The abuse of women by their male partners is common in many regions where women are tested for HIV-1 in prevention of mother-to-child HIV-1 transmission (PMTCT) programmes . In these settings, HIV-1 testing may place women at risk of domestic violence, and fear of domestic violence (physical, financial, or psychological abuse) may make it difficult for women to access interventions to prevent mother-to-child transmission of HIV-1. There is therefore a need to understand how domestic violence interacts with PMTCT programmes.
There is increasing evidence from studies in sub-Saharan Africa linking physical abuse by intimate partners to HIV infection among women [2–4]. However, the reasons for this association are not well understood.
Studies on domestic violence after HIV-1 testing in sub-Saharan Africa, have not consistently found an association between HIV-1 test results and domestic violence. Among HIV-1-positive women after testing some studies have reported high rates of domestic violence, divorce, or separation, 6% in 2 years in one study, and 7% in 14 weeks in another [5,6]. However, other studies have not found that partner notification of positive HIV-1 results increases the risk of post-test domestic violence, with less than 5% of HIV-1-positive women reporting domestic violence, divorce, or separation after testing in one year of follow-up [7,8]. Those studies have been limited by a lack of baseline data on domestic violence before testing, low rates of notifying partners of positive HIV-1 results, and a lack of documentation of the reasons for conflict.
We interviewed and followed participants in a study on couple HIV-1 counseling and testing in an antenatal clinic, to determine the prevalence of domestic violence before and after HIV-1 testing, and the impact of domestic violence on the uptake of interventions to prevent perinatal HIV-1 transmission .
Study population and clinic procedures
Mathare North City Council clinic, where this study was conducted, is a public antenatal clinic that serves a low socioeconomic population in Nairobi, Kenya. The recruitment and follow-up procedures have been described elsewhere . Briefly, at the first antenatal clinic visit, women in groups of five to 10 were given health education and invited to participate. Baseline data on sociodemographic characteristics, relationship with current partner, living conditions, and domestic violence by the current partner were collected. Women were asked to return, at their convenience, for counseling and HIV-1 testing, either alone or with their partners. Similar information was obtained from consenting men who presented for counseling. After HIV-1 testing, women were asked to return to the clinic after 2 weeks for further counseling. At this 2-week post-test visit, nevirapine to prevent mother-to-child transmission of HIV-1 was dispensed to HIV-1-positive women, and all women completed a questionnaire on interim domestic violence.
Domestic violence included physical, financial, or psychological abuse by the current male partner. Physical abuse was defined as being physically assaulted, hit, slapped, or pushed. Psychological abuse was defined as the use of abusive language, yelling, humiliation, or actions intended to cause emotional pain. Intentional withholding of financial support or being forbidden to earn were considered financial abuse. To ensure confidentiality and promote confidence of clients, women saw the same counselor at each visit to the clinic, and questionnaires were administered to men and women separately in private rooms. The study was performed in accordance with the World Health Organization recommendations on ethics and safety for research on violence against women .
Data were analysed using SPSS-PC (Chicago, Illinois, USA) version 11.5. Chi-square and Wilcoxon rank sum tests were used to compare categorical and continuous independent data, respectively. McNemar's test was used for paired categorical data. Multivariate logistic regression was used to determine whether domestic violence at baseline was an independent predictor of HIV-1 infection, the uptake of PMTCT interventions, partner involvement in PMTCT, partner notification of HIV-1 test results, and domestic violence after receiving HIV-1 test results.
Recruitment and follow-up
Of 3137 women invited to join the study, 2836 (90%) were enrolled, 2231 (71%) were counseled and 313 (10%) came for counseling with their male partners. After counseling, 2104 women (94%) and 301 men (96%) accepted HIV-1 testing. Among those tested, 314 women (15%) and 32 men (11%) tested HIV-1 positive. Two weeks after receiving HIV-1 test results, 1640 women completed the study; 58% of enrolled and 78% of tested women (Fig. 1). Compared with the 1640 women who completed the study, the 441 women lost to follow-up after receiving HIV-1 test results were more likely to be HIV-1 positive (21 versus 13%; P < 0.001), and to have reported domestic violence at baseline (32 versus 26%; P = 0.02).
Domestic violence before testing
At baseline, 804 of 2836 enrolled women (28%) reported having experienced at least one episode of domestic violence by their current partner; 570 (20%) reported physical abuse, 429 (15%) reported psychological abuse and 217 (8%) reported financial abuse. The prevalence of domestic violence by the current partner increased with the duration of the relationship, from 13% among women in a relationship for less than one year to 43% among women in a relationship for at least 5 years. Among the 570 women reporting physical abuse, the last episode was precipitated by disagreements caused by suspected infidelity by the male partner (27%), finances (23%), pregnancy (11%), alcohol or drug use (11%), care of children (9%), and disrespect by the wife (9%).
In a multivariate logistic regression model, after adjusting for the duration of the relationship, a history of a sexually transmitted disease, polygamous marriage, and living in a house with more than three people were associated with domestic violence. Using the same model, women with post-secondary school education, whose sexual debut was after age 17 years, whose partners had post-secondary school education, who were in formalized marriages, and who were paying a house rent of more than US$20 per month had significantly lower odds of reporting having experienced domestic violence (Table 1).
HIV-1 test results and partner involvement
Women who tested HIV-1 positive were significantly more likely to have experienced domestic violence before testing than women who tested HIV-1 negative (37 versus 26%; P < 0.001). This difference persisted as a trend after adjusting for the number of lifetime sexual partners, age of sexual debut, polygamous marriage and number of people in the household. Partners of women who reported domestic violence were less likely to come for counseling than partners of women who did not report domestic violence (8.3 versus 12%; P = 0.005). This difference remained significant after adjusting for education level and the duration of the relationship. Two weeks after HIV-1 testing 1486 of 1638 women (91%) reported that they had informed their partners of their HIV-1 test results. Compared with those who did not notify their partners, women who notified their male partners of their HIV-1 test results were significantly less likely to have reported domestic violence at baseline (38 versus 25%; P = 0.001; Table 2).
Uptake of prevention of mother-to-child transmission interventions
The proportion of women reporting domestic violence before testing was slightly lower among women who received HIV-1 test counseling than among those who did not (28 versus 31%; P = 0.1) and among women who received compared with those who did not receive HIV-1 test results (28 versus 31%; P = 0.1) and was comparable among women who accepted HIV-1 testing and those who declined (28 versus 30%; P = 0.2; Table 3).
Two weeks after receiving HIV-1 test results, of 967 women who reported having sex, 334 (35%) reported using condoms at least once. Women who used condoms tended to be less likely to have reported domestic violence at baseline (21 versus 27%; P = 0.06). This trend remained after adjusting for HIV-1 test results [odds ratio (OR) 0.7, 95% confidence interval (CI) 0.5–1.0; P = 0.06]. There was no difference in the prevalence of reported domestic violence at baseline between women who accessed nevirapine and those who did not (39 versus 36%; P = 0.6; Table 3).
Domestic violence after HIV testing
In the 2 weeks after HIV-1 testing, domestic violence was reported by 15 of 1638 women (0.9%): 10 reported financial abuse, 10 psychological abuse, and five physical abuse. Six of the women (40%) reporting abuse said it was related to HIV-1 testing, five (33%) said it was caused by financial disagreement, four (27%) said it was caused by suspected infidelity by the male partner, and five (33%) gave other reasons.
Women who experienced domestic violence after testing tended to be more likely to have reported domestic violence before testing compared with women who did not experience domestic violence after testing (47 versus 26%; P = 0.08). Testing HIV-1 positive was a predictor of experiencing domestic violence after HIV-1 testing that was independent of having reported abuse before testing (OR 4.0, 95% CI 1.4–12; P = 0.009). After adjusting for a history of domestic violence at baseline and partner notification, the odds of HIV-1-seropositive women experiencing domestic violence were five times those of HIV-1-seronegative women (OR 4.8, 95% CI 1.4–16; P = 0.01; Table 4).
Domestic violence reports by male partners
Of 2836 women enrolled, 313 (11%) had their partners come for counseling. Among the 313 couples, the proportion of male partners and women reporting domestic violence during the relationship was similar (23 versus 21%; P = 0.6). Men were more likely to report that they physically abused their partner than women reported being abused (19 versus 14%; P = 0.04), but were less likely to report that the abuse led to physical injuries than their female partners (1 versus 3%; P = 0.04). The frequency of reasons reported for the last episode of physical abuse were similar for men and women: unfaithfulness by the man (6% of women versus 4% of men; P = 0.1), misunderstanding over money (2% of women versus 5% of men; P = 0.1), disagreement over relatives (2% of women versus 1% of men; P = 0.5), pregnancy (2% of women versus 1% of men; P = 1.0), drunkenness (1% of women versus 1% of men; P = 1.0), and misunderstanding over care of children (0% of women versus 2% of men).
Our findings confirm those of previous studies that domestic violence is associated with HIV infection [3,4]. Our finding that male partner behavior placing women at risk of HIV-1 infection, such as infidelity and having multiple partners, is a source of conflict may partly explain this association. This finding also suggests that interventions to prevent domestic violence that target reasons for conflict may have an additional benefit of preventing HIV-1 infection.
Previous studies have found fear of domestic violence to be a reason for women refusing HIV-1 testing and failing to notify partners of HIV-1 test results [9–11]. In our study, domestic violence before testing did not compromise the uptake of HIV-1 testing or other PMTCT interventions, but it was a barrier to bringing male partners for counseling, notifying partners of results, and using condoms. Therefore, domestic violence may limit the effectiveness of interventions that require male partner participation such as couple counseling and using PMTCT as an entry point for the prevention of the heterosexual transmission of HIV-1. There is a need for research to identify other strategies to reach partners of women who experience domestic violence.
In this study, the frequency of reported domestic violence in the immediate period after HIV-1 testing was low, which is consistent with other studies from sub-Saharan Africa that find that domestic violence after testing is infrequent [7,8]. However, positive HIV-1 test results appear to increase this risk, by acting as a new source of conflict after testing. To address this increased risk of domestic violence among HIV-1-seropositive women after testing, domestic violence prevention and screening needs to be integrated into programmes that provide HIV/AIDS care.
This study was unique in that we collected data on domestic violence before testing and obtained data on domestic violence directly from some male partners. Our findings suggest that domestic violence after testing is part of a continuing pattern of abuse that may be acceptable to men in the cultural context in which this study was performed, unless it leads to injury. Interventions to prevent domestic violence that target men in this setting will have to be culturally sensitive and take into account that physical abuse may only be stigmatized when it leads to injury.
Our study has some limitations. We probably underestimated the frequency of abuse after testing because of the high rate of loss to follow-up, particularly among HIV-1-positive women and women who had reported domestic violence at baseline. As a result of the short follow-up period (2 weeks) we observed few domestic violence events after testing. Although domestic violence that is related to HIV-1 testing is expected to occur soon after notifying partners of test results, it is possible that domestic violence related to HIV-1 testing will persist for a long time after the diagnosis of HIV-1.
In conclusion, our study findings show that although after notifying partners of test results many women do not experience domestic violence in the immediate post-HIV-1 test period, there is an increased risk of domestic violence after women notify partners of HIV-1-positive results. We also found that domestic violence does not limit the uptake of PMTCT interventions, but does limit partner involvement, which is important for the prevention of the heterosexual transmission of HIV-1. There is a need for innovative long-term approaches both to prevent domestic violence and to integrate domestic violence prevention into HIV-1 prevention programmes.
Sponsorship: This research was funded by the Elizabeth Glaser Pediatric AIDS Foundation (EGPAF). J.N.K. was a scholar in the AIDS International Training and Research Program supported by the National Institutes of Health/Fogarty International Center D43 TW00007. C.F. was supported by the National Institutes of Health (K23-HD41879). G.C.J-S. is an EGPAF scientist.
1. World Health Organization. Violence against women: a priority health issue
. Geneva: World Health Organization; 1997.
2. Van der Straten A, King R, Grinstead O, Vittinghoff E, Selufilira A, Allen S. Sexual coercion, physical violence, and HIV infection among women in steady relationships in Kigali, Rwanda. AIDS Behav 1998; 2:61–73.
3. Maman S, Mbwambo JK, Hogan NM, Kilonzo GP, Campbell JC, Weiss E, Sweat MD. HIV-positive women report more lifetime partner violence: findings from a voluntary counseling and testing clinic in Dar es Salaam, Tanzania. Am J Public Health 2002; 92:1331–1337.
4. 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.
5. Temmerman M, Ndinya-Achola J, Ambani J, Piot P. The right not to know HIV-test results. Lancet 1995; 345:969–970.
6. Gaillard P, Melis R, Mwanyumba F, Claeys P, Muigai E, Mandaliya K, et al
. Vulnerability of women in an African setting: lessons for mother-to-child HIV transmission prevention programmes. AIDS 2002; 16:937–939.
7. Yacouba N, Nicolas M, Valeraine L, Laurent M, Seydou Y, Issiaka S, et al
. Sexual and reproductive life of women informed of their HIV seropositivity: a prospective cohort study in Burkina Faso. J Acquir Immune Defic Syndr 2001; 28:367–372.
8. Kissinger PJ, Niccolai LM, Magnus M, Farley TA, Maher JE, Richardson-Alston G, et al
. Partner notification for HIV and syphilis: effects on sexual behaviors and relationship stability. Sex Transm Dis 2003; 30:75–82.
9. Antelman G, Fawzi MCS, Kaaya S, Mbwambo J, Msamanga GI, Hunter DJ, Fawzi WW. Predictors of HIV-1 serostatus disclosure: a prospective study among HIV-infected pregnant women in Dar es Salaam, Tanzania. AIDS 2001; 15:1865–1874.
10. Kilewo C, Massawe A, Lyamuya E, Semali I, Kalokola F, Urassa E, et al
. HIV counseling and testing of pregnant women in sub-Saharan Africa: experiences from a study on prevention of mother-to-child HIV-1 transmission in Dar es Salaam, Tanzania. J Acquir Immune Defic Syndr 2001; 28:458–462.
11. Pool R, Nyanzi S, Whitworth JA. Attitudes to voluntary counselling and testing for HIV among pregnant women in rural south-west Uganda. AIDS Care 2001; 13:605–615.
12. Farquhar C, Kiarie JN, Richardson BA, Kabura MN, John FN, Nduati RW, et al
. Antenatal couple counseling increases uptake of interventions to prevent HIV-1 transmission. J Acquir Immune Defic Syndr 2004; 37:1620–1626.
13. World Health Organization. Putting women first: ethical and safety recommendations for research on domestic violence against women
. Geneva: World Health Organization; 2001.