Table 5.
Demographic differences in frequency of recent drug use before sex.
| Predictor | B (95% CI) | SE | aB (95% CI) | SE |
|---|---|---|---|---|
| Age | .01 (−.01, .04) | .01 | .02 (−.01, .04) | .01 |
| Bisexual | .08 (−.14, .29) | .11 | .06 (−.16, .27) | .11 |
| Race/ethnicity | ||||
| Black | .16 (−.02, .34) | .09 | .14 (−.04, .33) | .09 |
| Latino | .21 (.05, .38) | .09 | .20 (.02, .38) | .09 |
| Other | .04 (−.22, .29) | .13 | .02 (−.24, .28) | .13 |
| College | .03 (−.11, .17) | .07 | .04 (−.11, .19) | .08 |
| City | ||||
| Chicago | −.20 (−.35, −.04) | .08 | −.17 (−.33, −.01) | .08 |
| Atlanta | −.05 (−.22, .12) | .09 | −.02 (−.20, .17) | .09 |
Notes. B = unstandardized coefficient; aB = adjusted unstandardized coefficient (i.e., adjusted for all other predictor variables); CI = confidence interval; 95% CIs that do not include 0 are significant at p < .05 and indicated with bold font; reference groups: gay (for sexual orientation), White (for race/ethnicity), less than college (for education), and NYC (for city). In supplemental analyses (not presented), we changed the reference group for city in order to compare men in Chicago to men in Atlanta, but none of the additional between-city comparisons were significant.