Heart Rate

Heart Rate

Heart rate is the most widely tracked metric in endurance sport and the most frequently misunderstood. Every sports watch records it. Training Load, Body Battery, readiness scores, stress tracking and HRV status are all derived from it.

The data serves three distinct purposes:

  • heart rate paired with GPS is the foundation of training;
  • heart rate variability is the foundation of wellness and of the advanced physiological metrics built on it; and
  • the electrocardiogram (ECG/EKG) is for medical use.

Each purpose requires a different sensor, a different accuracy standard and a different measurement protocol. Most disappointment with a device comes from applying the standard of one purpose to a sensor built for another.

What Your Watch Is Actually Measuring

Three technologies measure the heart, each measuring different physical events.

  • Optical sensors, built into nearly every wrist device, use photoplethysmography. The watch shines light into the skin, and a photodetector reads how much returns. Blood volume rises and falls with each pulse, the reflected light changes with it, and the device infers a pulse rate from that fluctuation. It measures a blood volume pulse. It does not measure an electrical signal.
  • A chest strap measures the timing of the heart’s electrical activity directly, through electrodes held against the skin. It does not store a full diagnostic waveform, but it captures the timing of every beat, the R-R interval, with millisecond precision. That timing matches the electrical event recorded by a hospital ECG.
  • An ECG records the complete electrical waveform, the shape of the beat and not only its timing. Waveform shape is what allows a rhythm abnormality to be identified, rather than a beat simply counted.

The property that separates these is resolution. A chest strap captures beat timing with millisecond precision, while optical sensors operate at a lower effective temporal resolution and with much heavier filtering. By the time the heart rate reaches a standard FIT activity file, it is usually reduced to one recorded value per second (max of 5). The fine detail, the small variation from one beat to the next, is lost twice over: first at the optical sensor, which never captured it cleanly, and again at the file, which was never designed to store it. This runs through everything below. A sensor can be accurate enough to tell you your heart rate, and nowhere near accurate enough to tell you the variation between beats.

How and Where to Wear a Sensor

A large factor in accuracy is how you wear your sensor.

An optical sensor needs firm, stable contact with the skin. Avoid these

  • A loose strap lets the sensor lift and ambient light leak in under it.
  • Positioned over the wrist bone rather than a finger’s width up the forearm, contact is compromised.
  • Cold skin at the start of a session carries less blood near the surface, and the reading suffers until the body warms.

Two failure modes deserve to be recognised by name.

  • The first is cadence lock, where the sensor mistakes the steady rhythm of running or pedalling for the pulse and reports a stable, plausible but incorrect number that happens to match step rate.
  • The second is the grip artefact: gripping a handlebar flexes the wrist and restricts blood flow beneath the sensor, which is the main reason why wrist-based optical heart rate is the least reliable on the bike.

Ranked for accuracy, a chest strap is the most dependable, an upper-arm optical band is a strong middle option that removes most wrist-specific problems, and a wrist sensor is adequate for steady running and weakest for high-intensity intervals. Modern multi-wavelength wrist sensors have narrowed the gap for steady-state running, but the hierarchy holds. Clothing-integrated optical sensors at positions such as the calf and glute have also been tested, with mixed results that depend heavily on garment fit and activity type. For most runners, most of the time, the wrist is fine. For interval sessions, hard cycling and any data feeding a downstream calculation, it is not.

Heart Rate for Sport: The Training Foundation

Heart rate, with GPS, is the bedrock of every sports watch, and for training, its strength is that the accuracy bar is comparatively low. Steady-state work needs a sound average heart rate, and an optical sensor delivers that well for running below threshold. The difficulty is intervals and sprints, where optical accuracy falls away and where heart rate itself lags the effort, unable to rise or settle quickly enough to mark a short, sharp repetition. For that work, pace or power is the better guide to intensity, and heart rate is the slower confirming signal.

The central application is training zones. Whether five zones, the polarised training model associated with Stephen Seiler’s endurance research, or zones anchored to ventilatory thresholds, the principle is the same: matching effort to a physiological band rather than a number. Zones are the most durable and useful outputs of a heart rate sensor, and they are forgiving of sensor noise. What they are not forgiving of is a wrongly set maximum heart rate. A max value taken from the old 220-minus-age formula, rather than measured, will misplace every zone above it, and that error damages training quality far more than any sensor imprecision. The author recommends a similar method fr zone determination based on LTHR (Lactate Threshold, LT2), which some brands automatically detect.

Heart rate also underpins training load. TRIMP (training impulse) combines time spent with heart rate to produce a single figure representing the cardiovascular cost of a session; the figure is weighted higher when the athlete performs in higher heart rate zones. Garmin’s Training Status and the wider Firstbeat physiology stack rely heavily on heart-rate-derived load metrics like these. This is also where heart rate stops being only an intensity gauge: held against pace or power, it measures efficiency. Aerobic decoupling, the drift of heart rate upward relative to a steady pace or power over a long effort, is a recognised marker of both aerobic fitness and the onset of fatigue. Heart rate is a poor guide in one area in particular, resistance training, where it badly misrepresents muscular effort, merely reflecting the cardiovascular load. Velocity-based training exists because heart rate cannot do that job.

Heart Rate Variability: The Wellness Layer

Heart rate variability is a mathematical measure derived from small, continuous variations in the time between heartbeats. It is the foundational signal for wellness tracking, and it is widely misunderstood. HRV is not a direct measure of recovery or readiness. It measures the balance of the autonomic nervous system, how the body is coping with the total load placed on it, training included, but also sleep, illness, alcohol and stress. It is linked to readiness but is not readiness itself. For a deeper look at how this data translates into training decisions, see the full recovery tracking guide. The figure consumer devices report is almost always rMSSD, a time-domain measure; the frequency-domain measures, the high and low frequency bands, belong mostly to research and clinical contexts. This is one reason HRV numbers do not transfer between apps. (Note: Apple used SDNN)

The accuracy problem is real and well-documented. Independent studies, including a validation across 62 Garmin devices against a clinical ECG and a separate study of the Forerunner 265, have found wrist-derived HRV to be insufficiently reliable for clinical or research-grade interpretation. Because Garmin’s wider physiology stack, Body Battery and stress, among others, are built on HRV inputs, those derived metrics inherit the limitation. They are most useful when read as trends rather than precise values.

Reading HRV: When You Measure Changes, What You Learn

HRV means different things at different times of day, and the timing is not simply an inconvenient detail. A waking measurement, taken upon rising in a seated or standing position, is the research-grade protocol that apps such as HRV4Training are built around. It is the most controlled reading and the most useful for tracking adaptation over weeks; it also rewards a chest strap. An overnight measurement is what Garmin, Oura, and Whoop automatically take during sleep. It is convenient and asks nothing of the user, but it is influenced by sleep stages and by the accuracy ceiling of a wrist sensor worn through the night. Furthermore, overnight measurements reflect resting physiology, not exercise physiology; they cannot explain athletic readiness. A daytime measurement, taken as spot checks or continuously, sits behind stress scores. It is the weakest application, the most affected by movement, posture, and many other daily stressors. It is the least suited to firm conclusions.

HRV During Exercise: DFA Alpha 1 and DDFA

This is a complex topic, but the essence is that training zones are, in reality, dynamic during a workout, and these techniques identify zones at any given moment.

A newer use of HRV is to measure it during exercise rather than at rest, to find training thresholds without a laboratory. DFA alpha 1 uses a fractal property of the beat-to-beat signal: values around 0.75 are commonly associated with the aerobic threshold, and around 0.5 with the anaerobic threshold. The method has shown promising agreement with ventilatory and lactate thresholds in several studies. Its weaknesses are that the fixed 0.75 does not hold for everyone and that the figure is sensitive to breathing rate. DDFA, which Suunto implements as ZoneSense, is an evolution of that method, not the same thing. It replaces the fixed aerobic-threshold value with a baseline individualised to the athlete, a direct response to that criticism. Both methods need a high-quality beat-to-beat signal, which in practice means a Polar H10 or an equivalent chest strap. Neither works reliably from the wrist.

ECG: The Medical Layer

The ECG label covers three different things.

  • A single-lead, manual snapshot ECG, the type in an Apple Watch, a Garmin Fenix or an Amazfit watch, is taken on demand over about 30 seconds and is designed mainly to detect atrial fibrillation. Even within that narrow role, it is designed to flag possible atrial fibrillation for follow-up rather than provide a diagnosis. It does not detect blocked arteries, structural heart disease, or most of the conditions that underlie sudden cardiac events in athletes.
  • A continuous ECG chest strap, the Fourth Frontier X2 being the leading example, records the waveform throughout an activity, including during hard exercise, when many cardiac signs that a resting snapshot misses actually appear.
  • A clinical 12-lead ECG is the diagnostic standard and cannot be migrated to a wearable. A 12-lead recording requires electrodes to create 12 electrical circuits, placed at specific points around the chest and on the limbs. That is a question of body geometry, and no wrist watch will ever resolve it. A wearable ECG is a genuine and worthwhile screening tool. It is not a substitute for clinical testing, and it is honest to be clear about where the line falls.

Where Heart Rate Technology Is Heading

The core thrusts are towards medical adjacency (making medical-grade measurements available to consumers), increased optical accuracy, and richer metrics derived during sport.

Several developments are visible and emerging.

  • Multi-wavelength optical sensors include longer wavelengths alongside the standard green, improving readings during motion and on darker skin tones, where green-only sensors have been weakest. Higher sample rates in stored files would close part of the data-loss gap described above.
  • Chest straps are acquiring new measurements, with the Tymewear VitalPro adding a ventilation sensor, a sign of where the category moves next. And continuous ECG continues to shrink toward something a person will wear without thinking.

The direction of travel is consistent: more of the signal captured, more of it kept, and more of it turned into something a watch can act on.

Further Reading

Sensing, Accuracy and Wear Position

Heart Rate for Sport

Heart Rate Variability

HRV During Exercise

ECG and Cardiac Monitoring