You walk out of the grocery store, ready to head home and make dinner. But as you scan the crowded parking lot, a frustrating question comes to mind. Where did you leave your car? You try to picture the spot, but the memory refuses to surface.
Remembering personal experiences like these depends on episodic memory, which allows us to recall events from our lives. These memories help form the foundation of long-term memory, whether they involve a childhood milestone, such as the first day of elementary school, or something that happened just minutes earlier, like parking your car.
This ability tends to weaken as we grow older. The sights, sounds, and other details associated with past experiences can become increasingly difficult to retrieve. Scientists still do not fully understand why this happens. Although some memory decline is a common part of aging, the changes can become far more severe in people developing Alzheimer’s disease.
A new study from researchers at the University of California, Berkeley, offers important clues about what may be happening inside the aging brain. Neuroscientist Omer Sharon and his colleagues, working under the direction of former Berkeley sleep researcher Matthew Walker, discovered a connection between disrupted brain waves during deep sleep, declining memory performance, and the accumulation of tau, a protein associated with Alzheimer’s disease.
The findings suggest that tau buildup in the brain’s frontal cortex is linked to changes in the coordinated electrical activity that helps preserve memories overnight.
"This is the first time we’ve shown this relationship between tau and how it messes with memory consolidation by attenuating traveling slow waves that originate in the frontal cortex," said Sharon, a postdoctoral researcher at UC Berkeley’s Center for Human Sleep Science.
The research was published in the journal Nature Neuroscience.
How Deep Sleep Helps the Brain Store Memories
Sleep involves several distinct stages, each associated with different patterns of brain activity. Rapid eye movement, or REM, sleep is particularly recognizable because the eyes move beneath closed eyelids and most dreaming takes place during this stage. However, another stage, known as non-REM sleep, is especially important for maintaining healthy brain function and memory consolidation.
During this deeper stage of sleep, when dreaming is limited or absent, large groups of neurons briefly become inactive and then active again in a coordinated sequence. These synchronized changes produce slow electrical waves that spread across the brain, typically beginning in the frontal cortex and rippling outward to other regions.
"Each slow wave reflects an enormous population of neurons switching off and back on together," Sharon said. "Our study shows that for memory, it’s critical that these events cascade in sequence across large parts of the brain. This coordinated shutdown has been observed only during deep sleep, and may be one reason why we sleep."
The researchers found that this carefully organized activity becomes less coordinated as tau accumulates in the aging brain. Instead of traveling together across broad regions, the waves become more scattered and cover shorter distances. These changes are closely associated with weaker memory consolidation, declining cognitive abilities, and the pathology underlying Alzheimer’s disease.
Alzheimer’s Protein Linked to Disrupted Brain Waves
To investigate the relationship between tau and electrical activity during sleep, the scientists monitored sleeping brain activity using electroencephalograms, commonly known as EEGs. They compared cognitively healthy young adults in their early 20s with older participants ranging from their mid-60s to mid-70s.
The differences observed between the two age groups were striking. In younger participants, groups of slow waves moved together smoothly across distances roughly equivalent to the width of a human hand on the scalp. Among older adults, however, the waves typically traveled less far and appeared much more independently, rather than moving in the broad, coordinated groups necessary for optimal memory processing.
The research team also examined whether these structural changes in sleep brain waves were associated with tau accumulation. Working with Dr. William Jagust, a UC Berkeley professor emeritus of public health and neuroscience, the researchers utilized positron emission tomography, or PET, scans to closely examine the brains of older participants. This advanced imaging technique uses radioactive tracers introduced safely into the bloodstream to reveal biological activity and map the precise distribution of specific molecular substances, including pathological proteins.
The detailed scans showed that greater tau buildup in the frontal cortex was directly associated with a breakdown in the longer traveling waves normally seen during healthy deep sleep.
Importantly, these physical and electrical changes were present in people who had not developed clinical Alzheimer’s disease or dementia.
"These people did not have Alzheimer’s," Sharon noted. "They had tau in their brains, but with subclinical impacts. Their memory decline was within the normal range for their age."
Sharon explained that early Alzheimer’s-related changes can vary considerably in severity across the population. The amount of tau present in the brain is associated with subtle cognitive performance variations, even when a person’s overall memory abilities remain within the expected range for their chronological age.
Mysterious "Lonely Waves" Linked to Memory Loss
Beyond observing the physical presence of the protein and the altered brain waves, the researchers wanted to determine whether these disrupted sleep patterns were associated with actual, measurable differences in how well people remembered information in their daily lives.
To test this hypothesis rigorously, participants were asked to learn pairs of associated words in the evening under controlled laboratory conditions. The following day, after a full night of monitored sleep, they were tested on their ability to recall what they had learned. This experimental design allowed the scientists to evaluate how effectively the memories had been consolidated and retained overnight.
The behavioral results matched the neurological data. Participants whose slow brain waves traveled shorter distances and appeared more isolated generally remembered significantly fewer word associations the next day. The researchers colorfully referred to these disconnected, uncoordinated patterns as "lonely waves."
Furthermore, the research team tracked a subset of these participants over several consecutive years to see how their brains changed with time. Among the individuals whose frontal cortex tau levels increased over the monitoring period, the slow waves became progressively less coordinated, and their ability to retain newly learned information overnight steadily declined.
These observations provide compelling evidence of a longitudinal relationship between tau accumulation, changes in deep sleep brain activity, and worsening episodic memory. However, the researchers emphasize that these findings do not yet establish whether tau buildup directly causes memory loss or whether the disrupted electrical waves are the primary mechanism responsible for the cognitive decline.
Additional Tests Strengthen the Connection
Because PET scans are costly, resource-intensive, and require specialized medical facilities, the scientists sought another way to investigate the complex relationship between Alzheimer’s-related proteins and sleep activity.
To expand their approach, they collaborated with neurologist Yo-El Ju at Washington University in St. Louis to analyze spinal fluid samples from a completely separate group of older adults. The primary goal was to determine whether independent measurements of Alzheimer’s-associated proteins would reveal patterns similar to those seen with the brain imaging techniques.
Unlike PET scans, the spinal fluid measurements could not identify tau accumulation specifically within the frontal cortex alone. Nevertheless, the independent results pointed in the exact same direction.
People whose sleeping brains produced more isolated slow waves had significantly higher ratios of tau to amyloid, another hallmark protein closely associated with the pathology of Alzheimer’s disease, present in their spinal fluid.
The consistency between these independent study groups and different testing methods greatly strengthened the researchers’ confidence in the validity of the findings.
"It’s collected in a totally different place and in a different cohort, so I think that adds to the strength of the evidence," Sharon said.
Together, the comprehensive sleep recordings, brain scans, spinal fluid analyses, and memory assessments suggest that tau accumulation is consistently associated with a very particular disruption of brain activity during deep sleep.
"This longitudinal correlation suggests that Alzheimer’s disease pathology is associated with a disruption of this main sleep feature," Sharon said.
What Deep Sleep Reveals About the Aging Brain
Although memory often naturally declines to some degree with age, the new research indicates that the underlying process may involve much more than simply growing older. Changes subclinically associated with Alzheimer’s disease—particularly tau accumulation in brain regions responsible for generating widespread slow waves—may help explain why some older individuals experience significantly greater difficulty retaining memories than others.
"It’s not just about age," Sharon said. "It’s how much pathology you have in frontal areas where global waves originate."
Sleep quality also appears to be an integral part of the larger picture. Although sleep can be notoriously difficult to define or measure precisely in everyday settings, the organization and movement of slow brain waves may provide critical indicators of how effectively the human brain performs its essential nighttime memory functions.
The findings raise the possibility that tracking changes in deep sleep activity could eventually help researchers and clinicians better understand, detect, or identify early memory problems before they progress. They also point toward potential targets for future therapeutic treatments aimed at protecting and preserving memory as people age.
For now, however, a central question remains unanswered. Scientists still need to determine whether accumulating tau directly contributes to the breakdown of synchronized sleep waves, whether changes in those electrical waves influence tau accumulation over time, or whether the relationship involves other complex, underlying biological processes.
"We can see tau and lonely waves rising together over time," Sharon said. "What we cannot yet say is which leads—and that question will shape how we design interventions."
The study’s additional co-authors include Xi Chen, James Westphal, Chelsea Brown, and Vyoma D. Shah of UC Berkeley, along with Jason Dude of Washington University in St. Louis.
Funding for the research came from major grants awarded by the National Institute on Aging at the National Institutes of Health. Sharon also received individual support through a Glenn Foundation for Medical Research Postdoctoral Fellowship and a Weill Neurohub Fellowship.