Measuring Adolescent Boys' Physical Activity: Bout Length and the Influence of Accelerometer Epoch Length

Lonsdale C (2014) Measuring Adolescent Boys' Physical Activity: Bout Length and the Influence of Accelerometer Epoch Length. PLoS ONE 9(3): e92040. doi:10.1371/journal.pone.0092040 Measuring Adolescent Boys' Physical Activity: Bout Length and the Influence of Accelerometer Epoch Length Taren Sanders 0 Dylan P. Cliff 0 Chris Lonsdale 0 Michael Taffe, The Scripps Research Institute, United States of America 0 1 University of Western Sydney, School of Science and Health , Penrith , Australia , 2 University of Wollongong, Interdisciplinary Educational Research Institute, School of Education , Wollongong , Australia Objectives: Accurate, objective measurement is important for understanding adolescents' physical activity (PA) behaviour. When using accelerometry to objectively measure PA, a decision must be made regarding how frequently data is recorded (i.e., epoch length). The purpose of this study was to examine i) PA bout length, and ii) the effect of variations in accelerometer epoch length on PA estimates during physical education (PE) and leisure time in adolescent boys. Methods: Year 9 boys (N = 133; mean age 6SD = 14.3660.48 years) wore accelerometers during two PE lessons, and for a period of seven consecutive days. Data were reintegrated from 1s into longer periods of 2, 5, 10, 30, and 60 seconds. ANOVAs were used to test for differences in PA estimates between epochs in leisure time and PE. Results: The mean length of vigorous PA (VPA) bouts was 3.562.0 seconds for PE and 2.561.7 seconds for leisure time, and mean length of moderate PA (MPA) bouts was 2.360.5 seconds for PE and 2.960.5 seconds for leisure time. During PE, estimates of MVPA, MPA, and light PA (LPA) increased as epoch increased from 1 second to 60 seconds, while VPA and sedentary behaviour estimates decreased. During leisure time, estimates of all PA intensities decreased as epoch increased from 1 second to 60 seconds, with the exception of sedentary behaviour, which increased as epoch length increased. Conclusion: The context in which PA occurs can influence PA bout length measurement and the effect of variations in epoch length on PA estimates. Researchers measuring PA with accelerometry should be conscious of the possible influence of context on PA estimates. - Competing Interests: The authors have declared that no competing interests exist. Introduction Regular physical activity (PA) during adolescence provides a number of health benefits [13]. However, PA declines rapidly during this time [4,5], making adolescence a priority period for PA promotion and research. Accurately quantifying PA is necessary to conduct meaningful research involving adolescents. Due to their ability to provide objective, time-stamped data on the intensity, frequency, and duration of PA with low participant burden [6,7], accelerometers have become the instrument of choice for assessing habitual PA among adolescents. In order to accurately measure PA using accelerometry, a variety of decisions must be made, including the frequency at which data is recorded, known as an epoch [8,9]. An epoch represents the length of time that activity counts- a measure of activity magnitude - are summed before being stored by the accelerometer. The choice of epoch should be largely determined by the PA bout length (i.e. time spent in a single PA intensity [10,11]) of the population being studied [8]. When PA bout length is shorter than the epoch length, there is potential for misclassification, causing the estimations of PA to become skewed. Therefore, an understanding of PA bout length is important when assessing accelerometer epoch effects. To the authors knowledge, only two studies have measured PA bout length among young people. Both were conducted in a leisure time context with children aged 610 years [10,11]. Bailey et al. [10] used direct observation during leisure time to show that 610 year-old childrens PA was highly intermittent, with a median bout length for all PA intensities of #6 s. Baquet et al.[11] used accelerometers with 2 s epochs to measure bout length in children aged 810 years and reported similar results to Bailey et al. More than 95% of vigorous or very vigorous PA lasted ,10 s, with average bout length for all intensities #9 s. These two studies are routinely cited as a justification for using a short epoch [12,13]. However, there is a need for evidence from other populations, including adolescents, whose PA behaviour might potentially be less sporadic and intermittent than that of children. Varying the epoch length has been shown to influence estimations of youth PA volume and intensity [9,14,15]. In an adolescent sample, Edwardson and Gorely [8] found that using a 5, 15, or 30 s epoch produced significantly different estimates of moderate-to-vigorous PA (MVPA) compared to a 60 s epoch. When using the same equipment, comparing similar epoch lengths, and similar samples, Nilsson et al. [14] demonstrated that estimates of 7 to 8 year-old childrens MVPA decreased as epoch length increased from 5 s to 60 s during physical education (PE), while McClain et al. [9] found that estimates of 10 to 11 year-old childrens MVPA increased as epoch length increased during leisure time. McClain et al. hypothesised that the difference in results might have occurred because structured PA, such as PE, may have shorter, more intense bouts of PA compared to unstructured PA occurring during leisure time. To date, this hypothesis has not been tested using data gathered from the same sample. The first objective of this study was to compare adolescent boys PA bout length in two PA contexts; leisure time and physical education lessons. The second objective was to compare the effect of varying accelerometer epoch length on estimates of MVPA, vigorous PA (VPA), moderate PA (MPA), light PA (LPA), and sedentary behaviour in both contexts. It was hypothesized that: 1) the mean bout lengths of all PA intensities would be longer in a leisure time context, compared to the PE context; 2) estimates of MVPA, VPA, and MPA would be higher using a short epoch length compared to a longer epoch length, in both contexts; and 3) due to shorter, more intense PA bouts occurring in PE lessons, the impact of varying the epoch length on PA intensity estimation would be more pronounced in PE than in leisure time. A sample of 133 male Year 9 students (mean age 6SD = 14.3660.48 years; mean BMI6SD = 22.3663.83) were recruited from six PE classes within a boys Catholic school in Sydney, Australia. To be eligible to participate, students were required to have no existing conditions that would prevent their participation in PA. Before PA data were collected, trained research assistants used a portable stadiometer to measure height to the nearest 0.5 cm (Surgical and Medical Products No. 26SM, Medtone Education Supplies, Melbourne, Australia), and a set of digital scales to measure weight to the nearest 0.1 kg (UC-321, A&D Company LTD, Tokyo, Japan). Body m (...truncated)