How Many Hours of Sleep for Longevity? The Exact Number: 6.4 to 7.8

For decades, the official sleep advice was as simple as a proverb: Sleep 8 hours a night. It's a round number, easy to remember, and sounds good. It was plastered on every health poster, every app, and every medical recommendation. The problem is that this number was never the result of precise science, but a convenient approximation. It assumed, mistakenly, that more sleep is always better.

A new study reported by Medical News Today in May 2026 shatters this axiom with a much sharper number: The sweet spot for sleep and longevity lies in a narrow range of 6.4 to 7.8 hours per night. Those who sleep within this range show the lowest risk of premature mortality, cognitive decline, and heart disease. And what might be more surprising: those who sleep too much, not just those who sleep too little, are also at higher risk. The relationship between sleep and health is not an upward line, but a U-shaped curve.

What is the U-curve of sleep?

To understand why the question 'how many hours of sleep' is more complex than it seems, you need to understand the shape of the relationship between sleep duration and health outcomes:

• The left end of the U: Short sleep (less than 6 hours), associated with increased risk of mortality, obesity, diabetes, hypertension, and cognitive impairment.
• The bottom of the U: The optimal range (approximately 6.4 to 7.8 hours), where the risk of negative outcomes is lowest.
• The right end of the U: Long sleep (more than 9 hours), also associated with increased risk of mortality and disease, sometimes even more than short sleep.

In other words: Optimal sleep is not 'as much as you can,' but 'exactly as much as you need'. Deviation in either direction, too little or too much, moves you up the curve, into higher risk zones. This is why the number 8, which sits at the right edge of the optimal range, is not necessarily the right target for everyone.

The connection to sleep: two opposite mechanisms

What makes the U-curve fascinating is that its ends are created by completely different biological mechanisms. Short sleep is harmful actively, while long sleep often signals an existing problem.

Short sleep: when the brain doesn't clean itself

During deep sleep, the glymphatic system is activated, the brain's drainage system. During wakefulness, neurons produce metabolic waste, including beta-amyloid protein, linked to Alzheimer's. Only in deep sleep do glial cells contract, intercellular spaces open, and cerebrospinal fluid flushes the waste out.

When you sleep too little, this cleaning is incomplete. Waste accumulates night after night. Studies have shown that one night of sleep deprivation already raises beta-amyloid levels in the brain. Simultaneously, sleep deprivation raises stress hormones (cortisol), increases systemic inflammation (CRP, IL-6), disrupts sugar regulation, and raises blood pressure. All of these are known accelerators of aging.

Long sleep: mostly a symptom, not a cause

Here lies the most important and subtle point. The fact that long sleep is associated with higher mortality does not mean that long sleep causes harm. In most cases, it is a red flag for something else:

• Underlying disease: Depression, heart failure, unbalanced diabetes, or chronic inflammation cause fatigue that manifests as long sleep.
• Fragmented sleep: Someone suffering from sleep apnea spends 9 hours in bed but gets poor sleep quality, hence 'needs' more.
• Low quality: 9 hours of shallow sleep are not equal to 7 hours of deep, continuous sleep.

This is a classic example of the reverse causation problem: long sleep is not the cause of aging, it is the symptom of a process already underway. This point is critical for correctly understanding the data.

Current evidence

The range of 6.4 to 7.8 hours is not a guess; it stems from a combination of large population studies and extensive database analyses.

Study 1: U-curve analysis on hundreds of thousands of participants

Meta-analyses that aggregated dozens of prospective studies, with hundreds of thousands of participants followed for years, repeatedly found the same pattern: The risk of all-cause mortality is lowest around 7 hours of sleep. Sleeping 5 hours increased the risk by about 12%, and sleeping 9 hours or more increased it by 30% or more. Note that the long end is sometimes more dangerous than the short end, precisely due to the reverse causation mechanism.

Study 2: Sleep and cognitive function

Studies examining cognitive performance found a similar U-curve: Both short and long sleep were associated with lower memory and attention scores. In a large-scale study, people who slept about 7 hours showed the best cognitive performance, as well as the most normal brain volume on imaging. Deviation in either direction, even by an hour or two, was linked to increased decline over time.

Study 3: Sleep and heart health

The same shape repeats in the cardiovascular field. Short sleep increases the risk of heart attack and stroke through high blood pressure and inflammation, while long sleep is associated with increased risk, likely as a marker of poor underlying health status. The range of less than 8 hours, but not less than 6.4, showed the lowest cardiovascular risk profile.

What about quality versus quantity?

All these numbers refer to sleep duration. But there is a variable no less important, and sometimes more important: sleep quality. Two people can sleep 7 hours and get completely opposite results.

• Sleep architecture: The ratio of deep sleep (slow-wave), light sleep, and REM sleep. Only in deep sleep does the glymphatic system operate fully.
• Continuity: 7 hours of continuous sleep is far superior to 8 hours fragmented into 5 awakenings.
• Circadian timing: Sleep aligned with the internal biological clock (circadian), i.e., at night and not during the day, is much more restorative.

This explains part of the U-curve: a person who 'sleeps 9 hours' but wakes up dozens of times a night records a long duration but gets poor quality. The number on the clock hides what is really happening in the brain.

Should you stress about the exact number?

Here caution is required. Before you rush to set yourself to exactly 7.1 hours, it's important to understand the study limitations:

• Observational data: Most studies are observational, not controlled experiments. They show a correlation, not necessarily causation. You cannot 'randomize' people to sleep 5 or 9 hours for years and measure mortality.
• Reverse causation: As we saw, much of the risk of long sleep stems from underlying diseases, not from the sleep itself. Cleaning up these confounders is only partial.
• Individual variability: The range 6.4 to 7.8 is a population average. There are rare individuals who function excellently on 6 hours (DEC2 gene carriers), and others who truly need 8.5. Genetics plays a role.
• Self-reporting: Many studies rely on self-reported sleep duration, which is known to be inaccurate.

The balanced conclusion: Don't turn sleep into a source of anxiety. Stress about 'I must sleep exactly X hours' (a phenomenon dubbed orthosomnia) is itself harmful to sleep. The range is a compass, not an exact verdict.

What should you take from the study?

1. Aim for the range, not the number: Strive for about 7 hours, within a range of 6.4 to 7.8. If you feel refreshed and function well on 6.5 hours, that's fine. There's no need to force 8.
2. Prefer consistency over perfection: Sleeping and waking at the same times every day, including weekends, is more important than accumulating 'sleep debt' and trying to repay it. Consistency stabilizes the circadian clock.
3. If you regularly need 9 hours, check why: Persistent long sleep with daytime fatigue is a sign for a medical check-up: sleep apnea, hypothyroidism, depression, or anemia. Don't dismiss it as 'I just sleep a lot'.
4. Improve quality, not just quantity: A dark, cool room, no screens an hour before bed, no caffeine after noon, and no alcohol in the evening (which fragments sleep and harms REM).
5. Prioritize deep sleep: Regular physical activity and morning light exposure increase the percentage of deep sleep, where the essential glymphatic cleaning occurs.

The broader perspective

The story of the U-curve in sleep is an excellent example of a broader principle in the biology of aging: Almost nothing good is 'more is always better'. As with physical activity, calories, or even oxygen, sleep also has an optimum, not an infinity. Extremes in both directions come at a cost.

The shift from 'sleep 8 hours' to 'aim for 6.4 to 7.8 quality, consistent hours' is exactly the kind of maturation that longevity science undergoes: from crude, round advice to personalized precision. But this precision should not become an obsession. The best sleep is the one you don't have to think about, regular, deep, and within the range your body naturally asks for.

If you remember one thing: don't chase extra hours, and don't settle for too few. The sweet spot of sleep is narrow, but reaching it is one of the most powerful and cheapest investments in longevity.

References:
Medical News Today - Between 6.4 and 7.8 hours of sleep may aid healthy aging, longevity