Background:  Historically cognitive difficulties associated with HIV were most commonly evident among individuals with significant immunosuppression. In the context of chronic treatment there is concern that cognitive difficulties may be exhibited among individuals with higher CD4 cell counts. In the present study we examined cognitive function in a cohort of HIV⁺ individuals taking ART and HIV^­ control subjects.  

Methods:  We enrolled 88 HIV⁺ individuals and 49 healthy control subjects. The HIV⁺ individuals were subdivided by CD4 cell count (CD4 <200; median CD4 = 126; n = 17; CD4 200 to 500; median CD4 = 327; n = 41; and CD4 >500, median CD4 = 629, n = 30). The groups were matched for age and education. Cognitive function was examined using a computerized battery of tests that measured attention, reaction time, information processing speed, motor speed, verbal fluency and executive function. 

Results:  Comparisons revealed significant differences on the tests of attention (F (3, 122) = 18.2, p <0.01), reaction time (F (3, 114 = 34.1, p <0.01), and motor tapping (F(3, 122) = 4.6, p <0.05). All 3 groups of HIV⁺ patients exhibited significant impairments relative to the seronegative control subjects on each of these tests. Furthermore, the results revealed a dose-response relationship (performance was best for those with the highest CD4 cell count, followed by a linear decline in performance in the remaining groups according to CD4 cell count), though the differences between the patient groups were not significant. Severity of heroin use correlated minimally with performance on attention (r = –0.39, p <0.05) and motor tapping (r = –0.23, p <0.05), while severity of cocaine use correlated minimally with performance on motor tapping (r = –0.23, p <0.05); severity of alcohol use did not correlate with performance on any cognitive measure.

Conclusions:  HIV⁺ patients exhibited significant cognitive impairments regardless of CD4 cell count. The results were not mediated by severity of illicit drug use, which accounted for <15% of the variance in cognitive function. These preliminary results suggest that the degree of immunosuppression may not represent a clinically useful biomarker for neurocognitive impairment in the era of ART.