Background:  Combination ART (cART) typically suppresses HIV viral replication in plasma and cerebrospinal fluid (CSF), but its long-term effects on neurocognitive impairment (NCI) are unclear. This multi-site study examined the prevalence and predictors of NCI in the cART era, within an HIV⁺ sample. We hypothesized that cART would be associated with a lower prevalence of NCI.

Methods:  We gave comprehensive neuromedical and neuropsychological evaluations to 1555 HIV⁺ patients at 6 university clinics across the United States (age 43.2±8.6; education 12.7±2.5 years; 77% male; ethnicity = 49% African American; 9% Hispanic; 39% non-Hispanic white; 3% other; 28% intravenous drug users; 58% men who have sex with men) received. Participants were unselected in that there were no neuromedical or psychiatric exclusions. Participants were classified according to severity of their co-morbidity (factors that might contribute to NCI independent of HIV, e.g., epilepsy):  minimal, n = 843; moderate, n = 473; and severe, n = 240.

Results:  Overall, 45% of the cohort had NCI based on a global rating ≥5 derived from comprehensive neuropsychological testing, with rates increased in groups with greater co-morbidity (39% vs 54% vs 79%, respectively; p <0.01). NCI prevalence was higher with AIDS diagnosis and low nadir CD4 only in the minimal co-morbidity group. In this group, a higher rate of NCI was found in those currently on cART (43% vs 31%, p <0.01), probably because 74% of AIDS vs 27% of non-AIDS (p <0.01) were on cART. In contrast, the rates of NCI in the moderate or severe co-morbidity groups were 64% on cART and 58% off cART (p=.13). For treated patients in the minimal co-morbidity group with no history of severe immunosuppression (nadir CD4 >200; n = 184), those with HIV suppression on treatment had lower NCI rates (29% vs 45% for those with detectable plasma viral load, p = 0.02).

Conclusions:  A high prevalence of NCI was found in this large geographically and demographically diverse HIV⁺ population despite the majority receiving cART. Substantial impairment was seen even among participants without significant co-morbidity. Biological and clinical indicators of HIV disease severity were related to NCI only in patients without significant co-morbidities. Thus, both HIV and co-morbidity contribute to NCI in the cART era. The results also indicate the importance of clinical studies to determine whether cART-related HIV suppression may protect the central nervous system more effectively if started earlier in the course of HIV infection.