Participants were children, ages 6–12 yr, from schools in each country. Schools in each country were selected based on their accessibility to the researcher and children in each school volunteered to participate. This convenience sample represents a small portion of each country and caution should be taken when making generalizations regarding this data. Children representing America were from a moderate size (400,000) urban community in a Southwest state. Children representing Australia were from a large (1,000,000) urban city in an east coast Australian state, whereas participants from Sweden were from two smaller (60,000 and 20,000) Southeast communities.
American children consisted of 711 total participants (386 girls, 325 boys) of varying ethnic groups (53% white, 30% Hispanic, 4% Native American, 3% African American, 2% Pacific Island/Asian, and 8% other-mixture of ethnicities). Australian children consisted of 563 total participants (285 girls, 278 boys). Swedish children consisted of 680 total participants (324 girls, 356 boys) who were predominantly white. A grand total 1954 children participated in this study (995 girls, 959 boys). Appropriate approval from Institutional Review Boards for each country regarding research with human subjects was obtained. All participants returned a written informed assent form signed by their parents.
Yamax pedometers (MLS-2000) were used in this study to measure the movement counts of children over a period of four weekdays. The Yamax pedometer is made in Japan and marketed under different names in the United States (My Life Stepper MLS-2000; New Lifestyles Digiwalker SW-200). The Yamax pedometer has been validated in the literature extensively (5,7,9,14,17,24,25). Pedometers are unobtrusive and convenient for participants. They measure vertical movement, are less costly, and have been shown to have less error than heart rate monitors and some accelerometers (5). The pedometer is limited in that it is unable to measure intensity, duration, or frequency of activity and cannot store information other than total counts. However, when taking into consideration the intermittent activity patterns of children and the recent public health emphasis on promoting the accumulation of daily physical activity, these limitations are seen as acceptable (14). Particularly with children, it makes sense to collect a total step count at the end of the day rather than be concerned about the intensity and duration of short bouts of activity performed throughout the day.
Pedometers were fastened to the waistband of participants’ pants or shorts in line with the right knee. When no waistband was available, a small belt was worn with the pedometer attached to the belt. The pedometers were worn for four weekdays because this has been found to be an appropriate length of time for determining habitual activity levels in children (7,18,23). A total step count at the end of each measurement period (day) was recorded for statistical analysis.
Children put the pedometers on at the beginning of the school day and wore them until they went to bed, at which time they removed the pedometers. In the morning, they put the pedometer back on and wore it to school. During the first hour of school, a researcher collected the pedometers, recorded the step counts, resealed, and returned them to the children within 1 h to begin the next measurement period. This procedure continued for all participants for four consecutive days (Monday through Thursday).
Pedometers were sealed to assure they were not accidentally reset. Before participation in the study, participants were given a pedometer to examine and use during a physical education class. This helped dispel their curiosity about wearing a sealed pedometer during the data collection period. Participants were asked to maintain normal activity patterns during the study.
A validity check was implemented by asking 8- to 12-yr-old participants to fill out a survey (the survey was not developmentally appropriate for children under 8). The purpose of the survey was to help determine that participants wore the pedometer the entire time and not just for part of the day. Step counts for participants who reported having their pedometer off for longer than 1 hour’s time were not included in the analysis. The survey also asked participants to report the types of activity they did during the previous day.
Measures of height and weight were obtained for each participant to determine their BMI. BMI is used to assess weight relative to height and is an easy method for evaluating a large number of participants. BMI has recently been used to determine international cut points for overweight and obese children who will pass through a BMI of 25 and 30 at age 18 (4). Table 1 reflects the cut points that were used as the basis for comparison in this study.
Descriptive data (N′s, means and standard deviations) for step counts by sex, age, and country were calculated to determine activity levels of children. Figure 1 illustrates the mean step counts by sex, age, and country. Descriptive data (means and standard deviations) for BMI by sex, age, and country were calculated and are illustrated in Figure 2.
A three-way multivariate ANOVA on step counts and BMI with countries, age, and sex as the between factors was conducted. The null hypothesis was assumed and rejected when probability values of P < 0.05 were found. The omnibus F was significant for countries (F = 65.45, P < 0.001), age (F = 20.88, P < 0.001), and sex (F = 119.47, P < 0.001). In addition, the F for step counts was significant among countries (F = 98.11, P < 0.001), age (F = 2.20, P < 0.05), and sex (F = 230.76, P < 0.001), and the F for BMI was significant among countries (F = 45.42, P < 0.001) and age (F = 39.48, P < 0.001) but not for sex (F = 0.24, P = 0.622). Follow-up analysis on step counts and BMI among countries at each age and sex found that, in general, the Swedish children were significantly more active than the Australian and American children, and the American children had significantly higher BMI values than the Australian and Swedish children (Figs. 1 and 2). For boys, the mean step counts ranged from 15,673–18,346 for Sweden, 13,864–15,023 for Australia, and 12,554–13,872 for America. For girls, the mean step counts ranged from 12,041–14,825 for Sweden, 11,221–12,322 for Australia, and 10,661–11,383 for America.
A large amount of variability existed within the groups (by country, age, and sex) for both step counts and BMI, which led to further analysis. To more clearly view how the more active students varied from less active students, each group was divided into tertiles based on their step counts. For every age by sex group within each country, a significant difference was found between the most active and least active tertiles at P < 0.01.
Few significant correlations were found between step counts and BMI. Significant step count and BMI correlations (P < 0.01) were found for American boys ages 11 (r = −.389) and 12 (r = −.553) and American girls age 9 (r = −.364). American and Australian girls, age 8, had significant correlations (r = −.276 and r = −.331 respectively) at P < 0.05. No other significant correlations were found between step counts and BMI.
Tables 2 and 3 show the percentage of boys and girls by age classified as overweight or obese using Cole et al. (4) cut points. Tables 4 and 5 show the number of students classified as obese or overweight by activity tertile. Chi-square analysis showed significant frequency differences among boys and girls classified as overweight/obese or nonoverweight/obese in both the United States and Sweden. The direction confirms the intuitive expectation that children who are less active tend to have higher values of BMI.
Contrary to previous research (8,12,16) the activity curve (as defined by the average daily step count) is somewhat level during the preadolescent years. Figure 1 shows a relatively flat slope that reflects a consistent accumulation of step counts throughout the ages of 6–12. The similarity in steps counts across all age groups in all three countries was an interesting finding because it has been often assumed that youngsters become less active with age. Even if youngsters didn’t become less active with age, one would expect that the longer steps of older youths would result in fewer accumulated steps required to cover similar distances. This is one of the first studies to look at accumulated steps as a method of defining an “activity curve.” Other studies have used energy expenditure, activity recall, and accelerometry to establish activity curves. It may be that older students become more efficient movers and are more accurate in recalling their previous day’s activity. This possible increase in efficiency and accuracy might result in the decrease reported in other studies (8,12,16,19).
Tudor-Locke and Meyers (20) conducted a systematic review of the literature published on pedometer step counts and found only one study (15) that gave mean counts for preadolescent children. Rowlands et al. (15) studied 8- to 10-yr-old children (15 boys and 14 girls from North Wales) and found mean pedometer counts of 16,035 for boys and 12,728 for girls, which is similar to the pedometer counts found in the Swedish children in this study. The President’s Council on Physical Fitness and Sports sponsors an activity award that requires children to accumulate 11,000 steps for girls or 13,000 steps for boys in a day to earn the award (11). Applying this award standard to the mean step counts in this study would make most participants award winners.
Swedish children were the most active with American children accumulating the fewest number of steps. This pattern is seen consistently in all comparisons of step counts made in this study. Based on observations in each country, it is possible that different environmental factors may contribute to the differing activity levels. For example, in Sweden, participants lived in a moderate-size community that was designed for walking and biking. In many cases, it was much quicker to take the walking paths into the city rather than drive around the perimeter looking for parking.
Other factors that might have influenced activity differences are the SES and the structure of the school day. The SES for the families of Australian and Swedish students in the study was middle class with some upper middle class families. U.S. students came from middle and lower middle class homes. Studies with adults have shown that lower income groups are less active and have higher levels of body fat (21), and there is a strong possibility that this difference may occur with children also. Another factor that can impact activity levels is the amount of time students spending sitting in class versus the amount of time they are given for physical activity. An examination of in-class and free time showed that the U.S. schools in this study required 390 min of daily instruction time, gave students 70 min of lunch and recess time (free time), and 60 min of physical education instruction each week. In contrast, Sweden offered 300 and 100 min of daily instruction/free time with 80 min of physical education per week. Australia offered 285 and 75 min of daily instruction/free time with only 30 min of physical education time each week. Nearly 70% of Swedish children participated in daily after-school sport clubs in grades 5 and 6. In the U.S. schools, about 20% of fifth and sixth graders participated in after school sports activity. The majority of Australian children participated in an after school sport program on Fridays for 90 min. Finally, approximately 80% of children walked or biked to school in the Swedish schools, 50% in the American schools, and 5% in the Australian schools.
It is possible that the activity levels of Australian participants may have been underestimated. A substantial proportion of the participants (24%) of Australian youth reported swimming at least 30 min on at least three occasions per week. Because pedometers don’t measure biking or swimming activity, the differential in activity between American children and Swedish/Australian youth may actually be greater than that reported as pedometer counts.
Standard deviations for pedometer counts showed much variability within groups by country, age, and sex. To see whether this range of scores would reveal large differences in activity levels, each group was divided into tertiles. For every age by sex group within each country, a significant difference was found between the most active and least active tertiles at P < 0.01. These differences highlight how, even within a specific age group, children vary greatly in their step counts. In all three countries, the least active youths accumulated approximately 5000–7000 fewer steps than those youth in the most active tertile. This amounts to roughly 30–45% fewer steps each day for those youngsters in the least active tertile.
In Sweden, the 12-yr-old boys in the most active tertile accumulated 1979 more step counts than the 7 yr olds in the same tertile. Similarly, 12-yr-old American boys in the most active tertile accumulated 2985 more step counts than the 6 yr olds. In contrast, the oldest Australian boys in the most active tertile (12 yr) collected 511 fewer steps than the most active 6-yr-old boys. The pattern in the least active tertile is mixed; the oldest Swedish boys accumulated 1216 steps less than the youngest group, the oldest Australian boys accumulated 1540 steps more than the youngest group, whereas 12-yr-old American boys accumulated 531 steps less than the youngest group.
The pattern for girls from 6 to 12 yr (7–12 yr in Sweden) based on tertiles is more consistent. In both the most active and least active tertiles, the oldest Swedish girls collected fewer steps than the youngest (1539 less steps in the most active tertile, 2153 less steps in the least active tertile), whereas the older Australian and American girls had similar step counts (Australian-most active tertile collected 114 steps less, least active tertile collected 627 less steps while the American-most active tertile collected 320 more steps, and the least active tertile collected only 7 steps less).
For boys, it appears older students in the most active tertile collect more steps than younger students. The step counts of girls seem to be more stable for Australians and Americans in both tertiles. The Swedish girls differed, however, and appear to collect fewer steps in the older groups, particularly for girls in the least active tertile.
Body mass index.
When comparing the BMI of students in this study by country, age, and sex to the international BMI standards (3), four groups of children were classified as overweight/obese. These were American boys ages 10, 11, and 12 and American girls age 12. No groups from Sweden or Australia were classified as overweight/obese. Further analysis was warranted to see whether examining the groups by step count tertiles would give a better view of youth who might be classified as overweight/obese. BMI scores for participants in the most active step tertiles and least active step tertiles were reviewed. The mean BMI for each of the tertile groups showed that American boys, ages 7 and 9–12 yr, in the least active tertile are classified as overweight. The mean step scores of American girls in the least active tertile, for ages 8–9 and 11–12 yr, were also classified as overweight. It is interesting to note that no mean scores for boys or girls in the least active tertile from Sweden or Australia were above the overweight/obese cut point.
The difference in BMI between the youngest and oldest age groups is greater in American children than it is in Swedish or Australian children. Also, boys in the most active tertile showed much smaller BMI differences between the youngest and oldest age groups than did the least active participants. The oldest Australian boys differed by 0.73 points more than the youngest, Sweden differed by 0.32 more, and America differed by 1.67 more. Boys in the least active tertile showed an even greater difference than those in the most active tertile. American boys differed by 8.98 points more as compared with differences of 2.57 more for Sweden and 1.09 more for Australia.
For girls, the BMI pattern among age groups is similar to the boys. Girls in the most active tertile showed much smaller BMI differences between the youngest and oldest age groups than did the least active participants. The oldest Swedish girls differed by 2.72 points more than the youngest; Australia differed by 1.87 more, and America differed by 5.05 more. Differences were even greater for girls in the least active tertile. American girls differed by 8.04 points more as compared with differences of 5.89 more for Sweden and 0.82 more for Australia. The comparison of boys and girls in the most active BMI tertile reveals that girls in all three countries increase in BMI more than boys.
Over 33% of all U.S. children were classified as overweight/obese compared with no more than 16.8% of all youth in Sweden and Australia (Tables 2 and 3). Clearly there is cause for concern over the high percentage of children in the United States who are overweight/obese. Although there are not many significant correlations between step counts and BMI, the chi-square analysis does show a trend that the number of youngsters classified as obese or overweight increases as activity level (by tertile group) decreases (Tables 4 and 5). In the United States and Sweden, the frequency of overweight/obese youngsters significantly differs from the expected frequency. A significantly greater number of overweight or obese youngsters are found in the least active tertile, whereas in Australia the frequency differences are not significant.
Relationship of step counts to BMI.
Correlation analysis found little relationship between step counts and BMI. Of the 40 correlations calculated, only five showed a significant negative relationship between step counts and BMI, and those correlations were relatively low. Significant correlations at P < 0.01 were found for American boys ages 11 (r = −0.389) and 12 (r = −0.553) and American girls age 9 (r = −0.364). American and Australian girls age 8 (r = −0.276 and r = −0.331, respectively) had significant correlations at P < 0.05. The relationships found were in the direction expected (negative), and the greatest correlation (r = −0.553 for 12-yr-old American boys) accounted for 28.4% of the variance. Although this is a fairly substantial amount of variance accounted for in a behavioral study, few significant correlations were detected. Generalizations regarding these few significant correlations should be made with caution. When looking at the tertile analyses, a relationship does appear among Swedish girls in the least active tertile. As they age, they decrease the number of steps taken while showing an increase in BMI. It seems intuitive that there would be a relationship between physical activity and BMI scores; however, although these data show some relationship, it is not consistent across all comparisons in the study.
Several limitations should be noted regarding this study. First, this study utilizes a cross-sectional design. Children in this study were not measured on a longitudinal basis, which limits the conclusions that can be made. Second, the convenience sample restricts generalizations that might be made across countries and within countries. Children in an eastern state within the United States may be different from children in a southwest state. Similar issues apply in both Sweden and Australia. Third, this study did not look at differences in developmental stages of the children, which may have contributed to some of the differences across populations. Fourth, activity data was assessed during weekdays only; no weekend activity data were included in this analysis. Weekend data were gathered but was fraught with errors because participants forgot to put on their pedometer, lost it, or used it for only one of the two days. The large number of subjects made it impossible to go to the homes over the weekend to gather daily step counts. Thus, step counts were for the entire 2-d period and it was impossible to certify that participants had worn the pedometers for the both days. Therefore, activity counts in this study refer to weekday activity only. Finally, as indicated in the methods section, there are limitations to the use of BMI as an indicator of overweight and obesity. Although it can be used as an indicator of overweight and obesity, it is just a screening tool and generalizations regarding BMI data should be made with caution.
This cross-sectional study revealed that American children, as compared with Swedish and Australian children, have the lowest mean step counts and the highest BMI. The amount of increase in BMI between age groups among American youngsters is greater than their counterparts in Sweden and Australia. Swedish children accumulated the most daily steps with Australian youth close behind. Swedish girls clearly showed a decrease in step counts and an increase in BMI scores as they aged. Swedish and Australian children maintained a healthier weight throughout their prepubescent years than did American children (as evidenced by a greater percentage of American youth who are classified as overweight and who get heavier with age).
This study focused on prepubescent children in three countries. Such studies need to continue into the adolescent years. A replication of this study with youth, ages 13–18, would provide a clearer picture of what happens to youth as they grow older into adulthood.
1. Berg, I. M., B. Simonsson, B. Brantefor, and I. Ringvist. Prevalence of overweight and obesity in children and adolescents in a county in Sweden. Acta Paediatr. 90: 671–676, 2001.
2. Blair, S. N., and S. Brodney. Effects of physical inactivity and obesity on morbidity and mortality: current evidence and research issues. Med. Sci. Sports Exerc. 31: (Suppl.) S646–S662, 1999.
3. Booth, M., M. Wake, T. Armstrong, et al. The epidemiology of overweight and obesity among Australian children and adolescent, 1995–1997. Aust. N. Z. J. Public Health 25: 162–169, 2001.
4. Cole, T. J., M. C. Bellizzi, K. M. Flegal, and W. H. Dietz. Establishing a standard definition for child overweight and obesity worldwide: international survey. Br. Med. J. 320: 1–6, 2000.
5. Eston, R. G., A. V. Rowlands, and D. K. Ingledew. Validity of heart rate, pedometry, and accelerometry for predicting the energy cost of children’s activities. J. Appl. Physiol. 84: 362–371, 1998.
6. Goran, M. I., B. A. Gower, T. R. Nagy, and R. K. Johnson. Developmental changes in energy expenditure and physical activity in children: evidence for a decline in physical activity in girls before puberty. Pediatrics 101: 887–891, 1998.
7. Gretebeck, R. J., and H. J. Montoye. Variability of some objective measures of physical activity. Med. Sci. Sports Exerc. 24: 1167–1172, 1992.
8. Hovell, M. F., J. F. Sallis, B. Kolody, and T. L. Mckenzie. Children’s physical activity choices: a developmental analysis of gender, intensity levels, and time. Pediatr. Exerc. Sci. 11: 158–168, 1999.
9. Kilanowski, C. K., A. R. Consalvi, and L. H. Epstein. Validation of an electronic pedometer for measurement of physical activity in children. Pediatr. Exerc. Sci. 11: 63–68, 1999.
10. Morrow, J. R., A. W. Jackson, and V. G. Payne. Physical activity promotion and school physical education. President’s Council on Physical Fitness and Sports Research Digest 3: 1–8, 1999.
11. President’s Council on Physical Fitness and Sports. The President’s Challenge Physical Activity and Fitness Awards Program. Bloomington, IN: President’s Council on Physical Fitness and Sports, 2001, p. 9.
12. Rowland, T. W. Exercise and Children’s Health. Champaign, IL: Human Kinetics Books, 1990, pp. 31–45.
13. Rowland, T. W. Adolescence: a “risk factor” for physical inactivity. The President’s Council on Physical Fitness and Sports Research Digest 3: 1–8, 1999.
14. Rowlands, A. V., R. G. Eston, and D. K. Ingledew. Measurement of physical activity in children with particular reference to the use of heart rate and pedometry. Sports Med. 24: 258–272, 1997.
15. Rowlands, A. V., R. G. Eston, and D. K. Ingledew. Relationship between activity levels, aerobic fitness, and body fat in 8–10-yr-old children. J. Appl. Physiol. 86: 1429–1435, 1999.
16. Saris, W. H. M. Habitual physical activity in children: methodology and findings in health and disease. Med. Sci. Sports Exerc. 18: 253–263, 1986.
17. Sequeira, M. M., M. Rickenbach, V. Wietlisbach, B. Tullen, and Y. Schutz. Physical activity assessment using a pedometer and its comparison with a questionnaire in a large population survey. Am. J. Epidemiol. 142: 989–999, 1995.
18. Trost, S. G., R. R. Pate, P. S. Freedson, J. F. Sallis, and W. C. Taylor. Using objective physical activity measures with youth: how many days of monitoring are needed? Med. Sci. Sports Exerc. 32: 426–431, 2000.
19. Trost, S. G., R. R. Page, J. F. Sallis, et al. Age and gender differences in objectively measured physical activity in youth. Med. Sci. Sports Exerc. 34: 350–355, 2002.
20. Tudor-Locke, C. E., and A. M. Myers. Methodological considerations for researchers and practitioners using pedometers to measure physical (ambulatory) activity. Res. Q. Exerc. Sport. 72: 1–12, 2001.
21. U.S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996, p. 177.
22. U.S. Department of Health and Human Services. Healthy People 2010. Washington, DC: Department of Health and Human Services; 2001, p. 13.
23. Vincent, S. D., and R. P. Pangrazi. Does reactivity exist in children when measuring activity levels with pedometers? Pediatr. Exerc. Sci. 14: 56–63, 2002.
24. Vincent, S. D., and C. L. Sidman. Determining measurement error in digital pedometers. Meas. Phys. Educ. Exerc. Sci. 7: 19–24, 2003.
25. Welk, G. J., J. A. Differding, R. W. Thompson, S. N. Blair, J. Dziura, and P. Hart. The utility of the digi-walker step counter to assess daily physical activity patterns. Med. Sci. Sports Exerc. 32: S481–S488, 2000.