Brain Connectivity Patterns May Predict Alzheimer’s Disease Risk Years Before Memory Slips

Stanford Medicine researchers have identified resting brain activity patterns capable of predicting whether cognitively unimpaired adults are headed toward Alzheimer’s disease or cognitive resilience years before the emergence of memory loss or the accumulation of amyloid-beta plaques and tau tangles. In a study published in Science Advances, investigators analyzed functional magnetic resonance imaging data to map neural firing patterns across opposite ends of the cognitive spectrum. The resulting neuroimaging framework suggests a novel class of early, potentially modifiable biomarkers for late-life cognitive decline.

To construct the model, the research team examined resting brain activity in individuals aged 75 and older, contrasting patients diagnosed with Alzheimer’s disease against so-called super agers who preserve the memory capabilities of individuals decades younger. Extrapolating from these two distinct clinical groups, the investigators formulated a quantitative scoring metric designed to position adults aged 55 to 70 who show no overt signs of cognitive impairment along that physiological spectrum based solely on neural connectivity measurements.

Within the study’s cohort, super agers were characterized as adults aged 75 or older who exhibited memory capacity typical of a 45-year-old alongside normal executive function for their age group, a profile estimated to represent approximately 10 percent of older adults depending on applied diagnostic benchmarks. While prior investigations into cognitive resilience have conventionally contrasted super agers with typical peers, the authors noted that typical older populations display substantial underlying biological heterogeneity, with some individuals already developing early amyloid-beta pathology or other subclinical indications of brain aging.

To overcome this confounding biological variation, senior author Yu Zhang, an assistant professor of psychiatry and behavioral sciences, and colleagues avoided comparisons against heterogeneous cohorts and instead evaluated super agers directly against similarly aged patients diagnosed with Alzheimer’s disease. Drawing upon neuroimaging datasets from the Open Access Series of Imaging Studies, the Harvard Aging Brain Study, and the Alzheimer’s Disease Neuroimaging Initiative, the researchers analyzed resting-state functional magnetic resonance imaging scans from 180 super agers and 110 individuals with Alzheimer’s disease to chart regional functional connectivity.

The functional neuroimaging analysis revealed consistent configurations of regional underactivity and overactivity across the functional connectome that differentiated super agers from patients diagnosed with Alzheimer’s disease.

The investigators subsequently evaluated the ARC scoring model across a prospective sample of more than 1,300 adults aged 55 to 70 who presented with no baseline indicators of cognitive impairment. Initial neuroimaging yielded a continuous range of intermediate connectome scores across the cohort, wherein lower scores corresponded to super-ager-like connectivity patterns and higher scores reflected activity profiles resembling Alzheimer’s disease, with no discernible cognitive differences observed among participants at the initial time of imaging.

Over extended follow-up intervals spanning up to 12 years, with most participants monitored across five to six years, longitudinal data revealed distinct cognitive trajectories tied to baseline ARC measurements. Individuals presenting with higher, Alzheimer’s-like scores experienced accelerated cognitive decline on standardized assessments of memory, attention, and executive function, alongside increased likelihood of developing amyloid-beta plaques and receiving clinical Alzheimer’s diagnoses. Conversely, individuals exhibiting lower, super-ager-like baseline scores demonstrated the slowest rates of cognitive decline.

According to lead author Kanhao Zhao, a postdoctoral scholar in psychiatry, the predictive performance of baseline ARC scores met or exceeded that of traditional demographic characteristics, genetic factors, and established amyloid-beta and tau biomarkers, while identifying risk at earlier time points.

The study also documented a measurable correlation between the connectome metric and brain age, an established neuroimaging index derived from functional magnetic resonance imaging investigations of healthy populations. However, the authors noted that the ARC score provides a more targeted evaluation of cognitive resilience and vulnerability to dementia within aging cohorts than generalized measures of physiological brain maturation.

The authors emphasized that the ARC score currently functions as a monitoring guide rather than a standalone diagnostic tool, pointing out that an individual’s connectome profile represents a physiological snapshot rather than an immutable prognosis. Connectome scores often shift negatively over time as individuals age, sustain brain trauma, or develop neuropsychiatric disorders, underscoring the dynamic nature of functional neural networks in older adults.

Looking toward clinical applications, the researchers aim to investigate whether targeted neuromodulation or behavioral interventions can actively shift brain activity patterns toward the resilient profile identified in super agers. While techniques such as transcranial magnetic stimulation have demonstrated clinical utility in modulating brain activity to treat depression, such interventions have not yet been evaluated for reversing cognitive decline, prompting the team to plan prospective trials measuring ARC scores before and after structured cognitive training.

The collaborative research effort included investigators from Children’s National Hospital, the University of Texas at Austin, Lehigh University, Penn State College of Medicine, New York University Grossman School of Medicine, and the University of California, Los Angeles. Research support was provided by multiple grants from the National Institutes of Health, a joint initiative by the Rosenkranz Foundation and Knight Initiative, the Alzheimer’s Association, and Lehigh University.

Addressing a central challenge in Alzheimer’s research served as the primary motivation for the investigation, as senior author Yu Zhang noted that widely evaluated biomarkers such as amyloid-beta and tau are observable only at relatively late disease stages. “We want to address a major challenge in the Alzheimer’s field nowadays — most studies are looking at well-known Alzheimer’s disease biomarkers, such as amyloid-beta or tau. But those are observable at relatively late stages of the disease,” Zhang said. “What we really care about is defining early markers that can detect earlier signs of cognitive decline and potentially help us deliver timely interventions to people.” By analyzing functional firing patterns in the resting brain, the research team sought to identify predictive indicators long before physical pathology emerges.

The researchers also identified a fundamental issue in earlier resilience research, which typically compared super agers against cognitively typical older adults. Zhang realized that cognitively unimpaired peers remain biologically heterogeneous, with some individuals already developing early amyloid-beta accumulation or other neurobiological signs of aging beneath normal cognitive test scores. “Normal, cognitively unimpaired individuals are pretty heterogeneous in terms of their biological abnormalities,” Zhang noted. “Simply comparing super agers with such a heterogeneous population makes it challenging to define what is critical in terms of resilient or vulnerable brain patterns.” Comparing those cohorts was akin to comparing apples to a fruit basket. By instead contrasting super agers directly with patients diagnosed with Alzheimer’s disease, the investigators established distinct polarized baselines to isolate resilient neural networks.

Lead author Kanhao Zhao highlighted how the resulting scoring system applies to prospective patient monitoring and risk prediction. “We used the ARC score to make a prediction of how likely this patient will convert into mild cognitive impairment or Alzheimer’s disease,” Zhao explained. Across follow-up periods extending up to 12 years, baseline connectome scores demonstrated robust predictive validity. “We compared the ARC score with existing biomarkers in predicting people converting to Alzheimer’s disease,” Zhao said. “Our model shows comparable or better prediction ability, and it can do so at an earlier time point.

Senior author Zhang expressed enthusiasm regarding the broader scientific utility of the connectome signature, noting that it provides specialized clarity where general indicators fall short. “I’m excited we established a signature that can help us better understand the difference between people who can maintain excellent cognitive performance and people who show cognitive decline,” Zhang said. While connectome scores show some correlation with broader brain age metrics derived from resting-state imaging in healthy populations, the ARC score serves as a more specific gauge of resilience and vulnerability to cognitive decline in older cohorts. By charting regional underactivity and overactivity, the model captures specific neural configurations associated with lasting memory preservation.

Looking toward future therapies, the researchers highlighted that defining resilient connectomes provides a tangible biological target for emerging noninvasive interventions. “There are two main components: One is to know which part of the brain to target, the other is how we should change it,” Zhang explained. “That’s why identifying the super-ager brain resilience pattern is a critical component — so we know what a good pattern should be.” Modalities like transcranial magnetic stimulation have shown promise in altering brain activity to treat depression, but they have not been tested for cognitive decline, and it remains unknown whether changing connectivity can redirect disease trajectories. Nevertheless, Zhang emphasized the importance of timely action: “The critical part is that we have to intervene at an early stage. That’s why we focus on the cognitively unimpaired population to identify super-ager-like versus Alzheimer’s-like individuals. If we can precisely identify such as early indicator, I’m optimistic that we can reverse or prevent this process through advanced intervention technologies.”

Subject of Research: Functional connectome patterns and early prediction of Alzheimer's disease risk

Article Title: Brain activity patterns may predict Alzheimer’s years before memory slips

Article References: Brain activity patterns may predict Alzheimer’s years before memory slips. Original publication (Bibliographic details unavailable.)

Image Credits: AI Generated

DOI: Not provided

Keywords: Alzheimer's disease, functional connectivity, resting-state fMRI, ARC score, super agers, cognitive resilience, neuroimaging, Science Advances, Stanford Medicine, amyloid-beta, tau biomarkers, cognitive decline, brain age, transcranial magnetic stimulation, executive function, longitudinal neuroimaging