Pediatric Exercise Science, 2012, 24, 420-434

© 2012 Human Kinetics, Inc.

Locomotive and Non-Locomotive Activity

as Determined by Triaxial Accelerometry
and Physical Fitness in Japanese
Preschool Children
Chiaki Tanaka
J.F. Oberlin University

Yuki Hikihara
Chiba Institute of Technology

Kazunori Ohkawara and Shigeho Tanaka

National Institute of Health and Nutrition

This study examined the potential relationship between participation in physi-

cal activity (PA) assessed by triaxial accelerometry and physical fitness testing,
including health-related and skill-related parameters of fitness, in 136 Japanese
preschoolers (65 girls and 71 boys, 5.5 ± 0.6 years). In partial correlation analyses,
grip strength and 20m shuttle run test were positively correlated with time spent
in physical activity ratio (PAR) ≥ 4. Better scores on standing long jump distance
and jump over and crawl under tests were associated with lower sedentary time and
greater moderate-to-vigorous PA time and PAR ≥ 4 time, and increased physical
activity level. Moreover, 25m run speed was positively correlated with time spent
in PAR ≥ 4 and locomotive activity. These findings suggest that development of
both health-related (muscle strength and aerobic fitness) and skill-related fitness
(power, agility and speed) may make engagement in PA easier for preschool
children, although further research on the cause-effect relationship is needed.

Low physical activity (PA) among children and youths is a health and social
problem in many countries, including Japan (1,18,19,25). Mattocks et al. (17)
reported that time spent outside in early life (up to age 5) predicted later PA as
determined by objective measurement in 11- and 12-year-old children. There is
evidence that obesity, inactivity, and PA may track from childhood to adolescence

C. Tanaka is with the Division of Integrated Sciences, J.F. Oberlin University, Tokyo, Japan. Hikihara
is with the Faculty of Engineering, Chiba Institute of Technology, Chiba, Japan. Ohkawara is with
the Faculty of Informatics and Engineering, University of Electro-Communications, Tokyo, Japan. S.
Tanaka is with the Dept. of Nutritional Science, National Institute of Health and Nutrition, Tokyo, Japan.

420
 Physical Activity and Physical Fitness in Preschool Children   421

and adulthood, and thus, recent research has begun to focus on the importance of
PA in preschool years (15,23). Objective measurement using an accelerometer has
been used to estimate energy expenditure and to classify levels of PA relatively
accurately (2). Previous studies have also indicated that activity patterns tend to be
sporadic and omnidirectional in preschoolers (22). Measures capable of capturing
different intensities over multiple planes are likely to have the greatest accuracy for
this population. Studies by our group (10,27) showed that in preschool children,
discrimination between locomotive and nonlocomotive activities such as certain
types of play (e.g., ball tossing, building with blocks) using a triaxial acceler-
ometer (ActivTracer, GMS) enables more accurate evaluation of PA intensity.
These studies indicate that the relationship between PA intensity and accelera-
tion counts is different for locomotion than for other types of PA, as indicated in
adults (16). Our previous study showed that nonlocomotive time as measured by
triaxial accelerometry contributed more to PA than locomotive time in free-living
Japanese preschool children (28). Moreover, in free-living adults, nonlocomo-
tive activities such as household activities significantly contribute to total energy
expenditure (13,21). However, many methods have been used to measure PA in
preschool children and most of these do not discriminate between locomotive and
nonlocomotive activity (2). In addition, the relationship between physical fitness
and locomotive or nonlocomotive activity itself also should be taken into account,
considering the complexity of PA in young children.
Lubans et al. (14) systematically reviewed the relationship between funda-
mental movement skills competency and potential health benefits in children and
adolescents, and found strong evidence for a positive association, primarily in
elementary school children and adolescents. In preschool children, a review by
Hinkley et al. (9) indicated that many factors influence PA also. Movement skills
are one of the possible biological variables. The preschool years are character-
ized by significant changes in the acquisition of locomotor and object control
skills and nervous system maturation. Moreover, there are large interindividual
difference in both fundamental motor skills and physical fitness in young children
(15). Thus, the relationship between fundamental movement skills or physical
fitness and PA might differ from young children and adolescents. However, there
are fewer reports on the relationship between fundamental movement skills
or physical fitness and PA in preschool-aged children than in older children.
Recently, some studies examined the relationship between fundamental motor
skills and objectively evaluated PA (3,4,8,31). However, accelerometry without
with discrimination between nonlocomotive and locomotive PA mainly reflects
locomotor activity (10,16,27). To our knowledge, no previous study has assessed
the relationship between locomotive and nonlocomotive PA and health-related
(aerobic fitness, muscle strength, endurance, and flexibility) and skill-related
aspects of physical fitness (agility, balance, coordination, power, reaction time,
and speed) in preschool children.
The purpose of the current study was to examine and further define the potential
relationship between health-related and skill-related physical fitness, and habitual
PA in preschool children. We objectively measured PA by triaxial accelerometry
with discrimination between nonlocomotive and locomotive PA, and performed
comprehensive assessment of physical fitness.
422  Tanaka et al.

Methods
Subjects
Subjects were 136 4- to 6-year-old Japanese preschool children (65 girls and 71
boys; mean age 5.5 ± 0.6 years, range 4.3–6.9 years), living in the Tokyo metropoli-
tan area and attending kindergarten or nursery school. A total of 234 children from
four kindergartens and three nursery schools were invited to participate in the study.
However, there were missing data for the physical fitness tests and accelerometry
data due to equipment failure during the study period. Moreover, only subjects with
data from at least two weekdays and at least one weekend were used in the analysis.
As a result, PA and physical fitness were measured successfully in 136 children
(58%). All of the subjects reported being in good health, without any anamnesis of
conditions affecting energy expenditure, such as abnormal thyroid gland function.
Informed consent was obtained from a parent, and the Ethical Committee of J. F.
Oberlin University approved the study protocol.

Measurement Materials and Methods

Physical Activity Measurements.  Habitual PA was measured using a triaxial
accelerometer (ActivTracer, GMS, Tokyo). Subjects wore a 57 g ActivTracer on
the left side of the waist, as previously described (27). The ActivTracer was set to
record in 1-min epoch. PA was monitored continuously for 6 days (typically on 4
weekdays plus 2 weekend days). Subjects were requested to wear these devices
except under special circumstances such as dressing and bathing. Times that the
subjects did not wear the equipment and sleeping times were recorded by their
parents. Nonwearing time for the ActivTracer was based on the parents’ records and
counts of the ActivTracer. We excluded days in which more than 2 hr of nonwearing
time had accrued, not counting time allowed for the above-mentioned unavoidable
reasons, and days when subjects were absent from school. Subjects with data from
at least two weekdays and at least one weekend were included in the analysis. PAR
(physical activity ratio), a multiple of basal metabolic rate, was estimated from
synthetic counts recorded every 1 min as previously described (27). Anteroposterior
(x-axis), mediolateral (y-axis), and vertical (z-axis) movement were measured
using the triaxial accelerometer, and synthetic (using synthesized triaxes as vectors)
acceleration was calculated. In addition, because the triaxial accelerometer can shift
horizontally during measurement, the x- and y-axes were synthesized as ‘horizontal
acceleration’ for the analysis. When synthetic activity counts were in the range
corresponding to medium-intensity activity (between 130 and 600 mG), classification
criteria using the vertical/horizontal counts ratio (as previously described) were used
to discriminate different types of medium-intensity activity, because PAR for some
medium-intensity activities (i.e., nonlocomotive activities) are underestimated by
synthetic activity counts (27): The vertical/horizontal ratio for locomotive activity
was ≥ 1.19 and for nonlocomotive activity was <1.19.
Anthropometry.  Height and weight were measured to the nearest 0.1 cm and 0.1
kg, respectively. Body mass index (BMI) was also calculated. Assessors were trained
by the author (CT) and each measurement was conducted by the same assessor.
Physical Fitness.  Health-related physical fitness was evaluated based on 3 tests:
grip strength as an indicator of muscle strength, a 20m shuttle run test for endurance,
 Physical Activity and Physical Fitness in Preschool Children   423

and sitting trunk flexion to assess flexibility. Skill-related physical fitness was assessed
based on 4 tests: standing long jump to assess power, a jump over and crawl test for
agility, balancing on one foot with eyes open to assess balance, and a 25m run to
indicate speed. The jump over and crawl test was shown to have both validity and
reliability as a measure of agility in preschool children by the Research Center for
Physical Education in Japan (11). Subjects jumped over a 35 cm-high elastic string
and then immediately crawled, and repeated this movement 5 times. Time required
to perform these actions was measured. The other 6 items were measured based on
the New Physical Fitness Standards of Japanese People, version II (12) except for the
jump over and crawl test. Physical fitness assessment was performed in a playroom
and on a playground on the morning of the measurement day for each facility.

Analyses
Average number of weekday and weekend minutes spent in sedentary (PAR < 2),
light activity (2 ≤ PAR <3), MVPA (3 ≤ PAR < 6), and PAR ≥ 4 were calculated for
each individual, and then average weekly values were calculated by weighting for
5 weekdays and 2 weekend days (Weighed data = (Average for 2 weekdays × 5 +
average for 1 weekend day × 2) / 7). PAL was also estimated as the average of PAR,
because PAL refers to total energy expenditure (kcal/day) divided basal metabolic
rate (kcal/day). Individuals were excluded from analysis if no ActivTracer data were
obtained for at least 2 weekdays and 1 weekend day. Penpraze et al. (24) indicated that
the reliability of PA monitoring was nearly the same from 3 to 10 days in 4–5 year
old children. Therefore, Cliff et al. (5) suggested at least three days of monitoring in
young children. When synthetic activity counts corresponding to medium-intensity
activity (between 130 and 600 mG) were obtained, time spent in locomotive-type
and non-locomotive-type activity was calculated. Total z-scores for health-related
and skill-related physical fitness were calculated as a comprehensive health-related
and skill-related physical fitness index. Furthermore, total z-scores for both aspects
were calculated as a comprehensive physical fitness index. Student’s t test was carried
out to assess the influence of gender. The relationship between the two variables was
evaluated by partial correlation analysis controlled for gender, age, height, and body
weight. When analysis for each gender was performed, the controlled variables used
were age, height, and body weight. Analysis of covariance with PA (sedentary, light,
MVPA, or PAR ≥ 4) for the dependent variable was performed. Models were adjusted
for gender, age, height, and body weight. The association between PA variables and
weight status was analyzed by analysis of covariance adjusted for age and gender.
All results are shown as means ± SD (standard deviations). Statistical analyses
were performed with SPSS version 17.0J for Windows (SPSS Inc, Japan, Tokyo).
All statistical tests were regarded as significant when p-values were less than 0.05.

Results
The physical characteristics of the subjects are shown in Table 1. The numbers of
overweight girls and boys based on BMI (6) were 6 and 4, respectively. Only one
boy was obese. Morphological variables did not show a gender difference. Scores
for girls were higher than those for boys for sitting trunk flexion and one-foot bal-
ance with eyes open. Girls scored lower than boys in the standing long jump and
25m run tests.
424  Tanaka et al.

Table 1  Physical Characteristics and Physical Fitness of Subjects

All subjects
(n = 136) Girls (n = 65) Boys (n = 71)
Variable Mean ± SD Mean ± SD Mean ± SD P-value
Age (yr) 5.5 ± 0.6 5.5 ± 0.6 5.6 ± 0.5
Height (cm) 109.9 ± 5.0 109.2 ± 5.3 110.6 ± 4.6
Weight (kg) 18.6 ± 2.5 18.4 ± 2.7 18.9 ± 2.3
Body mass index 15.4 ± 1.3 15.3 ± 1.3 15.4 ± 1.3
(kg/m2)
Health-related
physical fitness
Grip strength (kg) 7.7 ± 2.1 7.3 ± 2.2 8.0 ± 1.9
20m shuttle run 18.1 ± 11.0 16.7 ± 8.9 19.4 ± 12.6
test (times)
Sitting trunk 27.7 ± 6.7 30.7 ± 5.8 25.0 ± 6.3 <0.01*
flexion (cm)
Skill-reated physical
fitness
Standing long 104.5 ± 15.7 101.1 ± 15.9 107.7 ± 14.9 0.01*
jump (cm)
Jump over and 17.7 ± 4.3 18.2 ± 3.8 17.3 ± 4.7
crawl test (sec)
One -foot balance 40.3 ± 40.8 49.7 ± 50.3 31.7 ± 27.2 0.01*
with eye s open
(sec)
25m run (sec) 6.8 ± 0.6 7.0 ± 0.6 6.7 ± 0.6 <0.01*
*: girls vs boys p < 0.05.

Time spent at different activity intensity levels, PAL, and locomotive and
nonlocomotive activity time are shown in Table 2. MVPA, PAR ≥ 4, PAL, and
locomotive activity were significantly greater in boys than in girls. Sedentary time
for boys was significantly lower than for girls.
Table 3 shows the partial correlations between physical fitness and duration
of PA intensity levels and types of activities. Grip strength and the 20m shuttle
run test were positively associated with PAR ≥ 4. Significant correlations were
found between the standing long jump and jump over and crawl tests and all PA
intensity levels except for light activity. The 25m run score was negatively cor-
related with PAR ≥ 4, and was also correlated with locomotive activity, but no
other physical fitness parameters were specifically correlated with locomotive or
nonlocomotive activity. Health related total Z-score was positively associated with
PAR ≥ 4. Skill related total Z-score was positively correlated with PAR ≥ 4 and
locomotive activity, and negatively associated with nonlocomotive activity. Thus,
comprehensive physical fitness total Z-score was positively associated with PAR
≥ 4 and locomotive activity.
 Physical Activity and Physical Fitness in Preschool Children   425

Table 2  Characteristics of Daily Physical Activity

All subjects
(n = 136) Girls (n = 65) Boys (n = 71)
Activity Mean ± SD Mean ± SD Mean ± SD P-value
Time in sedentary PA 1184 ± 49 1198 ± 43 1172 ± 50 <0.01*
(min/day)
Time in light PA 160 ± 30 156 ± 29 164 ± 30
(min/day)
Time in MVPA 95 ± 29 86 ± 23 104 ± 31 <0.01*
(min/day)
Time in PAR≥4 21 ± 11 17 ± 9 24 ± 12 <0.01*
(min/day)
PAL 1.54 ± 0.07 1.52 ± 0.06 1.56 ± 0.07 <0.01*
Time in locomotive 104 ± 57 90 ± 53 116 ± 58 0.01*
activity (min/day)
Time in nonlocomotive 140 ± 49 137 ± 42 142 ± 54
activity (min/day)
MVPA: moderate-to-vigorous physical activity, PAR: physical activity ratio, PAL: physical activity
level, *: girls vs boys p < 0.05.

Locomotive activity time was significantly correlated with sedentary time

(r=-.41), light-intensity PA time (r = .61), PAR ≥ 4 (r = .37), and PAL (r = .35).
Nonlocomotive activity time was significantly correlated with sedentary time (r=-
.45), light-intensity PA time (r = .17), MVPA (r = .58), and PAL (r = .45).
The results of analysis of covariance (Table 4) indicate that children in the
highest tertiles for grip strength and 20m shuttle run test spent significantly greater
time at PAR ≥ 4 compared with children in the lowest tertiles and children in the
intermediate tertile for 20m shuttle run test, also. Children in the highest tertile for
the 20m shuttle run test spent significantly greater time engaging in locomotive
activity and less time in nonlocomotive activity compared with children in the
lowest tertile. Children in the highest tertile for standing long jump had significantly
greater PAL than children in the lowest and intermediate tertiles and significantly
greater MVPA time than children in the lowest tertile. Children in the highest ter-
tile for the jump over and crawl test and 25m run spent significantly greater time
in MVPA, PAR ≥ 4 and locomotive activity and greater PAL than children in the
lowest tertile. Moreover, in the highest tertile for 25m run, time spent in PAR ≥ 4
was significantly greater than for children in the lowest and intermediate tertile.
For nonlocomotive activity, there was no significant group difference except for
20m shuttle run test.
We also examined partial correlations separately for each gender. For girls,
standing long jump and nonlocomotive activity were significantly correlated with
MVPA after adjusting for age, height, and weight. The jump over and crawl test was
significantly correlated with intensity of PA and PAL, except for light activity. For
boys, muscle strength, 20m shuttle run test, standing long jump, the jump over and
crawl test, and 25m run were significantly correlated with PAR ≥ 4. Standing long
426
Table 3  Partial Correlations Between Physical Activity or Type of Activities and Physical Fitness
Non-
Physical fitness Sedentary P-value Light P-value MVPA P-value PAR≥4 P-value PAL P-value Locomotive P-value locomotive P-value
Health-related physical fitness
Grip strength (kg) -0.03 -0.03 0.09 0.22 0.01* 0.08 0.05 -0.07
20m shuttle run test -0.08 -0.03 0.16 0.26 0.00* 0.10 0.08 -0.07
(time s)
Sitting trunk flexion (cm) -0.07 0.05 0.07 0.05 0.06 0.07 0.00
Skill-related physical fitness
Standing long jump (cm) -0.18 0.04* 0.09 0.21 0.01* 0.20 0.02* 0.20 0.02* 0.10 0.02
Jump over and crawl 0.22 0.01* -0.09 -0.28 <0.01* -0.31 0.00* -0.23 <0.01* -0.16 0.02
test (se c)
One-foot balance with 0.03 0.01 -0.05 0.02 0.00 0.08 -0.11
eyes open (sec)
25m run (sec) 0.10 -0.01 -0.17 -0.33 0.00* -0.15 -0.21 0.01* 0.17
Health-related total Z-score -0.09 0.00 0.16 0.26 0.00* 0.12 0.10 -0.07
Skill-related total Z-score -0.05 0.01 0.07 0.22 0.01* 0.10 0.18 0.03* -0.18 0.04*
Total Z-score -0.09 0.00 0.15 0.30 0.00* 0.13 0.17 0.05* -0.15

Adjusted for gender, age, height, and body weight.*: p < 0.05.
 Table 4  Physical Activity by Tertile of Physical Fitness
Sedentary Light MVPA PAR≥4 Non-
(min/day) (min/day) (min/day) (min/day) Locomotive locomotive
Physical fitness Mean ± Mean ± Mean ± Mean ± PAL Mean (min/day) (min/day)
tertiles (n) SD P-value SD P-value SD P-value SD P-value ± SD P-value Mean ± SD P-value Mean ± SD P-value
Health-related
physical fitness
  Grip strength
 (kg)
   Low 46 1181 ± 8 164 ± 5 95 ± 5 18 ± 2 0.04* 1.54 ± 0.01 108 ± 9 144 ± 8
   Intermediate 46 1188 ± 7 159 ± 4 92 ± 4 21 ± 2 1.54 ± 0.01 102 ± 8 136 ± 7
   High 44 1184 ± 8 156 ± 5 99 ± 5 24 ± 2 1.55 ± 0.01 101 ± 10 139 ± 9
  20m shuttle run test (times)
   Low 44 1189 ± 8 158 ± 5 93 ± 5 16 ± 2 <0.01* 1.54 ± 0.01 77 ± 9 0.01* 162 ± 8 0.02*
   Intermediate 49 1187 ± 7 162 ± 4 91 ± 4 0.03* 21 ± 2 0.03* 1.54 ± 0.01 121 ± 8 125 ± 7
   High 43 1177 ± 7 159 ± 5 103 ± 4 25 ± 2 1.55 ± 0.01 112 ± 8 134 ± 8
  Sitting trunk flexion (cm)
   Low 44 1184 ± 8 162 ± 5 94 ± 4 20 ± 2 1.54 ± 0.01 100 ± 9 143 ± 8
   Intermediate 52 1182 ± 7 161 ± 4 97 ± 4 22 ± 2 1.55 ± 0.01 109 ± 8 135 ± 7
   High 40 1187 ± 8 158 ± 5 95 ± 5 20 ± 2 1.54 ± 0.01 101 ± 9 142 ± 8
Skill-related
physical fitness
  Standing long
  jump (cm)
   Low 45 1190 ± 7 161 ± 5 89 ± 4 0.02* 20 ± 2 1.53 ± 0.01 0.04* 114 ± 8 127 ± 8
   Intermediate 46 1193 ± 7 0.03* 154 ± 4 93 ± 4 18 ± 2 0.01* 1.53 ± 0.01 0.03* 79 ± 8 <0.01* 154 ± 7
   High 45 1170 ± 7 166 ± 5 104 ± 4 24 ± 2 1.56 ± 0.01 118 ± 8 137 ± 8
  Jump over and
  crawl test (sec)
   Low 45 1200 ± 7 0.05* 153 ± 5 86 ± 4 0.03* 16 ± 2 <0.01* 1.52 ± 0.01 0.04* 85 ± 9 0.04* 144 ± 8
   Intermediate 46 1175 ± 7 166 ± 4 99 ± 4 22 ± 2 1.56 ± 0.01 114 ± 8 139 ± 7

  427
(continued)
428
Table 4  (continued)
Sedentary Light MVPA PAR≥4 Non-
(min/day) (min/day) (min/day) (min/day) Locomotive locomotive
Physical fitness Mean ± Mean ± Mean ± Mean ± PAL Mean (min/day) (min/day)
tertiles (n) SD P-value SD P-value SD P-value SD P-value ± SD P-value Mean ± SD P-value Mean ± SD P-value
   High 45 1179 ± 7 161 ± 4 101 ± 4 23 ± 2 1.55 ± 0.01 112 ± 8 135 ± 8
  One-foot balance
  with eyes open
 (sec)
   Low 45 1188 ± 7 158 ± 4 95 ± 4 20 ± 2 1.54 ± 0.01 94 ± 8 147 ± 7
   Intermediate 46 1172 ± 7 0.03* 165 ± 4 103 ± 4 0.02* 22 ± 2 1.56 ± 0.01 108 ± 8 0.03* 145 ± 7
   High 45 1193 ± 7 158 ± 4 89 ± 4 20 ± 2 1.53 ± 0.01 109 ± 8 127 ± 7
  25m run (sec)
   Low 46 1195 ± 7 158 ± 5 87 ± 4 0.01* 16 ± 2 <0.01* 1.52 ± 0.01 0.02* 90 ± 8 0.02* 146 ± 8
   Intermediate 47 1183 ± 7 161 ± 4 96 ± 4 20 ± 2 0.01* 1.54 ± 0.01 102 ± 8 140 ± 7
   High 43 1174 ± 8 162 ± 5 104 ± 4 26 ± 2 1.56 ± 0.01 120 ± 9 132 ± 8

Adjusted for gender, age, height, and body weight.*: p < 0.05, compared with the highest tertiles.
 Physical Activity and Physical Fitness in Preschool Children   429

jump was also significantly correlated with PAL. Moreover, the jump over and crawl
test and 25m run were positively correlated with locomotive activity time. Nonloco-
motive activity was significantly negatively correlated with 25m run performance.

Discussion
This study evaluated physical states ranging from sedentary behavior to vigorous PA
and classified locomotive and nonlocomotive activity using triaxial accelerometry in
preschool Japanese children. Associations between health-related and skill-related
physical fitness parameters and PA and activity types were then examined. To our
knowledge, no previous study has addressed the potential role of health-related
physical fitness parameters, including muscle strength and flexibility, in the early
development of a physically active lifestyle with discrimination between locomo-
tive and nonlocomotive PA during medium-intensity activity in young children.
Therefore, the current study focused on the relationship between physical fitness,
PA, and activity type in preschoolers.
Our previous study showed that in medium-intensity activity time, nonlocomo-
tive activity accounted for a greater proportion of time than locomotive activity in
Japanese preschoolers. Furthermore, the relationship between PA and environmental
factors in Japan may be different than those in other countries. In our previous study
(29), MVPA was significantly related to crime safety and nature in the neighborhood
after adjustment for age, gender, body height, and weight. The presence of hills
was significantly associated with less time spent in PAR ≥ 4. However, a review
by Hinkley et al. (9) showed that crime safety was consistently shown to have no
association with PA in preschool children. Thus, it is important to investigate the
relationship between PA, including nonlocomotive activity, and physical fitness
among preschoolers in Asia, including those in Japan. Previous studies in this area
found evidence for a positive association between fundamental movement skills
competency and PA (8,31). Williams et al. (31) assessed motor skill performance,
including the major fundamental locomotor and object-control skills, and found that
the relationship between PA and locomotor skills was stronger than the relationship
between PA and object-control skills. Fisher et al. (8) reported a modest positive
correlation between fundamental movement skills, including jumping, balancing,
skipping, and playing with a ball, and total PA (r = .10, p = .039) and percent time
in MVPA (r = .18, p < .001). Both sedentary and light-intensity activities were not
significantly correlated with physical skill in these studies. Recently, Bürgi et al.
(3) showed partial correlations of PA with physical fitness in terms of agility (total
PA, r = -.17; MPA, r = -.15; VPA, r = -.13), balance (total PA, r = .20; MPA, r = .22;
VPA, r = .20), and aerobic fitness (total PA, r = .38; MPA, r = .30; VPA, r = .37). In
the current study, these relationships were comparable to those in previous studies
for some skill-related physical fitness parameters, except for balance (Table 3).
Moreover, there was a discrepancy in the relationship with balance, as no significant
relationship was observed in the current study while significant relationships were
observed with MPA and VPA in the study by Bürgi et al. (3). Thus, health-related
fitness (muscle strength and aerobic fitness) as well as skill-related fitness (power,
agility, and speed) are correlated with PA, especially higher intensity PA (PAR ≥
4) in preschool children. These results are similar to those in the previous studies,
except for balance. On the other hand, Cliff et al. (4) reported a negative association
430  Tanaka et al.

between locomotive skills and MVPA only in girls aged 3–5 years, while a positive
association between objective-control skills and MVPA only in boys.
One possible reason for some of the variability in the relationship between PA
and physical fitness is due to differences in assessment methods among studies.
While the current study evaluated physical fitness using continuous variables, a
study by Fisher et al. (8) used the Movement Assessment Battery in young children
and that by Williams et al. (31) used the CHAMPS Motor Skill Protocol to assess
physical skills, observing the presence or absence of each skill. There were also
discrepancies in the correlation of balance with PA since the current study evaluated
static balance and Burgi et al. used a test of dynamic balance (3). Although both
methods are used to evaluate children’s equilibrium, in static balance, the subject is
supported by having one foot by ground, but dynamic balance uses more instabil-
ity. Thus, dynamic balance might more accurately reflect free-living conditions.
Another possible reason is the differences between process (in previous studies)
and product (in the current study) scores. As far as we are aware, the current study
is the first to investigate the relationship of PA with muscle strength (showing a
significant correlation) and flexibility (which showed no correlation with PA), and
more detailed measures of skill-related physical fitness.
In previous studies, habitual PA was evaluated using a CSA/MTI uniaxial
accelerometer (8) or an ActiGraph uniaxial accelerometer (3,4,31), using cutoffs
of the acceleration counts in each study. However, studies by our group showed
that discrimination between locomotion and play in preschool children (e.g., ball
tossing) using a triaxial accelerometer (ActivTracer, GMS) and a vertical/horizontal
counts ratio as a classification criterion contributes to better evaluation of PA inten-
sity (10,27). When PA was assessed by this method, much more time was spent in
nonlocomotive activity than in locomotive activity (104 ± 57 min/day and 140 ±
49 min/day, respectively) in the current study. To assess PA in preschool children
using triaxial accelerometry it is useful to use such an algorithm to discriminate
between different types of medium-intensity activities including locomotive and
nonlocomotive activity (10, 27) to evaluate PA intensity. The previous studies did not
discriminate between locomotive and nonlocomotive PA. Thus, that differences in
methods used to assess PA may also account, in part, for differences in the outcomes
of the three studies. Additional reasons, such as a lower proportion of overweight
children in the current study, preschool environment, etc., also may be responsible
for the differences in PA levels between previous studies and the current study.
Several papers, including Fisher et al. (8), Williams et al. (31), Cliff et al. (4),
and Burgi et al. (3) have reported much smaller amounts of time in MVPA than
the current study and our previous study (28). Our results appear to agree with the
results of other previous studies of youth activity levels measured using a heart-
rate monitor (7). This may be because the use of accelerometers and predictive
equations used in most previous studies of preschool children underestimate the
PA intensity of nonlocomotive activity, as described above. Other reasons, such as
a lower proportion of overweight children in the current study, preschool environ-
ment, etc., may be associated with the different PA levels between previous studies
and the current study.
In the current study, when synthetic activity counts corresponding to medium-
intensity activity were obtained, the amount of time spent in locomotive-type
and non-locomotive-type activity was calculated. Our previous studies (10,27)
 Physical Activity and Physical Fitness in Preschool Children   431

showed that for preschool-aged children, the average PAR value for normal walk-
ing, building with blocks, performing a ball toss, and running was 2.60 or 2.80,
2.98, 3.64 or 4.55, and 5.58 or 5.74, respectively. Therefore, in the current study,
medium-intensity activity would have consisted of activities such as normal and
brisk walking and performing a ball toss. Locomotive activity time had a relatively
strong correlation with light-intensity PA time (r = .61). The average PAR value
of normal walking corresponds to light-intensity PA. In addition, one of the main
components of light-intensity PA may be normal walking, which may also be a
major component of locomotive activity. Nonlocomotive activity time was relatively
strongly correlated with MVPA (r = .58), suggesting that MVPA may consist of
complex PA such as throwing a ball, rather than normal or brisk walking. However,
there was no significant correlation between nonlocomotive activity and physical
fitness parameters except the 25m run (r=-.21 p < .05). Surprisingly, there were
weak negative correlations between nonlocomotive activity and skill-related total
z-scores in spite of a positive correlation between MVPA and nonlocomotive activ-
ity. The reasons for this discrepancy are not clear. However, it is clear that better
physical fitness was associated with increased PA, especially higher-intensity PA.
To further investigate this point, it is necessary to determine the nature of nonlo-
comotive activities in future studies.
As mentioned above, although the results of this and other studies (3,8,31)
suggest that the association between PA and physical fitness or fundamental motor
skills may be relatively weak, there may be a threshold of motor proficiency at
which this relationship is most important. Thus, it is important to examine these
relationships by analysis of covariance. Fisher et al. (8) reported an association
between quartiles of motor skill performance and time spent in MVPA. Williams et
al. (31) also found that young children with better-developed motor skills (highest
tertile) spent significantly more time in both MVPA and vigorous PA and signifi-
cantly less sedentary time than children with less-developed motor skills. More-
over, they showed that children with higher levels of locomotor skill performance
spent significantly more time in both MVPA and vigorous PA and less sedentary
time than children with lower levels of locomotor skill performance. However,
these indices of locomotor skill consisted of behavioral descriptors or process
characteristics for six locomotor activities. Evidence from our study supports the
assertion that not only skill-related but also health-related physical fitness may be
important factors in promoting a physically active lifestyle in preschool children
(Table 4). There were no differences between the high and low tertiles for sitting
trunk flexion or one-foot balance with eyes open for PA of any intensity. Thus,
flexibility and balance may not influence PA in preschoolers. These data may be
particularly important for providing insight into improving health- and skill-related
physical fitness as a determinant of PA in preschoolers. These findings suggest that
development of muscle strength, endurance capacity power, agility, and speed could
make engagement in PA easier for preschool children, although further research
on the cause-effect relationship is needed.
Boys had significantly higher standing long jump and running speed scores,
but had lower flexibility and balance ability compared with girls (Table 1). Previ-
ous studies support gender differences in physical fitness and fundamental motor
skills, with boys’ performance in motor tasks of strength (running speed and
long jump) exceeding that of girls (15). On the other hand, MVPA, PAR ≤ 4 and
432  Tanaka et al.

PAL were greater for boys than for girls (Table 2). These data are coincident with
previous studies (2). Furthermore, locomotive activity duration for boys was also
greater than that of girls. Cliff et al. (4) reported that among boys, object-control
skills were associated with PA in MVPA and total PA, respectively, after control-
ling for age, SES and z-BMI. On the other hand, locomotor skills were inversely
associated with MVPA among girls after controlling for confounders. The present
study also showed gender differences, but such an opposite relationship between
fundamental motor skills and PA in young girls reported by Cliff et al. (4) were not
observed. Additional research is needed to further examine the gender differences
for the relationship between fundamental movement skills or physical fitness and
PA (2). Before puberty, height and body proportions in Japanese boys and girls are
rather similar (26). Thus, body composition, environmental influences, biological
factors, or their interaction are possible explanations for these gender differences
in physical fitness, including fundamental locomotive motor skills.
Overall, the strengths of our study include: an objective and quantitative
measure of PA classifying locomotive and nonlocomotive activity, the inclusion
of both health-related and skill-related physical fitness parameters, and the use of
a sample population of Japanese preschool children from different school types.
There are, however, several methodological points to be considered in interpreting
these results. Because of the cross-sectional nature of this research, the cause-effect
relationship between physical fitness and PA or activity type cannot be determined.
The results of the current study indicate that some measures of physical fitness
affect PA or activity type, but the opposite conclusion, that is, that PA or activ-
ity type may affect physical fitness, could also be drawn. The accelerometer is a
widely used tool to measure PA, but may not assess all PA such as swimming and
cycling. It should be noted that the present data were recorded in1-min epoch,
which may not be sensitive enough to pick up short bursts of vigorous activity
(20,30). Smaller epoch lengths should be used to better reflect movement patterns
of young children (20,30), but ActivTracer can’t record data for several days by
smaller epoch length than 1-min due to a limitation of the memory. It is a limita-
tion of the current study.

Conclusions
The findings of this study indicate that parameters of skill-related physical fitness
such as power, agility, and speed were significantly associated with moderate to high
PA and not associated with light activity. Development of power, agility and speed
could make engagement in PA easier for preschool children. Moreover, parameters
of health-related physical fitness such as muscular strength and endurance, and
skill-related physical fitness may contribute to higher-intensity PA. These measures
showed differential relationships with locomotive and nonlocomotive activity.
Additional research is needed to further examine the relationship between physical
fitness and PA and activity type. A longitudinal study could provide information on
the direction of this association. If physical fitness parameters are determinants of
PA, then strategies that develop physical fitness in childhood may help to promote
increased PA and health in young children.
 Physical Activity and Physical Fitness in Preschool Children   433

Acknowledgments
The authors would like to thank the subjects, their parents, and the staff of the participating
kindergartens and nursery schools for their cooperation in this study. We also wish to thank
the staff of the National Institutes of Health and Nutrition for their help with the experi-
ments. This work was supported by a Grant-in-Aid for Young Scientists (B) of the Japan
Society for the Promotion of Science (to C. Tanaka) and by a Sports Medicine and Science
Research grant from the Japan Sports Association (to S. Tanaka).

References
1. Biddle, S., J.F. Sallis, and, N. Cavill. Young and active? Young people and health-
enhancing physical activity evidence and implications. United Kingdom: Health
Education Authority, London, 1998, pp. 1-176.
2. Bornstein, D.B., M.W. Beets, W. Byun, and K. McIver. Accelerometer-derived physical
activity levels of preschoolers: a meta-analysis. J. Sci. Med. Sport. 14:504–511, 2011.
PubMed doi:10.1016/j.jsams.2011.05.007
3. Bürgi, F., U. Meyer, U. Granacher, et al. Relationship of physical activity with motor
skills, aerobic fitness and body fat in preschool children: a cross-sectional and longi-
tudinal study (Ballabeina). Int. J. Obes. Lond, 2011, ([Epub ahead of print])
4. Cliff, D.P., A.D. Okely, L.M. Smith, and K. McKeen. Relationships between fundamen-
tal movement skills and objectively measured physical activity in preschool children.
Pediatr. Exerc. Sci. 21:436–449, 2009. PubMed
5. Cliff, D.P., J.J. Reilly, and A.D. Okely. Methodological considerations in using accel-
erometers to assess habitual physical activity in children aged 0-5 years. J. Sci. Med.
Sport. 12:557–567, 2009. PubMed doi:10.1016/j.jsams.2008.10.008
6. 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. BMJ. 320:1240–1243,
2000. PubMed doi:10.1136/bmj.320.7244.1240
7. Epstein, L.H., R.A. Paluch, L.E. Kalakanis, G.S. Goldfield, F.J. Cerny, and J.N. Roem-
mich. How much activity do youth get? A quantitative review of heart-rate measured
activity. Pediatrics. 108:E44, 2001. PubMed doi:10.1542/peds.108.3.e44
8. Fisher, A., J.J. Reilly, L.A. Kelly, et al. Fundamental movement skills and habitual
physical activity in young children. Med. Sci. Sports Exerc. 37:684–688, 2005. PubMed
doi:10.1249/01.MSS.0000159138.48107.7D
9. Hinkley, T., D. Crawford, J. Salmon, A.D. Okely, and K. Hesketh. Preschool children
and physical activity: a review of correlates. Am. J. Prev. Med. 34:435–441, 2008.
PubMed doi:10.1016/j.amepre.2008.02.001
10. Kawahara, J., S. Tanaka, C. Tanaka, Y. Aoki, and J. Yonemoto. Estimation of daily
inhalation rate in preschool children using a tri-axial accelerometer: A pilot study. Sci.
Total Environ. 409:3073–3077, 2011. PubMed doi:10.1016/j.scitotenv.2011.04.006
11. Kurimoto, E., T. Asami, K. Shibukawa, Y. Matsuura, and A. Katsube. Physical coordi-
nation - the final form of the field examination. Report of Research Center in Physical
Education. 9:207–212, 1981. (in Japanese).
12. Laboratory of Physical Fitness Standards Tokyo Metropolitan University (Ed.). New
Physical Fitness Standards of Japanese People, version II. Tokyo: Fumaido Publishing
Co., Ltd., 2007, (in Japanese)
13. Leenders, N.Y., W.M. Sherman, and H.N. Nagaraja. Energy expenditure estimated
by accelerometry and doubly labeled water: do they agree? Med. Sci. Sports Exerc.
38:2165–2172, 2006. PubMed doi:10.1249/01.mss.0000235883.94357.95
434  Tanaka et al.

14. Lubans, D.R., P.J. Morgan, D.P. Cliff, L.M. Barnett, and A.D. Okely. Fundamental
movement skills in children and adolescents: review of associated health benefits.
Sports Med. 40:1019–1035, 2010. PubMed doi:10.2165/11536850-000000000-00000
15. Malina, R.M., C. Bouchard, and O. Bar-Or. Growth, Maturation, and Physical Activity,
2nd ed. Human Kinetics Pub, 2004.
16. Matthews, C.E. Calibration of accelerometer output for adults. Med. Sci. Sports Exerc.
37(11, Suppl):S512–S522, 2005. PubMed doi:10.1249/01.mss.0000185659.11982.3d
17. Mattocks, C., K. Deere, S. Leary, et al. Early determinants of physical activity in 11
to 12 year olds: cohort study. Br. J. Sports Med. 42:721–724, 2008. PubMed
18. National Association for Sport and Physical Education. Active start: a statement of
guidelines for children birth to age 5, 2nd ed. Reston, VA: Author, 2009.
19. National Association for Sport and Physical Education. Physical activity for children:
a statement of guidelines for children ages 5-12, 2nd Edtion. Reston, VA: Author, 2004
20. Nilsson, A., U. Ekelund, A. Yngve, and M. Sjostrom. Assessing physical activity among
children with accelerometers using different time sampling intervals and placements.
Pediatr. Exerc. Sci. 14:87–96, 2002.
21. Ohkawara, K., K. Ishikawa-Takata, J.H. Park, I. Tabata, and S. Tanaka. How much
locomotive activity is needed for an active physical activity level: analysis of total step
counts. BMC Res Notes. 4:512, 2011. [Epub ahead of print]. PubMed doi:10.1186/1756-
0500-4-512
22. Oliver, M., G.M. Schofield, and G.S. Kolt. Physical activity in preschoolers: understand-
ing prevalence and measurement issues. Sports Med. 37:1045–1070, 2007. PubMed
doi:10.2165/00007256-200737120-00004
23. Parsons, T.J., C. Power, S. Logan, and C.D. Summerbell. Childhood predictors of adult
obesity: a systematic review. Int. J. Obes. Relat. Metab. Disord. 23(Suppl. 8):S1–S107,
1999. PubMed
24. Penpraze, V., J.J. Reilly, C.M. MacLean, et al. Monitoring of physical activity in young
children: how much is enough? Pediatr. Exerc. Sci. 18:483–491, 2006.
25. Strong, W.B., R.M. Malina, C.J. Blimkie, et al. Evidence based physical activ-
ity for school-age youth. J. Pediatr. 146:732–737, 2005. PubMed doi:10.1016/j.
jpeds.2005.01.055
26. Tanaka, C., M. Murata, M. Homma, and T. Kawahara. Reference charts of body
proportion for Japanese girls and boys. Ann. Hum. Biol. 31:681–689, 2004. PubMed
doi:10.1080/03014460400019794
27. Tanaka, C., S. Tanaka, J. Kawahara, and T. Midorikawa. Triaxial Accelerometry for
Assessment of Physical Activity in Young Children. Obesity (Silver Spring). 15:1233–
1241, 2007. PubMed doi:10.1038/oby.2007.145
28. Tanaka, C., and S. Tanaka. Daily physical activity in Japanese preschool children
evaluated by triaxial accelerometry: the relationship between period of engagement
in moderate-to-vigorous physical activity and daily step counts. J. Physiol. Anthropol.
28:283–288, 2009. PubMed doi:10.2114/jpa2.28.283
29. Tanaka, C., S. Tanaka, and T. Ando. Relationship between daily physical activity and
neighborhood environment among Japanese preschool children. Jap. J. Human Growth
and Development Res. 51:37-45 (in Japanese)
30. Vale, S., R. Santos, P. Silva, L. Soares-Miranda, and J. Mota. Preschool children
physical activity measurement: importance of epoch length choice. Pediatr. Exerc. Sci.
21:413–420, 2009. PubMed
31. Williams, H.G., K.A. Pfeiffer, J.R. O’Neill, et al. Motor skill performance and physical
activity in preschool children. Obesity (Silver Spring). 16:1421–1426, 2008. PubMed
doi:10.1038/oby.2008.214