The Two Processes That Decide When You Fall Asleep

Sleep timing is usually described as the product of two independent processes acting at once. The model is decades old, and it still accounts for most of what people notice about their own nights.

Pressure that builds with time awake

The first process is a homeostatic one. From the moment of waking, a drive towards sleep accumulates, growing steadily through the day and dissipating during sleep itself.

This pressure has a physical correlate. Byproducts of neural activity build up while the brain is working and are cleared during sleep, which is why the drive rises with wakefulness rather than with the hour.

Because it is cumulative, a longer day produces a stronger drive. Skip a night entirely and the pressure keeps rising, which is why a second night without sleep feels qualitatively worse than the first.

A clock signal that ignores how long you have been awake

The second process is circadian. An internal clock produces a rhythm of alerting signal that peaks and troughs on roughly a daily cycle, regardless of how much sleep has been taken.

This signal is not a mirror image of sleep pressure. It rises through the afternoon and early evening, actively counteracting the accumulating drive, and falls away sharply in the hours before habitual bedtime.

Its timing is set by light, chiefly light in the morning and early evening, which is how the clock stays aligned with the outside world rather than drifting on its own period.

Why the interaction explains the evening

The combination produces the familiar shape of a day. Sleep pressure is high by early evening, but the alerting signal is near its peak, so most people feel reasonably awake.

When that signal drops, the accumulated pressure is suddenly unopposed and sleepiness arrives quickly. The sensation of a window opening, and of missing it, follows directly from the shape of these curves.

Staying awake past that point often produces a second wind. The pressure has not gone anywhere, but the clock has moved into a phase where sleep is harder to initiate.

What the model explains about naps

A nap discharges some accumulated pressure. That is what makes it restorative, and also what makes a long or late nap interfere with the following night.

The timing of the afternoon dip is a clock effect rather than a consequence of lunch. A shallow trough in the alerting signal lets existing pressure show through for an hour or two.

Short naps taken in that window discharge relatively little, which is the mechanism behind the common advice to keep them brief and early rather than a matter of habit.

Where the two-process account stops

The model describes timing well and content not at all. It says when sleep is likely, not what stages will appear, how consolidated the night will be, or what will be dreamt.

It also assumes an untroubled system. Conditioned arousal, pain, medication and shift schedules all disturb the relationship between the curves and what actually happens at bedtime.

Its value is as a frame. Most ordinary complaints about sleep timing turn out to be a mismatch between the two processes rather than a shortage of either one.