New research published in JAMA Neurology reveals that specific sleep and activity patterns, captured by wearable devices, may help predict a person's risk of developing dementia. The study, led by Clémence Cavaillès and her team from Université Paris Cité, found that certain disruptions in the daily sleep-wake cycle are associated with a significantly increased risk of the condition. These findings, published online on June 5, 2026, suggest a new, scalable way to identify individuals at higher risk earlier.
Objective Monitoring Reveals Key Patterns
The comprehensive study used data from over 53,000 participants in the UK Biobank and nearly 4,000 from the Whitehall II study, all aged 60 or older and free of dementia at the start. Researchers equipped participants with wrist-worn accelerometers, small devices similar to fitness trackers, to collect detailed information on their sleep, activity levels, and daily rhythms. This objective monitoring gathered 36 different metrics related to their sleep-wake cycles.
Using advanced machine learning, the research team identified nine key metrics that grouped into two main patterns. The first pattern showed reduced and less varied daytime activity. This included lower amounts of moderate-to-vigorous physical activity, more time spent sitting, and more frequent shifts between being active and resting. People who scored higher on this pattern faced a 43 percent increased risk of developing dementia over an average follow-up of about eight years.[medicaldialogues+2]
The second identified pattern pointed to disrupted sleep. This involved irregular sleep durations, fragmented sleep with longer periods of waking up during the night, unstable times for falling asleep, and earlier wake times. This pattern also showed a statistically significant, though smaller, link to a higher risk of dementia. These associations were confirmed in a separate group of participants, strengthening the study's findings.[medicaldialogues+1]
Enhancing Early Detection
The study highlights the potential for these objective sleep and activity measurements to improve current methods for predicting dementia risk. When added to existing prediction models that consider factors like age, lifestyle, and clinical history, the sleep-wake cycle data modestly but significantly enhanced the accuracy of risk prediction. The researchers noted that this improvement was comparable to the predictive power of genetic risk markers, such as the APOE status.
This suggests that monitoring daily sleep and activity patterns could become a practical and widespread tool for early identification. "Objective sleep and activity pattern monitoring may serve as a scalable tool alongside established risk factors for the early identification of individuals at increased risk of dementia," the study authors explained. Such tools could help doctors pinpoint those who might benefit from closer monitoring or early interventions.[medicaldialogues+2]
Broader Connections to Brain Health
These findings add to a growing body of research linking sleep patterns to brain health and dementia risk. Other studies published in JAMA-affiliated journals have also explored various aspects of this connection. For instance, a March 2026 study in JAMA Network Open found that a person's "brain age," estimated from sleep brain waves using EEG, could predict dementia risk. For every 10-year increase in brain age compared to actual age, dementia risk rose by nearly 40 percent.
Another study in JAMA Neurology from November 2024 indicated that a yearly decrease in deep, slow-wave sleep was associated with a 27 percent higher risk of all-cause dementia in people aged 60 and older. This deep sleep is crucial for clearing metabolic waste from the brain, including proteins linked to Alzheimer's disease.[news-medical]
Furthermore, research published in Neurology in March 2025 showed that women in their 80s who experienced increasing daytime sleepiness over five years had twice the risk of developing dementia. Even basic sleep duration plays a role; a September 2020 JAMA study found that people who consistently slept less than four hours or more than 10 hours per night showed lower cognitive function and faster cognitive decline. The ideal duration appeared to be around seven hours. More recently, a July 2026 study in JAMA found that older adults routinely sleeping 10 or more hours per night had elevated levels of p-tau181, a blood protein considered an early warning sign of dementia.[mdlinx+3]
While the recent JAMA Neurology study provides valuable insights, the researchers acknowledge some limitations. The accelerometer data only covered one week, which might not reflect long-term sleep habits. Participants were also generally healthier than the wider population, meaning the results might not apply to everyone. The study also did not assess specific types of dementia or existing sleep disorders. Theauthors emphasize the need for further research to fully understand how these non-invasive measures can best complement existing methods for identifying high-risk individuals.[inc+1]
Understanding the complex relationship between sleep-wake cycles and dementia is crucial for developing new strategies to prevent or delay cognitive decline. The continued research into objective sleep monitoring offers a promising path forward for earlier detection and improved care.[medicaldialogues]
