A component analysis of the effects of a cognitive-behavioral stress management intervention on the psychological, neuroendocrine, immunologic and health status of HIV-infected gay men

Date of Award




Degree Name

Doctor of Philosophy (Ph.D.)



First Committee Member

Michael Antoni, Committee Chair


This study examined distress levels, quality of life, sleep, immunologic functioning and health status in 100 asymptomatic and early symptomatic HIV-infected gay men assigned to either a ten-week cognitive behavioral stress management intervention (CBSM) or a ten-week waitlist control group. Changes in coping, social support, self-efficacy, dysfunctional attitudes, and physiological arousal, targeted for change by components of CBSM, were also investigated, as well as their relationship to the outcome variables measured. CBSM subjects, but not control subjects, showed changes in several outcome variables, including decreases in POMS depression, anxiety, anger, confusion, and total mood disturbance (TMD); decreases in BDI scores; and increases in MOS mental health scores. CBSM subjects not taking protease inhibitor medications also displayed decreases in HSV-2 IgG antibody titers. CBSM subjects also showed enhanced self-efficacy and social support, use of more adaptive coping strategies and decreases in cognitive distortions. There were no significant effects of group assignment on sleep, cortisol, norepinephrine, epinephrine, blastogenic response to PHA, NK number or cytotoxicity, IgG antibodies to CMV, EBV, HHV-6 or HSV-1, or six-month health status.Tests of mediation showed that changes in social support and cognitive coping composite scores acted as co-mediators of the effect of group assignment on changes in POMS anxiety and TMD and that changes in coping and self-efficacy co-mediated the effect of group assignment on changes in MOS mental health. Changes in guidance also partially mediated the effect of group assignment on HSV-2 IgG changes in individuals not taking protease inhibitors. Stepwise regression analyses were then conducted to determine which variables (e.g., coping, social support, dysfunctional attitudes, self-efficacy, and physiological arousal) were predictive of changes in the outcome variables measured. Results showed that changes in each outcome variable were associated with a unique pattern of predictor variables. Finally, interrelationships between psychological distress, neuroendocrine measures, HSV-1 and HSV-2, and six-month health status were investigated within the whole sample and the non-protease inhibitor subsample. Correlational analyses revealed positive associations between BDI change and HSV-2 IgG change, cortisol change and HSV-2 IgG change, and among the three neuroendocrine measures and a negative correlation between BDI change and epinephrine change.


Psychology, Clinical

Link to Full Text


Link to Full Text