Migration and HIV in India: Study of Select Districts31 Dec 2011
The report presents results of a study undertaken by the Population Council with support from UNDP and National AIDS Control Organisation that examines the role of migration in the spread of HIV in districts with high out-migration in India.
In India, efforts of the National AIDS Control Programme have been successful in reducing overall HIV incidence in the country by 50 percent with focused interventions with female sex workers (FSWs), men who have sex with men (MSM) and injecting drug users (IDUs).
Despite these achievements, HIV prevalence is still estimated at over 1 percent in some population sub-groups and some sub-regions in the country. Further, while HIV prevalence is estimated to be 0.31 percent among the general population in 2009, it varied considerably across states and population sub-groups. For example, the percentage of women who reported HIV in India increased from 25 percent in 2001 to 39 percent in 2009. More than 90 percent of the new infections among females were occurring among non-commercial sex work related relationships, mostly within marriage.
Additionally, some districts in the states that showed a low HIV prevalence a decade ago have shown a slow and steady increase since 2007. Many of these districts also experience high male out-migration to other states. Increasing HIV infection among married women and in districts with high out-migration provided the impetus for this study. The goal of the study was to examine the linkages between male out-migration and HIV transmission in married men and women and to explore other mechanisms by which HIV is transmitted within marital relationships in districts with high out-migration.
In order to understand the linkages between migration and HIV, a case-control study design was used. Cases were currently married HIVpositive persons and controls were currently married HIV-negative persons. Both cases and controls were tested for HIV six months prior to the survey. HIV-positive persons were recruited from integrated counselling and testing centres (ICTCs) and from antiretroviral therapy (ART) centres. HIV-negative persons were recruited from ICTCs.
Separate studies were conducted on married men and married women in three groups of districts with high out-migration. The districts included in the study were: (1) Ganjam district where HIV infection among women attending antenatal care (ANC) clinics was more than 3 percent and men migrated predominantly to states/districts with high HIV prevalence (Surat in Gujarat and Mumbai and Thane in Maharashtra); (2) districts in northern Bihar where HIV infection among women was on the rise and men migrated primarily to states with low HIV prevalence (Delhi, Haryana and Punjab); and (3) districts in Eastern Uttar Pradesh (UP) where HIV infection among women was on the rise and men primarily migrated to states with high HIV prevalence (Mumbai and Thane in Maharashtra).
The participants from three study areas (Ganjam district: 414 males, 407 females; Northern Bihar districts: 396 males, 402 females; Eastern UP districts: 396 males, 400 females) were interviewed by trained research assistants after obtaining written informed consent. The study protocol was approved by ethics committee of the National AIDS Control Organisation (NACO) as well as the Institutional Review Board of the Population Council.
The findings show that all three study areas had demonstrated an unprecedented out-migration xiv ¦Migration and HIV in India: Study of select districts of rural men for work to cities. The data from the men’s survey reveals that there were more migrant men among HIV-positive than among HIV-negative populations. Thus, in northern Bihar, migrants accounted for 89 percent of the HIV-positive group compared to 59 percent of the HIV-negative group. Similarly, in Ganjam district migrants accounted for 79 percent of the HIV-positive compared to 41 percent of the HIV-negative group. These proportions were 73 percent and 32 percent, respectively, in Eastern UP.
Separate studies of married women conducted in the three study areas also showed similar results. There was a significant association between spousal migration and women’s HIV status. Significantly more HIV-positive than HIV-negative women had migrant husbands. Female out-migration for work was, however, low. It ranged between 0.5 percent in Eastern UP to about five percent each in northern Bihar and Ganjam in Orissa. Since the number of female migrants in the study sample was very low, no statistical association could be observed between female out-migration and HIV status.
The study showed that in northern Bihar, odds of HIV infection were eight times higher among migrant men than nonmigrant men, even after controlling for age, education, source of referral and other possible confounding factors. In the Eastern UP and Ganjam districts, migrant men were almost four times more likely to contract HIV than non-migrants. More importantly, returned migrant men had a higher likelihood of being infected as compared to non-migrant men. In northern Bihar, the odds of HIV infection were 13 times higher among returned migrant men compared to non-migrant men. The survey of women from the three study areas also showed that the odds of HIV infection were higher among women with migrant husbands than among women with non-migrant husbands.
The comparison of data between men’s and women’s survey revealed that the association between husbands’ out-migration and women’s HIV status is weaker than the case of male outmigration and their own HIV status. The study showed that there was a higher likelihood of HIV exposure among migrants visiting certain destinations. For example, migrants who went to the corridors of Mumbai, Thane and Surat districts from Eastern UP and Ganjam districts and migrants from Northern Bihar who went to Kolkata and Mumbai/Thane districts had higher HIV rates than those migrants who went to other districts. The study suggests a considerable spread of HIV linked to migrants’ extramarital sexual behaviors. However, local sexual networks also probably played an important role in spread of HIV for women as well as men. The findings revealed a higher proportion of migrant than nonmigrant men reporting having had extramarital sex in all three corridors of migration.
Among migrants, higher proportions of returned migrants than active migrant men reported extramarital sex. The study on married women also showed that a considerable proportion of women across all three study areas reported having had extramarital sex. Current extramarital sex among women has been reported at these high levels in India for the first time and perhaps these levels are not generalizeable because these women in the study were recruited from ICTCs. Surprisingly, even women with non-migrant husbands reported having had extramarital sex in Ganjam district. The proportion was similar for women with currently active migrant husbands.
In all study areas, condom use during the last extramarital sex act was low among both men and women, underscoring the need for immediate programmatic attention to ensure safe sex practice in districts with high out-migration. HIV sero-positive concordance was high among married women in all the three study areas suggesting that women were being infected by their migrant husbands. HIV male sero-discordance (when the husband is HIVpositive and wife is HIV-negative) was high in all study areas suggesting that there is a need to focus on sero-discordant couples to prevent further transmission of HIV. Data from the men’s survey indicated that a significant proportion of men tested HIV-positive but their spouse’s HIV status was not known; this suggests that there is an urgent need to promote Executive Summary¦ xv partner notification for HIV-testing.
Without partner notification, if the wife’s status remains unknown and she is HIV-positive she could either fail to receive treatment or her treatment could be delayed. If the wife is HIV-negative and having unprotected sex with partner, then she is at risk of acquiring HIV particularly if the spouse is not receiving ART.