Lower CSF Amyloid-β levels are associated with worse neurocognitive functioning in HIV-infected adults with a family history of dementia.

Type: Published Abstract
Title: Lower CSF Amyloid-β levels are associated with worse neurocognitive functioning in HIV-infected adults with a family history of dementia.
Authors: Fazeli P, Moore DJ, Franklin DR, Heaton RK, Marra C, Gelman B, McCutchan JA, Grant I, Letendre S, for the CHARTER Group
Year: 2015
Publication: Conference on Retroviruses and Opportunistic Infections

Background: Studies of HIV-infected (HIV+) patients have shown that independently both family history of dementia (FHD) and lower levels of Aβ-42 are associated with worse neurocognitive functioning. We measured cerebrospinal fluid (CSF) levels of Aβ-42 in a convenience sample of 184 HIV+ adults currently on antiretroviral therapy (ART) (90 with FHD and 94 demographically-matched without FHD) and examined the relationship between these variables and HIV-associated neurocognitive disorders (HAND).

Methods: All participants underwent comprehensive neuropsychological and neuromedical assessments, and determination of CSF concentration of Aβ-42. FHD was defined as a self-reported first or second-degree relative with a dementia diagnosis. Univariate analyses were used to determine whether HAND status was associated with FHD, CSF Aβ-42, or any potential covariates (e.g., demographics, HIV disease characteristics) and to examine if CSF Aβ-42 levels differed by FHD. Multivariable logistic regressions then examined the association of CSF Aβ-42, FHD, and their interaction on HAND.

Results: FHD did not differ between those with and without HAND (p = 0.24); however, CSF Aβ-42 levels were significantly (p = 0.03) lower in the HAND group. Further, CSF Aβ-42 was not associated with FHD (p = 0.89). Multivariable models controlling for race and comorbidity rating showed a significant main effect of CSF Aβ-42 (p = 0.03) and a trend (p = 0.06) towards an interaction between FHD and CSF Aβ-42, such that lower CSF Aβ-42 was only associated with HAND in those with FHD (p < 0.01) as compared to those without FHD (p = 0.83). Examining inheritance type (maternal vs. paternal), type of dementia (AD vs. non-AD), or number of relatives with a dementia diagnosis did not strengthen these associations. A median split was conducted on raw CSF Aβ-42 values to create high and low CSF Aβ-42 groups and showed that those with low CSF Aβ-42 and FHD had the highest prevalence of HAND (74%), while the remaining groups were similar (high CSF Aβ-42 FHD: 49%; high CSF Aβ-42 no FHD: 52%; low CSF Aβ-42 no FHD: 57%). Post-hoc analyses showed that the CSF Aβ-42 X FHD interaction trend was driven by speed of information processing (p = 0.01).

Conclusions: FHD moderates the effect of CSF Aβ-42 on HAND. These findings highlight the complexities of analysis of biomarkers of age-related neurodegeneration in HAND.