Secondary Logo

Journal Logo

Epidemiology and Social Science

High HIV Prevalence Among a High-Risk Subgroup of Women Attending Sexually Transmitted Infection Clinics in Pune, India

Mehta, Shruti H PhD, MPH*; Gupta, Amita MD; Sahay, Seema PhD; Godbole, Sheela V MD; Joshi, Smita N MD; Reynolds, Steven J MD†§; Celentano, David D PhD*; Risbud, Arun MD; Mehendale, Sanjay M MD; Bollinger, Robert C MD, MPH†∥

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: January 1, 2006 - Volume 41 - Issue 1 - p 75-80
doi: 10.1097/01.qai.0000174653.17911.4a
  • Free


Recent projections suggest that approximately 5.1 million persons are living with HIV in India,1 leading some to speculate that India has surpassed South Africa and now has more HIV cases than any other country.2 The spread of HIV in India has predominantly been through heterosexual transmission, with an estimated 38% of HIV infections occurring among women.3,4 Initially, the Indian HIV epidemic in women was heavily concentrated among commercial sex workers (CSWs)4; however, studies from several cities in India have revealed the vulnerability to HIV of married women who have had only 1 lifetime partner.5-7 In a study from Manipur, 45% of wives of HIV-infected injection drug users were also HIV positive despite the fact that none of these women reported injection drug use themselves.6 In Chennai, among 134 HIV-positive women who sought care for HIV infection, 82% were married, monogamous women whose only risk factor for HIV infection was unprotected intercourse with their HIV-infected spouse.7 Finally, in Pune, a high HIV prevalence (14%) among married, monogamous women attending sexually transmitted infection (STI) clinics between 1993 and 1996 has been reported.5 These studies suggest that these women are at risk for HIV because of the high-risk behaviors of their husbands.

Over the past decade, India's governmental organizations, including the National AIDS Control Organization (NACO), nongovernmental organizations, and public-private partnerships have mounted numerous prevention and education programs directed at high-risk groups.8 Women attending STI clinics are considered one such high-risk group, but most risk reduction efforts have focused on female CSWs and not other female clinic clients. To design interventions for women who deny CSW, we need to better understand the prevalence and correlates of HIV in this population. The objective of this study was to characterize changes in HIV prevalence and correlates among women who denied a history of CSW who attended STI clinics in Pune (Maharashtra state), India over the decade 1993 to 2002.


Study Population

Between May 1993 and August 2002, 13,249 persons (2376 women) attending 3 STI clinics in Pune, Maharashtra (1 municipal STI clinic, 1 STI clinic in a state government teaching hospital, and 1 health care center in a “red light” district) were screened for HIV-1 and HIV-2 infection. The present study sample included the subset of 1020 women who denied a history of CSW (63% from the municipal STI clinic, 27% from the STI clinic in the government hospital, and 10% from the health center in the red light district). The project was approved by the Johns Hopkins University Joint Committee on Clinical Investigation and the Ethical Committee of the National AIDS Research Institute in Pune. Both institutional review boards have filed an assurance of compliance with US Health and Human Services Regulations for Protection of Human Research Subjects with the Office for Protections from Research Risks.

The methods for this study have been described in detail elsewhere.9 After informed consent was obtained, a structured interviewer-administered questionnaire was used to obtain data on demographics, medical and reproductive history, sexual behavior, and knowledge of HIV and STIs. The questionnaire was revised in 1998 to collect more detailed information on types of first and regular sexual partners. Participants also underwent a physical examination and were tested for STIs. Clinical diagnosis of genital discharge and ulcer was based on physical examination by a physician and clinic-based laboratory tests (Gram stain, wet mount, dark ground microscopy). Gram stains and cultures for Neisseria gonorrheae were performed on all women who consented to physical examination. Blood samples were screened for syphilis antibody using a venereal disease research laboratory test or a rapid plasma reagin test. Positive results were confirmed by Treponema pallidum hemagglutination assay using standard microbiologic methods.10 All patients received pre- and posttest counseling for HIV and other STIs and were offered treatment of STIs according to guidelines from the World Health Organization.11

HIV Serology

Serum samples were screened initially using a commercially available enzyme-linked immunosorbent assay (ELISA) kit for detection of HIV-1 and HIV-2 antibodies (HIV EIA, Labsystems OY, Helsinki, Finland; Detect-HIV, BioChem ImmunoSystems, Inc., Montreal, Canada; Innotest HIV-1/HIV-2 Ab s.p., Innogenetics NV, Zwijndrecht, Belgium). Specimens positive by ELISA were confirmed using the Rapid Test Device (Immunocomb II, HIV1 & 2 BiSpot, Orgenics, Yavne, Israel; HIV Tri-DOT, J. Mitra & Co., New Delhi, India; Capillus HIV-1/HIV-2, Cambridge Diagnostics, Galway, Ireland). Specimens that were discrepant by these assays were confirmed using either a 3rd ELISA or the commercially available HIV-1/HIV-2 Western blot (INNO-LIA HIV-1/HIV-2 Ab, Innogenetics NV, Zwijndrecht, Belgium; HIV BLOT 2.2 Western blot Assay, Genelabs Diagnostics SA, Geneva, Switzerland). Western blot results were interpreted according to Centers for Disease Control and Prevention criteria.12 For the purposes of this analysis, individuals with evidence of either HIV-1 or HIV-2 infection were considered to be HIV infected.

Statistical Analysis

The Mantel-Haenszel test for trend was used to determine whether prevalence of HIV, STIs, or risk behaviors changed over time. Calendar time was categorized into 3 periods: 1993-1996, 1997-1999, and 2000-2002 based on the patterns of HIV prevalence in each calendar year. Correlates of HIV infection were examined using χ2 tests for categorical variables, Mann-Whitney U tests for continuous variables, and univariate logistic regression. Multivariate logistic regression analysis was used to identify factors independently associated with prevalent HIV infection. Factors that were considered in the analysis included calendar time, demographics, history of tattoo, blood transfusion and injection drug use, lifetime number of sexual partners, age at first sexual intercourse and condom use, as well as history and current clinical findings of STIs. Potential interactions between independent correlates of HIV infection and calendar time were considered. Statistical analysis was performed using STATA version 8.0 (College Station, TX).


Description of Population

Characteristics of the 1020 women who attended STI clinics from May 1993 to August 2002 and denied a history of CSW are shown in Table 1. Of these 1020 women, 411 (40%) had no formal education, 853 (84%) were married, 123 (12%) were divorced, separated, or widowed, and 382 (38%) were employed. A total of 350 (34%) were 16 years of age or younger at the time of their first sexual intercourse. The majority of women reported only 1 lifetime sexual partner (n = 813, 86%) and no condom use (n = 770, 86%). Three-quarters (74%) of women reported that they came to the clinic because of STI symptoms, 12% reported coming in for a check-up, 13% reported coming in because of a referral, and 9% reported coming in because of contact with a person that had an STI. Only 40% had ever heard of HIV/AIDS. Six percent and 8% of women had laboratory-confirmed gonorrhea and syphilis, respectively.

HIV Prevalence by Sociodemographic and Risk Behaviors Among 1020 Women Attending STI Clinics From 1993-2002 in Pune, India Who Deny a History of Commercial Sex Work

Of the 1020 women, 839 (82%) consented to undergo a physical examination, of whom 766 (75%) also underwent a pelvic examination. Of the 839 women who had a physical examination, 210 (25%) were found to have evidence of genital ulcer disease. Of the 766 women who had a pelvic examination, physicians documented genital discharge in 521 (68%).

In a subgroup of 449 women screened between 1998 and 2002, after which more detailed questions were asked about sexual behavior, 410 (90%) reported that their first sexual partner was their husband and 376 (84%) reported having a regular partner. Of the 376 with a regular partner, 345 (92%) reported that their regular partner was their husband, and only 12 women (3%) reported having >1 regular partner.

Prevalence of HIV Infection

Overall, 216 (21%) of these 1020 women were HIV infected, 203 (94%) with HIV-1 and 13 (6%) with HIV-2. The prevalence of HIV increased significantly over time from 14% in 1993 to 29% in 2001-2002 (P < 0.001). Sociodemographic characteristics and sexual behavior also showed changes over time among this group of women (Table 2). Compared with women screened in the earlier years (1993-1996), women screened for HIV in later years (2000-2002) tended to be older (P < 0.001), were less likely to be currently married (P < 0.001), were more likely to be employed (P = 0.01), and were more likely to report some condom use (P < 0.001). Moreover, women screened in later years were more likely to report a history of genital discharge (P < 0.001) but less likely to report a history of genital ulcer disease (P < 0.001) than those screened in the earlier years. These differences persisted even when the analysis was restricted to women who were married and had only 1 lifetime sexual partner (n = 738); the prevalence of HIV increased steadily in this subgroup as well, from 12% in 1993 to 27% in 2001-2002 (P < 0.001).

Characteristics of Study Population by Calendar Period of Enrollment

Correlates of HIV Infection

In univariate analysis, a higher prevalence of HIV infection was associated with enrollment into the study at later calendar periods, older age, no formal education, having been divorced, separated, or widowed, being currently employed, and having initiated sexual intercourse before 16 years of age (Table 1). Higher HIV prevalence was also associated with history of genital discharge and genital ulcer disease as well as observation of genital ulcer disease on physical examination, but not current genital discharge. In multivariate analysis, independent correlates of a higher HIV prevalence included calendar period (odds ratio [OR], 1.9 for 1997-1999 vs. 1993-1996; 95% CI, 1.2-3.0; OR, 2.3 for 2000-2002 vs. 1993-1996; 95% CI, 1.5-3.6), lack of formal education (OR, 2.0; 95% CI, 1.4-2.9), having been divorced or separated (OR, 3.0; 95% CI, 1.5-5.9), widowed (OR, 3.1; 95% CI, 1.6-6.1), being currently employed (OR, 1.8; 95% CI, 1.2-2.6), and the observation of genital ulcer disease (OR, 1.8; 95% CI, 1.2-2.7) on examination (Table 3). Associations between education, marital status, employment, and genital ulcer disease and HIV did not differ by calendar period.

Unadjusted and Adjusted Odds Ratios of HIV Prevalence Among Women Attending STI Clinics in Pune, India*


We found a high prevalence of HIV among a subgroup of women attending STI clinics who denied engaging in CSW and generally did not report multiple partners. Moreover, the HIV prevalence among these women doubled over a 10-year period from 14% in 1993 to 29% in 2001-2002 and approached the HIV prevalence in men attending STI clinics during this same period (Fig. 1). Although these women are clearly not representative of the general population of women in India, who have a significantly lower HIV prevalence rate (<2%),1 they are an important hard-to-reach population whose high HIV prevalence rate is probably largely due to the high-risk behavior of their husbands. These women are predominantly of childbearing age, with 13% being pregnant at the time of interview, and thus prevention efforts are doubly important not only to reduce their own risk but also to stem the perinatal HIV epidemic.

Prevalence of HIV infection among women who denied sex work and men attending STI clinics in Pune, India between 1993-2002.

High rates of HIV infection among these women are not surprising given that 50% of these women had STIs, which facilitate HIV acquisition,9,13,14 and a high prevalence of HIV among this population of women was reported earlier.5 However, the trend of increasing prevalence over time is unexpected. National data suggest that HIV prevalence rates among STI clinics attendees are stable or decreasing overall and across most high-prevalence states including Maharashtra.1,15 It is important to note that these sentinel surveillance data are not stratified by risk group within STI clinics, nor by gender, and thus might mask differential trends over time in men and women. An examination of trends in our population also reveals a stable prevalence overall, which is being driven by men and CSWs, who constitute >90% of the patients attending these STI clinics and whose HIV prevalence rates have remained stable over 10 years.16,17 Incidence rates in this same population further underscore these trends. Over the same 10-year period, HIV incidence rates decreased among men and CSWs, whereas rates among women who denied CSW remained stable.18 Failure to examine trends separately among risk groups would have masked these differences.

We hypothesize that the women in our study are at risk for HIV infection largely because of the risk behavior of their husbands. Although we were not able to link behaviors of the women to their husbands, data suggested that ∼80% of ∼11,000 men attending these clinics reported ever having sex with a CSW, 65% of whom reported never using a condom during CSW encounters (unpublished data). These low rates of condom use with CSWs, and even lower rates of use with their wives (12%), may explain in part the dynamics of HIV spread in this population. Men in our study represent an important bridge population who bring HIV from CSWs home to their wives, a supposition that is supported not only by our data but also by other studies in India.6,7

These patterns have implications for designing prevention interventions. Early awareness and prevention efforts in India were largely focused on CSWs and their male partners because that is where the HIV epidemic in India began.3,8,19 Stable HIV prevalence over time among these risk groups, based on data from our study and at the national level, suggests that these campaigns have had some success. However, the increasing prevalence among women who denied sex work in our study suggests a need to expand these efforts to include these women when they access care for STIs-and the even harder-to-reach group of women who are at risk but do not seek care for STIs and thus might have no means to obtain HIV testing.

Reaching these women can be challenging. The women in this study initiated sex at young ages and were largely uneducated, both of which were associated with higher HIV prevalence. Furthermore, although many of the women in our study had heard of AIDS, there is ample evidence that awareness of HIV alone may not be sufficient to reduce HIV risk. The 1999 National Family Health Survey in India revealed that even among married women who were aware of AIDS, one-third did not know even one method to avoid infection.20 Similarly, the NACO National Baseline General Population Behavioral Surveillance Survey suggested that although 77% of women sampled in Maharashtra (site of the current study) had heard of AIDS, only 36% of women reported knowing that having one uninfected faithful partner and consistent condom use could protect against HIV.3

Lack of knowledge and education about HIV is further complicated by the fact that even women who are aware of HIV and know how to protect themselves probably have low self-perception of HIV risk. Moreover, even if a woman knows she is at risk, she is often not empowered to negotiate condom use with her husband because of gender norms and the tremendous emphasis on fertility and reproduction in this cultural setting, thus emphasizing the need for female-controlled prevention methods, including microbicides and female condoms. These data collectively suggest that women must be reached early, before they seek care for an STI or antenatal care, at which time they may already be infected with HIV. Because their risk appears to be largely linked to the risk behavior of their husbands, effective prevention initiatives should simultaneously target men and women. Until now, there has been little focus on the marital relationship as a point for HIV prevention/intervention. However, men represent an access point for bringing their wives into care and prevention of HIV. Men coming into care for their first STI should be strongly encouraged to bring their wives for HIV testing and risk reduction counseling. In addition to interventions at the individual-couple level, continued large-scale campaigns using various forms of media including television serials, Bollywood movies, local media, and health camps are also effective means of disseminating information among married women.21,22

We were limited in this study by only having cross-sectional data on correlates and prevalence of HIV infection and thus could not ensure temporality of exposure and outcomes. Moreover, although these changes in prevalence might also reflect parallel changes in incidence, prevalence is not independent of duration of infection, so the increase in prevalence might also represent increased survival of women with HIV. However, because antiretroviral therapy was not widely available in India during the study period, it is unlikely that there were substantial changes in HIV mortality during the study period. We must also consider the possibility of selection bias because women self-selected into this study in seeking care for STIs. Moreover, the underlying population of women attending these clinics changed over time. Referral patterns and awareness of HIV might have changed over time, and it is possible that women in later years were coming to the clinic already knowing they were HIV positive. Compared with those in the early years, women attending these clinics in the later years were more likely to be older, employed, and divorced, separated, or widowed, all of which were associated with higher HIV prevalence. However, inclusion of these factors in multivariate models could not fully explain the increase in HIV prevalence. Moreover, this shift in the population likely reflects a consequence of HIV rather than a cause, reflecting the devastating effect of HIV on families in India. Women in this population are probably becoming widowed or abandoned at an increasing rate over time because of HIV, leaving them alone without a support structure to provide for their children and families. Finally, these data are not generalizable to the general population of women in India but rather to high-risk women attending public STI clinics. Although we sampled women from only 3 public STI clinics in Pune, we do not have reason to believe that the women attending these particular clinics would be different from those attending other public clinics.

In conclusion, we observed a high and increasing prevalence of HIV infection among a subgroup of high-risk women attending STI clinics who did not report CSW or multiple sex partners. Awareness and education efforts in India that have successfully reached other high-risk groups, including CSWs and their partners, need also to be targeted toward this often-overlooked population. Efforts to prevent and detect HIV infection early among these women not only are crucial for the women themselves, but risk reduction of this subgroup of Indian women would also be expected to reduce perinatal infections in India.


The authors thank Lidia Propper, Radhika Brahme, and Dr. Manisha Ghate for their valuable contributions to this paper. We also thank all study staff of the National AIDS Research Institute, HIVNET, and PAVE for their participation in data collection, clinical care, and laboratory work. Finally, many thanks to our study participants.


1. National AIDS Control Organization. HIV estimates 2003. Available at: Accessed November 1, 2004.
2. Mahapatra. India overtakes South Africa as country with most HIV cases. Seattle Post-Intelligencer. September 16, 2004.
3. National AIDS Control Organization. National baseline general population behavioral surveillance survey. Available at: Accessed November 1, 2004.
4. National AIDS Control Organization. HIV/AIDS Indian scenario: HIV/AIDS surveillance in India (as reported to NACO on 31, September 2003). Available at: Accessed November 1, 2004.
5. Gangakhedkar RR, Bentley ME, Divekar AD, et al. Spread of HIV infection in married monogamous women in India. JAMA. 1997;278:2090-2092.
6. Panda S, Chatterjee A, Bhattacharya SK. Transmission of HIV from injecting drug users to their wives in India. Int J STD AIDS. 2000;11:468-473.
7. Newmann S, Sarin P, Kumarasamy N, et al. Marriage, monogamy and HIV: a profile of HIV-infected women in south India. Int J STD AIDS. 2000;11:250-253.
8. National AIDS Control Organization. NACO policies: national AIDS Prevention and control policy. Available at: Accessed August 11, 2003.
9. Mehendale SM, Rodrigues JJ, Brookmeyer RS, et al. Incidence and predictors of human immunodeficiency virus type 1 seroconversion in patients attending sexually transmitted disease clinics in India. J Infect Dis. 1995;172:1486-1491.
10. Larsen SA, Steiner BM, Rudolph AH. Laboratory diagnosis and interpretation of tests for syphilis. Clin Microbiol Rev. 1995;8:1-21.
11. World Health Organization. Guidelines for the management of sexually transmitted infections. Available at: Accessed November 1, 2004.
12. Interpretation and use of the Western blot assay for serodiagnosis of human immunodeficiency virus type 1 infections. MMWR Morb Mortal Wkly Rep. 1989;38:1-7.
13. Plummer FA, Simonsen JN, Cameron DW, et al. Cofactors in male-female sexual transmission of human immunodeficiency virus type 1. J Infect Dis. 1991;163:233-239.
14. Reynolds SJ, Risbud AR, Shepherd ME, et al. Recent herpes simplex virus type 2 infection and the risk of human immunodeficiency virus type 1 acquisition in India. J Infect Dis. 2003;187:1513-1521.
15. National AIDS Control Organization. State wise HIV prevalence (1998-2003). Available at: Accessed April 28, 2005.
16. Gupta A, Mehta S, Joglekar N, et al. Same sex behavior, sexually transmitted infections and HIV risks among men attending STD clinics in Pune, India, 1993-2002. Paper presented at: XV International AIDS Conference, Bangkok, Thailand; July 11-16, 2004. Abstract WePeC6109.
17. Brahme RG, Mehta SH, Sahay S, et al. Increased insistence of condom use and stable HIV prevalence among female sex workers attending sexually transmitted disease clinics in Pune, India. Paper presented at: XV International AIDS Conference, Bangkok, Thailand; July 11-16, 2004. Abstract WePeC6205.
18. Bollinger RC, Shepherd ME, Ghate MV, et al. Declining HIV-1 incidence rates among high-risk persons in Pune, India: 1993-2002. Paper presented at: XV International AIDS Conference, Bangkok, Thailand; July 11-16, 2004. Abstract MoPpC2016.
19. National AIDS Control Organization. Combating HIV/AIDS in India, 2000-2001. Available at: Accessed August 12, 2003.
20. National Family Health Survey-2, 1998-1999. HIV/AIDS awareness among women. Available at: Accessed September 22, 2003.
21. Gupta A, Saxena A, Rathore P, et al. Massive information education communication (IEC) campaigns as a strategy to reduce the spread of HIV in rural India. Paper presented at: XV International AIDS Conference, Bangkok, Thailand; July 11-16, 2004. Abstract MoPeE4253.
22. Chatterjee N. AIDS-related information exposure in the mass media and discussion within social networks among married women in Bombay, India. AIDS Care. 1999;11:443-446.

HIV; AIDS; women; sexually transmitted diseases; prevalence; India

© 2006 Lippincott Williams & Wilkins, Inc.