Background:  Studies of HAART-treated patients have demonstrated durability of viral load suppression, decreased incidence rates of HIV-associated neurological disease, and increased survival. However, with prolonged HAART exposure, a significant proportion of patients develop ART resistance and toxicities, such as insulin resistance and mitochondrial dysfunction, that could affect the central nervous system. The extent to which HAART influences the prevalence and incidence of HIV-associated neurocognitive impairment positively or negatively remains unclear. We sought to estimate the prevalence of neurocognitive impairment in patients on HAART and the relationship of virologic and immunologic factors to neurocognitive impairment.

Methods:  Trained study coordinators administered a brief neuropsychological battery of Trailmaking A, B, and Digit Symbol to 1498 subjects enrolled in ACTG A5001, a longitudinal study of subjects enrolled into randomized trials of ART. Neurological tests were adjusted based on normative corrections for age, race/ethnicity, sex, and education. Neurocognitive impairment was defined as performance at least 1 standard deviation below norms for 2 tests, or 2 standard deviations on 1 test. Logistic regression was used to examine the association between virologic and immunologic factors and neurocognitive impairment.

Results:  The study sample was:  54% white, 23% African American, 20% Hispanic; 85% male; median CD4 of 421, and HIV RNA 50; and age 40 years. We classified 645 subjects (43%, 95%CI = 41 to 46) as having neurocognitive impairment at the first testing visit at least 20 weeks after study entry—CD4 nadir <200 cells (p <0.05, OR = 1.23, 95%CI = (1.00 to 1.52). Nadir <200 (49%) was associated with prevalent neurocognitive impairment after adjusting for race, education, age, gender, and ART history. Concurrent viral load and CD4 counts were not significantly associated with neurocognitive impairment. Among 853 initially unimpaired subjects followed for a median of 93 weeks, 159 (18.6%) subsequently became neuropsychologically impaired.

Conclusions:  In this large, HAART-treated cohort, the prevalence of neurocognitive impairment, based on a screening battery, was notable (43%) suggesting that neurocognitive impairment is still frequent even in the era of HAART. In addition, the incidence of neurocognitive impairment during follow-up was relatively frequent. Low CD4 nadir (<200) was associated with an increased risk for of neurocognitive impairment. However, concurrent CD4 and viral load were not significantly associated with neurocognitive impairment, a finding that is distinctly different from pre-HAART cohorts. Future studies should evaluate potential predictors or risk factors for incident impairment to help identify possible interventions to reduce the effect of neurocognitive impairment.