Discrimination based on sexual orientation against MSM in Brazil: a latent class analysis

ABSTRACT Introduction: Discrimination based on sexual orientation can influence vulnerability to HIV, increasing exposure to risky sexual behavior among men who have sex with men (MSM). Objectives: To analyze data using latent class analysis (LCA) to identify groups of individuals with specific patterns of discrimination based on sexual orientation (DSO). Methods: Cross-sectional study using respondent-driven sampling in 12 Brazilian cities in 2016. LCA was used to characterize discrimination among MSM based on 13 variables in the survey questionnaire. The proportions of men reporting DSO and other variables of interest were estimated using Gile’s Successive Sampling estimator. Results: Most MSM were young, single, had a religion, had a high school or college degree, black or brown skin color, and socioeconomic status classified as average. More than half of the participants reported that they had been discriminated against during the last 12 months due to their sexual orientation (65%), more than a third said they had felt afraid of walking in public places during the past 12 months, and about one-fifth of participants reported having been victims of physical or sexual assault due to DSO. DSO was classified into four latent classes: “very high”, “high”, “moderate” and “low”, with estimates of 2.2%, 16.4%, 35.1%, and 46.19%, respectively. Conclusion: We observed a high proportion of discrimination against MSM in this study. The use of LCA differentiated parsimoniously classes of discrimination. RESUMO Introdução: A discriminação por orientação sexual (DPOS) pode influenciar a vulnerabilidade ao HIV aumentando a exposição a comportamentos sexuais de risco entre homens que fazem sexo com homens (HSH). Objetivos: Examinar dados utilizando a análise de classes latentes (ACL) para identificar grupos de indivíduos com padrões específicos de DPOS. Métodos: Estudo transversal com entrevistados recrutados pelo processo amostral respondent driven sampling em 12 cidades brasileiras em 2016. A ACL foi usada para caracterizar o DPOS entre HSH com base em 13 variáveis do bloco de discriminação do questionário da pesquisa. As proporções de DPOS e das variáveis de interesse, bem como seus intervalos de confiança (95%) foram ponderados usando o estimador de Gile. Resultados: A maioria era de jovens, solteiros, com alguma religião, escolaridade média ou superior, cor da pele preta ou parda e com nível socioeconômico médio. Mais da metade referiu ter sido discriminado nos úl