Several studies have demonstrated an inverse association between objectively measured physical activity (PA) with markers of insulin resistance, hypertension, hyperlipidemia, and clustered metabolic risk in children (1,3,10). These associations are observed in healthy children and seem to be partly independent of adiposity (1,3,10) and aerobic fitness (10). Furthermore, PA seems to track from childhood to adulthood, which may influence on adult health (34). Taken together, this suggests that promotion of PA in childhood not only influences current health status but also may have long-term implications.
Previous studies using objective measurements of PA have suggested that boys are more active than girls (8,9,16,17,27,29,39), and gender differences have been shown already in preschool children (7). In addition, PA has been suggested to decline by age at least after 9 to 10 yr of age (2,5,15,26,31-33), but it is still unclear at what age this decline appears. A decline of PA in children younger than 9 yr has been observed when comparing younger age groups with older age groups (35,37,39). Studies using objective measurements of PA in children younger than 9 yr of age are scarce. Furthermore, the patterns of children's PA across days (13,16,17,38) and within days (11,19,20,22) are less well examined in younger children. This knowledge is important when identifying possible arenas and settings when planning preventive initiatives aimed at increasing children's levels of PA.
Therefore, the aim of this study was to examine differences in patterns of objectively measured PA among weekdays and weekend days and among different periods during the day and how these patterns are influenced by age and gender in a large sample of 6- to 10-yr-old Swedish children.
The study was approved by the Regional Ethical Review Board in Stockholm. Parent or guardian consent was obtained, and each child verbally agreed to participate in the study. All children included in this cross-sectional study participated in a randomized school-based obesity prevention program (Stockholm Obesity Prevention Project). The children were equally distributed between intervention (n = 653) and control schools (n = 640). Because no differences in PA levels between intervention and control schools were observed (18a), all children were pooled for the present study. A total of 1538 children (5-10 yr, grades 1-4) from 10 schools within the Stockholm County area provided measurements of PA for seven consecutive days. Twenty children were randomly selected and measured every school week (2002-2005). No children declined participation in the measurement of PA. Each child was measured once. Of these, 245 children (16%) were excluded because of invalid PA registrations or technical failures with activity monitors. Therefore, the present report includes 1293 children (653 girls and 640 boys). The physical characteristics of the children are shown in Table 1. In Sweden, most 6- to 10-yr-old children attend after school care centers from around 2:00 p.m. and leave between 4:00 and 5:00 p.m. The after school care centers provide both structured and unstructured activities. The children spend a great deal of time outdoors in an environment that stimulates free play and PA. Height was measured using a transportable Harpenden Stadiometer (Holtain, Crymych, United Kingdom), and weight was measured using a digital scale (Tanita BWB 800S; Tanita, Tokyo, Japan). Weight and height data were missing in 53 children. The baseline weights and heights for the children with missing weight and height showed no difference in terms of being more overweight or obese compared with the rest of the children's initial weights and heights (P = 0.17). Body mass index (BMI) was calculated as follows: weight divided by squared height (kg·m−2). BMI SD score (BMI SDS) was calculated according to a reference standard (30).
PA was assessed using the accelerometer Actiwatch® (AW, Model 4; Cambridge Neurotechnology Ltd, Cambridge, United Kingdom). The accelerometer is small, lightweight, and worn on the nondominant wrist like a watch. The wrist-worn AW was found to be feasible for long-term studies of PA in children (23). We have recently compared AW output against PA energy expenditure (PAEE) measured by indirect calorimetry in 8- to 10-yr-old children. AW counts explained 86% of the variance in PAEE during various free-living activities (unpublished results). PA was calculated as total counts divided with the total recorded time and expressed as counts per minute (CPM). Total PA (between 8:00 a.m. and 9:00 p.m.) was calculated during the 7-d period. Overall mean PA was also compared among weekdays and weekends (between 8:00 a.m. and 9:00 p.m.) and among different periods of the day. School time was defined as 8:00 a.m. to 1:30 p.m., after school care time between 1:30 and 4:00 p.m., and evening time between 4:00 and 9:00 p.m. We only included periods where more than 80% of the time included registered movement data in our analyses. All children (n = 1293) who provided at least 600 min of activity registration per day for a minimum of 4 d including one weekend day were included in the analyses. The mean (SD) number of included days was 6.4 (0.7) days, and the mean (SD) minutes of recorded time were 753 (17) minutes per day. Seventy children (5%) had one valid weekend day and 1223 children (95%) had two valid weekend days. In accordance with previous studies using accelerometry (27), data were excluded from the analyses if sequences during the waking hours indicated 10 or more consecutive zero counts assuming that the monitor was removed from the body. The epoch length was set to 1 min. The children were instructed to wear the AW for the whole week except while swimming and bathing. The accelerometers were distributed individually to each child on a weekly basis, evenly across gender and age groups and between intervention and control schools. The accelerometers were collected and redistributed every Tuesday. Therefore, Tuesdays were excluded for the analysis examining the weekly pattern. Because few children (n = 58) were 10 yr old at examination, these children were combined with 9-yr-old children into one age group. Similarly, 5-yr-old children (n = 15) were included in the 6-yr-old age group. We also categorized children gender specifically into tertiles of low, medium, and high activity levels on the basis of their mean total activity counts.
All data were analyzed using Statistica (8.0 Statsoft, Inc., Tulsa, OK), and the level of significance was set as P < 0.05. Descriptive statistics for participant characteristics are displayed as means and SD, and results of PA are presented as least square means and SE. Multivariate ANOVA was used to analyze PA with independent (between) factors and repeated measures (within factors). Where interaction or main effect showed statistical significance, post hoc comparisons using the least square difference test were performed.
In the first model, multivariate ANOVA with three independent factors was used to examine differences in mean weekly PA across gender, calendar years (2002, 2003, 2004, and 2005), and age groups of children (ages 6, 7, 8, and 9 yr).
In the second model, the same ANOVA model was used to analyze mean daily PA during weekdays and weekend days, respectively.
In the third model, multivariate ANOVA with a repeated-measures design was used to examine the differences in mean daily PA across days, i.e., the within factors Monday, Wednesday, Thursday, Friday, Saturday, and Sunday and the between factors calendar years, age (continuous), and gender.
In the fourth model, multivariate ANOVA with a repeated-measures design was used to evaluate the differences in mean daily PA among weekdays and weekends (the within factor) and differences among the independent factors calendar years, age (continuous), and gender. In addition, a factor with three activity levels (low, medium, and high) was included into model 4.
In the fifth model, multivariate ANOVA with a repeated-measures design was used to analyze the mean (period) PA using the within factors daily periods (8:00 a.m. to 1:30 p.m., 1:30-4:00 p.m., and 4:00-9:00 p.m.) and weekdays-weekend days and the between factors calendar year, age (continuous), and gender.
One-way ANOVA with one independent factor was used to study differences in mean daily PA among number of days used for measuring PA.
Means and SE for PA (CPM) for girls and boys are presented in Table 2. Overall, boys were 8% more active than the girls (P < 0.001, model 1). Mean weekly PA was significantly higher in boys compared with girls (F = 46.50, P < 0.001) in all age groups, ranging between 7% and 13%. Further, mean PA was inversely related to age (F = 10.34, P < 0.001) independent of gender (Fig. 1). No interaction effect could be detected between gender and age (F = 1.64, P = 0.18). Post hoc tests revealed that mean weekly PA was 9% greater in 6-yr-old girls and boys compared with 9-yr-old girls and boys (P < 0.001).
Mean daily PA was lower across ages for weekdays (F = 9.86, P < 0.001) and weekends (F = 6.82, P < 0.001), respectively (model 2). No interaction effect could be statistically demonstrated between gender and age during weekdays (F = 1.65, P = 0.18) and during weekends (F = 0.78, P = 0.51).
A significant interaction effect across days and gender (F = 3.31, P = 0.005) was observed where the gender difference in mean daily PA was greater during the days in the beginning of the week compared with days at the end of the week (Fig. 2; model 3).
Further, a significant interaction effect for mean daily PA among weekdays-weekend and gender was observed (F = 7.0, P = 0.008; model 4). Mean daily PA declined from weekdays to weekend days in boys (post hoc test: P = 0.001) and in girls (P = 0.001), and this decline was more pronounced in boys. A mean difference between weekdays and weekend days of 17% in boys and 15% in girls was observed. There was no interaction effect among weekdays-weekend and age (F = 0.68, P = 0.41).
The decline in mean daily PA among weekdays and weekends and activity groups was similar when categorizing children into tertiles of overall PA (19%, 13%, and 14% for girls and 24%, 18%, and 13% for boys in the low-, medium-, and high-activity groups, respectively). No interaction effect was observed among weekdays-weekend, gender, and activity groups (F = 1.94, P = 0.14).
A significant three-way interaction effect across gender, periods, and weekdays-weekends was observed for mean period PA (F = 16.07, P < 0.001; model 5; Fig. 3). Post hoc tests showed that across age groups, boys were approximately 13% more physically active than girls during school time (i.e., between 8:00 a.m. and 1:30 p.m., P < 0.001) and during after school care time (i.e., 1:30-4:00 p.m., P < 0.001). Further, there was no difference between boys and girls in activity levels between 4:00 and 9:00 p.m. on weekdays. During weekends, there was a significant difference between boys and girls for all examined periods: 8:00 a.m. to 1:30 p.m. (P < 0.001), 1:30-4:00 p.m. (P = 0.02), and 4:00-9:00 p.m. (P = 0.003). Post hoc test of mean PA also showed that boys (P < 0.001) and girls (P < 0.001) were significantly more active during after school care time compared with school time on weekdays.
There was no statistically demonstrated difference (F = 0.82, P = 0.48) in mean daily PA across the number of days (4-7 d) of PA measurement.
The prevalence of overweight and obesity in this group of children was 21.6%, calculated according to Cole et al. (6).
The results from this study suggest that overall PA declines already at the age of 6 yr. Boys are consistently more physically active than girls across age groups, i.e., between ages 6 and 9 yr, and this difference is most pronounced during school time and after school care time. Irrespective of overall PA levels, significantly lower levels of PA were observed during weekends compared with weekdays.
Previous studies have suggested a decline in PA by age between 9 and 18 yr (2,5,15,16,26,27,29,33) and across age groups consistently suggesting lower activity in older age groups (4,18,35,39). Recent data using accelerometry from the National Health and Nutrition Examination Survey suggest that activity levels are lower in adolescents compared with young children (37). A recent longitudinal study (n = 1032) reported that moderate and vigorous PA (MVPA) declined between ages 9 and 15 yr, and this decline was similar for girls and boys (21).The steepest decline in PA has previously been suggested to occur during adolescence, that is, between the ages of 13 and 18 yr (32), and a recent review suggested that PA levels increase between the ages of 3 and 8 yr (7). Our results extend these previous observations suggesting a decline in objectively measured overall PA levels across age groups between ages 6 and 9 yr. Part of the decline in overall PA levels might be explained by more scheduled time spent at school with increased age. However, this is probably not the most important factor because our results also suggested a decline in PA by age during weekends. This may at least partly suggest a biological explanation for the decline by age in spontaneous PA as observed in other species (32). However, it is also possible that increased use of computer and television and increased sedentary behaviors contribute.
Previous studies have reported higher levels of PA during weekdays compared with weekend days in 8- to 11-yr-old children from various countries (16,17,28). In a French study with primary schoolchildren (mean age 9 yr, n = 64), mean PA levels were lower during school-free days compared with school days (13). Similarly, MVPA was higher in adolescent girls (n = 1603) on weekdays compared with weekends (36). On the contrary, in another study, a subgroup of children in grades 1-3 (n = 92) showed higher levels of MVPA during the weekends compared with weekdays (38). Finally, compared with this present study, data from the Avon Longitudinal Study of Parents and Children (n = 5595) suggested similar activity pattern between weekdays and weekend days in 11-yr-old boys and girls (28). These results also suggested that the daily activity pattern was similar between most and least physically active children (highest and lowest quintiles). Interestingly, our study also showed that the weekly activity patterns were very similar between high- and low-activity groups. The magnitude of decline between weekdays and weekend days was similar across activity tertiles, suggesting that this decline in activity is consistent regardless of overall activity levels. Our results suggest that this pattern is evident already in 6- and 7-yr-old children.
Our results corroborate previous observations consistently demonstrating that boys are more physically active than girls (9,16,17,27,29,39), but to our knowledge, only one previous study (n = 698) that used accelerometers to measure PA has demonstrated a gender difference among 6- to 7-yr-old children (9). The gender difference in mean PA in the Danish study was comparable with that of the present results.
Gender differences were explained by significantly higher activity levels in boys during school and after school care time. In contrast, there was no significant difference in PA between girls and boys in the evenings on the weekdays. PA levels during school time versus leisure time were recently examined in 9- and 15-yr-old children suggesting that sociocultural factors may influence on the amount of activity (22). A small study (n = 58) including 7- to 11-yr-old children from England showed that PA behaviors were more consistent in school compared with the period after school (11), whereas 8- to 15-yr-old Portuguese girls were more physically active during school periods while boys were more active after school (20). Further, a small study (n = 54) studied activity patterns within days between active and less active 10- to 15-yr-old girls (19). They observed that the active girls were significantly more engaged in MVPA outside school compared with the less active girls.
It is unclear whether the gender differences are because the school environment and physical education classes are more adapted toward boys or if the difference mirrors the biological or social variations in social interaction among boys and girls. Boys, but not girls, are physically active if play areas in schools are improved, suggesting different play styles between girls and boys (14). The gender differences in PA that were most pronounced during school time and after school time might be important information when planning interventions to increase girls' PA.
It was recently concluded that the amount of PA accumulated during school time seemed not to be enough to achieve the current recommendations for health-enhancing PA in children (12). We found an activity pattern with low activity levels during leisure time in children as young as 6-7 yr, and it is therefore possible that this is a relevant concern also for young children. It is likely that most children in this age group spend most of their leisure time with their family. Therefore, family-involved interventions may be a prerequisite to increase PA in this age group, although the results of family interventions are so far discouraging (40).
We used a wrist-worn activity monitor for assessing overall levels of PA in our study. Although speculative, one possible advantage with the AW compared with hip-worn accelerometers is that a high compliance is easier to achieve with a wrist-worn accelerometer. The children rarely took the AW off because it is not obstructing their natural way of living and thereby diminishing the possibility of forgetting to put the monitor back on. The validity of the AW monitor placed at the hip or lower leg has previously been examined in children with correlation coefficients ranging from 0.66 to 0.89 (24,25).
The following limitations should be considered when interpreting our results. First, because of the cross-sectional design, we cannot conclude an age-related decline in PA, only differences across age groups. Second, we did not examine time spent at different intensity levels, which may also be of importance when examining the PA patterns in children. This study is a cohort analysis of children being part of an obesity intervention. However, it is unlikely that this will bias our results because the intervention did not produce a significant difference in overall mean PA between intervention and control schools (18a).
Our results suggest a difference in objectively measured PA across ages between 6 and 9 yr. PA levels were consistently lower during weekends and evening time compared with school times. These activity patterns were similar across low, medium, and highly active children. Girls were consistently less active than boys, but the differences were most pronounced during school and after school care time. Weekends and evenings when children's PA levels are disproportionally low might be good targets for interventions aimed to increase PA in young children. Therefore, family involvement may be important for a successful PA intervention program in this age group.
We would like to thank all the participating schools. The authors thank the Stockholm County Council, Swedish Council for Working Life and Social Research, Swedish Research Council, Freemason's in Stockholm Foundation for Children's Welfare, and Signhild Engkvist Foundation for funding this study and Jan Kowalski for his statistical expertise and support.
Conflict of interest statement: We declare that we have no conflict of interest.
The results of the present study do not constitute endorsement by ACSM.
1. Andersen LB, Harro M, Sardinha LB, et al. Physical activity and clustered cardiovascular risk in children: a cross-sectional study (the European Youth Heart Study). Lancet
2. Armstrong N, Welsman JR, Kirby BJ. Longitudinal changes in 11-13-year-olds' physical activity. Acta Paediatr
3. Brage S, Wedderkopp N, Ekelund U, et al. Features of the metabolic syndrome are associated with objectively measured physical activity and fitness in Danish children: the European Youth Heart Study (EYHS). Diabetes Care
4. Butte NF, Puyau MR, Adolph AL, Vohra FA, Zakeri I. Physical activity in nonoverweight and overweight Hispanic children and adolescents. Med Sci Sports Exerc
5. Caspersen CJ, Pereira MA, Curran KM. Changes in physical activity patterns in the United States, by sex and cross-sectional age. Med Sci Sports Exerc
6. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ
7. Dencker M, Andersen LB. Health-related aspects of objectively measured daily physical activity in children. Clin Physiol Funct Imaging
8. Dencker M, Thorsson O, Karlsson MK, et al. Daily physical activity in Swedish children aged 8-11 years. Scand J Med Sci Sports
9. Eiberg S, Hasselstrom H, Gronfeld V, Froberg K, Svensson J, Andersen LB. Maximum oxygen uptake and objectively measured physical activity in Danish children 6-7 years of age: the Copenhagen school child intervention study. Br J Sports Med
10. Ekelund U, Anderssen SA, Froberg K, Sardinha LB, Andersen LB, Brage S. Independent associations of physical activity and cardiorespiratory fitness with metabolic risk factors in children: the European Youth Heart Study. Diabetologia
11. Fairclough SJ, Butcher ZH, Stratton G. Whole-day and segmented-day physical activity variability of northwest England school children. Prev Med
12. Floriani V, Kennedy C. Promotion of physical activity in children. Curr Opin Pediatr
13. Gavarry O, Giacomoni M, Bernard T, Seymat M, Falgairette G. Habitual physical activity in children and adolescents during school and free days. Med Sci Sports Exerc
14. Harten N, Olds T, Dollman J. The effects of gender, motor skills and play area on the free play activities of 8-11 year old school children. Health Place
15. Kimm SY, Glynn NW, Kriska AM, et al. Longitudinal changes in physical activity in a biracial cohort during adolescence. Med Sci Sports Exerc
16. Klasson-Heggebo L, Anderssen SA. Gender and age differences in relation to the recommendations of physical activity among Norwegian children and youth. Scand J Med Sci Sports
17. Kristensen PL, Korsholm L, Moller NC, Wedderkopp N, Andersen LB, Froberg K. Sources of variation in habitual physical activity of children and adolescents: the European Youth Heart Study. Scand J Med Sci Sports
18. Lopes VP, Vasques CM, Maia JA, Ferreira JC. Habitual physical activity levels in childhood and adolescence assessed with accelerometry. J Sports Med Phys Fitness
18a. Marcus C, Nyberg G, Nordenfelt A, Karpmyr M, Kowalski J, Ekelund U. A 4-year cluster-randomized, controlled childhood obesity prevention study: STOPP. Int J Obes
19. Mota J, Santos MP, Ribeiro JC. Differences in leisure-time activities according to level of physical activity in adolescents. J Phys Act Health
20. Mota J, Santos P, Guerra S, Ribeiro JC, Duarte JA. Patterns of daily physical activity during school days in children and adolescents. Am J Hum Biol
21. Nader PR, Bradley RH, Houts RM, McRitchie SL, O'Brien M. Moderate-to-vigorous physical activity from ages 9 to 15 years. JAMA
22. Nilsson A, Anderssen SA, Andersen LB, et al. Between- and within-day variability in physical activity and inactivity in 9- and 15-year-old European children. Scand J Med Sci Sports
23. Nyberg G, Ekelund U, Marcus C. Physical activity in children measured by accelerometry: stability over time. Scand J Med Sci Sports
24. Puyau MR, Adolph AL, Vohra FA, Butte NF. Validation and calibration of physical activity monitors in children. Obes Res
25. Puyau MR, Adolph AL, Vohra FA, Zakeri I, Butte NF. Prediction of activity energy expenditure using accelerometers in children. Med Sci Sports Exerc
26. Riddoch C, Savage JM, Murphy N, Cran GW, Boreham C. Long term health implications of fitness and physical activity patterns. Arch Dis Child
27. Riddoch CJ, Andersen LB, Wedderkopp N, et al. Physical activity levels and patterns of 9- and 15-yr-old European children. Med Sci Sports Exerc
28. Riddoch CJ, Mattocks C, Deere K, et al. Objective measurement of levels and patterns of physical activity. Arch Dis Child
29. Rizzo NS, Ruiz JR, Hurtig-Wennlof A, Ortega FB, Sjostrom M. Relationship of physical activity, fitness, and fatness with clustered metabolic risk in children and adolescents: the European Youth Heart Study. J Pediatr
30. Rolland-Cachera MF, Sempe M, Guilloud-Bataille M, Patois E, Pequignot-Guggenbuhl F, Fautrad V. Adiposity indices in children. Am J Clin Nutr
31. Sallis JF. Epidemiology of physical activity and fitness in children and adolescents. Crit Rev Food Sci Nutr
32. Sallis JF. Age-related decline in physical activity: a synthesis of human and animal studies. Med Sci Sports Exerc
33. Telama R, Yang X. Decline of physical activity from youth to young adulthood in Finland. Med Sci Sports Exerc
34. Telama R, Yang X, Viikari J, Valimaki I, Wanne O, Raitakari O. Physical activity from childhood to adulthood: a 21-year tracking study. Am J Prev Med
35. Thompson AM, Campagna PD, Rehman LA, Murphy RJ, Rasmussen RL, Ness GW. Physical activity and body mass index in grade 3, 7, and 11 Nova Scotia students. Med Sci Sports Exerc
36. Treuth MS, Catellier DJ, Schmitz KH, et al. Weekend and weekday patterns of physical activity in overweight and normal-weight adolescent girls. Obesity (Silver Spring)
37. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc
38. Trost SG, Pate RR, Freedson PS, Sallis JF, Taylor WC. Using objective physical activity measures with youth: how many days of monitoring are needed? Med Sci Sports Exerc
39. Trost SG, Pate RR, Sallis JF, et al. Age and gender differences in objectively measured physical activity in youth. Med Sci Sports Exerc
40. van Sluijs EM, McMinn AM, Griffin SJ. Effectiveness of interventions to promote physical activity in children and adolescents: systematic review of controlled trials. BMJ
Keywords:©2009The American College of Sports Medicine
ACTIWATCH; MOTION SENSORS; PHYSICAL ACTIVITY LEVELS; SCHOOL