Funding: NIAAA R21 AA014773
This capacity-building study involved collaborations between an interdisciplinary team of Indian and United States (US) researchers, and was designed to test hypotheses regarding risk factors for HIV infection in Goa, India. Recent research by our team, that included ongoing work in India, suggested that heavy drinking significantly increases health risks, including risky relationship behaviors. The study included detailed assessment of overall alcohol consumption patterns (not only average volume) and their association with risky behaviors known to increase risk for HIV infection (such as having relationships with high-risk partners or violence). An intensive formative measurement-refinement phase included key informants, focus groups and depth interviews with small samples. Measure selection was informed by methodological work in the US and India. These efforts enhanced survey and daily diary measures and adapted them for use in Goa to account for local cultural circumstances. Following this, established alcohol screeners validated in India (AUDIT and RAPS4) were used to select urban and rural community samples of 500 men and 500 women in each of two communities (total n = 2,000). These samples, comprised three-fourths of both women and men current drinkers, including over samples of hazardous drinkers, completed a detailed survey. A selected population at elevated risk was studied in more depth. Specifically, 30 men and 30 women heavy drinkers who also reported risky behaviors were studied using 28-day prospective event-based diaries and qualitative depth interviews. A unique feature of the study was the triangulation of multiple methods to examine relationships between drinking and risk behaviors. Using culturally adapted measures, the study investigated specific risks associated with components of the drinking pattern hypothesized to elevate likelihood of risky behaviors, such as drinking contexts and alcohol-related problems. Relationships were examined while estimating and statistically adjusting for additional variables, like personal characteristics such as impulsivity and notions about relationships and gender, expected to be influential on the basis of earlier research both in the US and in India. The study tested hypotheses about various relationships among a set of predisposing and temperament variables, expectancies, attitudes, and patterns of alcohol consumption and problems, with risks as the dependent variable. Risk analyses informed the epidemiology of HIV risk behaviors in India and provided a rich source of information for planning preventive interventions in a resource-poor society.
