Large Study Links REM Sleep Quality to Lower Risk of 83 Diseases

A large study tracking nearly 96,000 people over nearly nine years has found that the quality of sleep matters as much as the quantity—and that certain sleep stages offer measurable protection against dozens of serious diseases.

Researchers at Capital Medical University in China analyzed data from 95,559 participants in the UK Biobank who wore wrist accelerometers continuously for a week. Using a deep-learning algorithm to decode the accelerometer data, the team identified distinct sleep stages and tracked how variations in those stages correlated with disease outcomes over a median follow-up period of 8.9 years. The study, published September 17 in PLOS Medicine, found that greater amounts of REM sleep—the stage most associated with dreaming and memory consolidation—was linked to lower risk of 83 different diseases. Deep sleep, the restorative stage when tissues repair and growth hormones peak, showed protective associations with seven conditions. Conversely, sleeping fewer than five hours per night was tied to elevated risk across 37 conditions.

The protective effects were substantial. Each additional 47.6 minutes of REM sleep per night was associated with a 26 percent lower risk of heart failure, a 46 percent lower risk of dementia, and a striking 80 percent lower risk of Parkinson's disease. Deep sleep showed similar but more modest associations, particularly for type 2 diabetes and major depressive disorder. The researchers also found that sleep irregularity—night-to-night variation in sleep patterns—and frequent wakefulness after falling asleep were each linked to higher risk of anxiety and substance use disorders.

Perhaps most notably, the study identified an optimal sleep window. For 69 of the 86 disease conditions examined, the lowest disease risk clustered tightly between six and eight hours of sleep per night. This finding aligns with recommendations from the National Sleep Foundation and other health bodies, though it also suggests that the relationship between sleep and disease is not simply linear—sleeping more is not automatically better. The authors emphasized that sleep architecture—the composition and distribution of different sleep stages—appears to be as important as total sleep time for long-term health.

The methodological approach represents a shift from traditional sleep research. Laboratory polysomnography, the clinical gold standard, captures brain activity directly but typically measures only one or two nights in a controlled setting. The accelerometer method, while relying on algorithmic estimation rather than direct brain measurement, allowed the researchers to assess real-world sleep patterns across a massive population over years. This trade-off between precision and scale revealed patterns that smaller, more controlled studies might miss.

The findings arrive as sleep deprivation has become increasingly common. A 2026 China Sleep Health White Paper reported an average sleep duration of 6.97 hours among surveyed respondents, with 23 percent sleeping less than six hours nightly. Only 26 percent reported maintaining a consistent sleep schedule. The authors noted that sleep quality and quantity is now a priority for China's National Health Commission, reflecting broader recognition that sleep is a public health issue.

As an observational study, the research cannot prove that better sleep causes lower disease risk—only that the two are associated. The authors were explicit about this limitation: the study identifies correlations, not causation. Factors that influence both sleep quality and disease risk, such as underlying health conditions or lifestyle factors, could explain some or all of the observed associations. The study design cannot rule out such confounding variables.

Still, the findings add to a growing body of evidence that sleep stages matter. Earlier research has linked reduced REM and deep sleep to brain shrinkage in regions vulnerable to Alzheimer's disease, and other work has identified a U-shaped relationship between sleep duration and biological aging, with optimal outcomes between 6.4 and 7.8 hours per night. The current study's large scale and long follow-up period lend weight to these patterns.

The authors concluded that maintaining six to eight hours of sleep may reduce risk across multiple diseases, offering a concrete target for health promotion and disease prevention. They noted that lifestyle interventions—stress reduction, consistent sleep schedules, good sleep hygiene—can improve sleep quality. What remains unclear is whether interventions that specifically increase REM or deep sleep, beyond simply extending total sleep time, would produce the protective effects the study documents. That question will likely drive the next phase of sleep research.