Research review: WATCH OUT! Does your life depend upon your activity tracker?

Researchers set out to determine the accuracy of assessing heart rate during exercise using two of the most popular wearable devices. Here’s what they found. Review by Dr Mike Climstein PhD & Dr Joe Walsh PhD


  • Wearable technology continues to evolve and to grow in popularity, with a global revenue of around $34 billion
  • Researchers determined the validity of exercise heart rate at different intensities for two popular wearable devices
  • Using a treadmill test, the data collected by the Fitbit Charge and the Apple Watch were compared to that collected by medical-grade electrocardiogram (ECG) monitoring
  • The data collected by the wrist-worn tech was understandably not as accurate as that of the ECG, but accuracy was noticeably improved compared to earlier wearable devices.

Title: Heart rate measures from the Apple Watch, Fitbit Charge HR 2, and electrocardiogram across different exercise intensities

Authors: Dr’s Thompson and colleagues. (Dept of Health and Exercise Science, Colorado State University, USA) Source: Journal of Sports Sciences. 37 (12): 1411-1419.

PREFACE: I would like to congratulate my colleague, friend and co-author of these Research Reviews, Dr Joe Walsh, on completing his PhD. Joe recently attained his qualification at Charles Darwin University in very unique circumstances. His PhD (Physiological and health indices of masters athletes: Incorporating machine learning predictive models based upon psychological motivations) included a very impressive 14 publications, which I’m certain is a record. Congratulations on your great achievement Dr Walsh!

Introduction: I’m pretty fortunate in the teaching I do at Southern Cross University, as it’s all focused on the Master of Clinical Exercise Physiology. One of the units I teach is ‘cardiovascular health’, for which students must learn to conduct clinical graded exercise tests – an important component of which is preparation and monitoring of a 12-lead electrocardiogram (ECG). Now, not only is stress testing and ECG my area of clinical expertise, but I’ve also got state of the art wireless 12-lead ECGs to teach the students on and use in our student-led clinics. As a fitness professional, you’ve probably known someone who has undergone a stress test or had a resting ECG. For our students, it’s critically important that they are able to recognise a number of heart arrhythmias, and that they can properly identify these arrhythmias at rest (as this may preclude the patient from even attempting to complete the stress test) and during exercise and recovery (to ensure there are no arrhythmias that develop during exercise or post-exercise, particularly life threatening arrhythmias, and that no changes occur on the ECG which are indicative of cardiovascular disease). The students must also ensure the haemodynamic (heart rate and blood pressure) responses are normal pre, during and post stress test. We must note that it is an absolute requirement that stress tests (submaximal or maximal) are conducted with ECG monitoring to help ensure the safety of each patient. And hence, ECG monitoring is considered the gold standard. Imagine if you could offer your clients that level of monitoring when they are exercising... if only! And this takes us into this Research Review, where Dr Thompson and her colleagues investigated the accuracy of assessing heart rate during exercise using wearable physical activity monitors. Now, Dr Walsh and I are once again delving into the holy grail of publications, as our 2016 Research Review (Wrist-worn tech: investment or waste of money?) stimulated a huge response (understandable, as a number of our readers found out that their costly device was inaccurate, ouch!). So please, remember that we are only providing factual findings from a published, peer-reviewed scientific study. Let us reiterate again in this Research Review, our 12-lead wireless ECGs cost in excess of $14,000 each, whereas consumer-orientated wearable devices are considerably less expensive, i.e. not even 5% of the price! So, to an extent, you do get what you pay for!

Dr Thomas and her colleagues’ aim was to determine the validity of exercise heart rate at different intensities for two popular wearable devices; the Fitbit Charge 2 and the Apple Watch. These devices were compared to a simultaneous ECG monitoring. The protocol had participants wearing the Fitbit Charge 2 on their left wrist, the Apple Watch on their right wrist and a standard 12-lead ECG. Heart rate readings were taken each minute from each of the devices and ECG during the entire exercise protocol. The exercise protocol consisted of subjects completing a maximal exercise test using the Bruce Protocol treadmill test (Table 1). Heart rate was assessed in the last 10 seconds of every minute on both devices and the 12-lead ECG.

Results: The resting heart rates were similar between genders: males at approximately 70 beats per minute and females slightly lower at 66 beats per minute. The researchers broke the exercise intensities down relative to each participant’s heart rate reserve (HRR: age-predicted heart rate max minus resting heart rate). Intensity was categorised according to the American College of Sports Medicine definitions of very light (<20% HRR), light (20–40% HRR), moderate (40–60% HRR), vigorous (60–85% HRR), and very vigorous (>85% HRR).

As this study had a large number of results, we have chosen here to focus on the ‘Group’ results for each of the intensities, in which the males’ and females’ data was combined into one group.

Specific to each of the wrist-worn devices, the Apple Watch had relative error rates that ranged from 2.4% (overall) to 5.1% (moderate). The Fitbit Charge 2 had relative error rates that ranged from 4.9% (overall) to 13.4% (very vigorous). It should be noted that the overall relative error rate increased in the Fitbit Charge 2 as the intensity of exercise increased (very light 4.9% to very vigorous 13.0%). The authors note that a relative error rate of < 5% defines the acceptable limit.

The authors concluded that the Apple Watch matched the ECG more accurately for all exercise intensities overall, and for each gender. Also, neither the apple Watch or the Fitbit Charge 2 matched the actual HR during the vigorous and very vigorous intensity. The accuracy of the Apple Watch and Fitbit Charge 2 was reduced as the exercise intensity increased. As a clinician, it is encouraging to note that wrist-worn technology has made great improvements in accuracy over just 2 years. However, in instances where a very accurate exercise heart rate is required (i.e. exertional angina) caution must be exercised (excuse the pun) with regard to prescribing exercise, particularly aerobic exercise, in individuals where high exercise heart rates could become life threatening. Pros: We are impressed with this study, it was well designed and conducted. Revenue from wearable technology is predicted to exceed $247 million dollars here in Australia this year alone, with one in five Australians owning a device. Worldwide, this market is expected to exceed $34 billion by next year. Clearly this is not a short-term trend and, as we are seeing, the technology is improving dramatically. These devices are being used for more than exercise heart rates. A friend from Hawaii recently bought the new Apple Watch for his dad, exclusively as it can monitor the heart. Apple claims it captures the heart rhythm (referred to as irregular rhythm notification) when the wearer experiences symptoms (rapid or skipped heart beats), and it can provide this information to the individual’s doctor. No doubt we will see further improvements in this area over the coming years.

Cons: It would have been an interesting extension of this research if they also investigated the accuracy during resistance training.


Thompson, E., Comstock, A., Reinwald, S., Blake, S., Pimentel, R., Tracy, B., & Li, K. (2019). Heart rate measures from the Apple Watch, Fitbit Charge HR 2, and electrocardiogram across different exercise intensities. Journal of Sports Sciences. 37 (12): 1411-1419. Climstein, M., & Walsh, J. (2016). Research Reviews: Wrist-worn tech: Investment of waste of Money? Australian Fitness Network. Spring: 44-45. Pevnick, J., Birkeland, K., Zimmer, R., Elad, Y., & Kedan, I. (2019). Wearable Technology for Cardiology: An Update and Framework for the Future. Trends Cardiovascular Medicine. 28 (2): 144-150.


Dr Climstein is one of Australia’s leading Accredited Exercise Physiologists. He is a faculty member in Clinical Exercise Physiology, Sport & Exercise Science at Southern Cross University (Gold Coast). Dr Joe Walsh, PhD Joe is a sport and exercise scientist. As well as working for Charles Darwin and Bond Universities, he is a director of Fitness Clinic in Five Dock, Sydney.