Design and participants
The study was designed to determine the associations between internalized homophobia, mental health, sexual behaviors, outness, and substance use among the gay and bisexual male population in Southwest China. A cross-sectional survey was conducted online in 2014 and 2015. Participants were recruited over a period of three months using advertisements on gay chat-room websites, Blued and Zank application software (two popular gay social software applications in China), and QQ groups. Participants completed an Internet-based survey questionnaire via the professional survey website, Wenjuan xing. The initial dataset consisted of 513 respondents.
Participants were eligible for the study if they were male, age 18 years or older, gay or bisexual, and residents of one of the recruitment regions in Southwest China. The website recorded each user’s city address, and the addresses of the respondents living in Sichuan, Chongqing, Yunnan, and Guizhou Provinces were included. A total of 435 participants were eligible for this study.
Measures
Socio-demographic data
Basic socio-demographic information, including age, highest educational level attained, occupation, monthly salary, and current relationship status, was collected.
Internalized homophobia
Internalized homophobia was measured using the 9-item Internalized Homophobia Scale that assesses the extent to which gay/bisexual individuals reject their sexual orientation, uneasiness about their same-sex desires, and seeking to avoid same-sex personal and social involvement and sexual feelings [1, 6, 19, 27]. Sample items include “I have tried to stop being attracted to men in general” and “I would like to get professional help in order to change my sexual orientation from gay/bisexual to straight.” Responses are made on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Scores range from 9 to 45, with higher scores indicating higher levels of internalized homophobia. The Chinese version of the scale was translated by one English professional translator and two psychology graduate students independently, and then consensus was reached. Previous research has indicated that the scale has acceptable internal consistency reliability [1, 27]. Cronbach’s α for the Chinese version of the scale in the present study was 0.88.
Mental health
The 6-item Kessler Psychological Distress Scale (K6) was used to measure respondents’ non-specific psychological distress during the past month [28]. The items measure depression and anxiety symptoms. Examples of items are “How often did you feel worthless?” and “How often did you feel anxious?” Responses are made using a 5-point Likert-type scale ranging from 1 (none of the time) to 5 (all of the time). Scores range from 6 to 30, with higher scores indicating higher levels of psychological distress. The internal consistency reliability of the Chinese version of the scale was very good in this study (Cronbach’s α = 0.95).
Disclosure of sexual orientation
Outness was measured using two questions adapted from a previous study [4]. The first question concerned the extent to which the respondents had disclosed their sexual orientation to others, with three optional categories: never had disclosed, partially disclosed, and fully disclosed. The responses were coded dichotomously by occurrence. The second question pertained to the extent to which the respondents had disclosed their sexual orientation to their parents. The respondents used a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree); their responses were dichotomized to reflect their disagreement/agreement with disclosure to their parents.
Sexual compulsivity
Sexual compulsivity was measured using the 10-item Sexual Compulsivity Scale [29] that assesses compulsive urges to perform specific sexual acts. Responses are made using a 4-point Likert scale ranging from 1 (strongly disagree) to 4 (strongly agree). Scores range from 10 to 40, with higher scores indicating a higher degree of compulsivity. The Chinese version of the scale was used in a previous study of 436 self-identified gay and bisexual men in Southwest China (Cronbach’s α = 0.91) [30].
Sexual orientation identity, sexual attraction, and sexual partners
Sexual identity, sexual attraction, and sexual partners were assessed using questions that were adapted from a previous online survey by Vrangalova and Savin-Williams [31]. Sexual identity was assessed by asking participants to identify their sexual orientation. They answered using one of five response options: gay, bisexual, heterosexual, questioning/uncertain, and other. Participants who self-identified their sexual orientation as heterosexual or questioning/uncertain were excluded in this study.
Sexual attraction, which included same-sex and other-sex attraction, was assessed using two separate questions: “How sexually attracted are you to women?” and “How sexually attracted are you to men?” Participants rated their attraction on a scale of 1 (not at all) to 5 (very much) [31].
Sexual partners was assessed using two separate questions: the total number of (1) male and (2) female partners with whom the respondent had sexual experiences during the previous 6 months, including vaginal sex and anal sex. The response was coded as having multiple male sex partners if the number of partners was greater than one, and it was coded as having female sex partners if the number was greater than 0.
Transactional sex
Transactional sex was measured using a single question: “During the previous 6 months, did you trade sex to get money?” The four response options were: never, rarely, sometimes, and always. The responses were dichotomized to reflect at least one episode of recent transactional sex.
Condom use
Condom use was measured using two separate questions about the frequency of condom use with male and female sex partners. There were six response options: never, seldom, often, most of the time, every time, and not applicable. The response was coded as having condomless sex if never, seldom, often, or most of the time was selected. Similar questions and recoding procedures have been used in published studies [32].
Substance use
To assess binge drinking, participants answered a question that was translated from a previous study [33]: “During the previous 6 months, how often [have you] had five or more drinks of alcohol in 2 hours at least once?” To assess the use of methamphetamines and Rush poppers, participants answered two separate questions: “During the previous 6 months, how often have you used methamphetamines?” and “During the previous 6 months, how often have you used Rush poppers?” Ten frequency categories were used as response options for each type of substance use. Each response was dichotomized to reflect at least one episode of recent substance use.
Statistical analysis
The data were analyzed using IBM’s SPSS, version 17, Mplus, version 6.12, and AMOS software, version 17. Cronbach’s α was used to assess the internal consistency reliability of the instruments. To examine cultural differences, confirmatory factor analysis was performed on the Internalized Homophobia Scale. Descriptive statistics were calculated. Latent profile analysis, which is a superior statistical technique compared to the traditional method of cluster analysis and has a variety of fit indices, was used to determine the number of homogenous groups or levels based on data from the Internalized Homophobia Scale [34, 35]. Spearman’s rho correlations were computed to examine the relationships between internalized homophobia and other factors. Sexual compulsivity and psychological distress responses were later dichotomized by a median split for use in logistic regression. Univariate logistic regression was performed. Separate multivariate logistic regression analyses using block entry were then used to examine the independent association between internalized homophobia and the factors, controlling for sexual identity, current relationship status, and monthly salary.