Sie sind auf Seite 1von 6

Thoracic Hyperkyphosis in the

Young Athlete: A Review of the

Biomechanical Issues
James A. Ashton-Miller, PhD

Address Definition of Hyperkyphosis

Biomechanics Research Laboratory, Department of Mechanical Scheuermans disease is characterized by 5 or more of verte-
Engineering, 3208 G.G. Brown, 2250 Beal/Hayward, University bral wedging (measured radiographically using the Cobb
of Michigan, Ann Arbor, MI 48109-2125, USA.
E-mail: method) in at least three adjacent vertebrae [6]. The incidence
of Scheuermanns disease has been estimated to be 0.4% in a
Current Sports Medicine Reports 2004, 3:4752
Current Science Inc. ISSN 1537-890x prospective study of 1060 children followed for 3 years [7].
Copyright 2004 by Current Science Inc. How large must a thoracic kyphosis be to qualify as hyperky-
phosis? To determine the answer, we need to first define the
normal range of kyphosis. One set of values, 20 to 50, has
Can intense athletic training cause progression of a tho-
been published by Boseker et al. [8], who noted no age or sex
racic hyperkyphosis in the immature athlete? The lack of
differences in their sample of 121 normal children. Another
prospective, controlled trials addressing this issue pre-
well-regarded study enrolled several hundred children in each
cludes a definitive answer. Hyperkyphosis is defined
age strata and found mean (SD) thoracic kyphosis angles of
radiographically as a thoracic kyphosis exceeding 55 in
36.7 (6.9), 37.5 (8.0), and 38.5 (8.1) in 5- to 9-, 10- to 14-,
10- to 20-year-olds, and 65 using back surface mea-
and 15- to 20-year-olds, respectively [9]. The authors measured
surements. In this paper we review the biomechanical
the kyphosis angle, using the Cobb method, from T2 to T12,
scenarios that could lead to an increase in kyphosis in
when possible, and otherwise from T3 to T12. If one defines
the immature athlete via excessive mechanical loading.
the normal range of kyphosis as the mean plus or minus two
These include fall-related impacts, rapid acceleration/
SDs, then the normal range of kyphosis, from these radio-
deceleration of heavy weights, and maximum effort
graphic data, is from 22 to 51 in the 5- to 9-year-olds, and
trunk extensor muscle lengthening contractions.
from 22 to 55 in the 15- to 20-year-olds. Hence, we can
Research is needed on whether the number and magni-
define hyperkyphosis as being a kyphosis of 52 or more in the
tude of vertebral loading cycles, the recovery interval
youngest age group and a kyphosis of 56 or more in 10- to 20-
between loading cycles, and the frequency of training
year-olds. Unless a lateral radiograph is taken with the arms in
affect apophyseal growth during periods of rapid growth
a forward reach posture, the upper thoracic vertebrae may not
of the apical vertebrae in a hyperkyphotic spine.
be visible on the standard lateral radiograph. Therefore, a mea-
surement of thoracic kyphosis on a standard lateral radiograph
can sometimes underestimate the true angle of kyphosis.
Introduction For screening and research purposes one can estimate
Hyperkyphosis of the nonpostural (nonreversible) type kyphosis without using radiographs by measuring the mid-
involves a fixed deformity, usually as the result of a con- sagittal plane profile of the dorsal surface of the thorax. This
genital abnormality [1], a pathology, or an iatrogenic con- has been done using a variety of methods, including panto-
dition. Thoracic hyperkyphosis in the juvenile or graph, Moir fringe patterns, and rasterstereography. An alter-
adolescent is often associated with Scheuermans disease. native measurement method involves the Debrunner
Although its etiology is idiopathic, both genetic factors [2] kyphometer [10]. These surface shape methods often yield a
and mechanical factors are thought to play a role in the surface kyphosis angle that is slightly greater than that
development of Scheuermanns disease. The diagnosis and obtained radiographically. In a classic study of 40 adolescents,
clinical management of this condition have been reviewed Willner [11] found a mean (SD) surface thoracic kyphosis
elsewhere [3], as has its natural history [4,5]. In this review angle of 39.7 (12.2). Thus, using the plus or minus two SD
we focus on the biomechanical factors that might cause criterion to define normality, a kyphosis angle of 65 or more
progression of a hyperkyphotic spine under intense ath- can be classified as hyperkyphosis when using his panto-
letic training. graphic method of measuring back surface shape.
48 Spine Conditions

Vertebral Morphology cases, regions of the vertebra and apophyses were missing
In the axial plane, the thoracic vertebra has a relatively tri- entirely. This growth retardation led to altered ossification
angular shape, with its apex pointing ventrally in compari- of the vertebra along with wedging and unevenness at the
son with the more elliptically shaped lumbar vertebra. vertebral surface. Finally, studies in immature rat tail have
Thus, the thoracic vertebra has relatively little cross-sec- demonstrated that disc height can also be modulated by
tional area anteriorly with which to resist the compressive mechanical loading history [16]. In summary, these find-
stresses associated with a flexion bending moment. This ings suggest that in both immature animals and humans,
means that the compressive stresses in the narrowed ante- alterations in mechanical loading can affect both vertebral
rior region tend to be higher under a flexion moment than and disc shape and thereby, potentially, spine curvature.
those in the larger posterior region of the centrum. In the
sagittal plane, the thoracic vertebrae are naturally wedge-
shaped, with the amount of wedging being correlated with Athletic Training and Hyperkyphosis
the local kyphotic curvature in that region [12]. The height What is the risk that intense athletic training can cause pro-
of each vertebra is determined by heredity, whereas the gression of a thoracic hyperkyphosis in the immature ath-
diameters of the centrum are influenced by the level of lete? The lack of a prospective controlled trial in these
physical activity. patients precludes a definitive answer. Such a trial would
The centrum of juvenile and adolescent thoracic verte- require documentation of the age and sex of the partici-
brae is characterized by the apophyseal rings, cartilaginous pant, sport, level of play, as well as repeated measures of
secondary growth centers that form the anterolateral mar- kyphosis angle, training methods, training intensity, and
gins of its superior and inferior endplate surfaces. These duration. The study would require the cooperation of the
structures are most highly stressed when axial compression athletes, their coaches, and parents. Methods would have
loads are combined with a bending moment. Given the to be found to circumvent the possible bias engendered in
narrow anterior aspect of the thoracic vertebra, the most self-reported training methods, intensity, and duration.
ventral region of the apophyseal ring will experience the Strenuous physical activity is known to cause structural
highest compressive stress under a flexion bending abnormalities in the immature vertebral body. Concern
moment. that exposure to years of intense athletic training may
The length of the thoracic spine grows most rapidly increase the risk for developing adolescent hyperkyphosis
between the ages of 10 and 16 years [13]. It may be during in certain sports, as well as the known association between
this time that the columnar arrangements of cells in the ring hyperkyphosis and adult-onset back pain, led us to exam-
apophyses are most vulnerable to excessive loading and ine the association between cumulated hours of athletic
repetitive loading. Indeed, Scheuermann posited that the training and the magnitude of the sagittal curvature of the
abnormal thoracic wedging could be due to abnormalities in immature spine in a cross-sectional study [18]. We stud-
the growth regions of the immature vertebra based on radio- ied a convenience sample of 2270 children (407 girls)
graphic evidence that he noted. More recent work in animal between 8 and 18 years of age. An optical rasterstereo-
models has confirmed that sustained asymmetric loading graphic method was used to measure the mid-sagittal cur-
can cause wedging in the immature vertebra [14]. vatures of the surface of the back in the upright standing
position in order to quantify the angles of thoracic kypho-
sis and lumbar lordosis. These data were then correlated
The Influence of Asymmetric Loading on with self-reported hours of training measured by interview
Vertebral Growth and questionnaire. The possible effects of age, sex, sport,
The presence of wedged vertebrae in Scheuermanns and upper and lower body weight training were investi-
kyphosis suggests that asymmetric mechanical loading can gated. Hyperkyphosis was defined as 65 or greater on
alter vertebral endplate alignment. Studies of immature rat these back surface measurements. Gymnasts trained the
tail vertebrae have shown similar alterations in vertebral most each year (439 h/y), followed by swimmers (379 h/
growth [15,16]. Similarly, a study in immature rat verte- y), football players (282 h/y), and wrestlers (228 h/y); the
brae shows that a vertebral wedge deformity can be cor- remaining sports trained less than 200 hours per year. The
rected by reversing the load used to create it [14]. This results in these young athletes showed that larger angles of
finding provides a sound rationale for the Boston (or thoracic kyphosis and lumbar lordosis were associated
equivalent) brace to treat and reduce hyperkyphosis in the with greater cumulative training times (h/y; P < 0.001);
immature spine. gymnasts showed the largest curves (mean SD thoracic:
Excessive compressive loading of the spine is known to 42.4 13.4; lumbar: 52.1 16.7). Lack of sports par-
affect normal spinal development at the apophyses [17]. ticipation, on the other hand, was associated with the
For example, in biopsy specimens from seven patients with smallest curves (sedentary controls: thoracic 16.1 10.4;
hyperkyphosis, the endosteal vertebral bone growth was lumbar 17.6 15.6). Age and sex did not significantly
stunted beneath the abnormal apophyseal plates. In severe affect the degree of curvature.
Thoracic Hyperkyphosis in the Young Athlete: A Review of the Biomechanical Issues Ashton-Miller 49

How are Thoracic Vertebrae Loaded? when the extensor muscles develop an insufficient
Any vertebra is normally loaded by two types of forces: ten- moment to counter it. In general, the apical vertebra is also
sile forces in any muscles that directly insert onto the verte- placed under considerable axial compression because the
bra, and the forces and moments applied to the vertebral extensor muscle forces have to equilibrate the large gravita-
centrum by the intervertebral disc above and below it. Of tional flexion moment.
particular concern are flexion moments when they are The thoracic vertebra of interest can be acted on
applied to an already-wedged (kyphotic) vertebra, especially directly by a third type of force applied directly to the
if they are not equilibrated by the thoracic extensor muscles. spinous process via the overlying skin. This can occur when
Extensor muscle fatigue would be one scenario in which a the athlete falls and contacts the ground, a support surface,
flexion moment is not fully equilibrated. Flexion moments or is hit by another player on the dorsal surface of the neck
that are not equilibrated by the extensor muscles would tend or back. This type of loading would tend to translate the
to increase the angle of wedging. Similarly, compressive vertebra anteriorly, and load the inferior vertebrae in flex-
forces on the centrum of a wedged vertebra could also cause ion. Thus, a ventrally directed force applied to the head or
wedging to progress if they slow the growth of the apophy- spine above the vertebra of interest would exert an addi-
seal ring anteriorly. We now consider how these flexion tional flexion moment on the vertebra of interest. The fur-
moments and compression forces arise. ther cranial this force acts, the larger the flexion moment it
We start by considering the quiet, upright standing pos- applies about the vertebra of interest will be. Likewise, a
ture. The forces applied by the disc above the vertebra of caudally directed impact force on the head, neck, or shoul-
interest include axial compression and shear forces. The ders will also induce a flexion moment on the thoracic ver-
latter may be oriented in the anterior or posterior, left or tebra of interest.
right directions, or any combination of these. These forces Finally, inertial forces can also load the apical thoracic
include the effect of gravity acting on the superincumbent vertebrae. For example, flexion loading can occur when the
body mass as well as the muscle forces needed to equili- upright body, having forward momentum, collides with a
brate it. The moments applied by the adjacent disc above surface or object such that the body below the vertebra of
the vertebra include flexion or extension, lateral bending interest is abruptly decelerated. This type of loading will
and/or axial torsion, or any combination of these. Typi- apply a forward shear force to the vertebra via its upper
cally, the line of gravity of superincumbent body weight facet joints and superincumbent disc, along with a flexion
acts forward of a thoracic vertebra, and therefore applies a bending moment that must be equilibrated by the thoracic
flexion moment to it. This must then be equilibrated by extensor muscles. A similar argument holds for the case in
tensile forces developed by the thoracic extensor muscles. which the inverted athlete performs a handspring: now it is
As a result, the vertebra of interest must then resist a com- the downward momentum of the upper body that is
pression force that results from both the gravitational force abruptly decelerated by the ground. In this case the lower
and these muscle forces. The higher the level of torso mus- thorax and body will now load the apical thoracic verte-
cle co-contraction at the vertebral level of interest, the brae in compression and flexion, along with the thoracic
larger the muscles forces involved, and therefore the larger extensor muscle forces. Weight lifting, a ubiquitous form of
the compression force acting on the vertebra. In general, in athletic training, can cause significant thoracic compres-
quiet standing, less than half the total compression force sion and flexion loading when rapidly lifting a heavy
acting on a vertebra is derived from the effect of gravity on hand-held weight. The magnitude of the inertial force is
the superincumbent body above that level, and the other directly proportional to the product of the weight lifted
half comes from the muscle forces required to equilibrate multiplied by its upward acceleration (along with the
that effect. weight and acceleration of the arms themselves). Likewise,
The forces and moments applied by the disc below the weights having downward momentum that are then
vertebra of interest are simply the reaction forces and reac- abruptly decelerated using the upper extremities will exert
tion moments required (from Newtons third law, action = a similar load for the same reason. Finally, a large flexion
reaction) to equilibrate the aforementioned forces and moment and compression loading can be exerted on the
moments, as well as the negligible weight of the vertebra apical vertebrae by the rapid deceleration of the body
itself, and any muscle or rib forces acting directly on the below the vertebra of interest, as when landing from a drop
vertebra itself. jump. In this case, the upper body has downward momen-
In an upright individual, a flexion moment acts on all tum prior to the feet impacting the ground. After impact,
the thoracic vertebrae via the weight of the rib cage, and the upward force on the feet is transmitted up the skeleton
this is greatest at the apex of the thoracic curve because this to decelerate the vertebra of interest via the disc and facet
vertebra lies furthest from the line of gravity of the head joints below it. The straighter the knees and the more flat-
and the thorax. This apical vertebra is then the vertebra footed the landing, the larger the deceleration, and the
most exposed to the greatest flexion moment, especially larger the flexion moment exerted on the apical vertebrae.
50 Spine Conditions

What Determines the Magnitude of an Impact lengthening contraction are used here in preference to the
Force on the Body? less accurate corresponding terms, concentric and eccentric
The biomechanical factors that determine the magnitude contraction [23].
of an impact in fall have recently been determined [19 It is well known that lengthening contractions result in
22]. If the impact force were to act on the body above the significantly higher muscle forces than shortening contrac-
level of the target vertebra, for example the dorsal aspect tions. In fact, at sufficiently high velocities, the maximum
of the neck region, then that force would act to deceler- force developed by a muscle undergoing a lengthening
ate the neck from the impact velocity (just prior to contraction can reach 1.6 times the maximum volitional
impact) to the final velocity of zero after impact. That force developed isometrically [24,25]. Of course, the well-
impact force, in turn, acts to apply a bending moment known hyperbolic relationship between contractile force
on, and a shear force to, the thoracic vertebra of interest; and shortening velocity means that even a small shorten-
hence our concern that it not exceed the ability of the ing velocity causes a large decrease in the maximum force
vertebral tissues to resist it without injury. What deter- than can be developed by the muscle. As the shortening
mines the magnitude of the impact force? Apart from the velocity increases the force that can be developed further
stiffness of the impact surface (the higher the stiffness, decreases to very low values. In summary, the largest mus-
the higher the impact force), the magnitude of the cle forces occur in a maximally activated muscle when that
impact force will then depend upon two variables: the muscle is forcibly lengthened.
impact velocity and the effective mass of the body part It follows then that the largest muscle forces exerted on
impacting the surface. The product of those two variables the vertebra of interest will occur when the adjacent thoracic
is a measure of the momentum of the body part immedi- extensor muscles are maximally contracting and being
ately prior to impact, for instance, with the ground. At lengthened by forced flexion of the torso. This can happen,
impact, the ground exerts a reaction force on the neck, for example, when an athlete who is running forward trips.
slowing its momentum to zero. The magnitude of that In an effort to slow his or her forward and downward
impact force may be found by considering the impulse momentum of the head and upper body and arms, the tho-
(defined as force multiplied by the time that it acts) racic extensor muscles have to contract maximally. As they
required to decrease the neck momentum to zero. For gradually slow the forward-moving upper body, they are
example, if the mass of the neck is 5 kg and it impacts the placed under a lengthening contraction condition. As a
ground with a downward velocity of 3 m/sec, then its result, these muscles develop large tensile forces which in
momentum is 15 kg-m/sec. This momentum can be turn place the flexing thoracic spine under a considerable
completely arrested by an impact force of 1500 N acting compression. Because a single cycle of forceful flexion and
for 10 msec, 150 N acting for 100 msec or 15 N acting for compression loading can cause a mature lumbar disc to her-
1 second. Notice that the shorter the impact time, the niate, there is some concern that the anterior portion of an
higher the impact force must be. The higher the impact immature vertebral apophysis could also be at risk for injury
force, the more concerned we have to be with the poten- under such combined loading.
tial for it to injure the apophysis. Shorter impact times
are typically associated with collisions with a hard or stiff
surface or object. The higher the impact force, the more Repetitive Vertebral Loading
concerned we have to be with the potential for apophy- There is presently only indirect evidence that repetitive
seal injury in the hyperkyphotic region. loading can cause vertebral centrum abnormalities. The
original observations of Scheuermann were that approxi-
mately half of his patients who presented with three or
The Significance of Extensor Muscle more wedged vertebrae were engaged in heavy agricultural
Lengthening Contraction on labor. In a later study, Wassman [26] also noted an associa-
Vertebral Loading tion with heavy physical work. Blazek et al. [27] found in a
It is well known that the force developed by a striated cross-sectional study that Scheuermanns disease was
muscle depends both upon its neural activation level present in 41% of top athletes compared with 10% in non-
(number of motoneurons recruited and their firing fre- sporting controls. One can not differentiate between the
quency) as well as the external load. For a given neural possible effects of repetitive loading, excessive single loads,
activation level, when the external load is less than the or impact loading in these studies; all are possible in an
contractile force developed by the muscle, the muscle will agricultural setting. Working with an immature rat tail
shorten, a so-called shortening contraction. When the exter- model, Revel et al. [28] found that repetitive tail bending
nal load exactly equals the contractile force in the muscle, caused endplate abnormalities, along with disruptions of
it will contract isometrically (at a constant length). And the columnar arrangement of chondrocytes in the deep
when the external load exceeds the contractile force devel- zone of the cartilaginous endplate. These findings are con-
oped by the muscle, then the muscle will undergo a so- sistent with similar findings in Scheuermanns patients
called lengthening contraction. The terms shortening and [29]. Because ossification of the anterior apophysis has
Thoracic Hyperkyphosis in the Young Athlete: A Review of the Biomechanical Issues Ashton-Miller 51

been noted in gymnasts [30], athletes who are routinely 7. Nissinen M: Spinal posture during pubertal growth. Acta Pae-
exposed to repetitive impact loading via gymnastic jump diatr 1995, 84:308312.
8. Boseker EH, Moe JH, Winter RB, Koop SE: Determination of
landings, inadvertent falls, and lengthening contractions of normal thoracic kyphosis: a roentgenographic study of 121
their torso extensor muscles, one cannot exclude the possi- normal children. J Pediatr Orthop 2000, 20:796798.
bility that repetitive loading may play a role in causing ver- This is a classic study quantifying the normal range of kyphosis using
the Cobb method on lateral radiographs of healthy children. Drs.
tebral wedging. However, the experiments are required to Moe and Winter are widely recognized for their major contributions
determine apophyseal injury risk from a few cycles of to the assessment and treatment of spinal deformities
excessive loading versus many cycles of moderate loading. 9. Voutsinas AS, MacEwan GD: Sagittal profiles of the spine. Clin
Orthop Rel Res 1986, 210:235242.
10. Korovessis P, Petsinis G, Papazisi Z, Baikousis A: Prediction of
thoracic kyphosis using the Debrunner kyphometer. J Spinal
Handedness Disord 2001, 14:6772.
Using a threshold for hyperkyphosis of 45, 20 less than 11. Willner S: Spinal pantograph: a noninvasive anthropometric
device for describing postures and asymmetries of the trunk.
is defined here using back surface shape analysis, left hand- J Pediatr Orthop 1983, 3:245249.
edness has been found to be associated with a fourfold This study by a Swedish orthopedic surgeon pioneered the noncon-
higher risk for hyperkyphosis during the pubertal growth tact measurement of kyphosis in large samples of children using a
simple pantograph.
spurt in Finnish children from 10.8 to 13.8 years [31]. Why
12. Goh S, Price RI, Leedman PJ, Singer KP: The relative influence
this should be so remains unclear. It remains to be seen if of vertebral body and intervertebral disc shape on thoracic
the findings can be replicated using a more rigorous defini- kyphosis. Clin Biomech 1999, 14:439448.
tion of hyperkyphosis. 13. DiMeglio A, Bonnel F: La Rachis en Croissance. Paris: Springer-
Verlag; 1990.
14. Mente PL, Aronsson DD, Stokes IAF, Iatrides JC: Mechanical
modulation of growth for the correction of vertebral wedge
Conclusions deformities. J Orthop Res 1999, 17:518524.
15. Mente PL, Stokes IAF, Spence H, Aronsson DD: Progression of
A prospective randomized trial would help determine
vertebral wedging in an asymmetrically loaded rat tail
whether or not intensive athletic training can increase a model. Spine 1997, 22:12901296.
pre-existing hyperkyphosis in a juvenile or adolescent ath- 16. Stokes IAF, Aronsson DD, Spence H, Iatridis JC: Mechanical
lete. From a biomechanical perspective, both impact forces modulation of intervertebral disc thickness in growing rat
tails. J Spinal Disord 1998, 11:261265.
and large thoracic extensor muscle forces, associated with 17. Ippolito E, Bellocci M, Motanaro A, et al.: Juvenile kyphosis: an
lengthening contractions as a result of torso flexion, could ultrastructural study. J Pediatr Orthop 1985, 5:315322.
cause excessive stress on the anterior thoracic apophyses. 18. Wojtys EM, Ashton-Miller JA, Huston L, Moga P: The associa-
tion between athletic training time and the sagittal curvature
However, the effect on the vertebral apophyses of loading
of the immature spine. Am J Sports Med 2000, 28:490498.
type, magnitude, and number of repetitions in a given This study used a noncontact optical method for measuring kyphosis
interval remain to be clarified. Coaches, parents, and sports in several thousand adolescents who participated in regular sports
training programs.
physicians should, therefore, remain vigilant for any
19. DeGoede KM, Ashton-Miller JA: Fall arrest strategy affects
changes in hyperkyphosis as a result of intense athletic peak hand impact force in a forward fall. J Biomech 2002,
training regimens. 35:843848.
20. DeGoede KM. Ashton-Miller JA: Biomechanical simulations of
forward fall arrests: effects of upper extremity arrest strategy,
gender and aging-related declines in muscle strength. J Bio-
References and Recommended Reading mech 2003, 36:413420.
Papers of particular interest, published recently, have been 21. DeGoede KM, Ashton-Miller JA, Schultz AB: Fall-related upper
highlighted as: body injuries in the older adult: a review of the biomechani-
Of importance cal issues. J Biomech 2003, 36:10431053.
Of major importance 22. Lo J, McCabe GN, DeGoede KM, et al.: On reducing hand
impact force in forward falls: results of a brief intervention
1. McMaster MJ, Singh H: Natural history of congenital kyphosis in young males. Clin Biomech 2003, 18:730736.
and kyphoscoliosis. A study of one hundred and twelve 23. Faulkner JA: Terminology for contractions of muscles during
patients. J Bone Jt Surg 1999, 81-A:13671383. shortening, while isometric, and during lengthening. J Appl
2. Graat HC, van Rhijn LW, Scrander-Stumpel CT, van Ooij A: Physiol 2003, 95:455459.
Classical Scheuermann disease in male monozygotic twins: 24. McCully KK, Faulkner JA: Characteristics of lengthening con-
further support for the genetic etiology hypothesis. Spine tractions associated with injury in skeletal muscle fibers. J
2002, 27:485487. Appl Physiol 1986, 61:293299.
3. Tribus CB: Scheuermanns kyphosis in adolescents and 25. Warren GL, Hayes DA, Lowe DA, et al.: Mechanical factors in
adults: diagnosis and management. J Am Acad Orthop Surg the initiation of eccentric contraction-induced injury in rat
1998, 6:3643. soleus muscle. J Physiol (Lond) 1993, 464:457475.
4. Murray PM, Weinstein SL, Spratt KF: The natural history and 26. Wassman K: Kyphosis juvenilis Scheuermann an occupa-
long-term follow-up of Scheuermanns kyphosis. J Bone Jt tional disorder. Acta Orthop Scand 1951, 21:6574.
Surg 1993, 75-A:236248. 27. Blazek O, Streda A, Cermak V, Skallova O: The incidence of
5. Sachs B, Bradford D, Winter RB, et al.: Scheuermanns kypho- morbus Scheuermann in sportsmen. J Sports Med 1986,
sis. J Bone Jt Surg 1987, 69-A:5057. 26:5559.
6. Sorensen KH: Scheuermanns Juvenile Kyphosis. Copenhagen: 28. Revel N, Andre-Deshays C, Roudier R, et al.: Effects of repetitive
Munksgaard; 1964. strains on vertebral end plates in young rats. Clin Orthop
1992, 279:303309.
52 Spine Conditions

29. Scoles PV, Latimer BM, Digiovanni BF, et al.: Vertebral alter- 31. Nissinen M, Heliovaara M, Seitsamo J, Poussa M: Left handed-
ations in Scheuermanns disease. Spine 1991, 16:509515. ness and risk for thoracic hyperkyphosis in prepubertal
30. Tertti M, Paajanen H, Kujala UM, et al.: Disc degeneration in schoolchildren. Int J Epidemiol 1995, 24:117811181.
young gymnasts. Am J Sports Med 1990, 18:206208.