Change point analyses in prodromal Alzheimer’s disease

Change point analysis can reveal when a biomarker starts to diverge from the pattern of normal aging. This paper analyzes several biomarkers from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) to estimate the sequence and timing of their change points relative to a subsequent clinical diagno...

Full description

Bibliographic Details
Main Authors: Alvin H. Bachman, Babak A. Ardekani
Format: Article
Language:English
Published: Elsevier 2020-12-01
Series:Biomarkers in Neuropsychiatry
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2666144620300186
Description
Summary:Change point analysis can reveal when a biomarker starts to diverge from the pattern of normal aging. This paper analyzes several biomarkers from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) to estimate the sequence and timing of their change points relative to a subsequent clinical diagnosis of mild cognitive impairment (MCI) in subjects initially considered cognitively normal (CN). Data on 379 stable CN (sCN) and 98 progressive CN (pCN) subjects who progressed to an MCI diagnosis were used. Linear mixed-effects change point models were used to estimate when various biomarkers in pCN started to diverge from rates expected in normal aging. Our results indicate that in pCN, hippocampal atrophy rate diverges from normal aging 12.4 (±2.8) years before MCI diagnosis, followed by ventricular volume expansion and decrease in Rey Auditory Verbal Learning Test of immediate recall scores about 5 years later. Glucose metabolism decrease begins about 5 (±1.3) years before diagnosis, followed by deterioration in other cognitive test scores. Planned AD interventions should note that irreversible changes such as atrophy may occur a decade before possible diagnosis of MCI.
ISSN:2666-1446