A just-published Nature study, “Sleep chart of biological ageing clocks in middle and late life,” asked a simple but important question: does sleep duration line up with how quickly different parts of the body appear to be ageing?
The researchers used large biobank datasets, especially UK Biobank, and compared self-reported sleep duration with 23 biological ageing clocks built from three data layers:
- MRI imaging,
- plasma proteins, and
- metabolites.
In plain English, these clocks estimate whether a person’s brain, liver, immune system, pancreas, adipose tissue, and other systems look biologically “older” or “younger” than expected for their actual age.
The study focused on adults in middle and later life, with UK Biobank participants aged roughly 37–84 years.
The key concept is the biological age gap, or BAG. A higher BAG means a system looks older than expected for someone’s chronological age; a lower BAG means it looks closer to, or younger than, expected.
Sleep was measured by a questionnaire asking participants about average hours slept in a 24-hour day, including naps (so this is not a lab-based sleep study using polysomnography or wearable sleep tracking).
What the study found: sleep and ageing markers followed a U-shaped curve.
People at the very short end and the very long end of sleep duration tended to show higher biological ageing signals, while the lowest ageing signals clustered around a middle range. Across the nine biological ageing clocks that showed significant nonlinear associations, the lowest observed ageing burden fell around 6.5–7.8 hours for women and 6.4–7.7 hours for men. The authors then classified sleep into three broad groups: short sleep under 6 hours, normal sleep 6–8 hours, and long sleep over 8 hours.
The most interesting part is that this was not just a “brain sleep” story. Significant U-shaped patterns appeared across multiple systems and measurement types. Proteomic clocks showed associations in the brain, pulmonary, hepatic, immune, and skin systems. Metabolomic data showed a significant endocrine ageing pattern. MRI-based clocks showed significant patterns in the brain, adipose tissue, and pancreas. In other words, poor sleep duration was associated with ageing-like signals across the body, not only in cognition or mood.
The disease-risk data were also striking. The researchers reported 153 significant associations between abnormal sleep duration and disease endpoints after multiple-testing correction. Short sleep had the broader disease footprint, including associations with depression, anxiety, obesity, type 2 diabetes, hyperlipidaemia, hypertension, ischaemic heart disease, arrhythmias, COPD, asthma, reflux, gastritis, and functional intestinal disorders. Long sleep was also associated with several outcomes, but the authors suggest it may often act more like a marker of underlying disease, recovery burden, or neuropsychiatric vulnerability than a direct cause.
Mortality risk was another notable data point. Compared with 6–8 hours, short sleep was associated with a 50% higher all-cause mortality risk — hazard ratio 1.50, 95% CI 1.44–1.55 — while long sleep was associated with a 40% higher risk — hazard ratio 1.40, 95% CI 1.36–1.44. This does not prove that sleeping less or more directly caused death, but it does reinforce that persistent sleep duration outside the middle range is a meaningful health signal.
What this means for you: the practical takeaway is not “sleep exactly 7.3 hours.”
The better takeaway is: your habitual sleep duration is a whole-body health signal.
If you are consistently sleeping under 6 hours, this study adds more evidence that your body may be carrying higher cardiometabolic, inflammatory, neurological, and ageing-related strain. If you are consistently sleeping over 8–9 hours and still feel unrefreshed, that is also worth paying attention to, because long sleep may reflect underlying illness, depression, sleep fragmentation, sleep apnea, medication effects, or recovery debt.
A good target range might be consistent, high-quality sleep around 7–8 hours, with the exact number adjusted by how you feel and function. The study’s “sweet spot” varied by organ system and data type. For example, the brain proteomic clock looked lowest closer to about 7.7–7.8 hours, while the brain MRI clock had a lower point closer to about 6.4–6.5 hours.
The study also has important limits.
Sleep duration was self-reported, the design cannot fully illustrate causality, most analyses were in people of predominantly European ancestry, and reverse causality remains possible — especially for long sleep, where disease may drive longer time in bed rather than the other way around. The authors also note that single snapshots of proteins and metabolites can be affected by illness, medication, diet, and timing.
Action items
First, track your sleep for two weeks.
Write down bedtime, wake time, awakenings, caffeine timing, alcohol, exercise, naps, and how rested you feel. The pattern matters more than any single data point.
Second, aim for a stable wake time and a realistic sleep window. If you usually get under 6 hours, the highest-value move is not optimization; it is recovery of basic sleep opportunity. Move bedtime earlier by 15–30 minutes every few nights until you are reliably in the 6.5–8 hour zone.
Third, treat long sleep plus fatigue as a flag. Sleeping 9 hours after a hard week is normal. Sleeping 9–10 hours regularly while still feeling foggy, low, or unrested is different. That is when it is worth looking at depression, sleep apnea, thyroid issues, medication effects, alcohol, chronic stress, or other medical causes.
Fourth, protect sleep quality, not just quantity. Keep caffeine earlier in the day, get morning light, keep the room cool and dark, avoid heavy late meals, and reduce late-night screens or work stress. This study measured duration, but the authors explicitly note that sleep fragmentation and circadian misalignment were not directly assessed, meaning “8 hours in bed” is not automatically “8 hours of restorative sleep.”
Bottom line: this is a strong reminder that sleep is not a soft lifestyle variable.
It is tied to measurable ageing signals across the brain, immune system, liver, endocrine system, pancreas, adipose tissue, and mortality risk. Your best move is to make consistent, restorative sleep in the 7–8 hour range a default — and to investigate persistent short or long sleep rather than normalize it.
—Kat
P.S. If waking too early or sleeping short has become part of the week, I’m opening private early access to my science-backed sleep intelligence app built exclusively for men 40+ who care about brain health and longevity, coming to iOS and Android. Join the waitlist, or share this with someone you care about who could benefit.
References
– The MULTI Consortium., O’Toole, C.K., Song, Z. et al. Sleep chart of biological ageing clocks in middle and late life. Nature (2026). https://doi.org/10.1038/s41586-026-10524-5