Patterns, Trends and Tracking in Physical Activity Throughout Childhood and Adolescence

Aims Understanding trends and tracking of children’s physical activity (PA) could help to optimise the type and timing of interventions designed to make children more active. The aims of this longitudinal study were to establish the age at which PA levels begin to fall and to determine whether inactivity in childhood is a precursor to inactivity in adolescence.

Methods Physical activity was measured by 7-day Actigraph recordings every year from age 5y to 15y in a cohort of children (n=273-229, —50% boys) from Plymouth, UK. Total PA and intensity-specific PA (sedentary, light intensity and moderate-and-vigorous intensity [MVPA]) were extracted from the accelerometer output. Age-related trends in PA were assessed using multilevel modelling. Tracking of PA was assessed by correlation and Cohen’s Kappa statistic.

Results Total PA was relatively stable between the ages 5y and 8y for both genders though it was systematically lower in girls. It then fell consistently from 9y to 15y, more rapidly in girls than boys (boys: -16,419, p<0.001, girls: -22,771 counts/day per year, p<0.001, p<0.01 for ‘gender’ interaction). The decrease in MVPA from 9-15y accounted for only a quarter of the decrease in total PA counts during that period whereas the decrease in light intensity PA accounted for about half (boys: MVPA fell by 3221, light PA by 9801 counts/day per year p<0.001, girls: MVPA fell by 6149 p<0.001, light PA by 10,340 counts/day per year p<0.001). PA tracked moderately throughout childhood (from primary to secondary school years r=0.55, kappa=0.25, both p<0.001)

Conclusions Children’s activity levels decline from age 9y though the greater reduction observed in light intensity activity than in moderate-and-vigorous activity implies that habitual activity should be targeted for PA promotion. Identifying children that are inactive during early childhood could help interventions to target those who are likely to be inactive adolescents.