Risk of HIV infection but not other sexually transmitted diseases is lower among homeless Muslim men in Kolkata

CONCISE COMMUNICATION Risk of HIV infection but not other sexually transmitted diseases is lower among homeless Muslim men in Kolkata Arunansu Talukdar a, Mafafuzur R. Khandokar a, Subir K. Bandopadhyay a and Roger Detels b Objective: To compare the risk of HIV and sexually transmitted diseases (STD) among homeless Muslim (circumcised) and Hindu (uncircumcised) men in Kolkata, India. Background: Many observational studies and clinical trials in Africa have demonstrated that male circumcision provides protection against HIV acquisition, but there are sparse data on circumcision and HIV in India, which has the largest number of HIV cases in the world. Methods: Using a two-stage probability proportionate to size cluster design among homeless men aged 18 49 years in Kolkata, India, data were obtained on religion, behavioral risk factors, and HIV/STD prevalence, by administering an anonymous questionnaire. Rapid HIV tests and testing for syphilis were performed on blood, and urine samples were obtained to test for gonorrhea. Results: The odds ratio for HIV among Muslims (circumcised) compared to Hindus (uncircumcised) was 0.43 (95% confidence interval 0.29 0.67). Despite Muslims having more partners and visits to commercial sex workers, the rates of syphilis and gonorrhea were similar. The results suggest that a biological effect of circumcision protects against HIV infection. Conclusion: The beneficial effect of circumcision should be communicated to high-risk groups, as well as to the general population. ß 2007 Wolters Kluwer Health Lippincott Williams & Wilkins AIDS 2007, 21:2231 2235 Keywords: circumcision, HIV/AIDS, India, Muslims, sexual behaviour, sexually transmitted diseases Introduction Many observational studies have reported a correlation of male circumcision with a reduced risk for HIV infection. The Cochrane Collaboration published a review of the available studies suggesting that circumcision can be used as an intervention to prevent HIV infection [1]. The updated review included 37 observational studies, 18 conducted among the general population and 19 among high-risk populations. The results of clinical trials held in Kenya and Uganda have recently demonstrated the protective effect of male circumcision on HIV From the a Calcutta National Medical College, Calcutta, India, and the b UCLA School of Public Health, Los Angeles, California, USA. Correspondence to Roger Detels, School of Public Health, University of California, Los Angeles, CA 90095-1772, USA. Tel: +1 310 206 2837; fax: +1 310 206 6039; e-mail: detels@ucla.edu Received: 24 April 2007; revised: 7 June 2007; accepted: 14 June 2007. ISSN 0269-9370 Q 2007 Wolters Kluwer Health Lippincott Williams & Wilkins 2231

2232 AIDS 2007, Vol 21 No 16 infection [2]. The results of the two clinical trials support the findings of the South Africa Orange Farm Intervention Trial [3]. India has the highest number of HIV cases in the world according to UNAIDS, with an estimated 5.7 million people living with HIV/AIDS [4]. Although there is a high prevalence of HIV among sex workers and drug users, HIV prevalence (0.9%) among the general population is still relatively low. India has over one million inhabitants; therefore, even a small increase in the prevalence of HIV would represent a significant increase in the world s HIV/AIDS burden. The majority of the cases are in urban areas, fuelled by commercial sex workers (CSW) and their customers, and heterosexual spread beyond them to others. India is considered to be a next wave country; that is, it stands at a critical point, with HIV poised to expand. Circumcision is usually determined according to religion in India, and there are strong beliefs and opinions surrounding its practice. It is mostly practised in India among Muslims, who comprise less than 15% of India s population [5]. So far, only two studies from India have reported on circumcision and HIV risk. Both studies were conducted among attendees at sexually transmitted disease (STD) clinics, where there is a strong potential for selection bias [6]. National HIV sentinel surveillance is carried out annually to identify the sociodemographic characteristics of those infected, in addition to estimating HIV prevalence, but information on circumcision and religion are not captured in the surveillance database. The present population-based study investigated the association of the Muslim religion with HIV infection and other STD among homeless men in Kolkata, India. The findings of this study are relevant to the debate on circumcision and the risk of acquisition of HIV. Methods We obtained data on sociodemographic characteristics, including religion, behavioral risk factors, and HIV/STD prevalence, using a two-stage probability proportionate to size cluster design among homeless men in Kolkata, India. The study area included seven municipal wards located in the downtown area of Kolkata city, where the majority of homeless live. The survey targeted men aged 18 49 years who lived in public spaces within the study area and had been living there at least 30 days. An anonymous questionnaire was administered to each participant. The first part of the question included non-sensitive questions such as sociodemographic characteristics, knowledge about HIV/AIDS, condom use, etc. The sensitive questions relating to sexual behavior were included in the second part, administered by CD player with earphones and an answer sheet utilizing symbols rather than words, because many of the men were illiterate [7]. We followed World Health Organization strategy III for HIV diagnosis. The following three different rapid tests were used: (i) HIV Comb test; (ii) HIV Tridot test; and (iii) HIV enzyme immunoassay. Men seropositive on all three tests were considered to be HIV infected. All test kits were approved by the Ministry of Health, Government of India. The rapid HIV antibody testing kits were evaluated using an established panel of samples, and all showed 100% sensitivity and specificities in the range of 98.6 100%. To identify syphilis, the toludine red unheated serum test (Span Diagnostics Limited, Surat, India) and TPHA 500 (Newmarket Laboratories Ltd., UK) were used. Urine was tested for Neisseria gonorrhoea by the polymerase chain reaction nucleic acid amplification and nucleic acid hybridization technique (Amplicor NG Test; Roche Molecular Diagnostics, India). All the kits used for this study were approved by the Drug Controller Authority of India. All participants received counseling before and after testing, as well as free treatment for STD and access to free condoms. Those found to be HIV positive were referred to the Ante Retroviral Treatment Centre run by the government for follow-up and free HAART and other medications. The two-stage cluster design sample gave us a selfweighted sample. Analyses were performed using the SAS program (SAS 9.1.3; SAS Institute, Inc., Cary, North Carolina, USA). Descriptive analyses of all variables were done using proportions and means when applicable. A x 2 test was performed to compare proportions, and was considered significant if the P value was less than 0.05. Adjusted logistic regression was carried out to detect predictors of being HIV positive and having other STDs. The study was approved by the Institutional Review Boards of the University of California, Los Angeles, USA and the Indian Council of Medical Research. Results We approached 606 eligible participants for this study, of whom 112 refused (18.5%). Of these refusals, 85 were unwilling to participate, and 27 were willing but said they did not have time. One participant withdrew during the interview. Therefore, 493 (81%) homeless men completed the interview. An additional 19 participants refused to donate specimens. Blood and urine specimens were therefore obtained from 474 participants. The overall participation rate was 78.2%. Among them, 105

Risk of HIV in Muslim men in Kolkata, India Talukdar et al. 2233 participants were Muslim, 380 were Hindu, and eight were another religion. All Muslims are circumcised on religious grounds, and Hindus in this area do not usually opt for circumcision. We restricted our further analyses to the 485 Hindu and Muslim participants. We assumed all Muslims were circumcised and all Hindus were not. The mean age of the participants was 28.2 years. The study participants had been living in Kolkata from 0 to 32 years, with a mean duration of 4.7 years. Overall, 63.7% of the participants had not attended school; however, approximately half of them could read and write their names. Almost 43% were currently married, but only a quarter of the married participants lived with their wives. Ninety-four per cent of participants had ever used tobacco or consumed alcohol. HIV knowledge was poor. Only 10% of men perceived that they were at risk of HIV infection, although 90% were sexually active. The mean age for sexual debut for Muslim men was 19.5 years, and 18.9 years for Hindu men. Over four-fifths of the participants had ever visited a CSW. The average number of sex partners in the past 6 months was three per participant. Only a quarter of them had ever used condoms with non-regular partners. A comparison of sociodemographic and sexual risk behaviors of Hindu and Muslim men is presented in Table 1. The Hindu men were younger. There were more illiterate individuals among Muslim men than among the Hindu men. No significant difference was observed between the two groups with regard to income, home visits per year, substance abuse, and having sex with men. Visits to a CSW and multiple partners were, however, significantly more frequently reported by Muslim men. HIV prevalence was significantly (P < 0.001) higher among Hindu men (4.4 versus 0.98%), but the prevalence of syphilis (6.9 versus 7.8%) and gonorrhea (5.2 versus 9.8%) was not significantly different between Hindu and Muslim men (Table 2). The prevalence of gonorrhea among HIV-positive (5.9%) and HIV-negative (6.2%) men and the prevalence of syphilis among HIV-positive (11.8%) and HIV-negative (6.9%) men were not significantly different. The prevalence of HIV among gonorrhea-infected (3.5%) and gonorrhea-uninfected (3.7%) men among syphilis-infected (6.1%) and syphilis-uninfected (3.5%) men was also not significantly different. Crude analysis did not show any significant association between HIV or STD with religion. An adjusted analysis, however, demonstrated that Muslim men had a 57% less likelihood of being positive for HIV (odds ratio 0.43, 95% confidence interval 0.29 0.67) than Hindu men. Adjusted analysis did not reveal any significant association between syphilis or gonorrhea infection and religion. Adjusted logistic regression analysis for predictors of HIV (Table 3) indicated that visits to CSW, frequency of visits to CSW, condom use with CSW, having multiple sexual partners, and being married were independently associated with being positive for HIV in this study, whereas being Muslim was protective. Discussion This population-based study expands results from other studies in India. This result supports previous reports that male circumcision, which is routinely practised among Muslims, reduces the risk of HIV acquisition. Despite Table 1. Sociodemographic and behavioral risk factors of study population (n U 485). Characteristics Category Hindu Muslim P value Age in years, number (%) 18 24 98 (25.8) 14 (13.3) 0.04 25 29 149 (39.2) 42 (40.0) 30 34 86 (22.6) 33 (31.4) 35 39 30 (7.9) 8 (7.8) 40 17 (4.5) 8 (7.6) Educational level, number (%) Illiterate 218 (57.4) 90 (85.7) 0.001 Up to grade IV 75 (19.7) 9 (8.6) Above grade IV 87 (22.9) 6 (5.7) Home visits per year, number (%) Less than 3 times 191 (50.1) 54 (51.4) 0.83 3 times or more 189 (49.7) 51 (48.6) Monthly income, number (%) US$0 7 77 (20.2) 16 (15.2) 0.8 US$8 9 118 (31.0) 40 (38.1) US$9 11 116 (30.5) 28 (26.7) More than US$11 69 (18.2) 21 (20.0) Marital status, number (%) Never married 146 (38.4) 64 (60.1) 0.002 Currently married 92 (24.2) 14 (13.3) Formerly married 142 (37.4) 27 (25.7) Had consumed alcohol, number (%) Yes 263 (69.2) 75 (71.4) 0.73 Had visited a CSW in the past 6 months, number (%) Yes 266 (77.6) 90 (93.8) 0.001 Had male sex partners, number (%) Yes 52 (15.2) 16 (16.7) 0.71 Had multiple sexual partners, number (%) Yes 290 (84.6) 92 (95.8) 0.001 CSW, Commercial sex worker.

2234 AIDS 2007, Vol 21 No 16 Table 2. HIV/sexually transmitted disease prevalence and odds ratio (crude and adjusted) for religion and sexually transmitted diseases. Disease Prevalence (%) Crude OR AOR Hindu (n ¼ 364) Muslim (n ¼ 102) (95% CI) a (95% CI) b P value c HIV 4.4 0.9 0.21 (0.03 1.25) 0.43 (0.29 0.67) 0.001 Syphilis 6.9 7.8 0.87 (0.38 2.1) 1.11 (0.77 1.42) 0.82 Gonorrhea 5.2 9.8 0.53 (0.23 1.25) 1.25 (0.91 1.67) 0.17 AOR, Adjusted odds ratio; CI, confidence interval; OR, odds ratio. a Fisher s exact test performed when applicable. b Age, education, marital status, income, alcohol use, commercial sex worker visits, condom use, multiple partners, and AIDS awareness were adjusted. c P value for adjusted analysis. having more partners and visits to CSW, the Muslim participants had a lower prevalence of HIV, suggesting that a biological effect of circumcision protects against HIV infection. In an uncircumcised man, the cells in the inner foreskin and frenulum are directly exposed to vaginal secretions during intercourse, and this accessible location of HIV-1 target cells presumably increases the risk of infection. In contrast, in circumsized men, the penile shaft is thought to be covered with a thickly keratinized epithelium, providing protection from infection. The role of genital hygiene in the association between male circumcision and HIV infection is unclear. Several studies have shown that when compared with circumcised men, uncircumcised men are at greater risk of acquiring certain STD. Other studies, however, have found no association between STD and circumcision [8 11]. The first systematic review and meta-analysis of the association of male circumcision with ulcerative STD [12] reported that potential male circumcision interventions to reduce HIV in high-risk populations may provide an additional benefit by protecting against certain other STD (e.g. chancroid, syphilis). We did not, Table 3. Logistic regression analysis for sexual behaviors associated with being positive for HIV. Characteristics % HIV-positive AOR 95% CI P value Marital status Ever married 3.92 1.6 1.1 2.3 0.01 Unmarried 4.27 ref Visited CSW Yes 4.68 5.22 2.03 13.3 0.001 No 1.23 ref Multiple sexual partners 4.63 2.9 2.5 3.5 0.001 Frequency of CSW visits Few times a week 8.33 1.88 1.32 2.68 0.001 Few times a month 2.38 ref Condom use with CSW Always 0.0 0.63 0.44 0.9 0.01 Sometimes/never 4.85 ref Sex with male partner Yes 4.76 1.04 0.62 1.72 0.89 No 3.89 ref AOR, Adjusted odds ratio; CI, confidence interval; CSW, commercial sex worker. 80.85% (342 of 423) of sexually active study population had visited CSW. N is 423 for all variables except frequency of CSW visits and condom use with CSW (asked questions only for those who visited CSW). however, find any significant protection against STD among those who were circumcised. Any misclassification of circumcision would have lowered the observed association. HIV testing was performed after the interview by technicians who were unaware of the religion of the participant or the interview results. The reported status could thus not have differentially influenced the test results. Our data did not reveal a significant association between religion and syphilis or gonorrhea infection. It is possible that the higher level of infectiousness of the syphilis and gonorrhea organisms may have overcome the protective effect of circumcision in those men who had limited access to hygienic facilities. The validity of the study might have been jeopardized depending on the extent of non-participation and nonresponse. Low rates of participation (61 82%) by homeless individuals have been observed in other studies [13]. We were not able to compare the characteristics of non-participants with participants. To have biased the results, non-participating Muslims would have had to practice less risky sexual behaviors than Hindu participants, which is counterintuitive, or Hindu men who refused participation would have had to practice more risky sexual behaviors. The study suggests that circumcision reduces the risk of HIVacquisition among a high-risk group of men in India, the homeless. Awareness of the beneficial effect of circumcision must be propagated to high-risk groups, as well as to the general population. Data on male circumcision should be captured in the future on HIV sentinel surveys and voluntary counseling and testing records. Public health officials and policy makers should consider the practicalities of providing safe voluntary male circumcision services for high-risk groups, keeping in mind that religious beliefs are closely linked to the acceptability of circumcision. Sponsorship: This work was supported by National Institutes of Health Fogarty International Center grant D43 TW000013. Conflicts of interest: None.

Risk of HIV in Muslim men in Kolkata, India Talukdar et al. 2235 References 1. Siegfried N, Muller M, Deeks J, Volmink J, Egger M, Low N, et al. HIV and male circumcision a systematic review with assessment of the quality of studies. Lancet Infect Dis 2005; 5:165 173. 2. National Institute of Allergy and Infectious Diseases. National Institutes of Health News Release, 13 December 2006. Adult male circumcision significantly reduces risk of acquiring HIV trials Kenya and Uganda stopped early. Available at: http:// www3.niaid.nih.gov/news/newsreleases/2006/amc12_06.htm. Accessed: July 2007. 3. Gray RH, Kiwanuka N, Quinn TC, Makumbi F, Watya S, Nalugoda F, et al. Male circumcision and HIV acquisition and transmission: cohort studies in Rakai Uganda: Rakai Project Team. AIDS 2000; 14:2371 2381. 4. UNAIDS. 2006 Report on global AIDS epidemic. Geneva, Switzerland: UNAIDS; May 2006. 5. Census of India, 2001. New Delhi, India. Office of the Registrar General. Available at: www.censusindia.net. Accessed: July 2007. 6. Reynolds SJ, Shepherd ME, Risbud AR, Gangakhedkar RR, Brookmeyer RS, Divekar AD, et al. Male circumcision and risk of HIV-1 and other sexually transmitted infections in India. Lancet 2004; 363:1039 1040. 7. Liu H, Xie J, Yu W, Song W, Gao Z, Ma Z, et al. A study of sexual behavior among rural residents of China. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 19:80 88. 8. Smith GL, Greenup R, Takafuji ET. Circumcision as a risk factor for urethritis in racial groups. Am J Public Health 1987; 77:452 454. 9. Nicoll A. Routine male neonatal circumcision and risk of infection with HIV-1 and other sexually transmitted diseases. Arch Dis Child 1997; 77:194 195. 10. De Vincenzi I, Mertens T. Male circumcision: a role in HIV prevention? [Editorial review]. AIDS 1994; 8:153 160. 11. Donovan B, Bassett I, Bodsworth NJ. Male circumcision and common sexually transmitted diseases in a developed nation setting. Genitourin Med 1994; 70:317 320. 12. Weiss HA, Thomas SL, Munabi SK, Hayes RJ. Male circumcision and risk of syphilis, chancroid, and genital herpes: a systematic review and meta-analysis. Sex Transm Infect 2006; 82:101 109. 13. Robertson MJ, Clark RA, Charlebois ED, Tulsky J, Long HL, Bangsberg DR, et al. HIV seroprevalence among homeless and marginally housed adults in San Francisco. Am J Pub Health 2004; 94:1207 1217.