People with chronic HIV (PWCH) have an excess burden of comorbidities with a substantial adverse impact on mortality and quality of life. But their effects on long-term neurocognitive (NC) outcomes are less well-studied.
Inclusions: people with HIV for at least 2 years. Exclusions: active conditions sufficiently severe to interfere with the study evaluations. We constructed a simple comorbidity index (SCI, one point per comorbidity, range 0-5), selected based on univariable analyses predicting neurocognitive decline over 12 years, comprising the presence or absence of diabetes, hypertension, neuropathic pain, chronic obstructive pulmonary disease (COPD), and current major depressive disorder (MDD). Linear regression assessed the relationship between the SCI and a published, summary regression-based change score (sRBCS), adjusted for baseline, practice effect, and other factors, measuring change in overall NC performance. We compared the SCI to other well-established comorbidity indices: The Charlson, Veterans Administration Comorbidity Score (VACS) and Framingham cardiovascular risk.
Participants were 397 PWCH enrolled between 2003-2009, mean (SD) age at baseline 43.5 (7.68) years, 304 (76.6%) men, 164 (41.3%) white, education 13.0 (2.63) years, median (interquartile range, IQR) estimated duration of HIV 9.82 (4.44, 14.5), nadir and current CD4+ T-lymphocytes/uL 172 (30, 308) and 452 (278, 636), On ART 74.3%, 45.9% with undetectable plasma HIV RNA. Those with higher SCI at baseline had significantly worse cognitive change over 12 years (p=0.003, R2=0.03). There was a dose-response relationship such that those with 3 baseline comorbidities had worse decline than those with 2, 2 worse than 1, and 1 worse than none (indices of 2+ and 1 vs. none, effect sizes 0.64 and 0.29). SCI was significantly associated with declines in the specific domains of executive function, learning, verbal, and SIP. In separate multivariable models using previously published indexes (Charlson, VACS and Framingham CVD risk), only the SCI was significantly associated with NC worsening rate. Comorbidities with the largest effects at baseline were hypertension, COPD and current MDD.
Baseline comorbidities in PWCH predicted subsequent trajectories of global cognitive decline over extended follow-up, suggesting that any treatments for these comorbidities during the follow-up period were only partly effective, and that more successful treatment of them might ameliorate NC decline.