Accuracy Of Accelerometer, Activity Record, And Questionnaire By Comparison With The Doubly-labeled Water Method

Abstract

BACKGROUND: The need for developing a simple and accurate assessment method for physical activity is increasing both for the study of the relationship between physical activity and health, and for health policy making. PURPOSE: To determine the accuracy of estimates of energy expenditure using three types of commercial devices, a tri-axial accelerometer with our previously reported equation, activity record, and a simple questionnaire by comparing the energy expenditure as measured by the doubly-labeled water method. METHODS: Healthy sedentary women's college students (n=1 1, age=21.2 years, height=159.1 cm, weight 50.9 kg) and middle-aged housewives (n=12, age=40.0 years, height=156.7 cm, weight=49.6 kg) volunteered as subjects. They completed the measurement of basal metabolic rate, the 3-day activity record, and physical activity questionnaire for the Japan Arteriosclerosis Longitudinal Study (JALSPAQ). They wore a uni-axial accelerometer (Lifecorder, Suzuken Inc., Japan), an ankle-mounted accelerometer-based activity monitor (AMP331, Dynastream Innovations Inc., Canada), a tri-axial accelerometer (AC-210, GMS Inc., Japan), and a microcomuter-based portable physical activity measurement device (IDEEA, MiniSun LLC, USA). 14-day energy expenditure was measured using the doubly-labeled water method. RESULTS: Most of the subjects were at the moderate physical activity level (total energy expenditure (TEE)/basal metabolic rate=1 .60 to 1.90) according to the Dietary Reference Intakes for Japanese, 2005. All devices and the questionnaire underestimated TEE both for young and middle-aged women. The AMP331 monitor underestimated TEE the most. Accuracy of each method was similar for young women, but it varied greatly for the middle-aged women. CONCLUSION: Accuracy of the assessment method for physical activity differed with the physical activity pattern.