Validity of sports watches when estimating energy expenditure during running

Roos et al. BMC Sports Science, Medicine and Rehabilitation (2017) 9:22 DOI 10.1186/s13102-017-0089-6 RESEARCH ARTICLE Open Access Validity of sports watches when estimating energy expenditure during running Lilian Roos1,2* , Wolfgang Taube2, Nadja Beeler1 and Thomas Wyss1 Abstract Background: The aim of this study was to assess the accuracy of three different sport watches in estimating energy expenditure during aerobic and anaerobic running. Methods: Twenty trained subjects ran at different intensities while wearing three commercial sport watches (Suunto Ambit2, Garmin Forerunner920XT, and Polar V800). Indirect calorimetry was used as the criterion measure for assessing energy expenditure. Different formulas were applied to compute energy expenditure from the gas exchange values for aerobic and anaerobic running. Results: The accuracy of the energy expenditure estimations was intensity-dependent for all tested watches. During aerobic running (4–11 km/h), mean absolute percentage error values of −25.16% to +38.09% were observed, with the Polar V800 performing most accurately (stage 1: −12.20%, stage 2: −3.61%, and stage 3: −4.29%). The Garmin Forerunner920XT significantly underestimated energy expenditure during the slowest stage (stage 1: −25.16%), whereas, the Suunto Ambit2 significantly overestimated energy expenditure during the two slowest stages (stage 1: 38.09%, stage 2: 36.29%). During anaerobic running (14–17 km/h), all three watches significantly underestimated energy expenditure by −21.62% to −49.30%. Therefore, the error in estimating energy expenditure systematically increased as the anaerobic running speed increased. Conclusions: To estimate energy expenditure during aerobic running, the Polar V800 is recommended. By contrast, the other two watches either significantly overestimated or underestimated energy expenditure during most running intensities. The energy expenditure estimations generated during anaerobic exercises revealed large measurement errors in all tested sport watches. Therefore, the algorithms for estimating energy expenditure during intense activities must be improved before they can be used to monitor energy expenditure during high-intensity physical activities. Keywords: Wearables, High-intensity, Maximal accumulated oxygen deficit, Validation, Monitoring training Background The amount of energy spent on a specific activity – commonly known as energy expenditure (EE) – is important not only for athletes but also for patients suffering from obesity or diabetes [1–3]. The term EE is often used with regard to nutrition, sport science, occupational tasks, and athlete training, areas in which it is important to monitor the demands of various physical activities. Especially in clinical nutrition settings (e.g. monitoring the exercise activity of obese people), it is important to use devices that provide accurate EE measurements as these measurements are crucial in determining the amount of calories * Correspondence: 1 Section for Elite Sport, Swiss Federal Institute of Sport Magglingen SFISM, Hauptstrasse 247, 2532 Magglingen, Switzerland 2 Department of Medicine, Movement and Sport Science, University of Fribourg, Boulevard de Pérolles 90, 1700 Fribourg, Switzerland that a patient can consume without gaining weight [3]. Similarly, active and lean people may be interested in obtaining precise EE data during their training sessions. Therefore, devices that can accurately measure EE are useful. Indirect calorimetry can be performed by using stationary or portable spirometers to measure breath-bybreath gas exchange, which in turn is analyzed in order to estimate EE. This reference method measures activities performed over a duration of 1–3 h and has been found to be accurate during rest periods and various levels of exercise intensity [4, 5]. Indirect calorimetry is considered the most feasible method for attaining accurate data for short-term physical activity in a laboratory setting [6]. Another option is to estimate EE using heart rate (HR) data, due to the linear relationship © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Roos et al. BMC Sports Science, Medicine and Rehabilitation (2017) 9:22 Page 2 of 8 of oxygen consumption and HR [7]. Previous findings supported HR measurements to be a valid method to assess EE in a laboratory or field setting, EE estimations were even better when using percentage of HR reserve or difference between active and resting HR [8]. When considering different methods for assessing EE, it becomes obvious that there is a trade-off between accuracy, feasibility, and costs [9]. At the same time, factors such as device usability and movement constraints are important to consider. For example, sports watches could constitute the perfect solution as they are userfriendly, relatively low-priced, non-invasive, and can provide other important information during a training session, such as duration, HR, speed, distance and altitude covered [10, 11]. It is important to understand how accurate sports watches are in assessing EE during varying levels of exercise intensity. For researchers to make informed decisions about which products to include in a study or trial. This information is equally relevant for professional and recreational athletes who use the popular sports watches to monitor different variables during their training sessions. However, the accuracy of the newest sports watches (season 2015) in assessing EE is thus far unknown. The companies developing these devices use proprietary algorithms to estimate EE. Generally, these algorithms consider variables such as age, weight, height, sex, maximal heart rate (HRmax), and maximal oxygen uptake (VO2peak) in computing an individual’s EE. A recent study reported that prediction accuracy of EE during running was significantly increased when real-time running speed was included [12]. The newer generation of sports watches also have built-in accelerometers, so it is likely that acceleration data is factored into the algorithm as well. Even some earlier devices from different manufacturers had accelerometers implemented. However, sports watch developers prefer to keep their algorithms secret, and there exists only limited published research regarding the development, validity, and reliability of EE estimation algorithms in sports watches [8, 10, 13], especially with regard to vigorous physical activity and (...truncated)