Predictors of risky sexual behavior in university students and injection drug users: A test of the Health Belief Model

Date of Award




Degree Name

Doctor of Philosophy (Ph.D.)

First Committee Member

Michael H. Antoni, Committee Chair


The present study was an investigation of the efficacy of the Health Belief Model (HBM) to predict condom usage and risky sexual practices in university students and injection drug users. One hundred twenty-two university students (58 men, 64 women) and 123 injection drug users (62 men, 61 women) were recruited for this investigation. Data were analysed separately for men and women in each population. Substantial proportions of individuals across all four groups reportedly engaged in the following risky sexual practices: multiple sexual partnerships, unprotected oral sex, sex with a member of an AIDS risk group, unprotected vaginal intercourse, and sex while under the influence of alcohol or drugs. Sixty percent of university men, 43% of university women, 21% of injection drug using men, and 32% of injection drug using women reportedly "always or almost always" used condoms during the previous six months. Remaining participants varied in the extent to which condom use was reported. In each of the four populations, HBM components were regressed on condom usage and each of the risk behaviors, controlling for demographic variables and recreational drug use. HBM components did not predict condom usage in university men, university women, or injection drug using women. However, in injection drug using men the model significantly explained 25% of the variance in condom usage. There was a trend for model components to predict number of sexual partners in university men. Among university women, the HBM explained 22% of the variance in number of sexual partners and accounted for 18% of the variance in total number of risk behaviors endorsed. The model did not significantly predict risky sexual practices in either male or female injection drug users. Methodological issues in the assessment of HBM constructs were examined. Alternative models and additional variables which might explain AIDS-risk behaviors in these populations were explored, and suggestions for future studies were put forth. Based upon the results of the current inquiry, the HBM may have limited utility for predicting AIDS-risk reduction strategies in these populations.


Psychology, Social; Health Sciences, Public Health

Link to Full Text


Link to Full Text