Alzheimer signs emerge up to twelve years before clinical diagnosis, according to a study published in Science Advances.
Researchers built an artificial intelligence model that analyzes resting-state functional connectivity in the brain, WELT reported. The team formed the Alzheimer’s Resilience-Vulnerability Connectome score, known as the ARC-scale, by comparing brain scans of 180 super-agers with 110 Alzheimer patients. Super-agers are individuals over 75 with memory capacities comparable to 45-year-olds.
The model achieved roughly 80 percent accuracy when tested on an independent group of 143 people, ad-hoc-news.de reported. Scientists then applied the score to more than 1,300 cognitively unimpaired adults aged 55 to 70 over a twelve-year observation period. Adults whose brain resting patterns resembled the Alzheimer end of the scale showed faster memory decline and increased amyloid accumulation over time.
The team trained the system using magnetic resonance imaging scans that measured how different brain regions communicate while subjects lay still. Initial scans showed no differences in amyloid and tau protein levels between the two extreme groups, indicating the communication patterns shift before classical biomarkers appear.
“Our model shows comparable or better predictive power compared to established biomarkers, and at an earlier time point,” said Kanhao Zhao, first author of the study.
Participants with vulnerable brain patterns also displayed biologically older brains based on neural activity rather than chronological age. The ARC-score also linked poor memory profiles with apathy and depression.
The entorhinal cortex and the amygdala lose tissue more rapidly in early disease stages than the hippocampus, according to a study from Johns Hopkins University published in NeuroImage: Reports. These structural losses occur up to a decade before memory impairments manifest in patients, scientists noted during the Society for Neuroscience annual congress.
“We have early changes in the brain, which show up as tissue loss, up to a decade before there are clinical symptoms,” said Michael Miller, biomedical engineer at Johns Hopkins University.
Researchers caution that the ARC-score remains a snapshot rather than a definitive personal prognosis.
“An ARC-value is just a snapshot, not a verdict,” Yu Zhang, lead author of the study, told WELT.
The Stanford team plans to investigate whether targeted cognitive training can shift abnormal brain connectivity patterns toward a super-ager profile.
Meanwhile, regulatory bodies continue approving complementary diagnostic tools, including a blood test approved by Japan’s Ministry of Health for detecting amyloid-beta accumulation, ad-hoc-news.de reported.
“The critical thing is that we have to intervene early,” Zhang said.
Keep reading