Fitness trackers and smartwatches have become ubiquitous health accessories. Most people wearing one have a rough sense of what it tracks — steps, heart rate, sleep. Fewer have a clear picture of how those numbers are generated, how accurate they really are, and what health conclusions you can legitimately draw from them.
Here's an honest breakdown.
Steps and movement
Step counting is the oldest and most reliable wearable metric. Accelerometers detect arm swing and convert it into step counts — accuracy is generally within 10 percent for walking on flat ground. It's a reasonable proxy for overall daily movement.
Where it breaks down: cycling, swimming, driving on a bumpy road, and repetitive hand movements can all generate phantom steps. Step goals (10,000 is culturally entrenched but was never based on clinical evidence) are useful for building habits, not targets to be hit precisely.
Heart rate
Optical heart rate sensors use green light to detect blood flow changes at your wrist. At rest and during steady-state exercise, accuracy is reasonably good — within a few beats per minute. During high-intensity interval training, rapid transitions, or if the watch fits loosely, readings can be significantly off.
Heart rate variability (HRV), which some devices now track, measures the variation in time between heartbeats. Higher HRV generally indicates better cardiovascular fitness and recovery. The trend matters more than any single reading — daily fluctuations are normal.
Sleep tracking
Sleep tracking is where wearables are most useful as trend indicators and least reliable as precise measurements. Devices estimate sleep stages — light, deep, REM — using movement and heart rate data. They're reasonably good at distinguishing sleep from wake, but their ability to accurately identify specific sleep stages is limited compared to clinical polysomnography.
What wearable sleep data does well: it reveals your average sleep duration and consistency over weeks. It can flag whether you're consistently getting less than seven hours, or whether sleep quality is trending worse. Use it directionally.
Calories burned
Calorie estimates are the least reliable metric wearables produce. They're calculated from heart rate, movement, weight and age using statistical models — not direct measurement. Errors of 20–30 percent are common, and errors of 50 percent or more occur during non-standard exercise. Don't use your device's daily calorie burn to justify eating significantly more.
What wearables can't tell you
A fitness tracker cannot diagnose anything. It cannot tell you whether a high heart rate is caused by dehydration, anxiety, a heart condition, or simply running up stairs. It cannot measure blood oxygen accurately enough to diagnose respiratory problems without clinical validation. It cannot replace a blood test, an ECG, or a clinical examination.
Several wearables now claim to track blood glucose, blood pressure or body temperature continuously — approaches that are technically promising but currently lack the accuracy needed for medical decision-making.
Putting wearable data in context
The real value of wearable data is longitudinal. A single day's readings are noise. Months of step counts, sleep durations and resting heart rate create a personal baseline. Deviations from that baseline — a resting heart rate that's 15 bpm higher than usual for five days running — are worth noting and, if persistent, discussing with a doctor.
Syncing your wearable data to a health record platform like Looms means your activity, sleep and heart rate trends sit alongside your lab results, prescriptions and visit notes — so patterns that cross boundaries become visible.