The most recent physical activity recommendations from public health authorities in the United Kingdom (11), North America (24,25,29), and Australia (7) state that participation in at least 30 min of moderate intensity activity on most days of the week is sufficient to confer health benefits, such as a decreased incidence of coronary heart disease, hypertension, Type 2 diabetes, osteoporosis, colon cancer, anxiety, and depression. These moderate intensity activities, which are assumed to result in energy expenditures of 3.0–6.0 METs, include walking and possibly household and garden activities. Because many people spend substantial amounts of time engaged in these activities, their inclusion or exclusion from measures of physical activity participation will result in substantially different estimates of the prevalence of adequate physical activity in the population. For example, Active Australia (a federal coalition with a charter to increase physical activity levels in the community) conducted a random telephone survey in November and December 1997 (N = 4824) and while 56% of adults were classified as sufficiently active for health benefits based on standard leisure time activity items, this figure rose to 71% if household and garden activities were included (4). The nature and scope of health promotion strategies related to physical activity will therefore alter substantially if household and garden activities are classified as being of moderate intensity. Additionally, population prevalence estimates of “sufficient activity” and population attributable risk estimates would be influenced significantly.
In 1996, Norgan (21) concluded that the existing energy expenditure data had the following limitations: the sample sizes are small, no information is given on biological variability, and it is not clear whether values are based on continuous or intermittent work. Further concerns are that: much early data were collected using the Konfranyi-Michaelis calorimeter, which underreads the volume expired by 16–20% in the range of 2–60 L·min−1 (22); data on the precision and reliability of calorimetry systems are nonexistent; and in many publications, there are no reported checks that the subjects had attained a “steady state” when energy expenditure was measured. Another limitation is that the original 1993 compendium by Ainsworth et al. (1) relied heavily on the dated 1955 listings of Passmore and Durnin (23). However, there have been several recent publications (5,16,30) on energy expenditure during household/garden activities, and these have resulted in a revised compendium by Ainsworth et al. (2).
The present study was undertaken to expand the limited database using the criterion Douglas bag technique and also to address the reliability and precision of measuring energy expenditure during household/garden tasks. We selected sweeping, lawn mowing, window cleaning, and vacuuming because these are among the most strenuous of household and garden tasks. If these activities are not of adequate intensity to confer health benefits, then it is unlikely that tasks such as pruning, weeding, ironing, washing dishes, and dusting will be of adequate intensity. A further consideration is the expedient convention of calculating the MET rating for an activity by dividing the exercise O2 by an assumed resting constant of 3.5 mL O2·kg−1·min−1 (3,29). Although this constant may be representative of the average for lean athletes, it is too high for many persons because their greater relative fat mass has a much lower rate of resting metabolism than the fat free mass (FFM;10). The aims of our study were therefore to: a) calculate the reproducibility (intraclass correlation coefficient, ICC) and precision (technical error of measurement, TEM) during moderate paced walking and self-paced sweeping, lawn mowing, window cleaning and vacuuming; b) determine which of these five activities rate ≥ 3.0 when exercise intensity is calculated in METs and multiples of the measured RMR; and c) expand the limited database on energy expenditure during household and garden activities.
MATERIALS AND METHODS
Twelve men (mean ± SD: 38.8 ± 3.6 yr; 179.5 ± 3.5 cm; 86.0 ± 13.4 kg) and 12 women (39.9 ± 3.2 yr; 163.8 ± 6.6 cm; 75.9 ± 16.4 kg) in the 35–45 yr age range were recruited (Table 1). By using the data of the Australian Fitness Norms (14) as a guide, equal numbers of subjects were tested across three body mass bands with an attempt to recruit some who were light, medium, and heavy. Also, sampling within a limited age range of 10 yr controlled for the agewise decrease in O2max, which increases the relative intensity of absolute workloads.
The sample was screened to exclude smokers, persons suffering from diseases, or taking any medication known to affect energy metabolism and those who had a history of any clinical eating disorder. This project was approved by the Flinders Medical Center’s Human Ethics Committee. The experimental procedures, possible risks, and benefits were explained to the subjects before their written consent was obtained.
Each subject was habituated for RMR and moderate-paced walking, and the four household and garden tasks (self-paced sweeping, window cleaning, vacuuming, and lawn mowing) while wearing a head harness, nose clip, and respiratory valve (described in the subsection “Energy Cost of Activities”). The RMR and energy costs of the five activities for each subject were measured on two separate days, with an interval of 1–4 d (median = 2 d) between trials. On each day, duplicate Douglas bag samples were collected for RMR and the five activities. The order of the five activities was randomized in accordance with a Latin square design.
The classical Douglas bag method, with aluminized Mylar bags that have extremely low gas permeability, and the Geppert and Zuntz (13) transformation were used to measure oxygen consumption (O2). A Beckman LB-2 CO2 analyzer (Yorba Linda, CA) and Electrochemistry S-3A O2 analyzer (Pittsburgh, PA) were calibrated throughout the physiological range of mixed expirate by using gases that had been verified by Lloyd-Haldane analyses. Analyzer calibration was conducted immediately before measuring the gas fractions in the Douglas bags. The volume of the expirate, corrected for the 600 mL of gas extracted by the CO2 and O2 analyzers, was determined using a calibrated 350 L Tissot spirometer (15). The temperature of the expirate within the spirometer was determined as explained previously (15). Energy expenditure (kJ) was calculated from the respiratory exchange ratio (RER) and O2 data in accordance with the recommendations of Elia and Livesey (9). Each subject completed two RMR trials and the lowest value was used in further calculations.
RMR measurement for each subject was preceded by 50 min of bedrest in the supine position with the head and shoulders slightly elevated. All RMR measures were based on two 10-min collection periods while subjects breathed through a low dead space Hans Rudolph R2600 respiratory valve (Kansas City, MO) and wearing a nose clip. A blanket always covered the subject, and the temperature in their vicinity was maintained at 24.0 ± 0.5°C. All precautions (e.g., phone off the hook and only the subject together with the two experimenters allowed in the laboratory) were taken to eliminate disturbing influences that can affect the RMR. The subjects were requested to adhere to the following routine before the two experimental trials in order to control for the factors known to affect the RMR: a) no vigorous exercise during the preceding 36 h; b) no caffeine, alcohol and drugs during the preceding 12 h; c) consume a standardized evening meal between 1930 and 2000 h on the day before the test with only water to be consumed afterward; d) be transported to the laboratory by car to eliminate uncontrolled activity; e) self-reported mass stability (2.0 kg) during the preceding year; and f) measurements on the women were conducted during the late follicular to early luteal phase of the menstrual cycle (days 7–20 of the menstrual cycle) to control for the effect of the menstrual cycle on RMR.
Compliance with some of the foregoing criteria was determined by such procedures as noting resting heart rates, which were monitored continuously via a Polar X-Trainer Plus (Polar Electro OY, Kempele, Finland), and RERs. The most recent data on the reliability of our calorimetry system yielded an intraclass correlation coefficient and technical error of measurement of 0.99 and 1.4%, respectively, for repeated trials on 16 subjects.
Energy cost of activities.
Oxygen consumption and energy expenditure were measured as for RMR, but a Hans Rudolph R2700 respiratory valve (131 g) was used, and this was stabilized in the subjects’ mouths by a customized head harness (247 g). A 2-m length of corrugated tubing (ID = 3.8 cm) connected the expiratory port of the respiratory valve to aluminized Mylar lined Douglas bags via a three-way tap that activated two digital stopwatches. This Douglas bag assembly was bolted to the end of a pole, the other end of which fitted into a holster that was worn by the researcher. The total mass carried by the subject, including respiratory tubing, was 450 g. Mixed expirate was collected from minutes 5–10 and 10–15 of 15 min of continuous activity on each of the following activities: a) moderate paced walking; b) gardening/yard activities—sweeping (broom mass = 660 g) and lawn mowing (Victa Corvette four-stroke power mower; push variety; mass = 31.0 kg); and c) household chores—window cleaning (washer mass = 250 g; wiper mass = 170 g) and vacuuming (Hoover WindTunnel 1300; upright model; mass = 7.9 kg).
Walking was used as a marker of an activity whose intensity should range 3–6 METs because the subjects were requested to walk at what they considered to be a moderate pace. The volunteers were instructed to perform the four household and garden tasks at the pace they would normally do them at home. These standardized instructions were given to all subjects before each task. It was possible to estimate exercise intensity for walking (km·h−1 because the subjects walked around a quadrangle), window cleaning (m2·min−1 because the subjects cleaned a building whose windows consisted of 1.37 m2 panes of glass), and sweeping on a paved surface (m·min−1 for 500 g of sand).
Data were analyzed for reproducibility and precision using the ICC and TEM, respectively. Various dependent and single sample t-tests and interclass correlation coefficients were also conducted. The 0.05 probability level was used for all tests of statistical significance.
The descriptive statistics and metabolic data for the 12 men, 12 women, and combined group (N = 24) are presented in Tables 1 and 2, respectively. There were no mean energy expenditure differences (kJ·kg−1·h−1) between the men and women for RMR (P = 0.95), moderate paced walking (P = 0.86) and self-paced sweeping (P = 0.25), window cleaning (P = 0.58), vacuuming (P = 0.86), and lawn mowing (P = 0.40). The data for men and women were therefore combined for the statistical significance tests.
Table 3 contains the O2 comparisons between the two days and duplicate bags. Dependent t-tests between bags 1 (collection period = 5–10 min) and 2 (collection period = 10–15 min) on day 1 resulted in no significant O2 differences for sweeping (P = 0.84), window cleaning (P = 0.21), vacuuming (P = 0.17), and lawn mowing (P = 0.12); however, there was a significant difference for moderate paced walking (P = 0.01), but the two O2 means were 12.9 and 13.1 mL·kg−1·min−1, respectively. A similar trend emerged for the corresponding comparison on day 2 (sweeping:P = 0.91; window cleaning:P = 0.30; vacuuming:P = 0.78; lawn mowing:P = 0.22; walking:P = 0.053, mean O2(bag1) = 13.3 mL·kg−1·min−1, mean O2(bag2) = 13.5 mL·kg−1·min−1). Except for vacuuming (P = 0.001), there were also no statistically significant differences (P ≥ 0.09) between the overall means for days 1 and 2. The O2 and kJ·kg−1·h−1 data in Table 2 were therefore based on the pooled means of the measurements which were conducted on both days.
The reproducibility and precision data are reported in Table 4. The lowest interday TEM of 2.1% for the combined group occurred for RMR where physical activity was standardized at zero. The combined group’s interday ICCs and TEMs for the five activities ranged from 0.81 to 0.96 and from 3.8 to 7.0%, respectively. All the data in Table 4 furthermore emphasize that, with the exception of the data for sweeping by men, the intraday reproducibilities and precisions were better than the interday ones.
When O2 was divided by measured RMR, all subjects scored above 3.0 for moderate paced walking (3.3–8.7), window cleaning (3.0–6.0), and lawn mowing (4.9–7.5), whereas one subject (2.9) scored below 3.0 for sweeping (2.9–6.7), and five scored below three for vacuuming (2.6–4.4). Single sample t-tests furthermore emphasized that all the means for the men, women, and combined group were significantly greater (P < 0.001) than 3.0 (Table 2). However, division of measured O2 by the conventional O2 of 3.5 mL·kg−1·min−1 significantly decreased (P < 0.001) all the preceding values, and lawn mowing was the only activity where all subjects still scored above 3.0 METs (3.8–6.4). Nevertheless, single sample t-tests resulted in all three data groups for walking, window cleaning, and mowing having means which were significantly greater than 3.0 METs (P ≤ 0.03).
The interclass correlation coefficient of 0.73 between self-perceived moderate paced walking speed and energy expenditure was statistically significant (P < 0.001). These data are depicted in Figure 1. The interclass correlation coefficients between speed and energy expenditure for self-paced sweeping (r = 0.06;P = 0.68) and window cleaning (r = 0.13;P = 0.41) were not statistically significant.
Our results emphasize that the classical Douglas bag method remains ideal to assess O2 during walking and household and garden activities because O2 can be measured with good reliability and precision. If future studies can identify the means and variations in energy expenditure associated with various types of physical activities for different age, gender, and body mass groups, then there is the potential to develop more refined prevalence estimates for health-related physical activity. For example, prevalence estimates can be statistically adjusted for older adults to account for the agewise decrease in O2max, which increases the relative stress of an absolute workload. Our results also demonstrate that expressing the energy expenditure of some common household and garden tasks in conventional METs (RMR = O2 constant of 3.5 mL·kg−1·min−1) yields lower values than when it is presented relative to the measured RMR. Nevertheless, the means that we report for sweeping, window cleaning, and lawn mowing are each in the moderate intensity category of 3–6 METs when exercise O2 is stated in conventional METs. Except for vacuuming, these tasks therefore have the potential to confer health benefits if performed for adequate duration and frequency. However, this generalization must be tempered by the observation that some individuals perform these activities at less than 3 METs.
Precision and Reliability
It is not surprising that the second highest ICC and the lowest TEM in Table 4 were for RMR where the subjects were at rest so that energy expenditure was standardized. This was not so for the five activities where variations in self-paced energy expenditure would adversely affect the precision and reliability for both intraday and interday measurements compared with, for example, pedaling a cycle ergometer at a constant power output. The interday TEM of 3.8% for interday self-paced sweeping means that the probability is 0.68 that the person’s true energy expenditure is within ± 3.8% of the measured value. It is also noticeable that for the combined male and female data the intraday ICCs and TEMs for all variables in Table 4 are higher and lower, respectively, than the interday ones. This is because the intraday values were primarily influenced by tester and equipment error plus variation in self-paced exercise intensity for the five activities, whereas biological variability was superimposed on these three sources of measurement error for the interday ICCs and TEMs. Despite the preceding limitations, the results show that all variables can be measured with good reliability and precision by using the classical Douglas bag method. Although Hendelman et al. (16) tested for statistical significance between the means of repeated trials, recent studies of household and garden activities (5,16,30) have not reported the reliability and precision of their energy expenditure measurements. Hence, the uncertainty associated with interday variation is unknown for expedient methods of measuring energy expenditure using the CosMed K4b2 (Rome, Italy) and Aerosport KB1-C (Ann Arbor, MI).
Indexing energy expenditure to METs.
Howley (17) has researched the origin of the term MET. He traced it back to La Grange’s (18) 1890 text, which contained the conceptualization of the MET, but the term was not mentioned. La Grange (18) used the data of a Dr. Smith to quantify the O2 of sitting, standing, walking, and running in multiples of lying O2. This concept was subsequently supported by Dill (8) in 1936 when he proposed that exercise intensity be expressed as the ratio of the work metabolic rate to RMR and he furthermore recommended guidelines for classifying physical activity into moderate, hard, and maximal intensity categories. However, Gagge et al. (12) were probably the first investigators to use the term MET. They defined the MET as 50 kcal·h−1·m−2 BSA, which approximated the resting metabolism of a subject in a sitting position and under conditions of thermal neutrality. Although we were unable to locate the data on which the 1 MET constant is based, it is used to expediently quantify the intensity of exercise energy expenditure even though we know that the RMR is quite variable.
Our data emphasize that the means for each activity were significantly greater (P < 0.001) when energy expenditure was divided by the measured RMR compared with a constant O2 of 3.5 mL·kg−1·min−1 or 1 MET (Table 2). This demonstrates that there is a limitation to the convention of expressing the energy expenditure during physical activity in multiples of an assumed constant (25), which ignores the biological variability between individuals. An assumed value is adequate if the population mean approximates 3.5 mL O2·kg−1·min−1, but our results suggest that there will be subgroups for whom a different constant is likely to be appropriate. Nevertheless, if exercise intensity is reported in conventional METs, then O2 mL·kg−1·min−1 can be calculated easily by multiplying this value by 3.5; furthermore, the prevalence estimates for physical activity in the population (4), the physical activity recommendations (7,11,24,25,29), and the compendia (1,2) are based on the MET constant. It can be argued that the ACSM guidelines (3) stipulate that 1 MET approximates the O2 of a seated individual at rest with no mention of whether this is for the postabsorptive state whereas we measured the RMR when the subject was lying and postabsorptive. Male energy expenditure is 7.3% higher when sitting compared with lying (28) and the thermic effect of feeding will also elevate the metabolism. However, it is unlikely that a combination of these two factors would increase the resting metabolism of our males from 2.9 mL O2·kg−1·min−1 to 3.5 mL O2·kg−1·min−1. Variation in body composition is the major determinant of RMR because adipose tissue and the FFM have been estimated to have RMRs of 18.8 and 124.3 kJ·kg−1·d−1, respectively (10). An increase in % BF consequently decreases the RMR.
Both Ainsworth et al. (1) and McArdle et al. (20) produced compendia of contemporary physical activities and their associated MET values. However, these were not entirely new data. They relied heavily on the listings of Passmore and Durnin (23) and the other human energy expenditure researchers who preceded them. Much of the data, which were used to estimate the energy cost of different activities, were therefore collected more than 45 yr ago. Although the energy cost of walking and running has been thoroughly characterized as functions of speed and body mass (23), much less is known about the energy cost of walking at a self-selected pace (27). Furthermore, the sweeping paths data of Passmore and Durnin (23) came from a 1920 publication on one subject (mean of four trials = 1.7 METs;19) and their window cleaning data can be sourced to a 1920 (6) report also on one subject (mean of three trials = 3.4 METs) and a 1927 (26) one on three trials for each of two subjects (mean of six trials = 3.7 METs). Many of the MET values in the original compendium were therefore based on very small numbers, and the absence of data for some activities caused Ainsworth et al. (1) to estimate the METs from the energy costs of activities having similar movement patterns. It is therefore refreshing to note that there has been a resurgence in the literature (5,16,30) of contemporary data on energy expenditure during household and garden tasks. These data resulted in Ainsworth et al. (2) revising some of their MET values. Our overall METs, which were calculated using the assumed constant of 3.5 mL O2·kg−1·min−1, were therefore compared with those in the revised compendium of Ainsworth et al. (2) using single sample t-tests. Interestingly, our average METs for sweeping (3.2 vs 4.0), lawn mowing (5.0 vs 5.5), and vacuuming (2.8 vs 3.5) were all significantly lower (P ≤ 0.001) than those in the compendium, whereas our mean MET value for window cleaning was significantly higher (3.6 vs 3.0;P < 0.001). While the masses of the lawn mowers and vacuum cleaners would affect the energy expenditures, we are unable to assess this because other investigators have not included this information in their papers. Our preliminary data therefore suggest that only the vacuuming mean in the 2000 compendium (2) would be marginally downgraded to light intensity (<3 METs). The preceding therefore impacts little on current estimates of whether respondents are exercising at an adequate intensity to accrue health benefits. The 24 subjects in our study were also requested to walk at what they considered to be a moderate intensity. Their average speed of 5.0 km·h−1 resulted in an intensity of 3.7 METs which was significantly greater (P = 0.03) than the compendium value of 3.3 METs for walking at 5.0 km·h−1.
While our research expands the currently limited database on energy expenditure during household and garden activities, the biological variability for the five activities in Table 2 is emphasized by the METs whose ranges spanned from 1.9 (2.2–4.1 METs) for self-paced vacuuming to 4.1 (2.3–6.4 METs) for self-perceived moderate intensity walking. This is furthermore stressed by the larger coefficients of variation reported by Welk et al. (30) for O2 mL·kg−1·min−1 during sweeping (18.4% vs 40.6%) and vacuuming (15.3% vs 44.5%). Although some of these interindividual differences will be due to variability in mechanical efficiency, the major reason is simply that some persons perform specific tasks at a greater intensity than others. Taken together, these findings highlight the problems of estimating from questionnaires and interviews whether respondents are exercising at an adequate intensity to confer health benefits. The situation is further compounded when total daily energy expenditure is estimated.
Population surveys (4) that are now commonly used to track prevalence and trends in physical activity typically ask respondents to report on the moderate intensity activities (often including household and garden tasks) in which they engaged over the past week or past 2 wk. Survey items frequently describe these activities and examples are given to prompt respondent recall. However, if a 35- to 45-yr-old respondent’s perception of all household and garden tasks is that they are of vigorous intensity, then they would be misclassified because our preliminary data suggest that these activities are only of moderate intensity. If this occurs for a number of survey respondents, or is more frequent for particular groups such as the elderly, then overall or subgroup prevalence estimates may be biased.
Our control of potential confounding factors such as recent exercise, smoking, caffeine, and the thermic effect of food facilitated a good estimate of the true biological variability for the energy cost of the selected household and garden activities. Nevertheless, when measuring energy expenditure during physical activity, it is important to determine whether the subjects have reached a state where oxygen consumption equals oxygen requirement. If this stage has not been attained, then the measured value will be less than the true energy expenditure. We attempted to circumvent this problem by continuously monitoring heart rate and not collecting any expired gas until this variable had stabilized. The heart rates of all subjects had attained this plateau by the end of the fifth minute of exercise. Attainment of this “steady-state” was furthermore supported by the nonsignificant differences (P: 0.12–0.91) between bags 1 and 2 on each day for sweeping, window cleaning, vacuuming, and lawn mowing (Table 3). Also, the statistically significant differences (P: 0.01 and 0.053) of just 0.2 mL O2·kg−1·min−1 for walking could have been due to a small increase in walking speed during the final collection periods, even though the subjects were instructed to walk at a constant pace that they considered to be of moderate intensity.
A major problem with the present and similar investigations is the potential measurement error associated with being observed. For example, did the subjects work at a faster pace than normal just because they were being tested, or conversely did they slow down as a consequence of the measurement protocol? We attempted to alleviate this problem by constantly reminding the participants that all physical activities must be performed at the same intensity as if they were in their own house or garden and not being measured by experimenters. They were also informed that their honoraria, which were paid at the completion of testing, did not depend on their exercise intensity. Furthermore, they were not measured during occupational activities, so there were no employment implications.
In summary, our findings: a) demonstrate that energy expenditure during self-paced moderate intensity walking and self-paced sweeping, window cleaning, vacuuming, and lawn mowing can be measured with reliability and precision; b) show that expressing energy expenditure during physical activity in multiples of measured RMR yields higher values than using conventional METs; c) expand the limited database on energy expenditure during household and garden tasks; and d) highlight the biological variability in energy expenditure when different people perform the same task.
This study was supported by a grant from the National Health and Medical Research Council of Australia.
Address for correspondence: Professor R. T. Withers, Exercise Physiology Laboratory, School of Education, Flinders University, GPO Box 2100, Adelaide, South Australia 5001, Australia; E-mail: email@example.com.
1. Ainsworth, B. E., W. L. Haskell, and A. S. Leon, et al. Compendium of physical activities: classification of energy costs of human physical activities. Med. Sci. Sports Exerc. 25: 71–80, 1993.
2. Ainsworth, B. E., W. L. Haskell, and M. C. Whitt, et al. Compendium of physical activities: an update of activity codes and MET
intensities. Med. Sci. Sports Exerc. 32: S498–S504, 2000.
3. American College of Sports Medicine. Guidelines for Exercise Testing and Prescription, 4th Ed. Philadelphia: Lea & Febiger, 1991, pp. 287.
4. Armstrong, T., A. Bauman, and J. Davies. Physical activity patterns of Australian adults. Australian International Health and Welfare.
Canberra: AIHW cat no CVD10, 2000, pp. 24–27.
5. Bassett, D. R., Jr., B. E. Ainsworth, A. M. Swartz, S. J. Strath, W. L. O’Brien, and G. A. King. Validity of four motion sensors in measuring moderate intensity physical activity. Med. Sci. Sports Exerc. 32: S471–S480, 2000.
6. Cathcart, E. P., and F. J. Trafford. Energy expenditure in minor duties. J. Physiol. (Lond. ) 53: 99P, 1920.
7. Commonwealth Department of Health and Aged Care (CDHAC). National Physical Activity Guidelines for Australians. Canberra: CDHAC, 1999.
8. Dill, D. B. The economy of muscular exercise. Physiol. Rev. 16: 263–291, 1936.
9. Elia, M., and G. Livesey. Energy expenditure and fuel selection in biological systems: the theory and practice of calculations based on indirect calorimetry and tracer methods. In: Metabolic Control of Eating, Energy Expenditure and the Bioenergetics of Obesity, A. P. Simopoulos (Ed.). Basel: Karger, 1992, pp. 68–131.
10. Elia, M. Organ and tissue contribution to metabolic rate. In: Energy Metabolism: Tissue Determinants and Cellular Corollaries, J. M. Kinney and H. N. Tucker (Eds.). New York: Raven Press Ltd., 1992, pp. 61–79.
11. Fentem, P., and A. Walker. Setting targets for England: challenging measurable and achievable. In: Moving on: International Perspectives on Promoting Physical Activity, A. Killoran, P. Fentem, and C. Caspersen (Eds.). London: Health Education Authority, 1994, pp. 58–79.
12. Gagge, A. P., A. C. Burton, and H. C. Bazett. A practical system of units for the description of heat exchange of man with his environment. Science 94: 428–430, 1941.
13. Geppert, J., and N. Zuntz. On the regulation of breathing. Pfluegers Arch. 42: 189–245, 1888.
14. Gore, C. J., and D. A. Edwards. Australian Fitness Norms. Adelaide, South Australia: Health Development Foundation, 1992, p. 27.
15. Hart, J. D., and R. T. Withers. The calibration of gas volume measuring devices at continuous and pulsatile flows. Aust. J. Sci. Med. Sport 28: 61–65, 1996.
16. Hendelman, D., K. Miller, C. Baggett, E. Debold, and P. Freedson. Validity of accelerometry for the assessment of moderate intensity physical activity in the field. Med. Sci. Sports Exerc. 32: S442–S449, 2000.
17. Howley, E. T. You asked for it: question authority. ACSM’s Health and Fitness Journal
4(1):6, 40, 2000.
18. Lagrange, F. Physiology of Bodily Exercise. New York: D. Appleton, 1890, pp. 262–265.
19. Langworthy, C. F., and H. G. Barott. Energy expenditure in household tasks. Am. J. Physiol. 52: 400–408, 1920.
20. Mcardle, W. D., F. I. Katch, and V. L. Katch. Exercise Physiology: Energy, Nutrition and Human Performance, 2nd Ed. Philadelphia: Lea & Febiger, 1988, pp. 642–649.
21. Norgan, N. G. Measurement and interpretation issues in laboratory and field studies of energy expenditure. Am. J. Hum. Biol. 8: 143–158, 1996.
22. Orsini, D., and R. Passmore. The energy expended carrying loads up and down stairs: experiments using the Kofranyi-Michaelis calorimeter. J. Physiol. (Lond. ) 115: 95–100, 1951.
23. Passmore, R., and J. V. G. A. Durnin. Human energy expenditure. Physiol. Rev. 35: 801–839, 1955.
24. Pate, R. R., M. Pratt, and S. N. Blair, et al. Physical activity and public health: a recommendation from the Centres for Disease Control and Prevention and the American College of Sports Medicine. JAMA 273: 402–407, 1995.
25. Pollock, M. L., G. A. Gaesser, and J. D. Butcher, et al. 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. 30: 975–991, 1998.
26. Richardson, D. T., W. Campbell, R. Morrison, and J. F. Ford. Report on the investigation of the energy expenditure of the British soldier in India. Calcutta: Government of India Press, 1927.
27. Spelman, C. C., R. R. Pate, C. A. Macera, and D. S. Ward. Self-selected exercise intensity of habitual walkers. Med. Sci. Sports Exerc. 25: 1174–1179, 1993.
28. Strickland, S. S., and S. J. Ulijaszek. Energetic cost of standard activities in Gurkha and British soldiers. Ann. Hum. Biol. 17: 133–144, 1990.
29. U.S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996.
30. Welk, G. J., S. N. Blair, K. Wood, S. Jones, and R. W. Thompson. A comparative evaluation of three accelerometry-based physical activity monitors. Med. Sci. Sports Exerc. 32: S489–S497, 2000.