Share this article on:

Longitudinal Study of Physical Activity and Inactivity in Preschoolers: The FLAME Study

Taylor, Rachael W.1; Murdoch, Linda2; Carter, Philippa2; Gerrard, David F.3; Williams, Sheila M.4; Taylor, Barry J.2

Medicine & Science in Sports & Exercise: January 2009 - Volume 41 - Issue 1 - p 96-102
doi: 10.1249/MSS.0b013e3181849d81
Basic Sciences

Purpose: To investigate patterns of activity and inactivity in a birth cohort of children followed from 3 to 5 yr and to investigate whether changes in activity occurred over time.

Methods: Two hundred and forty-four children (44% female) were seen annually at 3, 4, and 5 yr. Physical activity and inactivity was measured by questionnaire (parent-proxy) and by Actical accelerometers for five consecutive days (24-h monitoring) each year in children and once in each parent for 7 d (69% with data).

Results: Retention of participants was high (92%). Viable accelerometry data were obtained for 76-85% of children at each age. Reliability estimates ranged from 0.80 (3 yr) to 0.84 (5 yr). Day of the week, season, sex, hours of childcare, or birth order did not affect daily average accelerometry counts (AAC) at any age. Parental activity correlated weakly with the child's activity at 3 and 4 yr (r values = 0.17-0.28), but only the father's activity remained a significant predictor of the child's activity after adjustment for confounders. Children spent approximately 90 min·d−1 in screen time (television, videos, DVD, and computers) with an additional 90 min in other sedentary activities (reading, drawing, and music). Physical activity was significantly reduced at 4 and 5 yr compared with 3 yr in both sexes, whether measured as AAC (24-h data, awake time only, weekend days, weekdays), time in moderate or vigorous activity, or from parental reports of activity.

Conclusion: Levels of physical activity declined in boys and girls between the ages 3 and 4-5 yr, whether using objective measures or parental reports of activity.

1Edgar National Centre for Diabetes Research, 2Department of Women's and Children's Health, 3Dunedin School of Medicine, and 4Department of Preventive and Social Medicine, University of Otago, Dunedin, NEW ZEALAND

Address for correspondence: Rachael Taylor, Ph.D., B.Sc., Edgar National Centre for Diabetes Research, c/o Department of Human Nutrition, University of Otago, PO Box 56, Dunedin; E-mail:

Submitted for publication August 2007.

Accepted for publication June 2008.

Recent rapid increases in the prevalence of obesity in preschool-aged children (38) has led to the call for further research in this understudied age group (7). It is increasingly acknowledged that physical activity and inactivity may represent distinct constructs with respect to body composition and health (26). Accurate assessment of physical activity and inactivity is complicated in this age group due to the intermittent nature of activity and the lack of structured play (15). However, use of objective measures such as accelerometry may obviate the need for surrogate measures by parents that may have limited reliability or validity.

Although several studies have used accelerometry to determine various components of activity in preschool-aged children (3,5,19,23,36), few have provided longitudinal data to investigate changes in activity over time (10,39). Jackson et al. (10) reported that activity levels increased in a small group of 3-yr-old children followed up at 4 yr, whereas no change in activity was observed in slightly older children followed for 1 yr (4.9 yr at baseline) (39). Interest in sedentary behavior has focused traditionally on television viewing in preschoolers and older children (9,37), despite the observation that television time may not provide an adequate representation of sedentary activity during growth (16). Little appears to be known about other sedentary activities chosen by preschool-aged children and whether these change with age (34).

This study used the waterproof Actical accelerometer, which has been validated for use in preschool-aged children (24) to investigate patterns of activity and inactivity in a birth cohort of children followed from 3 to 5 yr and to investigate whether changes in activity level occurred.

Back to Top | Article Outline



The Family Lifestyle, Activity, Movement, and Eating (FLAME) study recruited children just before their third birthday from a birth cohort born in Dunedin, New Zealand, at Queen Mary Maternity Unit. All births in Dunedin city (population 120,000) apart from home births (<3%) occur in this unit. The birth cohort was born between 19 July 2001 and 14 January 2002, and the families of these children were all resident in Dunedin. Exclusion criteria were gestation less than 37 wk, major congenital abnormalities, severe postnatal illness, multiple births, or children from families unlikely to be resident in Dunedin for the next 2 yr. After exclusion, 413 children were eligible of which 244 (44% female) agreed to participate (59% response rate). There were no differences in sex, gestation, or birth weight of participating and nonparticipating children. In addition, mothers were of similar age and ethnicity, but participating mothers were more socially advantaged (data not shown). The children were predominantly Caucasian (87%), with 10.8% identifying as Maori and 3.7% as Pacific Islanders, comparable with the wider Otago region (2001 Statistics New Zealand). The study was approved by the Lower South Regional Ethics Committee (Reference OTA/04/03/023), and signed informed consent was obtained from the parents or guardians of every participating child.

Participants were seen annually at 3, 4, and 5 yr as close as possible to their birthdays. At 5 yr, most children were seen just before starting school. At baseline, age at first visit ranged from 2.96 to 3.15 yr, with 86% of participants being seen within 4 wk of their birthday. At 4 and 5 yr, 97% and 99% of children attended clinic visits within 4 wk of their birthday.

Back to Top | Article Outline


Height was measured in duplicate to the nearest 0.1 cm using an electronic wall-mounted stadiometer (Heightronic; QuickMedical, Northbend, WA), and weight was measured by electronic scales (Toledo, Columbus, OH) in duplicate to the nearest 0.1 kg. Body mass index (BMI) was calculated (kg·m−2) and BMI-for-age z-scores were calculated using US reference data (14). Waist circumference was measured immediately above the iliac crest, directly against the skin following the procedures used in the recent National Children's Nutrition Survey (21). The reliability or intraclass correlations for height, weight, and waist circumference in our laboratory are 0.999, 0.999, and 0.993, respectively. Physical activity was measured using Mini-Mitter (Bend, OR) omnidirectional Actical accelerometers (24). The accelerometers were placed on the child at the clinic, and the parents were instructed to keep the monitors on at all times for five consecutive days. Additional belts were provided so that the child could wear the monitors while bathing and swimming. Activity counts were analyzed for 24-h periods and also during waking hours. Waking hours were calculated from parental reports of bed and awake times and from visual inspection of the accelerometry data. Due to calibration difficulties discovered after data collection, accelerometry data were only available for 85% of children at 3 yr, 76% of children at 4 yr, and 83% of children at 5 yr. Minutes per day spent in moderate or vigorous activity was calculated using the 0.04-kcal·kg−1·min−1 cut-off provided by Actical, which corresponds to the moderate cut-off (20 ml·kg−1·min−1 or 715 counts·15 s−1) developed by Pfeiffer et al. (24).

Physical and sedentary activities were also assessed by structured written questionnaires developed specifically for the FLAME study. Parents were asked to estimate the amount of time spent each week in various specific activities (dancing, biking, playground, walking, organized activities, trampoline, swimming, and others). These data was summed to give total reported activity (min·d−1). Parents were asked to rate their child's level of activity relative to other children of the same age and sex on a 10-point scale. Time spent watching television, videos, or DVD and using a computer was assessed separately then summed to provide an indication of total screen time. Time spent in other sedentary activities including reading, music, and art was also assessed and when combined with screen time gave total sedentary activity. Parents also indicated the number of hours per week that children attended childcare (kindergarten, home-based care, crèche), the number of siblings living in the same household, and the birth order of the child.

Physical activity was also assessed in both parents by using 7-d accelerometry (data available on 166 mothers and 170 fathers) and by questionnaire (perceived activity rating relative to others of the same age and sex on a 10-point scale, N = 207 mothers and 213 fathers).

Back to Top | Article Outline


Data were analyzed using STATA 9.0 (StataCorp LP, College Station, TX). Random coefficient models with participant as a random effect were used to analyze the repeated measures at each assessment and data for all three occasions simultaneously. These models take into account the underlying correlations between repeated observations. Models including terms for day of the week were used to examine differences between days of the week and weekdays and weekends for the replicated data collected on each occasion. Intraclass correlation coefficients (ICC), the ratio of the between-participant variance to the total variance, were estimated from data for each occasion. The ICC can have values between 0 and 1, with values close to 1 indicating good reliability or repeatability. Models, which included terms for age and sex, were used to examine differences between girls and boys and changes over time. Interaction terms between age and sex were considered. Further models examined the effect of father's and mother's activity, BMI z-score at age 3, and time awake on overall activity. Spearman rank-order correlations determined relationships between parental and child activity. The correlations between the accelerometry measures were adjusted for their reliability (6).

Back to Top | Article Outline


Table 1 demonstrates that retention of participants was high, with 97% of the original sample returning at 4 yr and 92% of both sexes attending clinic visits at 5 yr. There were no significant differences in baseline BMI z-score or waist circumference between dropouts (n = 20) and remaining participants (n = 224, data not shown). Anthropometric variables were similar in boys and girls at every age, except for a slightly higher weight in 3-yr-old boys compared with girls (Table 1).



Ninety-three percent of 3-yr-olds, 86% of 4-yr-olds, and 96% of 5-yr-olds wore the monitors for at least 5 d. The remaining children wore the monitors for 3 d (n = 5) or 4 d (n = 42) and were included in the data set. Daily activity did vary among individual children; within-child variation was large at each age (70-148 average counts) and similar to the between-child variation (60-132 average counts). ICC calculating the degree of association between different days of measurement were r = 0.45, 0.42, and 0.46 for ages 3, 4, and 5, respectively, indicating that the day-to-day variation in accelerometry counts within individuals was similar in magnitude to the variation in the whole sample. Replication meant that the reliability estimates showed that the reliability of the accelerometry data based on the mean of the observations for each child was acceptable, with estimates of 0.80 at age 3 (4.9 d of measurement), 0.79 at age 4 (5.1 d), and 0.84 at age 5 (6.1 d).

However, children were not more active on certain days of the week at any age (P = 0.940 at age 3, P = 0.233 at age 4, and P = 0.135 at age 5). Similarly, children did not have different levels of activity on weekdays compared with weekends at any age (P > 0.05; Table 2).



Table 2 also highlights that no gender difference was apparent in average accelerometry count (AAC), either as 24-h values or awake hours, at any age. Activity counts during sleep were small and varied little among individuals; the sleeping AAC were more than 30 in fewer than 10 children at each age (Table 2).

There was no consistent effect of season on activity levels. Linear regression demonstrated that season explained 8.7% of the variance in AAC at 3 yr (P < 0.001) with children being significantly less active in the spring (September to October) than they were in summer (November to February) or winter (June to August, P = 0.001). However, season was not a significant predictor of activity counts at either 4 (P = 0.974) or 5 (P = 0.383) yr.

Parental activity was only weakly associated with children's activity as measured by AAC at 3 and 4 yr but not 5 yr (r = 0.18, P = 0.034 and r = 0.17, P = 0.051 for mothers and r = 0.28, P = 0.001 and r = 0.23, P = 0.007 for fathers for 3 and 4 yr, respectively). Significant correlations between parental ratings of their own and their child's activity were only observed between mothers and children at 4 (r = 0.21, P = 0.003) and 5 (r = 0.18, P = 0.014) yr. Neither the number of hours per week childcare attendance (P = 0.069-0.806) nor the birth order (P = 0.161-0.704) influenced AAC or parental estimates of time spent active at any age. The number of siblings living in the same household was weakly correlated with average counts at age 5 only (r = 0.156, P = 0.033), but the difference was effectively negligible at 15 counts per sibling.

Table 3 describes the time spent in various sedentary activities as well as the total amount of time parents estimated children were active each day. Children spent approximately 90 min each day on accumulated screen time plus an additional 70-90 min·d−1 on other sedentary activities such as reading, music, and drawing. The sex differences were not significant. Although 5-yr-old girls used the computer less than similar-aged boys, total time in both sexes was very minimal. However, 3-yr-old girls spent 23 more minutes each day on sedentary activities compared with similar-aged boys, which was due to greater amounts of time on drawing and music rather than screen time (data not shown).



Regardless of the choice of activity measure, significant declines in physical activity were observed as the children grew older. AAC for all measures (24-h data, awake time only, weekend or weekdays) were significantly lower at 4 and 5 yr compared with that observed at 3 yr, with no further decline occurring between 4 and 5 yr (Table 2). Average counts at ages 3 and 5 are demonstrated in Figure 1. Similarly, mean time in moderate or vigorous activity declined significantly by approximately 50% between 3 and 4 yr and remained significantly lower at age 5 (Table 2). Parental reports of activity supported these independent observations. Reported active time decreased from 81 min·d−1 at 3 yr to 72 min in children aged 4 (P = 0.052) with a further decrease to 57 min·d−1 at 5 yr (P = 0.001, Table 3). Similarly, when parents were asked to rate their child's activity level on a 10-point scale, they rated their children as significantly more active at 3 yr (6.5) than they did at 4 (6.1, P < 0.001) or 5 (6.2, P = 0.001) yr. Total sedentary time remained similar in 3- and 4-yr-old children but decreased significantly between 4 and 5 yr (P = 0.005; Table 3).



Table 4 presents the tracking coefficients between various measures of sedentary and active time obtained from questionnaires and accelerometry data. Although all variables tracked significantly over the 2 yr of the study, time in sedentary activities appeared to track more strongly than time spent physically active, with similar results in boys and girls (data not shown). Correlations between questionnaire data on activity and accelerometry data were r = 0.094 (P = 0.182), r = 0.267 (P < 0.001), and r = 0.158 (P = 0.039) for parental rating of activity and average counts at 3, 4, and 5 yr, respectively, and r = 0.114 (P = 0.102), r = 0.116 (P = 0.133), and r = 0.080 (P = 0.282) for total time in activity and average counts at 3, 4, and 5 yr respectively. Total time in activity and parental rating of activity was correlated at each age; r = 0.344, 0.245, and 0.342 (all P values <0.001) at 3, 4, and 5 yr.



A mixed model was used to estimate the association between factors that might influence overall activity counts (using all data points). Sex (P = 0.436), weight status (P = 0.591), and time awake (P = 0.372) did not influence overall activity counts. Although the relative effect of the mother's activity was twice than that of the father's (β (SE) values of 0.21 (0.12) compared with 0.11 (0.05), respectively), only father's activity reached statistical significance (P = 0.024, mother's P = 0.062). The reduction in accelerometry counts between 3 and 4-5 yr remained significant (P = 0.030) after adjustment for these confounders.

Back to Top | Article Outline


Our data demonstrate that season and day of the week do not affect activity levels in preschool-aged children, and sex differences in activity are not apparent. However, significant declines in total and moderate or vigorous activity are apparent between 3 and 4-5 yr and occur in both sexes. Parental activity, particularly that of the father, was a weak predictor of a child's activity. Preschool-aged children are spending approximately 90 min·d−1 engaged in screen time and a further 90 min in other sedentary activities such as reading, drawing, and music.

Although some (29) but not all (28) studies in older children have found higher activity levels in the warmer months, our results support the single study in preschool-aged children that reported no seasonal differences in total activity counts (5). It may be that season is less of an influence on activity in younger children, although outdoor play time has been related to activity at this age (3), because their play tends to be more spontaneous (15) and they are not typically involved in organized sport, an identified determinant of activity in adolescents (2). The number of hours per week in childcare did not influence activity at any age in our sample. The only comparable study (5) reported that childcare center was an important predictor of activity in preschool children. However, this latter study only evaluated children in childcare settings during the childcare day, whereas the current study reported overall activity and included children who did not attend any childcare facility. It is feasible that the number of children in a given household may affect the level of activity in preschool-aged children, given the potential for role modeling and the influence of having a readily available playmate. However, although activity counts at age 5 were associated with the number of siblings, even once adjusted for birth order (data not shown), the effect was negligible. Studies in older children (5-6 and 10-12 yr) have shown that girls, but not boys, with siblings participate in more moderate to vigorous activity per day than girls who are an only child (1).

Whether participation in physical activity differs according to the day of the week appears to be somewhat variable in the literature. Most studies in school-aged children utilizing objective measures have reported lower activity in the weekend compared with weekdays (4,8,40) with one exception where greater activity on weekend days was reported in 7-yr-old children (35). Moreover, the relationship is not consistent across all age groups within individual studies (8,35). Reports in preschool-aged children are scant and report strong intraclass correlations between different days with no variation in weekday or weekend activity (10,39) consistent with this study. These findings support the observations from the Earlybird study (39) that variation in activity at this age is influenced less by the environment and more by the internal factors.

A recent review highlighted that less than 40% of studies report a significant relationship between parental and child activity (32). Limited studies have been undertaken in preschool-aged children, with some (20,31) but not all (30) reporting parental activity as a positive influence. No longitudinal studies appear to have used objective measures of physical activity to examine this potential. In our study, parental activity was only a weak determinant of the child's activity in cross-sectional analyses. However, multiple regression analysis demonstrated that father's activity remained a small but significant predictor of the child's activity once adjusted for multiple confounders. The observed weak relationship between parental and child activity may be a consequence of changes in social structure and familial relationships decreasing the strength of any observed effect (18).

No clear explanation is offered for the observed lack of sex differences in activity. Studies in older children, almost without exception (40), show that boys are more active than girls (4,12,27,33). Similarly, the few existing studies in preschool-aged children also report significant sex differences favoring boys (5,10,13). However, the results from our study consistently demonstrated that activity counts were not higher in boys compared with girls at any age or for any measure. Similarly, parental reports of activity time did not differ between sexes.

Accelerometry outputs demonstrate that preschool-aged children spend a large proportion of the day in sedentary activities (25). We found that a significant portion of the day was spent in screen time, even at the age of 3 yr, consistent with the existing literature (17,37). However, little data are available describing how young children spend the rest of their sedentary time (34). We found that 3- to 5-yr-old children spent another 70-90 min·d−1 drawing, listening to music (not dancing), and reading. Interestingly, sedentary time appeared to track more consistently over time than active time, consistent with that observed in older children (11).

Limited longitudinal data exist describing how patterns of activity may change in young children assessed by objective measures of activity. Jackson et al. (10) reported that AAC increased by 27% from 3 to 4 yr. However, two other studies report no time difference in 3- to 4-yr-old children followed for 3 yr (22) or a large sample of children aged 4.9 yr at baseline followed for 1 yr (39). In our sample, activity clearly decreased at 4 and 5 yr compared with that obtained at 3 yr, whether measured as 24-h data, awake time only, in terms of weekdays or weekends, and the amount of time spent engaging in moderate or vigorous activity. Moreover, parental reports of activity support the observed decline in activity counts, suggesting that the decrease was genuine. Neither initial weight status nor sex explained these changes. Baseline activity was important, but father's activity remained a small but significant predictor of later activity once confounders were adjusted for. It is difficult to explain the variation in results observed in the literature. Our study is a relatively large birth cohort, displays a reasonable response rate, and reports an excellent retention with 92% of the original sample still participating at 5 yr. In contrast, while representative, Jackson et al. (10) had a smaller sample, with lower retention and 3 d of accelerometry data. Our data support those of Wilkin et al. (39) who demonstrate no change from 4.9 to 5.9 yr. It is possible that our study may also support the remaining study (22), which observed no difference in activity between 3-4 and 6-7 yr. Although we observed a decline between 3 and 4 yr, children in New Zealand start to be involved in sporting activities from 6 yr. It is possible that data collected at this time may show an increase back to preschool levels. We also collected 5 d of accelerometry data, which demonstrated high reliability and should provide an adequate description of activity in our sample (35). Also, the use of waterproof Actical monitors meant that aquatic activity including swimming and bathing was captured (24). We also collected data for 24-h periods rather than just awake times, so that monitoring time was consistent for each child.

The number of children with viable accelerometry data was limited by a calibration error that was discovered after data collection was completed. Although data were obtained from 201 to 233 (86-95%) children at each age, 15 to 30 cases (7-14%) were excluded from the analysis due to unacceptably high readings relating to specific monitors. Rather than limiting analyses to only those children with valid data at every age of measurement (3, 4, and 5 yr, n = 135), we have presented data for all children with valid measurements (n = 180-208) at any time. However, children who had viable accelerometry data at every age (n = 135) did not differ from the remaining participants in terms of BMI z-score at any age, sex distribution, or reported physical activity (data not shown).

In conclusion, longitudinal analysis of our data set demonstrated that activity levels declined in both sexes between the ages 3 and 4-5 yr, whether using objective measures or parental reports of activity. Although the predominant determinant of later activity was baseline activity, the father's activity remained a small but significant predictor. Further follow-up of our data set is planned to ascertain what differences in physical activity and sedentary time might become apparent once this cohort starts school.

The authors thank the children and their families for taking part in the FLAME project. We acknowledge funding from the Child Health Research Foundation, The National Heart Foundation of New Zealand, The Caversham Foundation, an Otago University Research Grant, HS & JC Anderson Trust, the Deans Bequest, a University of Otago Equipment Grant, and the Freemasons of New Zealand Paediatric Fellowship.

Back to Top | Article Outline


1. Bagley S, Salmon J, Crawford D. Family structure and children's television viewing and physical activity. Med Sci Sports Exerc. 2006;38(5):910-8.
2. Biddle SJ, Gorely T, Stensel DJ. Health-enhancing physical activity and sedentary behaviour in children and adolescents. J Sports Sci. 2004;22(8):679-701.
3. Burdette HL, Whitaker RC, Daniels SR. Parental report of outdoor playtime as a measure of physical activity in preschool-aged children. Arch Pediatr Adolesc Med. 2004;158(4):353-7.
4. Duncan JS, Schofield G, Duncan EK. Pedometer-determined physical activity and body composition in New Zealand children. Med Sci Sports Exerc. 2006;38(8):1402-9.
5. Finn K, Johannsen N, Specker B. Factors associated with physical activity in preschool children. J Pediatr. 2002;140(1):81-5.
6. Fleiss JL. The Design and Analysis of Clinical Experiments. New York: John Wiley & Sons; 1986. pp. 1-448.
7. Flynn MA, McNeil DA, Maloff B, et al. Reducing obesity and related chronic disease risk in children and youth: a synthesis of evidence with 'best practice' recommendations. Obes Rev. 2006;7(Suppl 1):7-66.
8. 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. 2003;35(3):525-31.
9. Gorely T, Marshall SJ, Biddle SJ. Couch kids: correlates of television viewing among youth. Int J Behav Med. 2004;11(3):152-63.
10. Jackson DM, Reilly JJ, Kelly LA, Montgomery C, Grant S, Paton JY. Objectively measured physical activity in a representative sample of 3- to 4-year-old children. Obes Res. 2003;11(3):420-5.
11. Janz KF, Burns TL, Levy SM. Tracking of activity and sedentary behaviors in childhood: the Iowa Bone Development Study. Am J Prev Med. 2005;29(3):171-8.
12. Janz KF, Burns TL, Levy SM, et al. Everyday activity predicts bone geometry in children: the Iowa Bone Development Study. Med Sci Sports Exerc. 2004;36(7):1124-31.
13. Kelly LA, Reilly JJ, Grant S, Paton JY. Low physical activity levels and high levels of sedentary behaviour are characteristic of rural Irish primary school children. Ir Med J. 2005;98(5):138-41.
14. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat 11. 2002;246:1-190.
15. Lewicka M, Farrell L. Physical Activity Measurement in Children 2-5 Years of Age. Report No. CPAH06-002. Sydney: NSW Centre for Physical Activity and Health; 2007, pp. 1-22.
16. Marshall SJ, Biddle SJ, Gorely T, Cameron N, Murdey I. Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. Int J Obes Relat Metab Disord. 2004;28(10):1238-46.
17. Marshall SJ, Gorely T, Biddle SJ. A descriptive epidemiology of screen-based media use in youth: a review and critique. J Adolesc. 2006;29(3):333-49.
18. Martin M, Dollman J, Norton K, Robertson I. A decrease in the association between the physical activity patterns of Australian parents and their children; 1985-1997. J Sci Med Sport. 2005;8(1):71-6.
19. Metallinos-Katsaras ES, Freedson PS, Fulton JE, Sherry B. The association between an objective measure of physical activity and weight status in preschoolers. Obesity (Silver Spring). 2007;15(3):686-94.
20. Moore LL, Lombardi DA, White MJ, Campbell JL, Oliveria SA, Ellison RC. Influence of parents' physical activity levels on activity levels of young children. J Pediatr. 1991;118(2):215-9.
21. Parnell W, Scragg R, Wilson N, Schaaf D, Fitzgerald E. NZ Food NZ Children: Key Results of the 2002 National Children's Nutrition Survey. Wellington (New Zealand): Ministry of Health, 2003, pp. 1-267.
22. Pate RR, Baranowski T, Dowda M, Trost SG. Tracking of physical activity in young children. Med Sci Sports Exerc. 1996;28(1):92-6.
23. Pate RR, Pfeiffer KA, Trost SG, Ziegler P, Dowda M. Physical activity among children attending preschools. Pediatrics. 2004;114(5):1258-63.
24. Pfeiffer KA, McIver KL, Dowda M, Almeida MJ, Pate RR. Validation and calibration of the Actical accelerometer in preschool children. Med Sci Sports Exerc. 2006;38(1):152-7.
25. Reilly JJ, Jackson DM, Montgomery C, et al. Total energy expenditure and physical activity in young Scottish children: mixed longitudinal study. Lancet. 2004;363(9404):211-2.
26. Rennie KL, Wells JC, McCaffrey TA, Livingstone MB. The effect of physical activity on body fatness in children and adolescents. Proc Nutr Soc. 2006;65(4):393-402.
27. Riddoch CJ, Bo Andersen L, Wedderkopp N, et al. Physical activity levels and patterns of 9- and 15-yr-old European children. Med Sci Sports Exerc. 2004;36(1):86-92.
28. Ridgers ND, Stratton G, Clark E, Fairclough SJ, Richardson DJ. Day-to-day and seasonal variability of physical activity during school recess. Prev Med. 2006;42(5):372-4.
29. Rowlands AV, Hughes DR. Variability of physical activity patterns by type of day and season in 8-10-year-old boys. Res Q Exerc Sport. 2006;77(3):391-5.
30. Sallis JF, Nader PR, Broyles SL, et al. Correlates of physical activity at home in Mexican-American and Anglo-American preschool children. Health Psychol. 1993;12(5):390-8.
31. Sallis JF, Patterson TL, McKenzie TL, Nader PR. Family variables and physical activity in preschool children. J Dev Behav Pediatr. 1988;9(2):57-61.
32. Sallis JF, Prochaska JJ, Taylor WC. A review of correlates of physical activity of children and adolescents. Med Sci Sports Exerc. 2000;32(5):963-75.
33. 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. 2005;37(11):1902-8.
34. Tomopoulos S, Valdez PT, Dreyer BP, et al. Is exposure to media intended for preschool children associated with less parent-child shared reading aloud and teaching activities? Ambul Pediatr. 2007;7(1):18-24.
35. 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. 2000;32(2):426-31.
36. Trost SG, Sirard JR, Dowda M, Pfeiffer KA, Pate RR. Physical activity in overweight and nonoverweight preschool children. Int J Obes Relat Metab Disord. 2003;27(7):834-9.
37. Vandewater EA, Rideout VJ, Wartella EA, Huang X, Lee JH, Shim MS. Digital childhood: electronic media and technology use among infants, toddlers, and preschoolers. Pediatrics. 2007;119(5):e1006-15.
38. Whitaker RC, Orzol SM. Obesity among US urban preschool children: relationships to race, ethnicity, and socioeconomic status. Arch Pediatr Adolesc Med. 2006;160(6):578-84.
39. Wilkin TJ, Mallam KM, Metcalf BS, Jeffery AN, Voss LD. Variation in physical activity lies with the child, not his environment: evidence for an 'activitystat' in young children (EarlyBird 16). Int J Obes (Lond). 2006;30(7):1050-5.
40. Yamauchi T, Kim SN, Lu Z, et al. Age and gender differences in the physical activity patterns of urban schoolchildren in Korea and China. J Physiol Anthropol. 2007;26(2):101-7.


© 2009 American College of Sports Medicine