| Publication Abstract Display | | Type: Poster | | Title: The relationship between amnestic mild cognitive impairment and biomarkers of inflammation among adults living with HIV. | | Authors: Campbell L, Sundermann E, Letendre S, Kallianpur A, Hulgan T, Montoya J, Ellis R, Grant I, Heaton R, Moore D, Moore R | | Date: 09-13-2018 | | Abstract:Background: As people living with HIV (PLHIV) age, it becomes increasingly important to differentiate HIV-Associated Neurocognitive Disorders (HAND) from other common causes of neurocognitive impairment, such as Alzheimer’s disease (AD) and amnestic mild cognitive impairment (aMCI). While aging is the greatest risk factor for AD, little is known about how aging in the context of HIV affects risk for AD/aMCI. HIV may influence AD/aMCI risk via inflammation, as both HIV infection and aMCI/AD are associated with higher levels of chronic, low-grade inflammation. Therefore, we classified PLHIV into HAND or aMCI groups and hypothesized that plasma biomarkers of inflammation would be highest in older PLHIV with aMCI.
Material & Methods: Analysis included 244 PLHIV (age range: 24-68, 86% male, 59% non-Hispanic white) on antiretroviral therapy with undetectable viral load (i.e., <50 copies/ml plasma). Participants completed comprehensive neurobehavioral and neuromedical evaluations. Three plasma biomarkers of inflammation (i.e., TNF-, MCP-1, and IL-6) were measured by immunoassay, and all biomarkers were log-transformed. HAND was classified using Frascati criteria. We used an empirically-based neuropsychological diagnostic approach for aMCI (Jak/Bondi criteria) that was adapted to distinguish HAND from aMCI by focusing on recognition memory impairment which is characteristic of aMCI but not HAND. Participants that met criteria for aMCI, regardless of HAND classification, were classified into one aMCI+ group. Participants who were not classified as aMCI+ were separated into HAND (HAND+/aMCI-) and no HAND (HAND-/aMCI-) groups. Multinomial logistic regression modeling analyzed the relationship between classification group and age. Multivariable linear regressions predicting inflammatory markers were used to examine the relationship among classification group, age, and inflammation. Comorbidities and demographic variables related to the outcome at p<0.10 were included as covariates; age and sex were also included in models predicting inflammatory biomarkers.
Results: Using the articulated classification approach, 85 PLHIV were classified as aMCI+ (78% HAND), 66 HAND+/aMCI-, and 93 HAND-/aMCI-. Increase in age significantly increased the odds of being in the aMCI+ (odds ratio:1.05, p=0.01), but not in the HAND+/aMCI- group (p=0.33). There were no significant differences in inflammatory biomarkers (i.e., TNF-, MCI-1, and IL-6) between the three classification groups (all p’s>0.05). Separate multivariable models examined the associations between inflammatory biomarkers and classification group x age (i.e., <50 vs. ≥50 years) interactions. The interaction term when comparing the aMCI+ group to the HAND+/aMCI- group in the model predicting TNF- was significant, such that older aMCI+ PLHIV demonstrated higher TNF- levels (suggesting more inflammation) than younger aMCI+ PLHIV (=0.05, p<0.01). When comparing the aMCI+ group to the HAND-/aMCI- group, however, the interaction was not significant (p=0.16). A similar trend was seen with the classification group x age interaction in the aMCI+ group with MCP-1 (p=0.08), but not for IL-6.
Conclusions: These findings suggest that some inflammatory biomarkers are more strongly associated with older age in aMCI and may be a mechanism of aMCI particularly in the context of HIV. Longitudinal studies that include adults without HIV infection are needed to further examine biomarkers of inflammation and cognitive trajectories, such as conversion to aMCI or AD. |
return to publications listing
|