HIV-Associated Dementia: Cofactors as Predictors of Severity of Neurocoenitive Deficits

The objective of the present study was to evaluate the relationship between a set of cofactors and severity of cognitive impairment, to determine if there were any factors which significantly predicted more severe neurocognitive deficits in persons with AIDS. Twenty-four male volunteers recruited from community groups and physician referrals participated. Subjects completed several self-report questionnaires eliciting information regarding demographics and risk factor variables, in addition to a comprehensive battery of neuropsychological tests. A severity of cognitive impairment summary score was computed for each subject, reflecting both the number of impaired tests and their distance in the impaired direction from normative data. Neither CD4 count, number of months since diagnosis of AIDS, number of AIDS-related illnesses, number of recent stressors, history of head injury/LOC, history of substance use, current or past psychiatric disorder, history of learning disability nor history of other medical illness were found to be significantly related to severity of cognitive impairment in this sample, after controlling for the effects of age, level of education, estimated premorbid IQ and mood status. However, no reliable conclusions could be drawn from this study because the small sample size resulted in an unacceptably low level of statistical power for the desired regression analysis. Exploratory analyses of variance revealed no significant group differences for any of the covariate or cofactor variables when subjects falling at the low, middle, and high ranges of severity of impairment were compared, with the exception of a possible inverse relationship with CD4 count. This was consistent with an exploratory stepwise regression analysis in which only CD4 count entered the model. Some potential limitations of the operational definitions used for the variables in this study were identified, and modifications were suggested. The results of additional exploratory analyses comparing group differences between the "globally impaired" and "unimpaired" subjects (Maj et al., 1994 criteria) on both the covariate and cofactor variables, and neuropsychological test performance, were also discussed