Why do some older adults easily remember the name of a new acquaintance or the location of their keys, while others struggle with everyday recall—despite appearing equally healthy?
A new study affiliated with the Knight Initiative for Brain Resilience at Stanford University, recently published in Science Advances, suggests part of the answer lies along two distinct but converging pathways in the aging brain: one shaped by early Alzheimer’s disease pathology, the other by the brain’s capacity to focus.
The study was conducted by a team of Stanford scientists that includes postdoctoral researcher Jintao Sheng and cognitive neuroscientist Anthony Wagner. It draws on data from the Stanford Aging and Memory Study (SAMS), a landmark longitudinal project on cognitive aging co-led by Wagner and Elizabeth Mormino and supported by the National Institute on Aging. By following cognitively healthy individuals over many years, SAMS offers researchers a powerful window into why some people maintain sharper memories into their 70s and 80s, while others falter.
“We’re trying to understand which of these factors—or set of factors—explain why some older adults are more resilient, maintaining memory function late into life, while others show a decline,” said Wagner, Lucie Stern Professor in the Department of Psychology and a deputy director of Stanford’s Wu Tsai Neurosciences Institute, which houses the Knight Initiative.
Precision, Focus, and the Brain’s “Tuning Knob”
At the heart of the new research is the concept of “neural selectivity,” the tendency for certain neurons or brain regions to respond more strongly to specific types of information, such as faces or places. The selectivity helps form robust memories, much like tuning a radio to a clear signal. But as people age, tuning becomes less precise, resulting in “fuzzy” memory traces. Scientists call this phenomenon “neural dedifferentiation.”
“A memory of a life event consists of a collection of features—who, what, where, when—bound together in the brain,” explained Wagner. “As we age, the brain’s ability to sharply represent these features can begin to break down, even in people who seem cognitively healthy… But the mechanisms for why this occurs were less clear.”
To investigate, Sheng used functional MRI scans along with blood and spinal fluid samples from 166 older adults enrolled in SAMS. All participants were cognitively healthy at the time of testing. She found that memory performance was closely linked to how precisely the brain encoded details of what the participants were learning, or neural selectivity.
This precision, it turned out, was shaped by two largely independent factors.
Two Paths, One Destination
The first factor was the presence of early Alzheimer’s disease pathology. Participants with higher levels of tau protein—a key marker of Alzheimer’s—showed lower neural selectivity in scene-processing regions of the brain, even though they had no clinical symptoms of the disease. Participants with elevated tau had a harder time encoding memories, in part because their neural representations were less distinct, as though early signs of Alzheimer’s had subtly turned down the brain’s “tuning knob” for experience.
The second factor was attention. Specifically, the researchers looked at the dorsal attention network, a set of brain regions that help us focus on what matters most based on our goals. “This is the network that we use, given our goals, to engage with the world—to focus on things that are task-relevant,” Wagner said.
In the experiment, researchers scanned people’s brains as they studied pairs of words and images—either famous faces like Queen Elizabeth’s or landmarks like the Eiffel Tower—and asked them to make a meaningful link between the two, such as a brief mental story. Next, the team showed participants the words again and asked them to recall whether each had been paired with a face or a place. Brain activity revealed that people who more strongly engaged their attention network formed clearer, more detailed neural representations during learning and performed better on the later memory test. This effect of attention was independent of the effect of early Alzheimer’s pathology.
“Even more interesting,” said Sheng, a Brain Resilience Postdoctoral Research Fellow with the Knight Initiative, “was that moment-to-moment attention really mattered. The brain was encoding memories more clearly at times when participants were paying more attention to the task.”
The Implications for Healthy Aging
The study stands out because it identifies not one, but two distinct and largely independent pathways that influence how memories are formed in the aging brain: one related to early Alzheimer’s pathology, and the other to how effectively a person pays attention during learning.
This distinction matters. While we can’t yet alter the biological course of Alzheimer’s disease, attention is a cognitive function that may be more flexible. That makes it a potential target for supporting memory resilience, even in individuals in the early stages of Alzheimer’s.
“Older adults who showed stronger top-down attention—meaning they were better at focusing on what was important during learning—had higher encoding precision and better memory,” said Sheng. “Even in brains with early Alzheimer’s changes, attention still helped support memory. That’s promising, because it suggests attention might be one factor we could support to help maintain memory function as people age.”
Going forward, the Stanford team seeks to better understand the mechanisms that lead to altered neural selectivity as well as to identify other pathways that contribute to memory change. This includes plans to use tau PET—an imaging technique that shows where in the brain toxic tau proteins have built up. Under the leadership of Mormino, an assistant professor of neurology, tau PET scans have already been collected as part of SAMS and may help clarify how early Alzheimer’s pathology disrupts memory circuits.
The researchers are also exploring how the hippocampus, another memory-critical region, interacts with cortical areas during encoding. In collaboration with Knight Initiative Director Tony-Wyss Coray and other Knight Initiative researchers who bring tools like proteomics—the large-scale study of proteins—the team hopes to uncover molecular signatures of memory vulnerability or resilience. And as SAMS continues to collect long-term follow-up data, including repeat imaging and cognitive testing seven years after the initial study, the researchers will be able to ask a powerful new question: which early brain and biological signals predict memory decline—or resilience—over time.
For Wagner, the study is part of a growing body of research that shows just how multifaceted—and individual—memory aging is. “Different systems and processes interact to shape what we learn and remember,” he said. “And those systems change in different ways for different people. Understanding those patterns is a key step toward helping individuals thrive as they age.”
Publication Details
Research Team
Study authors were Jintao Sheng, Alexandra Trelle, America Romero, Jennifer Park, Tammy Tran, Sharon Sha, Katrin Andreasson, Edward Wilson, Elizabeth Mormino, and Anthony Wagner of the Stanford Department of Psychology, the Stanford School of Medicine's Department of Neurology and Neurological Sciences, and Wu Tsai Neurosciences Institute.
Research Support
This work was supported by the Knight Initiative for Brain Resilience at the Wu Tsai Neurosciences Institute, the National Institute on Aging (R01AG048076, R01AG074339, R21AG058859, K99AG075184, and NRSA F32AG071263), and The Alzheimer’s Association (AARFD-21-852597).
Competing Interests
Sharon Sha has disclosed additional potential conflicts of interest, including research support from Biogen, Eisai, Eli Lilly, Aribio, Janssen, Genentech, and Cognition Therapeutics. She has also participated in consulting activities for Guidepoint Global, delivered speaking engagements for Peerview and Medscape, and serves on the advisory board of Cognition Therapeutics. All other authors declare that they have no competing interests.
