Sleep, HRV and the overnight recovery window

The recovery window

Across a normal day, the autonomic nervous system spends most of its time managing demand. Sleep is the only extended period in which it can do the opposite: vagal tone rises, heart rate falls below its daytime floor, blood pressure dips, and the systems that have been running all day get several hours of low load.

This is why overnight HRV has become the standard measurement window on wearables. It is not just convenient — it is the period during which the signal is cleanest and most informative, because posture, activity and conscious breathing are all held roughly constant.

What happens across the night

Autonomic state is not uniform through the night. During deep, slow-wave sleep — concentrated in the first half — parasympathetic dominance is at its strongest and HRV is high. During REM sleep, which lengthens toward morning, sympathetic activity rises again and heart rate becomes irregular in a quite different way.

Two consequences follow. First, a night that is short at the end loses mostly REM, while a night that is disrupted at the start loses slow-wave sleep and with it most of the deepest autonomic recovery. Second, HRV averaged across a whole night depends on how much of each stage you got, which is one reason single-night values bounce around.

The two-way loop with stress

Sleep and stress form a loop that runs in both directions, and it is genuinely difficult to say which end is driving.

Elevated evening arousal delays sleep onset and fragments the first part of the night. The resulting short or broken sleep raises next-day cortisol and inflammatory markers and lowers HRV, which makes the following day's demands feel larger. That, in turn, raises evening arousal.

The practical implication is that you can enter the loop from either side. People who cannot address the stressor directly can often still fix the sleep, and the loop weakens from that end too. That is why the Calmspan test includes two recovery questions alongside the PSS-10 items — restedness carries information the stress questions miss.

Alcohol, the clearest signal in the data

If you own a wearable and want to see a large, reliable, same-night effect on your own data, drink two glasses of wine at 9pm.

Alcohol has one of the most consistent acute effects on overnight HRV of anything measurable at home. It raises heart rate and suppresses parasympathetic activity for hours, with the largest effect in the first part of the night — precisely the window in which deep sleep and the strongest autonomic recovery would otherwise be happening. The effect scales with dose, is present at quantities most people consider modest, and is often still detectable the following night.

Late training and late eating

Hard exercise elevates sympathetic activity for hours afterwards. A demanding session finishing at 9pm frequently shows up as suppressed HRV for the first half of the night — not because the training was harmful, but because the recovery started late. Moving the same session three hours earlier often resolves it entirely.

Large late meals have a smaller but similar effect: digestion raises heart rate and delays the nocturnal dip. Neither is a reason for anxiety about an occasional late dinner. Both are worth adjusting if your overnight numbers are chronically flat.

Regularity beats duration

The most useful recent finding in sleep research is that when you sleep may matter more than how long. In a large accelerometer-based cohort study, sleep regularity — the consistency of sleep and wake timing from day to day — predicted mortality risk more strongly than sleep duration did.

Duration still matters; the U-shaped relationship between hours slept and mortality is one of the most replicated findings in the field, with the flat bottom sitting around seven to eight hours. But regularity is the underrated variable, and it happens to be the one most people can actually control.

For the autonomic system specifically, a fixed wake time is the strongest single anchor. Circadian timing drives the cortisol awakening response, and a stable circadian signal is what allows the evening decline to be steep rather than flat.

A practical protocol

1. Fix the wake time first, seven days a week, and let bedtime follow. Weekend lie-ins are the most common single cause of an irregular schedule.
2. Get bright light within an hour of waking — outdoors if possible. This is what steepens the cortisol curve at the right end.
3. Hard stop on work input 90 minutes before bed. The problem is cognitive activation, not blue light; a spreadsheet on paper is as activating as one on a screen.
4. Four alcohol-free nights a week, and none within three hours of sleep.
5. Finish hard training by early evening where your schedule allows.
6. Keep a notebook by the bed. If you wake with racing thoughts, writing the thought down reliably shortens the episode — the mind stops rehearsing what it no longer has to hold.