Authors: Gilda E Ennis, Rebecca L Koscik, Yue Ma, Tobey J Betthauser, Erin M Jonaitis, Carol A Van Hulle, Shenikqua Bouges, Nathaniel A Chin, Corinne D Engelman, Rozalyn Anderson, Richard Batrla, Gwendlyn Kollmorgen, Cynthia M Carlsson, Sanjay Asthana, Sterling C Johnson, Henrik Zetterberg, Kaj Blennow, Barbara B Bendlin
Published: 2020-12-07
DOI: 10.1002/alz.047022
Source: Full article
AbstractBackgroundInsulin resistance (IR) has been related to increased risk for Alzheimer’s clinical syndrome. While some cross‐sectional studies have linked IR to Alzheimer’s disease (AD) pathology, others have been negative. We examined the relationship of IR to longitudinal cerebrospinal fluid (CSF) biomarkers of AD pathology, cerebral amyloid chronicity, and cognitive decline.MethodParticipants (Table 1) were enrolled in the Wisconsin Registry for Alzheimer’s Prevention (WRAP), a longitudinal observational study of non‐demented adults. We utilized data from biennial visits and CSF and positron emission tomography (PET) sub‐studies. For our first study, participants (n=148) had CSF biomarker data (Aβ42: Aβ40 ratio, p‐tau181, t‐tau, and neurogranin) from 1‐5 CSF samples collected every 2‐8 years and fasting insulin and glucose (collected during 1‐5 visits before first CSF) for calculating homeostatic model assessment of insulin resistance (HOMA‐IR). CSF was assayed using the Roche NeuroToolKit on the Elecsys® platform. Linear mixed effects models tested HOMA‐IR as a moderator of aging‐related change in CSF biomarkers. In study 2, participants (n=253) had [C11]PiB PET imaging and 3‐6 HOMA‐IR values for calculating a within‐person HOMA‐IR mean and SD (i.e., fluctuation). Amyloid chronicity was calculated by subtracting estimated age at PiB positivity (i.e., estimated age when PiB DVR ≥ 1.2) (Koscik & Betthauser, et al., 2020) from age at last HOMA‐IR. Hierarchical regression tested whether HOMA‐IR and an HOMA‐IR x APOE4 interaction predicted amyloid chronicity. In study 3 (n=1,298), linear mixed effects tested if baseline HOMA‐IR moderated age‐related decline in longitudinal measures of the preclinical Alzheimer’s cognitive composite (PACC). Age, APOE4, sex, education, and systolic blood pressure (SBP) were controlled in all study models.ResultHOMA‐IR was not a significant moderator of aging‐related change in any CSF biomarker (Figure 1 & 2). HOMA‐IR mean and SD and the HOMA‐IR x APOE4 interaction were not significant predictors of amyloid chronicity. Increased HOMA‐IR predicted greater age‐related decline in PACC (Figure 3).ConclusionHOMA‐IR was related to cognitive decline but not cerebral amyloid chronicity or age‐related change in CSF biomarkers of AD pathology and neurodegeneration. IR may not be directly related to preclinical AD pathology. Pathways linking IR to dementia require further investigation.