Training with the Apple Watch: what changes when logging moves from your pocket to your wrist

September 7, 2026 · 5 min read

Heart rate measured at the wrist during resistance training is, in fact, accurate — a controlled study shows near-perfect agreement with a chest strap. But that accuracy is not uniform: it varies from exercise to exercise, and drops considerably in movements that rotate the wrist. Calories, the metric every watch advertises, remain the weak point of any wearable — Apple Watch included.

Most of the literature validating a watch against a chest strap tests running and cycling — cyclical activities, with the wrist swinging back and forth in the same pattern on every stride. Resistance training is a different thing: the wrist rotates, stops, takes impact, and each exercise moves the arm in its own way. Two studies test resistance training specifically against a reference chest strap, and together they show a pattern that is more specific — and more useful for someone lifting weights — than the loose idea that "wearables are accurate" or "wearables are useless."

What the wrist measures well: heart rate, for most of the workout

Hajj-Boutros and colleagues (2023), in the European Journal of Sport Science, tested the Apple Watch 6, the Polar Vantage V and the Fitbit Sense against a Polar H10 chest strap in 60 healthy young adults (30 men, 30 women), across a session that moved through sitting, walking, running, cycling and resistance exercise. For the resistance exercise segment specifically, the Apple Watch correlated at 0.98 with the chest strap, with a coefficient of variation of 2.3% and a mean bias of only 1.3 beats per minute — with the wrist reading ranging from 6.9 beats below to 4.3 above the strap in the most discrepant cases. In a typical resistance-training set, that is a small enough difference to not confuse one effort zone for another.

But accuracy changes from exercise to exercise

Støve and Hansen (2022), in the Journal of Sports Sciences, tested exactly the movements that matter to someone lifting weights: 29 people (16 women, ages 19 to 37) performed the barbell back squat, barbell deadlift, dumbbell curl to overhead press, seated cable row and burpee, all against the same Polar H10 chest strap. In the squat, deadlift and cable row — movements where the wrist stays relatively still or traces a stable arc — agreement with the chest strap was high. In the curl to overhead press and the burpee — where the wrist rotates and takes impact — agreement dropped to moderate or low.

Why this likely happens. The watch's sensor reads the change in color of blood under the skin through light (photoplethysmography), and it depends on relatively stable contact to work well. A chest strap reads an electrical signal, which does not suffer from the wrist rotating or striking something. That is a plausible explanation for the difference between exercises — but the two studies together add up to 89 people across a single lab session each: the direction is consistent, the exact size of the effect is not yet well established, and neither one looked at what happens over weeks of real training.

What it measures poorly: calories

In the same Hajj-Boutros study, while the Apple Watch's heart rate scored a CV of 2.3%, the energy-expenditure estimate from the same watch — and from the other two devices tested — had a CV between 13% and 30% against a reference indirect calorimeter. Calories are not something the sensor measures directly: it is an estimate that multiplies the heart rate reading, an algorithm's guess about oxygen consumption, and a body weight someone typed in once — each step carries its own error, and the errors compound. The calorie count at the end of a workout is the watch's worst number, and precisely the one that draws the most attention on the screen.

What changes when logging moves from your pocket to your wrist

The wrist does not replace what gets logged by hand — load, reps, RPE. It adds one more signal: effort, seen through heart rate, continuously and without requiring another tap on the screen. In PulsoGym, the average and peak heart rate of workouts recorded by the Apple Watch comes in through the Apple Health connection, alongside resting heart rate and heart rate variability on the recovery card — a complement to what a person already logs, not a replacement for it. The watch does not know how much weight was on the bar; that is still what someone types in.