Sleep: what the research says

What you'll learn: what actually happens in your brain and body during a night's sleep, from light sleep to deep sleep and dreaming, and how those stages repeat in cycles.

This course is general education, not medical advice. If you are worried about your sleep or your health, talk to a doctor or another qualified health professional.

Meet Marcus

Marcus is 38, a project manager with two children aged four and seven and a 40-minute drive to work. On weeknights he usually turns the light off around 11:45 pm after scrolling his phone in bed, and his alarm goes at 6:15 am. At weekends he sleeps until nine if the children let him. He drinks three or four coffees a day and often has a glass or two of wine to unwind. He is not ill. He is just tired most afternoons, snappy by Thursday, and he has started to wonder whether the way he sleeps is the problem.

Over this course we will follow Marcus as he learns what sleep research actually says and tries some changes. His first step is a simple one: for a week he keeps a sleep diary, noting when he went to bed, roughly when he fell asleep, any wakings, when he got up and how he felt. Before he can interpret it, he needs to know what sleep is.

Sleep is not one thing

Sleep can look like a single state from the outside: eyes closed, still, unresponsive. Inside, the brain moves through distinct stages, which scientists identify mainly by recording brain waves, eye movements and muscle activity in a sleep laboratory. The current standard scheme has four stages.

N1, N2 and N3 are together called non-REM sleep. The percentages are typical figures for healthy adults; they vary between people and from night to night.

Deep sleep

N3 is the stage that seems most physically restorative. Growth hormone is released mainly during it, and research links it to recovery and to strengthening some kinds of memory. It is also when people are hardest to wake. If you are woken from deep sleep, you are likely to feel confused and heavy for a while, a state called sleep inertia.

REM sleep

REM sleep is when most vivid, story-like dreaming happens. The brain is about as active as when awake, but the body's major muscles are switched off, which is probably what stops people acting out their dreams. REM is thought to play a part in emotional processing and in some kinds of learning, though researchers still debate exactly what it does.

Cycles through the night

You do not go through the stages once. You cycle through them, typically N1, then N2, then N3, back up through N2 and into REM, and then round again. One cycle lasts about 90 minutes on average, though anything from roughly 70 to 120 minutes is normal, and most adults complete four to six cycles a night.

The mix changes as the night goes on. Think of a night's sleep as a playlist that repeats about every 90 minutes, but where the mix of tracks changes with each repeat. The early rounds are heavy on deep N3 sleep, with only a short burst of REM. The later rounds have little or no deep sleep and much longer stretches of REM, sometimes 30 minutes or more.

That pattern has a practical consequence. Cutting your sleep short at the end of the night removes a disproportionate amount of REM. Going to bed very late but waking at the usual time has a similar effect.

It is also normal to wake briefly several times a night, often between cycles. Most of these wakings last seconds and are never remembered.

What Marcus's diary showed

After a week, Marcus's diary looks like this on a typical weekday:

1. Lights out at 11:45 pm, asleep by around midnight.
2. Woke once around 3 am, needed the bathroom, back to sleep fairly quickly.
3. Alarm at 6:15 am: groggy, hit snooze once.
4. Roughly six hours of actual sleep.

Six hours is about four cycles. Because REM is concentrated late in the night, his early alarm is cutting into the REM-rich part of his sleep. And on mornings when the alarm happens to land in deeper sleep, he feels the heavy grogginess of sleep inertia.

Marcus also owns a fitness watch that reports his "deep" and "REM" sleep. Consumer wearables estimate stages from movement and heart rate rather than brain waves. They are reasonably useful for spotting patterns in total sleep time and timing, but much less accurate at telling the stages apart, so it is wise not to worry over one night's stage breakdown.

Some apps promise to time your alarm to the end of a 90-minute cycle. Because cycle length varies between people and through the night, that kind of timing is imprecise; the bigger lever for Marcus is simply the total amount of sleep he gets.

Recap

• Sleep has four stages: N1, N2, deep N3 and REM, each with different brain activity.
• Deep sleep is concentrated early in the night; REM sleep is concentrated later.
• Stages repeat in cycles of roughly 90 minutes, four to six times a night.
• Brief wakings between cycles are normal and usually forgotten.
• Cutting sleep short at the end of the night removes a lot of REM sleep.
• Wearables estimate sleep timing reasonably well but sleep stages much less accurately.

What you'll learn: how your body clock and the build-up of sleep pressure work together to decide when you feel sleepy and when you feel alert.

The slump and the second wind

Marcus's diary showed something else. Most days he hits a wall at around 3 pm, staring at spreadsheets and reaching for another coffee. Then in the evening, after the children are in bed, he gets a second wind at about 10 pm and feels too awake to sleep, so he picks up his phone. By the time he feels sleepy it is nearly midnight. He assumed this was just his personality. In fact it is a textbook example of two biological systems interacting, described by what sleep scientists call the two-process model, first set out by Alexander Borbély in the early 1980s and still the standard framework.

Process S: sleep pressure

The first process is sleep pressure, sometimes written as Process S (for sleep). From the moment you wake up, pressure to sleep starts to build. The longer you are awake, the stronger it gets. Sleep is what releases it.

A key chemical involved is adenosine, a by-product of the brain's energy use. It builds up in parts of the brain during waking hours and binds to receptors that promote sleepiness. During sleep, especially deep sleep, adenosine levels fall again. This is why, after a very long day, you can fall asleep almost anywhere, and why deep sleep is concentrated early in the night: there is the most pressure to discharge.

The usual way to picture it is as an hourglass. Waking up turns the hourglass over, and sand (sleep pressure) runs into the bottom all day. Sleep turns it back over and empties it. If you do not sleep enough, the hourglass is not fully reset, and you start the next day with some sand already in the bottom.

Process C: the body clock

The second process is the circadian rhythm, Process C (for circadian). Most cells in your body keep their own roughly 24-hour rhythm, coordinated by a master clock in a small region of the brain called the suprachiasmatic nucleus, just above where the optic nerves cross.

The clock does not keep perfect time on its own. In most people it runs a little longer than 24 hours, around 24.2 hours on average, so it needs resetting daily. Its most powerful reset signal is light reaching special light-sensitive cells in the eye. Meal times, exercise and social routines help to a lesser degree.

The clock drives many daily rhythms:

• Melatonin, a hormone from the pineal gland, rises in the evening when it is dark and signals "night" to the body. It does not knock you out; it helps tell your body when night is.
• Core body temperature falls in the evening and reaches its low point in the early morning, a couple of hours before your usual waking time.
• Alertness is promoted by the clock during the day, with the strongest alerting signal in the evening.

How the two interact

The clever part is how the two processes fit together. Sleep pressure rises steadily through the day. If it acted alone, you would get sleepier and sleepier from mid-morning on. The clock pushes back, sending an alerting signal that grows through the day and peaks in the evening, offsetting the growing pressure. Then, at night, the clock's alerting signal drops away while pressure is high, and sleep comes easily.

The early afternoon dip is a real feature of the clock and happens whether or not you eat lunch, although a big lunch may make it feel worse. The evening peak in alertness is so reliable that sleep scientists call the couple of hours before your usual bedtime the wake maintenance zone: it is one of the hardest times of day to fall asleep, even when tired.

What this means for Marcus

Marcus's pattern now makes sense:

1. The 3 pm wall is the circadian dip, made much worse by the sand left over from short nights.
2. The 10 pm second wind is his clock's evening alerting peak. Using it to scroll his phone means bright light in his eyes, which, as we will see in lesson 5, can push his clock later.
3. Weekend lie-ins shift his light exposure later, nudging his clock later still, so Sunday night is the hardest night of the week to fall asleep.

None of this means Marcus is broken. It means his habits are fighting his biology instead of working with it. He writes one change on a sticky note for now: on weeknights, notice the second wind and do something calm and dimly lit, rather than reaching for the phone.

Recap

• When you feel sleepy is set by two processes: sleep pressure (Process S) and the body clock (Process C).
• Sleep pressure builds through waking hours, partly via adenosine, and is released by sleep, like an hourglass turned over.
• The body clock runs at about 24.2 hours in most people and is reset mainly by light; it drives melatonin, body temperature and alertness.
• The clock's alerting signal peaks in the evening, creating a "second wind" a couple of hours before usual bedtime.
• The early afternoon slump is a real circadian dip, made worse by short sleep.