Heart-Rate Sensor Placement: Read the Version and Wear-Location Evidence

Heart rate sensor placement accuracy is not one number you can look up. Published evidence is tied to a specific hardware version, a specific place on the body, a specific activity and a specific reference device. If any of those four differ from your setup, the result may not carry over. This guide shows how to read the evidence that exists, and how to avoid a "most accurate brand" conclusion that no single study supports.

The short answer

• Version matters. A validation of one hardware generation does not describe another.
• Wear location matters. In one study, identical wrist-style bands worn on the wrist, forearm and upper arm did not perform alike.
• Activity and reference matter. Rest, steady walking, sleep and short burst exercise produce different results, and ECG and chest straps are different yardsticks.
• Ranking claims need the same protocol. A device that leads in one test can trail in another.

Step 1: Confirm your version and wear location

Before you read any study, write down two facts about your own setup.

1. Which hardware generation you wear. A finding should be applied only to the generation the study tested. Work that says only a brand name, with no generation, is incomplete. For placement guidance, follow the manufacturer's current instructions for your band.
2. Where the sensor sits. Wrist, forearm or upper arm, and how snug it is. Sensor placement affects signal quality, and a placement study notes that moderate skin pressure tends to help while too little causes weak contact and too much can disrupt the signal.

If your band is a different generation from the one tested, the findings below are context, not a verdict on your band.

Step 2: Read what each study actually tested

The table compares three studies, recording only details from the study text. Each row describes the specific setup in its linked study.

Several things stand out.

• Sleep is not exercise. The sleep result applies to heart rate during slow-wave sleep episodes only. It says nothing directly about burpees or intervals, and it should not be stretched to other sleep stages, newer hardware or other settings.
• Placement was compared in only one study. Three identical bands were worn at three sites at the same time, so that study offers a same-hardware, simultaneous comparison under its tested conditions. It points to a site effect, but it does not isolate every possible cause, such as unit-to-unit or fit differences.
• The reference changes the meaning. ECG is the yardstick in two of the studies. The placement study used a chest strap, which is a strong comparator but a different one.
• Upper-arm results describe that tested band. They do not tell you how any other band behaves on the upper arm, and they are not a reason to move a band that is designed for the wrist.

Step 3: Check the activity and the reference

Match your own use against the table before trusting a number.

Steady versus changing effort. The placement authors found accuracy was highest at rest and declined as movement intensity increased. Burpees produced the lowest accuracy across devices (concordance correlation coefficient below 0.50), which the authors attribute to motion artifacts in short, high-intensity, full-body work. The transient-state study found performance fell across all wrist-worn devices when heart rate changed rapidly, and that larger averaging windows improved accuracy. Its authors suggest wrist devices suit average and trend monitoring better than capturing acute dynamics. For a home strength session with hard sets and short rests, that is the closest parallel in this evidence, though no included study tested strength training.

Which statistic was used. Check whether a result reports correlation, bias, error or limits of agreement. A high correlation tells you two traces move together. It doesn't tell you they match in value. Bias and limits of agreement address how far apart the readings are. The sleep study contextualized its agreement against smallest worthwhile change and coefficient of variation, which is a stricter question than "do they correlate?"

What the reference can see. A study that compares second-by-second readings behaves differently from one that averages over 60 seconds. The transient-state study evaluated per-second, 10-second and 60-second resolutions, and larger windows improved accuracy. A device can look better simply because the analysis smoothed the data.

Step 4: An illustrative evidence-reading example

The scenario below is invented to show the reasoning. It is not from an actual test.

Suppose you wear a current-generation wrist band and mostly do interval-style strength work. A friend says, "This band is very accurate, a study said so."

1. Input: The claim cites the slow-wave sleep study.
2. Check version: That study tested an earlier generation, so it does not describe your band.
3. Check activity: The data came from slow-wave sleep, not intervals.
4. Check location: That study used a wrist-based sensor and did not compare body sites.
5. Output: The study supports "this earlier band agreed well with ECG during slow-wave sleep." It does not support "my wrist reading is accurate during intervals."

Then look for evidence closer to your case. The placement and transient-state studies are nearer, but their activities (burpees, treadmill protocol, walking) still differ from your session. The honest conclusion is partial support, with a gap you can name.

Step 5: Judge your own comparison fairly

If you compare your band against a chest strap or another watch, keep these cautions in mind.

• Same workload does not mean the same heart rate. Two sessions with identical effort can have different true heart rates. A logged difference supports a hypothesis such as a loose band or a hard transition. It does not prove an artifact versus real physiology.
• Pair the data in time. Compare matching intervals, not session summaries.
• Don't over-read a single disagreement. The studies above report group-level agreement. They don't predict any one workout.

Apply the Evidence to Your Own Setup

The studies above provide a framework for evaluating heart-rate accuracy, but their findings apply only to the tested devices, wear locations, activities, and reference methods. To assess your own setup, wear the device consistently, compare time-matched readings against an appropriate reference when possible, and treat results from rapid transitions with caution. If your device generation or activity differs from those in the studies, record the evidence as limited rather than assuming the same accuracy.

Acceptance criteria for a claim you'll rely on

Accept a heart-rate accuracy claim for your own use only when you can answer yes to all of these:

1. The study names the same hardware generation as your band.
2. The wear site matches yours, or the study shows the site effect.
3. The activity resembles what you do.
4. The reference is named (ECG, chest strap) and appropriate.
5. The report gives agreement measures such as bias and limits of agreement, not correlation alone.
6. The averaging window matches how you use the data (trend versus moment-to-moment).

If one fails, downgrade the claim from "accurate" to "supported only for what was tested."

Checklist: scenarios the cited evidence does not cover

The cited studies do not establish any of these scenarios:

• Heart rate during resistance or strength training with loaded sets and rest intervals.
• Wrist, forearm and upper-arm differences in sleep, since the placement study covered rest and exercise.
• The slow-wave-sleep findings applied to newer generations or to other sleep stages.
• Fit differences, such as band tightness, which these studies did not isolate.
• Long free-living days, rather than 20-minute or short lab protocols.
• Populations beyond the samples described, such as the 28 young healthy adults in the placement study.
• Any generation other than the one named in each study.

Where a scenario is on this list, the sensible stance is "unverified," not "inaccurate" and not "accurate." It also means no study here supports calling any brand the most accurate. For readers who use heart rate to guide intensity, the site's heart rate zones guide explains how zones are defined. Zones depend on a maximum heart rate, so a reading error can move you across a boundary.