Shoulder Joint Position Sense Improves with Elevation

Angle in a Novel, Unconstrained Task

David N. Suprak, Louis R. Osternig, Paul van Donkelaar, Andrew R. Karduna

Department of Human Physiology, 1240 University of Oregon, Esslinger 122,

University of Oregon, Eugene, Oregon 97403-1240

Received 15 June 2005; accepted 3 October 2005

Published online 6 February 2006 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jor.20095

ABSTRACT: Proprioception, encompassing the submodalities of kinesthesia and joint position

sense, is important in the maintenance of joint stability, especially in the shoulder. The purpose of
this study was to examine the effects of plane and elevation angle on unconstrained shoulder joint
position sense. Twenty-two subjects (12 male, 10 female) without a history of shoulder pathology
were recruited from a university campus. Subjects attempted to replicate, with respect to plane and
elevation angles, various target positions. Target positions consisted of five plane angles at 908 of arm
elevation and five arm elevation angles in the scapular plane. All target positions were tested twice to
assess the reliability of the measurement. Intraclass correlation coefficients were generally low
across target positions, possibly owing to the novelty and demanding nature of the task. No
differences in repositioning errors were observed between plane angles (p ¼ 0.255). Repositioning
errors decreased linearly as the elevation angle increased from 308 to 908 (p ¼ 0.007) and increased
again from 908 to 1108 of elevation (p ¼ 0.029). Our results suggest that unconstrained joint position
sense may be enhanced with increased muscular activation levels. Further, afferent feedback from
musculotendinous mechanoreceptors may dominate over that from capsuloligamentous sources in
unconstrained movements. ß 2006 Orthopaedic Research Society. Published by Wiley Periodicals,
Inc. J Orthop Res 24:559–568, 2006
Keywords: proprioception; joint position sense; muscle activation; shoulder

INTRODUCTION position sense, the ability of a person to identify

the position of a limb in space, and kinesthesia,
Joint stability is important for coordinated per- the ability to detect limb movement.2
formance of functional tasks of daily living The role of proprioceptive mechanisms in the
as well as for more demanding athletic skills. maintenance of joint stability is especially impor-
Stability is afforded via factors such as the tant for the shoulder, as compared to other
degree of bony congruity, integrity of capsuloliga- synovial joints, where stability is sacrificed for a
mentous supporting structures, and feedback large range of motion (ROM).3 This relative
loops involving joint and musculotendinous shoulder instability is due to low bony congruity
mechanoreceptors that are integrated with the and relative inefficiency of mechanical restraints
central nervous system. The effects of these provided by the joint capsule and ligaments in
feedback loops have collectively been termed dynamic conditions.4 It has been postulated that
proprioception, as defined by Sherrington in afferent signals arising from shoulder musculo-
1906.1 Since that time, proprioception has been tendinous and capsuloligamentous mechanore-
compartmentalized into submodalities and has ceptors enable the central nervous system to
recently been described as a combination of joint maintain muscle stiffness and coordination about
the joint and produce smooth movements while
minimizing the chance for joint injury.5
Correspondence to: A. R. Karduna (Telephone: 541-346-0438;
Fax: 541-346-0441; E-mail: karduna@uoregon.edu)
Joint position sense is commonly tested using
ß 2006 Orthopaedic Research Society. Published by Wiley Periodicals,
either active or passive reproduction of joint
Inc. positioning. In these paradigms, the shoulder joint

JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006 559

560 SUPRAK ET AL.

of a blindfolded subject is moved through a range of more intense contractions.18,19 However, the
rotation via a mechanical device to a predeter- extent to which muscle activation level affects
mined target position and held for a period of time joint position sense has not yet been explored.
(usually 5–10 s). Upon returning to the starting Several studies examined upper extremity
position, subjects attempt to replicate the target proprioception in an unconstrained model.20,21,38
position, either by means of active movement or by However, to date, no study has examined shoulder
passive movement induced by the mechanical joint position sense in multiple planes, using a
device during which subjects indicate when they testing paradigm employing active reproduction of
feel the presented position has been matched. The active positioning. Therefore, the purpose of this
difference between the presented and reproduced study was twofold: to examine the effect of plane
position is recorded as the repositioning error. angle on repositioning error at a constant eleva-
Movements in passive positioning conditions tion, and to examine the effect of arm elevation
are most commonly performed at speeds of 0.5 to angle on repositioning error in the scapular
28/s.6–9 However, functional activities are per- plane, in an unconstrained testing paradigm. We
formed predominately in the presence of active hypothesized that as the shoulder joint angle
muscle contraction. Further, the speeds chosen for approached the end range plane angle, position
testing in the passive paradigms are much slower sense would be enhanced, due to stretching of the
than those seen in functional tasks. For these capsuloligamentous structures. As the presented
reasons, active joint position sense testing meth- elevation angle increased toward 908, we further
ods are hypothesized to better indicate joint hypothesized that muscle activation, and there-
function than passive protocols.8 Joint position fore muscle spindle and golgi tendon organ activa-
sense is thought to be provided by the slowly tion, would increase due to the increased torque
adapting musculotendinous (muscle spindles, about the joint in that position. We expected that
golgi tendon organs) and capsuloligamentous this increase in muscle spindle and golgi tendon
(Ruffini, and golgi tendon organlike endings) organ activation would be manifested in decreased
mechanoreceptors.3 repositioning error.
Capsuloligamentous mechanoreceptors are sti-
mulated upon deformation of their parent tis-
sue.10,11 Several authors have hypothesized that MATERIALS AND METHODS
these receptors are stimulated more in the end
ranges of motion, compared to the mid ranges, due Twenty-two healthy individuals (12 males, 10 females)
to the elongation of their parent tissues in these with a mean age of 23.7 years (
4.8), a mean height
ranges.12–14 In the shoulder, this hypothesis has of 172.9 cm (
10.1), and a mean body mass of 72.4 kg
been supported by several studies examining joint (
14.3) agreed to participate in the study. Prior to
participation, all subjects signed an informed consent
position sense in one plane, which have reported
form approved by the university’s Institutional Review
enhanced position sense as the presented angle
Board (IRB). Subjects were included in the study only if
approaches end range.3,15,16 To our knowledge, they had no history of shoulder pathology requiring
however, this effect has not been studied in an surgery or physical therapy. Exclusion criteria included
unconstrained model. limited ROM in arm elevation and previous diagnosis of
Although the capsuloligamentous receptors are shoulder instability or other pathology that might alter
relatively inactive in the midranges of motion, neuromuscular control of the shoulder. However, no
when the tissues are slack, afferent information direct measurements of either shoulder or generalized
regarding joint position is still relayed to the joint laxity were made. In addition, no individuals
central nervous system from the periphery. There- involved in competitive or recreational overhand
fore, musculotendinous mechanoreceptors have throwing activities were included.
been hypothesized as the primary contributor to
joint position sense in the mid ranges of motion. Instrumentation
This hypothesis is supported by the pronounced
Kinematic data were collected using the Polhemus
detrimental effect that muscle fatigue exerts on Fastrak 3Space magnetic tracking system (Colchester,
joint position sense, both in active and passive VT). The Polhemus unit consists of a transmitter, three
reproduction paradigms.8,17 These mechanorecep- receivers, and a digitizer. The transmitter emits an
tors are more highly stimulated in the presence of electromagnetic field that is sensed by the receivers
and following muscular contraction and tension, and digitizer. Signal strength and orientation are used
with more intense activation being associated with to determine relative position and orientation of the

JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006 DOI 10.1002/jor

 UNCONSTRAINED SHOULDER JOINT POSITION SENSE 561

receivers in space. To track movement of the humerus

with respect to the thorax, receivers were placed on the
sternum, approximately 1.5 cm inferior to the jugular
notch,22 and on the humerus just above the lateral
epicondyle, via a custom-molded OrthoplastTM cuff and
VelcroTM strap. In addition, one receiver was fastened
to the acromion process23 for digitization purposes, but
was removed prior to testing. The transmitter was
positioned level to the thoracic receiver with the subject
seated.
Following attachment of the receivers, various bony
landmarks were digitized on the thorax and humerus to
establish anatomical coordinate systems, in accordance
with the standard endorsed by the International Society
of Biomechanics.24 The body segments and correspond-
ing digitization points were: thorax—C7, T8, jugular
notch, and xiphoid process (Fig. 1a); humerus—medial
and lateral epicondyles and humeral head (Fig. 1b). The
center of the humeral head was calculated using a least
squares algorithm and was defined as the point that
moved the least during several small arcs of motion.25
Euler angles were used to represent two sequence-
dependent humeral rotations with respect to the thorax,
consisting of the plane of elevation and degree of
elevation.26 According to our established anatomical
coordinate systems, this Euler sequence corresponded to
a z-x0 -z00 rotation sequence.
To occlude visual cues related to shoulder position, all
subjects were fitted with a head-mounted display (I-O
Display Systems, Sacramento, CA), modified with felt
attached to the top, sides, and bottom of the display unit
to eliminate the influence of external light sources. The
display permitted kinematic output from the computer
to be presented to subjects on a two-dimensional screen.
Therefore, subjects could view the computer output with
complete visual occlusion of the position and movement
of the shoulder joint.

Testing Procedures
All testing was completed in a single session and
performed on the dominant upper extremity. Subjects
performed a standardized warm-up procedure includ-
ing Codman’s pendulums and stretches for the rotator
cuff muscles. Codman’s pendulum exercises were
performed with subjects bent over with the nondomi-
nant hand on a table, and holding a 2.5 lb (11.1 N)
weight in their dominant hand with the weight hanging
down at arm’s length. Subjects performed one set of 15 Figure 1. Depiction of digitized points and anatomi-
repetitions of arm circles, both clockwise and counter- cal coordinate systems for the (a) thorax and (b)
clockwise, followed by one set of 15 repetitions of a back humerus.
and forth movement in the sagittal plane.27 Stretches
consisted of holding a static external and then internal
rotation position, both with the shoulder abducted to without back support to minimize cutaneous tactile
approximately 908, for two sets of 15 s each. Following cues from the lower back. The stool height was adjusted
the warm-up procedure, subjects removed their shirts such that their knees were flexed to approximately 908
(females wore sports bras) and all jewelry that may with their feet flat on the floor (Fig. 2).
have contributed to tactile cues during testing. Subjects The testing protocol was thoroughly explained to the
were seated on a fully adjustable pneumatic stool subject while watching the visual output, first on the

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006

562 SUPRAK ET AL.

Figure 2. Photograph of experimental set-up showing sensors and head-mounted

display.

computer monitor, then through the head-mounted Three seconds later, another computer generated
display. A gray screen with a black square in the voice instructed subjects to return. Subjects then
center was presented to the subject, via custom attempted to replicate the target position in both plane
written Labview software (National Instruments, and elevation angles. When subjects perceived that the
Austin, TX). The black square represented the target shoulder was at the target position, they used the
position for a given trial. On the four sides of the contralateral hand to push a trigger button interfaced
screen, rectangular boxes appeared to prompt subjects with the computer to time-stamp the reproduced
as to which direction to move their arm to arrive at the position. Subjects were instructed to maintain the
target position (Fig. 3a). shoulder in the reproduced position for 1 s after
All trials began with the arm at the side. Subjects pushing the trigger button, at which time an audible
were instructed to move their arms in the direction of the beep sounded and the trial ended.
rectangular boxes. When the actual shoulder position The procedure was explained and demonstrated to
was within 58 of the target position in both plane and subjects until they felt comfortable with the process.
elevation angles, all of the boxes disappeared and a red Prior to the start of testing, subjects performed at least
dot appeared on the screen, representing the instanta- five practice trials at a target position consisting of a
neous shoulder position (Fig. 3b). Subjects continued to plane of 458 anterior to the coronal plane and 458 of
position the arm until the red dot on the screen was elevation. The practice trials were repeated until
inside the black square, indicating that the shoulder was subjects felt comfortable and confident in performing
in the target position. Once the shoulder was in the the task. To address effects of plane and elevation on
target position for 1 s, an audible beep was heard, and unconstrained joint position sense, nine target positions
the screen turned black and remained so for the were presented: elevation angles of 308, 508, 708, 908, and
remainder of the trial. Subjects were instructed to 1108 in the scapular plane (defined as 358 anterior to the
maintain their shoulder in the target position for 5 s, coronal plane) and plane angles of 08, 208, 358, 608, and
during which time they were to concentrate only on the 808 at 908 of elevation. These nine trials were automated
position of the shoulder. After the subject maintained via the software, and separated by a 15-s rest interval.
the target position for 5 s, a computer-generated voice The target positions were presented in random order,
instructed subjects to relax, at which time the subject according to a balanced Latin square design.28 To
lowered the arm back to the side. establish reliability, the nine trials were repeated in a

JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006 DOI 10.1002/jor

 UNCONSTRAINED SHOULDER JOINT POSITION SENSE 563

Figure 3. Computer output seen through the head-mounted display (A) guiding the
subject to target position and (B) with the shoulder in the target position.

randomized order unlike that used for the first nine-trial Statistical Analysis
sequence, following a 10-min rest period. Thus, subjects
completed a total of 18 trials (two trials at each of the SPSS version 13.0 (Chicago, IL) was used for
nine positions). statistical analysis. Intraclass correlation coefficients
Kinematic data were converted into humeral plane [ICC (3,1)]2,29 and standard errors of measurement
and elevation angles, using transformation matrices values were calculated for repositioning error magni-
between the thoracic and humeral coordinate systems. tude at each target position using the observed errors
Three-dimensional vectors were calculated, using these from before and after the 10-min rest interval. Two
plane and elevation angles, as lines running from the repeated measures analyses of variance (ANOVA)
center of the humeral head through the midpoint were conducted to determine the effect of plane and
between the medial and lateral epicondyles at the elevation angles on the magnitude of repositioning
presented and reproduced angles. The angle between error. For analysis of the effect of plane on reposition-
presented and reproduced position vectors was calcu- ing error, the five target positions consisting of
lated for each trial and assumed to represent the various plane angles at 908 of elevation were
absolute magnitude of the repositioning error (Fig. 4) included. For analysis of the effect of elevation angle,

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006

564 SUPRAK ET AL.

tion, the data were divided into positions at or

below 908 elevation and positions at or above 908.
The linear contrast was repeated excluding the
data for 1108 of elevation, revealing a more
pronounced linear decrease (p < 0.001) in reposi-
tioning error as shoulder elevation increased to 908
(Fig. 5b). A pairwise comparison revealed a
. significant increase in error from 908 to 1108 of
Figure 4. Depiction of repositioning error calcula- elevation (p ¼ 0.029).
tion.

the five target positions consisting of various eleva- DISCUSSION

tion angles in the scapular plane were included. If a
main effect was found for plane or elevation angle, The purpose of the present study was to deter-
polynomial contrasts were conducted to uncover mine the effects of plane and elevation angles on
significant linear trends. An a priori a level of 0.05
repositioning error in an unconstrained joint
was set for all analyses.
position sense testing paradigm. In examining
the effect of plane, we hypothesized that, due to
RESULTS stretching of the capsule and ligaments as the
plane angle approached 08 (coronal plane), repo-
ICCs for repositioning error magnitude were sitioning error would decrease. However, this
variable at different presented shoulder positions, hypothesis was not supported, as no effect of
and most were found to be relatively low (between plane on repositioning error was observed. Often
0.11–0.69; Table 1). For statistical analysis, the when a nonsignificant effect is observed, the
error scores for each position from the two test power of the analysis becomes an issue. A power
sequences were averaged together, and the mean analysis conducted using pilot data, with a
errors were used in all further analyses. minimal detectable difference in error between
The ANOVA revealed no main effect for plane positions of 28, a standard deviation of 2.488, and
angle on repositioning error magnitude [F(4, 18) ¼ an a level of 0.05 revealed that only 12 subjects
1.36, p ¼ 0.255, Fig. 5a]. However, the ANOVA for were necessary to achieve a power level of 0.8.
elevation angle revealed a main effect on reposi- This result is in contrast to the results of several
tioning error [F(4, 18) ¼ 6.431, p < 0.001]. studies employing methods involving internal
A significant linear decrease (p ¼ 0.002) in reposi- and external shoulder rotation in 908 of abduc-
tioning error was observed with increasing eleva- tion. Rotational studies of joint position sense
tion angle, with the smallest error observed at 908. have reported enhanced repositioning accuracy in
A sharp increase in repositioning error occurred both active16 and passive3,7,16 repositioning of a
from 908 to 1108 of elevation. Because torque about passively presented position, as the presented
the shoulder theoretically peaks at 908 of eleva- position approaches the end ROM. In addition, in
studies of shoulder kinesthesia, most commonly
Table 1. ICC and Standard Error of Measurement tested using the threshold to detection of passive
Values for Repositioning Errors/Across Presented motion (TTDPM) method, decreased threshold to
Positions detection was found as the starting position
approached end range.7,29 In TTDPM testing,
Joint Position blindfolded subjects begin with their shoulder
(Plane/Elevation, Deg) ICC SEM (Deg)
abducted to 908 and positioned in some internal/
35/30 0.36 3.99 external rotation starting position, and then
35/50 0.69 3.03 are asked to indicate when they sense a passive
35/70 0.26 3.51 rotational displacement at the shoulder. The
35/90 0.49 1.90 TTDPM is measured as the displacement occur-
35/110 0.23 3.18 ring between the onset of the passive movement
0/90 0.11 3.72 and the subjects depressing a hand-held trigger.
20/90 0.10 4.07
The disparities between the results of these
60/90 0.45 2.55
80/90 0.43 2.39
studies and the present findings may have arisen
from multiple sources. In our study, the most

JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006 DOI 10.1002/jor

 UNCONSTRAINED SHOULDER JOINT POSITION SENSE 565

Figure 5. Vector error magnitude across (A) plane and (B) elevation angles.

extreme plane angle (08) was standardized for all 908 against gravity in all of these positions.
subjects and was not normalized to each subject’s Because afferent input from muscle spindles
ROM. The mean ROM in horizontal abduction for may be the primary contributor to joint position
our sample was 308 posterior to the coronal plane. sense,8 the information provided by these recep-
Therefore, subjects may not have sufficiently tors may have overridden that provided by the
approached their end range to result in significant capsuloligamentous receptors, leading to a non-
stretching of the capsular and ligamentous recep- significant effect of plane. In previous uniplanar
tors. Another possible explanation of this result studies reporting enhanced joint position sense
is related to the nature of the measurement. near the end ROM, the arm was supported by
Although the joint position with respect to the the testing apparatus,3,7,16 whereas in our
plane angle was varied from 08 to 808, subjects study, position was maintained by active muscle
were required to maintain the elevation angle of contraction.

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006

566 SUPRAK ET AL.

Our results indicate that joint position sense is as the elevation angle increases, the changes in
affected by elevation angle, as illustrated by the muscle length, capsular tightness, and scapular
significant linear decrease in repositioning error orientation also take place, which may affected
from 308 to 908 of elevation and the subsequent the observed findings. Further investigation is
increase in error as elevation further increased to needed to examine more directly the effect of
1108. As the shoulder elevation angle increases, muscle activation level on unconstrained joint
the torque applied to the shoulder due to gravity position sense.
increases with the increasing moment arm of the The ICC values calculated for most of the joint
center of mass of the upper extremity, theoreti- positions tested were quite low (Table 1). This may
cally peaking at 908 of elevation. Therefore, the be partly due to the novelty of the task. Studies
muscular effort required and tendon tension involving joint position sense testing in internal/
developed in both attaining and remaining in the external rotation have reported ICC values higher
target position for the 5-s period, and then than those reported here.3,35 However, subjects in
returning to that position, presumably increased our study were asked to find a joint position, to hold
as the elevation angle approached 908. We the shoulder in this position for 5 s, concentrating
hypothesize that, if positions higher than 1108 only on shoulder position, and to replicate the
elevation were tested, the repositioning error position in two planes. The observed variability in
would further increase as the torque decreased. the error scores from one testing interval to another
However, 1108 was the highest elevation angle may have been due to the demanding nature of the
tested in this study. protocol, making the measurement more variable
The g motor neurons innervating intrafusal within subjects. Visual inspection of the data for all
muscle fibers are activated simultaneously with a subjects across the two trial sequences revealed a
motor neurons innervating extrafusal muscle large variation in error scores between testing
fibers.30 This coactivation of both the intrafusal sequences within subjects. In contrast, the varia-
and extrafusal fibers maintains the sensitivity of bility of scores between subjects was noticeably
muscle spindle afferents over the full ROM.30 smaller, contributing to the low ICC values
Poppele and Quick reported that the sensitivity of observed. To account for some of the within-subject
muscle spindles in the cat tenuissimus muscle, a variability, the mean of the error scores for each
hip abductor, is directly related to the degree of joint position during both testing intervals was
muscle contraction.31 Myers and colleagues exam- used in all statistical analyses. Future studies
ined shoulder internal rotator muscle onset employing this testing method should include more
latency in response to an imposed external rota- trials at each joint position to obtain a more
tion perturbation at various levels of muscle representative error score.
activation.5 They found that muscle onset latency The magnitudes of the repositioning errors in
was significantly reduced in the presence of our study differ from those previously reported in
some level of muscle activation prior to perturba- the literature for uniplanar rotational studies. In
tion. They attributed this effect to a heightened rotational studies examining joint position sense
sensitivity of the muscle spindle afferents asso- at the shoulder and knee, the repositioning errors
ciated with the involved musculature. This result ranged from 28 to 58,8,35–37 whereas in our study,
has been corroborated by various authors examin- they ranged from 48 to 98. Therefore, there is some
ing muscle spindle afferent responses to passive overlap in the errors seen in our study and those
versus active stretching perturbations32 and to employing uniaxial testing methods. The differ-
increases in g motor neuron stimulation.33,34 ences may be due to the more challenging nature of
Group Ib golgi tendon organ afferents respond our measurement protocol, in that subjects were
to tension developed within the tendons associated required to reposition the joint in two planes
with contracting or stretched muscle fibers.30 As rather than only one.
tension within the tendon increases, Ib afferent In the present study, we found no significant
stimulation rises.18 Due to the increased gravita- difference between repositioning errors at differ-
tional torque as the elevation angle approached ent plane angles, but JPS did vary with alterations
908 in our study and the presumed increase in in elevation angle. These results suggest that
required muscular activation level, our results signals arising from musculotendinous mechan-
may have been due to heightened musculotendi- oreceptors are an increasingly important source of
nous mechanoreceptor sensitivity as the presented afferent feedback contributing to shoulder joint
shoulder position increased in elevation. However, position sense in multiple planes, and that

JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006 DOI 10.1002/jor

 UNCONSTRAINED SHOULDER JOINT POSITION SENSE 567

increased muscle activation results in enhanced levels of muscle contraction. Clin Orthop Rel Res
sense. Active muscle contraction is essential to 407:92–101.
maintaining shoulder joint stability. Our findings 6. Lephart SM, Myers JB, Bradley JP, et al. 2002.
may lend further insight into the important Shoulder proprioception and function following
thermal capsulorraphy. Arthroscopy 18:770–778.
proprioceptive role played by shoulder muscula-
7. Safran MR, Borsa PA, Lephart SM, et al. 2001.
ture during dynamic conditions. These findings
Shoulder proprioception in baseball pitchers. J
may also have implications for improving dynamic Shoulder Elbow Surg 10:438–444.
shoulder stability in the rehabilitation setting, 8. Voight ML, Hardin JA, Blackburn TA, et al. 1996.
advocating the prescription of exercises utilizing The effects of muscle fatigue on and the relation-
unconstrained movements in functional ranges ship of arm dominance to shoulder proprioception.
under conditions and positions of increased muscle J Orthop Sports Phys Ther 23:348–352.
activation to optimize proprioceptive feedback 9. Zuckerman JD, Gallagher MA, Lehman C, et al.
from musculotendinous mechanoreceptors. How- 1999. Normal shoulder proprioception and the
ever, further study is required to determine the effect of lidocaine injection. J Shoulder Elbow Surg
effects of muscle activation level directly on 8:11–16.
10. Blasier RB, Carpenter JE, Huston LJ. 1994.
unconstrained shoulder joint position sense.
Shoulder proprioception: effect of joint laxity, joint
Future research should focus on the effect of
position, and direction of motion. Orthop Rev
altering the muscle activation level involved in 23:45–50.
this type of experimental paradigm at a given 11. Grigg P. 1976. Response of joint afferent neurons in
target position on the repositioning error observed. cat medial articular nerve to active and passive
The results of such a study would provide more movements of the knee. Brain Res 118:482–485.
direct evidence of the behavior of musculotendi- 12. Salo PT, Tatton WG. 1993. Age-related loss of
nous receptors in functional situations. This line of knee joint afferents in mice. J Neurosci Res
research may be further extended to the investiga- 35:664–677.
tion of how joint position sense is affected by 13. Steinbeck J, Bruntrup J, Greshake O, et al. 2003.
conditions of external loading in patients with Neurohistological examination of the inferior gle-
nohumeral ligament of the shoulder. J Orthop Res
shoulder injuries to gain insight into the extent of
21:250–255.
damage to musculotendinous receptors in such
14. Vangsness CT, Ennis M, Taylor JG, et al. 1995.
conditions and provide more specific guidelines for Neural anatomy of the glenohumeral ligaments,
rehabilitative efforts in the clinical setting. labrum, and subacromial bursa. Arthroscopy 11:
180–184.
ACKNOWLEDGMENTS 15. Lonn J, Crenshaw AG, Djupsjobacka M, et al. 2000.
Position sense testing: influence of starting position
The authors would like to thank Brian Fedor for and type of displacement. Arch Phys Med Rehabil
technical assistance. 81:592–597.
16. Rogol IM, Ernst G, Perrin DH. 1998. Open and
closed kinetic chain exercises improve shoulder
REFERENCES joint reposition sense equally in healthy subjects. J
Athl Train 33:315–318.
1. Sherrington CS. 1906. On the proprio-ceptive 17. Lee H-M, Liau J-J, Cheng C-K, et al. 2003.
system, especially in its reflex aspect. Brain Evaluation of shoulder proprioception following
29:467–482. muscle fatigue. Clin Biomech 18:843–847.
2. Aydin T, Yildiz Y, Yanmis I, et al. 2001. Shoulder 18. Gregory JE, Brockett CL, Morgan DL, et al. 2002.
proprioception: a comparison between the shoulder Effect of eccentric muscle contractions on Golgi
joint in healthy and surgically repaired shoulders. tendon organ responses to passive and active
Arch Orthop Trauma Sur 121:422–425. tension in the cat. J Physiol 538:209–218.
3. Janwantanakul P, Magarey ME, Jones MA, et al. 19. Vallbo AB. 1974. Human muscle spindle discharge
2001. Variation in shoulder position sense at mid during isometric voluntary contractions: amplitude
and extreme range of motion. Arch Phys Med relations between spindle frequency and torque.
Rehabil 82:840–844. Acta Physiologica Scandinavica 90:319–336.
4. Nyland JA, Caborn DN, Johnson DL. 1998. The 20. Barden JM, Balyk R, Raso VJ, et al. 2004. Dynamic
human glenohumeral joint. A proprioceptive and upper limb proprioception in multidirectional
stability alliance. Knee Surg Sports Traumatol shoulder instability. Clin Orthop Relat Res
Arthrosc 6:50–61. 420:181–189.
5. Myers JB, Riemann BL, Ju Y-Y, et al. 2003. 21. Tripp BL, Boswell L, Gansneder BM, et al.
Shoulder muscle reflex latencies under various 2004. Functional fatigue decreases 3-dimensional

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006

568 SUPRAK ET AL.

multijoint position reproduction acuity in the over- 30. Kandel ER, Schwartz JH, Jessell TM. 2000.
head-throwing athlete. J Athl Train 39:316–320. Principles of neural science. New York: McGraw-
22. Borstad JD, Ludewig PM. 2002. Comparison of Hill; 2000.
scapular kinematics between elevation and low- 31. Poppele RE, Quick DC. 1985. Effect of intrafusal
ering of the arm in the scapular plane. Clin muscle mechanics on mammalian muscle spindle
Biomech 17:650–659. sensitivity. J Neurosci 5:1881–1885.
23. Karduna AR, McClure PW, Michener LA, et al. 32. Kallio J, Linnamo V, Komi PV. 2004. The effects of
2001. Dynamic measurements of three-dimen- muscle history on short latency stretch reflex
sional scapular kinematics: a validation study. J response of soleus muscle. J Electromyogr Kinesiol
Biomech Eng 123:184–190. 14:411–421.
24. Wu GE, van der Helm FCT, Veeger HEJ, et al. 33. Durbaba R, Taylor A, Ellaway PH, et al. 2001.
2005. ISB recommendation on definitions of Modulation of primary afferent discharge by
joint coordinate systems of various joints for the dynamic and static gamma motor axons in cat
reporting of human joint motion — Part II: muscle spindles in relation to the intrafusal fibre
shoulder, elbow, wrist, and hand. J Biomech types activated. J Physiol 532:563–574.
38:981–992. 34. Jami L, Lan-Couton D, Petit J. 1980. Glycogen-
25. Harryman DT 2nd, Sidles JA, Harris SL, et al. depletion method of intrafusal distribution of
1992. The role of the rotator interval capsule in gamma-axons that increase sensitivity of spindle
passive motion and stability of the shoulder. J Bone secondary endings. J Neurophysiol 43:16–26.
Joint Surg Am 74:53–66. 35. Lephart SM, Warner JJP, Borsa PA, et al. 1994.
26. An KN, Browne AO, Korinek S, et al. 1991. Three- Proprioception of the shoulder joint in healthy,
dimensional kinematics of glenohumeral elevation. unstable, and surgically repaired shoulders. J
J Orthop Res 9:143–149. Shoulder Elbow Surg 3:371–380.
27. Prentice WE. Rehabilitation techniques in sports 36. Bullock-Saxton JE, Wong WJ, Hogan N. 2001.
medicine. St. Louis: Mosby; 1994. The influence of age on weight-bearing joint reposi-
28. Portney LGW. 2000. Foundations of clinical siton sense of the knee. Exp Brain Res 136:400–406.
research: applications to practice, 2nd edn. Upper 37. Skinner HB, Wyatt MP, Hodgdon JA, et al. 1986.
Saddle River, NJ: Prentice-Hall Health; 2000. Effect of fatigue on joint position sense of the knee.
29. Allegrucci M, Whitney SL, Lephart SM, et al. 1995. J Orthop Res 4:112–118.
Shoulder kinesthesia in healthy unilateral athletes 38. Darling WG, Miller GF. 1995. Perception of arm
participating in upper extremity sports. J Orthop orientation in three-dimensional space. Exp Brain
Sports Phys Ther 21:220–226. Res 102:495–502.

JOURNAL OF ORTHOPAEDIC RESEARCH MARCH 2006 DOI 10.1002/jor