Using functional data analysis to understand daily activity levels and patterns in primary school-aged children: Cross-sectional analysis of a UK-wide study.

Francesco Sera,Mario Cortina-Borja,Lucy J Griffiths,Marco Geraci,Carol Dezateux

doi:10.1371/journal.pone.0187677

Abstract

BackgroundTemporal characterisation of physical activity in children is required for effective strategies to increase physical activity (PA). Evidence regarding determinants of physical activity in childhood and their time-dependent patterns remain inconclusive. We used functional data analysis (FDA) to model temporal profiles of daily activity, measured objectively using accelerometers, to identify diurnal and seasonal PA patterns in a nationally representative sample of primary school-aged UK children. We hypothesised that PA levels would be lower in girls than boys at play times and after school, higher in children participating in social forms of exercise (such as sport or play), and lower among those not walking to school.MethodsChildren participating in the UK-wide Millennium Cohort Study wore an Actigraph GT1M accelerometer for seven consecutive days during waking hours. We modelled 6,497 daily PA profiles from singleton children (3,176 boys; mean age: 7.5 years) by means of splines, and used functional analysis of variance to examine the cross-sectional relation of time and place of measurement, demographic and behavioural characteristics to smoothed PA profiles.ResultsDiurnal and time-specific patterns of activity showed significant variation by sex, ethnicity, UK country and season of measurement; girls were markedly less active than boys during school break times than boys, and children of Indian ethnicity were significantly less active during school hours (9:30–12:00). Social activities such as sport clubs, playing with friends were associated with higher level of PA in afternoon (15:00–17:30) and early evenings (17:30–19:30). Lower PA levels between 8:30–9:30 and 17:30–19:30 were associated with mode of travel to and from school, and number of cars in regular use in the household.ConclusionDiminished PA in primary school aged children is temporally patterned and related to modifiable behavioural factors. FDA can be used to inform and evaluate public health policies to promote childhood PA.

Highlights

In the UK, it is recommended that young people participate in activity of moderate to vigorous intensity for at least one hour every day, [1] only half of seven year old children in the UK achieve these recommended levels, with significant sex, ethnic and geographic variations. [2]Effective strategies are needed to increase physical activity (PA) levels in young children
We hypothesised that PA levels would be lower in girls than boys at play times and after school, higher in children participating in social forms of exercise, and lower among those not walking to school
The importance of social activities for physical activity is underscored by the time-specific differences in activity levels observed in those involved in sport or after school clubs or who played with friends on a regular basis

Summary

Introduction

In the UK, it is recommended that young people participate in activity of moderate to vigorous intensity for at least one hour every day, [1] only half of seven year old children in the UK achieve these recommended levels, with significant sex, ethnic and geographic variations. Effective strategies are needed to increase PA levels in young children. Accelerometer-based measures of physical activity (PA) are increasingly used to measure childhood activity levels and their intensity, and to investigate their determinants, in largescale epidemiological studies. We used functional data analysis (FDA) to model temporal profiles of daily activity, measured objectively using accelerometers, to identify diurnal and seasonal PA patterns in a nationally representative sample of primary school-aged UK children. We hypothesised that PA levels would be lower in girls than boys at play times and after school, higher in children participating in social forms of exercise (such as sport or play), and lower among those not walking to school.

Methods

Results

Discussion

Conclusion