Your Sleep Score Is Not a Verdict: What Wearables Get Right and Wrong

It is 6:52 a.m. Your eyes are not all the way open, but your hand has already found the phone. Before you have noticed how you actually feel, a ring of color tells you. Sleep score: 61. Recovery: low.

And just like that, you are tired. Or more precisely, you have been told you are tired. Your jaw tightens a little. The whole morning tilts.

If that has happened to you, nothing is wrong with you. You bought a wearable because you care: about your heart, your energy, about being here and capable for the people who count on you. Wanting to see inside your own body is a reasonable, even beautiful, impulse.

The trouble is that the device speaks every number in the same confident voice, whether it measured it well or mostly guessed. Some of what it tells you is close to clinical quality. Some is an estimate dressed up in decimals. Nobody prints the difference on the box.

I am an electrical engineer by training, and I build AI systems for a living, so I think about sensors the way engineers do. Every measurement has an error bar, and a reading is only useful if you know how wide that bar is. So here is the error bar, as best the research can draw it today. This is general information, not medical advice; a worrying symptom or alert is always a conversation for your doctor.

Are health wearables worth it?

For most people, a wearable is worth it as a motivation tool and a trend tracker, and not worth it as a diagnostic instrument or a nightly report card.

They are mainstream now. A 2023 analysis in JAMA Network Open of a national federal health survey estimated that about 29 percent of US adults used a wearable device in 2019 to 2020, and roughly half of users wore one every day.

The strongest case for them is movement. A 2022 umbrella review in The Lancet Digital Health, pooling 39 systematic reviews covering about 164,000 participants, found activity trackers increased physical activity by the equivalent of roughly 1,800 extra steps a day and about 40 more minutes of walking a day, with weight reductions of about 1 kg. Effects on blood pressure, cholesterol and quality of life were typically small and often not significant.

Those 1,800 steps matter more than they sound. In the research I walk through in what longevity science actually supports, more daily steps tracked with lower mortality up to roughly 6,000 to 10,000 a day, depending on age. A device that nudges you up that curve is doing real work on your behalf.

What do wearables measure accurately?

Heart rate is the thing wrist wearables measure best, and calories burned is among the things they measure worst.

A 2017 Stanford study in the Journal of Personalized Medicine tested seven wrist devices on 60 volunteers against continuous medical monitoring while they sat, walked, ran and cycled. Six of the seven had a median heart-rate error below 5 percent during cycling. But no device estimated energy expenditure with an error below 20 percent. Errors were also higher during walking, and for people with darker skin, higher body mass index, and for men.

So if your watch says you burned 600 calories on a walk, hold that number loosely. It is not a license for dessert, and it is not a reason for guilt either.

Overnight, the picture is similar. In a 2022 lab study in Sensors, 53 adults slept wearing six devices at once alongside clinical sleep monitoring and an ECG. Nighttime heart rate was generally close for most devices, while heart rate variability (HRV) agreement ranged from poor to excellent depending on the device.

I treat HRV the way I would treat any noisy sensor in an engineering system: as a trend over weeks against my own baseline, never as a single morning grade.

How accurate is wearable sleep tracking?

Wearables are good at telling whether you were asleep and roughly for how long, and much less reliable at telling which stage of sleep you were in.

In that same 2022 study, the six devices agreed with clinical sleep monitoring about 86 to 89 percent of the time on simple sleep versus wake. On specific stages, agreement fell to roughly 50 to 65 percent. A 2021 study in Sleep of seven consumer sleep trackers found the same shape: high sensitivity for detecting sleep, weaker detection of time awake, inconsistent staging, and worse performance on nights of poorer or disrupted sleep.

Read that last part again. The nights you most want an accurate answer, the restless ones, are the nights the device is most likely to be wrong.

Your deep sleep number is an estimate. Your total time asleep, and how you feel at 10 a.m., are often better information.

Can tracking your sleep make you sleep worse?

It can for some people, and sleep clinicians have a name for it: orthosomnia.

The term comes from a 2017 case report in the Journal of Clinical Sleep Medicine by clinicians at Rush University and Northwestern. They described patients seeking treatment for insomnia and short sleep they had inferred from their tracker data, a pursuit of ideal sleep that had turned perfectionistic. Striking detail: the tracker data often felt more true to these patients than validated lab testing did.

You can probably feel how this happens. You wake at 2 a.m. and the first thought is not about your body, it is about the number. What is this doing to my score? And the wondering itself keeps you awake. The tool you bought to protect your sleep is now standing between you and it.

If that is you, take this as absolution. You were not being neurotic. You were handed a grade with no error bar and asked to improve it, and anyone who cares would try. Some gentle relief lives in a humane news and phone diet for better sleep and in a two-minute breathing reset for the moment you wake up wired.

Should you trust heart alerts from a smartwatch?

Treat a heart-rhythm alert as a reason to call your doctor calmly, not as a diagnosis, and not as something to ignore.

The largest test so far, a 2019 study in the New England Journal of Medicine, enrolled 419,297 smartwatch users. Only 0.52 percent received an irregular pulse notification over a median of 117 days of monitoring. Among notified people who later wore an ECG patch, 34 percent had atrial fibrillation on the patch. And when a new alert fired while the patch was on, 84 percent of those alerts matched atrial fibrillation on the ECG.

So alerts are rare, often meaningful, and not the final word. That is exactly the kind of signal a clinician should interpret, not a search engine at midnight.

The emotional side is real too. A 2024 study in the Journal of the American Heart Association of 172 people already diagnosed with atrial fibrillation found wearable users reported more symptom monitoring and preoccupation, and 20 percent of them experienced anxiety and always contacted their doctors after irregular rhythm notifications. The study was small and observational, so it cannot say the device caused the worry. But if you recognize yourself in it, you are in good company.

How do you use health data without becoming anxious?

Use your wearable for trends and nudges, decide in advance what a number will make you do, and give your own body the final vote.

Here is the posture I suggest to anyone who asks me:

• Keep only the metrics that change what you do. If a number never changes a decision, it is just a source of feelings. Hide it.
• Read the week, not the night. One rough night is noise. A month of shorter sleep is a signal worth acting on.
• Feel first, then look. Notice how rested or wired you are before the app tells you what to think.
• Decide your response ahead of time. A rule written on a calm Sunday beats a reaction invented at 2 a.m.
• Let the device rest sometimes. If tracking is costing you sleep, a drawer is a perfectly good place for a watch.

One more thought about the decade we are walking into. I build AI systems every day, and I expect these sensors will increasingly be read for us by software that sounds even more confident than the watch does. That makes the error bar matter more, not less. If you hand your data to a chatbot, read how to use an AI health assistant safely first.

The goal was never a perfect score. The goal is a person who moves more, sleeps enough, notices early when something is truly off, and does not spend their mornings being graded. You get to be the one who reads the instrument. The instrument does not get to read you.

The Future of Wellness letter is education only. It is not investment, medical or financial advice, and it is not an offer to sell or a solicitation of an offer to buy any security. No business or practitioner mentioned in this letter paid to be included. If New Earth Capital ever has a financial interest in a business named here, the letter will say so. This is education, not medical advice. Talk with your doctor or provider about your own health.