Correlates of paid sex among men who have sex with men in Chennai, India.

OBJECTIVES: To assess correlates of paid sex among men who have sex with men (MSM) in Chennai, India.
METHODS: A randomised survey was conducted among 200 MSM recruited from public sex environments using time-space sampling. The association of predictors with paid sex was assessed with chi(2) tests and multiple logistic regression.
RESULTS: Participants' mean age was 28.5 years (SD 8.7). Most (71.5%) were kothis; 60% had less than high school education and two-thirds had a monthly income less than 2000 Indian rupees. More than one-third (35.0%) reported daily/weekly harassment; 40.5% reported forced sex in the past year. The prevalence of paid sex was 59.5% (95% CI 52.7% to 66.3%). Univariate analyses indicated that paid sex was associated with kothi identity (chi(2) = 14.46; p<0.01), less than high school education (chi(2) = 4.79; p<0.05), harassment (chi(2) = 11.75; p<0.01) and forced sex (chi(2) = 3.98; p<0.05). Adjusted analyses revealed that paid sex was associated with kothi identity (adjusted odds ratio (AOR) 2.62, 95% CI 1.34 to 5.10) and harassment (AOR 2.34, 95% CI 1.16 to 4.72). MSM who engaged in paid sex (versus no paid sex) had a mean of 31 partners in the past month (versus 4, t = 6.17, p<0.001) and 71.2% used condoms consistently (versus 46.4%, chi(2) = 18.34; p<0.01). Overall, 32.5% were never tested for HIV.
CONCLUSIONS: Epidemic rates of harassment and sexual violence against MSM who engage in paid sex, predominantly kothis, suggest that interventions should target structural factors placing these men at increased risk of HIV/sexually transmitted infections and other health-compromising conditions. The effectiveness of individual-level, knowledge-based and condom-focused preventive interventions may be constrained in the context of poverty, low education, harassment and sexual violence.