CONTEXT

Intimate partner violence is one dimension of relationships that may influence condom use, yet few studies have examined male‐ and female‐initiated violence in efforts to understand variation in condom use.

METHODS

Power dynamics and relationship conflict approaches were employed to examine the association between relationship violence and condom use. In a latent class analysis of 8,599 dating relationships from Wave 3 (2001–2002) of the National Longitudinal Study of Adolescent to Adult Health, when the respondents were aged 18–25, relationship violence was characterized by the severity and frequency of violence and the perpetrator's gender. Random‐effects logistic regression analyses assessed the association between violence classes and condom use.

RESULTS

One in five young adult dating relationships involved violence in the past year. Four violence classes were identified: one male‐dominant class (3% of relationships), in which many relationships had reciprocal violence; two female‐dominant classes, differentiated by the frequency and severity of violence (2% low‐intensity, 4% medium‐intensity); and one class with limited or no violence (91%). Male‐perpetrated violence was reported less frequently but was more severe than female‐perpetrated violence. Respondents in relationships in the male‐dominant/high‐intensity and the female‐dominant/medium‐intensity classes were less likely to report condom use than those in relationships in the no/low‐violence class (odds ratios, 0.4–0.5).

CONCLUSIONS

This work expands on research focusing only on male‐perpetrated violence and highlights the importance of capturing relationship violence in national data sets. Future surveys that collect information about violence from both partners can inform efforts to prevent violence and to support victims.