Accelerometers in Young Children: Methodological Considerations and Cross-sectional and Longitudinal Associations with Motor Abilities, Physical Fitness, and Cognitive Function

Physical activity plays an important role during early childhood and has favourable associations with numerous health outcomes. It is thought that lifestyle and physical activity behaviours may develop within the first five years of life, making the early years an optimal time for targeted intervention and observation. Accelerometers are valid and reliable devices that allow researchers to quantify large amounts of free-living data in young child populations. However, use of these devices are accompanied by many methodological decisions which can create inconsistencies and limit comparability within the literature. The focus of this thesis was to extensively review current methods being used to analyse accelerometer data from young children, examine how certain methodologies affect interpretation of accelerometer outputs, and observe subsequent cross-sectional and longitudinal associations between physical activity and health. An extensive review revealed that hundreds of published studies use accelerometers with a wide variety of data collection and analysis methods to track movement in young children. Applying different thresholds or cutpoints to quantify movement behaviours
into specific intensities of movement demonstrated that choice of cutpoint significantly impacts amounts of time spent in each intensity as well as the number of children categorized as meeting physical activity guidelines. A cross-sectional analysis showed that preschool-aged children who met physical activity guidelines did not have greater levels of individual motor abilities compared to children not meeting guidelines, however sports club membership may influence motor ability development. In a longitudinal analysis, children who met physical activity guidelines during early childhood were more likely to meet guidelines during later childhood/adolescence. Additionally, performance on certain physical fitness tests at baseline also predicted greater amounts of time in physical activity and meeting physical activity guidelines at follow-up. Interestingly, meeting physical activity guidelines showed no associations with cognitive function, and some physical fitness tests showed statistically significant associations, but models did not show a substantial goodness of fit. In summary, future work should continue to investigate associations between meeting physical activity guidelines and health outcomes in young children. Given the importance of this work, we strongly encourage researchers to adapt and apply our recommended, standardized accelerometry reporting practice.