Journal Logo

Measurement of Moderate Physical Activity: Advances in Assessment Techniques

Introduction: evaluation of some measurements of physical activity and energy expenditure


Author Information
Medicine & Science in Sports & Exercise: September 2000 - Volume 32 - Issue 9 - p S439-S441
  • Free

The notion that regular exercise may prevent or delay the onset or progress of disease is not new. Until fairly recently, however, few scientific studies investigated this hypothesis. Two developments in modern life have increased interest in this area as a topic for investigation. First, advances in technology have altered most of our occupations and modes of transportation so that we are now required to expend less energy in these daily activities. Therefore, if a sedentary life contributes to ill health, the evidence should be more manifest now than ever before. Second, chronic degenerative diseases have replaced many infectious and contagious diseases as causes of death and disability. It is logical to suppose that our lifestyles, including physical activity, are more closely associated with some of these chronic diseases than with those diseases that were formerly responsible for much ill health. Of course, there are other reasons for measuring physical activity (PA) or energy expenditure (EE) in individuals in a population besides the potential value of physical activity in the maintenance of health. Nevertheless, it is the latter association that has commanded most research interest. Thus, it is this application that should be kept in mind in the discussion of techniques for assessing PA and EE that follows.

With the greater emphasis on studies of the relationship of activity to health has come the need for more accurate and reliable methods of estimating PA and EE in the field. The papers that follow make significant contributions toward fulfilling that need.

Physical activity is a complex phenomenon characterized by its intensity (rate of EE), duration of one session, frequency (per hour, per day, etc.), and surrounding environmental and social conditions. With this in mind, which of these characteristics might be most closely related to health? Is a life history of PA important? Do daily or seasonal variations in PA influence health? The search for answers to these and similar questions affects the measurement methods to be selected for a particular study. Also, the size and other characteristics of the study population dictate, to some extent, the method of choice for estimating PA or EE. The direct measurement of EE by heat production or the indirect method by oxygen consumption, until recent years, has been confined to the laboratory. The need for a portable method of measuring oxygen consumption has long been recognized. Zuntz and Leowy (14) were probably the first to design such an apparatus. Improved devices by Kofranyi and Michaelis (7), Müller and Franz (9), Wolff (12), and Humphrey and Wolff (5) followed. These instruments were cumbersome and had limitations in accuracy. A new generation of portable equipment for measuring EE that overcomes to a considerable extent these limitations is now available; hence, the evaluation of one of these portable metabolic systems by King et al. (6) is noteworthy. Granted, because of the cost of measurements of this kind, their usefulness is confined to small populations. Also, the measurements provide data only for a limited period of time. Nevertheless, such measurements are useful as criteria for evaluating other methods.

A variety of methods for estimating PA and EE in the field are available. These include energy consumption, biomechanical techniques, observation and time/motion analyses, diaries, questionnaires, interviews, recording of physiologic response to activity, portable monitoring devices such as pedometers and accelerometers, and doubly labeled water. All of these have limitations but have been found to be useful in particular circumstances. A discussion of these methods can be found elsewhere (8).

Heart rate, ventilation, blood pressure, electromyographs, and body temperature are all roughly related to EE, and it is possible to record these physiological functions by portable equipment worn by a subject. Of these, heart rate is most easily recorded and has been found useful in estimating EE in some populations. Nevertheless, there are limitations to these techniques as well. Heart rate is affected by factors other than EE, for example, emotional and environmental conditions and training state. Strath and colleagues in a report that follows propose a new approach in the use of monitored heart rates, and their method appears to improve estimates of EE.

All of the methods mentioned above are limited in accuracy and/or feasibility for estimating PA or EE in particular populations or circumstances. As a result, there has long been a need for a simple, inexpensive device that could be worn by a subject for estimating PA or EE. The first of these took the form of a pedometer that counted the steps walked. According to the World Almanac of Presidential Facts, Thomas Jefferson, the third president of the United States, invented the pedometer. However, the first pedometer was probably designed by Leonardo da Vinci about 500 yr ago (3). In the museum at Leiden, The Netherlands, one can see a pedometer used in the 17th century to count steps, apparently to measure plots of land. Another pedometer was developed by the French physiologist Etienne-Jules Marey more than 100 yr ago (1). Modern technology has produced much more sophisticated pedometers. The report in this supplement by Hendelman et al. provides data on the validity of one of these pedometers for estimating EE.

Habitual physical activity, of course, is not limited to walking. A pedometer likely will not accurately register EE during activity not involving the lower limbs. Also, the impact of the feet on the ground or floor surface affects estimates of EE. Because of these and other limitations of pedometers, the search continues for portable devices that more accurately reflect EE in a variety of activities. Laboratory studies have indicated that the acceleration of body mass and/or limbs is related to EE (8, p. 79). Apparently, the first single-plane portable accelerometers used to estimate EE in human beings consisted of modified self-winding watches (4,10). Because we had difficulty with standardization of a self-winding watch, our first attempts at developing a single-plane portable accelerometer employed a modified monaural phonocartridge (13). A later version used a bender element consisting of two layers of piezoelectric material and a brass center layer as the transducer for recording acceleration (11). This device eventually developed into the Caltrac.

Inasmuch as these devices use single-plane accelerometers, it was thought that a three-plane accelerometer apparatus might more accurately reflect EE. To study this, a Caltrac was mounted in each of three planes on the body and the outputs integrated (2). This resulted in a small improvement in estimating EE compared with a single-dimension accelerometer. These data were provided to the company that produced the Caltrac. The Tritrac (a three-dimensional accelerometer device) was later produced by this company. This instrument and other accelerometers were used to estimate EE in papers in this supplement. In addition to providing validity and reliability of information for these accelerometers, the studies provide valuable data on the energy cost of various physical activities. This is possible because oxygen consumption during the activities was measured. These data are important for updating lists of energy costs of physical activities used in other methods of assessing EE, namely, diaries, questionnaires, and interviews.

There are instances in which a questionnaire, diary, or interview is the only feasible method for assessing habitual PA. To convert data obtained in these ways to estimate EE, a table of energy costs for various activities is needed. Such lists have appeared in the literature from time to time. Unfortunately, data for many activities have been derived from populations limited in 1) numbers of subjects, 2) gender (including subjects of only one gender), and 3) range (including only a narrow age range of subjects). Additionally, no data were available for some activities. However, energy cost data for particular activities continue to appear in the literature. In one of their papers in this supplement, Ainsworth and colleagues provide an updated compendium of energy costs, a much-needed addition to the literature.

The Centers for Disease Control and Prevention, the American College of Sports Medicine, and the U.S. Surgeon General’s Office have all made recommendations for the amount of moderate or vigorous exercise needed to maintain a healthy lifestyle. Methods are required to accurately assess the frequency and duration of exercise of these kinds in individuals. To this end, Ainsworth and coworkers, in a paper in this supplement, compare three techniques for estimating the amounts of light, moderate, and strenuous activity engaged in by members of an adult population. If there is good agreement in the results using various methods, we can experience a measure of confidence that the assessments are accurate. If, however, there is poor agreement, it is difficult to ascertain which method or methods are valid.

In summary, the research studies reported in this supplement increase our understanding of the use of several approaches in the measurement of PA and EE. They provide new ways of utilizing outputs of the devices and extend the body of information concerning their validity and reliability. Together, these papers represent a valuable addition to our knowledge of field measurements of PA and EE.


1. Amar, J. The Human Motor. New York: E. P. Dutton, 1920, p. 282.
2. Ayen, T. G., and H. J. Montoye. Estimation of energy expenditure with a simulated three-dimensional accelerometer. J. Ambul. Monitor. 1: 293–301, 1988.
3. Gibbs-Smith, C. The Inventions of Leonardo da Vinci. London: Phaidon Press, 1978, p. 84.
4. Groenewegen, J. D. Adiposity, activity and energy Intake. Doctoral dissertation. University of Leiden, Rotterdam, 1979.
5. Humphrey, S. J. E., and H. S. Wolff. The Oxylog (Abstract). J. Physiol. 267: 120, 1977.
6. King, G. A., J. E. McLaughlin, E. T. Howley, D. R. Bassett, and B. E. Ainsworth. Validation of Aerosport KB1-C portable metabolic system. Int. J. Sports Med. 20: 304–308, 1999.
7. Kofranyi, E., and H. F. Michaelis. A portable apparatus to determine metabolism (Ein tragbärer Apparat zur Bestimmung des Gasstoffwechsels). Arbeitsphysiologie 11: 148–150, 1940.
8. Montoye, H. J., H. C. G. Kemper, W. H. M. Saris, and R. A. Washburn. Measuring Physical Activity and Energy Expenditure. Champaign, IL: Human Kinetics Publishers, 1996, pp. 17–115.
9. Müller, E. A., and H. Franz. The measurement of energy consumption during occupational work with an improved respirometer (Energieverbrauchsmessungen bei beruflicher Arbeit mit einer verbesserten Respirationsgasuhr). Arbeitsphysiologie 14: 499–504, 1952.
10. Schulman, J. L., and J. M. Reisman. An objective measure of hyperactivity. Am. J. Ment. Def. 64: 455–456, 1959.
11. Servais, S. B., J. G. Webster, and H. J. Montoye. Estimating human energy expenditure using an accelerometer device. IEEE Frontiers of Engineering in Health Care 4: 309–312, 1982.
12. Wolff, H. S. The integrating pneumotachograph: a new instrument for the measurement of energy expenditure. Q. J. Exerc. Physiol. 43: 270–283, 1958.
13. Wong, T. C., J. G. Webster, H. J. Montoye, and R. Washburn. Portable accelerometer device for measuring human energy expenditure. IEEE Trans. Biomed. Eng. BME- 28: 467–472, 1981.
14. Zuntz, N., and A. Leowy. Textbook of Human Physiology(Lehrbuch der Physiologie des Menschen). Leipzig: Germany: F. C. W. Vogel, 1909, p. 141.
© 2000 Lippincott Williams & Wilkins, Inc.