05/17/2026
Really good reading again: I love the verbiage this guy uses to explain things. Most things in life aren’t just black/white or either/or.
Short sleep ages you. Long sleep usually means something is already aging you. Same curve. Different stories.
A new paper in Nature, led by Junhao Wen's lab at Columbia, mapped sleep duration against biological aging across 9 organ systems in half a million UK Biobank adults aged 37 to 84. They used 23 different aging clocks built from MRI scans, plasma proteins, and metabolites. The relationship is U-shaped. The slowest measured aging sat between 6.4 and 7.8 hours per night. Outside that window, in either direction, organs looked older than chronological age would predict.
The two arms of the U are not the same biology.
On the short-sleep side, the causal story is well established. Sleeping under 6 hours raises systemic inflammation, impairs glucose tolerance the next morning, suppresses NK cell activity, and associates with markers of poorer overnight brain waste clearance. Mendelian randomization analyses, including in this paper, support a direct causal effect of short sleep on aging biology. Short sleep drives the wear and tear.
On the long-sleep side, the picture flips. Consistently sleeping over 8 or 9 hours is a well-documented marker of underlying disease, not a damaging behavior in itself. It tracks with major depression, undiagnosed sleep apnea, hypothyroidism, chronic inflammation, and neurodegenerative disease. The authors of this paper note that Mendelian randomization could not strongly support reverse causality on the long arm, but they explicitly could not exclude it either. Decades of prior work in sleep medicine and psychiatry argue that for most long sleepers, the long sleep is the body compensating for something already wrong.
This matters because the practical advice for the two groups is opposite.
If you sleep under 6 hours, the levers are direct. Total sleep opportunity. Sleep timing consistency. Morning light exposure. Caffeine cutoff after lunch. Alcohol stopped three hours before bed at minimum. These are the highest-evidence behavioral interventions in sleep medicine. Sleep extension trials adding 45 to 90 minutes a night have shown improvements in metabolic and cardiovascular markers across small studies.
If you consistently sleep 9 or more hours and especially if you still wake unrefreshed, the right move is to investigate what your body is recovering from. The workup is straightforward. A home sleep study to rule out apnea. TSH, free T4, ferritin, CRP, vitamin D, vitamin B12. Depression and anxiety screening. A review of medications that increase sleep need, including antihistamines, gabapentinoids, mirtazapine, and beta blockers.
There is a third scenario worth naming because it gets lumped in with the long sleepers. Athletes in heavy training blocks, adolescents, people recovering from infection, and people in their first trimester of pregnancy genuinely need 9 to 10 hours and the curve does not apply to them in the same way. The paper looked at habitual sleep in adults aged 37 to 84, not acute recovery states.
The cleaner way to state the finding is this. There is a window in the middle where the body looks youngest on every clock the authors built. Both sides of that window correlate with faster organ aging. The reasons differ. Short sleep does the damage. Long sleep usually shows the damage is already underway.
Wen et al., Nature, 2026 Cappuccio et al., Sleep, 2010 Irwin, Nat Rev Immunol, 2019 Spiegel et al., Lancet, 1999 Tasali et al., JAMA Intern Med, 2022 Besedovsky et al., Physiol Rev, 2019
