Abstract
Medical male circumcision is a proven method of HIV risk reduction in men with no known direct benefit to women. We investigated the benefit of partner circumcision on women’s health. We conducted a secondary analysis of 5,029 women enrolled in the Vaginal and Oral Interventions to Control the Epidemic trial across 15 African sites, to look at the impact of partner circumcision status on sexually transmitted infections, pregnancy, frequency of sex, and condom use in women. Of 4,982 participants with a baseline response, 31% had circumcised partners. Women with circumcised partners had a significantly reduced risk of syphilis acquisition, hazard ratio 0.51 (0.26, 1.00), p value =] .05. Participants with uncircumcised partners were significantly less likely to have used a condom at the last sex act than the other two groups, adj. relative risk 0.86 (0.80, 0.92), adj. p value < .0001. We found no evidence of sexual risk compensation in women with circumcised partners.
Keywords: male circumcision, HIV, sexual risk compensation, women’s health
Medical male circumcision (MMC) has been shown to reduce the risk of human immunodeficiency virus (HIV) acquisition in men by 50–60% in three randomized controlled trials conducted in Africa (Auvert et al., 2005; Bailey et al., 2007; Gray et al., 2007). These findings resulted in scale up of voluntary medical male (VMMC) circumcision as a HIV prevention method in 14 high priority Southern and East African countries with high HIV prevalence and low levels of male circumcision (Terms, 2007). It is however important to understand that male circumcision offers only partial protection against female to male transmission of HIV as well as other sexually transmitted infections (STIs). This direct protective effect of MMC in men has been shown to be linked to the greater density of HIV-1 susceptible Langerhans cells and HIV-1 susceptible receptors located on the inner and outer foreskin as compared to the glands and penile shaft (McCoombe & Short, 2006; Szabo & Short, 2000). HIV transmission is also associated with balanoposthitis, the presence of smegma, ulceration caused by STIs and tearing of the foreskin or associated frenulum during sexual intercourse, thus explaining why male circumcision protects against HIV infection in men (Morris & Wamai, 2012).
Although the protective effect of male circumcision in men is well known, there is currently insufficient data to determine whether circumcision would directly reduce the risk of HIV acquisition in women (Weiss, Hankins, & Dickson, 2009). There are however, indirect beneficial effects on the health outcomes of their female partners. A recent systematic review conducted by Grund and colleagues provided strong and consistent evidence that male circumcision is protective against cervical cancer, cervical dysplasia, herpes simplex virus type 2, chlamydia, and syphilis in female partners (Grund et al., 2017). The authors also noted medium consistency evidence that women with circumcised partners were protected against human papillomavirus, a cancer-causing virus (Grund et al., 2017). Low consistency evidence was shown for gonorrhea, bacterial vaginosis, trichomonas, genital ulcers, and HIV infection; although at a population level, widespread male circumcision would benefit women by reducing their risk of exposure to HIV and STI’s (Grund et al., 2017; Weiss et al., 2009).
There are also some concerns about male circumcision and its impact on women’s health. Circumcision of HIV-infected men did not reduce HIV transmission to female partners (Wawer et al., 2009). Transmission of HIV to their female partners was noted to be higher in couples who resumed sex prior to complete wound healing post MMC surgery (Wawer et al., 2009). Hence, the evidence concludes that HIV risk reduction in female partners is not evident when the male partners are HIV infected.
There is also further concern regarding the potential risk of sexual risk compensation among women with circumcised partners. Risk compensation is defined as increased risky behaviors by the individual or their partner due to decreased perceived risk following the introduction of a partially efficacious biomedical intervention (Cassell, Halperin, Shelton, & Stanton, 2006). Women are being made aware by MMC promotion campaigns that circumcised men are potentially less at risk for HIV infection, which may reduce their perceived risks and hence their protective behaviors (Feuer, 2010). Women are also being educated on the indirect benefits of male circumcision on their health (Feuer, 2010).
Studies to date have reported conflicting results on the impact of MMC on sexual risk behavior among men. In Orange Farm, South Africa, circumcised participants reported a higher mean number of sexual partners during follow up visits. Despite this higher risk behavior, the protective effect of circumcision remained unchanged after adjustment for sexual behavior and was consistent with the results of the other two trials (Auvert et al., 2005). In a study in Kisumu, Kenya, there was an increase in condom use and a decrease from baseline in the number of sexual partners and unprotected sexual acts in both control intervention arms (Bailey et al., 2007). In a study in Rakai Uganda, rates of condom use increased in both the intervention and control arms during the study (Gray et al., 2007). A qualitative study conducted in a rural setting in South Africa identified beliefs among men that may suggest potential risk compensation. These beliefs included overestimation of the protective effect of male circumcision, perception that male circumcision allows them to have more sexual partners and more sex and that they could use condoms less consistently unless they wanted to protect against pregnancy (Humphries, van Rooyen, Knight, Barnabas, & Celum, 2015).
There is however, little evidence or research conducted on how MMC influences women’s sexual behaviors. A study among university students in KwaZulu-Natal, South Africa, noted that women perceived that MMC affords no direct benefit to women and felt that MMC could increase their risk for HIV acquisition as a result of men increasing their risky behavior such as having multiple partners and decreasing condom use (Mantell et al., 2013). A recent study on male circumcision and risk compensation and perceptions among South African women showed that awareness of men’s circumcision status resulted in a belief that they were more protected against HIV infection (Kalichman, Mathews, Kalichman, Eaton, & Nkoko, 2018). Many women participating in the WHiPT project (HIV prevention tracking report for women across five African counties), believed that they would be directly protected from HIV acquisition if their male partners were protected (Women’s HIV Prevention Tracking Project, 2010). A vast majority of women also perceived that male circumcision might lead to an increase in gender-based violence (GBV) and stigma for women living with HIV (Women’s HIV Prevention Tracking Project, 2010).
Further insights into the role women play in relation to MMC including the potential for sexual risk compensation will be helpful to understand aspects related to the roll out of the VMMC programs. In this study, we describe the prevalence of reported male partner circumcision and describe associations between male partner circumcision and incident STI and HIV infections, pregnancy, and risky sexual behavior in women.
METHODS
The Vaginal and Oral Interventions to Control the Epidemic (VOICE) trial was a multicenter, randomized, double blind, placebo-controlled trial of the safety and effectiveness study of oral and topical PrEP (1% Tenofovir vaginal gel, oral Tenofovir and oral Truvada) in preventing sexually acquired HIV-1 infection (Marrazzo et al., 2015). This study is a secondary analysis from the VOICE trial, which enrolled 5,029 HIV uninfected women, aged 18 to 45 years at 15 sites across three countries: South Africa (eight sites in KwaZulu-Natal, three sites in Johannesburg), Uganda (one site in Kampala), and Zimbabwe (two sites in Chitungwiza and one site in Harare). Details of the VOICE trial methods have been described elsewhere (Marrazzo et al., 2015) and includes the study eligibility criteria as well as the main study results.
PARTICIPANTS
Briefly the study initiated in September 2009 and enrolled HIV uninfected, sexually active women aged 18 to 45 years, who were neither pregnant nor breastfeeding, were willing to use effective contraception and who had a normal renal, hematologic and hepatic function at baseline (Marrazzo et al., 2015). All participants attended monthly study visits for study product provision together with individualized adherence counseling. All participants were tested for HIV-1 infection and pregnancy monthly and for STIs at specified study visits. Participants completed a behavioral risk assessment at specified study visits and were provided with risk-reduction counseling, male and female condoms, contraception, and treated for STIs as per World Health Organization (WHO) and local guidelines. Written informed consent was obtained from all participants prior to enrolment into the study. Local and international ethics and regulatory bodies in all participating countries, including the U.S. Food and Drug Administration, approved the protocol.
STATISTICAL ANALYSIS
Circumcision status of the male sexual partner was self-reported by women at baseline and quarterly during follow-up. Of the 5,029 participants randomized in VOICE, 4,982 (99.1%) reported the circumcision status (Yes, No, Do not know) of their male sexual partner at baseline. These participants were included in all the following analyses when relevant data were available. A preliminary baseline analysis was performed by comparing the distributions of various baseline characteristics across the levels of male sexual partner circumcision status at baseline. Differences in circumcision status were tested by a Chi-squared test for categorical variables and by a Kruskal-Wallis test for medians of continuous variables.
For incident HIV, STIs, and pregnancy during the study, incident rates were calculated to the time of first occurrence, if an event occurred, or to the time of the last negative test result, if no event occurred. Cox proportional hazards models were used to determine any associations between these outcomes and the male sexual partner circumcision status at baseline. For the three nominal levels of circumcision status (Yes, No, Do not know), pair-wise comparisons were performed to compare each level. All models were stratified by site, and adjusted models included participant age, participant education, and marital status.
The association between quarterly follow-up male sexual partner circumcision status and condom use at the last vaginal sex prior to a visit over study follow-up was evaluated using generalized estimating equations (GEE) with a logit link and exchangeable correlation structure. Similarly, the association between quarterly follow-up male sexual partner circumcision status and frequency of sex in the 7 days prior to a visit was evaluated using GEE with a log link (Poisson) and exchangeable correlation structure. All models were stratified by site, and adjusted models included participant age, participant education, and marital status. For both outcomes, we used all available follow-up data for the 4,982 participants with a baseline male sexual partner circumcision status.
RESULTS
Of the 4,982 participants with a baseline response, 1,561 (31%) reported having circumcised partners, 2,863 (58%) reported having uncircumcised partners and 558 (11%) did not know the circumcision status of their partner. Partner circumcision rates differed between countries, with a circumcision prevalence of 10% in Zimbabwe, 33% in South Africa, and 51% in Uganda.
Table 1 contains the baseline characteristics of participants and their male sexual partners. Women with circumcised partners had a median age of 24 years, a significantly higher proportion were unmarried (84%; p < .0001) and a significantly lower proportion were living with their partners (29%; p < .0001) than women with uncircumcised partners. Women with uncircumcised partners and unknown partner circumcision status were significantly more likely to be more worried about risk of HIV in the next year.
TABLE 1.
Participant and Partner Baseline Characteristics by Male Sexual Partner Circumcision Status at Baseline
| Baseline Characteristics | Total (n = 4,982) | Circumcised (n = 1,561) | Uncircumcised (n = 2,863) | Unknown Circumcision Status (n = 558) | p valuea |
|---|---|---|---|---|---|
| Participant Characteristics | |||||
| Median age, years (IQR) | 24 (21, 29) | 24 (21, 29) | 25 (21, 29) | 23 (20, 26) | < .0001 |
| n (%) | n (%) | n (%) | n (%) | ||
| Partner Status | < .0001 | ||||
| Married | 1,052 (21) | 257 (16) | 714 (25) | 81 (15) | |
| Unmarried, has primary sex partner | 3,930 (78) | 1304 (84) | 2,149 (75) | 477 (85) | |
| Currently living with partner | 1,606 (32) | 455 (29) | 1,042 (36) | 109 (20) | < .0001 |
| Condom use with partner, last sexb | 3,724 (75) | 1,142 (73) | 2,141 (75) | 441 (79) | .023 |
| Earns own income | 2,853 (57) | 844 (54) | 1,688 (59) | 321 (58) | .007 |
| Educationc | .005 | ||||
| Primary or less | 388 (8) | 128 (8) | 236 (8) | 24 (4) | |
| Secondary or more | 4,590 (92) | 1,432 (92) | 2,625 (92) | 533 (96) | |
| Median parity, live children (IQR)d | 1 (1, 2) | 1 (1, 2) | 1 (1, 2) | 1 (1, 2) | < .0001 |
| Gonorrheae | 163 (3) | 61 (4) | 79 (3) | 23 (4) | .059 |
| Chlamydiae | 608 (12) | 195 (12) | 334 (12) | 79 (14) | .231 |
| Syphilis | 70 (1) | 20 (1) | 45 (2) | 5 (1) | .408 |
| HSV-2f | 2,268 (46) | 739 (48) | 1,323 (46) | 206 (37) | < .0001 |
| Disclosed study participation to primary sex partnerg | 4,028 (81) | 1,245 (80) | 2,368 (83) | 415 (74) | < .0001 |
| Perceived risk of HIV in the next yearh | .002 | ||||
| Very worried | 3,016 (61) | 880 (57) | 1,770 (62) | 366 (66) | |
| Somewhat worried | 832 (17) | 280 (18) | 468 (16) | 84 (15) | |
| Not at all worried | 1,088 (22) | 377 (25) | 609 (21) | 102 (18) | |
| Partner Characteristics | |||||
| Median age, years (IQR)i | 28 (25, 33) | 29 (25, 34) | 28 (25, 34) | 26 (24, 31) | < .0001 |
| n (%) | n (%) | n (%) | n (%) | ||
| Has other sexual partners | < .0001 | ||||
| Yes | 764 (15) | 266 (17) | 429 (15) | 69 (12) | |
| No | 1,231 (25) | 392 (25) | 741 (26) | 98 (18) | |
| Participant does not know | 2,987 (60) | 903 (58) | 1,693 (59) | 391 (70) | |
| Provides financial supportj | 4,148 (83) | 1,294 (83) | 2415 (84) | 439 (79) | .004 |
| Educationk | < .0001 | ||||
| Primary or less | 181 (4) | 53 (3) | 115 (4) | 13 (2) | |
| Secondary or more | 4,480 (90) | 1,363 (87) | 2,602 (91) | 515 (92) | |
| Participant does not know | 320 (6) | 145 (9) | 145 (5) | 30 (5) | |
Note. Percentages may not equal 100 due to rounding. IQR: interquartile range.
p value for difference in circumcision status from chi-squared test for categories and from Kruskal-Wallis test for medians.
Based on n = 4,972 participants with baseline data for condom use at last vaginal sex act.
Based on n = 4,978 participants with baseline data for participant education.
The mean parity is highest among uncircumcised (1.61 live children, standard deviation 1.13), followed by circumcised (1.47, 1.21) and unknown (1.13, 0.90).
Based on n = 4,981 participants with baseline gonorrhea and chlamydia test results.
Based on n = 4,962 participants with baseline HSV-2 EIA test results.
Based on n = 4,974 participants with baseline data for study disclosure to partner.
Based on n = 4,936 participants with a response on question 7 on the Baseline ACASI Questionnaire.
Partner age was reported by 4,702 participants (circumcised, 1,437; uncircumcised, 2,725; unknown, 540).
Based on n = 4,979 participants with baseline data for provision of financial support to participant.
Based on n = 4,981 participants with baseline partner education status.
Incidence rates of HIV, STIs, and pregnancy are displayed by level of circumcision status in Table 2, and selected comparisons are displayed in Table 3. There was no significant difference in HIV incidence between women with circumcised and uncircumcised partners, adjusted (adj.) hazard ratio (HR) 1.14 (0.87, 1.48), adj. p value = .345. The pregnancy incidence rate between women with circumcised and uncircumcised partners was not significantly different (adj. p value = .479). There was no significant difference in gonorrhoea and chlamydia incidence rates between women with circumcised and uncircumcised partners (adj. p value = .281 and adj. p value = .652, respectively). Among women with circumcised partners we observed a trend toward a significantly reduced risk of syphilis acquisition compared to all other women, adj. HR 0.52 (0.27, 1.02), adj. p value = .055.
TABLE 2.
HIV, STIs, and Pregnancy Incidence Rates by Level of Circumcision Status at Baseline
| Total | Circumcised | Uncircumcised | Unknown Circumcision Status | |
|---|---|---|---|---|
| HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | |
| HIV incidencea | 5.67 (5.05, 6.33) | 5.52 (4.44, 6.78) | 5.29 (4.52, 6.15) | 8.00 (5.94, 10.54) |
| Pregnancy incidence | 7.70 (6.98, 8.48) | 8.33 (6.98, 9.85) | 7.25 (6.34, 8.25) | 8.36 (6.26, 10.94) |
| Gonorrhea incidenceb | 3.25 (2.80, 3.76) | 2.93 (2.17, 3.88) | 3.36 (2.76, 4.05) | 3.57 (2.26, 5.35) |
| Chlamydia incidenceb | 12.00 (11.10, 12.95) | 11.67 (10.08, 13.45) | 11.61 (10.46, 12.85) | 14.84 (11.97, 18.17) |
| Syphilis incidenceb | 0.92 (0.69, 1.21) | 0.71 (0.37, 1.25) | 1.06 (0.74, 1.48) | 0.77 (0.25, 1.79) |
Note.
Based on n = 4,960 participants with HIV-negative status at baseline.
Based on n = 4,876 participants with baseline and follow-up STI results.
TABLE 3.
Comparison of HIV, STIs, and Pregnancy Incidence Rates between Levels of Circumcision Status
| Circumcision Status Comparison | HR (95% CI) | Unadjusted p value | Adjusted HR (95% CI) | Adjusted p value |
|---|---|---|---|---|
| Circumcised vs. uncircumcised | ||||
| HIV incidence | 1.08 (0.83, 1.40) | .587 | 1.14 (0.87, 1.48) | .345 |
| Pregnancy incidence | 1.08 (0.86, 1.35) | .5 | 1.08 (0.87, 1.36) | .479 |
| Gonorrhea incidence | 0.82 (0.58, 1.16) | .253 | 0.83 (0.58, 1.17) | .281 |
| Chlamydia incidence | 0.94 (0.79, 1.13) | .519 | 0.96 (0.80, 1.15) | .652 |
| Syphilis incidence | 0.52 (0.26, 1.02) | .056 | 0.52 (0.26, 1.02) | .058 |
|
Circumcised vs. all othersa Syphilis incidence |
0.51 (0.26, 1.00) | .05 | 0.52 (0.27, 1.02) | .055 |
Note.
The group of participants with circumcised partners compared with the combined group of participants with uncircumcised partners or unknown circumcision status.
Frequency of vaginal sex in the past 7 days and the proportion of condom use with the last vaginal sex prior to a visit are displayed by level of quarterly follow-up circumcision status in Table 4, and selected comparisons are displayed in Table 5. Participants with circumcised partners tended to have significantly fewer sex acts in the past 7 days than participants with uncircumcised partners, (6% less likely per each act of sex), adj. mean estimate 0.94 (0.90, 0.97), adj. p value = .0005. Participants with uncircumcised partners were significantly less likely to have used a condom at last sex act than the other two groups, adj. relative risk 0.86 (0.80, 0.92), adj. p value < .0001
TABLE 4.
Mean Number of Vaginal Sex in the 7 Days Prior to a Visit and Proportion Reporting Condom Use During Last Vaginal Sex per Level of Partner Circumcision Status at Same Follow-up Visit
| Total | Circumcised | Uncircumcised | Unknown Circumcision Status | |
|---|---|---|---|---|
| Vaginal Sex in the 7 Days Prior to a Visit | ||||
| Number of participant-visitsa | 28,366 | 10,710 | 13,964 | 3,692 |
| Mean number of vaginal sex in past 7 days (SD) | 2.28 (2.43) | 2.18 (2.40) | 2.42 (2.51) | 2.02 (2.14) |
| Condom Use During Last Vaginal Sex | ||||
| Number of participant-visitsb | 28,646 | 10,827 | 14,078 | 3,741 |
| Number of participant-visits with condom use | 20,184 | 7,787 | 94,60 | 2,937 |
| Proportion of visits with condom use (%) | 70.5 | 71.9 | 67.2 | 78.5 |
Note. SD = standard deviation.
Based on n = 28,366 participants with Vaginal Sex in the 7 days prior to a visit.
Based on n = 28,646 participants who reported condom use during last vaginal sex.
TABLE 5.
Associations of Vaginal Sex Frequency and Last Vaginal Sex Condom Use with Partner Circumcision Status at Follow-up Visits
| Circumcision Status Comparison | Estimate (95% CI)a | Unadjusted p value | Adjusted Estimatea (95% CI) | Adjusted p value |
|---|---|---|---|---|
| Circumcised vs. uncircumcised | ||||
| Vaginal sex in past 7 days | 0.93 (0.90, 0.96) | < .0001 | 0.94 (0.90, 0.97) | .0005 |
| Condom use with last vaginal sex | 1.16 (1.08, 1.25) | < .0001 | 1.12 (1.04, 1.21) | .003 |
| Uncircumcised vs. all othersb | ||||
| Vaginal sex in past 7 days | 1.08 (1.05, 1.12) | < .0001 | 1.07 (1.04, 1.11) | < .0001 |
| Condom use with last vaginal sex | 0.83 (0.78, 0.89) | < .0001 | 0.86 (0.80, 0.92) | < .0001 |
Note.
For number of vaginal sex acts in past 7 days, this is a mean estimate, and for proportion of condom use during the last vaginal sex, this is a relative risk.
The group of participants with uncircumcised partners compared with the combined group of participants with circumcised partners or unknown circumcision status.
DISCUSSION
Women in this study reported an overall 31% prevalence of male partner circumcision, similar to the estimated global prevalence of circumcision among males aged 15 years and over (30%) reported by WHO (2007). The country-specific rate for South Africa of 33% in this study is similar to the circumcision rate of 35% as estimated by WHO. The rate of 51% for Uganda reported in our study exceeds the 14% circumcision rate estimated by WHO.
WHO did not report on Zimbabwe as the circumcision rate was estimated to be very low. The rate of 10% self-reported by Zimbabwean participants in our study is therefore very encouraging and greater efforts must be made to increase the uptake in a country where there are no cultural, religious, or traditional barriers to male circumcision. Even small gains in rates of uptake of male circumcision will have a huge impact on HIV prevention in both men and women.
The circumcision rates across the three countries are notably different. The data in this study was collected in a period immediately following on the success of three randomized controlled trials (RCTs) in 2006–2007 conducted in South Africa (Orange Farm, Johannesburg), Kenya (Kisumu), and Uganda (Rakai District), which showed a 60% reduction in female to male transmission of HIV-1 (Auvert et al., 2005; Bailey et al., 2007; Gray et al., 2007). The positive outcomes of these RCTs resulted in the WHO and UNIADS recommendations for VMMC program to be implemented in 13 priority countries in Southern and East Africa (Terms, 2007), including South Africa, Uganda, and Zimbabwe, where our study sites were located.
Uganda, although being a late adopter of the VMMC program, is a country where traditional circumcision is a cultural norm and hence the high rates of circumcision in this study (Dickson et al., 2011). The increase in circumcision rates in South Africa could be attributed to South Africa being one of the first countries to adopt and implement VMMC programmes as a public private partnership. This is corroborated by the fact that most of the sites in South Africa were not located in regions where traditional male circumcision forms part of their rituals into manhood. The lower circumcision rate in Zimbabwe is in keeping with WHO in-country estimates during the period of this study (2009 to 2012). However, due to efforts by the Zimbabwean Ministry of Health, circumcision rates increased from 9% in 2010 to 14% in 2015 (Zimbabwe National Statistics Agency & ICF International, 2016).
The study data did not differentiate between partners who may have been traditionally circumcised versus medical circumcision. This could have therefore impacted on the reported partner circumcision rates by women as traditional circumcision results in only partial or incomplete removal of the foreskin (Mark et al., 2012), which could have resulted in misreporting of circumcision cases.
There was no significant difference in HIV incidence between women with circumcised and uncircumcised partners (Tables 2 and 3), which is similar to previous studies that noted no increased risk of male to female transmission of HIV-1 (Baeten et al., 2010; Wawer et al., 2009). A systematic review conducted by Grund et al. in 2017, considered all evidence on associations between male circumcision and women’s health outcomes. The authors found both significant and non-significant protective association for HIV-1 in females whose partner was circumcised, thus resulting in the evidence being classified as low consistency evidence (Grund et al., 2017). Hence, further assessments of the effect of circumcision on male to female transmission of HIV-1 are required.
This study showed no difference in pregnancy incidence between women with circumcised and uncircumcised partners, suggesting that women with circumcised partners did not engage in riskier sexual behaviors than women with uncircumcised partners. While this study showed no difference in gonorrhoea and chlamydia incidence between women with circumcised and uncircumcised partners, the study did report a trend toward a significantly reduced risk of syphilis acquisition in women with circumcised partners. This finding corroborates the results of several other studies that also reported a statistically significant reduced risk of incident syphilis among female partners of circumcised men (Pintye et al., 2014). Additionally, as syphilis is a known co-factor of HIV-1 infection, lower rates of syphilis in circumcised men would result in secondary protection for women against HIV infection. Efforts to promote medical male circumcision should therefore emphasise the protective effect of male circumcision in decreasing incidence of syphilis in both men and their female partners.
Participants with circumcised partners tended to have significantly fewer sex acts in the past 7 days and used condoms more frequently at the last sex act than participants with uncircumcised partners. This important finding highlights the fact that women and their circumcised male partners in this study did not engage in riskier sexual practices due to being circumcised. This finding may also suggest that circumcised partners were possibly more educated about HIV prevention methods and hence were able to reduce their risky sexual behavior.
There were a few limitations in this study. Women in this study reported on their partner’s circumcision status and hence there could have been misclassification bias in reporting on their partner’s circumcision status. Furthermore, women were never asked to distinguish between traditional and medical circumcision practices in their partners. In terms of risk compensation, the fact these women were participating in the VOICE trial, received risk-reduction counseling at every study visit, and therefore, had a greater knowledge of HIV risk reduction must be considered.
CONCLUSION
Our findings indicate that the benefits of male circumcision extended to women in this study. Women with circumcised partners displayed less risky sexual behavior and showed a reduced risk of syphilis acquisition compared to women with uncircumcised partners. While there was no evidence of risk compensation, ongoing education of couples on safe sex practices and the benefits of medical male circumcision is essential to HIV prevention efforts. Our results also confirm a reduction in syphilis incidence in female partners of circumcised men, which has important implications for prevention of syphilis especially in high HIV Incident settings.
RECOMMENDATION
Include key messages on medical male circumcision in HIV prevention efforts to ensure that women understand that circumcised men have only partial protection from HIV acquisition. Empower women to discuss the benefits of medical male circumcision for both men and women with their male sexual partners.
Acknowledgments
The Microbicide Trials Network is funded by the National Institute of Allergy and Infectious Diseases (UM1AI068633, UM1AI068615, UM1AI106707), with co-funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Institute of Mental Health, all components of the U.S. National Institutes of Health.
The authors would like to thank all the women who participated in the VOICE Trial, the VOICE study teams at the South African, Ugandan, and Zimbabwean sites; the VOICE Protocol team and the MTN Publication Committee for assisting with the analysis of data, and for their valuable comments and review of the manuscript.
Contributor Information
Jayajothi Moodley, HIV Prevention Research Unit, South African Medical Research Council, Cape Town, South Africa.; Aurum Institute, Johannesburg, South Africa.
Sarita Naidoo, HIV Prevention Research Unit, South African Medical Research Council, Cape Town, South Africa.; Aurum Institute, Johannesburg, South Africa.
Cliff Kelly, SCHARP (Statistical Center for HIV/AIDS Research and Prevention), Fred Hutchinson Cancer Research Center, Seattle, Washington..
Tarylee Reddy, Biostatistics Unit, South African Medical Research Council..
Gita Ramjee, HIV Prevention Research Unit, South African Medical Research Council, Cape Town, South Africa.; Aurum Institute, Johannesburg, South Africa. London School of Hygiene and Tropical Medicine and the School of Medicine, University of Washington, Seattle, Washington.
REFERENCES
- Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R, & Puren A. (2005). Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: The ANRS 1265 trial. PLoS Medicine, 2, 1112–1122. 10.1371/journal.pmed.0020298 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Baeten JM, Donnell D, Kapiga SH, Ronald A, John-Stewart G, Inambao M, … Celum C. (2010). Male circumcision and risk of male-to-female HIV-1 transmission: A multinational prospective study in African HIV-1-serodiscordant couples. AIDS, 24, 737–744. 10.1097/QAD.0b013e32833616e0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bailey RC, Moses S, Parker CB, Agot K, Maclean I, Krieger JN, … Ndinya-Achola JO (2007). Male circumcision for HIV prevention in young men in Kisumu, Kenya: A randomised controlled trial. Lancet, 369, 643–656. 10.1016/S0140-6736(07)60312-2 [DOI] [PubMed] [Google Scholar]
- Cassell MM, Halperin DT, Shelton JD, & Stanton D. (2006). Risk compensation: The Achilles’ heel of innovations in HIV prevention? BMJ, 332, 605–607. 10.1136/bmj.332.7541.605 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dickson KE, Tran NT, Samuelson JL, Njeuhmeli E, Reed J, Dick B, … Hankins CA (2011). Voluntary medical male circumcision: A framework analysis of policy and program implementation in Eastern and Southern Africa. PLoS Medicine, 8, e1001133. 10.1371/journal.pmed.1001133 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Feuer C. (2010). A cautious nod to the cut … Women weigh in on medical male circumcision. ALQ, (July), 63–68. [Google Scholar]
- Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, Nalugoda F, … Wawer MJ (2007). Male circumcision for HIV prevention in men in Rakai, Uganda: A randomised trial. Lancet, 369, 657–666. 10.1016/S0140-6736(07)60313-4 [DOI] [PubMed] [Google Scholar]
- Grund JM, Bryant TS, Jackson I, Curran K, Bock N, Taliano J, … Davis SM (2017). Association between male circumcision and women’s biomedical health outcomes: A systematic review. Lancet Global Health, 5, E1113–E1122. 10.1016/S2214-109X(17)30369-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Humphries H, van Rooyen H, Knight L, Barnabas R, & Celum C. (2015). ‘If you are circumcised, you are the best’: Understandings and perceptions of voluntary medical male circumcision among men from KwaZulu-Natal, South Africa. Culture, Health and Sexuality, 17, 920–931. 10.1080/13691058.2014.992045 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kalichman S, Mathews C, Kalichman M, Eaton LA, & Nkoko K. (2018). Male circumcision for HIV prevention: Awareness, risk compensation, and risk perceptions among South African women. Global Public Health, 13, 1682–1690. 10.1080/17441692.2018.1427277 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mantell JE, Smit JA, Saffitz JL, Milford C, Mosery N, Mabude Z, … Stein ZA (2013). Medical male circumcision and HIV risk: Perceptions of women in a higher learning institution in KwaZulu-Natal, South Africa. Sexual Health, 10, 112–118. 10.1071/SH12067 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mark D, Middlekoop K, Black S, Roux S, Fleurs L, Wood R, & Bekker L-G (2012). Low acceptability of medical male circumcision as an HIV/AIDS prevention intervention within a South African community that practises traditional circumcision. South African Medical Journal, 102, 571–573. 10.7196/samj.5477 [DOI] [PubMed] [Google Scholar]
- Marrazzo JM, Ramjee G, Richardson BA, Gomez K, Mgodi N, Nair G, … Chirenje ZM (2015). Tenofovir-based preexposure prophylaxis for HIV infection among African women. Obstetrical and Gynecological Survey, 70, 444–446. 10.1097/01.ogx.0000466878.37011.6f [DOI] [PMC free article] [PubMed] [Google Scholar]
- McCoombe SG, & Short RV (2006). Potential HIV-1 target cells in the human penis. AIDS, 20, 1491–1495. [DOI] [PubMed] [Google Scholar]
- Morris BJ, & Wamai RG (2012). Biological basis for the protective effect conferred by male circumcision against HIV infection. International Journal of STD & AIDS, 23, 153–159. 10.1258/ijsa.2011.011228 [DOI] [PubMed] [Google Scholar]
- Pintye J, Baeten JM, Manhart L, Celum C, Ronald A, Mugo N, … Bukusi E. (2014). Decreased incidence of syphilis in both men and women associated with male circumcision: A prospective study among HIV-1 serodiscordant heterosexual African couples. Lancet Global Health, 2, 664–671. 10.1016/S2214-109X(14)70315-8.Decreased [DOI] [PMC free article] [PubMed] [Google Scholar]
- Szabo R, & Short RV (2000). How does male circumcision protect against HIV infection? BMJ, 320, 1592 10.1136/bmj.320.7249.1592 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Terms F. (2007). Roundtable: Male circumcision for HIV prevention: Research implications for policy and programming WHO/UNAIDS Technical Consultation 6–8 March 2007 Conclusions and Recommendations (Excerpts). Reproductive Health Matters, 15, 11–14. 10.1016/S0968-8080(07)29307-3 [DOI] [PubMed] [Google Scholar]
- Wawer MJ, Makumbi F, Kigozi G, Serwadda D, Nalugoda F, Buwembo D, … Gray RH (2009). Randomized trial of male circumcision in HIV-infected men: Effects of HIV transmission to female partners, Rakai, Uganda. Lancet, 374, 229–237. 10.1016/S0140-6736(09)60998-3.Randomized [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weiss HA, Hankins CA, & Dickson K. (2009). Male circumcision and risk of HIV infection in women: A systematic review and meta-analysis. Lancet Infectious Diseases, 9, 669–677. 10.1016/S1473-3099(09)70235-X [DOI] [PubMed] [Google Scholar]
- Women’s HIV Prevention Tracking Project. (2010). Making medical male circumcision work for women. https://www.avac.org/sites/default/files/resource-files/Making%20Medical%20Male%20Circumcision%20Work%20for%20Women.pdf
- World Health Organization. (2007). Male circumcision: Global trends and determinants of prevalence, safety and acceptability. https://apps.who.int/iris/bitstream/handle/10665/43749/9789241596169_eng.pdf;jsessionid=66C75AE3AAD2BC7B6E33C503047D357F?sequence=1
- Zimbabwe National Statistics Agency & ICF International. (2016). Zimbabwe Demographic and Health Survey 2015: Final Report. 10.1017/CBO9781107415324.004 [DOI]