This study showed that British 4-year old children spend on average 67% of their day being active, engaging predominately in LPA (~88% of active time). Although few differences in average activity levels were observed, the current study provides novel information about how temporal and demographic factors differentially influence children’s activity when segmented across the day. Child’s sex and weight status, age their mother left full-time education, attending childcare full-time, time of the week and season were all independently associated with children’s activity levels at different times of the day. These time-specific observations are likely to be important for intervention development, indicating that focussing on (specific subgroups in) periods when children are less active, for example girls in the mornings, or during winter months, may result in larger increases in children’s activity.
Based on the data presented here, all children met the current UK activity guidelines. Although these guidelines are comparable to Australian recommendations, previous estimates indicate that only 5.1% of Australian children met these guidelines . The proportion of time children spent active in this study (~67%) was also higher than reported previously in studies in Europe (~16%) [37, 38] and Australia(~16%) . Several factors, including the population studied, the measure of physical activity and wear and processing protocols, are likely to contribute to these differences across studies. These data were collected in a population-based sample, as opposed to a pre-school or care-setting based sample as is more commonly used [17, 38]. Moreover, previous studies have predominantly used Actigraph accelerometers worn on the hip during waking hours and taken off during water-based activities [17, 37, 38]. The waterproof Actiheart monitor, worn continuously for 24 hours each day, is therefore likely to have captured more of children’s daily activity. Further, the cut points used here, equivalent to those previous applied to Actigraph data [30, 31, 39], were derived experimentally and validated against doubly labelled water . As other studies have used higher cut points to classify active time [17, 37, 38], they are likely to report lower levels of activity. Taken together, varying study methods and lack of raw count per minute data to compare between studies , make comparison of prevalence estimates of young children’s activity, and their compliance with guidelines, problematic .
We observed that children spend the majority of their time in LPA. Provided that ≥100 cpm is a valid threshold for LPA [30, 31], and even accounting for differences in classification according to accelerometer cutpoints , children’s light activity levels are ‘sufficient’ to comfortably satisfy current activity guidelines. It may however be important to consider whether specifying activity intensity along with a total activity guideline is necessary for this age-group. LPA may confer limited health benefits for young children, if, as is the case in older children and adults, benefits to health are associated with more vigorous intensity activity [42, 43]. Indeed, in this sample of children, MVPA (and not LPA) has been found to be positively associated with bone density  and vigorous activity with lower fat mass . The relevance of this debate is further highlighted by the recent observation that more than 1 in 5 UK children are either overweight or obese on entering school at age 5 , which seems at odds with observations that all children are sufficiently active. The importance of activity intensity in younger children therefore requires further investigation to inform future activity guidelines for the under-5 s.
Increasingly, evidence suggests that even at a young age boys are more active than girls [17, 37, 47], which was also observed here. This study adds to the current literature by highlighting that these differences are not consistent throughout the day. Sex differences were most apparent in the mornings, with girls being relatively less active and more sedentary than boys. In contrast, children’s weight status did not appear to have an influence on children’s activity across the day. As similar finding have been seen previously in this age group , these patterns may be indicative of preschool-aged children’s activity across the day, and future interventions may need to consider a temporal focus to differentially target boys and girls.
Interestingly, children who attended nursery full-time compared to part-time were more sedentary and less active in the mornings. It has been suggested previously that childcare may influence children’s physical activity levels , but the measure used here provides only a basic idea of whether children were usually in childcare full- or part-time. As it was not possible to determine where children were during the measurement week, further work is required to determine whether activity levels differ by time spent in childcare and what factors in the childcare environment influence children’s activity.
The differences in daily patterns observed between week and weekend days are likely to reflect families’ normal working weeks, with children typically rising and going to bed earlier on weekdays. What is not clear is why children whose mothers left school at a later age were also more likely to be sedentary and less active in the evenings. As maternal employment status was not collected at the age four visit, it is not possible to determine whether this finding relates to more of these mothers being in full-time work. However, it has been shown previously that three-year-old children with more highly educated mothers are more likely to have a set bedtime and bedtime routine , which may plausibly account for the children of higher educated mothers being less active in the evenings here.
Although not studied directly, it is likely that the observed seasonal variation, with greater activity and less sedentary time in the evenings during the summer compared to winter, reflects the longer daylight hours of summer months. Contrary to previous observations in preschool children , children’s average daily MVPA was higher in spring compared to winter, which appears to be the result of a consistent accumulation of more activity throughout the day. Through this novel and detailed investigation of the temporal differences in behaviour, we are able to highlight opportunities to encourage activity in children, such as specifically targetting activity during the mornings for girls, or facilitating day-time activity opportunities for preschoolers during the autumn and winter months.
Strengths and limitations
Using data from a large population-based sample of four-year-old children, this study is one of the first to describe objectively measured physical activity levels of British preschool-aged children [50, 51], and highlight differences in activity across the day. As mothers were recruited before children were born, children were drawn from all socio-economic strata in the city of Southampton and surrounding areas. The sample is not therefore subject to biases seen in more commonly used (pre-)school-based samples, which by their nature exclude children who do not attend formal childcare (for sufficient amounts of time). In addition, children’s activity was measured throughout the year, rather than being bounded by school terms, allowing analysis of seasonal variation.
We included all children with at least one valid day of physical activity data to maximise our sample size and power. However, sensitivity analyses showed that restricting the sample to those with valid physical activity for ≥3 days or to those with at least one weekday and weekend day did not alter the results. We also found no significant differences between those who did and did not provide valid activity data at age 4, suggesting that the sample is likely to be representative of the Southampton study population and wider population. This said, participants were predominately white British in keeping with the Southampton region (~82%) , and fewer children in this sample were overweight or obese compared to the national average . Care should therefore be taken in generalizing these results to specific sub-populations, including those from ethnic minorities or populations with high levels of childhood overweight/obesity.
The novel use of time-stamped data, dividing days into three time periods, allows a more detailed description of the temporal patterns of children’s activity across the day. These segments reflect plausible subdivision of preschool-aged children’s days, with sessional care offered from Monday-Friday in the UK. Although most children in this sample attended childcare at least part-time, time-matched care attendance data was not available to determine what influence childcare attendance had on their physical activity levels. However, by segmenting the day in this way, clear times do appear during which public health interventions to increase activity in specific subgroups may be more likely to show a beneficial impact.
Actiheart monitors present a valid [21, 22] and feasible form of measuring young children’s activity, providing increased wear time and therefore enhanced characterization of children’s daily activity levels, in combination with a validated questionnaire . Validation studies have shown that the use of 60-second epochs, used here to allow sufficient memory to record for 7 days, may underestimate MVPA , whilst overestimating LPA . Whilst this may have led to attenuation of the associations found here, our results still suggest that factors influencing children’s sedentary, LPA and MVPA are likely to differ in the preschool-aged population.