OBJECTIVETo investigate gay identity status of men who have sex with men(MSM) in Xinjiang, and analyze its association with sexual behaviors.

METHODSFrom May to August in 2013, an anonymous questionnaire survey with convenience sampling was conducted among MSM in Urumqi, Kashgar, Aksu and Yining cities of Xinjiang. Those who were ≥ 16 years of age and claimed to have male-to-male sex behaviors were eligible for the survey. A total of 1 467 participants were recruited. Chi square analysis was performed for examining the association between ego-identity, social-identity and relevant sexual behaviors. Selected variables associated with ego-identity, social-identity in the univariate analysis were subsequently assessed in a multivariate analysis using logistic regression.

RESULTSThe MSM with positive ego-identity accounted for 44.7% (656/1 467), and positive social-identity accounted for 29.1% (427/1 467). For ego-identity, Uyghur MSM were less positive than Han MSM (OR (95%CI) was 0.43 (0.32-0.58) ) ; those who were over 20 years of age when accepting their sexual orientation were less positive than those who were younger than 20 years of age when accepting their sexual orientation (OR (95%CI) was 0.61 (0.47-0.80) );and homosexuals were more positive than bisexuals (OR (95%CI) was 1.98 (1.50-2.61) ). For social-identity, Hui MSM were less positive than Han MSM (OR (95%CI) was 0.61 (0.42-0.88) ); those whose education level was higher than college were less positive than those with junior high education (OR (95%CI) was 0.60 (0.40-0.89) ); famers/herdsmen were more positive than students (OR (95%CI) was 4.17 (2.13-8.17) ); the divorced/widowed were more positive than the singles (OR (95%CI) was 2.40 (1.34-4.29) ); those who were over 20 years of age when accepting their sexual orientation were less positive than those who were younger than 20 years of age when accepting their sexual orientation (OR (95%CI) was 0.59 (0.44-0.81) ). Among the MSM with negative ego-identity and negative social-identity, the proportion of planning to get married with women was 66.8% (267/400) and 76.5% (306/400), respectively, which were higher than those with positive ego-identity (33.2% (133/400)) and positive social-identity (23.5% (94/400)) ; their proportion of intending to have offspring was 62.0% (287/463) and 73.4% (340/463), respectively, which were higher than those with positive ego-identity and social-identity (38.0% (176/463), 26.6% (123/463)) (χ(2) = 39.61 and 7.90, respectively, both P values were <0.05). Among the MSM with negative social-identity, the proportion of looking for male sexual partners in toilets or parks was 18.1% (188/1 040), in bathhouse or sauna was 17.3% (180/1 040), and through internet was 82.0% (853/1 040), which were higher than those with positive social-identity (8.9% (38/427), 9.8% (42/427) and 61.6% (263/427)) (χ(2) = 66.78, P < 0.01).

CONCLUSIONIn 2013, the gay identity of MSM in Xinjiang was relatively poor, and the factors associated with it included ethnicity, occupation, marital status and education level. The MSM with negative identity tended to involve in risk sexual behaviors, and had increased risk of HIV infection.