ORIGINAL RESEARCH ARTICLE HIV-1 discordant couples in Karnataka, South India: is the sex ratio of index cases changing? S Shastri MBBS*, A Shet MD †‡ , B Rewari MD* and A De Costa MD PhD ‡ *National AIDS Control Organization, New Delhi, India; † St John’s National Academy of Health Sciences, Bangalore, India; ‡ Division of Global Health, Nobels Va ¨ g 9, Karolinska Institutet, Stockholm SE 171 77, Sweden Summary: India has an estimated 2.5 million HIV infections, most of which are heterosexually transmitted. Women comprise 40% of infected adults. In India, 90% of women between the ages of 15 and 45 years are married. Previous literature has suggested that sexual intercourse with an HIV-infected husband represents a married woman’s greatest risk for being infected. However, a recent meta-analysis of discordant couples from sub-Saharan Africa reported that women were the index case in half of all couples. Similar data are not available from India. This cross-sectional study describes the epidemiology of 925 discordant couples from five districts in Karnataka province, one of the high HIV prevalence provinces in India. Men were the index case in 74% of couples. However, in young couples (where the index case was aged ,30 years), women were more likely to be the infected partner (64% of couples). Condom use was reported by 46% of these discordant couples. These results suggest an emerging predominance of female index case infections among younger discordant couples in India, and point to the need for focusing HIV preventive messages on youth and couples before marriage. Keywords: HIV, transmission, discordant couples, sexual behaviour, India INTRODUCTION Recent analyses demonstrate that married or cohabitating discor- dant couples are an important source of new HIV infections in different regions of the world. Population surveys and math- ematical modelling studies suggest that the majority of new HIV-1 transmission in parts of Africa occurs within stable discor- dant partnerships. 1 In India, estimates suggest that 2.4 million persons are infected with HIV, 40% of whom are women. Approximately 87% of adult HIV infections in the country result from heterosexual transmission. 2 Early and universal mar- riage of women in the country is common, resulting in 90% of Indian women between the age group 15–44 years being married. 3 Literature has suggested that sexual intercourse with an infected husband represents a woman’s greatest risk for being infected with HIV, 4,5 assuming that men are more likely to practice high-risk sexual behaviours. 6,7 The dominance of men in sexual negotiations, greater opportunities for casual sexual encounters, existing social norms and gender relations lend support to this view. A recent report on the epidemiology of discordant couples in sub-Saharan Africa reported that contrary to the common per- ception of men mostly being the index case, 47% of index case infections were in women. 8 There have so far been few reports on the characteristics of discordant couples from India outside of studies that have reported such data from single clinics. 9,10 This cross-sectional study, from five districts of Karnataka is comprised of all discordant couples obtaining medical care in district clinics under the National AIDS Control Programme of the Government of India. A descriptive epidemiology of these discordant couples is presented. Besides the age–sex dis- tributions of index case infections, other characteristics includ- ing sociodemographic variables, contraceptive use, mode of transmission and HIV testing of children of these couples are also described. METHODS Study setting The study was based on data obtained from five districts of Karnataka province in South India. Districts are administrative units within a province; each district has approximately a million residents. The health-care delivery units of the National AIDS programme that provides diagnosis and treat- ment of HIV are located at the level of the district, and are attached to the respective district hospital; these units are known as antiretroviral treatment centres or ART centres. In four districts, i.e. Chikmagalur, Madikere, Gadag and Uttara Kannada districts, data were obtained from the district ART centres. In one district, Bangalore, data were obtained from a tertiary hospital, St John’s Medical College Hospital, which operates an ART centre in collaboration with the National Program. The districts are shown in Figure 1. Study design and period Secondary data were used for analysis. A review was per- formed of all case records of all discordant couples who were registered at the above ART centres as of August 2010. Correspondence to: A De Costa Email: Ayesha.de.costa@ki.se International Journal of STD & AIDS 2012; 23: 502–506. DOI: 10.1258/ijsa.2011.011352