Predicting Alzheimer’s via resting-state fMRI
Resting-state brain activity patterns can predict Alzheimer’s disease up to 12 years before cognitive symptoms or physical plaque accumulation appear, according to a study led by Stanford Medicine. Published on November 4, 2026, in the journal Science Advances, the research utilizes functional magnetic resonance imaging (fMRI) to map neural firing networks in cognitively unimpaired adults. This approach identifies functional connectome patterns that signal future neurodegeneration.
The study challenges traditional diagnostic benchmarks that rely on comparing older adults to an average peer group. Senior author Dr. Yu Zhang noted that such methods are flawed because the “cognitively normal” baseline harbors wide neurobiological heterogeneity. Instead, the research team contrasted two distinct cohorts: 110 adults with clinically diagnosed Alzheimer’s disease and 180 “super-agers” aged 75 and older who maintained the memory and executive function of 45-year-olds.
Establishing the ARC scoring model
By mapping regions of underactivity and overactivity between the Alzheimer’s and super-ager cohorts, researchers created the Alzheimer’s resilience-vulnerability connectome (ARC) score. This metric effectively distinguishes super-agers from patients with Alzheimer’s disease with 85% accuracy, according to reporting by UA.News.
The team validated the ARC score by applying it to a longitudinal cohort of over 1,300 cognitively intact individuals aged 55 to 70. At the start of the study, participants showed no obvious differences in cognitive testing. However, as the data matured, those with higher ARC scores—indicating brain activity closer to the Alzheimer’s profile—experienced faster declines in memory, attention, and executive function. The longest follow-up period reached 12 years, though most participants were tracked for five to six years.
Comparing ARC against traditional biomarkers
The ARC score demonstrated predictive accuracy that matched or exceeded standard demographic data, genetic risk factors such as APOE-ε4, and fluid or PET biomarkers for amyloid-beta and tau proteins. While high scores correlated with an increased likelihood of amyloid-beta deposits and eventual clinical diagnosis, researchers found that combining ARC scores with existing biomarkers yielded the highest predictive precision.
Despite these findings, the study authors emphasize that the ARC score is not currently a clinical diagnostic tool. Instead, it serves as a method for monitoring cognitive status and early risk assessment. The researchers suggest that functional connectivity signatures may represent modifiable neural circuits, unlike fixed genetic markers or late-stage protein plaques.
Future research into neural interventions
The research team is now shifting focus toward potential non-invasive interventions. Dr. Yu Zhang stated that the group plans to investigate whether methods such as targeted transcranial magnetic stimulation (TMS) or cognitive training can actively guide vulnerable brain activity patterns toward a more resilient profile. These specific interventions were not tested during the initial study, leaving the efficacy of such treatments as a subject for future investigation.
ARC score serves as research tool rather than clinical test
Is the ARC score available for clinical use today?
No, the authors explicitly state that the ARC score should not be used as a diagnostic test at this time. It is currently positioned as a research tool for monitoring cognitive status and assessing long-term risk.
How does the ARC score differ from amyloid-beta testing?
While amyloid-beta and tau biomarkers track physical protein accumulation in the brain, the ARC score measures functional connectivity—the patterns of neural firing and interaction between brain regions. The study found the ARC score could predict decline even before physical plaques were detectable.
What is a “super-ager” in the context of this study?
Researchers defined super-agers as individuals aged 75 and older who possess the memory and executive function performance of a typical 45-year-old. These individuals served as the baseline for “resilient” brain connectivity in the ARC model.
Related reading