Sleep and training: the recovery variable people skip

Training and nutrition get the attention because they are the parts you actively do. Sleep is the part that happens whether or not you plan it, which is precisely why it tends to be the recovery variable people leave unmanaged.

About those ninety-minute cycles

Sleep progresses through light sleep, deep slow-wave sleep and REM in a repeating cycle, and ninety minutes is the figure everyone quotes.

It is an average, and a fairly loose one. Real cycles run roughly seventy to a hundred and twenty minutes, vary between people, and vary within a single night — early cycles contain proportionally more deep sleep, later cycles more REM.

This is why cycle-based bedtime calculators, including ours, should be treated as a nudge rather than a schedule. Waking between cycles rather than out of deep sleep does tend to feel easier, so aiming for a whole number of cycles is a reasonable heuristic. But the precision implied by “go to bed at 11:15 exactly” is not something the underlying physiology supports.

What sleep loss does to training

The performance literature here is less settled than the confident version you usually hear. The major review of sleep loss in athletes concludes that findings conflict, and that the extent and mechanisms of the effect remain uncertain — noting specifically that some maximal efforts and gross motor tasks can be maintained despite sleep loss. Where the evidence is clearer is in sport-specific performance rather than raw physical output. ¹

What that leaves is a real but narrower claim: effects appear more reliably in sustained and skill-dependent efforts than in single maximal attempts.

One finding matters more than the raw performance numbers: perceived exertion rises. The same session feels harder after poor sleep even when output has not fallen much. That subjective cost is what erodes training consistency over weeks — sessions get shortened, optional work gets skipped, and intensity quietly drifts down.

Strength on a single heavy attempt appears relatively robust to one bad night. What suffers is the accumulation: multiple hard sessions across a week, technical work, and the willingness to finish a programme as written.

Why this matters more in a deficit

If you are dieting, sleep moves from useful to important.

Controlled work restricting sleep during calorie restriction found that participants lost a greater proportion of lean mass and less fat when sleep was short, despite matched calorie intake. On 5.5 hours in bed rather than 8.5, fat loss fell from 1.4 kg to 0.6 kg while lean-mass loss rose from 1.5 kg to 2.4 kg over two weeks. ²

That is a striking result and it is worth knowing how thin the evidence under it is: ten people, fourteen days, in a crossover design. The authors name the sample size and duration as limitations themselves. The direction is plausible and consistent with the appetite findings below, but the widely repeated “55 per cent less fat lost” figure rests on a study small enough that it should be treated as a signal to take sleep seriously, not as a precisely known quantity. Separate research on sleep curtailment in healthy young men found changes in appetite signalling — leptin down about 18 per cent, ghrelin up about 28 per cent, and the ratio between them up 71 per cent — accompanied by increased hunger and appetite, particularly for calorie-dense foods. ³

Read that one carefully too. It compared four hours in bed against ten in twelve young men over two days, with calories delivered by constant glucose infusion rather than by eating. That is a clean way to isolate the hormone signal and a long way from anyone’s actual week. What it establishes is that the mechanism is real; it does not tell you the size of the effect at, say, six hours versus eight while eating normally.

The practical translation is that short sleep makes adherence harder through a mechanism you cannot willpower your way around. You are not weak for finding a deficit harder after four hours of sleep; you are operating with altered appetite signalling.

This is also why “I’ll sleep less to fit in a morning session” is usually a bad trade during a cut. The session gains you a few hundred calories of expenditure and may cost you more than that in intake.

What sleep trackers can and cannot tell you

Consumer devices estimate sleep from movement and heart rate. The most thorough comparison tested seven of them against polysomnography — the laboratory standard — across three nights in 34 adults, and the result was kinder to the devices than their reputation suggests: most performed as well as or better than research-grade actigraphy at telling sleep from wake. ⁵ Comparisons like it have generally found that modern devices do a reasonable job of estimating total sleep time and a considerably poorer job of classifying sleep stages.

So the nightly “deep sleep score” deserves much less attention than it receives. Total duration and bedtime consistency are the outputs worth reading, and both are things you could track with a note on your phone.

There is also a documented irony worth knowing about: anxiety over tracker sleep scores can itself impair sleep. If your device is making you worry about sleeping, that is a reason to stop looking at the score, not to buy a better device.

The bottom line

Consensus guidance for adults is seven to nine hours — narrowing to seven to eight for older adults — and it is worth knowing what kind of number that is. It came from an 18-member expert panel working through the published literature by a structured consensus method, not from a single trial measuring outcomes at each duration. ⁴ It is informed judgement about where the healthy range sits, which is the right tool for the question and still a different thing from a measured result. The range exists because individual requirement genuinely varies. Find where in it you function well rather than treating a number as pass or fail.

Then prioritise consistency over cycle arithmetic. A regular seven hours beats an erratic nine, and going to bed and waking at similar times is the single highest-return change most people can make. Use the calculator to pick a bedtime, then keep it — the keeping is the part that works.