Accelerometry measurement of physical activity and sedentary behaviour in pre-school children

This thesis is based on six studies which address questions around the use of accelerometers to measure physical activity and sedentary behaviour of pre-school children: are shorter epochs more accurate? Which epochs are most accurate? Are there advantages to using triaxial accelerometers? Which cut-points are most accurate? Are different generations of Actigraph accelerometers comparable? What is the recommended wear time to provide a reliable estimate of habitual physical activity and sedentary behaviour? Analysis of 7-10 day accelerometry data, collected from 31 pre-school children (mean (SD) age 5.9 (0.7) y), suggests that shorter epochs (15 s) result in significantly greater estimates of time spent in moderate-to-vigorous physical activity (MVPA) in comparison to 60-s epochs (p <0.05). When compared against a direct observation method, Children's Activity Rating Scale (CARS), with 32 pre-school children (4.4 (0.8) y) during 1 hour of free-play, 15-s epochs were more accurate than 60-s epochs. Comparison of the triaxial RT3 against a uniaxial accelerometer, suggests no advantage of the RT3 accelerometer. The Puyau et al. (2002) cut-points had the 'best' agreement with estimates of sedentary behaviour, light intensity and MVPA against the CARS. Different generations of accelerometers were not comparable, however, application of a correction factor to the GT1M data (7164 = GT1M/0.91) may improve comparability of total physical activity. Finally, analysis of 7 day accelerometry data from 112 pre-school children (3.7 (0.7) y) suggests that 3 days of 7 hours provides a reliable estimate of habitual physical activity and that inclusion of weekend days is not necessary. This thesis highlights the implications that methodological decisions can have over apparent estimates of physical activity and sedentary behaviour and has made recommendations for accelerometry use. Ideally, there needs to be a move towards consensus, as, only by adopting standardised approaches to accelerometry use, will comparison between study outcomes become meaningful.sub_phyunpub1598_ethesesunpu