The consensus recommendation for adults is seven to nine hours, not eight. The range exists because sleep need varies between people, and the variable most people ignore, timing, predicts outcomes almost as well as duration does.
In 2015 the American Academy of Sleep Medicine and the Sleep Research Society convened fifteen sleep scientists and clinicians, put them through a formal consensus process across more than five thousand papers, and asked a single question: how much sleep does an adult need. Their answer was seven hours or more on a regular basis.
That is the evidence-based number. It has a floor and a soft ceiling, and the gap between them is two hours wide. Eight is roughly where the middle of the distribution sits, which is how it became a prescription.
A population average is useful for public health and misleading for an individual. Nobody would tell you to wear the average shoe size.
Time in bed is not
time asleep
Sleep runs in cycles. Mary Carskadon and William Dement's standard textbook overview describes four to six cycles a night, each passing through lighter stages into slow-wave sleep and then REM, averaging about ninety minutes. Early cycles carry most of the slow-wave sleep. Later cycles carry most of the REM. Five cycles at ninety minutes is seven and a half hours, and it is a sensible illustration rather than a rule.
The more useful number is sleep efficiency: total sleep divided by time in bed. Healthy adults usually sit above eighty five percent, and clinicians treat a figure below that as a marker of disturbed sleep. Eight hours in bed at seventy five percent efficiency is six hours of sleep broken across the night. Seven hours in bed at ninety two percent is six and a half, mostly consolidated.
The claim that fragmented eight hours produces worse recovery than seven uninterrupted is directionally right and has not been tested as that exact comparison. What is well established is that fragmentation degrades next-day function independently of total sleep. Michael Bonnet's experimental work in the 1980s showed that waking people briefly every few minutes produced impairment resembling sleep loss, even when total sleep time was preserved.
Which is the counterintuitive basis of the most effective treatment for chronic insomnia. Cognitive behavioural therapy for insomnia includes sleep restriction: you reduce time in bed to roughly the time you actually sleep, efficiency rises, sleep consolidates, and time in bed is then extended in small steps.
Why chasing the number
can keep you awake
Sleep onset needs arousal to fall. Watching the clock, calculating how many hours remain and worrying about tomorrow raise it. Allison Harvey's cognitive model of insomnia, developed at Oxford, puts that kind of sleep-related worry and monitoring at the centre of how insomnia is maintained, and it is a large part of why CBT-I targets beliefs about sleep alongside behaviour.
In 2017 Kelly Baron and colleagues at Northwestern described a clinical pattern they named orthosomnia: patients whose pursuit of perfect sleep-tracker data was itself worsening their sleep. It was a case series, not a trial, and the pattern is familiar to anyone who has looked at a poor sleep score at seven in the morning and felt worse for it.
Hours-chasing, worry at bedtime, raised arousal, delayed sleep onset. The arousal part is well supported. The specific claim of a cortisol spike at bedtime from this worry has not been cleanly measured, so we describe it as physiological arousal, which is what the evidence shows.
None of this means short sleep is fine. Francesco Cappuccio's 2010 meta-analysis in Sleep pooled sixteen prospective studies covering more than a million people and found higher mortality at both ends: under six hours and over nine. The point stands that the evidence points to a range, not a number.
“People ask me how to get to eight hours when the answer they need is how to stop lying awake for ninety minutes of the eight they already give it. Fix the efficiency and the wake time. The duration usually sorts itself out once the bed stops being the place where you worry.”
Yes, some people need less.
You are probably not one of them.
In 2009 Ying-Hui Fu's laboratory at UCSF published in Science on a mother and daughter who habitually slept around six hours and woke fully rested. Both carried a rare mutation in the DEC2 gene. Mice engineered with the same mutation slept less and recovered from deprivation more efficiently.
That is the real basis for short sleepers, and the key word is rare. The far larger group of people who sleep six hours and say they feel fine have mostly adapted to a deficit they no longer notice. Hans Van Dongen's 2003 restriction study showed exactly that: participants held at six hours a night for two weeks rated their own sleepiness as stable while their objective performance kept falling.
The DEC2 finding is He et al. 2009 in Science, not Mignot 2005 in Sleep. And the claim that slow-wave percentage, REM duration and HRV predict daytime performance far better than total hours is not supported. Consumer devices estimate sleep stages with modest accuracy, and stage-level predictions of performance are an active research question rather than a settled finding.
HRV deserves its own sentence. A rising or falling trend over weeks, measured the same way each morning, is a reasonable recovery indicator. A single night's value is noisy, varies widely between individuals, and is influenced by alcohol, illness, training load and the device itself.
Four things that matter
more than the number
A consistent wake time
Andrew Phillips and colleagues tracked sixty one Harvard undergraduates for thirty days. Irregular sleepers had melatonin onset delayed by around two and a half hours and lower academic performance than regular sleepers, at the same average sleep duration. The wake time is the anchor the circadian system sets itself against, and it is the one end of the night you can actually control.
Sleep efficiency above 85 percent
If you regularly lie awake for long stretches, the fix is less time in bed, not more. Get up if you are not asleep after roughly twenty minutes, do something dull in dim light, and return when sleepy. This is stimulus control, a core component of CBT-I, and it rebuilds the association between bed and sleep.
Judge by the morning, not the tracker
How alert you are in the first two hours, how much caffeine you need to function, and whether you could nap at three in the afternoon are better day-to-day indicators than a sleep score. If you track, watch HRV and sleep timing as weekly trends and ignore the nightly grade.
Take persistent insomnia to CBT-I
For chronic insomnia, CBT-I is the first-line treatment recommended by both the American College of Physicians and the European Sleep Research Society, ahead of medication. It is structured, typically six to eight sessions, and increasingly available digitally. If you have struggled for three months or more, this is the item that changes outcomes.
A 30-second decision rule
- Do you wake rested and stay alert until mid afternoon without leaning on caffeine?
- Does your wake time stay within thirty minutes across the whole week?
- Do you fall asleep within twenty minutes and spend little of the night lying awake?
Stop targeting a number. Start protecting a system: the same wake time, a bed you sleep in rather than worry in, and a morning that tells you how it went.
Send this to someone who woke up rested after 6.5 hours and immediately felt guilty about it.
- Watson, N.F. et al. (2015). Recommended amount of sleep for a healthy adult. Sleep, 38(6).
- He, Y., Jones, C.R., Fujiki, N. et al. (2009). The transcriptional repressor DEC2 regulates sleep length in mammals. Science, 325(5942).
- Van Dongen, H.P.A. et al. (2003). The cumulative cost of additional wakefulness. Sleep, 26(2).
- Baron, K.G. et al. (2017). Orthosomnia: are some patients taking the quantified self too far? Journal of Clinical Sleep Medicine, 13(2).
- Cappuccio, F.P. et al. (2010). Sleep duration and all-cause mortality. Sleep, 33(5).
