Palermo et al. respond to "Disclosure of gender-based violence".

In their commentary (1) on our article (2) examining reporting among survivors of gender-based violence, Beydoun and Beydoun provide an excellent overview of the current state of research on prevalence, disclosure, and health consequences of gender-based violence. We enthusiastically join their call for primary and secondary prevention efforts aimed at ending gender-based violence. However, to avoid oversimplifying a complex problem, further discussion is warranted surrounding their conclusion that men and women tend to be equally aggressive in intimate relationships. Specifically, there is little research to support or refute gender symmetry of intimate partner violence (IPV) in developing countries, because most research comes from developed countries (3–12). In addition, we highlight their important caveat that evidence globally suggests that the severity and consequences of IPV are not gender symmetrical. Despite the dual and complex nature of perpetration and victimization among intimate partners, the majority of largescale surveys, particularly from developing countries, collect victimization information fromwomen only (13, 14), or more recently, perpetration information from men only (15, 16). In one of the few studies to examine gender symmetry of IPV in a resource-poor setting, Kishor and Bradley (17) used data from nationally representative Demographic and Health Surveys in Ghana and Uganda. Prevalence rates of lifetime physical IPV were higher among women than men (19% vs. 10% in Ghana and 47% vs. 19% in Uganda), whereas lifetime rates of perpetration were lower among women than men (7% vs. 16% in Ghana and 6% vs. 41% in Uganda). Furthermore, violence perpetrated by men was more common, more severe, and more likely to result in injury (17). However, in a longitudinal study from the Cebu province in the Philippines, findings indicated higher rates of female-perpetrated versus male-perpetrated physical violence in the past 12 months (56% vs. 25%), yet comparable rates of victimization (28% among women vs. 31% among men); thus, the authors hypothesized that self-defense may have played a role in perpetration asymmetries (18). Reporting bias may also play a role; a meta-analysis indicated that individuals tend to report less perpetration than their partners would attribute to them, and this bias was larger for men (19). Lack of clear consensus on dynamics is also caused by data collection challenges; ethical standards recommend against conducting interviews with the man and woman in the same partnership to ensure the safety of the victim (20, 21). Therefore, we often lack generalizable evidence from both men and women experiencing the same dynamics and instead rely on separate samples drawn from the same population or triangulation through other data. Given the lack of consensus regarding the magnitude of gender asymmetries in IPV perpetration, we would like to reiterate the point made by Beydoun and Beydoun (1) regarding the severity and consequences of IPV. Studies consistently show that women are more likely than men to experience IPV-related injury (3, 4, 6–8, 17). Particularly telling is the study by Stöckl et al. (22), which estimated that the global proportion of female homicides committed by an intimate partner is 6 times higher than that of male homicides. Moreover, women are more likely than men to experience sexual IPV (6, 8, 17), putting women at risk of adverse reproductive health outcomes, including gynecological symptoms, unwanted pregnancy, pregnancy loss, poor birth outcomes, and decreased access to prenatal care (23–29). Taken together, the body of evidence described above and by Beydoun and Beydoun (1) illustrates the need for more investigation on whether the gender symmetry of IPV perpetration found in developed settings is in fact mirrored in resource-poor settings. More importantly, caution is needed when interpreting symmetry of IPV perpetration findings for programming and policy purposes, because the severity and consequences of IPV are clearly not gender symmetrical.