Background:  Illicit substance use is associated with abnormal cognition, and there are concerns that with HIV, substance use contributes to neuropsychological dysfunction. We hypothesized that in a large HIV⁺ cohort:  that those with a history of substance use would demonstrate worse baseline neuropsychological function than those without; that effects would be greatest for those not on ARVs; and that cumulative lifetime dose of substance would be negatively associated with neuropsychological function.

Methods:  Baseline data were taken from the central nervous system (CNS) HIV ARV Treatment Effects Research (CHART) study. A multi-method approach was used to determine substance use:  either lifetime DSM IV (Diagnostic and Statistical Manual of Mental Disorders, fourth edition) diagnosis of substance use disorder, or self-report of lifetime use, or positive urine toxicology. A comprehensive battery of neuropsychological tests determined neurocognition. Variables subjected to statistical covariance or group level matching included age, ethnicity, education, literacy, plasma and CSF HIV load, nadir CD4, ARV status, and depressive symptoms. No persons were acutely intoxicated or in active withdrawal.

Results:  We analyzed 1316 participants:  81% reported lifetime substance use and 73% met DSM IV criteria for lifetime abuse or dependence. In a subgroup of subjects carefully matched for important variables, a MANCOVA of neuropsychological summary scores, co-varying for urine toxicology stimulants (cocaine or methamphetamine) and depression revealed no significant effect of substance use (F(8, 505) = 1.48, p = 0.145). Similarly, substance use had no effect on the proportional functional dependence or number of functional complaints. These results did not change when inclusion was restricted to those with DSM IV substance use disorders, or when ARV status was considered. Nonparametric correlational analyses of quantified lifetime substance use exposure with neuropsychological function demonstrated no significant relationships and did not vary by ARV status. Subjects with positive urine toxicology stimulants were compared to a matched group with negative urine toxicology. Participants with positive urine toxicology performed significantly better on verbal tests (F(1, 437) = 7.95, p <0.01) with no significant performance differences in other neuropsychological domains.

Conclusions:  Substance use was not associated with compromised neuropsychological function at baseline in this HIV⁺ cohort when important co-factors were considered. The absence of relationship remained when substance use was determined by syndromic use, “casual” use, or urine toxicology, and did not vary by ARV status. This suggests that historic substance use and acute stimulant use do not require special consideration in cross sectional analyses of neuropsychological function in neuro-AIDS research.