A lipid storage-like disorder contributes to cognitive decline in HIV-infected subjects

Abstract

Objective: In this multicenter cohort study, we sought to identify prognostic and associative metabolic indicators for HIV-associated neurocognitive disorders (HAND).

Methods: A quantitative lipidomic analysis was conducted on 524 longitudinal CSF samples collected from 7 different performance sites across the mainland United States, Hawaii, and Puerto Rico. Subjects included HIV-infected individuals with longitudinal clinical and cognitive testing data and cognitively normal HIV-negative healthy controls.

Results: At baseline, HIV+ subjects could be differentiated from HIV- controls by reductions in a single ceramide species and increases in multiple forms of cholesterol. Perturbations in cholesterol metabolism and ceramide were influenced by combined antiretroviral therapy (cART) use. There were no cross-sectional baseline differences in any lipid metabolite when HIV+ subjects were grouped according to cognitive status. However, a single sphingolipid metabolite and reduced levels of esterified cholesterols were prognostic indicators of incident cognitive decline. Longitudinal patterns of these disturbances in sphingolipid and sterol metabolism suggest that a progressive disorder of lipid metabolism that is similar to disorders of lipid storage may contribute to the pathogenesis of HAND.

Conclusions: These findings suggest that HIV infection and cART are independently associated with a CNS metabolic disturbance, identify surrogate markers that are prognostic for cognitive decline, and implicate a lipid storage-like disorder in the progression of HAND.

Figure 1. Baseline sphingolipids that are prognostic indicators for cognitive decline

(A) Lower ratios of CSF sphingomyelin/ceramide C24:1 at baseline was a prognostic indicator for cognitive decline (odds ratio [OR] 0.06, 95% confidence interval [CI] 0.01–0.96, p = 0.047). (B) Elevated baseline levels of CSF ceramide C24:1 were associated with a greater probability of cognitive decline at the next visit (OR 1.31, 95% CI 0.91–1.88, p = 0.252). (C) Baseline levels of sphingomyelin did not predict cognitive decline (OR 0.94, 95% CI 0.70–1.26, p = 0.679).

Figure 2. Specific forms of sterols and triglycerides that are prognostic indicators for change in cognitive status

Decreased CSF levels in multiple forms of cholesterol esters at baseline were prognostic indicators for cognitive decline. (A) Scatterplot shows the relationship between baseline levels of the cholesterol ester C16:0 and the probability of cognitive decline at the next visit (odds ratio [OR] 0.51, 95% confidence interval [CI] 0.31–0.84, p = 0.008). This plot is representative of the 16 cholesterol esters with similar relationships to cognitive decline (see table 1 for a complete list). (B) Higher levels of triglycerides were associated with a greater probability of cognitive improvement at the next visit. Data show the relationship of triglyceride to the probability of cognitive decline at the next visit (OR 1.93, 95% CI 1.21–3.10, p = 0.006).

Figure 3. A progressive disturbance in sphingolipid metabolism is associated with cognitive decline

Increasing ratios for 3 sphingomyelin/ceramide species over time were associated with declining cognitive status. Scatterplots show the continuous relationships of sphingomyelin and ceramide to cognitive decline. Data are expressed as a ratio of SM/ceramide (A, D, G), individual ceramide (B, E, H), and SM (C, F, I) species to the probability of cognitive decline for the indicated species. (A–C) C16:0 (odds ratio [OR] 5.07, 95% confidence interval [CI] 1.06–24.15, p = 0.042), (D–F) C18:0 (OR 11.86, 95% CI 1.39–100.90, p = 0.024), and (G–I) C24:0 (OR 7.98, 95% CI 1.18–54.21, p = 0.034).