Vol.2, No.12, 1377-1381 (2010) Health
Copyright © 2010 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/
Change in stature by walking and running at a preferred
transition speed
Shinichi Demura1, Takayoshi Yamada2*, Tamotsu Kitabayashi3, Masanobu Uchiyama4
1Kanazawa University, Graduate School of Natural Science & Technology Kakuma, Kanazawa, Ishikawa, Japan;
2Fukui National College of Technology, Department of Health & Physical Education Geshi, Sabae, Fukui, Japan;
*Corresponding Author: takay@fukui-nct.ac.jp;
3 Tokyo University of Science, Department of Physical Education and Sports Science, 1-3 Kagurazaka Shinjuku-ku, Tokyo, Japan;
4Akita Prefectural University, Research and Education Center for Comprehensive Science 241-438 Kaidobata-Nishi Nakano Shimo-
shinjo Akita City, Japan. uchiyama@akita-pu.ac.jp
Received 12 September 2010; revised 2 October 2010; accepted 9 October 2010
This study aimed to measure stature changes
during and after walking and running at a pre-
ferred transition speed (PTS) and the recovery
period, and to examine differences caused by
loads imposed on the spinal column. Seven
males and three females aged 22-41 years took
part in this study. Subjects The subjects un-
derwent 15 minutes of walking or running on a
treadmill in a random order. Stature changes
were measured during each exercise at inter-
vals of 5 minutes and after a 20 minute standing
recovery period within units of 0.01 mm. Two-
way ANOVA revealed that both main factors,
gait (F = 5.250, P < 0.05) and elapsed time (F =
14.409, P < 0.05), had a significant effect on
stature. In the post hoc test, stature shrank w ith
time and its loss was found to be greater in
running than in walking, but recovered after
both exercises. In conclusion, the spinal load in-
creases with time during both walking and run-
ning at PTS, but is greater in running than in
walking. After both exercises, spinal shrinkage
shows a similar recovery process and recovers
faster in walking to its pre-exercise level.
Keywords: Preferred Transition Speed; Wal king;
Running; Change in Stature; Spinal Shrinkage
Walking and running are fundamental modes in hu-
man locomotion. Until now, the preferred transition
speed (PTS) between walking and running has been stu-
died from various viewpoints and it has been found that
the energy cost of locomotion would not be the trigger to
determine its speed [1]. In short, the energy consumed
by running that exceeds walking is at a faster level than
PTS [1]. It has also been also known that the burden of
the dorsiflexor muscles of the ankle [2,3], force produc-
tion of plantar flexors of the ankle [4], and peak ante-
rior/posterior and vertical ground reaction forces during
the propulsion phase [4] relate to the PTS. In other
words, the mechanical stress to the muscular- skeletal
system is considered to be an important factor in decid-
ing the transition speed between walking and running.
The PTS is a faster speed than walking, but slower
one than running. It was reported to be about 2 ms-1 in
adults according to Hreljac [1], and is very close to the
habitual walker’s pace (average of 1.78 ms-1; Spelman
et al.[5]) or brisk walking pace for fitness training (av-
erage of 1.76 ms-1; Hardman and Hudson. [6]). In ad-
dition, it would be comparable to the “jogging” pace.
Thus, PTS may be an exercise intensity that corresponds
to an aerobic training level in healthy adults. Recently,
there has been a focus on the necessity of appropriate
daily activity. A mode of walking or running is a useful
means of fitness training and can be easily performed by
people of all ages. Therefore, information on the safety
of these modes of exercise is very important. Even at the
same speed, walking has a smaller impact force as com-
pared to running [7-9]. In addition, walking has merits
such as lower intensity and low risks to th e cardiovascu-
lar system. These are undoubtedly reasons why walking
is widely recommended as a mode of exercise. On the
other hand, from the prevention of back injuries, many
researchers have studied an exact change of stature to
evaluate the spinal load imposed by various tasks or ex-
S. Demura et al. / Health 2 (2010) 1377-1381
Copyright © 2010 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/
Generally, stature is subjected to a typical pattern of
circadian variation. This variation is produced by com-
pression alterations of the intervertical disc height re-
sulting from the shrinkage in the viscoelastic properties
of the disc [10-13]. In addition, it has been found that
this shrinkage depends largely on the magnitude of the
spinal load and when the load of activities or tasks is
greater; also, acute shrinkage is more meaningful [10,
12-17]. From the above, clarifying the different burdens
imposed on the spinal column between walking and run-
ning will offer useful data to prevent back injuries.
This study aimed to clarify the change in stature and
the difference in spinal load during walking and running
at a preferred transition speed, and to compare the re-
covery process of the stature after both exercises.
2.1. Subjects
Seven males and three females aged 22-41 years took
part in the study. Their physical characteristics were
shown in Table 1. All were physically active and none of
them had a history of any disorder or injury affecting
their ability to walk. Before participating in this study,
written informed consent was obtained from them.
2.2. Determination of the Preferred
Transition Speed (PTS)
The same speed is the major premise for comparing
the spinal load between walking and running. In this
study, the PTS, which corresponds to a level of fitness
training [5,6], was used as the speed for both exercises.
The PTS was decided according to Hreljac’s method [1]
for all subjects on a day prior to the measurement of
stature change. For walking, the treadmill speed was
initially set as 1.38 ms-1 (5 kmh-1) and was increased
Tab le 1. Physical characteristics and preferred transition speed
(PTS) of subjects.
Subject Sex Age
(yr) Stature
(m) Body mass
(kg) PTS
KT M 23 1.72 73.0 2.00
YT M 24 1.73 63.7 1.75
TN M 35 1.88 102.3 2.00
AN M 36 1.75 50.9 1.94
JO M 33 1.62 53.6 1.86
AA M 40 1.63 56.9 1.94
AS M 42 1.73 82.9 1.75
EK F 22 1.62 60.9 1.94
ME F 23 1.64 54.2 1.89
YK F 26 1.64 47.7 2.03
Mean 30.4 1.70 64.6 1.91
SD 7.6 0.08 17.0 0.10
0.14 m/s (0.5 km/h) every 30 sec. Subjects were in-
structed to run wh en they felt running to be more natural
than walking on the treadmill. This speed was d efined as
the walk to run transition speed (WRTS). For running,
the treadmill speed was decreased from 2.22 m/s (8 km/
h) by about 0.14 ms-1 (0.5 kmh-1) every 30 seconds.
The entire process was repeated twice in a random order.
These measurements were defined as WRTS and RWTS,
respectively. The average of these values was used to
find the PTS based on Hreljac’s method [1].
2.3. App aratus
The stature change was measured using a stadiometer
(Figure 1) with a digital height gauge (MonotaRo Co.,
Ltd, #131-103, Japan) as described by Boocock et al.
[14] and Rodacki et al. [18], and followed their meas
urement protocol. Subjects remained in a standing posi-
tion for 2 minutes to minimize the effect of soft tissue
creep deformation of the lower limbs [19] in all mea-
surements. The stadiometer was sensitive within 0.01
mm. Measurements were completed when the subject
could reproduce ten consecutive measures with a stan-
dard deviation of less than 0.5 mm [18], and their mean
Digital Height Gauge
(MonotaRo Co.,Ltd, #131-103, Japan)
Figure1. Stadiometer, the apparatus for measuring
change in stature.
S. Demura et al. / Health 2 (2010) 1377-1381
Copyright © 2010 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/
used as a representative value. For example, when the
standard deviation of ten consecutive measures was
larger than 0.5 mm, the ten consecutive measurements
were repeated.
2.4. Experimental Procedure
Subjects participated in walking and running experi-
ments with PTS intensity at the same time slot (08:00-
12:00). They were required to refrain from stressful
physical activities which are more strenuous than walk-
ing for 24 hours before the experiment, to stand for a
period of 20 minutes to distribute their weight evenly
and to standardize and control any abnormal spinal
loading or unloading before the experiment. The subjects
then underwent walking or running for 15 minutes on
the treadmill in a random order. Their stature was meas-
ured prior to ( as a baseline) and during each exercise
at intervals of 5 minutes. In brief, three measurements
with a set of 10 consecutive measurements [20] were
carried out at walking or running for 5, 10, and 15 mi-
nutes intervals (, , and ). In addⅢⅣ ition, the stature
was finally measured () just after the 20 minute standing
recovery. Exercise for health is generally over 10 mi-
nutes [21]. Considering the above, we used 15 minutes
for walking or running time.
2.5. Statistical Analysis
Reliability coefficients were calculated from ten con-
secutive measurements taken at each measurement pe-
riod (, , , , and ). Stature changes were caⅠⅡⅢⅣ Ⅴl-
culated by subtracting individual values of , , , ⅡⅢⅣ
and from the value of (baseline), respectively. The
changes were analyzed using 2 factors, modes of exer-
cise (gait: walking and running) by 4 periods of elapsed
time (after 5, 10, 15 minutes, and recovery) analysis of
variance (ANOVA) with repeated measurements of the
last factor. A Scheffe post hoc test was used to clarify
significant differences between two means. A prob ability
level of less than 0.05 was set.
Higher reliability coefficients in measurements of sta-
ture changes were found with a range of 0.904-938 (Table
2). Figure 2 shows means and standard deviations of sta-
ture changes following walking and running in each time
interval including the recovery. ANOVA revealed that
both main effects of gait (F = 5.250, P < 0.05) and elapsed
time (F = 14.409, P < 0.05) were significant, but the int e-
raction was insignificant (F = 1.653, P > 0.05). In
short, the stature loss of walking was greater than that
of running, and stature changes in both walking and
running increased with time elapsed until the 15 minute
Table 2. Reliability coefficients of 10 consecutives measure-
ments at each period.
Walking Running
0.938 0.904
0.928 0.912
0.942 0.912
0.930 0.912
0.932 0.921
: before walking or running, : after walⅠⅡking or runnning for 5
min, : after walking or running for 10 min, : after walking or
running for 15 m in, : after recovery (20 min)
Change of stature (mm)
Time (min)
Exercise Recovery
Figure 2. Means and 1SDs of the change of stature af-
ter exercises (walking and running) and recovery.
interval. In the post hoc test, a significant difference in
stature change was found between time intervals except
for between the 10 and 15 minute intervals. On the other
hand, the stature loss was significantly recovered and got
closer to the baseline after the recovery period.
Most research that investigated the accurate stature
change have used a technique in which ten consecutive
measurements reached stable data within a standard
deviation of less than 0.5 mm [10,12-18]. However, the
reliability of measurements by this technique has been
not examined. Reliability coefficients of more than 0.9
were confirmed in this study. Thus, it was judged to be a
valid technique. Circadian variation of stature has been
known to be 19.3 mm or 1.1% of overall stature [11]. As
another perspective, the stature change has been widely
studied as an index of spinal load (spinal shrinkage) in
ergonomics [22]. For example, it was found that the
greater load (stimulus) of the gravity to the spine would
Walking Running
S. Demura et al. / Health 2 (2010) 1377-1381
Copyright © 2010 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/
afford the reasonable stature change from results of sev-
eral studies on stature changes in the drop jump [10,14]
or weightlifting [15,16] and walking in obese people
[20]. In addition, it was also reported that these acute
changes (losses) in stature would recover after exercise
or tasks [10,14]. The present stature losses following
walking and running at PTS were significant and marked
between the 5 and 10 minute interval, being also signifi-
cant between 5 and 15 minutes, but not between 10 and
15 minutes. Based on Reilly et al.’s report [11], the cir-
cadian variation of the stature of the present subjects is
estimated to be about 18.7 mm (170 cm × 0.011) on the
basis of mean height (1.7 m). Average losses of stature
while walking (2.5 mm) and running (4.2 mm) for 15
minutes at PTS were respectively 13.4% and 22.5% of
the circadian variation. In addition, the stature loss by
both exercises increased with time. Assuming active
times of day excluding sleep time as 16 hours, the ratio
of the stature loss in the circadian variation is estimated
to be 0.019 mmmin-1 (18.7 mm/960 min). In contrast, a
loss rate of the stature was 0.17 mmmin-1 in walking
and 0.28 mmmin-1 in running. In short, a larger loss
rate is considered to occur during exercise. The change
after 15 minutes of exercise is especially larger in run-
ning than in walking. It was reported that ground reac-
tion forces at the same speed are greater in running than
in walking [7-9]. In short, the spinal load is greater in
running. The difference of this impact force may pro-
duce a difference in stature loss. The recovery of the
absolute stature loss (spinal shrinkage) within the same
time (20 minutes), on average, was almost the same in
walking (1.5 mm) and running (1.6 mm). However,
when calculating the statu re loss after 15 minutes as cri-
teria, the recovery rate was estimated to be 58% in
walking and 39% in running. Thus, the recovery period
shows a similar recovery process in walking and running,
but, after 15 minutes, the recovery may be faster in
walking than in running because of the smaller spinal
load. Rodacki et al. [20] reported that the stature loss in
the non-obese (aged 22.4 yr on average) was found to be
3.00 mm (estimated from a figure) and 3.55 mm follow-
ing walking for 15 and 30 minutes, respectively. Ac-
cording to Garbutt et al. [17], the stature loss in mara-
thon runners decreased to 3.26 mm and 7.69 mm after
running at a normal marathon pace for 15 minutes and
30 minutes, respectively. The stature loss after running
for 15 minutes in this study was very similar to that in
the above studies. In addition, the stature loss after 15
minutes in the above studies tended to be relatively
small. A difference in stature change between 10 and 15
minutes in this study was not found either. Hence, the
stature change (loss) may be greater just after beginning
exercise. This may be associated with the viscoelasticity
of intervertical discs, in which the shape of the disc
changes rapidly in th e initial stage of the exercise or task,
and is followed by an exp elling of fluid from the nu cleus
pulposus [20,23]. It is known that reducing the interver-
tical disc height (spinal shrinkage) decreases the ability
of the disc to absorb and transmit forces, and the resis-
tance to failure [24]. In addition, the spinal shrinkage
associated with compressed intervertical discs is be-
lieved to increase susceptibility to spinal injuries. As
epidemiological data, it was reported that there is a sig-
nificant relationship between lumbar disorders and tasks
involving heavy manual lifting [25]. Regular sports par-
ticipants also have been id entified as a high risk popula-
tion for back injuries because of repeated high impact
forces [26].
In conclusion, the spinal load is greater in running
than in walking at PTS, and increases as time elapses
during exercise. The recovery process of the spinal load
is similar in both walking and running after exercise, but
the degree of recovery of the stature loss to the
pre-exercise level is faster in walking. Hence, walking
can be recommended as a mode of fitness training rather
than running for the safety of the spin e or for the preven-
tion of back injuries.
[1] Hreljac, A. (1993) Preferred and energetically optimal
gait transition speeds in human locomotion. Medicine
and Science in Sports and Exercise, 25, 1158-1162.
[2] Hreljac, A. (1995) Determinants of the gait transition
speed duri ng huma n l ocomoti on: ki nema tic f actors. Journal
of Biomechanics, 28, 669-677.
[3] Hreljac, A., Imamura, R.T., Escamilla, R. F., Edwards, W.
B. and MacLeod, T. (2008) The relationship between
joint kinetic factors and the walk-run gait transition
speed during human locomotion. Journal of Applied
Biomechanics, 24, 149-157.
[4] Neptune, R.R. and Sasaki, K. (2005) Ankle plantar flexor
force production is an important determinant of the pre-
ferred walk-to-run transition speed. Journal of Experi-
mental Biology, 208, 799-808.
[5] Spelman, C.C., Pate, R. R., Macera, C. A. and Ward, D. S.
(1993) Self-selected exercise intensity of habitual walk-
ers. Medicine and Science in Sports and Exercise, 25,
[6] Hardman, A.E. and Hudson, A. (1994) Brisk walking and
serum lipid and lipoprotein variables in previously se-
dentary women—effect of 12 weeks of regular brisk
walking followed by 12 weeks of detraining. British
Journal of Sports Medicine, 28, 261-266.
[7] Cavanagh, P.R. and Lafortune, M.A. (1980) Ground
reaction force in distance running. Journal of Biome-
chanics, 13, 397-406.
S. Demura et al. / Health 2 (2010) 1377-1381
Copyright © 2010 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/
[8] Chao, E., Laughman, R., Schneider, E. and Stauffer, R.
(1983) Normative data of knee joint motion and ground
reaction force in adult level walking. Journal of Biome-
chanics, 16, 219-233.
[9] Nilsson, J. and Thorstensson, A. (1989) Ground reaction
forces at different speeds of human walking and running.
Acta Physiologica Scandinavica, 136, 217-227.
[10] Fowler, N. E. and Reilly, T. (1997) Changes in stature
following plyometric drop-jump and pendulum exercises.
Ergonomics, 40, 1279-1286.
[11] Reilly, T., Tyrrell, A. and Troup, J.D.G. (1984) Circadian
variation in human stature. Chronobiology International,
1, 121-126.
[12] Reilly, T., Boocock, M.G., Garbutt, G., Troup, J.D. and
Linge, K. (1991) Change in stature during exercise and
sports training. Applied Ergonomics, 22, 308-311.
[13] Reilly, T. and Freeman, A. (2006) Effects of loading on
spinal shrinkage in males of different age groups. Ap-
plied Ergonomics, 37, 305-310.
[14] Boocock, M.G., Garbutt, G., Reilly, T. and Troup, D. G.
(1990) Change in stature following drop jumping and
post-exercise gravity inversion. Medicine and Science in
Sports and Exercise, 22, 385-390.
[15] Bourne, N.D. and Reilly, T. (1991) Effect of a
weightlifting belt on spinal shrinkage. British Journal of
Sports Medicine, 25, 209-212.
[16] Fowler, N.E. and Reilly, T. (1994) Spinal shrinkage in
unloaded and loaded drop-jumping. Ergonomics, 37, 133
[17] Garbutt, G., Boocock, M.G., Reilly, T. and Troup, J.D.G.
(1990) Running speed and spinal shrinkage in runners
with and without low back pain. Medicine and Science in
Sports Exercise, 22, 769-772.
[18] Rodacki, C.L.N., Fowler, N.E., Rodacki, A.L.F. and
Birch, K. (2001) Repeatability of measurement in deter-
mining stature in sitting and standing posture. Ergonom-
ics, 12, 1076-1085.
[19] Foreman, T.K. and Linge, K. (1989) The importance of
heel compression in the measurement of diurnal stature
variation. Applied Ergonomics, 20, 299-300.
[20] Rodacki, A.L.F., Fowler, N.E., Provensi, C.L.G., Rodacki,
C.L.N. and Dezan, V. H. (2005) Body mass as a factor in
stature change. Clinical Biomech an ic s, 20, 799-805.
[21] Haskell, W.L., Lee, I.M., Pate, R.R., Powell, K.E., Blair,
S.N., Franklin, B.A., Macera, C. A., Heath, G.W., Thompso n,
P.D. and Bauman, A. (2007) Physical activity and public
health: Updated recommendation for adults from the
American College of Sports Medicine and the American
Heart Association. Medicine and Science in Sports and
Exercise, 39, 1423-1434.
[22] Corlett, E.N., Eklund, J.A.E., Reilly, T. and Troup, J.D.G.
(1987) Assessment of workload from measurement of
stature. Applied Ergonomics, 18, 65-71.
[23] Adams, M.A. and Dolan, P. (1995) Recent advances in
lumbar spinal mechanics and their clinical significance.
Clinical Biomechani c s, 10, 3-19.
[24] Perey, O. (1957) Fracture of the vertebral end plate in the
lumbar spine: An experimental biomechanical investiga-
tion. Acta Orthopaedica Scandinavica Supplementum, 25,
[25] Troup, J.D.G. (1965) Relation of lumbar spine disorders
to heavy manual work and lifting. Lancet, 17, 857-861.
[26] Alexander, M.J.L. (1985) Biomechanical aspects of
lumbar spine injuries in athletes: a review. Canadian
Journal of Applied Sport Sciences, 10, 1-20.