Integrated amyloid, neurodegeneration, and vascular biomarkers estimate the risk of dementia progression in mild cognitive impairment

Abstract

Background: Progression from mild cognitive impairment (MCI) to dementia arises from heterogeneous mechanisms involving Alzheimer&apos;s disease (AD) pathology and vascular factors. Objective: This study aimed to develop a multimodal biomarker model to estimate the risk of dementia progression and to examine whether carotid atherosclerosis provides independent prognostic value, particularly in amyloid-beta (A beta)-negative MCI. Methods: We retrospectively analyzed 300 individuals with MCI who underwent baseline [F-18]florbetapir PET, structural MRI, carotid Doppler ultrasound, cognitive assessments, and APOE genotyping (2018-2021). Participants were followed for a total of 37 months to dementia based on longitudinal cognitive and functional decline, independent of follow-up amyloid PET findings. A beta positivity was defined using Brain Amyloid Plaque Load criteria. Multivariable logistic regression and receiver operating characteristic analyses were performed. Results: Among 189 A beta-positive individuals, 30.7% (n = 58) progressed to A beta-positive AD dementia, while 9.0% (n = 10) of 111 A beta-negative individuals developed non-AD dementia. Independent factors estimating A beta-positive AD dementia included higher A beta burden (OR 2.34, p < 0.001), smaller hippocampal volume (OR 0.71, p < 0.001), greater carotid plaque count (OR 1.45, p = 0.001), lower Mini-Mental State Examination (OR 1.11, p = 0.002), and APOE epsilon 4 carriage (OR 1.82, p = 0.021). The integrated model showed excellent performance (AUC 0.903; 95% CI: 0.814-0.968). In A beta-negative MCI, carotid plaque burden was the primary estimator of non-AD dementia progression. Conclusions: The prominent prognostic role of carotid plaques in A beta-negative MCI underscores the vascular contributions to non-amyloid cognitive decline and highlights the importance of evaluating both AD-related and vascular mechanisms in prodromal dementia.