Smartwatch Study Finds Older Adults Can Sense Their Own Mental Decline in Daily Life

A week of smartwatch data isn't going to tell someone whether they are developing dementia
Lead author Brandon Gavett stresses the limits of the study's short, one-week window.
Mark

Start with the basic finding. What did the watches actually show?

Mimi

That when people rated themselves less sharp than their own usual, their scores on short attention and speed tests tended to be lower too. It's a within-person comparison, against each person's own average.

Mark

Why does the real-time part matter so much?

Mimi

Because the usual method asks people to recall a whole year, and recall is biased. Doctors also tend to see patients on a day they may be feeling better. Asking four times a day sidesteps some of that.

Luke

Before we get carried away, it's 62 people over one week, all of whom tested within normal ranges. That's a small, short study.

Mimi

Fair, and the researchers say so. Gavett is explicit that a week of data won't tell anyone if they're developing dementia.

Mark

Farias said the mood result surprised her. Why?

Mimi

Depression is known to make people feel their memory is worse, so skeptics worried self-ratings just reflect mood. Here, the link between sharpness ratings and performance didn't appear to be explained by mood.

Luke

"Didn't appear to be" is doing real work there. It's one study's finding, and nothing here shows how it holds in people with depression or in larger groups.

Mark

And the hope of early detection?

Luke

That's the aspiration, not a result. Nobody has shown yet that these patterns predict future decline. That needs months or years of follow-up.

Mimi

Which Farias names as the biggest open question. Next come clinical trials, then larger populations.

  • A week of buzzing wrist prompts asks older adults, four times a day, how sharp they feel, and the answers are measured against quick cognitive tests.
  • Conventional memory check-ups rely on recall of the past year, a vulnerability that leaves clinicians uncertain about what patients truly experience.
  • Researchers have worked to separate genuine cognitive signals from depression, which can itself make people feel their memory is failing, and found the link persisted.
  • The findings are promising but early, and the team is moving toward clinical trials and multi-month or multi-year follow-up to learn whether these patterns predict dementia.

Somewhere in the long human effort to understand the slow erosion of memory and attention, a small study from the University of California, Davis suggests that the older person may know something their doctors have had trouble measuring. By prompting sixty-two cognitively healthy adults several times a day over a week, researchers found that moments of felt mental dullness tended to coincide with weaker results on brief tests, and that this link held apart from mood. The question it leaves open is less about the present than about the future: whether these small daily reckonings of the self can foretell the larger decline that many people fear.

Four times a day for a week, a smartwatch asked each of 62 older adults to rate how mentally sharp they felt and what their mood was. After the answers were logged, the watch offered a short test of processing speed and attention. Every participant began within normal range on standard cognitive tests, and the researchers accounted for age, mood, and time of day. When someone rated their sharpness below their own average, their scores on the objective tasks also tended to dip. Both ratings and performance drifted lower later in the day. The results appeared in Neuropsychology.

The approach matters because it measures thinking as people live it, rather than through memory or a single clinic visit. Sarah Tomaszewski Farias, a UC Davis neurologist and study author, explained that asking someone to look back over a year invites recall bias, and patients often get screened when they feel their best. Brandon Gavett, the first author, put the difficulty simply: measuring change in cognition is really hard.

Mood has long complicated any attempt to trust self-assessment, since depression can make memory seem worse. For Farias, the most surprising and encouraging finding was that the link between perceived sharpness and actual performance was not explained by mood.

Outside experts welcomed the work with caution. Heather Snyder of the Alzheimer's Association, who was not involved, said perceptions are sometimes the first indicator of change. The practical limits are real, though. Some older adults may not want continuous prompts, and a week of data cannot diagnose dementia. The central question now is whether months or years of such data can predict future impairment. The team plans further clinical trials before wider use, with Gavett saying the goal is getting people help for dementia as early as possible.

Four times a day for a week, a smartwatch on the wrist of an older adult buzzed with the same request: rate how mentally sharp you feel, and how your mood is. Once the answers were logged, the watch offered a short test of processing speed and attention. That routine, repeated across 62 participants, is the backbone of a University of California, Davis study that suggests older adults may be better judges of their own moment-to-moment thinking than scientists once assumed.

All 62 participants performed within normal ranges on standard cognitive tests when they started. The researchers accounted for age, mood and time of day. What they found was that when a person rated their sharpness below their own average, their results on the objective tasks also tended to be lower. Both the ratings and the performance tended to slip later in the day. The work was published in Neuropsychology, a journal of the American Psychological Association.

The design is unusual because it captures cognition as people live it, rather than relying on memory or a single office visit. Sarah Tomaszewski Farias, a study author and professor of neurology at UC Davis, described the problem with the usual approach. Doctors and researchers typically ask someone to look back over the past year and assess how their thinking has been, and those assessments are vulnerable to recall bias. When patients do report cognitive trouble, clinicians depend on the patient's recollection and often screen them when they may be feeling better. Asking repeatedly about the present moment could sidestep some of those limits. "Measuring change in a person's cognitive functioning is really hard," said Brandon Gavett, the study's first author.

There is also a long-standing skepticism about whether people can judge their own cognition at all, and mood is a central complication. Previous research has found an association between subjective cognitive concerns and depression, and depression can itself make people feel their memory is worse. So Farias said the most encouraging result was that the link between perceived sharpness and actual performance did not appear to be explained by mood. "That's one of the aspects of the study that surprised me the most," she said.

Heather Snyder, senior vice president of medical and scientific relations at the Alzheimer's Association, was not involved in the research. She called it an interesting step toward learning whether people's perceptions could give researchers additional information about cognitive health. "Our perceptions are sometimes the first indicator," Snyder said. She noted the study joins a growing body of work testing whether everyday technology can pick up subtle cognitive change, including wearables that track changes in speech and in patterns of activity.

The appeal is that measurement can leave the clinic. A traditional assessment captures one point in time, often in a quiet, controlled room. A wearable could in theory gather many observations while someone cooks dinner, runs errands or goes about the day.

The limits are practical as well as scientific. Gavett said some older adults may not want to wear the devices or be continuously prompted, and users must tolerate interruptions and remember to charge them. He was also direct about what a week of data can and cannot do: it will not tell anyone whether they are developing dementia.

The open question is whether patterns gathered over months or years can predict later impairment. "Researchers may need to follow people for months or even years to determine whether changes in these measurements provide meaningful information about someone's risk of larger cognitive decline," Gavett said. Farias called that one of the biggest questions now facing the team. The researchers plan to keep testing such approaches in clinical trials before moving to larger populations and judging how reliably they detect meaningful change. "Moving in the right direction to get people help for dementia as early as possible is the goal," Gavett said.

Measuring change in a person's cognitive functioning is really hard.
— Brandon Gavett, study first author
Our perceptions are sometimes the first indicator.
— Heather Snyder, Alzheimer's Association
Contáctanos FAQ