Evaluation of the lag sign tests for external rotator function
of the shoulder
Christof Hurschler, PhD,a Nikolaus Wülker, MD,b Henning Windhagen, MD,a Niels Hellmers,a and
Piet Plumhoff,a Hannover and Tübingen, Germany
Rotator cuff lesions pose a serious clinical challenge.
The objective of this study was to verify the biome-
chanical basis for the lag sign clinical tests for rotator
cuff dysfunction. The lag sign tests were simulated in
vitro by a sensor-guided robotic simulator configured
to reproduce in vivo testing conditions. The ability of
the test to isolate supraspinatus and/or infraspinatus
dysfunction was investigated from 20° to 90° of scap-
ular plane arm elevation. The test was 100% sensitive
for lack of all infraspinatus–teres minor force at all ele-
vations tested (6/6 specimens at 20°, 30°, and 60°
elevation; 5/5 at 90° elevation). The test was less sen-
sitive to complete loss of supraspinatus force (1/6
specimens testing positive at 20°, 0/6 at 30°, 3/6 at
60°, and 3/5 at 90° elevation). The results of this bio-
mechanical study suggest the lag sign tests to be
highly sensitive only for infraspinatus–teres minor mus-
cle dysfunction. (J Shoulder Elbow Surg 2004;13:
298 –304.)
C linical tests for rotator cuff function are designed to
allow the examiner quickly and easily to detect pa-
thologies of the cuff muscles of the shoulder. Espe-
cially important are tests used to diagnose partial and
complete tears of the tendons of the external rotators,
in particular of the supraspinatus (SSP) and infraspi-
natus (ISP) muscles. These muscles are usually tested
by assessing the patient’s ability to elevate the arm in
internal rotation,10 by evaluating the external rota-
tional power of the patient,6,12 or by assessing the
ability of the patient to hold an externally rotated
position.1 The latter tests have been referred to as
From the Department of Orthopaedic Surgery, Hannover Medical
School, Hannover,a and Orthopaedic Clinics and Polyclinics,
University of Tübingen, Tübingen.b
This study was funded by the German Research Foundation (DFG).
Reprint requests: Christof Hurschler, PhD, Medizinische Hochschule
Hannover, Orthopädische Klinik, Annastift, Anne-von-Borries-
Strasse 1-7, D-30625 Hannover, Germany (E-mail: hurschler@
annastift.de).
Copyright © 2004 by Journal of Shoulder and Elbow Surgery
Board of Trustees.
1058-2746/2004/$30.00
doi:10.1016/j.jse.2004.01.021
298
drop or lag sign tests (Figure 1,Table I); when the
patient is unable to hold an externally rotated posi-
tion, the resulting internal rotation, referred to as lag,
indicates external rotator dysfunction. Tests per-
formed at lower elevations have been proposed for
the diagnosis of patients with shoulders too painful to
test at higher elevations.7
To be effective, tests of external rotator function are
designed to minimize the confounding synergistic or
antagonistic effects of muscles other than those of
interest (such as the deltoid and pectoralis major).
This attribute is confirmed by clinical, electromyo-
graphic (EMG), and biomechanical studies. EMG
studies show that the SSP and ISP muscles are highly
activated in positions similar to those of the diagnostic
tests.9,12,13 Furthermore, biomechanical studies show
that, in addition to the ISP, an external rotator, the SSP
muscle, also possesses an external rotation moment
arm in positions of external rotation.14 Thus, in posi-
tions of external rotation, both the SSP and ISP can
contribute to the moment of external rotation.8,13
However, clinical investigations based on computed
tomography arthrographic evaluation of muscle deg-
radation and/or intraoperative verification of tendon
ruptures remain inconclusive. The external rotation
lag sign (ERLS) was thus found to be sensitive to
complete ruptures of the SSP independent of the con-
dition of the ISP in one study,7 whereas a test of
external rotational power in a similar position (0°
abduction, 45° external rotation) was found to be
sensitive to stage 3 to stage 4 ISP fatty degeneration
in the presence of SSP and ISP tendon tears (length
ranging from 2.5 to 5.0 cm).22 Fatty degeneration of
the ISP has been shown to occur in the presence of
rotator cuff tears.5 Furthermore, although positions of
higher elevation are advocated for testing of the ISP,7
Walch et al22 found this position to be specific for the
ISP only in combination with significant teres minor
(TM) fatty degeneration.
It is thus not clear whether the lag sign tests for the
rotator cuff can distinguish between ISP and SSP
ruptures and at what angles they should optimally be
tested. The goals of this study were, first, to measure
the sensitivity of the lag sign tests and, second, to
determine an optimal testing position for diagnosing
as well as differentiating defects of the SSP and ISP
J Shoulder Elbow Surg Hurschler et al 299
Volume 13, Number 3

Figure 1 A, Clinical examiner holding the test position for performing the ERLS test at 20° elevation, as described
in the text as well as in Table I. B, Photograph taken after release of the patient’s hand. Lag is defined as the amount
of internal rotation that occurs relative to the test position. A lag of greater than 5° is considered a positive test.

Table I Definitions of tests for external rotator function

Name Position Test criterion

Dropping sign 0° Abduction Forearm drops back to

(Walch et al,22 45° External rotation 0° external rotation
1998) 90° Elbow flexion
ERLS (Hertel et al,7 20° Elevation Angular drop or lag
1996) Max-5° External rotation recorded to nearest 5°
90° Elbow flexion
Drop sign (Hertel 90° Elevation Angular drop or lag
et al,7 1996) Max-5° External rotation recorded to nearest 5°
0° Elbow flexion

tendons of the rotator cuff. The clinical application of

the lag sign test was simulated in vitro with a sensor-
guided robot-assisted shoulder simulator. The method
used is unique because it allows simultaneous control
of both the muscle forces (representing the patient
attempting to hold the test position) and the motion of
the specimen as it would be during the performance
of the clinical test (representing the examiner holding
the elbow). The resulting internal rotation (ie, lag) that
occurred at different levels of ISP and SSP muscle
force over a range of elevation angles was precisely
measured in positions and under loading conditions
that closely mimicked those occurring during clinical
evaluation.
MATERIALS AND METHODS
Specimen preparation and mounting
Seven radiographically normal human shoulders (mean
patient weight, 84 ⫾ 15 kg; mean age, 60 ⫾ 21 years,
with three patients of unknown age) were harvested and
frozen before testing. The specimens were thawed at room
temperature for 12 hours before dissection of all muscles
except for the rotator cuff and deltoid. The muscle bellies of
Figure 2 Photograph of shoulder specimen mounted in the simu-
lator, demonstrating the wires sutured to the SSP and ISP-TM
muscles, the humerus holder, the scapula mounting tower, the
manipulator end of the robot arm, and the FMS.
the rotator cuff were bluntly dissected away from the scap-
ula, and particular care was taken to preserve the integrity
of the joint capsule and ligaments as well as the rotator cuff
tendons and their insertions. Ductile wire was sutured to the
musculotendinous junctions of the SSP, combined ISP and
TM (ISP-TM), and subscapularis (SSC) muscles to allow the
application of muscle forces. Specimens were mounted in
the testing machine (Figure 2) by use of 3 threaded steel
rods through the scapula according to a standardized
procedure as follows24: the scapula was first aligned with a
leveling gauge so that the tangent to the medial and inferior
angles was vertically oriented. The plane of the scapula
was then defined visually as the vertical plane passing
through both the medial and inferior angles of the scapula,
as well as the middle of the glenoid surface. Finally, the
scapula was tilted forward 10° to approximate its physio-
logic orientation on the thorax. The humerus was transected
approximately 20 cm distal to the center of the humeral
300 Hurschler et al
Figure 3 Schema of experimental setup illustrating the transfer of
the equivalent gravitational force (FEG) and moment (MEG) to the
geometric center of the humeral head, as well as the robot and FMS
to which the distal end of the humerus is attached.
head and potted in a brass cylinder (outer diameter, 38
mm; height, 5 cm) by use of cold-curing methyl methacrylate
resin (Technovit 4004; Heraeus Kulber, GmbH, Wehrheim,
Germany). The axis of the humerus was defined by aligning
the humerus and the cylinder with a specially designed jig
inserted into the marrow cavity. The humerus was attached
to the robot by inserting the brass cylinder into a mount
attached to the force-moment sensor of the robot (Figure 2).
Neutral rotation of the humerus was defined by palpation-
aligning the intertubercular groove midway between the
coracoid process and the acromion.
Simulation of clinical tests for external rotator function
The external rotator tests investigated in this study are
applied clinically as follows7,22: the patient flexes the el-
bow to 90°, the arm is elevated in the scapular plane, and
the examining clinician passively rotates the patient’s arm
to the maximum attainable externally rotated position while
holding the elbow (Figure 1, Table I). The actual position
tested is attained by derotating 5° to avoid excessive elastic
recoil of the joint capsule (caused by tension in the anterior
capsule). The test is defined as positive if the patient cannot
hold this position and the arm rotates more than 5° inter-
nally upon release of the hand (Figure 1, A and B). This test
has been described at different arm elevations: when per-
formed at 0° elevation, it is termed the dropping sign22; at
20° elevation, it is termed the ERLS7; and at 90° elevation,
it is termed the drop sign7 (Table I). In this study the relative
contribution of the SSP and ISP muscles to both of these tests
was simulated in vitro at 20°, 30°, 60°, and 90° elevation
in the scapular plane. The scapulothoracic-to-glenohumeral
angle relationship was assumed to vary according to Pop-
pen and Walker16 so that the corresponding glenohumeral
angles were assumed to be approximately 16°, 22°, 33°,
and 50°, respectively.
The robot-assisted shoulder simulator used in this study
(Figures 2 and 3) consisted of a previously described in
vitro dynamic shoulder model24,25 combined with a sensor-
guided industrial robot (KR/15; Kuka, GmbH, Augsburg,
Germany). A 6-component force-moment sensor (KMS-60;
J Shoulder Elbow Surg
May/June 2004
Figure 4 Definition of all coordinate systems: tool coordinates of
the robot (T), measurement coordinates of the FMS (S), global
coordinate system fixed in space (G), and humeral coordinate
system at the geometric center of the humerus (H).
IpeA, GmbH, Berlin, Germany) integrated into the control
algorithm of the robot allowed it to function in real time
under force-moment control. Muscle force was applied by
computer-controlled hydraulic cylinders (accuracy of ⬍1
N).25 The force-moment sensor (FMS) measured the result-
ant force and moment acting at the geometric center of the
humerus (resolution ⬍3.0 N and ⬍0.15 Nm force and
torsion, respectively; linearity ⬍1%). The robot (repeatabil-
ity of ⫾0.1 mm) and specially written control software were
used to simulate the weight of the complete arm during the
simulation (Figure 3).
The robot was programmed to permit motion only in the
same degrees of freedom as would the clinical examiner
supporting the arm at the elbow and releasing the hand
(internal-external rotation and medial-lateral and anterior-
posterior translation). The robot is, thus, not used in the
traditional sense (ie, to perform some predefined pro-
grammed motion) but rather as a device to apply loads and
control motions as they would occur during the clinical
examination. The resultant equivalent gravitational force
(FEG) and moment (M) vectors acting at the center of the
humeral head as a result of the weight of the arm were
computed and superimposed on the load measured by the
FMS (Figure 3). Anthropometric data for the upper arm,
forearm, and hand were used for these computations.21
The robot was programmed to respond to the forces and
moments acting on the specimen in a coordinate system that
moved with the specimen (Figure 4). A continuously running
subroutine (Figure 5), with a turnaround time of 144 milli-
seconds or less between reading of FMS data and execu-
tion of the motion commands, feedback-controlled the robot
(termed sensor-guided mode). This sensor-guided mode is
fast enough so that the resulting motion is perceived as a
smooth motion. The FMS and humerus mount self-weight
were compensated so that they did not affect the measure-
ment of force on the specimen.
The humeral coordinate system, located at the geometric
center of the humeral head, was defined such that it was
congruent with the global coordinate system when the
specimen was centered in the glenoid and hanging under
its own weight in the previously defined position of neutral
rotation (Figure 4). The X-axis of the humerus thus corre-
sponded to the anatomic flexion-extension axis, the Y-axis
corresponded to the axis of internal-external rotation, and
the Z-axis corresponded to the axis of abduction-adduction.
J Shoulder Elbow Surg
Volume 13, Number 3
Figure 5 Flowchart of the algorithm used to control the robot. The
resultant equivalent gravitational force (F) and moment (M) acting
on the specimen are measured by the FMS. The data are corrected
for the self-weight of the sensor and humerus-mounting bracket. The
force and moment vectors resulting from the weight of the arm are
then computed (as a function of humerus position) and added to the
measured values, which are subsequently translated into the ap-
propriate motion commands.
The term elevation used in this study refers to abduction in
the scapular plane.
The robot was used to simulate the boundary conditions
during clinical examination (1) by blocking rotation about
the humeral coordinate system X- and Z-axes (H in Figure 4),
preventing a change in flexion and elevation angles, re-
spectively, and (2) by blocking translation along the global
Y-axis (G in Figure 4), preventing superior-inferior transla-
tion. All other degrees of freedom—that is, anterior-poste-
rior, superior-inferior, and medial-lateral translation—re-
mained unconstrained. Muscle forces applied to the
tendons of the rotator cuff were simulated by 1.0-mm diam-
eter stainless steel cables routed over pulleys to computer-
regulated force-controlled hydraulic cylinders.25 The cables
were attached with wire suture and oriented to approximate
the anatomic lines of action of the muscles as closely as
possible: for the SSP through the SSP fossa; for the ISP-TM
along the ISP fossa, bisecting the angle between the spine
and the lateral margin of the scapula; and for the SSC
along the SSC fossa, also bisecting the angle between the
superior and lateral margins.
Experimental protocol
Physiologic cross-sectional areas of the individual mus-
cles were used as a basis for determining muscle forces
used in this simulation.4,15,18,20,21 Antagonistic SSC ten-
sion was first determined, and the ISP-TM force and SSP
force necessary to hold the test position were subsequently
determined as follows: minimum tension of the rotator cuff
necessary to prevent the arm from subluxing under its own
weight was found by increasing the force on the SSP,
ISP-TM, and SSC proportionally (16.7% for SSP, 39.9% for
ISP-TM, and 43.3% for SSC)21 until the arm was held
stabilized in the glenoid. The lag sign test position at each
elevation angle was attained by elevating the arm in the
scapular plane and maximally passively externally rotating
the arm while centering the humerus in the glenoid with a
20-N force. Maximum external rotation was defined as the
rotation at which a torque of 1.8 Nm was measured about
Hurschler et al 301
the rotational axis of the humerus. This torque was chosen
based on preliminary experiments; it was high enough to
produce repeatable results while not damaging the speci-
men. To avoid excessive elastic capsular recoil, the test
position was finally obtained by derotating the arm 5°,
corresponding to the procedure recommended for the clin-
ical test.7
As a basis for investigating the relative contributions of
the SSP and ISP-TM muscles, normal SSP force and ISP-TM
force were determined as follows: the basic SSC antago-
nistic force was set at the level defined above for each
individual specimen, and SSP force and ISP-TM force were
increased proportionately according to their respective
physiologic cross-sectional areas (29.5% for SSP and
70.5% for ISP-TM)21 until the test position was maintained
upon release of the arm. This was done at each elevation
angle tested. The muscle forces thus obtained (SSP, ISP-TM,
and SSC) were considered representative of those in a
normal patient actively holding the test position and were
thus used to define 100% SSP and ISP-TM force.
The relative contributions of the SSP and ISP-TM force in
holding the test position were then investigated by chang-
ing the forces in the ISP-TM and SSP muscles and releasing
the arm from the test position. If the arm internally rotated
more than 5°, the test was considered positive; if not, it was
negative. As the ISP and TM are considered the primary
external rotators, force levels were chosen to range in
increments of 25% from 175% to 0% of normal for the SSP
and from 100% to 0% of normal for the ISP-TM. At each
elevation angle investigated, a total of 15 combinations of
SSP and ISP-TM force were tested in a randomized order
(Table II). Two-tailed paired Student t tests were used to
compare means at a significance level of ␣ ⫽ 0.05.
RESULTS
The results of the lag sign simulations are tabulated
according to the number of positive tests versus the
number of specimens tested at each elevation angle
and combination of SSP and ISP-TM force tested
(Table II). One of the seven specimens tested was
removed from the series because it exhibited unusual
laxity and an extreme external rotation of 104° at 90°
elevation, and thus almost no SSP and ISP-TM force
was required to hold the test position. The SSP tendon
insertion of another specimen failed at 90° elevation,
reducing the number of specimens tested to 5 at that
elevation.
In general, the results showed that the force of both
the SSP and ISP-TM muscles can affect the outcome of
the lag sign tests. Decreasing the amount of ISP-TM
muscle force from 100% to 75% while holding SSP
force constant at 100% resulted in a clear rise in the
number of positive tests: further reducing ISP-TM force
to 50% or less resulted in all specimens testing posi-
tive (Table II). Decreasing SSP force while holding
ISP-TM force at 100% also resulted in an increase in
the number of positive tests; however, the effect was
less dramatic than with the ISP-TM and was almost
nonexistent at 20° and 30° elevation (Table II). Thus,
302 Hurschler et al J Shoulder Elbow Surg
May/June 2004

Table II Number of positive lag Sign tests at indicated elevation angle and combination of SSP and ISP-TM force

SSP force

ISP-TM force 0% 25% 50% 75% 100% 125% 150% 175%

20° Elevation (N ⫽ 6)
0% — — — — 6 — — —
25% — — — — 6 — — —
50% — — — — 6 6 6 6
75% — — — 6 5 3 1 —
100% 1 0 0 0 0 — — —
30° Elevation (N ⫽ 6)
0% — — — — 6 — — —
25% — — — — 6 — — —
50% — — — — 6 5 5 4
75% — — — 4 3 3 2 —
100% 0 0 0 0 0 — — —
60° Elevation (N ⫽ 6)
0% — — — — 6 — — —
25% — — — — 6 — — —
50% — — — — 6 6 6 5
75% — — — 3 3 2 2 —
100% 3 1 1 0 0 — — —
90° Elevation (N ⫽ 5*)
0% — — — — 5 — — —
25% — — — — 5 — — —
50% — — — — 5 5 5 5
75% — — — 5 5 5 4 —
100% 3 2 2 1 0 — — —

The underlined zero indicates the level of force defined as the normal SSP and ISP-TM force sufficient to hold the test position (ie, 100% force). Cells
with dashes indicate force combinations that were not tested.
*One specimen was lost at 90° elevation because the SSP tendon tore at its insertion.

more than half of the tests were negative at all eleva-
tions tested despite the fact that all SSP force had
been removed. This indicates that in some specimens,
100% ISP-TM force alone is sufficient to hold the test
position, whereas the inverse is not true—that is,
100% SSP force alone is never sufficient to hold the
test position in the absence of ISP-TM force.
The total amount of SSP and ISP-TM force required
to hold the test position increased with elevation
angle from 114.5 ⫾ 25 N at 20° elevation to 171.4
⫾ 54 N at 90° elevation (P ⫽ .051) (Figure 6). This
increase occurred despite the fact that the antagonis-
tic SSC force used for a particular specimen (mean,
36.3 ⫾ 9 N) was held constant for all elevations
tested. The maximum external rotation attained also
increased with elevation angle from 68.3 ⫾ 18° at
20° elevation to 81.3 ⫾ 22° at 90° elevation (P ⫽
.004) (Figure 7).
Although the test is more sensitive to ISP-TM than
SSP force, sensitivity to SSP force is increased when
ISP-TM force is reduced, particularly at lower eleva-
tions. Thus, at 75% ISP-TM force, an increase of SSP
force to 150% results in fewer positive tests, particu-
larly at 20° elevation, where four fewer positive tests
were observed (Table II). This effect is smaller at 30°,
60°, and 90° elevation, with one fewer positive test at
each elevation angle, respectively. Furthermore, at
50% ISP-TM force, the compensating effect of the SSP
is overwhelmed: increasing SSP force to 175% re-
sulted in no increase at 20°, two additional positive
tests at 30°, only one additional positive test at 60°,
and no increase at 90°.
DISCUSSION
The results of this study suggest that it is difficult to
distinguish between deficiencies of the SSP and
ISP-TM muscles by use of the lag sign tests based on
biomechanical considerations alone. Whereas com-
plete loss of ISP-TM force would result in 100% sen-
sitivity at all elevations tested (Table I, 100% SSP
force), complete loss of SSP force would be more
difficult to detect, especially at lower elevations. Al-
though increasing with elevation, sensitivity for the
isolated SSP rupture was still only 50% (3/6) at 60°
and 60% (3/5) at 90° elevation (Table II, 100%
ISP-TM force). The increased sensitivity to SSP force at
higher elevation is a result of the increase in total SSP
and ISP-TM force required to hold the test position
with increasing elevation angle. The increasing force
required to hold the test position at increasing eleva-
tion is itself due in part to the increasing internal
rotational torque due to the weight of the arm, as well
J Shoulder Elbow Surg
Volume 13, Number 3
Figure 6 Total external rotator force of the SSP and ISP-TM muscles
required to hold the test position at each elevation angle tested.
Figure 7 Maximum external rotation attained at each elevation
angle tested. The test position is defined as the maximum externally
rotated position minus 5° of internal rotation.
as the reducing rotational moment arm of the ISP,
which occurs with increasing elevation.14
The sensor-guided robot-assisted simulator used in
this study differs from previous joint-kinematic investi-
gations using industrial robots,3,23 primarily because
it was designed to evaluate both active and passive,
not just passive, structures of the joint. On the one
hand, the system’s ability to move passively and
precisely measure the movement of the arm allows
noninvasive definition of a humeral coordinate system
at the geometric center of the humeral head. On the
other hand, the ability to measure load and control
motion relative to a humeral coordinate system per-
mits simulation of the effect of the weight of the arm at
any position and orientation of the specimen. Finally,
and most importantly, the ability to allow or constrain
motion freely in translation or rotation in any direction
of the global (fixed) or humeral (moving) coordinated
systems allows realistic simulation of the boundary
conditions imposed on the joint during the perfor-
mance of the clinical lag sign test.
Although the advances and advantages of this
Hurschler et al 303
system are multiple, limitations characteristic of all in
vitro simulations still apply. Muscle forces were simu-
lated by use of cables attached to the muscle-tendon
junctions of the combined ISP and TM, SSP, and SSC
muscles. Although the cables were oriented along the
physiologic axis of the muscle, the simulation of mus-
cles that may have more than one functional unit as a
single force is an inherent limitation of this and similar
studies.2,11,17,19 Although it has a small external
rotational moment arm,14 the deltoid muscle was not
simulated; its external rotational power is minimal
because of the externally rotated position in which the
test is performed. Furthermore, any external rotational
power of the deltoid would reduce even further the
ability of the lag sign test to detect isolated SSP
ruptures. Finally, the ISP and TM muscles were simu-
lated as one functional unit. Although this assumption
is supported by EMG studies in normal patients,9
which indicate similar parallel activation patterns for
both muscles, observations of TM muscle hypertrophy
suggest that the TM may compensate for the reduced
external rotating power resulting from ISP tendon
tears.22 A strengthening of the TM would not contra-
dict the conclusions of this study, because it would
only serve to mask further the contribution of the SSP
muscle to external rotation strength.
It was impractical to simulate tendon ruptures di-
rectly by actually cutting the tendon, as we wanted to
investigate different combinations of SSP and ISP-TM
deficiency and the number of specimens available for
this study was limited. Dysfunction of tendon-muscle
units was thus simulated by reducing force compared
with the 100% force, defined as the force required to
hold the test position and based on physiologic cross-
sectional areas. Significant superior glenohumeral
translation may occur in some patients with advanced
rotator cuff disease. The changes in muscle moment
arms that may result from such translations were not
taken into consideration.
Although we tested the ISP and TM as one func-
tional unit (ISP-TM), the results of this study agree with
those of Walch et al,22 who found a test performed at
0° elevation and 45° external rotation to be positive
only in the presence of significant ISP degeneration in
patients with combined tears of the SSP and ISP
tendons. The authors suggested that, although exhib-
iting a tear in the ISP tendon, patients without signif-
icant fatty degeneration retain enough ISP function to
pass the test.22 This hypothesis is supported by our
study, which indicates that, at lower elevation, less
force is required to hold the test position and that
partial defects may be compensated by the other
muscles.
The clinical study of Hertel et al7 found the lag sign
test at 20° elevation (ERLS) to be sensitive to ruptures
of the SSP tendon, despite the presence of an intact
ISP tendon. The results of our study suggest that it
304 Hurschler et al
would be difficult to detect an isolated SSP defect in
the presence of a fully functional ISP-TM unit. As
muscle force cannot be directly measured in vivo, the
forces used in this study were based on physiologic
cross-sectional area and the assumption that muscles
with external rotating function would be maximally
activated to hold the test position. It is nonetheless
possible that we overestimated ISP muscle force; ISP
muscle shortening could result in a lower ISP force
relative to the SSP at the near-maximally externally
rotated position of the arm used in these lag sign
tests.7 Hypothetically assuming that ISP force is in fact
lower than simulated, the SSP/ISP force relationship
would be represented better by the forces tested when
the ISP-TM was set to 75% and the SSP was set to
150% of what we defined as normal (Table II). This
force combination results in an SSP/ISP-TM force ratio
of 0.84, rather than a ratio of 0.42. This force rela-
tionship does, in fact, result in more positive tests and
a higher sensitivity for the SSP at 20° elevation.
However, it should be noted that tears intraopera-
tively interpreted to involve only the SSP tendon may
also carry over into the ISP tendon. The results of this
study indicate that the lag sign test is highly sensitive
to reduced ISP-TM force, independent of SSP muscle
force; at ISP-TM forces less than 50%, the test has a
sensitivity of 100%. Reduction of ISP force as a result
of an overlooked partial tear of the ISP tendon could
lead to a positive lag sign test.
Although we have made every effort to represent
the physiologic situation as closely as possible, some
factors, such as the pain a patient experiences during
the test, cannot be simulated in an in vitro study such
as ours. The results of our study should, therefore, be
interpreted only in the context of existing clinical
investigations. As a result of this study, we expect the
lag sign tests to be highly sensitive only for the ISP-TM
muscles, especially at higher elevations, where the
SSP cannot compensate. Thus, if the test were posi-
tive, especially at higher elevation, it would predict a
deficiency of the ISP-TM but have little or no power to
diagnose isolated SSP defects in the presence of a
fully functional ISP-TM unit.
REFERENCES
1. Bigliani LU, Cordasco FA, McIIveen SJ, Musso ES. Operative
repair of massive rotator cuff tears: long-term results. J Shoulder
Elbow Surg 1992;1:120-30.
2. Debski RE, McMahon PJ, Thompson WO, et al. A new dynamic
testing apparatus to study glenohumeral joint motion. J Biomech
1995;28:869-74.
3. Debski RE, Wong EK, Sakane M, Warner JJ, Fu FH. In-situ forces
in the glenohumeral ligaments during anterior loading [abstract].
Trans Orthop Res Soc 1998;23:1031.
J Shoulder Elbow Surg
May/June 2004
4. Fick R. Handbuch der Anatomie und Mechanik der Gelenke unter
Berücksichtigung der bewegenden Muskeln. Jena: Fischer;
1910.
5. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC. Fatty
muscle degeneration in cuff ruptures. Pre- and postoperative
evaluation by CT scan. Clin Orthop 1994;304:78-83.
6. Gschwend N, Ivosevic-Radovanovic D, Patte D. Rotator cuff
tear—relationship between clinical and anatomopathological
findings. Arch Orthop Trauma Surg 1988;107:7-15.
7. Hertel R, Ballmer FT, Lombert SM, Gerber C. Lag signs in the
diagnosis of rotator cuff rupture. J Shoulder Elbow Surg 1996;5:
307-13.
8. Hughes RE, An KN. Force analysis of rotator cuff muscles. Clin
Orthop 1996;330:75-83.
9. Jenp YN, Malanga GA, Growney ES, An KN. Activation of the
rotator cuff in generating isometric shoulder rotation torque. Am J
Sports Med 1996;24:477-85.
10. Jobe FW, Moynes DR. Delineation of diagnostic criteria and a
rehabilitation program for rotator cuff injuries. Am J Sports Med
1982;10:336-9.
11. Karduna AR, Williams GR, Williams JL, Iannotti JP. Kinematics of
the glenohumeral joint: influences of muscle forces, ligamentous
constraints, and articular geometry. J Orthop Res 1996;14:986-
93.
12. Kelly BT, Kadrmas WR, Speer KP. The manual muscle examina-
tion for rotator cuff strength. An electromyographic investigation.
Am J Sports Med 1996;24:581-8.
13. Kronberg M, Nemeth G, Brostrom LA. Muscle activity and coor-
dination in the normal shoulder. An electromyographic study. Clin
Orthop 1990;257:76-85.
14. Otis JC, Jiang CC, Wickiewicz TL, et al. Changes in the moment
arms of the rotator cuff and deltoid muscles with abduction and
rotation. J Bone Joint Surg Am 1994;76:667-76.
15. Poppen NK, Walker PS. Forces at the glenohumeral joint in
abduction. Clin Orthop 1978;135:165-70.
16. Poppen NK, Walker PS. Normal and abnormal motion of the
shoulder. J Bone Joint Surg Am 1976;58:195-201.
17. Sharkey NA, Marder RA, Hanson PB. The entire rotator cuff
contributes to elevation of the arm. J Orthop Res 1994;12:699-
708.
18. Shiino K. Schultergelenksbewegungen und Schultermuskelarbeit.
Arch f Anat und Physiol Anat Abt 1913;(Suppl):46-87.
19. Soslowsky LJ, Malicky DM, Blasier RB. Active and passive factors
in inferior glenohumeral stabilization: a biomechanical model. J
Shoulder Elbow Surg 1997;6:371-9.
20. Strasser H. Die obere Extremität: Schulter und Oberarm. In:
Strasser H, editor. Spezieller Teil: Die obere Extremität. Berlin:
Springer; 1917. p. 16-29.
21. Veeger HE, van der Helm FC, Van der Woude LH, Pronk GM,
Rozendal RH. Inertia and muscle contraction parameters for
musculoskeletal modelling of the shoulder mechanism. J Biomech
1991;24:615-29.
22. Walch G, Boulahia A, Calderone S, Robinson AH. The ’drop-
ping’ and ’hornblower’s’ signs in evaluation of rotator-cuff tears.
J Bone Joint Surg Br 1998;80:624-8.
23. Woo SL, Debski RE, Wong EK, Yagi M, Tarinelli D. Use of
robotic technology for diathrodial joint research. J Sci Med Sport
1999;2:283-97.
24. Wuelker N, Schmotzer H, Thren K, Korell M. Translation of the
glenohumeral joint with simulated active elevation. Clin Orthop
1994;309:193-200.
25. Wuelker N, Wirth CJ, Plitz W, Roetman B. A dynamic shoulder
model: reliability testing and muscle force study. J Biomech
1995;28:489-99.