Disclosure: The authors declared no conflict of interest.
Funding agencies: This work was supported by a National Health and Medical Research Council of Australia project grant (No. 533526), a Senior Research Fellowship (No. 425513 for L.M.S.), and a University of Gavle international research fellowship (for R.A.A.).
Sedentary behavior in children is positively associated with an increased risk of both obesity and insulin resistance. Children spend a considerable portion of their awake time in sedentary behavior; however, the pattern of accumulation is not known. Thus the objective of this study was to describe the patterning of sedentary behavior of children at and away from school.
Design and Methods:
The patterns of sedentary time in a sample of 53 children (28 girls) aged 10-12 years during school-term time were examined. Children wore an accelerometer for 1 week. Total sedentary time, prolonged sequences (bouts) of sedentary time, and frequency of active interruptions to sedentary were examined on school days and weekends and within school time and non-school time on school days.
The data did not support our hypothesis that children accumulated more sedentary time on school days when compared with weekend days (mean [SD]: 64.4% [5.3] vs. 64.9% [9.0], P = 0.686). However, when comparing school time with non-school time on school days, children accumulated more sedentary time at school (66.8% [7.3] vs. 62.4% [5.2], P < 0.001) and spent more time at school in sustained sedentary sequences, that is, uninterrupted sedentary time for 30 min or more (75.6 min [45.8] vs. 45.0 min [26.8], P < 0.002). The children also recorded less breaks per sedentary hour within school time when compared with non-school time (8.9 h−1 vs. 10.2 h−1, P < 0.001).
Reducing total sedentary time spent both in and out of school remains an important challenge. Interrupting sedentary time more often in the “working” (school) day could also reap important musculoskeletal and metabolic health rewards for children.
In adults, time in sedentary behavior has been shown to be an independent risk factor for obesity, diabetes, and all cause mortality (1). Sedentary behavior refers to any waking behavior characterized by an energy expenditure ≤1.5 METs while sitting or reclining (2). How sedentary time is accumulated has also been shown to relate to metabolic health, with interruptions to this very low-intensity activity associated with favorable effects independent of overall sedentary time (3). In children as young as 9 years, sedentary time has been associated with increased insulin resistance (4). Children can spend up to 9 h/day in activities of a sedentary nature (5), and although we know more about the type of sedentary activities that children engage in and their temporal patterns (5, 6), less is known about how sedentary time is actually accumulated—in short bouts or through extended uninterrupted sequences. Suggestions have been made that a child's exposure to sedentary time, particularly extended sedentary time, may be more pronounced in the school setting (7); however, there has been no research published to date that has objectively examined this. Here, we address the hypothesis that children are more sedentary and have less interruption to sedentariness on school days when compared with weekend days and in school time when compared with to non-school time. Thus, we provide novel information regarding the patterning of children's sedentary time.
We investigated the sedentary patterns of children during term time on school days and weekend days and within school days to examine school time and non-school time. The data were obtained from 66 children who were recruited by mass media and local school notices for a randomized controlled study (Australia and New Zealand Clinical Trials Registry ACTRN 12609000279224) investigating the effects of electronic games on physical activity (8). The data were obtained from their baseline assessment. Inclusion criteria for the study were children aged 10-12 years at the start of the study and able to play electronic games on most days, with exclusions being diagnosis of a movement disorder, unable to access games, and living remotely. Height and weight were measured, and the children were given an Actical accelerometer to wear on their hip for 7 days. The children were instructed to put on the accelerometer as soon as they awoke each morning and remove just before bed time. School time was taken as the time that schools (n = 48) officially started and finished, and it did not include the commute to school or after school care.
Data Reduction and Statistical Analyses
As per the established standard practices with accelerometry (9), data were individually checked, and only children with 4 valid days of accelerometry data (at least 10 h recording per day and at least one weekend day) were included for analyses. The cutoff of 100 cpm validated by Puyau et al. (10) was used to define sedentary time. Total sedentary time was calculated and also expressed as a percentage of total weartime. Patterning of sedentary time was calculated using exposure variation analysis (11), which for the purposes of this study was configured to give the duration of uninterrupted sequences (bouts) of sedentary time of 30 min or more. “Active” breaks in sedentary time, that is, sequences of any duration with an accelerometer count larger than 100 cpm, were counted and expressed as the average frequency of breaks per hour of total sedentary time, as suggested by Healy et al. (3).
Differences between the school days and weekend days and the school time and non-school time (on a school day) were examined with paired t-tests.
Thirteen children (eight boys and five girls, not different in age or BMI from the remaining 53) did not have sufficient valid accelerometry data. Fifty-three children (28 girls), mean age 11.2 years [SD 0.8], and BMI z-score of −0.04 [1.24] had valid accelerometry data, which included 243 school days and 90 weekend days. The average time of accelerometer recording per day was 822 min , that is, 13.7 h, slightly longer on school days than weekend days (835 min  vs. 809 min , P = 0.004). On a school day, recording of non-school time was significantly more than school time (460 min  vs. 375 min , P < 0.001).
Children spent approximately two-thirds of their day be it a school day (538 min, 64.4% [5.3] wear time) or weekend day (525 min, 64.9% [9.0] wear time, P = 0.686) in sedentary time. Moreover, children recorded more than 2 h in uninterrupted sedentary sequences lasting more than 30 min (132 min, 16.0% [7.7] on school days and 129 min, 15.9% [7.7] on weekend days, P = 0.692; see Figure 1. The number of breaks per sedentary hour was similar on a school day and on a weekend day (9.1 h−1 [1.6] vs. 9.4 h−1 [2.4], P = 0.304).
When school days were examined in more detail, children not only recorded proportionately more sedentary time during school time as opposed to non-school time (66.8% [7.3] vs. 62.4% [5.2] weartime, P < 0.001) but they also spent twice as much of their time in uninterrupted sedentary sequences longer than 30 min (20.2% [12.2] vs. 10.0% [6.3] weartime, P < 0.001). This extended sedentary time equated to just more than 75 min in school time alone. Children also recorded significantly less breaks per sedentary hour within school time when compared with non-school time (8.1 h−1 [2.1] vs. 10.2 h−1 [1.8], P < 0.001). Across all days, breaks per sedentary hour were highly negatively correlated with sustained sedentary behavior (r = −0.815, P < 0.001).
We found that children spent a considerable proportion of their day sedentary. Furthermore, more than 2 h of each day was spent in sustained sequences of sedentary time, where their accelerometers did not register any significant hip movement for 30 min or more. School time in particular was associated with more uninterrupted sedentary time and fewer breaks than non-school time; however, the total school-day pattern was similar to weekend days. If the findings observed in adults are true for children (3), such uninterrupted sedentary time may be having a detrimental effect on children's health.
The dilemma is how to intervene in an effective manner. Salmon (12) concluded that it may be more important to focus on the accumulation of sedentary time across the day, as opposed to targeting a narrow range of behaviors that occur in children's discretionary time. School offers great potential to intervene, as not only does this account for a considerable portion of a child's waking time but it also should be an environment where healthy habits are formed. If the very act of learning was less sedentary, then this may become more of an established norm as children progress into adulthood. Indeed, the tracking of sedentary behavior into adulthood has been suggested to be stronger than that of physical activity (13). Our study highlights for the first time the extent of prolonged sequences of uninterrupted sedentary time that a child engages in at school, just more than one-fifth of school time. Although we do not know if this has always been the case, or has increased over the years, its magnitude warrants that we find novel ways to educate our children without exposing them to extended sedentariness. The results from the Australian Transform-Us study, which aims to interrupt class time with light activity and standing lessons, will help to answer whether interrupting school sedentary time is sustainable and associated with improved health (14). Theoretically, in terms of energy expenditure alone, replacing 20 min of sedentary behavior (<1.5 MET) per day with light-intensity activity (∼2.5 MET) could result in an increase of approximately 10-30 kcal/day (depending on the child's body weight), which could have a significant impact on energy balance let alone other potential benefits of simply reducing sedentary behavior per se. Indeed, Reiff et al. (15) from their observations of caloric expenditure while sitting and standing suggested that merely standing for one-fifth of classroom time could result in an energy deficit of 20 kcal/day.
Although our sample size was small, the children came from a range of schools and socioeconomic circumstances, and we believe that they are fairly representative of 10- to 12-year-old children from affluent countries. Moreover, the study was originally powered to detect differences in total time spent in different activity intensities, including sedentary time. Although sedentary time was measured with an objective method, the study design neither allowed an analysis of the actual postures, tasks, and behaviors nor the influence of teacher, class, and school on behaviors. Further research on the individual and environmental determinants of sedentary time is needed to support informed interventions among school children.
Current advice for children to limit leisure screen-based activities to help reduce overall sedentary time remains essential; however, it is the very least we should be doing. We suggest that there is a need to look beyond this and propose minimizing not only total sedentary time but also sustained sequences of sedentary time during leisure and school. This will involve exploring the possibilities of changing the way children “work” at school. Replacing some of the daily sedentary time with light-intensity activity and breaking up sedentary time more often could reap important musculoskeletal and metabolic health rewards for children.
This work was supported by a National Health and Medical Research Council of Australia project grant (No. 533526) and Senior Research Fellowship (No. 425513 for LS) and a University of Gavle international research fellowship (for R.A.). The authors thank the children and families of the study and the project staff without whom this research would not have been possible.