Publication Date




Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Nursing (Nursing)

Date of Defense


First Committee Member

Victoria Mitrani

Second Committee Member

Nilda Peragallo Montano

Third Committee Member

Daniel Feaster

Fourth Committee Member

Joseph De Santis


In the United States, men who have sex with men (MSM) remain at greatest risk for acquiring HIV infection. MSM are also at increased risk for depression, and depression and sexual risk behavior among MSM appear to be linked. Stigma, in the forms of gay related stigma and HIV-related stigma, have been associated with depression and high-risk sexual behavior among MSM living with HIV infection, as has the internalization of these stigmata over time. As stigma is socially constructed, the stress process model provides a useful framework for understanding the influence of stress and contextual factors on depressive symptoms and high-risk sexual behavior. The purpose of this study was to test a stress process model examining the mechanisms by which discrimination-related stress influences depression and high-risk sexual behavior among MSM living with HIV infection, and to understand the role of internalized gay related and HIV-related stigma as intervening variables. A cross-sectional observational design was used. A convenience sample of 102 men living with HIV infection, aged 18 and over, self-identifying as MSM, who reported sex with another man in the past 12 months, spoke and read English, and resided at least part time in South Florida were recruited in Broward and Miami-Dade Counties. Participant ages ranged from 24 to 69 years, with mean participant age of 48.03 (SD = 10.3) years. In terms of ethnicity or culture, 23.5% of the sample identified as African American (n = 24), 9.8% as Hispanic/Latino (n = 10), 55.9% as White (n = 57), and 10.8% as Other (n = 11). Variables conceptualized as stressors were gay related stigma and HIV-related discrimination. Outcome variables included depressive symptoms and high-risk sexual behavior (i.e., instances of unprotected sex in past year). Intervening variables included internalized homophobia and internalized HIV-related stigma. Contextual variables, controlled in primary hypotheses, and not controlled in exploratory analyses, included life experiences and social support. Other control variables included age, ethnicity, education, and employment. Demographic variables included county of residence, income level, sexual orientation, age since participant knew he was gay, bisexual, or interested in sex with other men, relationship status, year of HIV diagnosis, if ever diagnosed with AIDS, if currently on medication for HIV infection, if on medication, number of doses of HIV medication missed in previous week, last CD4+ count, and last viral load. The original study analysis plan was to use structural equation modeling, which failed to converge on solutions. Multiple and logistic regression were used as an alternative, and the study hypotheses were revised to follow the original hypotheses as closely as possible. The two stressor variables, and the two intervening variables, were each combined to a single variable due to significant shared variance. It was hypothesized that perceived/enacted stigma, the single stressor variable, would predict the outcomes, and that internalized stigma, the single intervening variable, would mediate or moderate the relationship between perceived/enacted stigma and the outcomes. The study hypotheses were not supported, with the exception of a significant relationship between perceived/enacted stigma and internalized stigma; the relationship between internalized stigma and depressive symptoms approached significance. Exploratory analyses were conducted to examine effects of covariates/controls and no significant relationships were found. The participants in this study experienced depressive symptoms and engaged in high-risk sexual behavior, but these outcomes were unrelated to the proposed predictor and intervening variables. It appears that other factors may be driving these findings. Possibilities include the influence of unmeasured components of the stress process, such as coping, characteristics of the sample, and participant bias. Recommendations for future nursing research are discussed.


HIV; men who have sex with men; depression; high-risk sexual behavior; stigma; stress process model