Sleep Beats Diet and Exercise for Lifespan

You have probably put real effort into what you eat. Maybe you track protein, or you have cut back on the things you know are not helping. You might train several times a week.

And then you stay up until half past midnight, because the evening is the only part of the day that belongs to you.

That last decision may be undoing more than the first two are building. Researchers ranking behavioural predictors of life expectancy found insufficient sleep came second only to smoking, ahead of both diet and exercise. Not as an equal partner to them. Ahead of them.

Key Takeaways

• Insufficient sleep is the second strongest behavioural predictor of a shorter life expectancy, behind only smoking
• In that analysis it outweighed both diet and exercise as a predictor of lifespan
• Consistency matters nearly as much as duration, because irregular bedtimes disrupt circadian rhythm even when total hours look fine
• The relationship is a U-shape: too much sleep is associated with faster biological ageing, not just too little
• Small changes stack. Cutting 30 minutes of sitting a day was associated with a 9% lower mortality risk for people sitting eight or more hours

What the sleep research actually ranked

The finding comes from work published in SLEEP Advances by researchers at Oregon Health & Science University, which set out to compare behavioural predictors of life expectancy against each other rather than studying one in isolation.

That comparison is the useful part. Plenty of studies establish that sleep matters. Far fewer put it in a ranking next to the behaviours people actually spend their effort on.

Insufficient sleep landed second, behind smoking and ahead of diet and exercise. For anyone who has optimised their meals and their training while treating sleep as the flexible item, that ordering is worth sitting with.

The target most sleep researchers converge on is seven to nine hours. The framing that matters is treating it as fixed rather than as the leftover time after work, training and screens have taken their share.

Why sleep outranks things that feel more active

Diet and exercise feel like effort, and effort feels like it should count for more. Sleep feels like an absence of activity.

Biologically it is the opposite. Sleep is when the body clears metabolic waste from the brain, consolidates memory, regulates the hormones governing appetite and stress, and carries out tissue repair. Short sleep raises cortisol, disrupts glucose handling, and increases the hunger signalling that makes the next day's eating harder to control.

Which means poor sleep does not merely sit alongside diet and exercise. It actively degrades both. A badly slept body is more insulin resistant, hungrier for fast carbohydrate, and less able to recover from training. You cannot out-train a sleep deficit, because the training is what the sleep was supposed to repair.

Consistency matters nearly as much as duration

The second finding surprises people who think they have this covered.

Total hours are not the only variable. Irregular bedtimes disrupt circadian rhythm even when the weekly total looks adequate. Sleeping five hours on weeknights and eleven on Saturday does not average out to a healthy pattern, though the spreadsheet says seven.

Your circadian system runs on prediction. It begins preparing for sleep hours in advance, lowering body temperature and releasing melatonin on a schedule it learned from your behaviour. Move that schedule around by three hours twice a week and the preparation keeps arriving at the wrong time.

The practical implication is unromantic: a consistent seven hours beats a chaotic eight. If you change one thing, change the variability before you chase the total.

More is not better

The relationship is not a straight line where more sleep keeps improving the outcome.

Research published in August found that both too much and too little sleep were associated with the body ageing faster. The curve is a U, and both ends of it carry risk.

The usual caution applies about what causes what. Consistently long sleep is often a symptom of an underlying condition rather than a cause of harm, including depression, sleep apnoea and chronic inflammation. Someone sleeping ten hours and still exhausted has a problem to investigate, not a habit to be proud of.

Practically: if you need well over nine hours to function, that is worth raising with a doctor rather than treating as your natural requirement. This is the same interpretive care needed when reading any single biomarker, a point covered in what biological age tests actually measure.

A worked example of a realistic change

Nobody reorganises their life around a study. Here is what a modest shift looks like in actual hours.

Say you currently get to bed at 00:30 and the alarm goes at 06:30. That is six hours, five nights a week, with a long lie-in at the weekend.

• Move bedtime to 23:45. That is 45 minutes earlier, which is one episode of something or roughly one scroll session
• Weeknight sleep goes from 6 hours to 6.75 hours
• Across five weeknights that is 3.75 extra hours a week, or about 195 hours a year
• Hold the wake time steady at weekends and the variability problem shrinks at the same time

You have not bought a device, taken a supplement or joined anything. You moved one decision 45 minutes earlier. Since the evening screen habit is usually what holds that boundary in place, the tactics in our guide to reclaiming your evening from screens do most of the practical work here.

The sitting finding, while you are at it

One more number from the same body of research, because it is unusually cheap to act on. Among people sitting eight or more hours a day, reducing sitting time by 30 minutes was associated with a 9% reduction in mortality risk.

Thirty minutes. Not a gym membership or a programme, just less continuous sitting. Findings like this one, and the sleep ranking above it, consistently point the same direction: the boring structural habits outperform the purchasable ones, which is roughly the conclusion of our look at which longevity supplements actually work.

Frequently Asked Questions

How many hours of sleep do I actually need?

Seven to nine hours for most adults. Individual variation exists but is smaller than people assume, and those who believe they thrive on five hours usually perform worse on objective testing than they report feeling. Judge it by how you function without caffeine rather than by how you feel with it.

Can I catch up on sleep at the weekend?

Partially, and it is not equivalent. Weekend recovery restores some short-term deficit but does not reverse the circadian disruption caused by a shifting schedule. Sleeping five hours on weeknights and eleven on Saturday is worse than a steady seven every night, despite a similar weekly total.

Is sleeping too much bad for you?

Research published in August found both short and long sleep associated with faster biological ageing. Long sleep is frequently a symptom of another condition rather than a cause of harm in itself, so persistently needing well over nine hours is worth discussing with a doctor.

Does sleep really matter more than diet and exercise?

In the ranking of behavioural predictors from the SLEEP Advances analysis, insufficient sleep outweighed both. That does not make diet and exercise unimportant. It means that if you are choosing where to spend limited effort, and your sleep is currently poor, sleep is likely the higher-return change.

What is the single most effective change for better sleep?

A fixed wake time, including at weekends. It anchors the circadian rhythm more reliably than a fixed bedtime, because you control when you get up far more directly than when you fall asleep, and a steady wake time gradually pulls the natural sleep onset earlier on its own.

Protect the hours

The appeal of diet and training is that they are visible. You can log them, track them, and feel that you did something.

Sleep offers none of that. It is unglamorous, unmeasurable in the moment, and the first thing sacrificed when the day runs long. Ranked against every other behaviour, it came second only to smoking.

If you are going to change one thing, make it the boundary that closes your day. Not a supplement, not a tracker, not a new protocol. A consistent wake time and a bedtime that is treated as fixed rather than negotiable.