Abstract: |
A longitudinal study of patterns of sexual behavior among asymptomatic, homosexual males in New York City was conducted. Participants were interviewed at two time points, 6 months apart. Based on their reports of sexual behavior during a recent 'typical' month, respondents were classified at each time point as engaging in safer (or low-risk) sexual practices versus high-risk sexual behaviors. Discriminant analysis was employed to distinguish the 53 males classified as risky at both time 1 and time 2 from the 47 males classified as safer in both periods. Statistically significant discrimination was achieved with 10 psychosocial predictor variables, four of which were significant while controlling for all other variables in the model. Of these predictors, drug use within sexual contexts was particularly noteworthy, since it provided the greatest relative contribution to the discriminant function and appears to be an important candidate for educational intervention. Other significant variables included perceived adequacy of emotional support, number of years engaged in regular sexual intercourse with other males, and perceived difficulty in modifying sexual behavior. Self-esteem and alcohol consumption approached significance. © 1989. |