Predicting time to dementia using a quantitative template of disease progression

Area of Science:

• Neuroscience
• Biomarker Research
• Alzheimer's Disease Pathophysiology

Background:

• Understanding Alzheimer's disease (AD) progression before clinical diagnosis is crucial for effective prevention and monitoring.
• Longitudinal biomarker trajectories in preclinical AD are not well-characterized due to data limitations.

Purpose of the Study:

• To develop a quantitative method for characterizing the natural history of AD from preclinical stages.
• To estimate the temporal evolution of key AD biomarkers and cognitive measures.

Main Methods:

• Utilized a multivariate Bayesian model to align biomarker data from 1369 participants in the Alzheimer's Disease Neuroimaging Initiative.
• Estimated a quantitative template of biomarker evolution (CSF Aβ, p-tau, t-tau, hippocampal volume, glucose metabolism, cognition).
• Calculated biomarker trajectories relative to AD dementia onset and predicted onset age in a separate sample.

Main Results:

• Early changes were observed in verbal memory, cerebrospinal fluid (CSF) amyloid-beta (Aβ) and p-tau, and hippocampal volume.
• The model accurately predicted AD dementia onset age with a mean error of [insert value] years.
• Established a quantitative template illustrating the temporal sequence of biomarker changes.

Conclusions:

• The developed method offers a quantitative approach to studying AD natural history from preclinical stages.
• This approach overcomes limitations of individual-level longitudinal data spanning the entire disease timeline.
• Provides a framework for understanding and potentially intervening in early AD pathogenesis.