Researchers develop a model to estimate Alzheimer’s risk long before symptoms
A team at the Mayo Clinic has developed a predictive model that estimates an individual’s 10-year and lifetime risk of cognitive decline — including mild cognitive impairment (MCI) and dementia — using decades of data from the Mayo Clinic Study of Aging. The work, published in The Lancet Neurology, analyzed brain imaging, genetics and medical records from more than 5,800 adults to build the tool.
How the model works
The model quantifies amyloid protein accumulation in the brain using specialized imaging and reports that measurement on a scale of 0–100, where higher values indicate greater amyloid burden and greater “biological severity” of Alzheimer’s pathology. The team combined amyloid scores with age, sex and the presence of the APOE ε4 genetic variant, then applied advanced statistical methods to estimate each person’s probability of developing MCI and progressing to dementia over time.
"This kind of risk estimate could eventually help people and their doctors decide when to begin therapy or make lifestyle changes that may delay the onset of symptoms," said Ronald Petersen, M.D., Ph.D., co-author and director of the Mayo Clinic Study of Aging. "It's similar to how cholesterol levels help predict heart attack risk."
Key findings
Researchers found a clear relationship between amyloid level and estimated risk: higher amyloid scores were associated with larger 10-year and lifetime risks of cognitive decline. Women had higher lifetime risk estimates than men overall, and carriers of the APOE ε4 variant were more likely to experience decline than non-carriers. As an example, a 75-year-old woman in the study who carried the APOE ε4 gene and had high amyloid levels faced an estimated lifetime risk of more than 80% for developing MCI.
Limitations and next steps
The authors note several limitations: the cohort was predominantly older, white adults from a single geographic region, which may limit generalizability. The current model relies on costly brain imaging that is not widely available and did not incorporate lifestyle or health behaviors that can influence cognitive trajectories. For now, the risk calculator is intended for research use only.
Mayo Clinic researchers say the work is an important step toward personalized Alzheimer’s prevention. Future versions of the tool may incorporate accessible biomarkers such as blood-based measures of amyloid or other indicators to broaden clinical utility without the need for specialized imaging.
Funding: National Institute on Aging, the GHR Foundation, Gates Ventures and the Alexander Family Foundation.
