LAUSR

Background: Much knowledge has been accumulated on individual-level risks and protective factors of violence against women. However, the influence of factors operating at the community level, such as community cohesion, remains unclear, especially in low- and middle-income countries. This study examined whether community cohesion, a combined measure of mutual trust and tolerance, affects women’s likelihood of experiencing intimate partner violence, violence perpetrated by a family member, and violence occurring in public spaces. Methods: Data on 4785 women aged between 18 and 60 years in Ghana, Pakistan, and South Sudan, initially collected by the ‘What Works to Prevent Violence Against Women and Girls’ Research and Innovation Programme, were used for this study. Binary logistic regression analyses were used to assess the association between a composite measure of community cohesion and recent experience of physical, sexual, and psychological violence while controlling for different individual-, community-, and state-level variables. Results: Multivariate analyses revealed that community cohesion was associated with lower risks of public spaces violence (adjusted odds ratio = 0.396, 95% confidence interval = 0.312–0.503, P < 0.001) and family member violence (adjusted odds ratio = 0.839, 95% confidence interval = 0.754–0.934, P < 0.001). There was no statistically significant association between intimate partner violence and community cohesion, but adjusted models showed that women with more developed social networks were at higher risks of experiencing intimate partner violence (adjusted odds ratio = 1.104, 95% confidence interval = 1.062–1.148, P < 0.001). Conclusion: Our findings suggest that community cohesion may have a protective effect against the most visible forms of violence against women. However, this effect may be attenuated or even eliminated by other individual- or household-level mechanisms in the case of violence between intimates or family members.