Sleep inertia

Sleep inertia is the impaired alertness and performance of the first minutes after waking. It is not a mood or a failure of will: it is a measurable transitional state, present in people who slept perfectly well, and in the first few minutes it is a heavier impairment than a full day and night without sleep.

Most of it clears inside half an hour. A measurable tail can run an hour or two. How severe it is depends mainly on two things, and neither of them is how good your night was: the stage of sleep you were pulled out of, and where in your body clock's day the waking happened.

The number that makes the point comes from Wertz and colleagues, who tested people three minutes after waking from a normal night and compared the scores with the same people after more than a full day of continuous wakefulness. Performance just after waking was worse. Briefly, waking up puts you below where sleep deprivation does.

The decay has a shape. Jewett's group tracked alertness and performance across the hours after waking and found the recovery runs as an exponential: steepest in the first fifteen to thirty minutes, most of the way back within an hour, with a small tail still detectable after two on sensitive measures. Tassi and Muzet's review puts the range for simple tasks at minutes and for demanding ones at considerably longer, which is the honest version of the answer — how long inertia lasts depends on what you are asking yourself to do.

What is affected is specific and recognisable: reaction time, holding something in mind for a few seconds, arithmetic, and decisions with more than one option in them. Subjective sleepiness tracks it only loosely, so the feeling of having come round is not proof of having come round.

This is also where the practical risk sits. The first ten minutes after an alarm is a bad window for a car, a consequential reply, or anything requiring you to remember what you were told.

The single best predictor of how rough a waking is turns out to be what the brain was doing at the moment it happened.

Slow-wave sleep produces the worst of it. Dinges and colleagues showed this with abrupt awakenings in the 1980s, and it has held up since: pulled out of deep sleep, people are substantially more impaired than they are from light sleep, and the effect is large enough to be visible within a single nap.

That one fact explains several familiar experiences.

Short naps feel different from long ones. Twenty minutes rarely reaches deep sleep. Forty minutes usually does, and the waking is worse than the nap was worth — the dose-response curve, and the reason the sensible ceiling exists, is in the entry on napping.

A night of too little sleep makes the morning worse, not just shorter. Deep sleep rebounds when it has been missed, so the alarm has a higher chance of landing in the middle of it.

Waking in the small hours is the worst case of all. Scheer and colleagues showed inertia follows a circadian rhythm of its own, peaking when the waking happens during the biological night. Being woken at three is not simply early. It is the hardest moment the clock offers, which is part of why a night waking at that hour feels so unlike one at six.

And snoozing re-enters sleep. Each new descent gives the alarm another chance to interrupt something deeper than the last.

Here is the inference almost everyone makes and it does not hold: I could not get up, so I slept badly.

What the first twenty minutes reports is mostly when and out of what you were woken, not how much or how well you slept. A good seven hours ending in the middle of a deep cycle produces a dreadful waking. A mediocre six hours ending at a light moment near your natural rise time produces an easy one. Read as a verdict on the night, grogginess gets the answer wrong in both directions regularly enough to be useless.

The clock version of this is the commonest case of all. If the alarm sits well before the window your body would have chosen, every weekday morning is an interruption rather than an ending, which is the ordinary experience of a late chronotype on an early schedule rather than evidence of disordered sleep — the trait itself is in the entry on chronotype.

None of which makes the state uninformative. Inertia is genuinely worse after short sleep and after a fragmented night, so a bad waking is weak evidence of a bad night. It is simply far weaker evidence than it feels like, and there are better instruments.

What actually tells you something about the night: the shape of the day as a whole rather than its first half hour, how long sleep took to arrive, how long you were awake in the middle, and whether a free morning with no alarm produces a different person. If the deficit turns out to be opportunity rather than ability, that is the question in the entry on sleep debt.

Less than people hope, and the things that work are unglamorous.

• A consistent rise time. The strongest lever, because it lets the body finish a cycle in the right place rather than being cut out of one at a different point every day.
• Light, modestly. Santhi and colleagues found morning light improved some aspects of post-waking performance and not others, so bright light is worth having and is not a cure.
• Cold water on the face, and caffeine taken at the start rather than the end. Hayashi and colleagues tested exactly this combination after a short nap. Caffeine needs around twenty to thirty minutes to arrive, so the coffee you drink on waking works on the tail of the inertia, never on its peak.
• Capping a daytime nap under about twenty-five minutes, for the staging reason above.
• Not snoozing, for the same reason.
• Not deciding anything. The cheapest intervention on this list is to move consequential things out of the first twenty minutes rather than trying to be sharper inside them.

Two sentences on the version of this that is not a nuisance but a hazard. The operational research on inertia exists because of on-call work: people woken to act — on a ward, in a cab, behind a wheel on the way home from a night shift — are impaired at precisely the moment they are being asked to perform, and the countermeasure is a buffer before the task rather than willpower inside it.

One scope note. Persistent, heavy difficulty waking, with long sleep and confusion that lasts well beyond an hour, is a recognised clinical picture — sleep drunkenness, described in Trotti's review — and it sits alongside idiopathic hypersomnia and untreated apnoea rather than alongside character. If that is the morning most mornings, it is a question for a doctor, and the fortnight of notes is what to take in. MoodyWonder's Sleep Quality test reads the night and the day rather than the first ten minutes, which is the distinction this page is about.