Background: Few longitudinal studies have examined predictors of dating violence, a public health issue that may be more prevalent among African Americans. Our objective was to examine the prevalence, incidence, and predictors of dating violence in an African American sample using the theory of gender and power. Methods: A longitudinal design with a 1-year follow-up period was used. Recruiters screened adolescents from a variety of venues, including school health classes, county health department clinics, and adolescent health clinics. Study participants were 522 African American females 14-18 years of age. This study achieved an 85.7% participation rate. Dating violence was defined as ever having experienced verbal or physical abuse perpetrated by a boyfriend. Results: At baseline, 28% of adolescents reported a history of dating violence. To calculate the 1-year incidence of dating violence, adolescents reporting a history of dating violence at baseline were excluded from subsequent longitudinal analyses. In longitudinal analyses, the 1-year incidence of dating violence was 12%. Logistic regression analyses identified four factors at baseline that were predictive of dating violence. This four-factor model correctly classified 87.6% of adolescents according to whether or not they experienced dating violence during the 1-year follow-up. Controlling for public financial assistance, relative to adolescents not experiencing dating violence, those who did were 2.0 times more likely to report less understanding of healthy relationships (CI = 1.1-3.9, p = 0.02), 2.0 times more likely to report using drugs (CI = 1.0-3.7, p = 0.02), and 1.9 times more likely to have viewed X-rated movies (CI = 1.0-3.7, p = 0.03). Conclusions: These findings suggest that secondary prevention of dating violence necessitates educating clinicians on the importance of screening and training practitioners in clinical settings on how to effectively screen adolescents for dating violence, including risk factors for victimization.
ASJC Scopus subject areas