Nature of children’s activities
Abstract
Introduction: Regular physical activity (PA) is an important part of a healthy lifestyle for children. The difficulty of measuring PA objectively in this population causes, in part, the lack of quantifiable conclusions regarding possible health effects and physical activity (PA) in children. Promising results in measuring PA objectively were recently published using heart rate monitoring and accelerometry simultaneously to measure energy expenditure in adults. A recent project at the EHSM institute followed this attempt to develop a measurement system that assesses furthermore the type of activities in adults. To apply this approach to children, preliminary data are required as activity patterns in children are different from those of adults. Therefore, duration, intensity, frequency and type of children’s physical activity should be determined in the present study.
Method: Physical activity of 41 children was recorded with two accelerometers (wirst and hip) and an activity log simultaneously during one week. In the measurement week, the physical activity of each child was recorded by video during two sessions of 4 hours. Observers labelled the accelerometer data with activity codes based on the video data. The labelled accelerometer data were used to develop a classifying system that allowed categorizing all accelerometry data over the whole measurement week. Validity of the data labelling system was verified by comparing the energy expenditure estimation based on video data with corresponding value based on accelerometer data. Data of classifying system on the recognised activities during the measurement week were compared to the data of the activity log.
Results: Means of accelerometry data (wrist and hip) correlated with means of energy expenditure estimated with video data (R= 0.77, p < 0.05). The recognition rate of activity categories perfomed by all children was moderate to high (stationary: 93%, walking: 69%, running: 75% and jumping: 60%). Activities that were less popular among the study sample were not recognised by the used classifier system. According to the classification system, weekly mean time spent at low, moderate and vigorous intensity levels were 605±83.3 min, 130±30.4 min, 16.7±8.5 min, 4.5±5.0 min, respectively. According to activity log the time spent in these activities was 555.3 ± 126.9 min, 212.0 ±183.3 min, 65.8 ± 69.98 min, respectively. The duration of activities was significantly different between activity log and classification system on the low and the vigorous level. Median durations of the activities registered by accelerometry were 4 s, 2 s, 1s and 1 s for stationary, walking, running and jumping, respectively.
Discussion: Short, intermittent bouts of vigorous physical activity are typical in children. It is suggested that intermittent bouts of walking and running in children are shorter than so far assumed. Most stationary bouts are suggested to be longer than ‘active’ bouts. As some activity could not be distinguished from other activity classes, more information of objective sensors are needed to complete the objective data on physical activity behaviour of children. However, the sensor system must be us unobtrusive as possible to minimise its interference with the physical activity behaviour.
