ORIGINAL ARTICLE

Juha E. Peltonen Simo Taimela Minna Erkintalo

Jouko J. Salminen Airi Oksanen Urho M. Kujala
Back extensor and psoas muscle cross-sectional area,
prior physical training, and trunk muscle strength
a longitudinal study in adolescent girls
Accepted: 10 June 1997
Abstract The association between physical training, low
back extensor (erector spinae plus multidus muscles)
and psoas muscle cross-sectional areas (CSA) and
strength characteristics of trunk extension and exion
were studied in adolescent girls. A group of athletes
(n 49) (age range 13.716.3 years) consisting of gym-
nasts, gure skaters and ballet dancers was age-matched
with non-athletes (n 17) who acted as a sedentary
control group. The CSA of psoas muscles and multidus
plus erector spinae muscles were measured from lumbar
axial images by magnetic resonance imaging. Maximal
trunk extension and exion forces were measured in a
standing position using a dynamometer and trunk
musculature endurance was evaluated using static
holding tests. When CSA were adjusted with body mass,
the athletes showed signicantly greater CSA in both
muscles studied (psoas P < 0.001; erector spinae plus
multidus P < 0.05) than the non-athletes. The athletes
also had a greater absolute psoas muscle CSA
(P < 0.01) and trunk exion force (P < 0.01) com-
pared to the controls. When the forces were expressed
relative to body mass, the athletes were superior both in
trunk exion (P < 0.001) and extension (P < 0.001).
There was a signicant correlation between muscle CSA
and strength parameters, but the force per muscle CSA
did not dier signicantly between the athletes and the
non-athletes. In addition, the athletes showed a better
body mass adjusted muscle endurance in trunk exion
(P < 0.05) than the non-athletes. Our study indicated
that regular physical training enhances trunk muscula-
ture hypertrophy, force and endurance in adolescent
girls, and that there is an association between muscle
CSA and strength parameters.
Key words Magnetic resonance imaging
Muscle cross-sectional area Force
Athletes Non-athletes
Introduction
The relationship between previous physical loading and
present trunk muscle function has two dimensions.
Firstly, it has been shown that adequate physical loading
is required to achieve and maintain normal muscle and
joint function, and to prevent deconditioning due to
disuse (Nachemson and Lindh 1969; Thorstensson and
Arvidsson 1982; Mayer et al. 1985a). On the other hand,
it has been established that physical overloading or ac-
cidental injury may cause damage to spinal tissues, acute
back pain and subsequent risk of chronic pain (Kujala
et al. 1996a, b). It has recently been noticed that the rst
episode of low back pain (LBP) may lead to multidus
muscle atrophy within weeks, and that the recovery of
the muscle is not automatic (Hides et al. 1996). It has
also been widely accepted that chronic LBP leads to
impaired muscle function due to the detrimental eects
of relative immobilization (Nachemson and Lindh 1969;
Thorstensson and Arvidsson 1982; Mayer et al. 1985b)
probably because of atrophy of the low back muscles
(Mattila et al. 1986; Zhu et al. 1989; Rissanen et al.
1995). However, the studies analysing the association
between the history of physical loading, trunk muscle
performance factors and future back problems have
given somewhat conicting results (Kujala et al. 1996b).
Thus, it is to some extent doubtful whether theoretically
advantageous large trunk muscles and good trunk
muscle function would in reality protect from low back
pain.
Parkkola et al. (1992) have found that a strength
training programme performed by sedentary students
Eur J Appl Physiol (1998) 77: 6671 Springer-Verlag 1998
J.E. Peltonen (&) S. Taimela U.M. Kujala
Unit for Sports and Exercise Medicine, Institute of Biomedicine,
University of Helsinki, Mannerheimintie 17, FIN-00250, Helsinki,
Finland
M. Erkintalo
Diagnostic Imaging Center, University of Turku, Turku, Finland
J.J. Salminen A. Oksanen
Department of Physical Medicine and Rehabilitation,
Turku University Hospital, Turku, Finland
increased both low back muscle cross-sectional area
(CSA) and static trunk extension and exion force.
However, no signicant correlation between CSA and
force was detected before or after training. Therefore,
the existing evidence for the association between low
back muscle CSA and trunk muscle strength has been
found to be weak (Parkkola et al. 1992). In particular we
are lacking knowledge on the training eects of low back
muscles in adolescents. There has been no general
agreement as to whether strength training induces an
increase in muscle CSA among adolescents at all, al-
though an improvement in muscle strength is normally
found. Ramsay et al. (1990) and Blimkie (1993) have
argued that resistance training programmes in children
do not induce increase in muscle CSA. In their judg-
ment, strength gain amongst adolescents that is smaller
than amongst young adults, is likely to have been due to
improved neural function. On the other hand, Fukunaga
et al. (1992) have reported an increment of muscle CSA
of upper arm amongst prepubescent boys and girls that
was about 50% of that normally found in adults after a
similar strength training period.
The purpose of the present study was to analyse the
association between prior physical training, back ex-
tensor and psoas muscle CSA, and trunk exion and
extension strength and endurance in adolescent girls
without severe LBP. The hypothesis was that girls who
participated regularly in gure skating, gymnastics or
ballet would show greater maximal trunk extension and
exion forces than their sedentary age-matched controls.
Whether the expected superiority in strength tests would
be explained by trunk musculature CSA was studied by
magnetic resonance imaging (MRI). Another purpose of
the study was to serve as the baseline for an intended use
of the same subjects in the future to analyse the occur-
rence of LBP and other low back disorders in relation to
physical activity, strength and low back muscle CSA in
adolescence.
Methods
Subjects
A group of 80 girls (18 non-athletes, 20 gymnasts, 20 gure skaters,
22 ballet dancers; age range 10.7 to 13.3 yr.) had been selected into
this longitudinal study 3 years before the MRI and muscle strength
examinations (Kujala et al. 1992). Informed written consent was
obtained from each subject and from a parent or guardian. The
experiment protocols were approved by the local Ethics Commit-
tee. At the beginning the athletes had been regularly training a
minimum of twice a week for at least 2 years, supervised by coaches
in various sports clubs. Mean weekly training hours and frequency
were recorded yearly by a questionnaire. The subjects who declined
to participate in any of the three follow-up examinations were ex-
cluded from our study, as well as the athletes who gave up com-
petitive sports during the 1st follow-up year. The subjects whose
level of physical activity changed during the 2nd or 3rd follow-up
years were included in the nal study groups. Each of the athletes
and dancers had performed physical exercise corresponding to her
grouping for at least 3 years before the 2nd follow-up year.
Thus, during the 3rd follow-up year, the group of 49 athletes
consisted of physically active girls (17 gymnasts, 14 gure skaters,
18 ballet dancers). The physically sedentary group (n 17) were
pupils of the same age attending two elementary schools who
participated in recreational sports at most sporadically (less than
twice a week). Ages, anthropometric data and training status of the
control and experiment groups are presented in Table 1.
Strength tests
Maximal trunk extension and exion forces (in newtons) were
measured in a standing position by a dynamometer (Digitest,
Muurame, Finland). When measuring the force, a collar was xed
around the subject's chest and a strain-gauge transducer was at-
tached horizontally between the collar and a vertical support bar.
Two horizontal bars, one above the patella and another at the level
of the iliac crest, gave support to the subject. After warming up, the
subjects performed two maximal isometric trunk extensions and
exions, each trial lasting 35 s, with rests of 60 s between them.
The better of the two trials was used for further analysis. The
relative maximal forces were calculated by dividing the force values
by body mass (newtons per kilogram body mass). The reproduc-
ibility of the tests of maximal isometric strength of the trunk ex-
tension and exion expressed as the coecient of variation between
duplicate measurements has been shown to be between 2% and 5%
(Viljanen et al. 1991). To measure the endurance strength of the
abdominal and hip extensor muscles, the subject tested was asked
to curl up from the supine position, with straight arms besides the
knees, toward the knees until the line drawn between the crests of
the ilia could be seen. This posture was held as long as possible, but
maximally for 240 s (see Kujala et al. 1992). While performing the
test, the subject was not supported from her ankles.
Endurance strength of the back muscles and trunk extensor
muscles was examined with the subject lying prone, the upper body
down to the crista iliaca being o the examination table. The lower
body and the legs of the subject were xed to the table, and the
subject was asked to hold the upper body and head horizontally as
long as possible, but maximally for 240 s (see Biering-So rensen
1984; Kujala et al. 1992). The repeatability of the measurements
has been tested in an earlier study by Salminen et al. (1992).
Muscle CSA
The MRI examinations were performed using a 0.1 T imager
(Merit, Picker Nordstar, Helsinki, Finland). The CSA of psoas
muscles and multidus plus erector spinae muscle groups were
measured from lumbar axial images, which were obtained parallel
Table 1 Age, height, body mass,
the number of training years
and the amount of training
weekly during the last 12
months
Non-athletes
Mean SD
Figure skaters
Mean SD
Gymnasts
Mean SD
Ballet dancers
Mean SD
n 17 17 14 18
Age (years) 14.9 0.4 14.6 0.7 14.7 1.0 14.8 0.6
Height (cm) 164.5 7.2 160.8 6.7 159.6 4.8 164.5 5.6
Body mass (kg) 57.0 7.8 51.6 8.0 50.0 3.6 49.2 6.2
Training years 8.4 1.7 6.2 1.5 6.4 1.9
Training (min week
)1
) 116 129 768 362 510 594 627 244
67
to the L4/L5 disc space (Fig. 1). Two axial 10-mm slices were
produced using a gradient echo 120/20-ms sequence with an
imaging time of 2 min 3 s. Using a eld-of-view of 307 307 mm
and a matrix of 256 256 a pixel size of 1.2 1.2 mm was
achieved. The CSA was measured from the slice adjusted to the
middle of the L4/L5 disc space.
The actual CSA of psoas muscles and multidus/erector spinae
muscle groups were measured using a freehand area calculation
facility provided by the MRI system (Fig. 1). The mean values of
the left and right psoas muscles as well as those of multidus/
erector spinae muscle groups were used for further statistical ana-
lyses. The measurements were performed by a radiologist (M.E.)
who had no knowledge of the subjects' history, grouping or other
study ndings. The area measurements were repeated by a second
analyst (J.P.) for seven randomly picked subjects to calculate the
correlation coecients in inter-observer comparisons. The inter-
observer correlation coecients for psoas CSA and for multidus
plus erector spinae CSA were 0.91 and 0.93, respectively.
Statistical analysis
An a priori selection of athletes versus non-athletes was made to
evaluate the eects of physical training on the variables measured.
The establishment of an athlete group was justied by searching for
dierences amongst gymnasts, gure skaters and ballet dancers
using the Student-Newman-Keuls test. Although gymnasts showed
higher absolute and relative trunk exion forces and ballet dancers
a higher back musculature endurance than the other two groups,
we persisted with the a priori selection of the groups of athletes and
sedentary controls.
A student's t-test was then introduced to examine the dierences
between the athletes and the non-athletes. In addition, analysis of
covariance (ANCOVA), body mass as a covariate, was used to
examine the dierences between the athletes and the non-athletes in
low back muscle CSA and in static strength endurance tests. Ac-
cording to the normality of the data, either Pearson or Spearman
correlation coecients of muscle CSA with strength parameters
and training time were calculated. Partial correlation coecients
were used to analyse the eects of athletic training. The appropriate
procedures in the SAS/STAT (SAS Institute Inc., Cary, N.C., USA)
program system were used for statistical computations.
Results
Anthropometry
There were no dierences between the groups in height,
but the body mass of the non-athletes was greater than
that of the athletes (P < 0.01).
Muscle CSA
The athletes showed a greater absolute psoas muscle
CSA (P-CSA) than their sedentary controls (P < 0.01).
In contrast, such a dierence was not found in the
summed absolute area of erector spinae plus multidus
muscles (M+ES-CSA). When the CSA was adjusted by
body mass, the athletes showed both a greater P-CSA
(F
2
30:0, P < 0.001) and a greater M+ES-CSA
(F
2
6:4, P < 0.05) than the non-athletes (Fig. 2). The
muscle CSA correlation coecients with force parame-
ters, anthropometric data, age and training time are
given in Table 2. We found a statistically signicant
correlation of P-CSA with trunk exion and extension
force, relative force (newtons per kilogram body mass)
of trunk exion, weekly training time, age, height and
body mass. The CSA of multidus plus erector spinae
muscles correlated signicantly with trunk exion and
extension force, trunk exion endurance, training time
and body height. A partial correlation analysis, adjusted
for participation in athletic training, showed a signi-
cant correlation between M+ES-CSA and trunk ex-
tension and exion forces, body mass and weekly
training time.
Fig. 1 Psoas (P-CSA) and mul-
tidus plus erector spinae (M +
ES-CSA) muscle cross-sectional
areas were analysed from mag-
netic resonance images ob-
tained at the L4/L5 level
parallel to the disc space
68
Strength tests
The athletes showed a signicantly greater maximal
trunk exion strength compared to the control group
(Fig. 3). The dierence was signicant both in trunk
exion and extension when the force was expressed rel-
ative to body mass (newtons per kilogram body mass;
Fig. 4). When the force was expressed relative to the
muscle CSA (newtons per centimetre squared), the ath-
letes did not show any statistically signicant dierence
when compared to the non-athletes (Fig. 5). Further-
more, the correlation coecient of force per CSA
(newtons per centimetre squared) with training time
(minutes per week) was not statistically signicant.
The athletes showed better muscle endurance in
trunk exion than the non-athletes. When the results
were adjusted by body mass (ANCOVA), the athletes
maintained their superiority in trunk exion but no
dierence was found in the trunk extension endurance
test (Fig. 6).
Discussion
Our results indicated that intensive physical training
amongst adolescent girls promotes hypertrophy of the
trunk musculature, strength and muscle endurance.
Both the psoas muscle and trunk extensor (erector
spinae plus multidus) muscle CSA were greater
amongst the athletes than amongst the sedentary age-
matched controls when normalised for body mass. The
purpose of the analysis adjusted to body mass (ANC-
OVA) was to clarify the eect of physical training on
trunk musculature CSA as the body mass per se may act
as a stimulus for the postural muscles of upper body,
leading to a muscle hypertrophy. The correlation anal-
ysis indicated a positive correlation between psoas CSA
and training time. Thus, it would seem that regular
physical training promotes both trunk extensor muscle T
a
b
l
e
2
C
o
r
r
e
l
a
t
i
o
n
c
o
e

c
i
e
n
t
s
(
C
C
)
a
n
d
p
a
r
t
i
a
l
c
o
r
r
e
l
a
t
i
o
n
c
o
e

c
i
e
n
t
s
(
P
C
C
)
(
a
d
j
u
s
t
e
d
f
o
r
p
a
r
t
i
c
i
p
a
t
i
o
n
i
n
a
t
h
l
e
t
i
c
t
r
a
i
n
i
n
g
)
b
e
t
w
e
e
n
p
s
o
a
s
(
P
-
C
S
A
)
m
u
l
t
i

d
u
s
p
l
u
s
e
r
e
c
t
o
r
s
p
i
n
a
e
m
u
s
c
l
e
(
M
+
E
S
-
C
S
A
)
c
r
o
s
s
-
s
e
c
t
i
o
n
a
l
a
r
e
a
s
a
n
d
o
t
h
e
r
m
e
a
s
u
r
e
d
d
a
t
a
T
r
u
n
k
e
x
t
e
n
s
i
o
n
(
N
)
T
r
u
n
k

e
x
i
o
n
(
N
)
T
r
u
n
k
e
x
t
e
n
s
i
o
n
(
N

k
g
)
1
b
o
d
y
m
a
s
s
)
T
r
u
n
k

e
x
i
o
n
(
N

k
g
)
1
b
o
d
y
m
a
s
s
)
E
x
t
e
n
s
i
o
n
e
n
d
u
r
a
n
c
e
(
s
)
F
l
e
x
i
o
n
e
n
d
u
r
a
n
c
e
(
s
)
A
g
e
(
y
e
a
r
s
)
H
e
i
g
h
t
(
c
m
)
B
o
d
y
m
a
s
s
(
k
g
)
T
r
a
i
n
i
n
g
t
i
m
e
(
m
i
n

w
e
e
k
)
1
)
T
r
a
i
n
i
n
g
(
y
e
a
r
s
)
P
-
C
S
A
(
c
m
2
)
C
C
r

0
:
4
6
*
*
*
r

0
:
4
6
*
*
*
r

0
:
1
6
r

0
:
2
7
*
r

0
:
1
0
r

0
:
2
7
*
r

0
:
2
8
*
r

0
:
4
1
*
*
r

0
:
3
9
*
*
r

0
:
4
0
*
*
r

0
:
2
6
P
C
C
r

0
:
4
0
*
*
r

0
:
3
7
*
*
r

0
:
0
8
r

0
:
0
8
r

0
:
1
8
r

0
:
1
7
r

0
:
2
4
r

0
:
1
3
r

0
:
5
6
*
*
*
r

0
:
1
4
r

0
:
2
6
M
+
E
S
-
C
S
A
(
c
m
2
)
C
C
r

0
:
4
6
*
*
*
r

0
:
3
7
*
*
r

0
:
2
0
r

0
:
2
1
r

0
:
2
5
r

0
:
2
7
*
r

0
:
2
5
r

0
:
3
1
*
r

0
:
2
4
r

0
:
3
2
*
r

0
:
1
4
P
C
C
r

0
:
4
3
*
*
r

0
:
3
2
*
r

0
:
0
9
r

0
:
1
1
r

0
:
2
2
r

0
:
2
2
r

0
:
1
5
r

0
:
2
4
r

0
:
4
1
*
*
r

0
:
3
6
*
*
r

0
:
1
4
*
P
<
0
.
0
5
,
*
*
P
<
0
.
0
1
,
*
*
*
P
<
0
.
0
0
1
Fig. 2 Multidus plus erector spinae and psoas muscle cross-sectional
areas at L4/L5 level adjusted for body mass (mean and SD) within
non-athletes (NA) and athletes (A). ***P < 0.001, *P < 0.05
69
and especially psoas muscle hypertrophy in adolescent
girls.
It should be emphasized that the muscles measured
are not solely responsible for trunk extension and exion
but several other muscles are also involved. This is the
case especially with psoas muscle, the ability of which to
produce trunk exion or extension has been demon-
strated to be limited (McGill et al. 1988). It has been
shown that the function of the psoas muscle is to sta-
bilize the spine and to provide thigh exion (Nachemson
1968; Andersson et al. 1995). On the other hand, back
extensor and psoas muscles are involved in the normal
function of the lower lumbar spine. They might also
have some role in the prevention of LBP and their role in
the rehabilitation (functional restoration) of chronic
back pain has been well documented (Parkkola et al.
1992; Rissanen et al. 1995).
A particularly good support for the vertebrae is
needed in gymnastics, gure skating and ballet, as trunk
exion and extension both in the sagittal and frontal
plane are common movements in these sports. There-
fore, it is not surprising that muscles respond with hy-
pertrophy to the increased physical activity. A partial
correlation analysis (adjusted for body mass) between
psoas CSA and training time did not achieve statistical
signicance. A strong positive correlation existed be-
tween multidus plus erector spinae CSA and training
time. In addition, the partial correlation coecient ad-
justed for training time was statistically signicant
showing a linear hypertrophy with increasing training
time.
The results of our study shows that regular physical
training may increase low back muscle CSA and muscle
strength in adolescent girls. Factors such as coordina-
tion, neuromuscular activation and changes in intrinsic
contractile characteristics of the muscle were beyond the
scope of the present study and their eects on force
production were not evaluated. However, we cannot
totally ignore the eect of selection bias on the results as
the muscle CSA were not measured before our subjects
Fig. 3 Maximal static trunk extension and exion forces (mean and
SD) measured in a standing position by a dynamometer. NA Non-
athletes, A athletes. **P < 0.01
Fig. 4 Maximal static trunk extension and exion forces relative
to body mass (mean and SD). NA Non-athletes, A athletes.
***P < 0.001
Fig. 5 Maximal static trunk extension and exion forces relative to
muscle cross-sectional area (mean and SD). NA Non-athletes, A
athletes
Fig. 6 Trunk musculature strength endurance measured by a static
holding test and adjusted for body mass (mean and SD). NA Non-
athletes, A athletes. *P < 0.05
70
started to participate in sports or dance. However, se-
lection bias as an explanation for our ndings is unlikely
because maximal strength is not a decisive factor when
choosing the sport to be followed.
The signicant correlations between absolute force
values and muscle CSA were in accordance with pre-
vious studies (see MacDougall et al. 1977; Ha kkinen
et al. 1981), showing an increased ability to produce
force with increasing CSA. When the force values were
expressed relative to the body mass, a signicant corre-
lation was found only in trunk exion and psoas CSA.
No statistically signicant dierence was found between
the groups as the force values were related to muscle
CSA. This result supports the idea that maximal force of
the trunk muscles is dependent on their CSA.
In conclusion, when low back muscle CSA was ad-
justed for body mass, the athletes showed both greater
psoas and erector spinae plus multidus muscle CSA
than their age-matched sedentary controls. The athletes'
ability in trunk extension and exion was superior to
that of the controls when force was related to body
mass. No dierence was found in force per muscle CSA
between the groups. Concurrently, a positive correlation
existed between muscle CSA and absolute force values.
This further conrms the existence of a close connection
between CSA and the maximal force also in low back
muscles. The present study indicated that regular phys-
ical training in adolescent girls favours the growth of
trunk musculature and development of both maximal
force and muscle endurance. It remains to be seen
whether these functionally favourable changes in ado-
lescence protect these girls from low back disorders in
adulthood. Our intention is to use the same subjects in
the future to analyse the occurrence of LBP and other
low back disorders in adulthood.
References
Andersson E, Oddsson L, Grundstro m H, Thorstensson A (1995)
The role of the psoas and iliacus muscles stability and move-
ment of the lumbar spine, pelvis and hip. Scand J Med Sci
Sports 5:1016
Biering-So rensen F (1984) Physical measurements as risk indicators
for low back trouble over one-year period. Spine 9:106119
Blimkie CJR (1993) Resistance training during preadolescence.
Issues and controversies. Sports Med 15:389407
Fukunaga T, Funato K, Ikegawa S (1992) The eects of resistance
training on muscle area and strength in prepubescent age. Ann
Physiol Anthropol 11:357364
Ha kkinen K, Komi P, Tesch P (1981) Eect of combined con-
centric and eccentric strength training and detraining on force-
time, muscle bre, and metabolic characteristics of leg extensor
muscles. Scand J Sports Sci 3:5058
Hides JA, Richardson CA, Jull GA (1996) Multidus muscle re-
covery is not automatic after resolution of acute, rst-episode
low back pain. Spine 21:27632769
Kujala UM, Salminen JJ, Taimela S, Oksanen A, Jaakkola
L (1992) Subject characteristics and low back pain in
young athletes and nonathletes. Med Sci Sports Exerc
24:627632
Kujala UM, Taimela S, Erkintalo M, Salminen JJ, Kaprio J
(1996a) Low back pain in adolescent athletes. Med Sci Sports
Exerc 28:165170
Kujala UM, Taimela S, Viljanen T, Jutila H, Viitasalo JT, Vide-
man T, Battie MC (1996b) Physical loading and performance as
predictors of back pain in healthy adults. Eur J Appl Physiol
73:452458
MacDougall J, Elder G, Sale D, Moroz J, Sutton J (1977) Skeletal
muscle hypertrophy and atrophy with respect to bre type in
humans. Can J Appl Sport Sci 2:229
Mattila M, Hurme M, Alaranta H, Palja rvi L, Kalimo H, Falck B,
Lehto M, Einola S, Ja rvinen M (1986) The multidus muscle in
patients with lumbar disc herniation. A histochemical and
morphometric analysis of intraoperative biopsies. Spine 11:732
738
Mayer TG, Gatchel RJ, Kishino N, Keeley J, Capra P, Mayer H,
Barnett J, Mooney V (1985a) Objective assessment of spine
function following industrial injury. Spine 10:482493
Mayer TG, Smith SS, Keeley J, Mooney V (1985b) Quantication
of lumbar function. 2. Sagittal plane strength in low back pain
patients. Spine 10:765772
McGill SM, Patt N, Norman RW (1988) Measurement of the trunk
musculature of active males using CT scan radiography: im-
plications for force and moment generating capacity about the
L4/L5 joint. J Biomech 21:329341
Nachemson A (1968) The possible importance of the psoas muscle
for stabilization of the lumbar spine. Acta Orthop Scand 39:47
57
Nachemson A, Lindh M (1969) Measurement of abdominal and
back extension strength with and without low back pain. Scand
J Rehabil Med 1:6065
Parkkola R, Kujala UM, Ryto koski U (1992) Response of the
muscles to training assessed by MRI and muscle strength. Eur J
Appl Physiol 65:383387
Ramsay JA, Blimkie CJ, Smith K, Garner S, MacDougall JD, Sale
DG (1990) Strength training eects in prepubescent boys. Med
Sci Sports Exerc 22:605614
Rissanen A, Kalimo H, Alaranta H (1995) Eect of intensive
training on the isokinetic strength and structure of lumbar
muscles in patients with chronic low back pain. Spine 20:
333340
Salminen JJ, Maki PT, Oksanen A, Pentti J (1992) Spinal mobility
and trunk muscle strength in 15-year-old school children with
and without low-back pain. Spine 17:405411
Thorstensson A, Arvidsson A (1982) Trunk muscle strength and
low back pain. Scand J Rehabil Med 14:6975
Viljanen T, Viitasalo JT, Kujala UM (1991) Strength characteris-
tics of a healthy urban adult population. Eur J Appl Physiol
63:4347
Zhu XZ, Parniapour M, Nordin M, Kahanovitz N (1989) Histo-
chemistry and morphology of erector spinae muscle in lumbar
disc herniation. Spine 14:391397
71