Intimate Partner Violence and Population Mental Health: Why Poverty and Gender Inequities Matter

Steven Pinker [1] suggests that the world is becoming less violent. Yet, although it has been nearly two decades since the United Nations issued its Declaration on the Elimination of Violence against Women, the global burden of violence against women still remains alarmingly high [2]. Violence commonly results in physical injury, and other serious physical or psychological sequelae may also result, compounding its contribution to the overall global burden of disease [3]. The systematic review and meta-analysis by Karen Devries and colleagues [4] in this week’s PLOS Medicine brings us one step closer toward understanding the mental health liabilities associated with violence against women. The public health impact of their contribution is substantial, given the lack of resources devoted to the prevention and treatment of mental disorders worldwide [5,6,7]. Their article is not the first to summarize the impacts of various forms of violence and abuse against women on mental health [8,9,10] (Table 1). Karen Devries and colleagues [4] contribute to this body of literature reviews by restricting their focus to longitudinal studies, thereby examining the temporal relationship between intimate partner violence and depression. Although the literature contained few studies examining the impacts of violence on incident depressive disorders, the authors identified enough studies to support a conclusion that violence was associated with incident symptoms of depression. They had intended to also examine the association between victimization and depression among men, but lack of data limited their ability to draw firm conclusions. Notably, Karen Devries and colleagues [4] also found that, among women, symptoms of depression were associated with incident experiences of violence. Whether the relationship between violence and depression is truly causal in both directions is unknown. For example, in examining the association between exposure to intimate partner violence during adulthood and incident symptoms of depression, an unmeasured confounder preceding both variables such as childhood abuse—not measured in most of the studies—could be causally related to both irrespective of their temporal relationship with each other, inducing a spurious association that could be mistakenly interpreted as bidirectional. Statistical adjustment for baseline depression symptom severity, which was done in most of the studies, could potentially help with this problem if it is a key pathway through which the confounding effects of childhood abuse are transmitted. So what do these findings suggest for clinical practice, programming, or policies to prevent violence against women? If depression is causally related to violence, one might consider evidence-based collaborative care management of depression [11,12] in order to prevent subsequent experiences of violence. If violence is causally related to depression, one might suggest counseling interventions for women with histories of partner abuse in order to prevent subsequent episodes of depression, but high-quality randomized controlled trials have shown mixed results [13,14,15,16]. The only randomized trial of universal screening for intimate partner violence yielded null findings [17]. It is possible that screening may exert substantive benefits only in settings where providers also have the ability to refer their clients to a broad array of services (e.g., case management, crisis services, legal advocacy, emergency shelters, transitional housing, and/or parenting and childcare support) [18], but the effectiveness of such a multipronged approach is as of yet unknown. This particular gap in the literature redirects our attention to more systemic determinants of the excess burden of violence and depression among women. The simplified conceptual model depicted in Figure 1, informed in part by previously published work [19,20], suggests not only that violence against women is a consequence of gendered norms about the use violence against intimate partners and power relations [21,22], but also that its