More Sleep Isn't Better. A New Study Mapped the Sweet Spot Organ by Organ.

There's a quiet assumption baked into most sleep advice: more is better. Hit eight hours. Bank it on the weekend. If you're tired, sleep longer.

A study published in Nature in May 2026 makes that harder to keep [1]. Measured against how old the body looks on the inside, sleep has a ceiling, not just a floor. Sleep too little and the body shows more wear. Too much, and it shows more wear too. The lowest-wear point sits in a narrow band in the middle, and that band is not eight hours. It isn't even a single number. It moves depending on which organ you ask, and on whether you're a woman or a man.

What they actually measured

The study is called Sleep Chart. A team based largely at Columbia, led by Cliodhna O'Toole and Junhao Wen, pulled data from the UK Biobank: just under 500,000 people aged 37 to 84.

What makes it different is the ruler. Instead of asking whether sleep predicts death, the team analyzed 23 separate "biological aging clocks." Each one estimates how old a single organ looks on the inside, drawn from MRI scans, plasma proteins, and metabolites in the blood. Your brain has an apparent age. So does your liver, your heart, your metabolic system. The gap between that apparent age and your real age is the number they tracked.

Then they asked, for each organ: at what sleep duration is that gap smallest?

The answer was a U, not a line

Across nine of the clocks, and all three types of measurement, one shape kept turning up: a U.

Picture a chart with hours of sleep along the bottom and biological aging up the side. It dips to a low point in the middle and rises again at both ends. Under six hours, aging ran high. Over eight hours, it ran high again. The bottom of the dip, where the body looked youngest, fell between 6.4 and 7.8 hours, depending on the organ.

That needs context so it doesn't get misread. The familiar "seven to nine hours" advice comes from decades of research on a different question: who lives longest. A large 2010 meta-analysis by Francesco Cappuccio's group found higher mortality at both short and long sleep, with roughly seven to nine hours as the healthy middle [2], and that still stands. Sleep Chart measures biological aging instead of death, and against that ruler the low point sits closer to seven than to nine. The two agree on the floor. They part ways at the top, where the aging clocks start climbing past about eight hours, earlier than mortality risk does.

At the edges, the two measures agree again. Compared with people sleeping six to eight hours, both short and long sleepers had higher rates of systemic disease and earlier death. The reach differed by direction, though: short sleep spread broadly across the body, into metabolic and cardiovascular conditions, while long sleep clustered more around the brain and mood.

Too little and too much sat on the same curve, and in this study both extremes carried more biological-aging wear.

Different organs wanted different amounts. And women ran a little longer than men.

This is the part that matters most for the people reading this site. Not as perimenopause-specific dosing, but as sex-aware sleep science. So, the numbers.

The clearest sex difference was metabolic. The endocrine-metabolic clock looked youngest at about 6.7 hours for women and about 6.1 for men, the widest gap of any organ, and the strongest sex difference in the study. More than half an hour, for the same inner system.

The brain told a stranger story, because the brain has two clocks. Measured by the proteins it sheds into the blood, it looked youngest at about 7.8 hours, the longest optimum of any organ. Measured by its shape on an MRI scan, it bottomed out closer to 6.5. Same organ, two measures, two answers. The brain, true to form, declined to commit. The researchers suspect the brain's chemistry may simply need more sleep than its shape does.

The liver, lungs, immune system, skin, fat and pancreas all traced the same U, though not every one of the 23 clocks did. What changed from organ to organ was where the bottom of the dip fell.

In the clocks the study put exact numbers on, women's low point sat slightly later than men's. And across all nine significant clocks, the female optimal range ran a little higher: about 6.5 to 7.8 hours, versus 6.4 to 7.7 for men.

There is no single right number. There's a personal one, and sex is one of the things that nudges it.

Sex differences in sleep aren't new, just often averaged away in big studies. The exercise scientist Stacy Sims built two books, ROAR and Next Level [3], on a blunter version of the point: women are not small men. Much of what's known about exercise and the body was studied in men and assumed to carry over. A study this large finding that the lowest-aging amount of sleep differed by sex is worth noticing.

Here's the honest part

This study was not about perimenopause. That needs saying plainly, because it would be easy to read the sex difference and assume the rest.

The words menopause, perimenopause, and estrogen never appear in the paper. Hormones do come up, but only in passing, as a guess for why the sexes might differ, never something the study measured. The cohort was sorted by sex, not hormonal stage, and those are not the same thing: a 41-year-old in early perimenopause and a 71-year-old well past it were both just "female" here.

So the paper does show something real: the least-aging amount of sleep isn't eight hours, and isn't the same for women and men. What it can't say is what perimenopause specifically does to that number. That study hasn't been done. With perimenopause, that sentence comes up a lot.

It also can't promise that changing your sleep reverses aging. It's a snapshot, so it shows links, not cause. But here's the useful part. The pattern tracked more with how people slept than with their inherited risk. When the researchers looked at genetic risk of faster aging, the U-shape flattened out. That makes sleep something you can likely work on, not proof that changing your hours turns back a clock. The sleep was self-reported, and the group was almost all of European descent, so it won't fit everyone.

So what do you do with this?

The honest answer is not "sleep exactly 6.7 hours." A population average isn't a prescription, and the study never followed anyone over time to prove a personal target. Two things are worth taking from it.

Find your own number, and stop chasing eight. The research points to a band, roughly six and a half to eight hours against this aging-clock measure, not a magic figure. Your sweet spot is the one where you wake up restored and stay steady through the day, not the one a tracker congratulates you for. The floor matters more than the ceiling: consistently dipping under six to seven hours is where the harm reliably shows up. This is not a license to cut sleep, though. Broken nights often need more time in bed, not less, and if you only feel human on eight or nine hours, that's your number. Shorting yourself to chase a clock reading would be exactly the wrong lesson.

Then work on quality, because duration is only half the story. Sleep Chart measured how long people slept, not how broken that sleep was, and for a lot of women broken is the real problem. Other research keeps finding that when you sleep matters as much as how long. A 2024 UK Biobank analysis by Windred and colleagues found that sleep regularity, consistent bed and wake times, predicted living longer more strongly than duration did [4].

A study by Brindle and colleagues put a number on it: keeping your sleep midpoint within about an hour, night to night, is roughly where "regular" begins [5]. That's a more actionable target than any duration figure, and for most people a bigger lever.

So the takeaway isn't a new number to hit. It's a change in the question. Not "did I get my eight hours," but "is this the right amount for my body, and is it consistent." The first part is personal and takes some honest attention to find. The second you can start tonight.

References

[1] The MULTI Consortium (O'Toole CK, Wen J, et al.). "Sleep chart of biological ageing clocks in middle and late life." Nature, 2026. Read it (open access) · interactive Sleep Chart

[2] Cappuccio FP, D'Elia L, Strazzullo P, Miller MA. Sleep duration and all-cause mortality (meta-analysis of prospective studies). Sleep, 2010. PubMed

[3] Sims ST. ROAR (2016) and Next Level (2022). drstacysims.com

[4] Windred DP, et al. Sleep regularity as a stronger predictor of mortality than duration (UK Biobank, n=60,977). Sleep, 2024. PubMed

[5] Brindle RC, Yu L, Buysse DJ, Hall MH. Sleep-regularity cutoffs in midlife adults (MIDUS). Sleep, 2019. PubMed