Journal of neurovirology 2017 04 20() doi 10.1007/s13365-017-0526-z
HIV-infected smokers lose more years of life to tobacco-related disease than HIV. Since neurocognitive deficits are common among those with HIV and are associated with smoking persistence, these deficits may be a unique barrier to smoking cessation among HIV-infected smokers. Documenting unique differences in and correlates of cognition among HIV-infected smokers is a critical step towards developing a population-specific tobacco cessation treatment. We compared neurocognitive function between HIV-infected (n = 103) and HIV-uninfected smokers (n = 70), accounting for demographic and smoking-related variables. We also evaluated whether HIV-related health outcomes (e.g., CD4 count, viral load, depression ratings, quality of life [QoL]) and HAART adherence were associated with cognition. Participants completed neurocognitive tasks (N-back and Continuous Performance Task [CPT]) measuring working memory, attention, and processing speed, and intra-individual variability. Stepwise regression models were conducted and validated with resampling techniques. HIV-infected smokers performed worse than HIV-uninfected smokers on working memory, processing speed, and intra-individual variability (all p < 0.01). ROC analysis for the model including cognitive measures demonstrated 85% area under the curve, which indicates "good prediction" for distinguishing between HIV-infected and HIV-uninfected smokers. This was a significant improvement over the model including demographic and smoking-related variables only (p = 0.0003). Among HIV-infected smokers, neurocognitive performance was negatively associated with QoL and depression ratings. Smoking cessation interventions for HIV-infected smokers should consider cognitive neurorehabilitation as a potential strategy to decrease the likelihood of nicotine relapse and decrease tobacco-related morbidity in this population.