Since early 1983, the incidence and prevalence of heterosexually transmitted AIDS among women have increased at an alarming rate. However, due to the conceptualization of AIDS as a "gay male" disease, little research attention has been devoted to the prevention of HIV infection among women. The purpose of the current intervention was to test the utility of extending a behavioral social influence/diffusion of innovation approach to a group of heterosexual women. A randomized experimental field design was used to compare a community intervention (AIDS education materials plus the training of individuals identified as key opinion leaders to serve as peer behavior change agents) with a comparison intervention (AIDS education materials alone). The study was conducted at a small liberal arts college for women.

Two dormitories were randomly assigned to either an intervention or comparison condition. Twenty-four women, living in the intervention dormitory, were identified as key opinion leaders among their female peers. These key opinion leaders then received information concerning the basic epidemiology of AIDS and other STDs, misconceptions about the transmission of AIDS/STDs among heterosexual women, gender constraints that impact health behavior change among women, and practical risk reduction strategies. Opinion leaders also received training in specific conversation skills to endorse HIV-protective behavior and to convey a change in normative sexual behavior to women living in the intervention dormitory.

At pre- and post-intervention, 580 surveys were collected from both the intervention and comparison dormitories. Using an anonymous identification code, survey data were matched for 192 comparison and intervention participants. The major dependent variables included (a) AIDS/STD risk behavior knowledge, (b) perceived risk, (c) perception of peer norms for HIV-risky and HIV-protective behaviors, (d) stage of health behavior change, (e) intentions to practice safer sex, (f) socially and sexually assertive behavior, (g) HIV-risky sexual behavior, and (h) alcohol and drug use. Condom-taking behavior provided a nonreactive measure of behavioral intentions.

A number of direct training effects were found for the key opinion leaders, including an increase in AIDS/STD risk behavior knowledge, conversation skills, and empathic assertion. However, there was no change in behavioral intentions, stage of change, perceptions of peer norms for HIV-protective behavior, or reported HIV-risky behavior.

Analysis of covariance on posttest scores, using pretest scores as covariates, showed that, relative to the comparison participants, the intervention participants increased their AIDS/STD risk behavior knowledge and the number of AIDS/STD related peer conversations. No other treatment effects due to diffusion of innovation were found. The rate of risky sexual behavior for the intervention period was relatively low. However, descriptive statistics revealed a pattern of HIV-risky behavior in the current sample suggesting that HIV/STD risk may increase over time as a function of unprotected vaginal and oral intercourse and serial monogamy. The implication of these findings are discussed in terms of future interventions targeting heterosexual women.