Eur J Appl Physiol (1986) 54:585--590 European Journal of Applied Physiology and Occupational Physiology 9 Springer-Vedag 1986 Correlations of pedometer readings with energy expenditure in workers during free-living daily activities Hiroshi Kashiwazaki 1, Tsukasa Inaoka 1, Tsuguyoshi Suzuki 1, and Yasuyuki Kondo z Department of Human Ecology, School of Health Sciences, University of Tokyo and 2 Section of Health Care, Mitsubishi Electric Co Summary. In a total of 23 subjects consisting of 10 clerical and 13 assembly workers in a factory, the pedometer readings during a day of free-living activity were analyzed for the relation with energy expenditure as determined by the simultaneously recorded 24-hour heart rate. The 24-hour energy expenditures in the clerical and assembly workers were 9515 kJ (2274 kcal) and 9698 kJ (2318 kcal) respectively. The whole day readings of the pe- dometer for all the subjects moderately correlated (r= 0.438,p < 0.05) with the net energy cost (NEC) as determined by subtracting the sleeping meta- bolic cost from the energy expenditure (clerical workers: r=0.781, p<0.01; assembly workers: r = 0.188, p > 0.05). The correlation analysis of the pedometer readings with the NEC in three activ- ity phases in a day (work, commuting and staying at home), showed that the extent of the relation- ship differed by job types and activity phases. The best correlation was obtained during commuting in both of the job types (clerical workers: r=0.843, p<0.01; assembly workers: r=0.743, p < 0.01). During work, a quite strong correlation (r= 0.889, p < 0.01) was obtained with the clerical workers but not with the assembly workers. No significant correlations were found in the data while the subjects were at home. The capacity of the pedometer to detect the impacts of body mov- ements, and the characteristics of activity, are re- sponsible for the differences in correlation. The limitations of the pedometer suggested in the present study must be taken into account if the device is to be used for measuring physical activi- ty. A particular advantage of the device appears Offprint requests to: H. Kashiwazaki, Department of Human Ecology, School of Health Sciences, University of Tokyo, 7-3- 1 Hongo, Bunkyo-ku, Tokyo, 113 Japan in its use for a sedentary population without regu- lar srenuous exercise or static contractions. Key words: Pedometer -- Energy expenditure -- Free-living daily activity -- Work conditions -- 24-h heart rate Introduction There is great concern among public health per- sonnel that the low daily energy expenditure in certain individuals may be further reduced by such events as automation, mechanization and widespread use of electric apparatus. However, we have little information on physical activity lev- els within a population. Indirect calorimetry and time-and-motion studies are reliable methods for estimating physical activity, but they are inappro- priate for a study with a large number of subjects, and cannot be used to gather data over a period of several days. Because these methods impose la- borious efforts on both subjects and investigators, and measurements of oxygen consumption inter- fere with free-living activity, the application of such methods has to be limited to a small number of subjects, and the results obtained are often the subject of criticism for being unrepresentative. A simple and quantifiable method for evaluat- ing physical activity is needed. The pedometer is the simplest device of a mechanical counter and has been used for recording activity as early as 1930s (Montoye and Taylor 1984; Stunkard 1960). For example, the comparison of physical activity as measured by pedometer for a week or so be- tween obese and non-obese subjects revealed that the obese subjects were far less active than their non-obese controls (Stunkard 1960). But until