics of the neighborhood social environment are important and potentially modifiable determinants of drug use.

Publication types

  • Comparative Study
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Adolescent
  • Adult
  • Child
  • Crack Cocaine*
  • Ethnicity*
  • Female
  • Humans
  • Logistic Models
  • Male
  • Population Surveillance
  • Racial Groups*
  • Socioeconomic Factors
  • Substance-Related Disorders / epidemiology*
  • Substance-Related Disorders / ethnology
  • United States


  • Crack Cocaine