Our findings from this first population-based investigation of the effect of IPV on HIV among husband-wife dyads indicate that abusive husbands increase their wives’ HIV risk in the Indian context via two distinct, yet convergent, pathways, specifically abusive husbands’ heightened risk of HIV infection and heightened risk of infection transmission in the presence of IPV. Compared with non-abusive husbands, abusive husbands demonstrated almost twice the odds of acquiring HIV outside their marital relationship. This finding supports concerns that abusive men are more likely to acquire HIV and subsequently introduce infection into their marital relationships; thus, IPV may be considered a risk marker for wives’ HIV infection via abusive husband’s greater odds of HIV infection. Moreover, compared to wives’ exposure to their husbands’ HIV infection in the absence of violence, wives’ exposure to both IPV and husbands’ HIV heightened their odds of contracting HIV sevenfold. This evidence supports the hypothesis that IPV may facilitate HIV transmission from an infected partner, rendering IPV a risk factor (i.e., direct mechanism) for women’s HIV infection. Taken together, these results support the hypothesis that abused women’s elevated HIV prevalence reflects their being subject to “double jeopardy”, i.e., abused women are more likely to have an HIV-infected male partner, with whom HIV transmission may be enhanced based on abuse in the relationship.
These findings advance previous efforts devoted to understanding both elements of abused women’s double jeopardy for HIV infection, with prior work informing the mechanisms likely responsible. Current evidence that abusive men are more likely to acquire HIV outside of the marital relationship (i.e., IPV as a risk marker for HIV) advances prior investigations across diverse settings demonstrating elevated sexual risk behavior and self-reported STI among abusive men
12, 13, 16, 24 by confirming these findings via utilization of laboratory testing for HIV among a population-based sample. Analyses restricted to the sample of husbands to those whose wives were not HIV infected afforded greater empirical clarity than previously possible in testing hypotheses that abusive husbands are at greater risk for acquiring HIV outside the marital relationship and subsequently introducing it to their wives. Both men’s IPV perpetration and their sexual risk behavior are increasingly considered to stem from a common source, specifically socially sanctioned norms of masculinity that prioritize sexual entitlement and multiple partnering, and physical and sexual domination of female partners.
15, 33,34 As such, the India National Aids Control Organization (NACO) has recognized social norms endorsing men’s sexual entitlement and power over women as a factor in women’s HIV vulnerability.
35 Supporting the import of modifying such factors, recent efforts targeting the intersection of men’s sexual risk and abusive behavior have been demonstrated effective in reducing both men’s violence perpetration and incident STI in the African context;
36 similar efforts underway in India should be prioritized as they hold promise in modifying masculinity norms to reduce both IPV and HIV/STI.
34Current evidence of facilitated HIV transmission to female partners in the presence of IPV (i.e., IPV as a risk factor), similarly advances the current state of knowledge. Use of matched husband-wife dyads with integrated HIV test results allowed assessment of the differential impact of wives’ exposure to their husbands’ HIV across violent and non-violent relationships, with women whose husbands were violent suffering approximately seven times the risk of becoming infected with HIV based on exposure to their husband’s infection. Prior work suggests a number of mechanisms for women’s greater likelihood of HIV infection based on exposure to husband HIV in the presence of violence. The role of unwanted sex and subsequent physical trauma (i.e., tearing or lacerations)
23, 29 associated with IPV is likely critical in increasing opportunity for infection transmission. Cultural norms dictate low levels of sexual communication between spouses concerning sexuality among this population generally;
27 abused women’s limited ability to negotiate or refuse sex in the face of violence, coupled with unwanted sex obtained via physical force and a range of coercive tactics,
13,26, 27,35, 37 likely facilitate trauma and HIV transmission. Qualitatively riskier sexual practices, e.g., anal sex, found more common among male IPV perpetrators
38 may also constitute an enhanced transmission mechanism within abusive relationships. While condom use was relatively rare within the current sample of married couples, evidence from India of violence in response to condom requests by wives
26, 27 suggests that abused women’s limited ability to negotiate condom use (e.g., in the context of wives’ knowledge of his HIV infection or suspicion of extramarital sexual risk behavior) may increase risk of HIV transmission, even within this low-use setting. Further investigations using husband-wife dyads as the unit of analysis across other national and high-risk contexts are recommended to advance the empirical basis for these hypothesized mechanisms and clarify the present findings.
Limitations of the current study include the inability to establish a temporal relationship based on the cross-sectional nature of the study, thus incident and secondary infections among concordant husband-wife dyads cannot be disentangled. While the analyses were based on a conceptual framework positing that husbands may be more likely to introduce HIV to the relationship, it is not possible to rule out the possibility that wives introduced HIV infection within concordant couples, although this appears unlikely based on the current findings that wives’ sexual risk was not associated with their HIV infection. Low levels of prior HIV testing in this sample
4 suggests that the majority of participants were unaware of their HIV status at the time of testing, limiting concerns that IPV could have been caused by HIV disclosure.
39, 40 Although procedures entailing multiple testing and external quality control enhanced the precision of HIV assessment, misclassification is possible and the low prevalence of HIV renders analyses sensitive to potential misclassification. While potential misclassification could limit the precision of estimates, misclassification is not likely to be differential relative to IPV, thus minimizing bias. The IPV assessment was dichotomized for ease in interpretation; further investigation is needed to evaluate the extent to which patterns identified may vary across severity levels and types (i.e., physical and sexual) violence. The self-reported nature of sexual risk behaviors by both male and female participants renders these measures imprecise.
The findings from this population-based study of IPV and HIV in India bolster the hypothesis that abused women face “double jeopardy,” i.e., compounded risk for HIV infection based on both abusive husbands’ greater likelihood of HIV infection and facilitated HIV transmission within abusive relationships. Patterns identified likely extend to other STIs that, while treatable, demonstrate higher population prevalence within the Indian context and elsewhere.
25, 41, 42 The current evidence that IPV serves both as a risk marker for greater likelihood of HIV infection and as a direct HIV transmission mechanism serves to echo calls for simultaneous modification of men’s sexual risk behavior and reduction of violence perpetration against female partners both within South Asia and elsewhere.
12, 15 As IPV may function to facilitate HIV transmission, the reduction of men’s sexual risk in the absence of reducing their abuse of female partners may fall critically short of stemming the secondary transmission of HIV and other STIs. Given evidence that over one in three women face abuse at the hands of their husbands both in the current sample and worldwide,
43 the need for prevention to stem the interwoven threats of IPV and HIV to women’s health and well-being cannot be overstated.
The publisher's final edited version of this article is available at 