| dc.description.abstract | In the recent years, accelerometers has become the most common way to assess physical activity, and serval activity type classification models have been developed based on accelerometry. These models aim to determine individual’s activity behavior by identifying the engagement in activity types, with main focus on common daily activities. In youth, physical activity is often performed as complex activities such as handball and football. To be able to include these activities into the classification models we need a detailed description of the different components (sub-activities) that constitute handball and football. The aim of this study was therefore to describe characteristics of complex activities such as handball and football in youth, by developing reliable activity definitions and investigate its relation to intensity. For this purpose 12 adolescents performed 90 minutes of complex activities. Six participants completed one football training and six participants completed two handball trainings, while they wore two AX3 Axivity accelerometers (on the lower back and mid-right thigh), a heart rate monitor and were video recoded. A total of 19 new activity definitions were developed and used as guidelines for annotation of the video-recordings. Two persons, using 31 different category labels, labeled the videos of the participants´ movements during football and handball. The mean inter rater reliability was high (Cohen´s kappa 0.87). When assessing the relation between the intensity of the activity (% of max heart rate), the activity pattern (change in activity per minute) and the vector magnitude of the acceleration signal (VM), we found that VM and change in activity per minute increased with increasing intensity when playing football. When playing handball, the same relation was found between low, moderate and high intensity, but not for sedentary intensity. VM showed the strongest relation to intensity for both handball and football. However, the relation was stronger when playing football compared to playing handball.
Based on these results we conclude that the activity definitions in this study were reliable for the different components (sub-activities) that constitute handball and football, and that VM can be used to estimate change in intensity during these complex activities. | nb_NO |
