Objectively measured physical activity and sedentary time in youth: the International children's accelerometry database (ICAD)

Abstract

Background: Physical activity and sedentary behaviour in youth have been reported to vary by sex, age, weight status and country. However, supporting data are often self-reported and/or do not encompass a wide range of ages or geographical locations. This study aimed to describe objectively-measured physical activity and sedentary time patterns in youth.

Methods: The International Children's Accelerometry Database (ICAD) consists of ActiGraph accelerometer data from 20 studies in ten countries, processed using common data reduction procedures. Analyses were conducted on 27,637 participants (2.8-18.4 years) who provided at least three days of valid accelerometer data. Linear regression was used to examine associations between age, sex, weight status, country and physical activity outcomes.

Results: Boys were less sedentary and more active than girls at all ages. After 5 years of age there was an average cross-sectional decrease of 4.2% in total physical activity with each additional year of age, due mainly to lower levels of light-intensity physical activity and greater time spent sedentary. Physical activity did not differ by weight status in the youngest children, but from age seven onwards, overweight/obese participants were less active than their normal weight counterparts. Physical activity varied between samples from different countries, with a 15-20% difference between the highest and lowest countries at age 9-10 and a 26-28% difference at age 12-13.

Conclusions: Physical activity differed between samples from different countries, but the associations between demographic characteristics and physical activity were consistently observed. Further research is needed to explore environmental and sociocultural explanations for these differences.

Fig. 1

Physical activity level by age and sex, according to different activity metrics. a Total physical activity (accelerometer counts per minute); b Percentage time in Moderate to Vigorous Physical activity (MVPA); c Percentage time in light intensity physical activity; d Percentage time sedentary. CI = confidence interval, cpm = counts per minute. In all graphs, boys aged 5–6 are the reference population. All graphs present mean differences in physical activity variables, adjusting for study population and season.

Fig. 2

Association between weight status and physical activity. a Total physical activity (accelerometer counts per minute) by age, sex and weight status; b Dose response relationship between total physical activity and BMI by sex CI = confidence interval, owt = overweight, cpm = counts per minute

Fig. 3

Average activity counts per minute across selected countries. a Level of physical activity across selected countries at age 9–10. b Level of physical activity across selected countries at age 12–13. CI = confidence interval. The Australian sample pools the two ICAD studies collected in Melbourne; both produced similar findings when analysed separately

Fig. 4

Achievement of physical activity guidelines across selected countries. a Average percentage of participants achieving ≥60 min of MVPA each measurement day. b Average percentage of days on which ≥60 min of MVPA were recorded

Fig. 5

Comparison between selected countries of the relative effect of sex, age and weight status. a Difference in physical activity by sex in selected countries at ages 9–10 and 12–13. b Longitudinal change in physical activity per year of increasing age. c Difference in physical activity associated with higher BMI in selected countries at ages 9–10 and 12–13. SD = standard deviation. Forest plots display random-effects meta-analysis. The estimates for each country were calculated from regression models adjusted for age, sex and season, fitted to each country in turn