The relationships between substance abuse, psychosocial variables, and natural killer cell enumeration and function in HIV-infected and high-risk uninfected adolescents

This report examines the relationship between substance use, psychosocial stressors, and natural killer (NK) cell enumeration and function in HIV-infected and high-risk uninfected adolescents. We studied the association of demographic characteristics; self-report measures of alcohol, tobacco, and marijuana use; and self-report measures of psychosocial stressors ( depressive symptoms, anxiety) with three immune outcomes: NK ( CD3(-)CD16(+)CD56(+)) absolute counts, lytic units per peripheral blood mononuclear cells (PBMCs), and lytic units per NK cell. In addition, we determined the association of HIV disease stage, antiretroviral therapy ( ART), CD4(+) T-cell count, and viral load with these outcomes in the subset of HIV-infected adolescents. Methods: This cross-sectional analysis reports on data collected during a longitudinal observational study of adolescents ( the REACH Study). A cross-sectional analysis was performed with data from the first visit for each subject that met criteria for concurrent ( within 3 days) assessment of NK number and function, substance use, and psychosocial data. The data set represented 501 subjects. Analyses were performed separately for the HIV-seropositive and seronegative adolescents. In the HIV-seronegative population, there were no significant predictors of NK cell count and only female gender was significantly associated with CD3(-)CD16(+)CD56(+) NK lytic units per PBMC. Analysis of the HIV-seronegative cohort also showed that black race was significantly associated with higher lytic units per NK cell. Results: In HIV-seropositive adolescents, we observed an association of female gender with lower NK cell number and lytic units per PBMC, but not with lytic units per NK cells. Current use of one or two antiretroviral drugs was predictive of lower NK numbers. This drug effect was also noted in the functional assay per PBMC but not per NK cell. Increasing worry scores and no marijuana use over the past 3 months were associated with lower functional NK measures per PBMC in HIV-seropositive youth. Laboratory-confirmed recent marijuana use was highly predictive of increased lytic activity calculated per NK cell. These effects were not observed in similar analyses of data from HIV-seronegative adolescents. Depressive symptoms, assessed with an epidemiologic screening tool, were not found to be predictive of NK cell number or function in either the HIV-seronegative or the HIV-seropositive subset. These findings document associations between substance abuse, psychosocial variables, and NK numbers and function in adolescents.