What Your Sleep Tracker Can and Cannot Tell You

Almost everyone who takes sleep seriously ends up wearing something to bed. A ring, a watch, a band, a mattress sensor. Then they wake up, look at a number, and let that number decide how they feel about the night they just had.

That last step is where it goes wrong. A tracker is a genuinely useful instrument, but only for a narrow set of questions, and the questions most people ask it are the ones it answers worst.

What your tracker is actually measuring

No consumer device measures your sleep directly. Measuring sleep properly means recording brain waves, eye movement and muscle tone, which requires electrodes on your head and face.

What a wearable records instead is:

• Movement, from an accelerometer.
• Heart rate, from a green light shining into your skin and reading the reflection.
• Heart rate variability, the variation in time between beats, calculated from that signal.
• Skin temperature, and on some devices an estimate of blood oxygen.
• Breathing rate, inferred from the heart signal.

Everything else is an inference. When your ring reports that you spent 58 minutes in deep sleep, no deep sleep was observed. An algorithm looked at how still you were and what your heart was doing and estimated a number.

That is not a scandal. It is just important to know what kind of number you are looking at.

What they get right, and what they do not

Trackers are good at telling asleep from awake. Modern devices agree with laboratory testing most of the time on the basic question of whether you were asleep. They also do a decent job with how long you slept and roughly when.

They are much weaker at sleep stages. Splitting the night into light, deep and REM is the hardest thing they attempt, and it is where agreement with laboratory measurement falls off sharply. Deep sleep and REM estimates are the least reliable numbers your device produces, and they are the two people fixate on most.

They tend to flatter you. Because stillness reads as sleep, devices generally overestimate how long you slept and underestimate the time you spent awake in the night. If you lie still and frustrated at 3 am, your tracker may well score that as sleep.

Resting heart rate and heart rate variability are the most trustworthy things on the screen, because they are measured rather than inferred. That is a useful fact, and most people ignore those two in favor of the sleep score.

Orthosomnia, or how tracking makes sleep worse

In 2017 a group of sleep researchers gave a name to something they kept seeing in clinic: patients whose sleep problems were being driven by their pursuit of perfect tracker data. They called it orthosomnia.

The pattern is familiar. Someone sees a poor score, worries about it, goes to bed anxious about tomorrow's number, and sleeps worse. Then they chase the number harder. Anxiety about sleep is one of the most reliable ways to damage sleep, and a device that grades you every morning is very good at producing it.

If your first act on waking is to check your score, and that score sets your mood for the morning, your tracker has stopped being an instrument and become a judge. That is the point to put it away for a few weeks.

How to actually use one

The device is at its best as a detector of causes, not a scoreboard.

Look at weeks, not nights. A single night tells you almost nothing. A seven day average tells you something real. Anything you conclude from one night's data is noise.

Change one thing at a time, for two weeks. This is the whole value proposition. Move your last drink earlier and watch what happens to resting heart rate and heart rate variability across two weeks. Drop the bedroom temperature and watch what happens to wakings. That is an experiment. Comparing tonight to last night is not.

Trust the measured numbers over the estimated ones. Resting heart rate, heart rate variability and breathing rate are closer to the truth than the stage breakdown. If you drank, your heart rate will run high and your variability will fall, and that shows up clearly. I went through why in what alcohol actually does to your sleep.

Compare yourself only to yourself. Heart rate variability in particular varies enormously between people, by age, fitness and simple anatomy. A number that is excellent for one person is unremarkable for another. Your own baseline is the only useful comparison.

Use it to catch the obvious. Most of what a tracker reveals is not exotic. It is a warm bedroom, a late meal, a glass of wine, an inconsistent wake time. Those are the same four things I check first in why you wake up at 3 am, and a tracker is a good way to prove to yourself that they matter. The temperature one is the most commonly missed, and the numbers behind it are in the cold room rule.

The one thing you must not use it for

Do not use a tracker to rule out sleep apnea.

This is the part I care about most, because I see the consequences. Some devices now report a blood oxygen trace or flag breathing disturbances. Those features are interesting. They are not diagnostic, they miss a great deal, and a normal looking reading is not reassurance.

If you snore habitually, if anyone has seen you stop breathing or gasp, or if you wake unrefreshed after a full night no matter what your score says, that needs a proper evaluation rather than a wearable. The difference between harmless snoring and something that needs treating is in snoring, when it is harmless and when it is a red flag.

A good score has talked plenty of people out of getting checked. That is the most expensive mistake this category enables.

Where it fits

A tracker sits in the tools layer, which means it comes after the free things rather than instead of them. Light, temperature, timing and alcohol are still where the returns are, and the order to work through them is in the sleep biohacking guide. A slightly different framing of the same idea is in sleep optimization 101.

And if your device has told you something that contradicts what you have read about sleep, be careful. A good deal of received wisdom is wrong, which I collected in ten sleep myths, but the answer to a confusing number is rarely to trust the number harder.

Used well, a tracker is a way to run experiments on yourself and see the results. Used badly, it is a nightly grade that makes you anxious about the one thing you cannot force.