A large-scale study maps sleep duration against multi-organ aging clocks and reveals why the healthiest biological aging profiles are clustered around the well-known 6-8 hour sleep window.
Research: Sleep graphs of the biological aging clock in midlife and late life. Image credit: Gorodenkoff / Shutterstock
In a recent study published in the journal natureIn , researchers from the MULTI consortium describe the development and potential of the Sleep Chart, a comprehensive framework developed using large-scale population data to assess the correlation between self-reported sleep duration and 23 biological aging clocks.
Results of this study revealed a U-shaped relationship between sleep and biological age differences that extends across nine of 23 aging clocks spanning brain and body systems, with the lowest observed biological age differences occurring within the range of 6.4 to 7.8 hours of self-reported daily sleep time. Furthermore, the results show that both insufficient and excessive sleep are associated with increased risk of systemic disease and biological aging, as well as increased all-cause mortality, highlighting that sleep duration may be associated with long-term health outcomes.
Background of sleep duration and biological aging
Traditional indicators of aging, particularly chronological age and the number of candles on a birthday cake, often fail to capture the organ-specific granular aging that precedes clinical disease, collectively referred to as “biological age.” Previous neuroimaging studies have identified nonlinear associations between sleep and brain phenotypes, but it remained unclear whether these patterns generalize to peripheral body systems and molecular layers.
The relatively recent fusion of magnetic resonance imaging (MRI) and high-throughput next-generation (“next-generation”) plasma proteomics and metabolomics has enabled researchers to quantify the biological age of an individual’s organs relative to their chronological age.
Researchers now aim to identify sample-specific minima in biological age difference curves and develop personalized and potentially modifiable targets to extend human lifespan and reduce the burden of age-related systemic diseases.
UK Biobank Sleep Chart Study Design
The study was conducted by the MULTI consortium, used data from the UK Biobank (UKBB), and included more than 500,000 participants (aged 37-84). The primary exposure for this study was participants’ self-reported sleep duration, which was conducted based on a questionnaire (Field ID: 1160). To minimize the effect of outliers, the analysis was specifically limited to individuals with UKBB who reported 4–10 hours.
The MULTI consortium then developed 23 organ-specific biological age gaps (BAGs) using a nested cross-validation machine learning framework.
- An MRI-based clock (MRIBAG; n = 7) was used to quantify the structural integrity of the brain, heart, liver, pancreas, spleen, adipose tissue, and kidney.
- The Proteomic Clock (ProtBAG; n = 11) utilizes Olink-based plasma proteomics to track aging signatures in circulating proteins, providing organ-specific resolution.
- The Metabolomic Clock (MetBAG; n = 5) was used to analyze plasma metabolomic profiles obtained from the Nightingale Health dataset.
A generalized additive model (GAM) with cubic regression splines was then used to model the nonlinear association without making any prior assumptions about the shape of the curve. Nonlinearity was statistically quantified using the effective degrees of freedom (EDF) curve complexity metric.
Research results on U-shaped sleep and aging
GAM analysis in this study revealed a U-shaped relationship for 9 out of 23 clocks (p < 0.05)。これは、睡眠時間が多すぎる (> 8 hours) or too little (< 6 hours) are associated with higher biological age gaps in these BAGs. The "youngest" biological age was observed among participants who reported 6.4 to 7.8 hours of sleep.
Brain ProtBAG showed the strongest U-shaped association with sleep (edf = 3.61, P1 < 1 x 10-20). The sample-specific minimum values representing the “youngest” biological state were 7.82 hours for women and 7.70 hours for men. Endocrine MetBAG also showed a significant U-shaped relationship (edf = 1.04, P1 = 3.97 x 10-5), with an estimated minimum value of 6.67 hours in women and 6.06 hours in men.
At the same time, brain MRIBAG was observed to reach its minimum value at approximately 6.48 h in women and 6.42 h in men (edf = 1.94, P1 = 3.85 x 10-7).
Furthermore, short sleep (<6 hours) was genetically correlated with heart failure (gc = 0.31), depression (gc = 0.37), and type 2 diabetes (T2D), whereas long sleep showed a more intensive genetic correlation profile, mainly related to brain-related and psychiatric traits.
In contrast, informed by structural equation modeling (SEM) data, it has been hypothesized that long sleep duration acts as a “marker” of underlying subclinical disease, suggesting that excessive sleep may be a marker of underlying physiological compensation or subclinical disease processes, potentially including neurodegeneration.
Finally, mortality and morbidity assessments revealed that both extremes increased the risk of all-cause mortality by approximately 40–50% (hazard ratio (HR) for long sleep = 1.40, HR for short sleep = 1.50, P < 1 x 10-20).
Optimizing sleep and its impact on healthy aging
The results of this study and the resulting sleep charts demonstrate that sleep time is associated with systemic patterns of biological aging across different organ systems and omics technologies. The consistent U-shaped associations observed with both structural imaging and circulating molecular markers suggest that maintaining sleep within 6-8 hours may be associated with a healthier organ aging profile.
The study specifically revealed that while women generally require slightly more sleep than men to reach their biological age minimum in certain categories, such as the brain’s proteomic clock (7.82 hours for women and 7.70 hours for men), optimizing sleep is a potential goal for promoting systemic health management and healthy aging trajectories for both men and women.
Click here to download your PDF copy.
