HIV and the neuropsychology of everyday life.

Early in the history of the HIV epidemic, investigators noted that carriers of HIV who were asymptomatic had poorer cognitive performance than matched HIV2 controls. These HIV1 patients in early cohorts, mostly men who had sex with men, were medically asymptomatic but still showed performance deficits in verbal memory, executive function, and language. An early report concluded, “The constellation of subjective and objective neuropsychological and neurologic findings suggests the possibility of a definable syndrome associated with HIV infection in asymptomatic individuals.” Availability of combination antiretroviral therapy (CART) has reduced the incidence of HIV dementia, but mild neurocognitive disorders remain common in people with HIV. In the large CNS HIV AntiRetroviral Therapy Effects Research (CHARTER) study reported here, 47% had some form of neurocognitive impairment, with the vast majority falling into the range of asymptomatic neurocognitive impairment (ANI). The functional import of this level of cognitive deficit has been difficult to establish. Early studies found that mild cognitive deficits in asymptomatic patients were associated with increased risk of incident work disability and poorer medication management, but the studies were suboptimal because they compared HIV1 and HIV2 participants. The early studies were also unable to track linked transitions in cognitive and functional status using an appropriately wide range of self-reported and performance-based assessments of occupational skills. The design of CHARTER allows this definitive study and the answer is now clear. In models that adjusted for education, estimated verbal IQ, indicators of HIV progression, CNS medication penetration, and comorbid status, people with ANI at baseline, that is, mild cognitive impairment without functional limitations, were more likely to reach disability endpoints than HIV1 people with unimpaired cognition. Just over 50% of people with ANI declined in function, compared to 21.7% of neurocognitively normal (NCN) patients. In time-dependent analyses examining cognitive change, the risk of meeting disability endpoints among people with ANI was more than 3-fold higher than transitions in the NCN group, and thus the greater risk of disability in this group was likely related to worsening cognitive status. Because people with ANI had greater immunosuppression (lower CD4 nadir and greater likelihood of AIDS diagnosis), Grant et al. conclude that ANI is HIVdriven and not benign. ANI was a risk factor for incident disability even when analyses were restricted to people who at baseline were virally suppressed. Establishing the clinical import of subtle cognitive impairment can be challenging. Self-reports have their weaknesses (because of denial, loss of insight, or alteration of behavior to avoid challenges), while performancebased measures often lack ecologic validity (because participants are asked to perform tasks that are not typical behaviors). CHARTER got around this problem by developing a series of self-report, performance, and combined outcomes. In the absence of norms for the disability outcomes, the research team developed a criterionbased measure for the self-report measure (difficulty with 3 or more daily tasks) and a statistical criterion for the performance measure (scores for medication management and vocational skills at least 1 SD below the sample mean, or one test 2 SDs below the mean). The risk of poorer functional outcome associated with ANI was mostly consistent across the outcomes. One missing component of this approach is information on the stability of disability. Grant et al. do not report whether such disability persisted or perhaps remitted on subsequent assessments. This may also be important for assessing the risk of disability in ANI. On the other hand, time to first follow-up with disability may also be clinically important as an outcome in its own right. Also notable in this research are differences in the risk of disability by sex. Women were nearly 3 times more likely to reach the disability endpoint than men. This difference provides strong support for reports of greater vulnerability to HIV-associated neurocognitive impairment among women. Risk factors for poor neurocognitive performance, such as lower education, were more prevalent among women and indicate the importance of identifying potentially sex-