No abstract

BackgroundExploring the relationship between physical activity, cognition and academic performance in children is an important but developing academic field. One of the key tasks for researchers is explaining how the three factors interact. The aim of this study was to develop and test a conceptual model that explains the associations among physical activity, cognition, academic performance, and potential mediating factors in children.MethodsData were sourced from 601 New Zealand children aged 6–11 years. Weekday home, weekday school, and weekend physical activity was measured by multiple pedometer step readings, cognition by four measures from the CNS Vital Signs assessment, and academic performance from the New Zealand Ministry of Education electronic Assessment Tools for Teaching and Learning (e-asTTle) reading and maths scores. A Structured Equation Modelling approach was used to test two models of variable relationships. The first model analysed the physical activity-academic performance relationship, and the second model added cognition to determine the mediating effect of cognition on the physical activity-academic performance association. Multigroup analysis was used to consider confounding effects of gender, ethnicity and school socioeconomic decile status.ResultsThe initial model identified a significant association between physical activity and academic performance (r = 0.225). This direct association weakened (r = 0.121) when cognition was included in the model, demonstrating a partial mediating effect of cognition. While cognition was strongly associated with academic performance (r = 0.750), physical activity was also associated with cognition (r = 0.138). Subgroups showed similar patterns to the full sample, but the smaller group sizes limited the strength of the conclusions.ConclusionsThis cross-sectional study demonstrates a direct association between physical activity and academic performance. Furthermore, and importantly, this study shows the relationship between physical activity and academic performance is supported by an independent relationship between physical activity and cognition. Larger sample sizes are needed to investigate confounding factors of gender, age, socioeconomic status, and ethnicity. Future longitudinal analyses could investigate whether increases in physical activity can improve both cognition and academic performance.

BackgroundActive transport (e.g., walking, cycling) to school (ATS) can contribute to children’s physical activity and health. The built environment is acknowledged as an important factor in understanding children’s ATS, alongside parental factors and seasonality. Inconsistencies in methodological approaches exist, and a clear understanding of factors related to ATS remains equivocal. The purpose of this study was to gain a better understanding of associates of children’s ATS, by considering the effects of daily weather patterns and neighbourhood walk ability and neighbourhood preferences (i.e., for living in a high or low walkable neighbourhood) on this behaviour.MethodsData were drawn from the Understanding Relationships between Activity and Neighbourhoods study, a cross-sectional study of physical activity and the built environment in adults and children in four New Zealand cities. Parents of participating children completed an interview and daily trip diary that assessed their child’s mode of travel to school, household and individual demographic information, and parental neighbourhood preference. Daily weather data were downloaded from New Zealand’s national climate database. Geographic information systems-derived variables were calculated for distance to school and neighbourhood walkability. Bivariate analyses were conducted with ATS and potential associates; factors related to ATS at p < 0.20 were considered simultaneously in generalized estimation equation models, and backwards elimination of non-significant factors was conducted; city was treated as a fixed effect in all models.ResultsA total of 217 children aged 6.5-15 years participated in this study. Female sex, age, city, household income, limited/no car access, residing in zone of school, shorter distance to school, neighbourhood self selection, rainfall, and sunlight hours were simultaneously considered in multivariate generalised estimation equation modelling (all p < 0.20 in bivariate analyses). After elimination of non-significant factors, age (p = 0.005), shorter distance to school (p < 0.001), city (p = 0.03), and neighbourhood self selection (p = 0.04) remained significantly associated with ATS in the multivariate analysis.ConclusionDistance to school is the prevailing environmental influencing factor on children’s ATS. This study, in conjunction with previous research, suggests that school siting is likely an important associate of children’s ATS.

Background: Application of machine learning for classifying human behavior is increasingly common as access to raw accelerometer data improves. The aims of this scoping review are (1) to examine if machine-learning techniques can accurately identify human activity behaviors from raw accelerometer data and (2) to summarize the practical implications of these machine-learning techniques for future work. Methods: Keyword searches were performed in Scopus, Web of Science, and EBSCO databases in 2018. Studies that applied supervised machine-learning techniques to raw accelerometer data and estimated components of physical activity were included. Information on study characteristics, machine-learning techniques, and key study findings were extracted from included studies. Results: Of the 53 studies included in the review, 75% were published in the last 5 years. Most studies predicted postures and activity type, rather than intensity, and were conducted in controlled environments using 1 or 2 devices. The most common models were support vector machine, random forest, and artificial neural network. Overall, classification accuracy ranged from 62% to 99.8%, although nearly 80% of studies achieved an overall accuracy above 85%. Conclusions: Machine-learning algorithms demonstrate good accuracy when predicting physical activity components; however, their application to free-living settings is currently uncertain.

When previous wear time compliance results are taken together with our findings, it represents a promising step forward for monitoring and understanding 24-h time-use behaviors. The next step will be to examine the generalizability of these findings in a free-living setting.