You are here

Article Text

Article menu

Download PDFPDF

Original article
Polar Activity Watch 200: a new device to accurately assess energy expenditure
  1. J V Brugniaux1,
  2. A Niva2,3,
  3. I Pulkkinen2,
  4. R M T Laukkanen2,3,
  5. J-P Richalet1,
  6. A P Pichon1
  1. 1Laboratoire Réponses cellulaires et fonctionnelles à l’hypoxie, ARPE, Université Paris, Bobigny, France
  2. 2Polar Electro Oy, Professorintie 5, Kempele, Finland
  3. 3Department of Medicine, University of Oulu, Oulu, Finland
  1. Correspondence to Julien V Brugniaux, UFR SMBH, 74 rue Marcel Cachin, 93017 Bobigny cedex, France; jbrugniaux{at}free.fr

Abstract

Objectives Energy expenditure (EE) based on movement detection is calculated by a new device, the Activity Watch 200 (AW200). The aim of this study was to validate EE measured by this device against indirect calorimetry (IC) and to assess the reproducibility of AW200 measurements.

Design EE was assessed during a 9.7 km hike. 10 men and 10 women in the age range 35–45 years, and 5 men and 6 women in the age range 50–55 years were tested. One in five participants of each age- and sex-matched group was equipped with a portable metabograph (Oxycon Mobil) for IC measurements. Data were collected every 30 min during the hike, and IC was extrapolated for the remaining four other participants of the group.

Results During the total hike, there was a high correlation between EE obtained from the AW200 and the IC calculation (r = 0.987, p<0.001). Identical values of EE were calculated by both methods during the first 90 min of the hike. However, EE calculated by the AW200 at 120 min and at the end of the hike was lower (p<0.05). Bland–Altman analysis showed limits of agreements between 105 and 279 kJ after 30 and 120 min, respectively. EE measured by the AW200 was well correlated with IC measurements, and limits of agreement between devices were below 10% of the measured values for hike durations longer than 60 min.

Conclusion The AW200 appears to be a very useful and accurate device for measuring EE during exercise in recreational hikers and provides a useful tool for keeping track of personal EE.

https://doi.org/10.1136/bjsm.2007.045575

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Chronic diseases and disabilities account for most of the illness among residents in developed countries. Many of these contemporary health problems, including obesity, Parkinson disease, coronary heart disease and osteoporosis, appear to be associated with our lifestyle, including low level of physical activity and energy expenditure (EE).1,–,6 Since total physical activity—that is, the amount of physical activity performed daily during both work and leisure time, appears to be a critical factor for achieving adequate amounts of daily EE for health, the evaluation of EE in field conditions has been of increasing interest.

    Because of its convenience, indirect calorimetry (IC) has been more frequently used over the past few decades.7 It is a method that allows for noninvasive measurement of EE and substrate utilisation in humans.8 Recently, the advancement of various mathematical algorithms along with the diminished size of electronics has permitted the commercial development of small devicesfor approximating EEsuchaswatches. These products do not calculate the EE as does an indirect calorimeter, but provide an accurate estimation, owing to the continuous monitoring of heart rate (HR) during the activity.9 Because of the market interest in such devices, many companies are attempting to simplify the use and to improve the convenience of their products.10,–,12 The Polar Electro Oy Company has developed a new device, the Activity Watch 200 (AW200),which,among otherfunctions,calculates EE. The novelty of this product resides in the development of a new algorithm in which HR monitoring is not necessary to assess EE. The purpose of this study was to verify the accuracy of this new device.

    We aimed to validate EE calculation by the AW200 during a field experiment (9.7-km hike) against IC, assessed by an already validated portable metabograph (Oxycon Mobil). In conjunction with validating the AW200, we also assessed the interchangeability of these devices.13

    Methods

    The AW200 is a new “Activity Watch” device designed by Polar Electro Oy. The AW200 measures the user’s EE through an accelerometer (arm activity) and does not include any chest belt to measure HR. This constitutes the main innovation of this product. Besides EE, the AW200 also measures the altitude, the barometric pressure, the maximal altitude reached and the total ascent and descent during the exercise. The AW200 measures EE based on body activity, altitude changes and user information (weight, height, age, sex). The AW200 also gives the duration of the activity and the active time portion of the total time. This active time is presented as the amount of time spent in five activity zones. Finally, the AW200 has a pedometer which calculates the number of steps.

    Participants

    In total, 31 participants were included in the study: 10 men and 10 women in the age range 35–45 years and 5 men and 6 women in the age range 50–55 years. There were several inclusion criteria: the participants had to be regular hikers, non-smokers, healthy and free of medication. The participant characteristics are presented in table 1. The reproducibility of the measurements was also assessed in the two experimenters participating in all of the experimental sessions (table 1). Both wore the AW200 and reported the total EE measured at the end of each hike by the device. All participants gave their informed consent to participate in the study.

    View this table:
    Table 1

    Participant characteristics Group Age (years)

    Measurements

    The hiking measurements were completed over 4 days with similar climatic conditions from the end of June to early July 2006. Each group covered the same round trip hike of approximately 9.7 km at varying terrains in southern Ile de France. The mean duration was 2 h and 15 min (approximately 1 h and 50 min of active time). The altitude at departure/arrival location was 45 m above sea level, with a maximal hike altitude of 95 m. A short break was taken 1 h after the departure for each group. There were three main ascents of about 30 to 40 m over the hike (two of the three were before the break). The total ascent and descent during the hike amounted to 130 m. About 60% of the time was spent on relatively flat ground, 20% for ascent and 20% for descent.

    All participants were also equipped with an HR monitor (Polar S625X, Polar Electro Oy, Kempele, Finland) and an accelerometer (Omron Walking Style II, Omron Healthcare, Kyoto, Japan). One of the members of each group was equipped with a portable metabograph (Oxycon Mobil, VIASYS Healthcare, Jäger, Würzburg, Germany). All participants were carrying a backpack. The total weight (ie, backpack alone or backpack and metabograph) was the same (4–5 kg). In every group, all of the participants walked at the same speed (∼5.0 km/h).

    Energy expenditure measurements by IC

    Reference measures of EE were assessed by IC. The participant equipped with the metabograph breathed through a facemask (Hans Rudolph, Kansas City, Missouri, USA). Expired gas flows were measured by a pneumotachograph and analysed breath by breath. The pneumotachograph was calibrated automatically before each session with high and low flows. Oxygen uptake (V̇o2), carbon dioxide production (V̇co2), ventilatory flow (V̇E), HR, and other standard respiratory parameters were monitored continuously and averaged every 15 s. Gas analysers were calibrated with a known gas mixture before each experiment. The data were recorded continuously on an integrated flashcard (128 Mb) and were transferred after each hike to a laptop computer.

    The non-proteic respiratory exchange ratio (RER) and the oxygen uptake were determined in order to calculate EE by IC. We used the energetic equivalent for O2 from the Zuntz table.14 The formula used was:

    EE IC (kJ) = V̇o2 (L/min)×time (min)×Energetic equivalent of O2 for non-proteic RER×4.1865

    We considered that the oxygen uptake for 1 kg of body weight was the same for every participant of the same age range and sex. These values allowed us to calculate the EE of the participants without using the metabograph. These analyses were performed every 30 min during the hike, and the results were compared to the EE obtained with the AW200 device.

    Statistics

    The results are expressed as mean values (SD). Linear regression was used to attempt to draw a line describing the relationship between EE assessed by the AW200 and the calculation from IC in all the participants and in the different groups. Pearson’s table was used to assess the significance of these interindividual relationships. Student t tests were used to assess significant differences between the values obtained with the different devices.

    Bland–Altman analysis was used to study the agreement between EE assessed by the AW200 and the EE calculated by IC from Oxycon Mobil measurement.15 The Bland–Altman method consists of plotting the difference between the more direct and the less direct measurement against their mean for each participant. These differences are averaged to determine mean bias. Mean bias variability is used to calculate the limits of agreement between the two measurements. Confidence intervals (CIs) are determined for the mean bias and for the upper and lower limits of agreement. Bland and Altman developed the following formulas for the CIs needed in the analysis: for 95% CIs, t = 0.05 is the critical value for 5% two-sided tests drawn from tables of t distribution with n21 degree of freedom, where n is the sample size.15 We decided a priori that for the AW200 and the IC to be deemed interchangeable, the acceptable limit of agreement was±10% of the measured EE. p Values<0.05 were considered statistically significant.

    Results

    The results of EE calculation from the two methods were not different during the first 90 min of the test for all participants (table 2). Considering the entire hike, there was a very high correlation between the value of EE obtained from the AW200 and the IC calculation (r = 0.987, p<0.001). However, even if the EE was similarly altered during the entire hike (according to coefficient of correlation), the EE values calculated at 120 min and at the end of the hike were lower for the AW200 compared with the IC measurements (p<0.01). The Bland–Altman analysis showed limits of agreements between 105 and 279 kJ after 30 and 120 min of hiking, respectively (table 3 and fig 1).

    View this table:
    Table 2

    Energy expenditure (kJ) calculated during the hike by the AW200 and by indirect calorimetry for all participants (n = 31)

    View this table:
    Table 3

    Bias, lower and upper limits of agreement and confidence intervals15 calculated for the energy expenditure (kJ) dissipated during the hike by the AW200 and by indirect calorimetry (n = 31)

    Figure 1
    Figure 1

    Bland–Altman plots showing relationship between mean calculated values of energy expenditure (EE) and the difference calculated by AW200 device and indirect calorimetry (IC) after (A) 30 min, (B) 60 min, (C) 90 min and (D) 120 min of hiking (n = 31).

    When we assessed the EE expenditure for each age and sex group, we obtained a similar tendency as for all of the participants. Indeed, the AW200 tends to underestimate the EE compared with the IC calculation from the Oxycon Mobil in all groups. We observed significant differences between EE calculated by the AW200 and by IC at the end of the hike for the women aged between 35 and 45 years and for the men aged between 50 and 55 years (p<0.05) (fig 2).

    Figure 2
    Figure 2

    Energy expenditure (EE) during the hike for the four age and sex groups. (A) men aged 35–45 years (n = 10), (B) men aged 50–55 years (n = 5); (C) women aged 35–45 years (n = 10); (D) women aged 50–55 years (n = 6). *p<0.05 AW200 vs indirect calorimetry (IC).

    The same trend was observed for the average of the five participants equipped with the portable metabograph (fig 3). The correlations between the two methods of EE assessment were significant after 30 min (r2 = 0.89, p = 0.01), 90 min (r2 = 0.90, p = 0.01) and 120 min (r2 = 0.84, p = 0.03), with a similar trend after 60 min (r2 = 0.72, p = 0.08). The lack of significance after 60 min is probably due to our sample size. For all the participants equipped with the metabograph, the EE calculated by the AW200 was lower than the EE obtained using the IC method.

    Figure 3
    Figure 3

    Energy expenditure during the hike measured by the AW200 and by indirect calorimetry for participants wearing the portable metabograph (n = 5).

    We observed variations of less than 5% (2.8% and 4.8%) in all the parameters (time, active time, active steps and calories) measured by the AW200 during eight repeated hikes (one done before experimental hikes and the seven experimental hikes) for the two experimenters.

    Discussion

    This study shows that the new Activity Watch 200 allows an accurate assessment of EE during hiking compared with IC calculation.

    Repeatability of the hikes and reproducibility of the results

    The results displayed in fig 2 demonstrate that the different sessions of hikes performed during the experiment were very comparable and that relative energetic cost of these hikes was the same regardless of the group of individuals studied. Moreover, the results of the two experimenters studied during all the sessions show a high reproducibility with a coefficient of variation less than 5% for EE. These results prove the usefulness of the AW200 as a non-invasive tool to assess EE on an ongoing basis for personal or scientific use.

    Energy expenditure

    The results show a good accordance between the EE expenditure calculated by the AW200 and IC according to the high correlation coefficient obtained (p<0.001). However, the Bland– Altman analysis, which allows assessment of the interchangeability between two methods, and is necessary for validation of the new device,13 reveals large limits of agreement between the AW200 and IC at the beginning of the hike. Nonetheless, as the duration of the hike increased, the limits of agreement decreased and thus were improved progressively with the duration of the hike to reach the limit fixed a priori. Indeed, we considered that a reasonable limit of agreement could be about 10% of the EE measured, and the limits of agreement observed are between 105 kJ, after 30 min, to 279 kJ after 120 min—that is, 14% and 10% of the mean EE values. Therefore, the AW200 could be considered as more interchangeable for a long hike. Conversely, when we studied absolute values of EE, we observed that the difference between the two methods grows with hike duration from 2.5% to 10.5% of the EE measured at the end of the hike. Consequently, it seems that the AW200 device systematically underestimates EE during hike compared with the IC method (figs 2 and 3) even if the interchangeability between the Oxycon Mobil and the AW200 increases with exercise duration.

    Recently, Perret et al16 showed that the Oxycon Mobil underestimates V̇o2 and overestimates RER calculations compared with a laboratory device only for exhaustive exercise on ergocycle and workloads higher than 200 W.16 As walking cannot be considered as an exhaustive activity, the V̇o2 measured by the Oxycon Mobil can obviously not be responsible for the difference of EE values observed between the two devices.

    One could argue that fatigue may have influenced EE calculation from the two devices. The more the fatigue increases, the more the coordination of the motor functions decreases, and this factor could have enhanced the EE calculation by the AW200, since its calculation is based on arm movements.17 However, the decrease in individuals’ efficiency with the duration of the hike “overestimates” oxygen uptake measurements made by the metabograph.18 Moreover, a switch in substrate (from carbohydrate to lipid) could also explain the divergence over time in the EE assessment between the two devices.

    Additionally, experimental sessions were performed in a warm environment (around 30uC). Therefore, it is possible that thermoregulatory adjustments (such as peripheral vasodilatation and rise in cardiac output) occurring during the hike7 could have increased the energetic cost of walking measured by the portable metabograph up to 10%.19 In that case, the overestimation of oxygen uptake during the hike would have increased the EE assessed by IC, whereas this environmental parameter was not taken into account by the AW200 algorithm. Furthermore, each participant wore a backpack, and this extra load was not taken into account in the individual characteristics menu of the AW200 needed to assess EE (height, weight, age). Therefore, the difference between the AW200 and the EE assessed by the IC method might be linked to the extra cost of walking due to the increased load carried by the participants.

    Consequently, the significant difference in EE observed at the end of the hike appears to have multiple potential origins. Another explanation could be that for a given group, the comparison between AW200 and IC measurements was indirect, since only one or two participants of the whole group had V̇o2 recordings and IC measurements. If we look at this relationship between AW200 and IC, made only with the data of the participants wearing the Oxycon Mobil, the significant difference disappears even if the AW200 still underestimated EE (fig 3).

    Usefulness of the AW200

    This new product has been designed more for recreational trekkers than for highly trained athletes. However, we think that this new easy-to-use product may become a very useful tool in rehabilitation programmes thanks to the EE calculation, pedometry and activity zone determination features of the AW200. Recently, Corrà and Giannuzzi stated the therapeutic value of exercise in cardiovascular prevention and rehabilitation.20 They indicated that a slight increase in V̇o2 (1 ml/min/ kg) could be translated into a 9–10% lowering of cardiac mortality. They also insisted on the fact that an accurate control of exercise is crucial. In this scope, the use of a device like the AW200, providing activity time and zones, is relevant. TudorLocke et al have concluded that an exercise could be considered as moderately intense around 3000 steps in 30 min for both sexes.21 These results have been recently confirmed in adolescent girls by Treuth et al.22 Therefore, even if the combination of HR with accelerometer data could furnish better results,23 the AW200, with its pedometry and activity zone features, can also be considered as an opportunity for the users to quickly evaluate their exercise/training.

    This validation study demonstrates that the Polar AW200 is a valid activity watch for hike durations longer than 60 min, suggesting that this easy and non-invasive device is very usefulfor assessing EE in individuals. Indeed, even if the mean differences between the two devices increases with the hike duration, the inter-individual reliability improves with time. Hence, besides its convenience for hikers, this accurate device could constitute a novel and useful tool for the easy control of exercise training during rehabilitation in patients or sedentary individuals.

    What is already known on this topic

    • Various wrist computers calculating EE based on HR measurement are already available.

    • Carefully evaluating exercise is a key point for rehabilitation and more generally for public health.

    What this study adds

    • This study validates a new device calculating EE directly from wrist movement obtained by an accelerometer included into the watch.

    • The AW200 appears to be very useful to appropriately assess exercise over rehabilitation follow-up.

    Acknowledgments

    We wish to thank all the individuals for their participation as well as the members of the MSR-SEBAC company for their assistance.

    References

    1. ↵
      American College of Sports Medicine Position Stand. The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults. Med Sci Sports Exerc 1998;30:975–91.
      OpenUrlPubMedWeb of Science
    2. ↵
      American College of Sports Medicine Position Stand. Exercise and physical activity for older adults. Med Sci Sports Exerc 1998;30:992–1008.
      OpenUrlPubMedWeb of Science
    3. ↵
      1. Crizzle AM,
      2. Newhouse IJ
      . Is physical exercise beneficial for persons with Parkinson’s disease? Clin J Sport Med 2006;16:422–5.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Jakicic JM,
      2. Otto AD
      . Treatment and prevention of obesity: what is the role of exercise? Nutr Rev 2006;64:S57–61.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Suominen H
      . Muscle training for bone strength. Aging Clin Exp Res 2006;18:85–93.
      OpenUrlPubMedWeb of Science
    6. ↵
      1. Warburton DE,
      2. Nicol CW,
      3. Bredin SS
      . Health benefits of physical activity: the evidence. CMAJ 2006;174:801–9.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Ainslie PN,
      2. Campbell IT,
      3. Frayn KN,
      4. et al
      . Physiological and metabolic responses to a hill walk. J Appl Physiol 2002;92:179–87.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Jequiery E,
      2. Felber JP
      . Indirect calorimetry. Baillieres Clin Endocrinol Metab 1987;1:911–35.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Crouter SE,
      2. Albright C,
      3. Bassett DR Jr
      . Accuracy of polar S410 heart rate monitor to estimate energy cost of exercise. Med Sci Sports Exerc 2004;36:1433–9.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Nichols JF,
      2. Morgan CG,
      3. Sarkin JA,
      4. et al
      . Validity, reliability, and calibration of the Tritrac accelerometer as a measure of physical activity. Med Sci Sports Exerc 1999;31:908–12.
      OpenUrlPubMedWeb of Science
    11. ↵
      1. Fruin ML,
      2. Rankin JW
      . Validity of a multi-sensor armband in estimating rest and exercise energy expenditure. Med Sci Sports Exerc 2004;36:1063–9.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. King GA,
      2. Torres N,
      3. Potter C,
      4. et al
      . Comparison of activity monitors to estimate energy cost of treadmill exercise. Med Sci Sports Exerc 2004;36:1244–51.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Bland JM,
      2. Altman DJ
      . Regression analysis. Lancet 1986;1:908–9.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Zuntz N
      . Ueber die bedeutung des verschiedenen nährstoffe als erzuger des muskelkraft. Pflugers Arch Physiol 1901;83:557–71.
      OpenUrlCrossRef
    15. ↵
      1. Bland JM,
      2. Altman DG
      . Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Perret C,
      2. Mueller G
      . Validation of a new portable ergospirometric device (Oxycon Mobile) during exercise. Int J Sports Med 2006;27:363–7.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Apriantono T,
      2. Nunome H,
      3. Ikegami Y,
      4. et al
      . The effect of muscle fatigue on instep kicking kinetics and kinematics in association football. J Sports Sci 2006;24:951–60.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Kalis JK,
      2. Freund BJ,
      3. Joyner MJ,
      4. et al
      . Effect of beta-blockade on the drift in O2 consumption during prolonged exercise. J Appl Physiol 1988;64:753–8.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Yanagimoto S,
      2. Kuwahara T,
      3. Zhang Y,
      4. et al
      . Intensity-dependent thermoregulatory responses at the onset of dynamic exercise in mildly heated humans. Am J Physiol Regul Integr Comp Physiol 2003;285:R200–7.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Corrà U,
      2. Giannuzzi P
      . Role of cardiopulmonary exercise testing in today’s cardiovascular prevention and rehabilitation. Eur J Cardiovasc Prev Rehabil 2006;13:473–4.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Tudor-Locke C,
      2. Sisson SB,
      3. Collova T,
      4. et al
      . Pedometer-determined step count guidelines for classifying walking intensity in a young ostensibly healthy population. Can J Appl Physiol 2005;30:666–76.
      OpenUrlPubMedWeb of Science
    22. ↵
      1. Treuth MS,
      2. Schmitz K,
      3. Catellier DJ,
      4. et al
      . Defining accelerometer thresholds for activity intensities in adolescent girls. Med Sci Sports Exerc 2004;36:1259–66.
      OpenUrlPubMedWeb of Science
    23. ↵
      1. Brage S,
      2. Brage N,
      3. Franks PW,
      4. et al
      . Branched equation modeling of simultaneous accelerometry and heart rate monitoring improves estimate of directly measured physical activity energy expenditure. J Appl Physiol 2004;96:343–51.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Funding This study was supported by Polar Electro Oy, Finland.

    • Competing interests None.