Sleep as a Longevity Practice: What to Monitor, What to Ignore, and Wh

Dr. Daniel ChongHuman Performance & Longevity Medicine

Written and medically reviewed by Dr. Daniel Chong. Last reviewed 2026-10-24. 19 min

By Dr. Daniel Chong · Human Performance and Longevity

Sleep sits awkwardly in longevity medicine, because it is simultaneously the intervention with the broadest effect and the one nobody can sell you.

It affects glucose regulation, appetite, blood pressure, immune function, mood, cognitive performance, hormonal rhythm and recovery from training. Nothing else in this field touches that many systems at once. And the core practices cost nothing, which makes them easy to describe and remarkably difficult to follow.

So this is an article about reading your own sleep honestly. What the data on your wrist genuinely tells you, what it does not, which number deserves more of your attention than the one everybody fixates on, and where a device stops being useful and a doctor starts.

What sleep is actually doing

Not an exhaustive account, but enough to see why fragmented sleep has such wide consequences.

Metabolic regulation. Short and disrupted sleep measurably reduces insulin sensitivity and disturbs the hormones governing appetite, which is why poor sleep and central weight gain travel together so reliably. This connects directly to the metabolic picture we have written about here.

Autonomic recovery. Overnight is when parasympathetic activity dominates, heart rate falls to its lowest point of the day and blood pressure dips. That dip is a meaningful physiological event rather than an incidental one.

Hormonal rhythm. Growth hormone release is concentrated in slow wave sleep early in the night. Cortisol follows a rhythm that normally peaks shortly after waking. Testosterone production in men is sleep-dependent, and chronic short sleep lowers it measurably.

Glymphatic clearance. Clearance of metabolic waste from the brain increases during sleep. The mechanism is still being worked out and the longevity claims attached to it have run ahead of the evidence, so I would hold this one more loosely than the others.

Immune function and consolidation of memory, both well established.

The practical summary: sleep is not downtime that could be trimmed. It is an active maintenance window, and the consequences of shortening it show up in systems that look unrelated.

The hierarchy almost nobody gets right

If you are going to change one thing, this section tells you which.

Regularity first

A prospective cohort study using accelerometer data from the UK Biobank, with around sixty thousand participants, compared a sleep regularity index against sleep duration as predictors of all-cause mortality. Regularity was the stronger predictor.

That finding deserves to be better known than it is. Going to bed and waking at consistent times appears to matter more than the total hours, within reason.

It is also the more achievable target. Asking someone with a demanding job to find another ninety minutes is often unrealistic. Asking them to keep their wake time within half an hour across the week, including weekends, is usually possible.

Wake time is the anchor. Your circadian system is set largely by the timing of light exposure after waking. Fixing wake time and letting bedtime follow works considerably better than the reverse.

Duration second

Most adults need somewhere in the region of seven to nine hours, with genuine individual variation. Habitual short sleepers who feel well and function well are less common than the number of people who believe they are one.

A useful self-test: on a holiday with no alarm, how long do you sleep by the fourth or fifth night? That is closer to your requirement than your weekday average.

Continuity third

Eight hours broken into four fragments is not eight hours. Frequent awakenings matter independently of total time, and they are one of the clearest pointers towards a sleep disorder rather than a habit problem.

Timing fourth

Chronotype is partly genetic. A genuine night owl forced into a six o'clock start accumulates a mismatch that behaviour changes only partly solve. Working with your chronotype where your job allows is better than fighting it.

And sleep stages last

Which brings us to the thing everybody asks about, and the thing their watch is worst at measuring.

What your wearable actually gets right and wrong

I am going to be specific here, because vague reassurance about wearables is not useful and the published validation data is clear enough to quote.

Reasonably reliable: duration and timing

Wrist devices are decent at estimating when you fell asleep, when you woke, and roughly how long you slept. A 2024 systematic review in JMIR mHealth and uHealth comparing three popular devices against polysomnography found mean differences in total sleep time of about one and a half minutes for one device, under six minutes for another, and around forty-seven minutes for a third. So accuracy varies by device, but the better ones are close enough for tracking trends.

Caveat worth knowing: these devices are much better at detecting sleep than at detecting wake. In that same review, sensitivity to sleep ran between about 91 and 98 per cent, while specificity, meaning the ability to correctly identify wakefulness, ran between roughly 30 and 61 per cent. Lying still and awake looks a great deal like sleeping to an accelerometer.

That matters for anyone with insomnia. If you lay awake for an hour without moving, your watch may well have scored it as sleep, and the summary you see in the morning will not match your experience. Your experience is the more reliable of the two.

Do not trust: the sleep stage breakdown

This is the number people screenshot and worry about, and it is the weakest output on the device.

In the same review, agreement with polysomnography on multistate sleep classification, measured by Cohen's kappa, came out at 0.20 for one device and 0.46 for another. For context, those values sit in the slight to moderate range. Sensitivity for detecting deep sleep ranged from about 45 to 75 per cent depending on the device, and for REM from about 34 to 86 per cent. Light sleep was overestimated by around half an hour in some cases.

My practical advice: ignore the percentages for deep and REM sleep entirely. If the number is low, it may reflect your sleep or it may reflect the algorithm, and you cannot tell which. I have seen patients genuinely distressed by a deep sleep figure that carries very little information.

Useful as a trend: resting heart rate

One of the most informative numbers on the device, and one of the least discussed.

What matters is not the absolute value but your own baseline and any sustained drift upwards. A rise of several beats sustained over a week or two commonly reflects alcohol, accumulated training load, insufficient recovery, an illness coming on, or a stretch of poor sleep.

It is also worth noting that the overnight low is a cleaner signal than a daytime resting value.

Useful but heavily over-interpreted: heart rate variability

HRV reflects beat-to-beat variation in heart rate, which is influenced by autonomic balance. Higher is generally associated with better recovery status, within an individual.

Three things to understand before you take it seriously.

It is highly individual. Absolute values are not comparable between people, and genuinely not comparable between devices, because measurement windows and algorithms differ.

Daily variation is large. A single night's reading is noisy. A seven-day rolling average is far more informative than this morning's figure.

And it responds to things other than sleep. Alcohol, late eating, hard training, illness and stress all move it.

Used as a weekly trend alongside your resting heart rate, it has real value. Used as a daily score to feel anxious about, it becomes one more source of poor sleep, which is a genuinely perverse outcome.

Treat as a nudge: readiness and recovery scores

Every manufacturer has one, under a different name, and each is a proprietary composite of resting heart rate, HRV, sleep duration, sometimes temperature and recent activity. The algorithms are not published and are not comparable across brands.

They are reasonable prompts. A low score on a morning when you already feel flat is a sensible reason to train lightly. A low score on a morning you feel fine is not a reason to cancel your day.

The score is a summary of inputs you can read yourself. When it disagrees with how you actually feel, trust how you feel.

Worth watching: overnight temperature and respiratory rate

Both are usually reported as deviation from your own baseline rather than absolute values, which is the correct way to use them. A rise in either often precedes a recognised illness by a day or so, and for women, overnight temperature tracks the menstrual cycle reasonably well.

Interpret carefully: oxygen saturation and breathing disturbance

Wrist-based oxygen saturation measurement is considerably less accurate than a finger probe and is affected by fit, skin tone and movement. Treat a single reading as close to meaningless.

Patterns are different. A device reporting frequent overnight oxygen dips or a high breathing disturbance figure is producing a signal worth discussing with a doctor, and that brings us to the most important section in this article.

The one thing a wearable cannot settle

Obstructive sleep apnoea is substantially underdiagnosed, including here, and it is the single most common reason someone sleeps eight hours and still wakes unrefreshed.

Some devices now offer sleep apnoea notifications, and one such feature received United States regulatory clearance in 2024. This is a genuine step forward and it is worth being precise about what it is: a notification that your breathing pattern during sleep shows disturbance consistent with apnoea, prompting you to seek assessment. It is a screening prompt, not a diagnosis. Diagnosis requires a sleep study, either at home or in a laboratory, with a measured apnoea-hypopnoea index.

Equally important, and less often said: the absence of a notification does not rule it out. A device that has not flagged anything is not a clearance.

Signs that need a doctor rather than a programme

Loud snoring, particularly with witnessed pauses in breathing or gasping. Waking unrefreshed despite adequate time in bed. Significant daytime sleepiness, especially falling asleep when inactive. Waking repeatedly to urinate. Morning headaches. A larger neck circumference. Poorly controlled blood pressure.

Your partner usually knows more about this than you do. Ask them.

And why I am emphatic about it: untreated sleep apnoea raises cardiovascular risk, worsens insulin resistance, lowers testosterone in men, and is a cause of exactly the fatigue and low mood that people attribute to stress or ageing. It also will not respond to a better pillow, a darker room or any amount of sleep hygiene, which is the honest limit of everything in the next section.

If any of the above describes you, the right first step is an assessment. We cover what belongs in one here.

Sleep hygiene: what it is, and what it is not

Now the practices. But first a correction that matters, because this is widely misunderstood and the misunderstanding leads people to spend money in the wrong place.

Sleep hygiene education on its own is not an adequate treatment for chronic insomnia. The American Academy of Sleep Medicine's behavioural treatment guideline gives cognitive behavioural therapy for insomnia, CBT-I, its only strong recommendation, and advises against using sleep hygiene as a standalone therapy. One of the guideline authors has described sleep hygiene as almost never enough to resolve chronic insomnia disorder, and as general guidance useful for everyone rather than a treatment for the condition.

So the honest framing is this. If you sleep reasonably but could sleep better, the practices below are where the benefit is. If you have genuine chronic insomnia, meaning persistent difficulty falling or staying asleep with daytime consequences over months, you need CBT-I, which is a structured course rather than a list of tips, and the practices below are supportive rather than sufficient.

Knowing which of those two you are is most of the value of a consultation.

The practices that carry the most weight

Morning light, promptly. The strongest signal available for setting your circadian clock. Fifteen to thirty minutes of outdoor light within an hour or so of waking does more than any supplement. In Penang this is pleasant at seven in the morning and punishing by ten, so early is both better physiologically and more comfortable.

A fixed wake time, including at weekends. The single highest-value habit, for the regularity reason above. Within half an hour, seven days a week, is a reasonable target.

Light management in the evening. Dimming the environment for the last hour or two matters more than blue-light filters on screens, where the evidence is weaker than the marketing. Reduce overall brightness rather than just shifting the colour.

Temperature. Core body temperature needs to fall for sleep to initiate. A cooler room helps, which in this climate means air conditioning is a legitimate sleep intervention rather than a luxury. Somewhere around 18 to 20 degrees suits most people, which is cooler than most Malaysian bedrooms are set.

A warm shower an hour or so before bed works through the same mechanism, by drawing blood to the skin and accelerating subsequent heat loss.

Caffeine timing. Caffeine has a half-life of roughly five to six hours, with considerable individual variation, so a four o'clock coffee leaves a meaningful amount in your system at bedtime. A cut-off eight to ten hours before bed is a reasonable default. This is worth mentioning specifically in Malaysia, where afternoon kopi is close to a social institution.

Alcohol. It helps you fall asleep and reliably degrades the second half of the night, suppressing REM and fragmenting sleep. This is the change people least want to hear and most reliably notice. If you wear a device, alcohol is the intervention whose effect on your resting heart rate and HRV is most visible the next morning, which is occasionally more persuasive than anything I can say.

The last heavy meal, earlier. Late large meals raise core temperature and worsen reflux, and in Malaysia supper is a genuine cultural obstacle rather than an oversight.

Bed for sleep. The behavioural core of CBT-I, and the one most people get wrong. If you are awake and frustrated for more than about twenty minutes, get up, do something dull in low light, and return when sleepy. Lying in bed awake trains your brain to associate the bed with wakefulness.

A wind-down that is actually a wind-down. Thirty to sixty minutes of low-stimulation activity. The specific activity matters less than its consistency, because the routine itself becomes the signal.

Movement during the day, which improves sleep quality reliably. Vigorous exercise very late at night is unhelpful for some people and fine for others, so observe your own pattern rather than following a rule.

And naps, which are fine when short, before mid-afternoon, and not if you have insomnia, where they reduce the sleep drive you need that night.

The Malaysian evening, and what to move

Most sleep advice is written for a culture that eats at seven and goes quiet by nine. Ours does not, and pretending otherwise produces advice nobody follows.

So rather than a list of prohibitions, here is what genuinely interferes, roughly in order of how much it matters, with the realistic adjustment rather than the ideal one.

Supper

The largest obstacle, and the most cultural. Mamak at eleven, supper after a late shift, a second dinner because the first was at six and you were hungry again.

A substantial late meal raises core body temperature at the point it needs to fall, worsens reflux when you lie down, and shifts the metabolic day later. Heavy, oily and spicy food makes all three worse, which unfortunately describes a good deal of what is open at that hour.

The realistic adjustment: aim for the last substantial meal around three hours before bed. If supper is a social fixture, make it the lighter option and the earlier sitting rather than giving it up. A bowl of soup noodles at ten is a different proposition from nasi kandar at midnight.

Afternoon and evening kopi or teh

Caffeine has a half-life of roughly five to six hours with considerable individual variation, so a four o'clock cup leaves a meaningful amount circulating at bedtime. Many people who insist caffeine does not affect them sleep, which is true, and sleep less deeply, which they cannot feel.

Worth knowing that this is not only about coffee. Teh tarik, teh o, Chinese tea with dinner, and cheng teng or bottled drinks containing tea all contribute.

The realistic adjustment: a cut-off eight to ten hours before bed. For most people that means the afternoon cup moves to late morning. The social part of kopi is the sitting down, not the caffeine, so kopi o kosong decaf or a teh with less caffeine keeps the ritual.

Alcohol, where it applies

It shortens the time to fall asleep and reliably degrades the second half of the night, fragmenting sleep and suppressing REM. People wake at three, attribute it to stress, and do not connect it to the two drinks at dinner.

The realistic adjustment: earlier rather than later, and fewer. If you wear a device, this is the one intervention whose effect on your overnight heart rate is most visible the next morning, which tends to be more persuasive than advice.

The bedroom temperature most of us accept

Air conditioning set at 24 or 25 degrees because that is the default, or a fan alone on a humid night. Core temperature needs to fall for sleep to begin, and a warm humid bedroom works directly against that.

The realistic adjustment: try 18 to 20 degrees, which is cooler than most Malaysian bedrooms and is the single most underrated change on this page. If sharing a room makes that difficult, a lighter blanket on one side is easier to negotiate than the thermostat.

Evening traffic and the late commute

A late drive home in heavy traffic is a sympathetic stimulus arriving at precisely the wrong hour, and for many people in Penang it is non-negotiable.

The realistic adjustment: treat the twenty minutes after you get home as a decompression period rather than the start of the evening's tasks. Changing clothes, a shower, lower lighting.

Screens and the second shift

Phones in bed, work messages at ten, and the particular Malaysian habit of the family group chat becoming active late in the evening.

The brightness matters less than the engagement. A stimulating conversation at eleven is a problem whatever colour the screen is.

The realistic adjustment: a charging point outside the bedroom, and permission to answer in the morning.

Napping

Common here, and genuinely useful when short and before mid-afternoon. A long nap at five in the evening reduces the sleep drive you need at eleven, and in anyone with insomnia naps are generally counterproductive.

Evening exercise

Gym at nine after work is reality for a great many people, and for most it is better than not training at all. Some find vigorous late exercise delays sleep onset and others do not, so observe your own pattern instead of following a rule.

If it does affect you: keep the last hour before bed free of it, finish with lower-intensity work, and let the post-training shower do the passive heating job described below.

And the heat itself

A genuine local factor rather than a habit. Falling asleep is harder in a warm humid room, and nights are warmer than they were. This is why the environment section further down matters more here than it would in a temperate country.

Biohacking the bedroom, sorted by evidence

Sleep has become the most heavily marketed area in the optimisation world, and the practices on offer range from genuinely useful to inert to actively unwise. Here they are sorted honestly.

Worth doing, and mostly free

Morning outdoor light. The strongest circadian intervention available, and nothing sold in a box outperforms it.

A cool room. Discussed above. In this climate, the air conditioner is doing real physiological work.

A warm shower or bath an hour or two before bed. Passive body heating draws blood to the skin and accelerates the subsequent fall in core temperature, which is the signal sleep onset depends on. Reasonable evidence, and pleasant.

Fixed wake time, caffeine cut-off, less alcohol. The unglamorous three that outperform everything below them.

Slow breathing before bed. Extended-exhale patterns shift autonomic balance towards parasympathetic activity. The mechanism is sound, the effect is modest, and the risk is zero.

Reasonable, with modest evidence

Sauna. Finnish cohort data has associated regular sauna use with favourable cardiovascular outcomes, and many people sleep well afterwards. The cardiovascular association is observational, and the sleep-specific evidence is softer than the enthusiasm suggests. Worth doing if you enjoy it, and timed earlier in the evening rather than immediately before bed.

Weighted blankets. Modest supporting evidence, mainly where anxiety contributes to poor sleep. Harmless to try.

Magnesium. Most useful where intake is genuinely low, which is not uncommon. The effect in people with adequate intake is small, and the various salts differ mainly in tolerability rather than in efficacy.

Nasal breathing and nasal strips. Helpful for mild nasal obstruction, and they do nothing for obstructive sleep apnoea.

Weak, unproven, or marketed well beyond the data

Red light and photobiomodulation devices for sleep. The sleep-specific evidence is thin.

Blue-light blocking glasses. Weaker than the marketing implies. Reducing overall evening brightness does more than filtering one wavelength.

Grounding or earthing mats. Small, preliminary literature, and not something I would spend money on.

Most sleep supplement stacks, including glycine, L-theanine and apigenin. Plausible mechanisms, limited human sleep outcome data. Not harmful, not transformative.

CBD for sleep. The evidence is unsettled, and the regulatory position in Malaysia is a separate matter entirely.

Sleep tracking as optimisation. Which brings us to the one genuine harm in this list.

Where I would urge caution

Orthosomnia. There is a recognised pattern of people developing sleep anxiety from their own tracking data, worrying about a readiness score until the worry itself degrades their sleep. If your device is making you anxious, the correct intervention is to stop wearing it for a fortnight. I have suggested this more than once and it has worked every time.

Mouth taping. Currently fashionable and without meaningful supporting evidence. More importantly, in anyone with undiagnosed obstructive sleep apnoea or significant nasal obstruction, deliberately occluding the mouth is not a neutral experiment. I would not do it, and certainly not before apnoea has been excluded.

Melatonin. Worth being clear about this for Malaysian readers, because it is widely misunderstood. Melatonin is a prescription medicine in Malaysia, not a supplement you can buy over the counter here, and most pharmacies do not stock it. People frequently bring it back from abroad or order it online without realising its status.

Beyond the regulatory point, it is also commonly used for the wrong purpose. Melatonin is primarily a chronobiotic, meaning it shifts circadian timing. It has a legitimate role in jet lag, shift work and delayed sleep-wake phase disorder. It is a weaker tool for staying asleep through the night, which is what most people take it hoping for. If you think you need it, that is a conversation with a doctor rather than a purchase.

Alcohol as a sleep aid, which is the most widely used sleep intervention in the world and one of the least effective.

And long-term sleeping tablets without an assessment, where the underlying cause has never been established.

Recovery, and the one fashionable practice that works against you

Sleep is the primary recovery intervention. Everything else in this section is marginal by comparison, which is worth saying before listing the marginal things.

What the evidence supports

Sleep itself, for the reasons set out above. Growth hormone release concentrated in early slow wave sleep, autonomic recovery, and consolidation of motor learning.

Protein intake and resistance training, which is where the actual adaptation happens.

Active recovery, meaning light movement rather than complete rest, which has reasonable support for perceived recovery.

Heat exposure, with the cardiovascular association noted above.

HRV-guided training, meaning adjusting your training load according to your weekly trend rather than a fixed plan. The evidence is modest but it points the right way, and it is a more intelligent use of the data on your wrist than scoring your mornings.

The cold plunge problem

Cold water immersion is the most fashionable recovery practice in circulation, and there is a genuine conflict between what it does for how you feel and what it does for what you are training for.

A seven-week resistance training study found that cold water immersion after each session reduced post-exercise mTORC1 signalling, raised a marker of protein degradation, and produced smaller increases in type II muscle fibre cross-sectional area, while strength gains were similar between groups. The authors concluded that post-exercise cold water immersion should be avoided where muscle hypertrophy is the goal.

That matters a great deal in this context. Everything we write about ageing well rests on protecting and building lean mass, because muscle is where glucose is disposed of and muscle mass tracks with function in later decades. If you are training to preserve muscle and plunging immediately afterwards, the two are pulling against each other.

The practical resolution is timing rather than abstinence. Cold immersion has a legitimate role where rapid recovery matters more than adaptation, between sessions in a competition, for example. For everyone else, separating it from resistance training by several hours, or using it on non-training days, keeps whatever benefit you value without blunting the adaptation you are actually there for.

The same logic applies loosely to high-dose antioxidant supplementation around training, where suppressing the oxidative signal can suppress part of the response to it.

Hyperbaric oxygen, stated honestly

We operate a hyperbaric facility, so I should be straightforward here.

Hyperbaric oxygen therapy has well-established evidence for a defined set of approved medical indications. Its evidence for routine athletic recovery, for general wellbeing, and for longevity specifically is considerably more limited, consisting largely of small studies and mechanistic reasoning rather than outcome trials.

I would describe it as a legitimate therapy with specific indications, and as an area where the longevity claims circulating online run well ahead of what has been demonstrated. Anyone offering it to you as a general anti-ageing intervention is overstating the evidence, including if that is us.

A necessary word on anything prescribed

Nothing in this article is a recommendation to take any medicine, and I have deliberately named no doses anywhere in it.

That applies to everything with a sedative or circadian effect: melatonin, which as noted is a prescription medicine in Malaysia rather than a supplement; the benzodiazepines and the so-called Z-drugs; sedating antihistamines, which are freely available and widely used for sleep despite next-day impairment and a poor fit for long-term use; and anything prescribed for an associated condition such as mood or pain that happens to affect sleep.

These are prescription decisions and they belong with a registered medical practitioner who has assessed you. Not with an article, not with a pharmacist handing over a box, and not with a friend who found something that worked for them.

Four reasons this matters more in sleep than in most areas.

A sedative treats the symptom and conceals the cause. Sleeping tablets will produce sleep in someone with untreated obstructive sleep apnoea, which does nothing about the apnoea and in some cases makes the respiratory picture worse. The window in which the actual diagnosis would have been made closes quietly.

Tolerance and dependence are real with several of these agents, and withdrawal frequently produces rebound insomnia worse than the original complaint, which drives people straight back to the medicine. Unwinding that takes months and is considerably harder than not starting.

Interactions and individual risk matter. Age, liver and kidney function, other medicines, alcohol intake, a history of falls, driving and shift work all change what is appropriate and what dose is appropriate. None of that can be judged from a general article.

And chronic insomnia has a better first-line option anyway. As set out above, CBT-I carries the strong recommendation in the behavioural treatment guideline, not medication. Medicine has a legitimate place, often short term and often alongside, and it is not where a well-run plan starts.

Please also tell your doctor what you are already taking, including anything bought abroad, ordered online, obtained without a prescription, or taken as a supplement. A great many sleep consultations change direction once that list is on the table, and nobody is going to judge you for it.

The same principle applies across our writing. Prescription treatment should be assessed, prescribed and monitored by a registered practitioner who knows your history, and an article is a reason to start a conversation rather than a substitute for one.

Why the environment is worth taking seriously

Most sleep advice assumes you can simply implement it. In practice people fail at sleep hygiene not because they disagree with it but because their environment works against them, and they have never experienced the alternative to know what they are missing.

A bedroom with street light through thin curtains, a room set at 25 degrees because that is what the air conditioner defaults to, a phone on the bedside table, traffic noise, and a mattress chosen eleven years ago. Each of those is minor. Together they set a ceiling on the sleep available to you, and because it has been that way for years, it feels like your normal.

This is the thinking behind HAYAT Sleep Sanctuary, a controlled sleep environment programme we are preparing. Details of what it includes will follow when it opens.

I want to be straightforward about what a controlled environment can and cannot achieve in general, because the distinction matters more than any marketing would.

What a controlled environment can do. Show you what your sleep looks like when the obstacles are removed, which is genuinely useful information and surprisingly hard to obtain otherwise. Separate the environmental component of your poor sleep from the rest. Teach practices in a setting where they are easy rather than effortful, which is a better place to learn a habit. And give you a baseline worth taking home.

What it cannot do. Treat sleep apnoea, which needs diagnosis and medical management. Resolve chronic insomnia, which needs CBT-I. Or produce a lasting change on its own, because a good night somewhere pleasant is a data point rather than a repaired sleep pattern. What matters is what you change at home afterwards.

Anyone presenting a comfortable night's sleep as a treatment is selling the wrong thing. What is worth having is the measurement, the diagnosis if there is one to make, and the practices that survive the journey home.

What improving your sleep actually gets you

Stated carefully, because this is an area where claims run freely.

Well supported. Improved insulin sensitivity and glucose regulation. Better appetite regulation and easier weight management. Lower blood pressure, particularly where sleep has been short or fragmented. Improved mood and cognitive performance. Better recovery from training and better preservation of lean mass. In men, improved testosterone where short sleep has been suppressing it.

Reasonably supported. Reduced systemic inflammation. Better immune response. Reduced cardiovascular risk over time, from observational data rather than from randomised trials of sleep extension.

Plausible but not established. Specific claims about dementia risk reduction through improved overnight brain clearance. The biology is interesting and the human outcome evidence is not there yet.

And one that is often overclaimed. Nobody can tell you that fixing your sleep will add a defined number of years to your life. Sleep duration and regularity are associated with mortality in observational cohorts, which is meaningful and is not the same as a demonstrated causal effect of an intervention.

What I can say is that of everything in longevity medicine, sleep has the broadest reach, the strongest mechanistic basis, and the lowest cost. That is a reasonable case without any need to overstate it.

The short version

Regularity beats duration, and wake time is the anchor. Keep it within half an hour, seven days a week, and get outside soon afterwards.

Your wearable is reasonably good at how long and when, poor at telling sleep from quiet wakefulness, and genuinely unreliable on sleep stages. Watch your resting heart rate trend and your weekly HRV trend. Ignore the deep sleep percentage.

Sleep hygiene is useful for everybody and is not a treatment for chronic insomnia, which needs CBT-I. And no amount of it will touch sleep apnoea, which needs diagnosing.

Locally, the two changes that do the most are moving supper earlier and lighter, and setting the bedroom cooler than you currently do. The afternoon kopi comes third.

Of the optimisation practices, morning light, a cool room and a warm shower beforehand do more than anything sold in a box. And if you lift weights to protect your muscle, the cold plunge belongs several hours away from the session rather than straight after it.

Anything prescribed, melatonin included, belongs with a registered doctor who has assessed you, not with an article or a box on a shelf.

A better environment shows you what you have been missing. What you do at home afterwards is the part that counts.

Which problem do you actually have

If you are sleeping badly, the useful first step is working out which problem you have: an environment and habit problem, a circadian timing problem, insomnia, or a breathing problem that needs investigating. Those are four different answers and only one of them responds to better practices alone.

References

Frequently asked questions

A large UK Biobank analysis using accelerometer data found sleep regularity to be a stronger predictor of all-cause mortality than duration. Both matter. If you can only fix one, consistency of timing is the better target and usually the easier one.

No. Validation studies show poor agreement with polysomnography on sleep staging, with multistate agreement in the slight to moderate range. Use the device for duration, timing and trends in resting heart rate, and ignore the stage breakdown.

There is no useful general answer, because absolute values are not comparable between individuals or between devices. What is informative is your own seven-day trend, interpreted alongside resting heart rate.

Two common reasons. Devices identify sleep well but wakefulness poorly, so quiet wakeful time may have been scored as sleep. Or your sleep was fragmented in a way the summary does not capture, which can point to a sleep disorder. If it persists, it warrants assessment.

Some devices now provide sleep apnoea notifications, and one such feature received United States regulatory clearance in 2024. It is a prompt to seek assessment rather than a diagnosis, and no notification does not mean you are clear. Diagnosis requires a sleep study.

If you have chronic insomnia, generally no. The American Academy of Sleep Medicine guideline advises against sleep hygiene as a standalone therapy and gives CBT-I its only strong recommendation. Good sleep practices help everyone; they are not a treatment for the disorder.

Cooler than most Malaysian bedrooms. Around 18 to 20 degrees suits most people, because core temperature needs to fall for sleep to begin.

It shortens the time to fall asleep and degrades the second half of the night, fragmenting sleep and suppressing REM. If you track your data, it is usually the single most visible influence on your overnight heart rate.

In order of how much it matters here: a heavy late supper within three hours of bed, afternoon or evening kopi and teh, alcohol, a bedroom warmer than about 20 degrees, and engaging screen use in bed. Of those, supper timing and bedroom temperature are the two most Malaysian and the two most often overlooked.

Make it earlier and lighter rather than abandoning it. Soup noodles at ten affects you less than nasi kandar at midnight, and keeping the social part while changing the timing and the dish is more sustainable than a rule you will break.

That is a conversation with a doctor rather than a purchase. Sedatives can mask an underlying problem such as sleep apnoea, several carry tolerance and dependence risk with rebound insomnia on stopping, and suitability depends on your age, other medicines and medical history. Any prescription medicine for sleep should be assessed, prescribed and monitored by a registered medical practitioner.

Melatonin is a prescription medicine here rather than an over-the-counter supplement, and most pharmacies do not stock it. It is also primarily a circadian timing agent, more useful for jet lag and shifted sleep phase than for staying asleep, which is what most people take it for.

It depends what you are recovering for. A resistance training study found cold water immersion after each session produced smaller gains in muscle fibre size, though strength gains were similar, and the authors advised avoiding it where hypertrophy is the goal. If you train to preserve muscle, separate the cold from the session by several hours or keep it for non-training days.

I would not recommend it. There is no meaningful evidence that it improves sleep, and deliberately occluding the mouth in someone with undiagnosed sleep apnoea or nasal obstruction is not a neutral thing to do.

Stop wearing the device for two weeks. Sleep anxiety driven by tracking data is a recognised pattern, and when the data is making your sleep worse, the data has stopped being useful.

If you snore loudly or have witnessed breathing pauses, if you wake unrefreshed despite adequate time in bed, if you are significantly sleepy during the day, or if difficulty sleeping has persisted for months. Those are assessments, not habit problems. ---

Book a consultation Longevity Clinic Published prices

Sleep as a Longevity Practice: What to Monitor, What to Ignore, and What a Wearable Cannot Tell You.