Key Questions Answered
Q: What did this study reveal about sleep and health?
A: An analysis of objective sleep measurements from more than 88,000 adults linked poor sleep patterns—particularly irregular bedtimes and unstable circadian rhythms—to higher risk for a wide range of diseases across many body systems.
Q: Which sleep traits were most harmful?
A: Late bedtimes (after 12:30 a.m.) and low sleep regularity (unstable daily rhythms) were among the strongest risk factors, associated with substantially higher incidence of severe conditions including liver cirrhosis and gangrene.
Q: What about long sleep—was it harmful?
A: Using objective device data, researchers found that long reported sleep is often misclassified; many people who report long sleep actually have short sleep when measured. The study found little evidence that objectively measured long sleep itself increases disease risk in most cases.
Summary: In the largest study of objectively measured sleep traits to date, researchers analyzed accelerometer data from 88,461 UK Biobank participants and found that irregular sleep timing and disrupted circadian rhythms were strongly linked to elevated risk for 172 diseases. The results highlight sleep regularity and rhythm stability as critical, and sometimes overlooked, components of long-term health—beyond just how many hours someone sleeps.
Individuals with irregular bedtimes or low circadian stability showed markedly higher risks for a range of serious conditions. The study also clarifies earlier confusion about “long” sleep by showing that self-reported long sleep often reflects time spent in bed, not true sleep time, which can lead to misleading associations in previous research.
Key Facts:
- Broad impact: Poor sleep regularity was associated with 172 diseases across multiple organ systems.
- Elevated risks: Bedtimes after 00:30 were linked to a 2.57-fold higher risk of liver cirrhosis; low interdaily stability raised the risk of gangrene by 2.61-fold.
- Misclassification explained: Over one-fifth of people who reported long sleep actually slept less than six hours when measured objectively, suggesting self-report can inflate perceived risks.
Source: Health Data Science
Overview of the study
A large international team led by researchers at Peking University and Army Medical University used accelerometer (actigraphy) data collected from 88,461 adults in the UK Biobank to evaluate multiple dimensions of sleep. Participants wore devices for several days and were followed for an average of 6.8 years to assess incident disease outcomes. The analysis focused on objective measures including sleep duration and timing, rhythm metrics (relative amplitude and interdaily stability), and fragmentation measures (sleep efficiency and number of awakenings).
The researchers estimated associations between these objective sleep traits and diseases coded using the International Classification of Diseases, 10th Revision (ICD-10), applying Cox proportional hazards models and adjusting for relevant confounders. They compared these objective-sleep associations with findings from prior literature based on subjective sleep reports and validated newly identified links in independent U.S. data sets.
Major findings
Across nearly seven years of follow-up, the study identified 172 diseases associated with one or more objective sleep traits. Of these, 42 conditions showed at least a doubling of risk in the worst versus best sleep trait categories. Examples include markedly higher risks of age-related physical debility, gangrene, and liver fibrosis/cirrhosis linked to poor rhythm stability and late sleep onset.
Overall, 92 diseases had more than 20% of their risk burden attributable to poor sleep characteristics. Notable high-burden conditions included Parkinson’s disease, type 2 diabetes, and acute kidney failure. Importantly, nearly half of the identified associations were specific to sleep rhythm measures, demonstrating that rhythm disruption—rather than duration alone—accounts for many health risks.
Reanalysis showed that subjective self-reports can introduce misclassification: many self-identified “long sleepers” are actually short sleepers by objective measures, which can produce false-positive links reported in prior meta-analyses (for example, with ischemic heart disease and depression). Independent validation in U.S. datasets replicated several novel rhythm–disease associations, and mediation analyses implicated inflammatory markers (white blood cells, eosinophils, C-reactive protein) as potential biological pathways.
About this sleep and health research news
Author: Mai Wang
Source: Health Data Science
Contact: Mai Wang – Health Data Science
Image credit: Neuroscience News
Original research: Phenome-wide Analysis of Diseases in Relation to Objectively Measured Sleep Traits and Comparison with Subjective Sleep Traits in 88,461 Adults — Yimeng Wang et al. (Open access)
Abstract (concise)
Background: Most prior evidence linking sleep to disease relies on subjective reports. This study examines how objectively measured sleep traits relate to disease risk across physiological systems and compares these findings with associations reported for subjective sleep measures.
Methods: Accelerometer-derived sleep metrics in 88,461 UK Biobank participants were used to assess nocturnal duration and timing, rhythm (relative amplitude and interdaily stability), and fragmentation (sleep efficiency and awakenings). Associations with ICD-10–coded diseases were estimated using Cox models, with validation in U.S. survey data and targeted reanalyses to investigate discrepancies with subjective-sleep literature.
Results: Over a mean follow-up of 6.8 years, 172 diseases were associated with objective sleep traits; 42 showed at least doubled risk in adverse sleep categories. Ninety-two diseases had over 20% of their burden attributable to poor sleep. Many associations were specific to rhythm disruption and distinct from results based on self-reported sleep duration. Misclassification of self-reported long sleep explained some previously reported associations. Inflammatory markers partially mediated several newly identified links.
Conclusions: Objective sleep measures reveal a disease spectrum that overlaps with but differs from subjective-sleep findings, emphasizing the importance of sleep regularity and rhythm stability. Objective measurement can reduce biases from self-report and improve understanding of sleep-related disease risk. Addressing multiple dimensions of sleep—timing, rhythm, duration, and fragmentation—may be important for preventing chronic disease.