Summary: Among Kenyan men recruited as sex partners of women with genital symptoms, 22 of 150 were HIV seropositive. Because male HIV infection and male hygiene were unexpectedly found to be associated with each other, we examined the relationship of 5 hygiene variables with HIV infection in the men in a principal components analysis, controlling for socioeconomic status and other potential confounders. By multivariate analyses, HIV infection in men was not only independently associated with previous illness (odds ratio [OR], 5.1; 95% confidence interval [CI], 1.4-19.1) and inversely associated with being circumcised (OR, 0.12; 95% CI, 0.02-0.91), but also independently associated with a combined measure of hygiene (OR, 0.41; 95% CI, 0.19-0.90).
From the Departments of *Laboratory Medicine, †Gynecology, and ‡Medicine, and the §Center for AIDS and STD, University of Washington, Seattle; the ∥Program in Biostatistics, Fred Hutchinson Cancer Research Center, Seattle; the ¶Center for Microbiology Research, Kenya Medical Research Institute, Nairobi, Kenya; and the #Department of Obstetrics, Gynecology, and Reproductive Science, University of California, San Francisco.
Received for publication November 23, 2005; accepted April 18, 2006.
Supported by the University of Washington Center for AIDS Research (AI 27757), STI-Topical Microbicide Cooperative Research Center (AI 31448), and AIDS International Research & Training Program (NIH FIC D43 TW00007).
Reprints: King K. Holmes, MD, PhD, Harborview Medical Center, Box 359931, 325 9th Ave, Seattle, WA 98104 (e-mail: email@example.com).
In assessing genital hygiene beliefs and practices in Kenyan men1 and the relationship of such practices to bacterial vaginosis in their female partners,2 we discovered an unexpected inverse association of hygiene with prevalence of HIV infection in men. This report uses principal components analyses3 to further explore the relationship of hygiene to HIV infection in these men.
As described in detail in other works,1,2 we enrolled 219 women, aged 18 and 45 years, with lower genital tract symptoms from a family planning clinic and an STD clinic in Nairobi, Kenya. Only women who thought their partners would participate were enrolled, and were asked to come back with them to the clinic or to have the partner return to the clinic with the woman's study card. One male partner for each of 175 female participants was recruited, whereas the partners of 44 women did not come, or came but declined participation. The men all completed standardized face-to-face interviews; 150 men also agreed to undergo testing for HIV-1, syphilis, trichomonas infection, and gonorrhea seropositivity, and represent the subjects for the present study.
Analyses were performed using the software SPSS 11.5 for Windows (SPSS, Inc, Chicago, IL). We examined the socioeconomic status (SES) and hygiene practices (HYG) of the participants using principal components analyses.3 The SES principal components were computed as linear combinations of years of education, marital status, employment status, number of rooms and people living in the house, and amenities, including indoor plumbing, electricity, television, and cement flooring. Hygiene components were computed from bathing frequency and duration, bathing immediately after sex, specific bathing of genitalia, amount of water used, and frequency of change of underwear (Bukusi et al2 provides a more complete description of how we applied principal components analysis). The following components resulted: SES1, SES2, SES3, SES4, and HYG1 and HYG2. Univariate and multivariate logistic regression analyses were performed to assess potential risk factors for HIV-1 infection. Risk factors associated with HIV-1 infection at P < 0.2 in univariate regression were included in multivariate analysis, and backward elimination was used to obtain only significant associations (P < 0.05).
The men were aged an average of 33 years (SD, 7); 84% were married. Their median lifetime number of sex partners was 6 (range, 1-10). The number of completed years of schooling averaged 10 (SD, 3). Of 148 men examined, 97% were circumcised. Three were syphilis seropositive, 6 had gonorrhea, none had trichomonas infection, and 22 of the 150 who consented to HIV testing were HIV-1 seropositive.
By univariate analysis, the factors significantly associated with HIV-1 infection in the men included syphilis seropositivity, SES component 1, and hygiene component 2. Factors marginally associated with HIV-1 infection at 0.05 < P < 0.2 were history of serious illness, circumcision status, and education. Factors not associated included trichomonas infection, age, occupation, religious affiliation, years spent in Nairobi, lifetime number of sexual partners, frequency of sex, or length of relationship with current partner.
By multivariate analysis, the odds of HIV-1 infection in men was 5.1 times higher if ever treated for a serious illness (95% confidence interval [CI], 1.4-19.1; P = 0.02), and only 0.12 as high if circumcised (95% CI, 0.02-0.91; P = 0.04). Also predictive of low HIV-1 prevalence was hygiene component 2 (OR, 0.41; 95% CI, 0.19-0.90; P = 0.03). The coefficients used to compute hygiene component 2 are shown in Table 1, which shows that this component is largest when bathing occurs immediately after sex. The decreased odds of HIV-1 infection associated with large values of hygiene component 2 suggest an inverse association between good hygiene and prevalence of HIV.
Although the primary objective of this study of sexual partners was to identify both male and female risk factors (including genital hygiene practices) for bacterial vaginosis in the female partners, we unexpectedly found that male genital hygiene was associated with the man's risk of HIV infection. Here, we have confirmed this association by principal components analyses.
Washing the penis had been thought to be associated with a reduction in the incidence of sexually transmitted infections, such as chancroid, during the first and second world wars,4 and some still advocate penile hygiene for sexually transmitted infection and HIV prevention.5,6 Our recent study of genital hygiene practices of men in Kenya showed that bathing the genitalia immediately after sex was common (74%) and that this practice was associated with access to water,1 demonstrating the potential acceptability of such an intervention even in resource-poor settings.
Limitations of this study include the fact that these men were enrolled because their partner had bacterial vaginosis, a fact that may provide information about their hygiene practices, and that hygiene practices reported by HIV-positive men may not reflect their specific behaviors at the time of HIV acquisition.
The recent intervention trial demonstrating a relative risk of 40% for HIV seroconversion among South African men after randomization to undergo circumcision7 invites the question about the mechanisms that might have explained the protective effect of circumcision. Beyond the removal of penile tissue highly susceptible to HIV infection, circumcision might also improve penile hygiene, reduce the risk of subpreputial inflammation and chancroid, and reduce the duration and size of the HIV inoculum to which the penile epithelium is exposed after intercourse-all factors that might also be influenced by postintercourse penile washing. In any case, further evaluation of the relatively inexpensive and simple practice of improved genital hygiene for reducing the risk of HIV acquisition by men seems warranted.
The authors would like to thank the director of KEMRI, Dr Davy Koech, and the study staff for their generous assistance on this project.
1. Steele MS, Bukusi E, Cohen CR. Male genital hygiene beliefs and practices in Nairobi, Kenya. Sex Transm Infect
2. Bukusi EA, Cohen CR, Meier AS, et al. Bacterial vaginosis: Risk factors among Kenyan women and their male partners. Sex Transm Dis.
March 16, 2006; [Epub ahead of print].
3. Seber GAF. Multivariate observations. Wiley, New York; 1984.
4. Moore JE. The value of prophylaxis against venereal diseases. JAMA
5. O'Farrell N. Soap and water prophylaxis for limiting genital ulcer disease and HIV-1 infection in men in sub-Saharan Africa. Genitourin Med
6. Short RV. The HIV/AIDS pandemic: new ways of preventing infection in men. Reproduction, Fertility and Development
7. Auvert B, Taljaard D, Lagarde E, et al. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 trial. PloS Med