Silverman, Jay G PhD*; Decker, Michele R MPH*; Gupta, Jhumka MPH*; Maheshwari, Ayonija MD, MPH†; Patel, Vipul‡; Raj, Anita PhD§
The trafficking of human beings, defined by United Nations as “…the recruitment, transportation, transfer, harboring or receipt of persons, by means of the threat or use of force or other forms of coercion, of abduction, of fraud, or deception…for the purpose of exploitation [including]…prostitution of others or other forms of sexual exploitation,”1 has emerged as a major international health and human rights concern.1,2 Although the underground nature of trafficking makes accurate assessment of its magnitude difficult,3,4 it is conservatively estimated that between 600,000 and 800,000 people are trafficked across international borders annually, with 80% of these being women and girls.2 Women and girls are at particular risk for sex trafficking, that is, trafficking for purposes of sexual exploitation (eg, forced prostitution, sexual slavery). Once considered a more general form of economic and community violence, the United Nations Special Rapporteur on Violence Against Women has underlined sex trafficking as a critical aspect of the global epidemic of violence against women.1
Of the hundreds of thousands of women and girls sex trafficked annually, the largest number originate in Asia, with 150,000 estimated to come from South Asia alone.5 In India and worldwide, it is estimated that half of all trafficked women are <18 years of age.2,6,7 This condition represents a clear violation of Indian federal statutes protecting minors from sex trafficking, sexual exploitation, and sex work8 as well as a violation of the the United Nations Convention on the Rights of the Child.9 Within Asia, India has been designated as a major destination country for sex-trafficked women and girls,2,10 with large numbers of Nepali and Bangladeshi women and girls, as well as Indian women and girls from more rural and impoverished areas trafficked to major Indian cities annually.2,7,10 The city of Mumbai is considered to be the largest center of sex trafficking within India.2
India's HIV epidemic, currently second only to sub-Saharan Africa in its total number of HIV infections and annual AIDS deaths,11 is thought to be uniquely propelled by the sex trade.12 Contact with commercial sex workers poses a 70% increased risk of HIV among Indian men,13 likely because of inconsistent condom use demonstrated among men seeking sex in this context.7,14-18 Despite recent practical and research attention devoted to commercial sex work as a mechanism for the epidemic,12,13 evidence of high rates of sex trafficking to this region, and numerous HIV risks associated with the experience of sex trafficking (eg, physical and sexual abuse, including forced unprotected sex, widely recognized as inherent to the experience of trafficking),2,10,19 little investigation has been conducted among known victims of sex trafficking regarding HIV.
The prevalence of HIV among sex workers in South Asia varies widely, with higher rates identified in clinic-based samples (30%-50%)20-23 as compared to community-based or brothel-based studies (6%-47%).24-26 The sole investigation specific to sex-trafficking victims in India documented HIV prevalence rates of 17.8% among those rescued from sex work.7 Reliance on self-reported HIV status in this study may have underestimated HIV rates among participants, however, given the low rates of HIV testing27 documented among sex workers.
Beyond the lack of data specific to victims of trafficking, young sex workers (ie, those <20 years of age) are underrepresented within most South Asian studies of HIV prevalence and risk. Despite the aforementioned high rates of sex trafficking among this age group and research indicating that sex workers in this age category may be at increased vulnerability for HIV infection,25,26 fewer than 10% of participants <20 years of age have been included in major studies reviewed.20,21,23,25,26,28 This underrepresentation of young sex workers greatly limits our ability to generate reliable HIV prevalence estimates for this vulnerable population. Poor representation of younger sex workers may reflect a lack of access to outside assistance for this group.
Further, the posited barriers to seeking health services or aid of any kind faced by sex workers (eg, stigma)27 are likely compounded for sex-trafficking victims via mechanisms reported among this population such as psychologic abuse, restrictions on movement, and social isolation.6,19 These factors may be particularly salient for younger women29 and may diminish the likelihood of young trafficked women participating in investigations or surveillance attempts. Additionally, major studies of HIV-related interventions have excluded minors because of ethical concerns,28 further preventing collection of data, and perhaps assistance, regarding the health of this large sector of individuals made vulnerable through sex work.
In sum, although young women trafficked to India for sex work may represent a distinct and vulnerable population regarding HIV risk, little data on this group currently exist within the public health literature. The purpose of the present study is to examine the prevalence and predictors of HIV infection among a sample of sex-trafficked women and girls recently rescued from brothels in Mumbai, India.
Case records and medical documentation of residents at a major nongovernmental organization (NGO) providing rescue, shelter, and care of minor girls and women held against their will in brothels in Mumbai were reviewed. Residents were rescued in brothel raid operations across 8 major “red light” areas in Mumbai, in which minor girls and individuals residing in brothels against their will were removed by local police authorities in collaboration with NGO staff. Individuals appropriate for rescue were identified through a complex network of informants, including local pimps, police, undercover NGO staff members, and formerly rescued trafficking victims. Once taken into police custody and processed, women and girls were brought to the NGO for temporary shelter, acute health care, counseling, legal assistance, and protection from brothel owners before being repatriated to a longer term rehabilitation facility in their Indian state or nation of origin. Upon intake and pending their consent, all residents received a basic medical examination and HIV testing conducted by medical professionals affiliated with the NGO. HIV testing typically occurred 4 to 7 days after rescue. Individuals testing positive for HIV were provided with HIV education and counseling to prevent transmission, medical treatment when NGO funding permitted, and hospice care as appropriate in Mumbai or in their region of origin. Additionally, all residents were interviewed by NGO case managers at intake to assess and document their trafficking history and conditions of work within the brothel.
Case and medical records for all residents of the NGO who were rescued between December 2002 and July 2005 were reviewed (N = 227). Records eligible for inclusion in the current study were those indicating (1) girls and women trafficked into sex work and (2) HIV laboratory test results. Of the original 227 records collected across the study period, 16 records were excluded based on their referring to men or individuals sheltered for reasons other than sex trafficking (eg, homelessness, child labor trafficking). Of the remaining 211 records, an additional 36 were excluded because of lack of inclusion of HIV laboratory test results; the resulting 175 records constituted the final sample. All 175 individuals indicated to police and NGO staff that they had been coerced or forced into commercial sex work and had been transported to Mumbai for this purpose.
Demographics, experiences of sex trafficking and sex work, and HIV status (as determined through documented enzyme-linked immunosorbent assay [ELISA] or rapid testing for HIV type I and type II) among rescued women and girls were systematically recorded by NGO staff into case and medical records; these records were abstracted by research staff for purposes of the present analyses. Demographic data (region and/or country of origin, marital status, and educational attainment) were obtained from case records. Age at the time of rescue was obtained from the case and medical records; a medical practitioner contracted by police verified age via radiographs of joints and assessment of secondary sexual characteristics and dentition. Duration of service in brothels and typical number of clients per day were also abstracted from case records; when a range was reported, the mean was taken. Age at trafficking was calculated based on duration of brothel stay and age at rescue. Protocols described were approved by the Harvard School of Public Health Human Subjects Committee.
Descriptive statistics were calculated for all demographic variables. Prevalence rates for HIV-seropositive status were calculated for the total sample. Differences in HIV status based on nationality, Indian state of origin (where relevant), and marital status were assessed via χ2 tests based on the categoric nature of these variables. Differences in HIV status based on continuous variables (age at trafficking, brothel duration, and clients per day) were assessed via Student t tests. Logistic multivariate regression models were constructed to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for HIV prevalence based on those variables found to relate to HIV prevalence in bivariate analyses.
Sample Demographics and Experiences of Sex Work
Slightly more than half of rescued women and girls (52.6%) were trafficked within India, most often from West Bengal (63.0% of those from India) (Table 1). The next largest nation of origin was Nepal (30.3%). More than one quarter (28.6%) reported being married, with another 10.8% reporting being separated, abandoned, divorced, or widowed. The age of women and girls at rescue ranged from 9 to 30 years (mean = 18.8 years). The mean time served in brothels at rescue was 15.8 months, with more than two thirds (69.3%) serving in brothels for 1 year or less and 40.9% reporting brothel confinement for <6 months. Age at trafficking ranged from 8 to 29 years (mean = 16.8 years); 65.9% of rescued women and girls were trafficked to brothels while <18 years of age.
HIV Prevalence and Associations With Demographics and Experiences of Sex Work
Approximately 1 in 4 (22.9%) rescued sex-trafficked women and girls tested positive for HIV (Table 1). Indian women and girls trafficked from the states of Karnataka and Maharashtra were more likely to test positive for HIV than those trafficked from other Indian states (P = 0.003). Marginal significance was indicated for the association of HIV status and age at trafficking; mean age at trafficking was younger for HIV-positive (15.9 years) as compared to HIV-negative women and girls (17.2 years; P = 0.06). Women and girls who tested positive for HIV were significantly more likely to report longer duration in brothels; the mean brothel duration for HIV-positive individuals was 27.6 months as compared to 11.6 months for those who tested negative (P < 0.001). No differences in HIV status based on nationality, marital status, or number of sex work clients per day were identified.
Multivariate Models Predicting HIV Status
No collinearity was detected among the variables of Indian state of origin, age at trafficking, or brothel duration. Thus, these variables were entered singly and in combination into logistic regression equations predicting HIV status (Table 2). Based on similarly high HIV prevalence estimates for the states of Maharashtra and Karnataka and the small numbers of individuals included from states other than these 2 states or West Bengal, Indian state of origin was dichotomized as Maharashtra or Karnataka versus West Bengal for purposes of regression analyses. Brothel duration emerged as an independent predictor of HIV status in models adjusted for age at trafficking (OR = 1.04, 95% CI: 1.01 to 1.06) and, among Indian trafficking victims, for age at trafficking and Indian state of origin (OR = 1.03, 95% CI: 1.00 to 1.07), indicating a 3% to 4% increased risk of HIV for each additional month of brothel duration. Indian state of origin was also found to predict HIV status in a model adjusted for age at trafficking and brothel duration, with Indian women and girls trafficked from Maharashtra or Karnataka far more likely to be infected than those trafficked from West Bengal (OR = 7.35, 95% CI: 2.23 to 24.21).
Post Hoc Analyses
Tests for collinearity revealed a strong inverse correlation of age at trafficking to duration of sex work (Pearson correlation = −0.438; P < 0.001); that is, young age at trafficking was associated with longer brothel duration. Post hoc analyses of this relation indicated that girls trafficked while minors (ie, <18 years of age) reported significantly longer periods of brothel service (mean = 18.5 months) as compared to those trafficked at the age of 18 years or older (mean = 9.6 months; P = 0.007).
The current study indicates a high prevalence of HIV infection (22.9%) among this sample of sex-trafficked South Asian young women and girls, with increased risk for HIV among those trafficked from Indian states with a high HIV prevalence, those reporting longer duration in brothels, and those trafficked at the age of 17 years or younger. Our finding of a positive association between brothel duration and HIV status, with an increased HIV risk of 3% to 4% for each additional month in brothel captivity, contrasts with prior work demonstrating no relation between duration and HIV status.25 The discrepancy in these findings is likely related to differences in sampling; the earlier study was not specific to trafficked women and, importantly, included an older sample (5% reported to be 20 years of age or younger) reporting longer brothel tenure. Within the previous sample, 53% had worked in brothels for more than 5 years; in the current sample, 69% had served in brothels for 1 year or less. We hypothesize that the inclusion of greater numbers of younger individuals and those serving extremely short periods as well as the overall higher prevalence of HIV detected among the current sample (22.9% vs. 9.6% for the previous study) allowed for better detection of the relation of HIV to length of exposure. A critical implication of this finding is demonstration of the benefit of rescue of sex-trafficked young women and girls. Greater efforts to identify and rescue sex-trafficked individuals, particularly those recently trafficked, may substantially reduce HIV infection rates among this population. Currently, relatively few programs in India focus on the identification and rescue of trafficked sex workers.
Building on previous research indicating increased HIV risk among young sex workers,25 results of the current study suggest that those trafficked at a younger age may be at increased risk for HIV infection. Although the bivariate relation observed was no longer significant when brothel duration was included in the multivariable model, this loss is likely attributable to the finding that girls who were trafficked before the age of 18 years (ie, as minors) were trapped in brothels for longer periods than those trafficked at an older age. Heightened risk of HIV infection among younger sex workers may stem from the qualitatively unique experiences of younger girls within a brothel setting and biologic vulnerability. Younger girls may have diminished power to demand and/or negotiate condom use by clients and are often kept carefully hidden in the recesses of brothels, restrained through confinement, and regularly moved from brothel to brothel to avoid detection.7 Hence, it seems that rescue of younger girls typically occurs after a relatively longer period, leading to increased risk for HIV among these individuals. Longer brothel captivity represents a possible mechanism for the increased HIV vulnerability among younger sex workers documented in prior work.25 Further, although younger women are kept hidden from authorities, prior research indicates that they are in high demand among clients.7 Younger girls and virgins are perceived as more likely to be free of HIV and other infections.7 The prevalent misconception that having sex with a virgin is curative of such conditions7 further illustrates potential HIV risk mechanisms for young women. Finally, as posited in a recent study that identified increased HIV risk among younger sex workers,25 biologic vulnerability based on larger areas of cervical ectopy attributable to repeated trauma to an immature genital tract during sexual intercourse may also be a factor in the current findings of increased risk for HIV infection among younger sex-trafficking victims.
Within the currently examined Indian majority subsample, women and girls trafficked from the states of Karnataka or Maharashtra were >7 times more likely to be HIV-seropositive subsequent to rescue from brothels as compared to those trafficked from West Bengal. Explanations for this great disparity in HIV infection likely stem from the relatively higher HIV prevalence rates for the formerly mentioned Indian states. According to sentinel surveillance conducted in 2004 by the National AIDS Control Organization (NACO) of India,30 10.4% to 12.0% of sexually transmitted disease (STD) patients and 1.3% of antenatal care patients in the states of Karnataka and Maharashtra were seropositive for HIV. In contrast, 0.9% of STD patients and 0.5% of antenatal care patients in West Bengal were found to be infected during this same year. Women and girls trafficked from states with high HIV prevalence may have been more likely to be infected based on exposure in their state of origin previous to or during trafficking. These individuals may also be more vulnerable to HIV infection based on the possible greater number of brothel clients from their home states, who would thus be more likely to be seropositive. Currently, no data exist on whether male brothel clients are more likely to select sex workers from their home states. Finally, women and girls from these high-prevalence states may be treated differently than other trafficking victims while in brothels, such that they are placed at higher risk for HIV infection (eg, being made to serve larger numbers of clients or to have intercourse without a condom more often as compared to those from other states). Such differential treatment may be based on state-based bigotry on the part of brothel owners or clients. Again, no data currently exist on whether sex workers experience such discriminatory treatment. Future research should include examinations of the existence and nature of state-based disparities in HIV prevalence among trafficked sex workers, including Indian state of origin as a potential factor in HIV risk.
The current study also documented a relatively higher rate of potential HIV exposure based on the number of clients reported per day. Trafficked women and girls in the current sample reported an average of >7 sex clients on a typical day as compared to <4 clients per day reported in previous work among older sex workers.25 Thus, increased vulnerability to infection among young trafficked sex workers may stem from multiple sources.
As discussed previously, most previous research into the health of sex workers in India has failed to assess whether individuals were trafficked into sex work. Distinguishing sex workers who have been trafficked from those not trafficked is essential to understanding the health impact of this coerced or forced transport and servitude as well as to guiding appropriate prevention strategies.4,31 Further investigation is needed to assess the role of sex trafficking in HIV risk among young sex workers in comparison with other types of entry into sex work, with particular attention to the role that sexual and physical violence in the process of trafficking may play in potentially increased HIV risk among trafficking victims. Given the present international focus on the elimination of trafficking and the prevention of HIV, collection of data to answer these basic questions regarding the relevance of trafficking to the health of sex workers must be prioritized.
Finally, the behavior of male clients and resultant health implications for trafficked women cannot be overlooked. Prior research indicates that the aforementioned demand for younger girls and virgins among male brothel clients is so great that traffickers are reportedly paid more than double the price for a virgin compared with other girls.7 Further, given consistent evidence of client-perpetrated violence against sex-trafficked women19 and client condom nonuse,7,14-18 investigations focused on male clients must be undertaken to elucidate the behavior of men seeking sex in this context so as to better understand health risks posed to young victims of sex trafficking and to guide primary prevention of this human rights abuse.
The major limitation of the current study is reliance on a rescued sample (eg, it is possible that individuals rescued differ in experiences of sex work and HIV risk as compared to those trafficked but not rescued). Although future research on sex-trafficked women and girls may assist in assessing this concern, investigations involving nonrescued trafficked individuals are neither feasible nor ethical, given the criminal nature of this circumstance. A further notable limitation relates to the practice of testing for HIV within 1 week after rescue. Given exposure to sex work of periods of 2 months or less among 22.2% of the current sample and an estimated average time after contact to production of detectible antibodies of several weeks, it is likely that testing was premature for many of these individuals. Thus, the prevalence of HIV infection currently reported likely underestimates the true rate of infection for this sample. Future research in this area should include testing a sufficient period after exposure to improve the validity of such estimates. Data concerning condom use, the major method for HIV prevention, were not available within the current study. Nevertheless, given prior evidence of inconsistent condom use in commercial sex work encounters7,14-18 and the likely diminished condom negotiation power among young sex-trafficked women, condom use is likely to have been low within the current sample. Further, factors identified as relevant to HIV risk may not be unique to trafficking victims but, instead, may apply to the larger population of sex workers. Continued and expanded research is needed into the conditions and risks faced by trafficking victims and whether and how these may differ from those for nontrafficked sex workers. Finally, the current study is based on a sample from a single large agency in a single major Indian city. Samples from multiple cities comprising greater numbers of individuals should yield greater statistical power to detect differences as presently attempted and greater generalizability of findings to the larger population of sex-trafficking victims.
In sum, the findings of the present study highlight the complexities of HIV vulnerability among young women trafficked for sex work and underscore the need for greater attention within research, surveillance, and programming to inclusion of younger sex workers and victims of sex trafficking so as to expand the state of knowledge and practice concerning the prevention of sex trafficking and HIV infection. In contrast to the historically narrow focus of public health on victims of sex trafficking as sex workers involved in the spread of the HIV epidemic, a comprehensive research, policy, and practical agenda is required. Future efforts should include (1) elucidation of pretrafficking vulnerabilities (ie, risk factors) and mechanisms for trafficking; (2) development and evaluation of primary trafficking prevention efforts based on such knowledge; (3) investigation into the epidemiology of and programmatic efforts to reduce the demand aspect of sex trafficking (ie, male brothel clients); and (4) improved provision of assistance to victims of sex trafficking through greater efforts to identify and rescue such individuals, expanded short-term and longer term programs to assist in their recovery, and, where possible, social reintegration. Such programs currently exist in small numbers in only a few locales and without adequate support to reach significant numbers of trafficking victims. Through strengthening the body of empiric data and programmatic support in these critical areas, we greatly enhance the potential of practitioners and policy makers to develop legal and public health strategies for effectively reducing the intertwined epidemics of HIV and sex trafficking plaguing India and many other nations.
The authors acknowledge the late Balkrishna Acharya and the staff of the Rescue Foundation for their provision of data and extensive cooperation in facilitating this investigation.
1. United Nations. United Nations Protocol to Prevent, Suppress, and Punish Trafficking in Persons, Especially Women and Children. Supplement to the United Nations Convention Against Transnational Organized Crime
. Article 3 (a-d), GA res. 55/25, annex II. 55 US GAOR Suppl (No. 49) at 60, UN Doc A/45/49, vol. I. United Nations, Geneva, 2000.
2. US Department of State. Victims of Trafficking and Violence Protection Act of 2000: Trafficking in Persons Report
. US Department of State, Washington, DC; 2005.
3. Beyrer C. Is trafficking a health issue? Lancet
4. Loff B, Sanghera J. Distortions and difficulties in data for trafficking. Lancet
5. Congressional Research Service. Trafficking in Women and Children: The U.S. and International Response
. Washington, DC: Library of Congress; 2001.
6. Hennink M, Simkhada P. Sex trafficking in Nepal: context and process. Asian Pac Migr J
7. Nair PM. A Report on Trafficking of Women and Children in India: 2002-2003
, vol. I. New Delhi, India: Institute of Social Sciences, National Human Rights Commission, United Nations Development Fund for Women; 2004.
8. Indian Penal Code. Immoral Trafficking Prevention Act
. New Delhi, India, 1986.
9. United Nations. Convention on the Rights of the Child
, vol. Article 34. United Nations, Geneva, 1989.
10. Human Rights Watch. Rape for Profit: Trafficking of Nepali Girls and Women to Indian Brothels
. New York: Human Rights Watch; 1995.
11. United Nations. 2004 Report on the Global AIDS Epidemic, Fourth Global Report
. Geneva: United Nations; 2004.
12. 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
13. Rodrigues JJ, Mehendale SM, Shepard ME, et al. Risk factors for HIV infection in people attending clinics for sexually transmitted diseases in India. BMJ
14. Sharma V, Sharma A, Dave S, et al. Sexual behaviour of adolescent boys-a cause for concern. Sex Marital Ther
15. Bansal RK. Sexual behaviour and substance use patterns amongst adolescent truck cleaners and risk of HIV/AIDS. Indian J Matern Child Health
16. Chakraborty AK, Jana S, Das A, et al. Community based survey of STD/HIV infection among commercial sex workers in Calcutta (India). Part 1. Some social features of commercial sex workers. J Commun Dis
17. Madhivanan P, Hernandez A, Gogate A, et al. Alcohol use by men is a risk factor for the acquisition of sexually transmitted infections and human immunodeficiency virus from female sex workers in Mumbai, India. Sex Transm Dis
18. Hosain GM, Chatterjee N. Beliefs, sexual behaviours and preventive practices with respect to HIV/AIDS among commercial sex workers in Daulatdia, Bangladesh. Public Health
19. Zimmerman C, Yun K, Shvab I, et al. The Health Risks and Consequences of Trafficking in Women and Adolescents: Findings from a European study
. London: London School of Hygiene and Tropical Medicine; 2003.
20. Brahme R, Mehta S, Sahay S, et al. Correlates and trend of HIV prevalence among female sex workers attending sexually transmitted disease clinics in Pune, India (1993-2002). J Acquir Immune Defic Syndr
21. Gangakhedkar RR, Bentley ME, Divekar AD, et al. Spread of HIV infection in married monogamous women in India. JAMA
22. 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 Center in Manipur, India. Int J STD AIDS
23. 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
24. Bhave G, Lindan CP, Hudes ES, et al. Impact of an intervention on HIV, sexually transmitted diseases, and condom use among sex workers in Bombay, India. AIDS
. 1995;9(Suppl 1):S21-S30.
25. Sarkar K, Bal B, Mukherjee R, et al. Young age is a risk factor for HIV among female sex workers-an experience from India. J Infect
. 2005 [Epub ahead of print].
26. Sarkar K, Bal B, Mukherjee SK, et al. Epidemiology of HIV infection among brothel-based sex workers in Kolkata, India. J Health Popul Nutr
27. Dandona R, Dandona L, Kumar AG, et al. HIV testing among female sex workers in Andhra Pradesh, India. AIDS
28. Gangopadhyay DN, Chanda M, Sarkar K, et al. Evaluation of sexually transmitted diseases/human immunodeficiency virus intervention programs for sex workers in Calcutta, India. Sex Transm Dis
29. Willis BM, Levy BS. Child prostitution: global health burden, research needs, and interventions. Lancet
30. NACO. State wise HIV prevalence 1998-2004. Available at: http://www.naco.nic.in
. Accessed September 19, 2006.
31. Saunders P. Traffic violations: determining the meaning of violence in sexual trafficking versus sex work. J Interpers Violence
© 2006 Lippincott Williams & Wilkins, Inc.