Drug use and unprotected anal intercourse among gay men.

Conducted a longitudinal study of 604 New York City gay men to examine the link between unprotected insertive and receptive anal intercourse and drug use with sex. Data spanning four 12-month periods from 1980 to 1987 indicated that the link between drug use and high-risk sex diminished as the acquired immunodeficiency syndrome (AIDS) epidemic progressed. Strong and significant cross-sectional associations (average r = .30, p less than .0001) were found between drug use with sex and both types of anal intercourse in the year prior to the onset of AIDS. By 1987 the magnitude of these associations was markedly reduced (average r = .09, p less than .05). Longitudinal results indicate that, although cessation of drug use with sex is associated with subsequent lower rates of unprotected anal intercourse, initiation of drug use with sex is not associated with subsequent increases in this sexual activity. These findings support both causal and noncausal interpretations of the link between substance use and high-risk sex among gay men.