Gender attitudes, sexual violence, and HIV/AIDS risks among men and women in Cape Town, South Africa.

HIV/AIDS is devastating South Africa, with an estimated 1,600 new HIV infections and 600 people dying of an AIDS-related illness each day. The overall HIV prevalence in South Africa is 11%, with HIV prevalence highest in townships and informal urban settlements, where as many as one in five people may be HIV-positive (Shisana & Simbayi, 2002). There is an urgent need for HIV-prevention interventions in South Africa, and the most effective interventions will be informed by careful analysis of social and behavioral factors associated with increased risks for HIV infection. The South African Human Sciences Research Council estimates that 9.5% of South African men are HIV-positive, while almost 13% of the country's women are believed to be infected (Shisana & Simbayi). As is the case in most of sub-Saharan Africa, women greatly surpass men in the number of people living with HIV/AIDS, and in many areas women double the number of men with the virus (World Health Organization/UNAIDS, 2002). Reasons for women's increased risks include biological and social risk factors (Maman, Campbell, Sweat, & Gielen, 2000). Among the factors that are most important in understanding HIV risks in South Africa are those associated with gender and power in relationships.

Gender-power differences are linked to an array of factors that increase risks for sexually transmitted infections (STI), including HIV infection. Studies conducted in the U.S. show that women in violent and abusive relationships are less likely to use condoms, more likely to incur abuse as a result of requesting condoms, and more likely to contract an STI than women who have not been in violent relationships (Kalichman, Williams, Cherry, Belcher, & Nachimson, 1998; Wingood & DiClemente, 1997). Associations between violence, particularly sexual assault, and risks for HIV/STI have also been observed in southern Africa (Ajuon et al., 2002; Garcia-Moreno & Watts, 2000). Research conducted in Kigali, Rwanda, for example, found that HIV-positive women experienced more unwanted sexual contact than HIV-negative women (Van der Straten, King, & Grinstead, 1998). Culturally-sanctioned gender roles foster power imbalances that facilitate women's risks for both sexual assault and STI/HIV (Farmer, Conners, & Simons, 1996; Jewkes, Penn-Kekana, Levin, Ratsaka, & Schreiber, 2001; Morrell, 2002; Pitcher & Bowley, 2002). When gender-power imbalances place women in subordinate roles, women can have few options for exercising personal control in their sexual relationships (Ajuwon et al., 2002; Jewkes & Abrahams, 2002; Wojcicki & Malala, 2001).

The association between control in sexual relationships and STI/HIV risks was supported by research conducted in a township in Cape Town (Kalichman & Simbayi, 2004). Women who had been sexually coerced (a total of 40% of women surveyed) were significantly more likely to have exchanged sex to meet survival needs and to have had multiple male sex partners, greater rates of unprotected vaginal intercourse, lower rates of condom use, more sexual contacts involving blood, and greater rates of STI. Women who had been sexually assaulted were also more likely to have been physically (non-sexually) abused by relationship partners and were more likely to be afraid to ask sex partners to use condoms.

Cross-cultural research suggests that South African men often hold strong traditional gender beliefs (Glick et al., 2000). Consequentially, South African women may experience multiple barriers in their efforts to reduce their risks for HIV infection, including risks for sexual violence in gender-power imbalanced relationships. The potential for women to reduce their risks for STI/HIV is, therefore, seriously limited by socially-constructed gender roles and sexual scripts. Women who suggest using condoms with a resistant sex partner may experience adverse consequences, including threatening the masculinity of their partner, raising partner suspicions about their monogamy or sexual histories, finding themselves vulnerable to further violence, being rejected, and potentially losing their partner's financial support (Ackermann & de Klerk, 2002; Maman et al., 2000; Wood & Jewkes, 2001; Wood, Maforah, & Jewkes, 1998). A study conducted by the International Center for Research on Women found that South African women do not initiate discussions about safer sex or tell their partners to use condoms because it is culturally inappropriate and also because it brings their own sexual behavior into question (Hadden, 1997).

This study used a social constructionist approach to studying gender. We conceptualized gender as a system of social classification that influences access to power, status, and material resources (Crawford, 1995; Crawford & Unger, 2004). All human societies make social distinctions based on gender, and virtually all allocate more power and higher status to men. Only recently have perceptions of gender and gender roles been examined in the context of HIV risks among men and women in southern Africa (Kalichman & Simbayi, 2004). This study examined socially-constructed gender roles at the individual level among men and women receiving STI clinic services in Cape Town, South Africa. Previous research suggested that South African men may hold negative attitudes toward women, including attitudes that may promote sexual violence, such as the acceptance of rape myths (Glick et al., 2000; Jewkes et al., 2001; Jewkes & Abrahams, 2002). Although rape myth acceptance and rape-supportive attitudes are related to sex role stereotyping, adversarial sexual beliefs, and a greater acceptance of relationship violence (Hinck & Thomas, 1999), less is known about gender attitudes in relation to STI/HIV risks. We therefore examined gender attitudes and sexual-violence supportive beliefs (i.e., acceptance of culturally-defined rape myths) in a sample of South African men and women at high risk for HIV transmission. We hypothesized that gender attitudes and rape myth acceptance would be related to greater risks for HIV infection in both men and women.

METHOD

Participants and Setting

Participants were 415 men and 127 women receiving services from an STI clinic in Cape Town, South Africa. In terms of demographic characteristics, 47 (9%) were 20 years old or younger, 159 (30%) were between 21 and 25 years old, 276 (52%) were between 26 and 35, and the remaining 53 (10%) were over age 35. In addition, 521 (95%) participants identified as Black, and 7 (1%) were White, 2 (4%) Indian, and 11 (2%) Colored or of mixed race. Nearly half of the sample had not matriculated high school (N = 235, 43%), 42% (N = 226) were unemployed, and 16% (N = 85) indicated that they were married.

The public STI clinic that served as the site for this research treats over 1,800 patients with STIs per month. Approximately half of all patients seen at the clinic have previously received STI services. In 2003, the clinic conducted over 5,000 HIV antibody tests, of which 25% were HIV-positive.

Measures

Measures were administered in English and Xhosa (an indigenous African language), the two languages spoken by nearly all clinic patients. Measures included demographic characteristics, history of sexual violence experiences, attitudes and beliefs pertaining to gender and sexual assault, sexual behaviors and STI risks, substance use, and AIDS-related knowledge.

Demographic characteristics. Participants reported their age, race, years of formal education, employment status, and marital status.

History of sexually violent experiences. To assess participants' history of having experienced sexual violence, we adapted items from the Sexual Experiences Survey, a widely used measure of sexually coercive experiences (Koss & Gidycz, 1985; Koss, Gidycz, & Wisniewski, 1987). Participants were asked to indicate "Yes" or "No" to three questions regarding having perpetrated sexual assault: "Have you ever had sexual intercourse with a person when he or she really did not want to?"; "Have you ever threatened to use force to get a person to have sex when he or she did not want to?"; and "Have you ever forced or pressured a person to have sex when he or she did not want to?" In addition, three items measured history of having been sexually assaulted: "Have you ever had sexual intercourse against your will because a person threatened to use physical force to make you?", "Has someone ever forced you to have sex when you did not want to?", and "Has someone ever pressured you to have sex when you did not want to?". All sexual violence history questions were responded to by men and women as either "Yes" or "No."

Gender attitudes. Participants responded to seven items adapted from the Attitudes Toward Women Scale (Spence & Helmreich, 1972). The items selected and adapted for this study represented beliefs about traditional gender roles and gender relations. The items correspond with gender themes identified in qualitative research with South African men, such as defining masculinity in terms of number of sex partners, the sexual desirability of one's partners to other men, and the ability to "control" girlfriends (Wood & Jewkes, 2001). The results section presents the items, which were responded to on 4-point scales from 1 (Disagree) to 4 (Agree). Reliability analyses indicated that the scale showed slight heterogeneity among items, alpha = .63.

Rape myth acceptance. The rape myth acceptance measure consisted of six items adapted from the Rape Myth Acceptance Scale (Burt, 1980). Items used in this study are presented in the results section. The rape myth acceptance items have been used in international research (e.g., Nayak, Byrne, Martin, & Abraham, 2003), and they reflect beliefs revealed in qualitative research with South African men (Wood & Jewkes, 2001). Items on the rape myth measure were responded to on 4-point scales, from 1 (Disagree) to 4 (Agree), and the instrument was marginally internally consistent, alpha = .70.

Sexual behaviors. To assess current sexual behaviors, participants were asked how many male and female sex partners they had in the past three months. In addition, participants reported the number of times that they engaged in vaginal and anal intercourse with and without condoms during that time period. Participants also indicated whether they had engaged in sexual activity in the past three months that involved direct contact with blood. Because sexual contact during menstruation may increase a risk of HIV infection for men and women (Lazzarin, Saracco, Musicco, & Nicolosi, 1991), as may genital bleeding during intercourse (Seidlin, Vogler, Lee, Lee, & Dubin, 1993), we thought it was appropriate to include questions relating to sexual activity involving direct blood contact. We collected responses on a range representing the number of partners and number of acts occurring in the preceding three months. In addition, we asked participants if they had ever exchanged sex for money or a place to stay and whether they had ever had a sex partner they believed had injected illegal drugs.

Participants also reported whether they had ever been diagnosed with an STI. History of genital ulcers was assessed by asking participants whether they had ever experienced an open sore on their genitals. We also asked whether participants were currently experiencing STI symptoms.

Substance use. To assess lifetime history of substance use, participants indicated whether they had ever used alcohol, dagga (marijuana), or mandrax (a sedative).

AIDS-related knowledge. An 11-item test assessed HIV risk and prevention-related knowledge. Items were adapted from a measure reported by Carey and Schroder (2002) and previously used in South African research (Kalichman & Simbayi, 2003). The items reflected information about HIV transmission, condom use, and AIDS-related knowledge. Participants responded to each item by checking "Yes," "No," or "Don't Know." The AIDS knowledge test was scored for the number of correct responses, with "Don't Know" responses scored as incorrect; the proportion of correct responses was calculated. The AIDS knowledge test demonstrated heterogeneity among items, alpha = .65, suggesting a representation of multiple aspects of AIDS knowledge.

Procedure

Participants were recruited to complete an anonymous survey of sexual behavior and health. Potential participants were referred to the study recruiter by their doctor or nurse following their clinical services; 90% of patients referred to the study agreed to complete surveys. Sampling occurred during all hours of clinic operation over a three-month period. We limited participant enrollment to no more than 12 individuals per day to assure a range of participants recruited over time. All persons who agreed to complete the survey were able to self-administer the measures with minimal assistance. Participants received 15 South African Rand ($3 U.S.) as compensation.

Data Analysis

To examine the associations of gender attitudes and STI/HIV risks, we created a composite to represent sexual risks among men and women. Risk behaviors included in the composite index were having two or more sex partners in the previous three months; engaging in sexual intercourse that involved blood contact in the previous three months; lifetime history of sexually transmitted infection; lifetime history of genital ulcers; having exchanged sex for money or a place to stay; and having had an injection drug-using sex partner. The six behavioral risk factors were summed to form an unweighted risk composite that ranged from 0 to 6. Our initial analyses, therefore, describe levels of HIV risks among participants. The risk composite index demonstrated significant divergence from normality (Mean = 2.1, SD = 1.1, Skewness = 0.8, Standard Error = 0.1) and was therefore transformed using the formula [log.sub.10](X + 1), as recommended by Winer (1971) and Tabachnik and Fidell (1996). The distribution of transformed index values was no longer skewed (Mean = 0.2, SD = 0.2, Skewness = 0, Standard Error = 0.1).

We conducted initial analyses to describe participants' HIV risk factors, sexual assault history, and endorsement of gender attitudes and rape myth acceptance items. Men and women were compared on each of the measures using logistic regressions adjusted for demographic characteristics. Logistic regression was selected for these analyses because of the dichotomous dependant variable (e.g., gender) in relation to a set of categorical and continuous predictor variables. Odds ratios adjusted for age, education, and marital status are reported with associated 95% confidence intervals.

We performed further analyses to test the associations between the HIV risk composite and gender attitudes, rape myth acceptance, and sexual violence history, as well as commonly studied HIV risk-related factors (e.g., substance use and MDS-related knowledge). To examine associations among variables, we conducted bivariate correlations for transformed risk composite scores, gender attitudes, rape myth acceptance, substance use, history of sexual violence, and AIDS-related knowledge. To test multivariate associations, we performed linear regressions with transformed risk composite index scores entered as the continuous dependent variable and gender attitudes, rape myth acceptance, substance use, history of sexual violence, and AIDS-related knowledge entered as independent variables. Analyses were performed separately for men and women; for all analyses, significance was defined as p < .05.

RESULTS

HIV risk factors were prevalent among both men and women STI clients. For men, risk was most commonly indicated by having multiple sex partners, a history of genital ulcer disease, and recently engaging in sexual intercourse that involved genital bleeding. For women, risks were also indicated by genital ulcers and sexual contact involving blood, as well as having had a sexual partner believed to have been an injection drug user. Analysis of the six-indicator risk composite showed that men were significantly more likely to report multiple HIV risks than women; 32% of men indicated three or more risk factors, compared to 11% of women (see Table 1).

Sexually Violent Experiences

Sexual violence was reported by men and women as indicated by endorsing any of the sexual violence items; over 16% of men and 40% of women indicated that they had been victims of sexual assault as adults. As shown in Table 2, 10% of men and 29% of women had been forced to have sexual intercourse. Women frequently reported multiple sexual assault experiences, with 13% of women experiencing at least two different types of sexual assault. In terms of having perpetrated sexual assault against others, 23% of men endorsed at least one of the sexual assault items, indicating that they had sexually assaulted a woman. In addition, 16% of women endorsed at least one item reporting that they had perpetrated sexual assault.

Gender Attitudes and Rape Myth Acceptance

Both men and women endorsed gender attitudes that represent traditional, submissive, and passive roles of women, with nearly all men and women stating that women should obey their husbands. Two out of three men and women agreed that there are many jobs that men can do much better than women, and one in three participants stated that women should not talk to men about sex (see Table 3). Although there were select gender differences on gender attitude items, there was a general pattern of response rates for men and women being more alike than different.

We observed a similar pattern of results for the rape myth acceptance items, with as many as one in five participants across genders agreeing that rape usually results from things that a woman says or does (27% across genders), that some cases of rape involve a woman who wants to have sex (18% across genders), and that rape is often a woman's fault (29% across genders). There were, again, few differences between men and women in terms of acceptance of rape myths (see Table 3).

Factors Associated with Sexual Risks for HIV Infection

Bivariate correlations among the transformed HIV risk composite index, gender attitudes, rape myth acceptance, sexual assault variables, substance use, and AIDS knowledge scores for men and women are presented in Table 4. For men, the HIV risk composite was significantly related to use of multiple substances, histories of having perpetrated sexual assault and having been sexually assaulted, and rape myth acceptance scores. Among women, HIV risk composite scores were significantly correlated with alcohol use and having been sexually assaulted. These results supported conducting multiple regressions on the HIV risk composite for both men and women.

We performed multiple regression analyses with the transformed HIV risk composite index entered as the dependent variable and the following measures entered as predictor variables: participant education, AIDS knowledge test scores, alcohol, mandrax and dagga use, sexually violent experiences (both perpetrated and victimized), gender attitudes, and rape myth acceptance. For men, the regression analysis showed that education, AIDS knowledge, alcohol and mandrax use, having perpetrated sexual assault, and rape myth acceptance significantly predicted HIV risk factors, F (9, 367) = 11.7, p < .01, adjusted [R.sup.2] =.16 (see Table 5).

Among women, the regression analysis also resulted in a significant multivariate predictive model of HIV risk composite scores, F (9, 118) = 2.0, p < .05, adjusted [R.sup.2] = .07. Only lower education and alcohol use were significantly associated with HIV risk composite scores when all other variables in the model were included (see Table 5).

DISCUSSION

Sexual risk behaviors observed in this study occurred in the context of co-factors for HIV transmission, particularly substance use and sexual violence. For both male and female STI clinic clients, it was common to have a history of sexual contacts involving blood and to report multiple sex partners. A majority of men and one third of women reported at least two of the risk factors included in our HIV risk composite. In addition, we found that substance use was associated with HIV risk behaviors, even after other risk-related factors were taken into account. Given that the estimated HIV prevalence in the clinic where we conducted the study is 25%, the potential for HIV transmission in our sample is alarmingly high.

Previous studies have shown that sexual coercion and sexual assault occur at astounding rates in South Africa (e.g., Buga, Amoko, & Ncyiyana, 1996; Wood et al., 1998). Unfortunately, this finding was repeated in this study, where more than 40% of women reported at least one sexual assault experience and more than one in five men openly admitted to having perpetrated sexual assaults against women. Men were also victims of sexual assault, a finding that has been reported in the U.S. (Struckman Johnson & Struckman-Johnson, 1994). These findings may suggest more generalized and pervasive sexual assault across gender lines in some South African communities. However, our data do not provide insight into the situations in which men are the victims of sexual assault and women are the perpetrators.

With regard to gender roles, we expected and found that men often viewed women as passive, subservient, and as fulfilling traditional gender roles. This finding is consistent with previous research that suggests South African women are expected to fulfill a stereotypical female gender role by being docile, especially in sexual relationships (e.g., Ackermann & de Klerk, 2002; Maman et al., 2000; Wood & Jewkes, 2001; Wood et al., 1998). However, we found that women endorsed these attitudes at rates that generally did not differ from men. Although representing a minority of men and women, a similar pattern of results occurred for responses to the rape myth acceptance items. As many as one in five participants across genders agreed that rape usually occurs as a result of the actions of a woman and that she can often be blamed for it. These results indicate a generalized view of women as subordinate to men and suggest the importance of looking at the gender-power differentials in South African relationships, particularly in sexual relationships where a lack of power among women could lead to increased risks for HIV.

For men, our multiple regression analyses showed that sexual assault and rape myth acceptance, along with substance use, were significantly related to cumulative risks for HIV transmission, pointing toward several avenues for risk-reduction interventions with men. Information and education campaigns are currently underway in South Africa to raise awareness of the link between sexual violence and STI/HIV, combining these issues under a common rubric of sexual health. Behavioral interventions that target men for STI/HIV risk reduction are also becoming more common in South Africa (Morrell, 2002). Given the independent association of substance use to sexual risks, efforts to prevent HIV transmission by focusing on men may be most efficient when targeting men who drink alcohol and use other drugs. Because men often control much of what happens in the sexual relationships, significant advances in reducing STI/HIV risks in South Africa may result from targeting men.

In contrast, our findings for women suggest few inroads for STI/HIV prevention by targeting women. Although we found that women were at substantial risk for STI/HIV, women's risks were only associated with lower education and history of alcohol use in this study. Women endorsed the view that women are subordinate and passive in their relationships with men and that women are often to blame for rape. Nevertheless, these attitudes were not independently associated with STI/HIV risks. We speculate that women's risks for STI/HIV are the product of partner characteristics in male-dominated relationships. Women who are victims of, or threatened with, violence are often unable to negotiate life-saving strategies in HIV prevention (Heise & Elias, 1995), and so the design and implementation of risk-reduction interventions focused on men may prove successful in ultimately reducing HIV risks for women.

These findings are limited by the study measures relying on self-reports of sensitive and private behaviors. In particular, results for admitting to having sexually violated a person or having been sexually assaulted must be considered conservative, as it is likely that these behaviors were under-reported. It is worth noting that the relationship between sexual coercion and risk of STI/HIV could be more prominent than is indicated by the current data. In addition, self-reports of substance use and sexual behaviors should be considered lower-bound estimates. Although our measures of gender attitudes and rape myths have been used in international research, and the items we selected were based on previous qualitative studies of men and women in South Africa, the scales themselves showed item heterogeneity, as indicated by low to moderate internal consistency and have not yet been validated in the South African context. Also, the amount of variance accounted for in our analyses was small, leaving much of the variance unexplained. Our study is also limited by its focus on STI clinic patients, reducing the generalizability of the findings to other populations. Our focus on STI clinic patients likely biased the study results by over-representing men and women who engage in considerable risk behaviors and are at some of the highest risks for HIV infection. Further research is needed to confirm these results before allowing for generalizability. Nevertheless, it is clear from our results that gender-tailored STI/HIV risk reduction interventions are urgently needed in South Africa.

Manuscript accepted May 17, 2005

REFERENCES

Ackermann, L., & de Klerk, G. W. (2002). Social factors that make South African women vulnerable to HIV infection. Health Care for Women International, 23, 163-172.

Ajuon, A., McFarland, W., Hudes, E., Adepado, S., Okikiolu, T., & Lurie, P. (2002). HIV risk-related behavior, sexual coercion, and implications for prevention strategies among female apprentice tailors, Ibadan, Nigeria. AIDS and Behavior, 6, 229-235.

Buga, G., Amoko, D., & Ncayiyana, D. (1996). Sexual behavior, contraceptive practice and reproductive health among school adolescents in rural Transkei. South African Medical Journal, 86, 523-527.

Burt, M. (1980). Cultural myths and support for rape. Journal of Personality and Social Psychology, 38, 217-230.

Carey, M. P., & Schroder, K. E. (2002). Development and psychometric evaluation of the brief HIV knowledge questionnaire (HIV-KQ-18). AIDS Education and Prevention, 14, 174-184.

Crawford, M. (1995). Talking difference: On gender and language. London: Sage.

Crawford, M., & Unger, R. (2004). Women and gender: A feminist psychology (4th ed.). Boston: McGraw Hill.

Farmer, P., Conners, M., & Simmons, J. (1996). Women, poverty, and AIDS: Sex, drugs, and structural violence. Monroe, ME: Common Courage Press.

Garcia-Moreno, C., & Watts, C. (2000). Violence against women: Its importance for HIV/AIDS. AIDS, 17 (Suppl 3), $253-$265.

Glick, P., Fiske, S., Mladinic, A., Saiz, J., Abrams, D., Masser, B.. et al. (2000). Beyond prejudice as simple antipathy: Hostile and benevolent sexism across cultures. Journal of Personality and Social Psychology, 79(5) 763-775.

Hadden, B. R. (1997). An HIV/AIDS prevention intervention with female and male STD patients in a peri-urban settlement in KwaZulu Natal, South Africa. Washington, DC: International Center for Research on Women.

Heise, L. L., & Elias, C. (1995). Transforming AIDS prevention to meet women's needs: A focus on developing countries. Social Science and Medicine, 40(7), 931-943.

Hinck, S. S., & Thomas, R. (1999). Rape myth acceptance in college students: How far have we come? Sex Roles, 40, 815-832.

Jewkes, R., & Abrahams, N. (2002). The epidemiology of rape and sexual coercion in South Africa: An overview. Social Science and Medicine, 55, 1,231-1,244.

Jewkes, R., Penn-Kekana, L., Levin, J., Ratsaka, M., & Schreiber, M. (2001). Prevalence of emotional, physical, and sexual abuse of women in three South African Provinces. South African Medical Journal, 91, 421-428.

Kalichman, S. C., Williams, E., Cherry, C., Belcher, L., & Nachimson, D. (1998). Sexual coercion, domestic violence, and negotiating condom use among low-income African-American women. Journal of Women's Health, 7, 371-378.

Kalichman, S. C., & Simbayi, L. (2003). HIV testing attitudes, AIDS stigma and voluntary HIV counseling and testing in a Black Township in Cape Town, South Africa. Sexually Transmitted Infections, 79, 442-447.

Kalichman, S. C., & Simbayi, L. (2004). Sexual assault history and risks for sexually transmitted infections among women in an African township in Cape Town, South Africa. AIDS Care, 16, 681-689.

Koss, M. P., & Gidycz, C. A. (1985). Sexual experiences survey: Reliability and validity. Journal of Consulting and Clinical Psychology, 53, 422-423.

Koss, M. P., Gidycz, C .A., & Wisniewski, N. (1987). The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. Journal of Consulting and Clinical Psychology, 55, 162-170.

Lazzarin, A., Saracco, A., Musicco, M., & Nicolosi, A. (1991). Man-to-woman sexual transmission of the human immunodeficiency virus: Risk factors related to sexual behavior, man's infectiousness, and woman's susceptibility. Archives of Internal Medicine, 151, 2,411-2,416.

Maman, S., Campbell, J., Sweat, M. D., & Gielen, A. C. (2000). The intersections of HIV and violence: Directions for future research and interventions. Social Science and Medicine, 50, 459-478.

Morrell, R. (2002). Changing men in South Africa. University of Natal Press, Durban, South Africa.

Nayak, M., Byrne, C., Martin, M., & Abraham, A. (2003). Attitudes toward violence against women: A cross-nation study. Sex Roles, 49, 333-342.

Pitcher, G., & Bowley, M. (2002). Infant rape in South Africa. Lancet, 359, 274-275.

Seidlin, M., Vogler, M., Lee, E., Lee, Y. S., & Dubin, N. (1993). Heterosexual transmission of HIV in a cohort of couples in New York City. AIDS, 7, 1,247-1,254.

Shisana, O., & Simbayi, L. (2002). Nelson Mandela/HSRC Study of HIV/AIDS: South African National HIV Prevalence, Behavioral Risks and Mass Media, Household Survey 2002. Cape Town, South Africa: Human Sciences Research Council.

Spence, J. T., & Helmreich, R. (1972). The attitudes toward women scale: An objective instrument to measure attitudes toward the rights and roles of women in contemporary society. Catalog of Selected Documents in Psychology (Spr.), 66-67.

Struckman-Johnson, C., & Struckman-Johnson, D. (1994). Men pressured and forced into sexual experiences. Archives of Sexual Behavior, 23, 93-114.

Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd edition). New York: Harper Collins.

Van der Straten, A., King, R., & Grinstead, O. (1998). Interpersonal violence and HIV infection in women in steady relationships in Kigali, Rwanda. AIDS & Behavior, 2, 61-73.

Winer, B. J. (1971). Statistical principles in experimental design. New York: McGraw Hill.

Wingood, G. A., & DiClemente, R. J. (1997). The effects of an abusive primary partner on the condom use and sexual negotiation practices of African American women. American Journal of Public Health, 87, 1,016-1,018.

Wojcicki, J., & Malala, J. (2001). Condom use, power, and HIV/AIDS risk: Sex workers bargain for survival in Hillbrow/Joubert/Park/Bera, Johannesburg. Social Science and Medicine, 53, 99-121.

Wood, K., & Jewkes, R. (2001). "Dangerous" love: Reflection on violence among Xhosa township youth. In R. Morrell (Ed.), Changing men in southern Africa (pp. 317-336). New York: Zed Books Ltd.

Wood, K., Maforah, F., & Jewkes, R. (1998). "He forced me to love him": Putting violence on adolescent sexual health agendas. Social Science and Medicine, 47(2), 233-242.

World Health Organization/UNAIDS. (2002). AIDS Epidemic Update. World Health Organization.

Seth C. Kalichman

University of Connecticut

Leickness C. Simbayi

Human Sciences Research Council, Cape Town, South Africa

Michelle Kaufman, Demetria Cain, and Chauncey Cherry

University of Connecticut

Sean Jooste

Human Sciences Research Council, Cape Town, South Africa

Vuyisile Mathiti

University of the Western Cape, South Africa

The National Institute of Alcohol Abuse and Alcoholism (NIAAA) Grant R21-AA14820, National Institute of Mental Health (NIMH) Grant R01-MH071160, and World AIDS Foundation Grant 278 supported this research.

Address correspondence to Seth C. Kalichman, Department of Psychology, 406 Babbidge Road, University of Connecticut, Storrs, CT, 06269; e-mail: [e-mail].

Table 1. Sexual Risk Factors for HIV Transmission Among Men
and Women Receiving STI Clinic Services in Cape Town, South Africa

Risk Factor                    Men            Women      Adjusted
                            (N = 415)       (N = 127)      OR
                            N       %       N       %
Two or more
  sex partners (a)         199     49      23      19     4.1 (c)
Sexual exposure
  to blood (a)             123     30      41      33     0.9
History of STI (b)         304     74      58      46     3.1 (c)
Genital ulcers (b)         124     30      20      16     2.3 (c)
Exchanged money OR
  material goods
  for sex (b)              41      10      12       9     1.1
Injection drug use or
  injection drug using
  sex partner (b)          23       6       5      18     1.4
Total risk factors
  No risk factors          37       9      35      28
    1 risk factor         127      31      44      35
    2 risk factors        119      29      34      27
    3 risk factors         92      22      10       8
    4 risk factors         27       7       3       2
    5 risk factors         13       3       1       1
Mean (SD)                 2.0     1.2     1.3     1.1     1.8

Note. Adjusted OR = odds ratio controlling for age,
education, and marital status.

(a) rate for the past 3 months.

(b) lifetime history.

(c) p < .01.

Table 2. Sexual Violence Among Men and Women Receiving
STI Clinic Services in Cape Town, South Africa

                                   Men          Women
Sexual Assault                  (N = 415)     (N = 127)    Adjusted
                                N      %      N      %        OR
Victim of assault
  Threatened to use
    force for sex               18     4      23     18      4.4 (a)
  Pressured to have sex         19     5      38     30      3.5 (a)
  Forced to have sex            45     10     37     29      8.2 (a)
  Any one of the 3 victim
    experiences                 63     16     52     41      3.7 (a)
  Perpetrated assault
  Had sex with an
    unwilling person            67     16     15     11      0.6
  Threatened to use
    force for sex               41     10     8      6       0.6
  Forced someone to have sex    55     16     7      6       0.4 (a)
  Any one of 3 perpetrator
    experiences                 93     23     21     16      0.6

Note. Adjusted OR = odds ratio controlling for age, education, and
marital status.

(a) p < .01.

Table 3. Gender Attitudes and Rape Myth Acceptance
Among Male and Female STI Clinic Patients

                                    Men        Women
                                  N = 415)   (N = 127)    Adjusted
Attitudes and beliefs             N     %     N     %        OR

Gender Attitudes
  Women should not talk
    about sex with men.          147   36    37    30       0.9
  Women should obey
    their husbands.              386   94    110   89       1.3 (a)
  Women should only
    be concerned about
    being a wife and mother.     256   63    69    55       1.1
  Women and men should
    have the same rights.        316   76    112   90       1.4 (b)
  Women should not expect to
    have the same
    freedom as men.              198   48    58    46       0.9
  Mothers should
    have most of the
    responsibility in caring
    for children.                160   39    49    39       1.0
  There are many
    jobs that men can
    do much better than women.   288   70    78    62       1.1

Rape Myth Acceptance
  When a women is raped it is
    usually because of
    things she said or did.      98    24    21    17       1.2
  In some cases of rape,
    women actually
    want the sex.                84    21    13    10       1.3 (a)
  Rape does not happen
    as often as people say.      162   40    40    32       1.2
  It is human nature for men
    to need sex and take it
    from women.                  129   32    29    23       1.1
  Rape mainly happens when
    a woman has sent a
    man sexual signals.          164   40    34    27       1.2 (a)
  Rape is often
    the woman's fault.           119   29    38    30       0.9

Note. Adjusted OR = odds ratio adjusted for
age, education, marital status.

(a) p < .05.

(b) p < .01.

Table 4. Correlation Coefficients Among Participant
Characteristics and Risk Composite Scores for Men
(panel above the diagonal) and Women (panel below diagonal)

                      HIV-AIDS     Alcohol        Mandrax
                      knowledge

HIV-AIDS knowledge       --          .06            .03
Alcohol use              .13          --            .02
Mandrax use              -.08        -.04           --
Dagga use                .03          .24 (b)      -.03
Sexually assaultive      .06         -.03           .11
Sexually assaulted       .12          .15           .02
Rape myths              -.30 (b)     -.05           .10
Gender attitudes        -.22 (b)     -.14          -.05
Risk composite (c)       .05          .29 (b)       .07

                        Dagga      Sexually      Sexually
                                  assaultive     assaulted

HIV-AIDS knowledge       .05         .01            -.10
Alcohol use              .21         .08            -.01
Mandrax use              .35         .13 (a)         .08
Dagga use                --          .23 (b)         .14 (a)
Sexually assaultive      .11          --             .30 (b)
Sexually assaulted       .19 (a)     .32 (b)          --
Rape myths               .01         .09             .04
Gender attitudes        -.22 (b)    -.14            -.07
Risk composite (c)       .09         .07             .16 (b)

                        Rape        Gender         Risk
                        myths     attitudes    composite (c)

HIV-AIDS knowledge     -.21 (b)     -.32 (b)       .01
Alcohol use            -.08          .09           .21 (b)
Mandrax use             .04         -.02           .17 (b)
Dagga use               .04          .02           .23 (b)
Sexually assaultive     .07          .02           .28 (b)
Sexually assaulted      .21 (b)      .10 (a)       .26 (b)
Rape myths               --          .38 (b)       .19 (b)
Gender attitudes        .56 (b)        --          .05
Risk composite (c)      .01          .01            --

(a) = p <.01.

(b) = p <.05.

(c) Transformed composite score using
formula [log.sub.10] (X + 1).

Table 5. Results of Multiple Regression Analyses Predicting
HIV Risk Composite Scores Among Male and Female STI Clinic
Patients

                                    Men                   Women
Variable                        Standardized           Standardized
                               Beta       t          Beta        t

Education                      -.05     -1.0        -.21       -2.1 (a)
AIDS Knowledge scores           .06      1.3         .06        0.6
Alcohol use                     .23      5.3 (b)     .31        3.2 (b)
Mandrax use                     .08      1.9         .11        1.2
Dagga use                       .11      2.4 (b)    -.01       -0.1
Sexually assaultive history     .15      3.5 (b)     .05        0.6
Sexually assaulted history      .08      1.8         .10        1.0
Gender attitudes               -.05     -1.0         .03        0.2
Rape myth acceptance            .12      2.7 (b)     .06        0.6

(a) p < .05.

(b) p < .05.