Sleep: the most underrated thing you can do

Sleep gets treated as background advice — the thing mentioned at the end of an article, after the useful parts. In the context of changing how you look and perform, it belongs near the front.

The study people should know about

A controlled trial put participants on identical calorie deficits and varied one thing: sleep. One group slept 8.5 hours a night, the other 5.5.

Both groups lost a similar amount of total weight. The composition of that weight differed sharply — the sleep-restricted group lost considerably more of it as lean tissue and less as fat.

Same food, same deficit, materially worse outcome, from sleep alone.

It is one study and the sample was small, so do not treat the exact ratios as settled. But it is consistent with a broader body of work, and the direction is not seriously disputed.

What short sleep does to appetite

Two hormones do most of the visible work.

Ghrelin, which signals hunger, rises with sleep restriction. Leptin, which signals sufficiency, falls. The result is a body that is hungrier and less easily satisfied on the same amount of food.

This is not a small effect and it is not a willpower problem. Short-sleep studies consistently find participants voluntarily eating several hundred additional calories a day, weighted toward calorie-dense and carbohydrate-rich food.

There is also a decision-making component. Sleep deprivation reduces activity in the prefrontal regions involved in impulse control while increasing reward-related responses to food. You are hungrier, and simultaneously less equipped to do anything about it.

If you are dieting and find that your resolve collapses reliably in the evenings, look at what time you are going to bed before you look at your character.

What it does to training

Reduced strength and power output, particularly on multi-joint movements. Higher perceived exertion, so the same weight feels heavier. Slower reaction time and worse coordination, which matters for technique-dependent lifts. Slower glycogen resynthesis.

Growth hormone release is also concentrated in the deep sleep of the first half of the night, and that window is the first thing lost when total sleep shortens.

You can train on poor sleep — and should, at reduced volume — but you will not adapt to it as well.

The number

Seven to nine hours for most adults. Some people genuinely need less, and a very small proportion carry a genetic short-sleep variant. Most people who believe they are in that group are simply accustomed to being tired.

A useful test: if you sleep substantially longer when you have no alarm — on holiday, at weekends — you are running a deficit during the week.

Fixing it, in order of what actually works

1. A consistent wake time, including weekends. The single highest-yield change and the one people resist. Your circadian rhythm is anchored primarily by when you wake and get light, not by when you go to bed. A fixed wake time drags bedtime into place behind it. Lying in on Saturday and Sunday produces something functionally similar to jet lag every Monday.

2. Morning light. Ten to fifteen minutes outside within an hour of waking. Outdoor light is many times brighter than indoor lighting even on an overcast day, and it is the strongest signal available for setting the clock. It makes falling asleep at a sensible time that night considerably easier.

3. Caffeine cut-off. Caffeine has a half-life of roughly five to six hours, so a 4pm coffee leaves a meaningful dose in your system at midnight. It shortens deep sleep even in people who fall asleep fine — the absence of difficulty falling asleep is not evidence that it is not affecting quality. Nothing after about 2pm if you have any sleep difficulty at all.

4. Alcohol. It gets you to sleep faster and makes the sleep worse — suppressed REM early, fragmented waking later. If you are drinking most evenings and sleeping badly, those two facts are related.

5. A cool, dark room. Core temperature needs to fall for sleep onset. Somewhere around 16–19°C suits most people. Genuinely dark, not nearly dark.

6. A wind-down that is not a screen. The blue light argument is weaker than commonly claimed — the larger problem is that phones are engineered to be engaging, and engagement is the opposite of sleep onset. Thirty minutes of something dull beats any filter setting.

What about naps

A 20-minute nap is genuinely restorative and does not usually interfere with night-time sleep. Beyond about 30 minutes you enter deep sleep and wake groggy, and late-afternoon naps reduce the sleep pressure you need at bedtime.

Nap early and briefly, or not at all.

Sleep trackers

Consumer devices are reasonably good at total sleep duration and bedtime consistency. They are considerably less accurate at sleep staging — the light, deep and REM breakdown is estimated from movement and heart rate rather than measured.

Use them for duration and consistency. Do not treat the stage percentages as precise, and be alert to the genuine phenomenon of anxiety about tracked sleep scores making sleep worse. If the number is stressing you, stop looking at it.

When it is not a habits problem

Some sleep problems are medical, and no amount of routine adjustment resolves them.

Loud snoring with pauses in breathing, waking gasping, or profound daytime sleepiness despite adequate time in bed can indicate sleep apnoea, which is common, underdiagnosed, and has significant cardiovascular consequences. It is also very treatable.

Persistent insomnia lasting months is likewise a medical matter, and the first-line treatment — cognitive behavioural therapy for insomnia — is more effective than medication in the long run.

If you have done the sensible things for a month and still sleep badly, see a doctor rather than buying another supplement.

Shift work and irregular schedules

Some people cannot keep a consistent sleep schedule, and the standard advice can read as unhelpful when your rota changes weekly.

What still applies: total sleep duration matters most, so protect the hours even when the timing is not ideal. Darkness matters enormously — blackout blinds and a sleep mask for daytime sleeping are not luxuries, they are the main intervention. And light exposure at the start of your waking period, whenever that falls, still helps anchor the rhythm.

What changes: expect to need somewhat more total sleep, accept that sleep quality will be lower, and be more conservative with training volume than someone on a settled schedule. Shift work is a genuine physiological stressor and treating it as neutral leads to accumulated fatigue.

Caffeine timing becomes more important, not less. Working out your own cut-off relative to when you intend to sleep, rather than to the clock, is the adjustment that matters.

Sleep and injury risk

The connection is less obvious than the appetite one and reasonably well documented in athletic populations.

Studies in adolescent and young adult athletes have found meaningfully higher injury rates in those sleeping fewer than eight hours a night. The plausible mechanisms are reduced coordination and reaction time, impaired tissue repair, and greater accumulated fatigue.

For anyone lifting heavy loads, that matters. A technical error under a loaded barbell is a different proposition from a technical error while typing. If you are chronically underslept and about to attempt a personal best, the sensible decision is to postpone it.

The framing

If you would not skip your training sessions, do not skip sleep. It is the same category of input — one of the small number of things that determines whether the effort you are putting in produces a result.