Abstract

AbstractBackgroundDysfunctional insulin signaling and elevated hemoglobin‐A1c (HbA1c) ‐ both associated with type‐2 diabetes mellitus (T2DM) ‐ may result in increased tau hyperphosphorylation, thereby increasing the risk of Alzheimer’s Disease (AD)‐related neuropathology in T2DM patients. Individuals with both AD and T2DM show different profiles of tauopathy compared to individuals with just AD. Here, we evaluated voxelwise tau‐PET patterns with and without T2DM to clarify T2DM‐associated patterns of tau deposition in cognitively intact older adults.MethodWe examined 502 cognitively intact adults from the Health and Aging Brain Study: Health Disparities who had available 18F‐PI‐2620 tau‐PET, florbetaben‐PET (Aβ) (n = 462), MRI T1‐weighted scans (Siemens‐3T), and blood HbA1c levels (n = 492) (Table1). Voxelwise regression models were run in SPM12 using warped 18F‐PI‐2620 standardized uptake value ratio (SUVR) images as the outcome variable (gray matter of the posterior cerebellum reference region) and covarying for age, sex, and education. The first model assessed the effect of T2DM consensus diagnosis on tau uptake. The second model examined the relationship between HbA1c and tau uptake. Both voxelwise regression models were rerun adding global Aβ‐SUVR, obesity, and hypertension as covariates (Figure1).ResultVoxelwise regressions were corrected at voxel p<0.001, cluster FWE p<0.05. T2DM diagnosis was associated with higher 18F‐PI‐2620 uptake in the medial and inferior occipital regions, posterior cingulate, and parahippocampus. Higher HbA1c levels were associated with more 18F‐PI‐2620 uptake in parietal and frontal regions. When covarying for global Aβ‐SUVR, obesity, and hypertension, T2DM diagnosis was associated with 18F‐PI‐2620 uptake in the posterior cingulate. In the HbA1c model, no clusters survived FWE correction after covarying for global Aβ‐SUVR, obesity, and hypertension (Figure1). ConclusionIndependent of common T2DM comorbidities and brain amyloid, T2DM subjects demonstrated a 18F‐PI‐2620 tau uptake pattern that was different from the characteristic age‐associated medial temporal lobe pattern. Our voxelwise approach allowed us to identify regions that may distinguish T2DM‐related pathology from AD‐related changes. The difference in significant clusters between models could be due to HbA1c levels being controlled by medication in some T2DM patients – an area for future study.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call