Go, Vivian F PhD*; Solomon, Suniti MD†; Srikrishnan, Aylur K BA†; Sivaram, Sudha DrPH*; Johnson, Sethulakshmi C MSW†; Sripaipan, Teerada MPH*; Murugavel, Kailapuri G PhD†; Latkin, Carl PhD‡; Mayer, Kenneth H MD§; Celentano, David D ScD*
India has the third highest number of HIV-infected people in the world. Since the first AIDS case was reported in 1986,1 by November 2005, the cumulative number of reported AIDS cases had risen to almost 117,000.2 An estimated 2.5 million Indians are living with HIV/AIDS,3 representing more than 80% of South Asia's infected population.
Although the HIV epidemic continues to expand in India, there have been conflicting reports on the extent of HIV in the country, particularly in southern India. Although the number of surveillance sites is expanding, the data may still be skewed and inadequate.4-6 The United Nations Program on HIV/AIDS (UNAIDS)/World Health Organization (WHO) defines countries to have a generalized HIV epidemic if the national HIV prevalence in pregnant women extends beyond 1%.7 Most studies conducted in India among general and high-risk populations support the inference that India may be on the edge of a generalized HIV epidemic.8-12 Although the reported prevalence of HIV in the country is 0.7%,13 national surveillance and other studies reported that the HIV prevalence rate increased to >1% in antenatal clinics in the 4 southern states of Maharashthra, Karnataka, Andhra Pradesh, and Tamil Nadu in 2001.13,14 Recent studies suggest that national surveillance may overestimate the burden of HIV in southern India5 up to 2- to 3-fold, however. Given India's population of more than 1 billion, the health and cost implications of a generalized HIV epidemic are enormous; understanding the stage of the epidemic is critical to health care planning and HIV prevention strategies.
The continued transformation of the HIV epidemic in India is supported by the presence and interconnection of at least 2 elements in male and female populations: high HIV prevalence concentrated in high-risk populations13,15,16 and reports of elevated high-risk sexual behaviors among the female and male general population.17 The linkage of these elements creates the epidemiologic condition to support widening HIV transmission in the general population. These 2 elements have been explored more fully in the female population than in the male population. The HIV prevalence among certain high-risk populations, such as sex workers and sexually transmitted disease (STD) clinic patients in Maharashtra State and injecting drug users (IDUs) in Manipur State, reached >5% in 1990.13 Heterosexual sex accounts for 86% of all new HIV-1 infections in India,18,19 and it is estimated that 50% to 75% of new infections are acquired through male use of female sex workers.8,10,12 In the mid-1990s, studies reported that 21% of men attending STD clinics in Pune,10 and more than 30% of female sex workers in Mumbai and Pune15,16 were HIV-positive. Research has shown that a significant proportion of Indian men report premarital and extramarital sexual activity in India; a recent study of 2901 men aged 18 to 40 years old living in rural districts of 5 states found that 15% to 47% of men reported extramarital sex.17 Despite active surveillance in antenatal clinics20,21 and several studies that have focused on high-risk groups, such as those attending STD clinics,22,23 sex workers,24,25 and IDUs,26 few studies have assessed HIV prevalence in the general male population using community-based random probability samples in India. More HIV data are needed for male high-risk and male general populations to gauge the stage of the epidemic.
To understand the burden of HIV in southern India better, we aimed to characterize the HIV epidemic among men using 2 cross-sectional surveys: we first conducted a probability household survey in low-income areas of Chennai, India to provide an indicator of the HIV epidemic in the general population. These data indicated low rates of HIV/STDs and sexual risk behaviors in the general population, but HIV prevalence and risk behaviors were associated with alcohol use. We then conducted a probability survey among men recruited in wine shops to examine if HIV/STD risk was concentrated in these venues.
In India, wine shops are community-based, licensed, commercial establishments that sell alcohol for consumption on the premises or on a take-out basis. A more detailed explanation of wine shops is provided elsewhere.27 In brief, wine shops primarily sell distilled spirits and beer. On average, a wine shop in Chennai serves approximately 200 regular (at least 3 times a week) customers. Most wine shop patrons are men, and previous ethnographic research has suggested that after going to wine shops, men frequently engage in unprotected sex with sex workers.28
We compare results from each survey and discuss implications for the stage of the epidemic in southern India and useful HIV prevention strategies for each population segment.
This study was conducted in the southern city of Chennai, in Tamil Nadu State, India between March 2001 and June 2002. Findings from our study were used to inform an HIV prevention intervention trial (National Institute of Mental Health [NIMH] Collaborative HIV/STD Prevention Trial). The research team identified approximately 900 self-contained urban residential areas (“slums”) in Chennai, designated by the Tamil Nadu Slum Clearance Board as sites for future permanent housing. To maximize intervention effectiveness and minimize contamination between experimental and control arms in the trial, 30 slums that housed 100 to 300 families, had stable populations, and had discrete boundaries were purposively selected as study sites. In 24 of the 30 slums, we conducted a household survey and a wine shop survey.
Sample Selection for General Population Survey
Local study staff first conducted a household enumeration in the project slums. Using a systematic random sampling scheme, 65 households from each slum with at least 1 individual aged 18 to 40 years were selected to be approached for the study. From each selected household, 1 eligible participant was randomly selected to participate in the study, bringing the total sample to 1950 slum residents. Study staff visited households of selected individuals and gave them a color-coded invitation to participate in the study.
All residents of the selected slums, regardless of study status, were invited to attend health camps offering free physical examinations and prescriptions by local physicians. In the health camps, sampled participants holding color-coded invitation cards from each slum were informed of the risks and benefits of participation and given a copy of the informed consent to read and sign.
Sample Selection of Wine Shop Patrons
Of the 700 licensed wine shops in Chennai city in 2001, we purposively selected 100 wine shops in 24 clusters of 4 to 5 wine shops based on the presence of sex work services in the wine shop vicinity and high-risk sexual behavior among wine shop patrons indicated by previously collected ethnographic data.28 We invited every third person who entered a project wine shop bar to participate, for a total of 55 men per cluster.
The research protocol, questionnaire, and consent forms were reviewed and approved by several ethical review committees: Indian Council on Medical Research, YRG CARE's Institutional Review Board (IRB), the Johns Hopkins Bloomberg School of Public Health's Committee on Human Research, and the National Institutes of Health's Office for Protection from Research Risks.
To be eligible for the study, participants had to be lucid and capable of providing voluntary informed consent at the time of the interview. Each participant was informed of the risks and benefits of participation and advised of his or her rights as a study participant. At the end of the consent procedure, participants were asked if they had any questions. If a person agreed to participate, a copy of the informed consent form was given to that person to read and sign (or mark an “X”). Consenting participants were administered a 30-minute interview by 1 of 4 interviewers, using the computer-aided personal interview (CAPI) method. In the household survey, interviews were conducted in a separate area of the health camp in private sound-proofed booths. In the wine-shop survey, participants were offered free transportation and interviewed at an assessment site with private sound-proofed rooms. After participants were given HIV pretest counseling, laboratory personnel drew 10 mL of blood and clinicians collected urine samples and vaginal swabs (as appropriate) for HIV and selected STD testing.
Trained laboratory personnel tested all biologic specimens in the study laboratory in Chennai. Twenty percent of specimens, selected randomly, were sent to the central laboratory at the Johns Hopkins University School of Medicine and retested for quality control. HIV testing was performed on serum using HIV enzyme-linked immunosorbent (ELISA) 1.2.0 (Abbott Murex Biotech Limited, Kent, England), repeated using Genscreen HIV ½ Version 2 ELISA (BioRad, Marnes LaCoquette, France), and confirmed using Western blot (BioRad). Serum was tested for herpes simplex virus 2 (HSV-2) antibodies using Herpeselect 2 enzyme immunoassay (EIA; MRL; Focus Technologies, Los Angeles, CA) and was tested for syphilis using the Treponema pallidum hemaglutination assay (TPHA; Fujirebio, Tokyo, Japan). Urine was tested for chlamydia and gonorrhea DNA using Amplicor CT/NG polymerase chain reaction (PCR; Roche, Totowa, NJ).
Test results were made available to participants 2 weeks after the interview. Those who had antibodies to HIV were confidentially contacted and provided retesting, posttest counseling, and referrals. Those who were diagnosed with an STD were provided treatment on site and/or referrals to local services.
Given the small number of female participants interviewed in the wine shop survey (n = 91) and key differences between female participants in the wine shop survey (female participants were all sex workers in the wine shop survey) and the household survey, we excluded female participants in these analyses. To maximize comparability across surveys further, we limited our analysis to the 24 slums in which household and wine shop surveys were both conducted.
Simple logistic regression analysis was used to conduct exploratory analysis of the associations between independent variables and the dependent variable of interest. We were unable to conduct multivariate analysis with HIV as an outcome because of the small number of individuals with HIV. Because STDs and HIV may be transmitted through the same sexual behaviors, we developed models to assess factors associated with any STD, including HSV, HIV, chlamydia, gonorrhea, and syphilis among men from the general population and men from wine shops in Chennai slums.
Risk factors significantly associated with the outcome (P < 0.10) in the household or wine shop sample, or hypothesized to be associated with STDs, were entered into a multiple logistic regression model among men sampled from the household survey. To adjust for potential confounders, we used forward stepwise logistic regression analysis (P < 0.10 to enter, P < 0.05 to retain). Using a parallel analysis, we then identified factors associated with having any STD among men sampled from wine shops. We used χ2 analysis (Fisher exact 2-sided P value was used when 1 or more cells had an expected count <5) to assess differences across demographic and behavioral characteristics and across STD outcomes between the general population and the wine shop population. Generalized estimating equations29 were used in all analyses to control for intraslum correlations, and all analyses were conducted using STATA software version 9 (StataCorp, College Station, TX).30
From April through June 2001, 1631 of 1950 sampled adults were interviewed (84% response rate) through the household survey. Approximately 47% (n = 774) of the sampled population was male, of whom 685 were residents of the 24 slums in which the wine shop survey was also completed.
In the household sample (n = 685), the mean age was 28.5 (SD = 6.76) years and most were married (63%). Most (96.1%) had 1 or no sex partners in the past 3 months, and most (60%) drank alcohol less than once a week. HIV prevalence in this population was 1.2%, and 11.8% had any STD (HIV, HSV-2, syphilis, gonorrhea, or chlamydia).
Table 1 shows the odds ratios for risk factors associated with any STD in univariate and multivariate analyses among men from the general population survey. Men 34 years of age or older and men who had ever been married were significantly more likely to have an STD than those who were not. Men who had a greater number of sexual partners or who had exchanged money or other goods for sex in the past 3 months were also more likely to have an STD. Increased alcohol consumption was associated with a 6-fold increased risk of a prevalent STD. Mobility was protective of any laboratory-diagnosed STDs.
Variables that remained statistically significant after multivariate adjustment (see Table 1) were having been married, greater number of sexual partners, exchange of money for sex in the past 3 months, and binge drinking. No interactions were found to be statistically significant.
Wine Shop Men in Slums
From March through June 2002, 1196 male participants (91% response rate) were interviewed in the wine shop survey; of those, 654 were sampled from wine shops in 1 of the 26 slums included in these analyses.
Men in the wine shop sample were less likely to have a higher education (12+ years) and more likely to be mobile (Table 2) than men in the general population. They also had a higher HIV risk profile, with 70% having 2 or more sexual partners and 58% having unprotected sex with a casual partner in the past 3 months compared with 4% and 2%, respectively, in the general population. As expected, men from the wine shop drank alcoholic beverages more frequently and in higher quantity than their counterparts in the general population. Men from the wine shop sample also had significantly higher rates of HIV (3.4%) and STDs (21.6%) compared with men from the household survey (1.2%, P = 0.007 and 11.8%, P < 0.0001, respectively; Table 3).
Table 4 shows the odds ratios for risk factors associated with any STD in univariate and multivariate analysis among wine shop men. As was seen in the general population, the oldest men, those 34 years of age or older, or those who had ever been married or tested for HIV in past were more likely to have an STD. Having unprotected sex with a casual partner in the past 3 months was also associated with increased risk for an STD. Again, mobility was protective of STD prevalence. Variables that remained statistically significant after multivariate adjustment were older age, ever married, ever tested for HIV, and unprotected sex in the past 3 months.
Because chlamydia, gonorrhea, and syphilis are bacterial and only infectious for a few months, we removed these infections from the outcome in the household and wine shop analyses and found that the multivariate models were unchanged (data not shown).
The prospect of an HIV epidemic in India poses enormous challenges to the Indian health infrastructure and, because of the large underlying population, is likely to contribute dramatically to the size of the global epidemic. Indian officials recognize the potential enormity of the health problem, and starting in April, a new phase of the National AIDS Control Program (NACP) aims to reduce the number of new HIV infections, improve clinical management, and provide antiretroviral therapy to more people.31 The current paucity of reliable data on the stage of the epidemic in India remains a barrier to planning cost-effective prevention and treatment strategies, however.
The rate of HIV in the general population of men is 1.2% in Chennai-just greater than the level of a generalized epidemic. We found that high-risk behaviors in the general population are limited to a relatively small group of men, however. Approximately 4% of men had sex with more than 2 partners over the past 3 months, 4% had exchanged sex for money or goods in the past 3 months, and 4% had 5 or more drinks on the days they consumed alcohol (see Table 2). Men who engaged in these behaviors were more likely to be diagnosed with an STD, including HIV.
The survey among wine shop patrons revealed that HIV/STDs may be concentrated in alcohol venues. Studies have shown that alcohol consumption is associated with increased HIV sexual risk behaviors.32-36 In contrast to men from the general population, most men in wine shops had more than 2 partners in the past 3 months (69.6%), had unprotected sex with a casual partner (not including paid partners) in the past 3 months (58.4%), and had exchanged sex for money or goods for sex in the past 3 months (54.1%). Given the pervasiveness of high-risk behaviors in this population, less traditional risk factors were associated with an STD. In addition to men who had unprotected sex with a casual partner in the past 3 months, men who had previously tested for HIV were also more likely to have an STD, indicating that individuals with STDs in this subpopulation were aware of their high-risk behaviors.
Age and marital status were associated with STD prevalence in both populations, reflecting the cumulative nature of viral STDs, which accounts for a substantial proportion of the STD burden.
It is interesting to note that the odds ratios are much higher in the low-prevalence population. This may be attributable, in part, to the fact that those at high risk have been in a high-risk environment for a long period; hence, the behaviors in the past 3 months among wine shop patrons is not as predictive as among the household survey.
This study has several limitations that should be considered when interpreting results. First, because all behavioral data relied on self-report, associations of high-risk behavior and disease may be underestimated across risk behaviors. When answering questions about sensitive behaviors, participants may have given what they perceived to be socially desirable responses. Although differential misclassification of the exposure (HIV risk behaviors) may have occurred, it is expected that this bias would dilute the estimates of association between risk behaviors and STD prevalence, and thereby result in conservative estimates. In addition, because women rarely go to wine shops in India, we were unable to include women in this analysis. Our data were collected in 2001 and 2002 and may not reflect the current situation. However, data collected between 2003 and 2004 among male wine shop patrons (n = 2914) found STD prevalence rates similar to our wine shop sample, suggesting that rates continue to be high among this population. The results of this study may have limited generalizability to other cities in India. Because the participants of this study were randomly selected from the general population living in slums, however, conclusions and recommendations may be applicable to other slum-dwelling communities in Chennai and to southern India.
With several reports of increases in HIV in antenatal clinics,13,14 there has been wide-spread speculation that the HIV epidemic has spread to the general Indian population. Our study found that HIV, STDs, and sexual risk behaviors are relatively uncommon among most men in the slums of Chennai, India. Within these slums, however, there are pockets of men who frequent wine shops; who are practicing high-risk behaviors; and have high rates of STDs, including HIV. These men may be immersed in a high-risk environment over a longer period. Given the widespread practice of high risk behaviors, interventions that include all male patrons from wine shops would be effective. In contrast, interventions set in slum communities, where high-risk behaviors are more uncommon, should focus on high-risk subgroups in the community, perhaps using a screening tool to identify men who frequent wine shops and are high risk for HIV/STDs.
The study was supported by a grant (1U10 MH61543) from the NIMH, National Institutes of Health. The authors express their gratitude to the study participants, whose commitment and cooperation made the study possible.
1. Simoes EA, Babu PG, John TJ, et al. Evidence for HTLV-III infection in prostitutes in Tamil Nadu (India). Indian J Med Res
2. World Health Organization. Summary Country Profile for HIV/AIDS Treatment Scale-Up: India. 1
. Geneva, Switzerland: WHO; 2005.
4. Chandrasekaran P, Dallabetta G, Loo V, et al. Containing HIV/AIDS in India: the unfinished agenda. Lancet Infect Dis
5. Dandona L, Lakshmi V, Sudha T, et al. A population-based study of human immunodeficiency virus in south India reveals major differences from sentinel surveillance-based estimates. BMC Med
6. Steinbrook R. HIV in India-a complex epidemic. N Engl J Med
7. UNAIDS/WHO. Guidelines for Second Generation HIV Surveillance
. Geneva, Switzerland: UNAIDS; 2000.
8. Nagelkerke NJ, Jha P, de Vlas SJ, et al. Modelling HIV/AIDS epidemics in Botswana and India: impact of interventions to prevent transmission. Bull World Health Organ
9. National Intelligence Council. The Next Wave of HIV/AIDS: Nigeria, Ethiopia, Russia, India, and China. Intelligence Community Assessment
. Central Intelligence Agency, 2002.
10. Rodrigues JJ, Mehendale SM, Shepherd ME, et al. Risk factors for HIV infection in people attending clinics for sexually transmitted diseases in India. BMJ
11. Shefalee V. What men want. India Today
. September 26, 2004.
12. Venkataramana CB, Sarada PV. Extent and speed of spread of HIV infection in India through the commercial sex networks: a perspective. Trop Med Int Health
13. Solomon S, Chakraborty A, Yepthomi RD. A review of the HIV epidemic in India. AIDS Educ Prev
14. National AIDS Control Organization. Programme Implementation Guidelines for a Phased Scale Up of Access to Antiretroviral Treatment for People Living With HIV/AIDS
. New Delhi, India: Ministry of Health and Family Welfare, Government of India; 2003.
15. Bollinger RC, Tripathy SP, Quinn TC. The human immunodeficiency virus epidemic in India. Current magnitude and future projections. Medicine (Baltimore)
16. Jain MK, John TJ, Keusch GT. Epidemiology of HIV and AIDS in India. AIDS
. 1994;8(Suppl 2):S61-S75.
17. Verma RK, Lhungdim H. Sexuality and sexual behaviors in rural India: evidence from a five state study. In: Verma RK, Pelto PJ, Schensul SL, et al, eds. Sexuality in the Time of AIDS: Contemporary Perspectives from Communities in India
. New Delhi, India: Sage Publications; 2004:156-176.
18. Bhattacharya G. Sociocultural and behavioral contexts of condom use in heterosexual married couples in India: challenges to the HIV prevention program. Health Educ Behav
19. Solomon S, Buck J, Chaguturu SK, et al. Stopping HIV before it begins: issues faced by women in India. Nat Immunol
20. UNAIDS. 2004 Report on the Global AIDS Epidemic
. Geneva, Switzerland: UNAIDS; 2004.
21. National AIDS Control Organization. Annual Report 2002-2003, 2003-2004
. Ministry of Health and Family Welfare, Government of India. New Delhi, India; 2004.
22. Arora DR, Gautam V, Gill PS, et al. Haryana state in India, still a low HIV prevalence state. Sex Transm Infect
23. Kumar R, Jha P, Arora P, et al, International Studies of HIV/AIDS (ISHA) Investigators. Trends in HIV-1 in young adults in south India from 2000 to 2004: a prevalence study. Lancet
24. Desai VK, Kosambiya JK, Thakor HG, et al. Prevalence of sexually transmitted infections and performance of STI syndromes against aetiological diagnosis, in female sex workers of red light area in Surat, India. Sex Transm Infect
25. Singh TN, Kananbala S, Thongam W, et al. Increasing trend of HIV seropositivity among commercial sex workers attending the Voluntary and Confidential Counseling and Testing Centre in Manipur, India. Int J STD AIDS
26. Panda S, Kumar MS, Lokabiraman S, et al. Risk factors for HIV infection in injection drug users and evidence for onward transmission of HIV to their sexual partners in Chennai, India. J Acquir Immune Defic Syndr
27. Sivaram S, Johnson S, Bentley ME, et al. Exploring “wine shops” as a venue for HIV prevention interventions in urban India. J Urban Health
28. Sivaram S, Srikrishnan AK, Latkin CA, et al. Development of an opinion leader-led HIV prevention intervention among alcohol users in Chennai, India. AIDS Educ Prev
29. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a generalized estimating equation approach. Biometrics
30. StataCorp. STATA Statistical Software: Release 9
. College Station, TX: StataCorp LP; 2005.
31. Steinbrook R. HIV in India-the challenges ahead. N Engl J Med
32. Bagnall G, Plant M, Warwick W. Alcohol, drugs and AIDS-related risks: results from a prospective study. AIDS Care
33. Caetano R, Hines AM. Alcohol, sexual practices, and risk of AIDS among blacks, Hispanics, and whites. J Acquir Immune Defic Syndr Hum Retrovirol
34. Graves KL. Risky sexual behavior and alcohol use among young adults: results from a national survey. Am J Health Promot
35. McEwan RT, McCallum A, Bhopal RS, et al. Sex and the risk of HIV infection: the role of alcohol. Br J Addict
36. Seage GR III, Mayer KH, Wold C, et al. The social context of drinking, drug use, and unsafe sex in the Boston Young Men Study. J Acquir Immune Defic Syndr Hum Retrovirol
© 2007 Lippincott Williams & Wilkins, Inc.