• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
Am J Public Health. Author manuscript; available in PMC Dec 6, 2012.
Published in final edited form as:
PMCID: PMC3515795
EMSID: EMS50556

Sexual Violence and Reproductive Health Outcomes Among South African Female Youths: A Contextual Analysis

Abstract

Objectives

We studied whether female youths from communities with higher sexual violence were at greater risk of negative reproductive health outcomes.

Methods

We used data from a 2003 nationally representative household survey of youths aged 15–24 years in South Africa. The key independent variable was whether a woman had ever been threatened or forced to have sex. We aggregated this variable to the community level to determine, with control for individual-level experience with violence, whether the community-level prevalence of violence was associated with HIV status and adolescent pregnancy among female, sexually experienced, never-married youths.

Results

Youths from communities with greater sexual violence were significantly more likely to have experienced an adolescent pregnancy or to be HIV-positive than were youths from communities experiencing lower sexual violence. Youths from communities with greater community-level violence were also less likely to have used a condom at their last sexual encounter. Individual-level violence was only associated with condom nonuse.

Conclusions

Programs to reduce adolescent pregnancies and HIV risk in South Africa and elsewhere in sub-Saharan Africa must address sexual violence as part of effective prevention strategies. (Am J Public Health. 2009;99:S425–S431. doi: 10.2105/AJPH.2008.136606)

Approximately two fifths of new HIV infections in sub-Saharan Africa are among youths aged 15 to 24 years,1 and females within that age range are 3 times as likely to be infected with HIV as are their male peers.2 In 2003, the prevalence of HIV among South African women visiting antenatal clinics was estimated to be 15% among women aged 15 to 19 years and 30% among women aged 20 to 24 years.3 Sexually active youths are also at risk for teenage pregnancy; in much of the developing world, more than one fifth of women of reproductive age have had a first birth by age 18 years.4 In South Africa in 2003, 32% of women aged 15–19 years had ever been pregnant.5 Many teenage pregnancies in sub-Saharan Africa are unintended, either mistimed or unwanted.6

Programs to prevent sexual transmission of HIV often ignore the mediating influence of sexual and physical violence on women’s lives. The threat and experience of sexual and physical violence are particular barriers to young women’s ability to practice safe sexual behaviors (including discussing HIV, remaining abstinent, or using condoms), given power imbalances between young women and their partners.7 For example, a study of South African female youths aged 15–24 years that used the same data we used showed that young women with lower relationship power and young women who experienced forced sex in the past 12 months with their current partner were less likely to use condoms consistently than were women with more power or who had not experienced forced sex.8 No association was found between relationship power or forced sex and HIV status.8 Conversely, a South African study of antenatal clients found that pregnant women who experienced intimate partner violence and reported male control over relationships were significantly more likely to be HIV positive than were pregnant women with no reported intimate partner violence or less reported male control.9 Another study of youths from KwaZulu-Natal showed that 46% of sexually experienced young women reported coerced first sex, including being persuaded, tricked, physically forced, or raped.10 Youths who experienced coerced first sex were significantly more likely to have ever been pregnant, to report their pregnancy as unwanted, and to have experienced a sexually transmitted infection.10 Finally, a study that compared pregnant youths with nonpregnant youths from parts of Cape Town, South Africa, found that unequal power relations within a partnership are reinforced by violence, putting young women at risk for pregnancy and possibly HIV.11

We were interested in studying the sexual behaviors that put young persons aged 15–24 years at risk for unintended pregnancies or sexually transmitted infections, including HIV infection. We examined whether both individual-level and community-level violence were associated with condom use at last sexual encounter, HIV status, and adolescent pregnancy. We hypothesized that youths from communities with higher sexual violence, regardless of whether they had personally experienced such violence, would be less likely to practice safe sex and more likely to be exposed to negative health outcomes because of the threat or fear of violence.

METHODS

We obtained data from a nationally representative household survey of South African youths aged 15–24 years conducted in 2003. The survey used a 3-stage, disproportionate, stratified design based on the 2001 national census to select a representative sample of participants; the census enumeration areas were the primary sampling unit. Households within randomly sampled enumeration areas were enumerated, and one eligible youth per household was randomly selected to take part in the interview. All participants were asked to provide an oral fluid specimen to test for HIV antibodies and to complete a face-to-face interview. Details of the questionnaire content, data collection strategies, and laboratory procedures were reported previously.12

Analysis Sample and Dependent Variables

For this analysis of sexual behaviors and outcomes, we reduced the sample to include only sexually experienced female youths. Given that unmarried, sexually active youths are the most in need of condoms for HIV prevention and that only 3% of this South African female sample was currently in union, we also limited the sample to include only unmarried female youths. The full sample was 6217 female youths (unweighted). Eliminating sexually inexperienced and married youths resulted in a sample size of 3865 female youths (unweighted); this represented 62% of the full sample of female youths surveyed.

Three dependent variables were used for this analysis: condom use at last sexual encounter, adolescent pregnancy experience, and HIV status. Two of these dependent variables were self-reported (condom use and pregnancy experience), whereas HIV status was based on HIV testing, as mentioned earlier. For the analyses of condom use, the sample was limited to youths who had sex in the past year. For the analyses of HIV status, all sexually experienced, unmarried female youths were included, whether they had been sexually active in the past year or not.

Adolescent pregnancy was defined as self-reported pregnancy before age 18 years. The age cutoff of 18 years was selected because the median age of first pregnancy in South Africa is 18 years, and we were interested in examining which youths were having early pregnancies. For the analyses of adolescent pregnancy, the sample was reduced to include only unmarried, sexually experienced female youths who were aged 18 years or older. This reduced the risk of left-censuring youths not old enough to have experienced an adolescent pregnancy.

Independent Variables

The key independent variable of interest was whether the young woman was ever threatened or forced to have sex. This variable was developed from 2 questions on violence asked to all sexually experienced youths: “Have you ever had sexual intercourse because someone threatened you to have sex with him or her?” and “Have you ever had sexual intercourse because someone used physical force to make you have sex with him or her?” All female youths who reported that they were ever threatened or forced to have sex were coded 1; all other respondents were coded 0.

Other covariates in the analyses of youth risk-taking and outcomes included age (continuous variable), residence (urban or rural), race (Black African, colored, White/Indian), and educational attainment (completed high school versus did not complete). These control variables were selected because of prior research from South Africa and elsewhere showing that these are key factors associated with youth risk-taking.12

Community-Level Variables

The definition of the community for this analysis merits mentioning. The lowest geographical level of sampling was the enumeration area from the South African national census; the sample included a total of 656 identifiable enumeration areas. Because in some cases the enumeration area included only a handful of individuals, we decided to group enumeration areas into “municipalities,” a natural grouping used by the South African census.

Youths were excluded if they were from enumeration areas that did not have a municipality code or if their municipality contained fewer than 5 respondents (a total of 261 female youths were dropped, 4% of the full sample). Steps were taken to redistribute the largest municipalities into smaller units. Municipalities containing 100 or more sexually experienced female respondents (the eligible sample) were split into “main places” (falling between the municipality and enumeration area in the census unit hierarchy and for which we have codes that link municipalities to main places), and the sample size of eligible respondents of these main places was verified to fall between 5 and 99. A total of 172 municipalities (main places) were included in the final analysis. The size of the 172 municipalities based on all female youths surveyed (used to calculate the community-level variable for violence) ranged from 6 to 146 respondents. Models were performed removing participants from communities with less than 10 female youths surveyed; results were similar to those presented, and the larger sample was retained to reduce further bias by dropping individuals from the smaller communities.

Community-level variables related to violence, sexual experience, and high school education were calculated by aggregating the weighted data at the level of the municipality (main place) to create a continuous variable that measured the mean in the community. All participating youths from the communities contributed to the community-level variables, whether or not they were in the analysis samples. For multivariate analyses, the community-level variables were centered and standardized by subtracting the overall mean and dividing by the standard deviation (providing z scores). Community-level variables were categorized into high, medium, and low categories to test the linearity assumptions of the models. Each demonstrated a continuous pattern, indicating the appropriateness of using the single centered and standardized mean value in the multivariate models presented.

Community-level violence experience was the key independent variable of interest. Community-level sexual experience was included to control for the fact that some youths came from communities where there was high sexual activity, which was associated with the outcomes of interest. Community-level high school completion was included to provide information on the socioeconomic context within which the participants lived. Unfortunately, no measures of wealth or poverty were available at the individual or community levels, so education was the closest proxy for this.

Analysis Methods

Univariate, bivariate, and multivariate analyses were performed. The multivariate logistic regression analyses that included both individual and community-level variables were performed using Huber–White standard errors in Stata version 9.2 (StataCorp, College Station, Texas). This modeling approach allowed for clustering of individuals within the communities and avoided underestimation of the standard errors.13,14 All analyses were weighted to represent the South African adolescent population. All multivariate models of the reproductive and behavioral outcomes were controlled for age, race, residence, and high school completion as well as the community-level variables for sexual experience and high school completion.

RESULTS

The demographic characteristics of the full sample of female youths surveyed and the reduced analysis samples of unmarried, sexually experienced youths are presented in Table 1. In general, the sexually experienced, unmarried youths were older, more educated, and more likely to be Black than were youths in the full sample.

TABLE 1
Distribution of Female Youths Aged 15 to 24 Years: South African Youth Survey, 2003

Among all sexually experienced female youths, about 11% reported that they had ever been threatened or forced to have sex. When the threat or force variable was aggregated to the community (municipality) level by using all females in the community, we found that the prevalence of threat or force in the communities ranged from 0.0% in some communities to 47.6% in other communities (mean=7.5%; median=6.1%; data not shown).

The prevalence of condom use at last sexual encounter, HIV status, and adolescent pregnancies are presented in Table 2, as are the unadjusted odds ratios (ORs) from logistic regressions examining the association between the individual and community-level violence variables and the outcomes of interest. Among unmarried female youths who had sex in the past 12 months, approximately one half (52%) reported that they had used a condom at their last sexual encounter. Female youths who had ever been threatened or forced were significantly less likely (41%) to have used a condom than were their peers who had never been threatened or forced (53%; unadjusted OR=0.62; 95% confidence interval [CI]=0.44, 0.87). Model 2 presents the unadjusted OR of the association between community-level violence experience and condom use at last sexual encounter. For every 1-point increase in the mean, standardized community-level threat or force variable, young women were 16% less likely to have used a condom at last sexual encounter (OR=0.84; 95% CI=0.75, 0.93). The results remained the same with the inclusion of both the individual and community-level violence measures (model 3).

TABLE 2
Prevalence of Condom Use at Last Sexual Encounter, HIV Positive Status, and Adolescent Pregnancy, by Threat or Force Experience at the Individual and Community Level: South African Youth Survey, 2003

As reported previously, 21% of the female sexually experienced youths tested positive for HIV.8 The percentage of youths who were HIV-positive was similar between sexually active female youths who had ever been threatened or forced to have sex (22%) and their peers who were never threatened or forced (21%); the unadjusted OR was not significant. The unadjusted analysis of community-level violence and HIV status showed that for every 1-point increase in the mean, standardized community-level threat or force variable, young women were 18% more likely to be HIV-positive (OR=1.18; 95% CI=1.03, 1.37); similar results were found in model 3, which included both the individual and community-level variables.

The examination of adolescent pregnancy experience before age 18 years among female youths 18 years and older is also presented in Table 2. Nearly 13% of sexually experienced female youths 18 years and older had ever had a pregnancy before age 18 years. Sixteen percent of unmarried female youths 18 years and older who had ever been threatened or forced to have sex had experienced an adolescent pregnancy compared with only 12% of female youths who had never been threatened or forced to have sex; this difference was not significant in cross-tabulations and unadjusted logistic regression. The unadjusted logistic regression showed that for every 1-point increase in the mean, standardized community-level threat or force variable, young women were 16% more likely to have ever experienced an adolescent pregnancy (OR=1.16; 95% CI=1.03, 1.29); the results were similar in model 3 in which the individual and community-level violence variables were included simultaneously.

The results of the multivariate analyses of condom use at last sexual encounter and HIV status are presented in Table 3. For each multivariate analysis, 3 models are presented. The first model included just the individual-level threat or force variable, with control for the demographic factors. The second model included the community-level variables, with control for the demographic factors. The third model included both levels of violence to determine whether, with control for an individual’s own exposure to violence, being from a community with a higher prevalence of violence affected sexual behaviors and health outcomes.

TABLE 3
Odd Ratios (ORs) and 95% Confidence Intervals (CIs) From Multilevel Logistic Regression Analyses of Condom Use at Last Sexual Encounter and HIV Status Among Unmarried Sexually Experienced Female Youths: South African Youth Survey, 2003

Model 1 for condom use at last sexual encounter showed that youths who had ever been threatened or physically forced to have sex were significantly less likely to have used a condom than were youths who had never been threatened or forced (OR=0.64; 95% CI=0.46, 0.88). Model 2 showed that youths from communities with a higher prevalence of threatened or forced sex were significantly less likely to use condoms than were youths from communities with a lower prevalence of threatened or forced sex (OR=0.86; 95% CI=0.77, 0.96). Finally, in model 3, which included all variables, individual-level experience with threats or force was marginally associated with nonuse of condoms at last sexual encounter (OR=0.71; 95% CI=0.49, 1.02) as was community-level violence (OR=0.89; 95% CI=0.78, 1.00).

The HIV models are also presented in Table 3. Model 1 showed that individual experience with threats or force was not associated with HIV status. Model 2 showed that youths who were from communities with a higher percentage of female youths reporting violence experience were significantly more likely to be HIV-positive than were youths from communities with a lower proportion of youths who experienced threats or force. Finally, model 3, which controlled for all variables, showed that with both individual and community-level variables in the model, the community-level violence measure was significantly associated with HIV status (OR=1.17; 95% CI=1.03, 1.32).

The same modeling approach for the analysis of adolescent pregnancy experience is presented in Table 4. In model 1 (and model 3), individual-level violence experience was not significant. Conversely, in model 2 (and model 3), female youths who were from communities where a higher proportion of female youths experienced threats or force were significantly more likely to have experienced an adolescent pregnancy than were female youths from communities with lower threat or force experience (OR=1.16; 95% CI=1.01, 1.33). Model 3 also showed that unmarried female youths from communities with a higher percentage of youths who had ever had sex were less likely to have ever been pregnant than were youths from communities with a lower percentage of youths who were sexually experienced (OR=0.86; 95% CI=0.72, 1.02).

TABLE 4
Odd Ratios (ORs) and 95% Confidence Intervals (CIs) From Multilevel Logistic Regression Analyses of Adolescent Pregnancy Experience Among Unmarried Sexually Experienced Female Youths Aged 18 Years and Older (Unweighted n = 3015): South African Youth Survey, ...

DISCUSSION

Our analyses, similar to those from studies of adults,1518 show that community-level factors are associated with reproductive health outcomes in youths. By controlling for individual-level experience of violence, we showed that youths from communities with greater sexual violence were significantly less likely to use condoms at their last sexual encounter and were more likely to be HIV positive or to have experienced an adolescent pregnancy than were youths from communities with lower violence experience. Individual-level violence was associated with condom nonuse only. The finding that community-level violence was significant but individual-level violence was not for HIV status and adolescent pregnancy experience is similar to the results of a 2007 study involving modern contraceptive use in 6 countries. In 2 of the 6 countries, community-level family planning approval was more important than individual-level perceived partner approval.15 The contraceptive use study and our study illustrate that social norms and community influences are important, especially in terms of reproductive health behaviors and outcomes.

Community experience with violence may be indicative of multiple contextual factors at play. Communities where there is greater sexual violence may have a higher underlying prevalence of HIV such that all youths in sexual partnerships are at greater risk of HIV than are youths in communities with a lower HIV prevalence. Alternatively, communities with a higher prevalence of violence may be places where there is a greater imbalance in gender power and women have less decision-making power. In these communities, women may have little say over the timing and circumstances of sex, thus increasing their risk of HIV, especially if their main partners have outside sexual partners. Similarly, communities with higher sexual violence may also be communities with less structural interventions for HIV and adolescent pregnancy prevention. For example, if these communities have fewer youth-friendly family planning and HIV counseling and testing centers, then youths may have less access to prevention programs and messages, as well as less access to condoms and other contraceptives.

This study had limitations. First, we could not include all of the sampled enumeration areas because some were too small and were not easily grouped with other enumeration areas into municipalities. This means that although we started with a representative sample of female youths, the analysis sample was somewhat less representative. That said, only a small number of enumeration areas were ultimately dropped, and the final analysis sample was 5956 young women (unweighted) rather than 6217 young women (unweighted). Overall, less than 5% of the sample was dropped. Second, the regrouping into municipalities was based on the South African census sampling plan and was not based on natural neighborhood or community groupings as identified at the local level. This means that the grouped community-level variables were somewhat artificial. To undertake this analysis more accurately, it would be necessary to (1) identify specific boundaries to communities as defined by community members, and (2) sample enough individuals in each community to determine the influence of community norms on youths’ behaviors.

Third, the community-level variables presented were based on a weighted aggregation of responses of a youths-only sample. Therefore, the community-level violence experience represents the experience of violence among youths and not the population-level experience (i.e., the prevalence if adult women were included). Similarly, the community-level variables were based on weighted aggregate measures of individual experiences and were not meant to represent community characteristics (e.g., number of schools, health centers, infrastructure). Moreover, there was no community-level measure of poverty or socioeconomic status or measures of migration, community cohesion, and access to condoms; these missing variables may bias the models presented. Additional data from multiple sources would be needed to include these types of community-level factors. Finally, because this study was based on cross-sectional data, it is not possible to know the true direction of effects between such things as HIV status and violence experience. It may be that when a violent act occurred, HIV was also transmitted (or a pregnancy happened). Conversely, youths who are HIV positive (or pregnant) may experience threats or force upon revealing the health outcome to their partner.19,20 A longitudinal study would be needed to determine the timing of these key sexual and reproductive health events.

Our study provided greater depth on the influence of violence on reproductive health behaviors and outcomes by including the community-level violence experience in the analysis. In particular, individual-level and community-level violence were associated with condom use (behavior), but only community-level violence was significantly associated with HIV status and adolescent pregnancy experience (health outcomes). Although individual-level violence discourages a young woman’s condom use, it does not directly lead to risk of pregnancy or HIV infection. Given that multiple factors influence adolescent pregnancy and HIV risk (including contraceptive use, coital frequency, partner status, number of sexual partners, type of partnerships, and age of sexual debut), it is not surprising that individual-level violence had a weak or nonsignificant effect on the health outcomes. Finally, it is important to note that community-level violence was associated with the health outcomes and individual-level behavior (condom use). This likely reflects social norms that are supportive of violence and structural factors, such as lack of HIV counseling and testing and lack of access to condoms and family planning at the community level, that are associated with less condom use and a greater risk of a pregnancy or HIV in these communities.

Future studies that include qualitative data to answer the why and how questions of the effect of sexual violence on risk-taking and outcomes are needed. In addition, multiple data sources should be used to obtain greater depth on structural factors that are protective (or lead to risk) at the community level. This type of information is crucial for expanding our understanding of how to prevent sexual violence and its associated negative outcomes (HIV infection and adolescent pregnancy). Furthermore, future programs for HIV and pregnancy prevention among youths need to address community-level norms that increase young women’s risk of these negative outcomes. This includes undertaking mass media and community-level drama programs to change gender-based violence norms with the objective of reducing sexual violence in South Africa and elsewhere in sub-Saharan Africa.

Acknowledgments

This secondary data analysis was supported by the US Agency for International Development (USAID) under cooperative agreement GPO-A-00-03-00003-00 with the MEASURE Evaluation project.

An earlier version of this article was presented at the American Public Health Association Meetings in Boston, MA, in November 2006.

Footnotes

Contributors I. S. Speizer originated the study, oversaw all data analyses, and led the writing. A. Pettifor assisted in interpretation and writing of results. S. Cummings undertook data analyses and assisted with interpretation of the data. C. MacPhail assisted with results interpretation and writing. I. Kleinschmidt provided coordinates to help develop the community-level variables and assisted with results interpretation. H. V. Rees and A. Pettifor led the original Reproductive Health and HIV Research Unit National Youth Survey. All authors contributed to critical revisions of the article.

Note. The views expressed here are those of the authors and not necessarily those of USAID or the United States government.

Human Participant Protection The study was approved by the Committee for the Protection of Human Subjects, University of Witwatersrand, South Africa. This secondary data analysis was approved by the institutional review board of the University of North Carolina at Chapel Hill.

Contributor Information

Ilene S. Speizer, Department of Maternal and Child Health, Gillings School of Global Public Health, University of North Carolina, Chapel Hill.

Audrey Pettifor, Department of Epidemiology, Gillings School of Global Public Health, University of North Carolina, Chapel Hill.

Stirling Cummings, Carolina Population Center, University of North Carolina, Chapel Hill.

Catherine MacPhail, Reproductive Health and HIV Research Unit, University of the Witwatersrand, Johannesburg, South Africa.

Immo Kleinschmidt, Tropical Epidemiology Group, Department of Epidemiology and Public Health, London School of Hygiene and Tropical Medicine, London, England.

Helen V. Rees, Reproductive Health and HIV Research Unit, University of the Witwatersrand, Johannesburg, South Africa.

References

1. Joint United Nations Programme on HIV/AIDS (UNAIDS) World Health Organization (WHO) AIDS Epidemic Update: December 2006. UNAIDS and WHO; Geneva: 2006.
2. Lamptey PR, Johnson JL, Khan M. The global challenge of HIV and AIDS. Popul Bull. 2006;61(1):1–28. Available at http://www.prb.org/pdf06/61.1GlobalChallenge_HIV AIDS.pdf.
3. National HIV and Syphilis Antenatal Seroprevalence Survey in South Africa 2005. South Africa Department of Health; Pretoria, South Africa: 2006.
4. Nugent R. Youth in a Global World. BRIDGE Project Policy Brief, Population Reference Bureau. Population Reference Bureau; Washington, DC: 2006.
5. Pettifor AE, Kleinschmidt I, Levin J, et al. A community-based study to examine the effect of a youth HIV prevention intervention on young people aged 15-24 in South Africa: results of the baseline survey. Trop Med Int Health. 2005;10(10):971–980. [PubMed]
6. Casterline J, Chalasani S. Young women and childbearing in Africa: the new skepticism. Presented at: 2007 Annual Meeting of the Population Association of America; New York, NY. March 29–31, 2007.
7. Jewkes RK, Levin JB, Penn-Kekana LA. Gender inequalities, intimate partner violence and HIV preventive practices: findings of a South African cross-sectional study. Soc Sci Med. 2003;56:125–134. [PubMed]
8. Pettifor AE, Measham DM, Rees HV, Padian NS. Sexual power and HIV risk, South Africa. Emerg Infect Dis. 2004;10(11):1996–2004. [PMC free article] [PubMed]
9. Dunkle KL, Jewkes RK, Brown HC, Gray GE, McIntryre JA, Harlow SD. Gender-based violence, relationship power and risk of HIV infection in women attending antenatal clinics in South Africa. Lancet. 2004;363:1415–1421. [PubMed]
10. Maharaj P, Munthree C. Coerced first sexual intercourse and selected reproductive health outcomes among young women in KwaZulu-Natal. South Africa. J Biosoc Sci. 2007;39(2):231–244. [PubMed]
11. Jewkes R, Vundule C, Maforah F, Jordaan E. Relationship dynamics and teenage pregnancy in South Africa. Soc Sci Med. 2001;52:733–744. [PubMed]
12. Pettifor AE, Rees HV, Kleinschmidt I, et al. Young people’s sexual health in South Africa: HIV prevalence and sexual behaviors from a nationally representative household survey. AIDS. 2005;19(14):1525–1534. [PubMed]
13. Huber P. The behavior of maximum likelihood estimates under non-standard conditions. In: Le Cam LM, Neyman J, editors. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability; Berkeley, CA: University of California Press; 1967. pp. 221–233.
14. White H. Maximum likelihood estimation of misspecified models. Econometrica. 1982;50:1–25.
15. Stephenson R, Baschieri A, Clements S, Hennink M, Madise N. Contextual influences on modern contraceptive use in sub-Saharan Africa. Am J Public Health. 2007;97:1233–1240. [PMC free article] [PubMed]
16. Pallitto CC, O’Campo P. Community level effects of gender inequality on intimate partner violence and unintended pregnancy in Colombia: testing the feminist perspective. Soc Sci Med. 2005;60:2205–2216. [PubMed]
17. McQuestion MJ. Endogenous social effects on intimate partner violence in Colombia. Soc Sci Res. 2003;32:335–345.
18. Koenig MA, Stephenson R, Ahmed S, Jejeebhoy SJ, Campbell J. Individual and contextual determinants of domestic violence in north India. Am J Public Health. 2006;96:132–138. [PMC free article] [PubMed]
19. Maman S, Campbell J, Sweat MD, Gielen AC. The intersections of HIV and violence: directions for future research and interventions. Soc Sci Med. 2000;50:459–478. [PubMed]
20. Van der Straten A, King R, Grinstead O, Vittinghoff E, Serufilira A, Allen S. Sexual coercion, physical violence, and HIV infection among women in steady relationships in Kigali, Rwanda. AIDS Behav. 1998;2(1):61–73.
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...