HIV This Week

Dandona L, Dandona R, Kumar GA, Reddy GB, Ameer MA, Ahmed GM, Ramgopal SP, Akbar M, Sudha T, Lakshmi V. Risk factors associated with HIV in a population-based study in Andhra Pradesh state of India. Int J Epidemiol. 2008;37(6):1274-86. Epub 2008.

Population-based data on risk factors associated with HIV are not readily available from India. This understanding, and an estimate of the impact of addressing behavioural factors on reducing HIV, would be useful. Dandona and colleagues interviewed a population-based sample of 12,617 persons 15-49 years old from 66 rural and urban clusters in Guntur district in the south Indian state of Andhra Pradesh and tested their dried blood spots for HIV. They used multiple logistic regression to assess the association of risk factors with HIV, and calculated population impact numbers for HIV reduction if behavioural factors were addressed. Among men, there was significant association between HIV and history of sex with men, blood transfusion, having ever visited sex worker or multiple lifetime women sex partners, consuming alcohol before sex, recreational drug use, male non-circumcision, and tattooing (odds ratios 5.74-1.97, P < 0.03, R(2) = 0.11). Among women, the only identified behavioural factor associated with HIV was multiple lifetime men sex partners (P = 0.001, R(2) = 0.10). Taking into account the relative risk and prevalence of risk factors, the highest impact on reducing the HIV number per unit population was for male circumcision. Among the identified factors, male circumcision was estimated to have the highest relative impact on reducing HIV per unit population, but the feasibility of this intervention in India needs further investigation. The low explanatory power in the regression models of the usually considered risk factors for HIV suggests that better understanding of HIV dynamics at the population level in India is needed.

Editors’ note: In this analysis, behavioural risk variables could explain only a small fraction of the variability of prevalent HIV in the Andhra Pradesh population – 11% for men and 10% for women. This may be because there are associations other than recognized risk factors that were not explored or perhaps there was incomplete reporting by respondents about sensitive risk behaviour (8.8% of women and 43.4% of men reported having had sex with more than one person in their life). That male circumcision in urban men in India would have the biggest impact in reducing HIV prevalence provides food for thought in this country where male circumcision is associated with religious identity. Acceptability studies, community conversations, and situational analyses would be needed to assess the relevance of male circumcision for HIV prevention, along with prevention of human papilloma virus infection and genital ulcer disease in India.