Spinal
and shoulder
complex
posture.
II:
thoracic
alignment
and shoulder
complex position
in normal
and
osteoporotic
women
Elsie Culham and Malcolm Peat School of Rehabilitation Therapy, Queen’s University, Kingston, Ontario
Address for correspondence: Elsie Culham, School of Rehabilitation Therapy, Louise D Acton Building, Queen’s University, Kingston, Ontario K7L 3N6, Canada.
Thoracic
spine
and shouldercomplex
posture were measured in 57 women overthe age of 50. On the basis of thoracic spine measures, the women were divided into normal posture (n = 27), thoracic kyphosis (n = 18) and thoracolumbar
kyphosis (n = 12) posture
groups. Analysis of variance was used to determine if
shoulder complex postural measures differed in the three groups. In the
sagittal
plane, both abnormal posture groups hadsignificantly
greater forward angulationof the scapula and humeral extension compared with the normal posture group. In the transverse plane, scapular protraction and the angle between the scapular
spine
and clavicle weresignificantly
greater in the thoracickyphosis
groupcompared with the other two groups. In the coronal
plane,
the abduction angle ofthe humerus was significantly less in the thoracic kyphosis group compared with the normal posture group. No difference was found in scapular position in this
plane. Although
differences were apparent between the posture groups, manyof the clinical
hypotheses regarding
the effect of spinal posture on shoulder complex position were not substantiated by the results of this study.Introduction
It is
generally recognized
that abnormalsagittal
plane alignment
of the cervical and thoracicspine
can contribute topain
anddysfunction
in the neck and shoulder
region.l-6
Specifi-cally,
several authors havesuggested
that anexaggerated
thoracickyphosis
alters theresting
position
of the shoulderedcomplex resulting
ina forward- or round-shoulder
posture.2,6,7
Thescapula
is described asfollowing
the contour of the thorax and is said to shiftanterolaterally
orto assume a more
protracted position
insubjects
with increased thoracic
kyphosis.
Inaddition,
thescapula
isthought
to rotate such that theglenoid
cavity
has an increased downward inclination.6-8 Kessler andHertling7
stated that thefreely
hang-ing
humerus assumes a more abductedposition
relative to the
scapula
and that tension is lost in theinter-nally
rotatedposition
of the arms is said to occurdue to the
protracted
position
of thescapula.9
The consequences of the alteredposition
arepostulated
to includelengthening
orshortening
of
muscle,
muscle weakness andchanges
in the direction of musclepull
as well as limitation ofupper
extremity
range of motion.2,3 Rehabilita-tion efforts are often aimed at restoration of the’normal’
position
and functionby
stretching
struc-tures assumed to be contracted and
strengthen-ing
musculature assumed weak due toprolonged
stretch.l°However,
despite
the various clinical observationsregarding
the detrimental effects of abnormalspinal
posture
on shouldercomplex
position,
no studies could be found in which theclaims were substantiated
by objective
measure-ment.
The purpose of this part of the
study
was to determineif,
and to what extent, shouldercom-plex position
was affectedby alignment
of thethoracic
spine
in thesagittal plane.
An olderfemale
population
was chosen forstudy
aspre-vious research has demonstrated that
postural
changes
are age related and aregreater
in femalesthan males.11,12 In
addition,
the incidence ofspi-nal
osteoporosis,
which is known to affectsagittal
plane spinal alignment,
ishigher
in women. 13~--15Methodology
A detailed
description
of thetechnique
used tomeasure
spinal
and shouldercomplex
posture isfound in part I*
(Figure 1).
Since the linearmeas-ures of
scapular position
werecompared
acrosssubjects
in thispart
of thestudy
it wasneces-sary to normalize these measures to account for differences in
subject
size. Both measures(Yabs
and
Zabs)
were normalizedby
dividing by
thelength
of the clavicle(three-dimensional
distancebetween the landmarks on the medial and lateral ends of the
clavicle).
Theresulting
measures werelabelled Ynor and Znor
respectively.
*
Clinical Rehabilitation 1993; 7: 309-18.
Figure 1 Angular measures of shoulder complex position.
(A) Sagittal plane. ScSg, forward angulation of the medial border of the scapula; HSg, flexion angle of the humerus
relative to the vertical.
(B) Coronal plane. ScCo, abduction angle of the scapula; CICo, elevation of the clavicle from medial to lateral; HCo. abduction angle of the humerus.
(C) Transverse plane. ScTr, protraction angle of the scapula; CITR, retraction angle of the clavicle; HTr, internal rotation angle of the humerus.
Subjects
Thoracic
posture
and shouldercomplex position
were measured in a total of 57 women between
the ages of 50 and 85 years.
Thirty-six
of thesub-jects
were recruited from thecommunity.
Two ofthese
subjects
werebeing
treated forosteoporosis.
Twenty-one
women with adiagnosis
of osteopor-osis were recruited from anoutpatient
metabolicclinic of a local
hospital.
Fifteen of thesesubjects
had
radiographic
evidence of vertebral fracture at at least onespinal
level. Exclusion criteriaincluded presence of a cardiac
pacemaker,
his-tory
ofneurological
disorder,
history
of surgeryinvolving
thespine
and shouldercomplex
and asignificant
scoliosis of thespine. Subjects
had to bemedically
stable and be able to assume and maintain astanding
position
for a minimum of 10minutes without discomfort. The
procedure
wasexplained
to thesubjects
and all gave informed consentprior
to theirparticipation
in thestudy.
Postural
subgroups
Subjects
were classified into one of threepostural
groups on the basis of the values obtained for thoracickyphosis
and uppertho-racic
slope
(Figure 2).
The groupsresulting
from this classification were:1)
Thoracickyphosis:
18subjects
had a thoracickyphosis
value ofgreater
than 42degrees,
higher
than any of the values found in normalyoung women in a
pilot study.
Theseindivid-uals were
placed
in the thoracickyphosis
pos-ture group.2)
Thoracolumbarkyphosis:
this group includedsubjects
in whom the location of thekyphotic
curvature waslow,
at the thoracolumbarjunc-tion or in the lumbar
region.
Criteria forinclusion in this group included an upper
thoracic
slope
ofgreater
than orequal
to 20degrees
and akyphosis
value of less than35
degrees.
Thoracolumbarkyphosis
wasevi-dent in the
osteoporotic
subjects
whose frac-turesprimarily
involved the lower thoracic orlumbar
vertebrae,
but was also seen insev-eral of the women recruited from the
com-munity.
Twelvesubjects
met the criteria for thoracolumbarkyphosis
and were included inthis group.
3)
Normal posture: thoracicposture
wasconsid-ered to be normal in the
remaining
27subjects
who were
subsequently placed
in the normalposture group.
The two abnormal posture groups are similar to the ’hollow round back’ and ’lower acute
kyphosis’
postures described
by
Itoil6 for women withspinal
osteoporosis.
Descriptive
statistics for age,height
andweight
of thesubjects
in the threepostural
groups are found in Table 1.
Figure 2 Classification of spinal posture into three groups
on the basis of values obtained for thoracic kyphosis and
upper thoracic slope
Statistical
analysis
’ ..The mean values of age,
height
and mass aswell as means of the
spinal
and shouldercom-plex
measures werecompared
across the threepostural
groupsusing
a one-wayanalysis
ofvari-ance
(ANOVA).
In the case of asignificant
F-ratioa post hoc Scheffe test was
completed
to determinewhere the differences were. A level
of p
< 0.05was chosen as the maximum level for statistical
sig-nificance for all ANOVA
analyses.
Systat
(Systat
Inc., Evanston,
IL)
statistical software was usedfor all
analyses.
Results
Posture
subgroups
Although
the mean age was greater and the mean mass less in the thoracickyphosis
group thanin the other two
postural
groups the differenceswere not
statistically significant.
Similarly,
therewere no
significant
differences in meanheight
among the three
postural subgroups
(p
>0.05).
The
spinal
posture
measures and results of sta-tisticalanalyses
arepresented
in Table 2. The meankyphosis angle
was muchgreater
in thethoracic
kyphosis
group than in the other twopostural
groups(F
=38.46;
p <
0.001).
Upper
thoracicslope
wasgreatest
in the thoracickypho-sis group followed
by
the thoracolumbarkyphosis
group. The mean wassignificantly
different for allcomparisons.
The lower thoracicslope
wasgreat-est in the thoracic
kyphosis
group and lowest in the thoracolumbarkyphosis
group. Thediffer-ence between groups was
significant (F
=16.20;
p <
0.001).
Post hocanalyses
revealed that the difference wassignificant
for allcomparisons.
Shoulder
complex
position
Sagittal plane
Results of
analyses
ofsagittal plane
measures arepresented
in Table 3. The medial border of thescapula
wasangled
forward 12.24degrees
onaver-age in the normal
posture
group. Thisangle
wassignificantly
greater, 18.93 and 17.88degrees
in the thoracolumbar and thoracickyphosis
groups,respectively.
The relativeangle
between the upper thoracicspine
and the medial border of thescapula
(ScSgR)
wassignificantly
increased in thetho-racic
kyphosis
groupcompared
with the other twopostural
groups.The humerus was flexed a mean of 1.85
degrees
from the vertical in the normal
posture
group. Thiscompared
with a meanangle
of 5degrees
exten-sion in both the abnormalposture
groups. Therewas no difference in the humeral
angle
relative to the medial border of thescapula
among the three groups(p
>0.05).
Transverse
plane
Results of
analyses
of transverseplane
Table 2 Thoracic spine measures in postural groups
All values in degrees.
SD = standard deviation; UTS =
slope of the upper thoracic spine measured from T1 downward; LTS = slope of the lower
thoracic spine measured from T12 upward; K = thoracic
measures are
presented
in Table 4. The meanretraction
angle
of the clavicle(CITr)
was 23.87degrees
in the normalposture
group. This wasnot
significantly
different from a mean of 21.57and 24.01
degrees
in the thoracolumbar andtho-racic
kyphosis
grouprespectively.
Theprotraction
angle
of thescapula (ScTr)
and the relativeangle
between the
spine
of thescapula
and clavicle(CITrR)
weresignificantly
greater
in the groupwith thoracic
kyphosis compared
with the other groups.The humeral
angle
in the transverseplane
wasthe most variable of all the shoulder
complex
measures as indicated
by
the standard deviations.This
angle
was less in the thoracickyphosis
group,compared
to the other two groups, but thedif-ference was not
statistically significant.
Humeral rotation relative to thespine
of thescapula
waslower in the thoracic
kyphosis
group andhigher
in the thoracolumbarkyphosis
groupcompared
with the normalposture
group. The difference betweenthe two abnormal posture groups was
significant.
Coronal plane
Results of
analyses
of coronalplane
measuresare
presented
in Table 5. Theangle
of the medialborder of the
scapula
to the horizontal was 92.14degrees
in the normal posture group. Thisangle
was lower in the two abnormal posture groups but the difference was not
significant
(p
<0.05).
The
angle
of elevation of the clavicle frommedial to lateral was 7.13
degrees
in thenor-mal posture group. This
angle
was increased inboth abnormal
posture
groups with the greatestdifference
occurring
between the normal and thoracolumbarkyphosis
groups.The humeral abduction
angle
was less in both abnormalpostural
groupscompared
with thenor-mal group. Post hoc
analysis
revealed that thedif-ference was
significant
only
between the thoracickyphosis
and normal group. The abductionangle
of the humerus relative to thescapula
(HCoR)
waslowest for the thoracic
kyphosis
group.However,
the difference among groups was not
statistically
significant.
Linear measures
Results of
analysis
of the linear measures arepresented
in Table 6. The normalized lineardis-tance from Tl to the
scapular
centrealong
the y
(Ynor)
and z(Znor)
axes were notsignificantly
different in thepostural
groups.Table 3 Sagittal plane shoulder complex measures in postural groups
All values in degrees.
SD = standard deviation; ScSg =
angle of forward tilt of the medial border of the scapula to the vertical; ScSgR = relative angle between the medial border of the scapula and the upper thoracic spine derived by subtracting ScSg from UTS;
HSg = angle of the long axis of the humerus to the vertical, a positive value indicating flexion; HSgR = angle of the humerus
Table 4 Transverse plane shoulder complex measures in postural groups
All values in degrees.
SD = standard deviation; ScTr = the protraction angle of the scapular spine relative to a coronal axis defined by a line connecting the roots of the right and left scapular spines; CITr = the retraction angle of the clavicle relative to the coronal
axis as defined above; CITrR = the angle between the clavicle and the spine of the scapula in the transverse plane;
HTr = the angle of internal rotation of the humerus; HTrR = internal rotation angle of the humerus relative to the spine of the
scapula.
Table 5 Coronal plane shoulder complex measures in postural groups
All values in degrees.
SD = standard deviation; ScCo = scapular abduction angle measured as the lateral angle formed between the medial border
of the scapula and the horizontal; CICo = the medial to lateral angle of elevation of the clavicle; HCo = the abduction angle
of the humerus relative to the vertical; HCoR = the abduction angle of the humerus relative to the medial border of the
Table 6 Linear shoulder complex measures in postural groups
All values in cm.
SD = standard deviation; YNor = the linear distance in centimetres from T1 to the centre of the three scapular landmarks
along the y axis, normalized by dividing by the length of the clavicle; ZNor = linear distance from T1 to the centre of the
three scapular landmarks, normalized by dividing by the length of the clavicle. Discussion
It has been
hypothesized
thatchanges
insagittal
plane
posture,specifically
thoracickyphosis
and forward-headposture,
result inchanges
in theresting position
of the shouldercomplex.
Theresults of this
study
indicate that an increase inanteroposterior
curvature of thespine, regardless
of the level at which it occurs, causes an increase
in upper thoracic
slope
which could contribute to a forward headposition.
However,
it is alsoevident from the results that the
position
of theskeletal
components
of the shouldercomplex
wasdependent
on the location of the curve.It is not
surprising
thatchanges
insagittal plane
spinal
posture
affected thesagittal plane
shouldercomplex
position
measures,although
no clinicalreports
regarding
the effect ofpostural changes
onthe shoulder
complex position
in thisplane
havebeen found. Forward
angulation
of the medial border of thescapula
wasgreater
in both of theabnormal posture groups
compared
with thenor-mal
posture
group.However,
the relativeangle
between the medial border of thescapula
andthe upper thoracic
spine
was increasedonly
inthe
subjects
with an increase in thoracickypho-sis
angle.
Thesefindings
can beexplained
by
thelocation of the
kyphotic
curve and theresulting
effect on theshape
of the thorax. Insubjects
with a thoracolumbar curve thescapula angled
forwardto
approximately
the samedegree
as theupper thorax and the relative
angle
between the upper thoracicspine
and medial border of thescapula
(ScSgR)
was notsignificantly
altered.When the curvature is in the thoracic
region
the ribs become
prominent dorsally
and theanteroposterior
diameter of the thorax appears to be increased. Theprominent
dorsal ribs mayprevent the
scapula
fromtilting
forward to thesame
degree
as the upper thoracicspine
result-ing
in the marked increase in the relativeangle
between the upper thoracicspine
andscapular
medial border
(ScSgR)
in these individuals. Thiscould
potentially
result in agradual elongation
of structureshaving
attachment to both the cervicalspine
and thescapula.
Forexample,
this abnormalposture
may lead topain
and irritation at the siteof insertion of the levator
scapula
as describedby
Cailliet.s The fibres of the upper
trapezius
andrhomboid muscles and the
suprascapular
nervemight
besimilarly
affected.The humerus was in more extension in both
the abnormal
posture
groupscompared
with thenormal group, but there was no
significant
dif-ference in the
angle
of the humerus relative to thescapula
(HSgR)
in thisplane.
This indicatesthat the humerus moved with the
scapula
or thehumerus extended as the forward
angulation
of thescapula
increased. Both abnormalpostures
resulted in an increase in upper thoracicslope.
Extension of the arms may be a compensatory
movement to
help
offset the anteriordisplacement
of the upper thorax and maintain balance.
The
protraction angle
of thescapula (ScTr)
inthe transverse
plane
wassignificantly
greater
inthoracic
spine.
Thisfinding
can also beexplained
by
the increasedprominence
of the ribsdorsally
and the increased
anteroposterior
diameter of the thorax evident insubjects
with a mid-thoraciccurves The
angulation
of thescapula
and/or the clavicle must increase to accommodate thegreater
anteroposterior
thoracic diameter. The results arein
agreement
with the clinical literaturedescrib-ing
increasedscapular protraction
with thoracickyphosis.7,9,17
The relativeangle
between thescapular spine
and clavicle(CITrR)
was alsogreatest
insubjects
with thoracickyphosis again
suggesting
an accommodation to an increasedanteroposterior
diameter of the thorax.The results of this
study
suggest
thatprotracted
scapulae
or ’round shoulderposture’
does notnecessarily
accompany a forward headposition
assuggested
in the literature. 7.9,17Subjects
inboth the thoracolumbar and thoracic
kyphosis
posture
groups had an increase in upper thoracicslope
and aclinically
apparent
forward headpos-ition
compared
with the normalposture
subjects.
However,
subjects
with a thoracolumbarkyphosis
did not have
protracted scapulae.
Theposition
of thescapula
in the transverseplane
appears to be affected moreby
theanteroposterior depth
ofthe thorax than
by sagittal plane
posture of thespine.
It would appear,therefore,
to be incorrect toequate
aparticular
shouldercomplex position
with a forward head posture.
The relative
angle
between the humerus andscapula
in the transverseplane (HTrR)
was lessin the thoracic
kyphosis
groupcompared
with theother two
posture
groups,indicating
external rota-tion of the humerus relative to thescapula.
Thisfinding
does notsupport
theconcept
ofincreas-ing
internal rotation of the humerus askyphosis
increases. The relative external rotation of the
humerus may be
compensatory
toprotraction
of thescapula
in order to maintain an upperextrem-ity position
more conducive to function.The
angle
of thescapula
in the coronalplane
wasless in the two abnormal posture groups
compared
with the normal posture group,
indicating
aten-dency
towardsscapular
adduction.However,
the differences in the meanangles
were small andnot
statistically
significant.
Thus,
there is little evidence from thisstudy
to indicate that scapu-lar adduction(downward
rotation of theglenoid)
accompanies
an increase in thoracic curvatureas
proposed by previous
authors.7,9 Downwardscapular
rotation as a cause ofelongation
ofleva-tor
scapula
and irritation at its site of insertionproposed by
Cailliet5 also appears doubtful basedon these
findings.
The mean elevation
angle
of the clavicle was greater in both the abnormalposture
groups.Subjects
in both of these groups had an increasein upper thoracic
slope.
As thisslope
increases the sternum and medial end of the clavicle wouldtend to become
depressed possibly resulting
in theobserved increase in the
angle
of elevation from medial to lateral in the abnormalposture
groups.The humerus was less abducted in the thoracic
kyphosis
groupcompared
with the other two groups. The abductionangle
of the humerus to thescapula
(HCoR)
was also less in this groupthough
the difference was notstatisti-cally
significant.
Thus,
there was no evidencefrom this
study
tosupport
thehypothesis
that the humerus abducts relative to thedownwardly
rotatedscapula
insubjects
withkyphotic
postureas
proposed
by
Kessler andHertling.7
The trendwould appear to be the reverse with adduction of the humerus relative to the
scapula occurring
insubjects
withincreasing
thoracickyphosis.
The linear distance from Tl to the
scapular
centre was notsignificantly
different in the threepostural
groups. There is no evidence from thisstudy,
therefore,
tosupport
thetheory
that thescapula
moveslaterally
on the chest wall insubjects
withkyphosis leading
toelongation
of the rhomboid and middle and lowertra-pezius
muscies.4,10 The ’stretch-weakness’ of themiddle and lower
trapezius
muscle in persons with’kyphosis
and forward shoulders’ asproposed by
Kendall and
McCreary’O
and, thus,
the need forstrengthening
of thescapular
retractors inper-sons with
kyphosis
should bequestioned
inlight
of these
findings.
Although
mean values of Znor indicated thatthe
scapula
wasdepressed
on the thorax in the thoracickyphosis subjects
and elevated in thethoracolumbar group
compared
with the normalsthis difference was not
statistically
significant.
Conclusions
The
resting position
of the shouldercomplex
wasaltered in the abnormal
posture
groupsthoracic posture. The shoulder
complex
posi-tion was
dependent
on where the curve waslocated;
thoracic versus thoracolumbarregion.
The effect of posture on the
resting position
of the shouldercomplex
may be related more to theresulting changes
in theshape
of the thoracic cage rather than to thesagittal plane
curvature alone.References
1 Nicholas JA, Wilson PD. Osteoporosis of the aged
spine. Clin Orthop 1963; 26: 19-33.
2 Bowling RW, Rockar PA, Erhard R. Examination of the shoulder complex. Phys Ther 1986; 66: 1866-77. 3 Braun BL, Amundson LR. Quantitative assessment
of head and shoulder posture. Arch Phys Med Rehabil
1989; 70: 322-29.
4 Darnell MW. A proposed chronology of events for forward head posture. J Craniomandib Pract 1983; 1: 50-53.
5 Cailliet R. Soft tissue pain and disability. Philadelphia: F.A. Davis Company, 1988.
6 Cailliet R. Neck and arm pain. Philadelphia: F.A.
Davis Company, 1981.
7 Kessler RM, Hertling D. Management of common
musculoskeletal disorders: physical therapy principles and methods. Philadelphia: Harper and Row, 1983:
274-310.
8 Kendall HO, Kendall FP, Boynton DA. Posture and pain. Baltimore: Williams and Wilkins, 1952. 9 Cailliet R. Shoulder pain. Philadelphia: F.A. Davis
Company, 1966.
10 Kendall FP, McCreary EK. Muscles: testing and
function. Baltimore: Williams and Wilkins, 1983.
11 Milne JS, Lauder IJ. Age effects in kyphosis and lordosis in adults. Ann Hum Biol 1974;1:327-37.
12 Fon GT, Pitt MJ, Thies AC. Thoracic kyphosis:
range in normal subjects. Am J Roentgenol 1980; 134: 979-83.
13 Albanese AA, Edelson AH, Lorenze EJ, Woodhull
ML, Wein EH. Problems of bone health in elderly:
ten year study. N Y State J Med 1975; 75: 326-36. 14 Riggs BL, Wahner HW, Dunn WL, Mazess RB, Offord KP, Melton LJ, III. Differential changes in bone mineral density of the appendicular and axial skeleton with aging: relationship to spinal osteoporosis. J Clin Invest 1981; 67: 328-35.
15 Aloia JF, Vaswani A, Ellis K, Yuen K, Cohn SH. A model for involutional bone loss. J Lab Clin Med 1985;
106: 630-37.
16 Itoi E. Roentgenographic analysis of posture in spinal
osteoporotics. Spine 1991; 16: 750-56.
17 Ayub E. Posture and the upper quarter. In: Donatelli R ed. Clinics in physical therapy: physical therapy
of the shoulder. New York: Churchill Livingstone, 1987 : 69-78.