There is increasing evidence that modern preschool children might be less physically active than expected (9,11,19), and there is increasing emphasis on early childhood lifestyle as a determinant of later disease risk (6,10,20). It is widely believed that fundamental movement skills and habitual physical activity are related in childhood and adolescence (6,12,14,16), and elucidating relationships between these two variables was highlighted as a research priority by the CDC in 2001 (6). However, Pate (16) questioned the premise that these two variables are related, and asked whether the degree of emphasis being placed on motor skills in the pediatric physical activity community was excessive.
In fact, very few studies have examined the relationship between habitual physical activity and fundamental movement skills (basic motor skills) in children. Okely et al. (14) found a significant relationship between fundamental movement skills and self-reported participation in organized physical activity in adolescents, although only a small proportion (3%) of participation in organized physical activity was explained by fundamental movement skills score. Butcher and Eaton (1) reported a significant positive relationship between fundamental movement skills and participation in vigorous activity in preschool children. However, these authors used free play behavior as their single measure of physical activity, and running speed and agility were the only movement skills assessed. Ulrich (22) found a significant relationship between parent-reported “movement competence” and participation in organized sport in 5- to 10-yr-old children.
There is no consensus in the literature on the methods that should be used to assess/report physical activity or the definition of fundamental movement skills, and measurement of both variables presents practical problems, particularly in young children. Consequently, studies in this area have been scarce. We are not aware of any published studies that have tested empirically whether there are relationships between objectively measured habitual physical activity and global fundamental movement skills in preschool children.
A pilot study by our group in 60 preschool children (20) found a significant positive correlation (r = 0.30, P < 0.05) between habitual physical activity, as measured by Computer Science and Applications (CSA) accelerometer over 2 d, and fundamental movement skills test scores in preschool children. The aim of the current study was therefore to test whether there was any significant relationship between these two variables in a larger and more socioeconomically representative sample of contemporary preschool children, in which physical activity was measured over a longer period (6 d).
SUBJECTS AND METHODS
We aimed to study preschool children recruited for baseline measurements (preintervention) of the Movement and Activity Intervention in Glasgow Children (MAGIC) randomized, controlled trial (20). For inclusion in the current study (for analysis purposes), children had to be apparently healthy, having no chronic disease relating to energy expenditure or physical activity. The children all lived in urban or suburban settings in and around Glasgow, Scotland. The study was approved by the Yorkhill Hospital ethics committee. Informed written consent was obtained from the parent/guardian of each child.
The MAGIC randomized controlled trial entered 545 children. Insufficient numbers of accelerometers and research staff during the study period meant that a pragmatic decision was made to measure habitual physical activity (a secondary outcome measure in the trial) in 482 of these 545, and we selected these 482 children randomly (using a computer-generated random number sequence) from the entire sample of 545. Of these 482, 58 children were excluded because their accelerometry data did not sufficiently meet our criteria established before the study began (a minimum of 9 h·d−1 over at least 3 d (9). Physical activity was therefore measured successfully in 424 of 482 children. A further 30 of these 424 children were excluded from analysis due to absence from nursery school on the day of movement skills test (N = 27) or inability/unwillingness to perform the movement skills test (N = 3). This left data from 394 children for analysis in the present study.
Physical activity measurements.
Habitual physical activity was measured using the CSA WAM −7164 accelerometer (MTI, Fort Walton Beach, FL). This is a small, lightweight, uniaxial device (measures movement largely in the vertical plane) that was worn on the right hip under clothing as previously described (4,9,18,19). A number of studies have now reported favorably on the validity of the CSA/MTI in children, by comparisons with direct observation of behavior or energy expenditure (3–5,15,18,21). The CSA/MTI produces output in activity counts per unit of time. This output can be considered in “raw” form (counts per minute (cpm)) as a valid index of total “volume” of physical activity (4,13,15). Alternatively, output can be interpreted using cut points that define different intensities of physical activity (13,15,18). Pediatric cut points have been published recently that have been validated for young children carrying out unrestricted activities based on both of the criterion measures for physical activity, that is, energy expenditure (15) and direct observation of behavior (18). We applied these cut points in the present study: sedentary behavior <1100 cpm (18); light-intensity physical activity 1100–3200 cpm (15); moderate and vigorous physical activity (MVPA) >3200 cpm (15). Alternative cut points are less suitable for pediatric use at present because they do not meet a number of criteria (17), are unpublished, involve extrapolations from adult studies (which may be questionable), and involve extrapolation from treadmill exercise to free-living activity, which is probably inappropriate.
We aimed to measure habitual physical activity in participating children by asking families to fit the accelerometers in the early morning and remove them at night (before bedtime) for a period of six consecutive days during October and November 2002. Parents and nursery staff were asked to record when accelerometers were fitted and removed and why they were removed (e.g., for swimming).
Fundamental movement skills assessment.
Fundamental movement skills were measured using 15 tasks based on the Movement Assessment Battery. Reliability and concurrent validity of the Movement Assessment Battery in young children has been documented previously by Croce et al. (2); intraclass correlations for test-retest reliability over a 1-wk period for children aged 5–6 yr ranged from 0.92 to 0.98. The measurements in the current study were carried out by the same trained observer, positioned to allow full view of the tests, but far enough away to avoid obstructing or distracting children. The test involved a set of 15 tasks: jumps (vertical jump, running jump, and standing jump from a measured distance of 33 inches), balance (standing on one foot for exactly 1 s and standing on one foot for exactly 6 s, timed by stopwatch), skips (four different forms of skipping), ball exercises (kicking a rolled ball from 72 inches, catching a ball from 33 inches (overarm throw), catching a bounced ball from 72 inches); and throwing a beanbag into a target from 72 inches. These tasks were selected because they had been successfully measured in other studies (23), were age appropriate for the children in our study (7), and, as a set, favored neither boys nor girls (8).
A member of the staff from each participating nursery was shown how to perform and demonstrate each task. To make the children feel more comfortable, they were taken out of class for testing in groups of 4–5 children (distraction caused by interactions between children is common in larger groups), but each child was tested and scored individually. Each group was given a single demonstration of the exercise. Children then attempted each task individually, and the single trained observer (A.F.) checked a box YES (if the task was performed correctly) or NO (if the task was not carried out/carried out incorrectly). The test was scored YES = 1, NO = 0. Each child was therefore given a total score from 0 to 15. This single measure of overall movement skills was used for analysis, though analysis by subsets of tasks (locomotive, manipulative, and balance) was also carried out; these results are discussed below.
For descriptive purposes, we measured height to 0.1 cm and body mass to 0.1 kg in all children. Body mass index (BMI) was calculated as mass (kg) divided by height (m2). BMI was expressed as a SD score (SDS) relative to 1990 U.K. reference data.
We first tested for associations between fundamental movement skills score and various indices of habitual physical activity (total physical activity summarized as accelerometry cpm, percent of monitored time spent in light-intensity physical activity; percent of monitored time spent in MVPA) using correlations. We also tested for differences in these indices of habitual physical activity by quartiles of movement skills score using Kruskal–Wallis tests. Where Kruskal–Wallis tests were significant, we used Mann–Whitney tests to assess the significance of differences in physical activity between the upper and lower quartiles for movement skills score. Statistical significance was set at P < 0.05.
The mean duration of accelerometry in the 394 children studied was 56.0 h (SD 13.3). Physical characteristics, summary data for total physical activity, percent time spent sedentary, in light-intensity physical activity, and MVPA, and fundamental movement skills are shown in Table 1. Total physical activity and percent monitored time spent in light activity and in MVPA were slightly but significantly higher in boys than in girls (Mann–Whitney U-test, P < 0.001). Percent time spent in sedentary behavior was slightly but significantly lower in boys than in girls (Mann–Whitney U-test, P < 0.001). There was no difference in fundamental movement skills score between boys (median 8, minimum 0, maximum 14) and girls (median 8, minimum 0, maximum 14) (Mann–Whitney U-test, P = 0.56).
Relationship between Fundamental Movement Skills and Physical Activity
Correlations in the entire sample.
Total movement skills score was weakly but significantly positively correlated with total physical activity (accelerometry output (cpm)): r = 0.10, P = 0.039. Total fundamental movement skills score was not significantly correlated with the percent time spent in light-intensity activity: r = 0.02, P = 0.625. Total movement skills score was weakly but significantly positively correlated with the percent of monitored time spent in MVPA: r = 0.18, P < 0.001. The correlations were very similar whether carried out in boys and girls separately, or with data from both sexes combined.
Relationships between quartiles for movement skills score and habitual physical activity.
We found no significant association between quartiles of movement skills score and total physical activity (accelerometry count (cpm); Kruskal–Wallis, P = 0.16), or time spent in light-intensity physical activity (P = 0.54). These observations were essentially the same whether analyzed for the entire sample or treating the sexes separately in the analysis.
There was a significant association with quartiles of movement skills score and the percent of monitored time spent in MVPA (P = 0.001). Comparison of the percent time spent in MVPA by quartiles for fundamental movement skills score are summarized in Figure 1. In girls, the time spent in MVPA was significantly higher in the upper quartile (Q4) for fundamental movement skills than the lower quartile (Q1) (median difference 0.9%, 95% confidence interval (CI) 0.2–1.6%). In boys, the time spent in the upper quartile (Q4) for fundamental movement skills score was significantly higher by the Mann–Whitney U-test than the lower quartile (Q1) (median difference 0.9%; 95% CI 0.0–0.2%, P < 0.04).
Main findings in context.
The present study suggests that there are significant but weak cross-sectional relationships between habitual physical activity (specifically time spent in MVPA and total physical activity) and fundamental movement skills in modern preschool children. Children who spent more time in MVPA tended to have higher fundamental movement skills scores, but each explained only a small amount of variation in the other variable in the present study. Our observations tend to support the view expressed by Pate (16) that habitual physical activity and movement skills need not be closely associated in childhood. However, our observation of marked differences between time spent in MVPA and motor skills in upper and lower quartiles (Fig. 1) perhaps suggests that, at the extremes of the distribution, these associations might be more important. This possibility should be explored in future research. For example, children with the most limited engagement in MVPA also had the poorest performance in the motor skills assessment, and it is possible that limited engagement in MVPA might hinder motor development or that limited motor development might restrict participation in MVPA.
Our findings are also similar to those of Okely et al. (14), who found a weak but statistically significant relationship between time spent in organized physical activity and fundamental movement skills in Australian adolescents. The amount of variance in time spent in organized physical activity explained by fundamental movement skills in the Australian study was small (3%).
In children in the present study, engagement in MVPA was relatively low, and engagement in sedentary behavior was high. We have previously reported this observation in other samples of young Scottish children (9,19). The relatively low levels of habitual physical activity and time spent in MVPA might have reduced our ability to detect associations with movement skills score.
It is possible that gender differences might exist in the relationships between physical activity and motor skills (7,8,12,14). We did not find any marked gender differences in the main analyses described above; this observation might have been due to the use of data reduction to a single global movement skills score. More subtle observations such as gender differences might have been detectable within categories of the movement skills score (for locomotive, manipulative, or balancing tasks), but our subsequent analyses using these three components of the movement skills score failed to indicate any clear gender differences.
Despite being the largest empirical investigation of relationships between movement skills and physical activity to date, the present study had a number of limitations. We used a uniaxial accelerometer, which in principle should provide less accurate measurement than bi- or triaxial instruments. However, in practice this might be relatively unimportant as long, as a valid uniaxial instrument is used. Empirical studies that have compared uni- versus bi- or triaxial accelerometers with reference methods have either reported little difference in accuracy or higher accuracy for the uniaxial systems (5,11,15,24). Furthermore, the uniaxial CSA/MTI accelerometer that we used has been the subject of several pediatric studies that have reported high validity relative to direct observation of behavior or energy expenditure as criterion measures (3–5,11,18,21), and validated pediatric cut points for accelerometry output are available that are based on free-living activity and sedentary behavior (15,18). Alternative accelerometry cut points are in pediatric use, but they have not yet been validated, are not published, are based on treadmill exercise rather than free-living activity, and are based on extrapolations from adult data (17).
The use of a 1-min accelerometry sampling interval (epoch) has been regarded by some as problematic, and it has been argued on theoretical grounds that shorter epochs might measure vigorous-intensity activity more accurately. Again, the empirical evidence suggests that this is a relatively minor source of error, with the main practical consequence of using 1-min epochs in children being a slight misclassification of some vigorous activity as moderate-intensity activity (13).
The fundamental movement skills test has certain limitations. There may be gender differences in the ability to perform certain tasks within the test, and indeed this has been demonstrated previously in preschool and in older children, with girls typically performing more successfully in balancing tasks and boys engaging in tasks that involve running and jumping (7,8,12). It is possible that relationships between movement skills and physical activity might have been stronger if these subcategories of movement skills were considered, rather than reducing the data to a single global score. In an attempt to address this question, we analyzed using three components of the overall movement skills score separately, but in practice the results were very similar to those obtained when the overall combined score was considered.
Our use of a product-oriented rather than process-oriented assessment of fundamental movement skills may be viewed as a weakness (14); however, the age of the children in our study did not allow for a process-oriented measurement. These limitations may have masked the relationship between physical activity and movement skills to some extent.
The present study found that in our sample and setting, measures of habitual physical activity and movement skills were significantly related, but the relationships observed were generally weak, and largely with MVPA. Future studies should test these relationships in other settings. Alternative study designs also should be considered. In particular, longitudinal studies and intervention studies might provide greater ability to detect associations between these variables.
1. Butcher, J. E., and W. O. Eaton. Gross and fine motor proficiency in pre-schoolers: relationships with free play behaviour and activity level. J. Hum. Mov. Stud
. 16:27–36, 1989.
2. Croce, R. V., M. Horvat, and E. McCarthy. Reliability and concurrent validity of the movement assessment battery for children. Percept. Mot. Skills
. 93:275–280, 2001.
3. 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
. 84:362–367, 1998.
4. Fairweather, S. C., J. J. Reilly, S. Grant, and J. Y. Paton. Using the CSA activity monitor in pre-school children. Pediatr. Exerc. Sci
. 11:413–420, 1999.
5. Finn, K. J., and B. Specker. Comparison of Activwatch activity monitor and children’s activity rating scale in children. Med. Sci. Sports Exerc
. 32:1794–1797, 2000.
6. Fulton, J. E., C. R. Burgeson, G. R. Perry, et al. Assessment of physical activity
and sedentary behaviour in pre-school age children: priorities for research. Pediatr. Exerc. Sci
. 13:113–126, 2001.
7. Gabbard, C. P. Lifelong Motor Development
, 4th Ed. San Francisco: Pearson Benjamin Cummings, 2004, pp 284–347.
8. Garcia, C. Gender differences in young children’s interactions when learning fundamental motor skills
. Res. Q. Exerc. Sport
. 65:213–215, 1994.
9. Jackson, D. M., J. J. Reilly, L. A. Kelly, C. Montgomery, S. Grant, and J. Y. Paton. Objectively measured physical activity
in a representative sample of 3 to 4 year old children. Obes. Res
. 11:420–425, 2003.
10. Janz, K. F., S. M. Levy, T. L. Burns, J. C. Torner, M. C. Willing, and J. J. Warren. Fatness, physical activity
and television viewing in children during the adiposity rebound period. Prev. Med
. 35:563–571, 2002.
11. Kelly, L. A., J. J. Reilly, S. C. Fairweather, S. Barrie, S. Grant, and J. Y. Paton. Comparison of two accelerometers
for assessment of physical activity
in pre-school children. Pediatr. Exerc. Sci
. 16:324–333, 2004.
12. McKenzie, T. L., J. F. Sallis, and S. L. Broyles. Childhood movement skills: predictors of physical activity
in Anglo American and Mexican American adolescents? Res. Q. Exerc. Sport
. 73:238–244, 2004.
13. Nillson, A., U. Ekelund, A. Yngve, and M. Sjostrom. Assessing physical activity
in children with the CSA accelerometer: effect of different time sampling intervals and placements. Pediatr. Exerc. Sci
. 14:87–96, 2002.
14. Okely, A. D., M. L. Booth, and J. W. Patterson. Relationship of physical activity
to fundamental movement skills among adolescents. Med. Sci. Sports Exerc
. 33:1899–1904, 2001.
15. Puyau, M. R., F. A. Adolph, N. F. Vohra, and N. F. Butte. Validation and calibration of physical activity
monitors in children. Obes. Res
. 10:150–157, 2002.
16. Pate, R. R. Assessment of physical activity
and sedentary behavior in pre-school children: priorities for research (Response). Pediatr. Exerc. Sci
. 13:129–130, 2001.
17. Reilly, J. J., D. M. Jackson, C. Montgomery, et al. Physical activity
in young children (Correspondence). Lancet
. 363:1163–1164, 2004.
18. Reilly, J. J., L. A. Kelly, G. Burke, S. Grant, and J. Y. Paton. An objective method for measurement of sedentary behavior in 3–4 year olds. Obes. Res
. 11:1155–1158, 2003.
19. Reilly, J. J., D. M. Jackson, C. Montgomery, et al. Total energy expenditure and physical activity
in young children: mixed longitudinal study. Lancet
. 363:211–212, 2004.
20. Reilly, J. J., and Z. C. McDowell. Physical activity
interventions for the treatment and prevention of childhood obesity. Proc. Nutr. Soc
. 62:611–619, 2003.
21. Trost, S. G., D. S. Wart, S. M. Moorehead, P. D. Watson, W. Riner, and R. Burke. Validity of the CSA monitor in children. Med. Sci. Sports Exerc
. 30:629–633, 1998.
22. Ulrich, B. D. Perceptions of physical competence, motor competence and participation in organised sport: their interrelationships in young children. Res. Q. Exerc. Sport
. 58:57–67, 2004.
23. Van Beurden E., L. M. Barnett, A. Zask, U. C. Dietrich, L. O. Brooks, and J. Beard. Can we skill and activate children through primary school physical education lessons? Prev. Med
. 36:493–501, 2003.
24. Welk, G., and C. B. Corbin. The validity of the Tritrac-R3D activity monitor for assessment of physical activity
. Res. Q. Exerc. Sport
. 66:202–209, 1995.