Relationship Factors and Condom Use Among Abused Women

Introduction: Women account for 23% of the HIV population and 20% of new infections in the U. S. Intimate partner violence (IPV) increases women’s risk for HIV infection by nearly 22%. Urban women of color are at high risk for both IPV and HIV infection. HIV transmission occurs through multiple pathways, including inconsistent condom use resulting from difficulties negotiating safer sex practices. To achieve health equity in HIV/AIDS prevention, we must investigate and eliminate factors that lead to health inequalities among at-risk populations. To understand factors that may impact condom use among abused women, we explored the associations among relationship factors [relationship power, fear of abuse when negotiating condoms (FOA), and partner dependence], Information-Motivation-Behavioral Skills (IMB) model constructs (information, motivation, and behavioral skills), and condom use. The IMB model has been widely used in HIV interventions.

Methods: Participants (n = 177; M age = 35 years, SD = 10.2; 45% African-American, 36% White) were recruited from community-based agencies to a clinical trial designed to evaluate an integrated IPV-HIV risk reduction intervention for abused women. Predictor variables consisted of the IMB model constructs.  Moderators were relationship power, FOA, and dependence. The criterion variable was condom use (i.e., proportion of episodes of unprotected sex in the last 3 months). Aggregated Pearson scaled logistic regression analyses with interaction terms in SAS were used.

Results: Participants reported high levels of IPV on Abusive Behavior Inventory (M = 54.6; SD = 29.3). As expected, all three relationship factors were associated with condom use: relationship power (β = -0.89, SE = 0.30, p = 0.003), FOA (β = -0.08, SE = 0.02, p < 0.001), and partner dependence (β = -0.49, SE = 0.11, p < 0.01). Results from aggregated Pearson scaled logistic regressions with interaction terms found that FOA (β = 0.37, SE = 0.08, p < 0.0001; β = 0.07, SE = 0.02, p < 0.0001; and β = 0.03, SE = 0.01, p < 0.0001) and partner dependence (β = 2.84, SE = 0.66, p < 0.0001; β = 0.48, SE = 0.17, p = 0.005; and β = 0.25, SE = 0.05, p < 0.0001) moderated the association between information, motivation, and behavioral skill constructs and condom use, respectively. Contrary to expectations, relationship power did not moderate these relationships.

Conclusions: The results highlight the critical importance of addressing relationship factors as well as violence in HIV prevention programming for women in heterosexual relationships characterized by IPV.