Project 458353

Sexual minority men, which are gay, bisexual, and other men who have sex with men, experience a high mental health burden due to societal stigma. However, our understanding remains limited due to challenges in recruitment and sampling. Population surveys from Statistics Canada are representative of 97% of Canadians, but up to 60% of sexual minority men have reported that they would not report their sexual orientation to a government employee. Linking these surveys to hospital records is another common way to identify mental health problems, although this would be missing out on those who do not seek care. Among sexual minorities in particular, this is compounded by the fact that many may choose not to seek care due to fear of discrimination, past negative experiences in the healthcare system, or lack of cultural competence from providers. Community-based surveys that are conducted by groups of sexual minorities could have fewer misclassification and underreporting biases, however, these will only capture those sexual minorities who are the most connected to the community. To remedy these issues, this thesis will propose the use a Bayesian approach, an alternative statistical method that will allow us to combine strengths and weaknesses of different datasets to produce more accurate estimates. Using this same approach, we will also calculate individual risk scores to predict improvement in mental health through factors such as seeking care, social support, outness, and resilience using community-based and population-based surveys. Building risk scores for mental health improvement among sexual minority men will help individuals to make more informed decisions regarding their care, especially in situations where care is not easily accessible (e.g., during pandemic times). It also allows public health practitioners to better target groups who need help. These methods can potentially be used in other hard to reach populations, which will improve future studies.