sleep researchOptimal sleep time: when to sleep and how long, by age
Optimal sleep time for adults is 7 to 9 hours, with bedtime between 10 and 11 PM offering the strongest cardiometabolic signal. Here is the evidence.
Adults who sleep 6 hours per night for two weeks score the same on cognitive tasks as adults kept awake for 24 hours straight. Yet in the same Van Dongen study from 2003, those sleep-restricted adults rated their own performance as unchanged. The brain quietly accepts impairment as the new normal, which is why "I feel fine on 6 hours" is one of the least reliable self-reports in medicine.
Optimal sleep time is one of the most studied and most misunderstood numbers in human performance. The answer is not 8 hours flat. It is a range, gated by age, individual variation, and circadian timing, with mortality curves that bend upward at both ends.
The 7 to 9 hour range, and why the U is so sharp
The National Sleep Foundation consensus panel, drawing on 312 studies, set the recommended range for adults aged 18 to 64 at 7 to 9 hours per night. The modal optimum across large cohort studies sits near 7.5 hours, with confidence intervals that tighten substantially when sleep is measured by actigraphy rather than self-report.
The mortality data is U-shaped. A 2016 meta-analysis of 67 cohort studies covering more than 3.5 million participants found minimum all-cause mortality at 7 hours, with risk rising 6 percent for every hour below and 13 percent for every hour above. Cardiovascular mortality follows the same shape. The penalty for chronic 5-hour sleep is real and measurable; the penalty for chronic 10-hour sleep is also real, though more likely a marker of underlying disease than a direct cause.
Adults over 65 show a slightly different curve. Sleep need declines modestly with age, and the optimum shifts toward 7 to 8 hours. Below 6, the same penalties apply.
Why bedtime, not just duration, matters
A 2021 study from the European Society of Cardiology, using accelerometer data from 88,000 UK Biobank participants, found that sleep onset between 10:00 and 10:59 PM was associated with the lowest incidence of cardiovascular disease over the 5.7-year follow-up. Going to bed between 11 and midnight raised risk by 12 percent. After midnight, by 25 percent. Before 10 PM, by 24 percent. The window is narrower than most people assume.
The mechanism appears circadian. Endogenous melatonin begins to rise around 9 PM in most adults. Sleep onset that aligns with this rise produces the cleanest transition into the first slow-wave cycle. Sleep onset that occurs hours later, after the melatonin curve has flattened, produces a shallower first cycle and compresses the deep sleep that should dominate the first third of the night.
This is also why shift workers, who sleep adequate total hours but at the wrong circadian phase, accumulate cardiovascular and metabolic risk equivalent to chronic short sleepers. Hours matter. Timing matters separately.
The body does not measure sleep in hours alone. It measures alignment between sleep and the circadian phase that the brain expects sleep to occupy.
What happens to the body below 6 hours
The cost of chronic short sleep is not subtle, even at one week. In Leproult and Van Cauter's 2011 study, young healthy men restricted to 5 hours per night for a single week dropped total testosterone by 10 to 15 percent, equivalent to 10 to 15 years of normal aging. Insulin sensitivity falls 20 to 30 percent over the same window. Inflammatory markers including CRP rise. Natural killer cell activity, the immune system's first line against transformed cells, drops 70 percent after a single night of 4-hour sleep.
The detailed mechanism on the hormonal axis is covered in the insights piece on sleep deprivation and testosterone. The short version: sleep is when the testicular Leydig cells receive the LH pulse that drives morning testosterone, and that pulse is gated by slow-wave sleep that does not occur in compressed schedules.
Cognitive deficits accumulate without subjective awareness. By day 10 of 6-hour sleep, attention lapses, reaction time, and working memory have all degraded to levels seen after 24 hours of total sleep deprivation. The participants in these studies consistently rate themselves as "a little tired but functioning."
Sleep regularity may matter more than total hours
A 2023 analysis of 60,000 UK Biobank participants found that sleep regularity, defined as the consistency of bed and wake times across the week, predicted all-cause mortality more strongly than total sleep duration. The most regular sleepers had a 20 to 48 percent lower mortality risk than the least regular, even controlling for total hours.
The practical implication is that 7 hours every night, with bedtime varying by less than 30 minutes, outperforms 8 hours on weekdays and 10 hours on weekends. Social jetlag, the chronic mismatch between weekday and weekend sleep timing, behaves like a low-grade jet lag the body never adjusts to. For most adults, the highest-leverage sleep intervention is not adding an hour. It is anchoring the existing hours to a fixed clock time.
How to find your own optimum
Population averages set the range. Individual variation sets the point inside it. A short sleep test, three weeks of going to bed at the same time and waking without an alarm, will reveal natural sleep need within 15 minutes. Most adults land between 7.25 and 8.5 hours. Genuine short sleepers, the 1 to 3 percent of the population with rare DEC2 or ADRB1 variants who function well on 6 hours, are vanishingly rare. Almost everyone who claims to be one is wrong.
For tracking, a wearable that measures sleep stages plus a daily log of sleep onset time and morning subjective alertness will surface the true optimum within a month. The sleep analyzer tool walks through the calculation. The broader framework is in the sleep optimization protocol.
Protocol: setting optimal sleep time
- Target 7.5 hours of actual sleep, not time in bed. Add 30 minutes of buffer for the time it takes to fall asleep, so 8 hours in bed.
- Anchor sleep onset between 10:00 and 11:00 PM, with bedtime varying by less than 30 minutes across the week including weekends.
- Wake without an alarm where possible. If an alarm is needed, set wake time first and back-calculate bedtime, not the reverse.
- Treat any night below 6 hours as a deficit, not a one-off. Two consecutive short nights produce measurable cognitive and hormonal decrements that persist for several days.
- Track sleep regularity, not just duration, in any sleep app or wearable. Aim for a regularity index above 85.
- Re-test the optimum every 6 months. Travel, training load, illness, and seasonal light exposure all shift the target by 15 to 45 minutes.