Five Distinct Patterns of Brain Aging Uncovered by 50,000 Scans
A comprehensive analysis of nearly 50,000 brain scans has identified five unique patterns of brain atrophy linked to aging and neurodegenerative diseases. The study also connected these patterns to lifestyle factors such as smoking and alcohol consumption, as well as to genetic and blood markers related to health and disease risk.
Andrei Irimia, a gerontologist at USC, calls this research a major advance in understanding brain aging. While not involved in the study, he notes that it provides deeper insights into how brain anatomy changes with age and disease.
The researchers published the findings in Nature Medicine on August 15.
The Brain’s Aging Process
As we age, our brain undergoes subtle anatomical changes visible on MRI scans, such as shrinkage or structural alterations in certain areas. These changes are hard to detect without advanced technology. “The human eye can’t easily spot the systematic patterns of brain aging,” says Christos Davatzikos from the University of Pennsylvania.
Previous machine-learning studies detected aging signs from MRI data but were limited in scale and included data from few individuals.
To identify comprehensive patterns, Davatzikos’s team spent eight years developing and training a deep-learning method called Surreal-GAN on MRIs from 1,150 healthy individuals and 8,992 older adults with cognitive decline. The algorithm learned to identify recurring features of aging brains and modeled how anatomical structures change together or independently.
The researchers applied the model to MRI scans from nearly 50,000 participants, revealing five distinct brain atrophy patterns. They linked various types of age-related degeneration to these patterns, noting individual variability even among those with the same condition.
Patterns of Aging and Disease
The study linked dementia and mild cognitive impairment to three of the five patterns and found that these patterns could also predict further brain degeneration. Davatzikos noted that one pattern was the strongest predictor of cognitive decline, with an additional pattern improving predictions at later stages. Other patterns were associated with Parkinson’s and Alzheimer’s diseases, and a combination of three patterns strongly predicted mortality.
The study linked brain atrophy patterns to factors like alcohol use, smoking, and genetic markers, suggesting that overall physical health impacts neurological well-being.
Despite the findings, Davatzikos warns that the study does not simplify brain aging into just five patterns. His team plans to expand their research using data from a broader range of neurological conditions and a more diverse population.
Read the original article on: Nature
Read more: Brain Found to Retain Three Copies of Each Memory
