Mullins · Human brain mapping 2017 · Cross-sectional observational study · n=24

Exosomal biomarkers of brain insulin resistance associated with regional atrophy in Alzheimer's disease.

Cited 131 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional biomarker and neuroimaging correlation study in a single clinical cohort.

PubMed 28105773 · doi:10.1002/hbm.23494 · record verified 2026-08-31

What was done

Neural-origin plasma exosomes were isolated via L1CAM immunoprecipitation in 24 patients with biomarker-confirmed probable Alzheimer's disease (low CSF Aβ42). Investigators measured two insulin-receptor-substrate-1 (IRS-1) phosphotypes: pSer312-IRS-1 (ineffective insulin signaling) and p-panTyr-IRS-1 (effective insulin signaling). Voxel-based morphometry (VBM) on structural MRI was conducted to assess associations between gray matter volume and both biomarker phosphotypes. Spatial patterns of VBM associations were compared against normal brain IRS-1 expression data from the Allen Human Brain Atlas.

What was found

Brain gray matter volume was positively associated with p-panTyr-IRS-1 and negatively associated with pSer312-IRS-1 in a matching regional pattern involving the bilateral parietal-occipital junction and the right middle temporal gyrus. This volumetric association pattern spatially correlated with normal human brain IRS-1 expression from the Allen Human Brain Atlas. The abstract reports directional findings and anatomical regions but contains no exact correlation coefficients, effect sizes, or p-values.

Why it matters

This study provides preliminary evidence that blood-derived neural exosomes can reflect central molecular insulin signaling abnormalities associated with localized brain atrophy in Alzheimer's disease.

Limits

The study is limited by a very small sample size (n = 24), a cross-sectional design that cannot determine causality, and the absence of a healthy control comparator group for imaging correlations. Spatial comparisons relied on a healthy reference brain atlas rather than patient-specific tissue measurements.

Cited by