Manual Therapy 20 (2015) 28e37

Contents lists available at ScienceDirect

Manual Therapy
journal homepage: www.elsevier.com/math

Systematic review

The effects of shoulder injury on kinaesthesia: A systematic review

and meta-analysis

a Gustavsson a, Craig Wassinger b, Gisela Sole c, *
Charlotte Fyhr a, Linne
a

Institution of Community Medicine and Rehabilitation, Physiotherapy, University of Ume, Sweden

Department of Physical Therapy, East Tennessee State University, USA
c
Centre for Health, Activity and Rehabilitation Research, School of Physiotherapy, University of Otago, New Zealand
b

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 1 May 2014
Received in revised form
8 August 2014
Accepted 26 August 2014

The aim of this systematic review was to synthesize the evidence for changes for proprioceptive variables
consisting of movement and position sense in participants with glenohumeral musculoskeletal disorders.
Five databases were searched until 13th August 2013. Methodological quality was assessed and metaanalyses were performed for active and passive joint reposition sense (AJPS and PJPS) and movement
sense, determined with threshold to detection of passive motion (TTDPM). The search yielded 17 studies,
four of which were classied as having high methodological quality, seven as moderate and six as low
quality. For participants with post-traumatic glenohumeral instability, pooled ndings indicate moderate
evidence for higher TTDPM for involved shoulders compared to control groups and the contralateral
uninvolved side, indicating decreased movement sense. For AJPS and PJPS there was moderate to limited
evidence for signicant increased errors for involved compared to uninvovled shoulders, but not when
compared to the control groups. Limited evidence was found for decreased AJPS acuity for patients with
chronic rotator cuff pain and for patients with unspecied shoulder pain compared to healthy controls.
Movement sense is most likely to be impaired after shoulder injury involving post-traumatic instability
when compared to the contralateral shoulder and to controls, while decits for AJPS and PJPS are more
likely to be evident compared to the contralateral shoulder in participants with glenohumeral musculoskeletal disorders.
2014 Elsevier Ltd. All rights reserved.

Keywords:
Shoulder
Proprioception
Kinestheasia
Position sense
Movement sense

1. Introduction
Changes in proprioception have been considered for various
musculoskeletal injuries and pain disorders such as low back pain
(O'Sullivan et al., 2013), cervical disorders (Lee et al., 2006), and
knee (Relph et al., 2014) and shoulder injuries (Myers and Oyama,
2008). Proprioception has been dened as the afferent information from peripheral segments and contributes towards joint stability, postural control and balance, and motor control (Myers and
Oyama, 2008). Thus, information from capsuloligamentous mechanoreceptors, Golgi tendon organs, muscle spindles, visual and
cutaneous sensors is mediated and integrated by the central nervous system, modifying neuromuscular control systems (Proske,
2006; Myers and Oyama, 2008; Fortier and Basset, 2012).

* Corresponding author. Centre for Health, Activity and Rehabilitation Research,

University of Otago, Box 56, Dunedin, 9054, New Zealand. Tel.: 64 3 4797936;
fax: 64 3 4798414.
E-mail address: [email protected] (G. Sole).
http://dx.doi.org/10.1016/j.math.2014.08.006
1356-689X/ 2014 Elsevier Ltd. All rights reserved.

Various classications for proprioception are available. While

Riemann et al. (2002) dened three domains of joint position sense,
kinesthaesia and sense of tension, others have dened four domains: kinaesthesia (joint position sense and movement sense),
sense of tension, sense of effort and the sense of balance (Fortier
and Basset, 2012). Assessment of joint position sense (JPS) involves passively moving a joint or limb to a target angle and
returning to the starting position. The subject then voluntarily
moves the limb back to the target angle (active joint position sense,
AJPS), alternatively the limb is moved passively by a device (passive
joint position sense, PJPS). The angular error between the target
and the replicated position is determined (Riemann et al., 2002).
Movement sense is assessed by measuring the angular threshold to
detection of passive motion (TTDPM) (Riemann et al., 2002; Fortier
and Basset, 2012). The limb is moved passively by a device and the
person is asked to indicate the rst instance the motion of the joint
is perceived (Riemann et al., 2002). The angular threshold for
detecting movement is recorded.
It has been suggested that injuries lead to proprioceptive decits, and in combination with tissue damage and pain, these

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

alterations may affect outcomes (Myers and Oyama, 2008). In line

with this, proprioceptive measures have been investigated as a
dependent variable to determine efcacy of treatment interventions. For example, shoulder JPS and TTDMD were found to
be restored after surgery for patients with recurrent anterior
instability (Lephart et al., 1994; Rokito et al., 2010). Shoulder JPS
was also improved following arthroplasty in patients with glenohumeral arthritis (Cuomo et al., 2005). More recently, a programme of neurocognitive therapeutic exercises that involves
proprioceptive and motor problem-solving tasks were described
for patients with shoulder impingement syndrome (Marzetti et al.,
2014).
Proprioceptive decit associated with musculoskeletal disorders has also been suggested to be a contributing factor towards
recurrence, persistent disability and secondary changes such as
osteoarthritis (Lephart et al., 1994; Marzetti et al., 2014). However,
conicting results have been reported for cross-sectional studies
reporting proprioceptive changes of injury of other joints, such as
the knee (Relph et al., 2014). The aim of this study was thus to
systematically review the literature to investigate whether position
and movement sense differ in injured shoulders compared to the
contralateral uninjured shoulders, and compared to uninjured
control participants.
2. Materials and methods
2.1. Search
PRISMA guidelines were followed where appropriate for
observational studies (Moher et al., 2009). The electronic databases
MEDLINE, EMBASE, CINAHL, PubMed and SportDiscus were
searched up to 13th August 2013. Key words and Boolean operators

29

were modied in different combinations as relevant for the individual database (Supplementary File 1). Reference lists of included
articles were checked for additional published papers.
Inclusion criteria were participants with glenohumeral musculoskeletal disorders; comparison to control groups (between-group
comparisons) or contralateral uninvolved shoulders (within-group
comparisons); proprioceptive domains: kinaesthesia, active or
passive joint position or movement sense; and written in English,
Swedish, German or Dutch. Exclusion criteria were acromioclavicular joint injuries, glenohumeral hypermobility without
injury, rheumatologic disorders and osteoarthritis, osteoporosis,
whiplash injury, congenital abnormalities or neurological disorders, and reviews. There was no restriction for participants' sex or
age, and studies investigating effects of surgery or other treatment
interventions on shoulder proprioception were included if preintervention baseline results were presented.
2.2. Quality assessment and data extraction
The 27-question Downs and Black (1998) checklist was modied
to assess methodological quality of the included studies. Questions
relating to randomized clinical trials were excluded, similar to
previous systematic reviews including non-randomized observational studies (Munn et al., 2010). The modied checklist had 14
yes/no questions considered relevant for this review and the
questions were grouped into reporting (Q 1 e 7), external validity
(Q 8), internal validity (Q 9 e 13) and power (Q 14) (Table 1).
Question 4 (distribution of confounders) had a maximum score of 2,
awarded as described in Table 1. The nal question (Q14) was
reworded to Were appropriate power calculations reported? and
the score was changed from 0 to 5 to a scale of 0e1 (Ratcliffe et al.,
2014).

Table 1
Methodological quality assessment, modied from Downs and Black (1998).
Category

Question

Application to this review

Reporting

1. Is the hypothesis/aim/objective of the study clearly described?

2. Are the main outcomes to be measured clearly described in
the Introduction or Methods section?
3. Are the characteristics of the patients included in the study
clearly described?
4. Are the distributions of principal confounders for each group
to be compared clearly described?

Time since injury, diagnosis

External validity
Internal validity:
bias

Internal validity:
selection bias

Power

5. Are the main ndings of the study clearly described?

6. Does the study provide estimates of the random variability in the
data for the main outcomes? (SD, SE or CI)
7. Have actual probability values been reported (e.g. 0.035 rather
than <0.05) for the main outcomes except where the probability
value is less than 0.001?
8. Were the subjects asked to participate in the study representative
of the entire population from which they were recruited?
9. If any of the results of the study were based on data dredging,
was this made clear?
10. Were the statistical tests used to assess the main outcomes
appropriate?
11. Were the main outcome measures used accurate
(valid and reliable)?
12. Were the participants of the two groups recruited from the
same population?
13. Was there adequate adjustment for confounding in the analyses
from which the main ndings were drawn?

14. Were appropriate power calculations reported?

Age, gender, level of physical activity (including overhand activity).

A score of 2 was applied when all three were reported,
and 1 when one or two were reported. This question was
considered not applicable for studies reporting only within-group
comparisons (injured to contralateral uninjured shoulders).

It was considered important that the occupation or sports-related

activities were reported for patients and control groups.
A score of 1 was applied when absence or presence of symptoms
during proprioception testing or on the day of testing was reported.
A zero score was applied when current symptoms for injured
participants were not mentioned.
A score of 1 was applied when a power or a sample size calculation
was provided. If these were not given or there was no explanation
whether the number of participants was appropriate a 0 was applied

30

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

Two reviewers (CF, LG) assessed each study independently. A

third reviewer (GS) was then included in discussion to conrm the
results and to reach consensus when there was uncertainty or
disparity between reviewers. An overall score was provided and for
the purpose of this review and quality was classied as 75% as
high, 60e74% as moderate, and <60% as low (Munn et al., 2010;
Hoch and McKeon, 2014).
Data were extracted for each study, including demographic data,
shoulder disorder, measuring equipment and tasks. Means and
standard deviation (SD) for injured and control groups were
extracted for active or PJPS and movement sense (TTDPM). If nonparametric data (medians) were presented, means and SD were
estimated (Hozo et al., 2005). All reported movement directions
were included, and sub-groups for each (abduction, exion, internal and external rotation) were assessed. Where more than one
starting position or target angle was reported, data for the one
closest to end-range was extracted as proprioceptive measures
have been shown to be most accurate closer to the end than the
inner or mid-range of motion (Janwantanakul et al., 2003; Yang
et al., 2008; King et al., 2013). The corresponding authors were
contacted when relevant data were not presented or when only
graphs were published. If the author did not reply, the results were
measured from the graphs by two reviewers (CF, LG). Proprioceptive data were extracted by one author, entered into Review Manager (RevMan) (Computer program, version 5.2. Copenhagen: The
Nordic Cochrane Centre, The Cochrane Collaboration, 2012) and
were veried for correctness by a second reviewer.
Meta-analyses were performed when more than one study was
found for a specic proprioceptive domain and shoulder disorder.
Individual movement directions (exion, abduction, internal and
external rotation) for each proprioceptive domain (AJPS, PJPS and
TTDPM) were rst combined in a sub-group analysis, and an overall
analysis was then performed, combining all movement directions
for respective domains. Standardized mean differences (SMD, 95%
condence intervals, CIs) were calculated in RevMan for dependent
variables between groups of participants (patients versus controls)
and between sides (involved side versus contralateral uninvolved
side). SMD were used due to expected variability in the methods
and devices used to assess a specic type of proprioceptive domain.
Where measures for the dominant and non-dominant shoulders
were provided for the control group, the dominant shoulder was
used as a reference, unless the average for both sides was the only
information available. Random effect models were used for the
meta-analyses and to calculate I2 (Israel and Richter, 2011).
Between-group comparisons were made when involved shoulders
of patients were compared to control persons, whereas withingroup comparisons were made for the involved to the contralateral uninvolved shoulders.
A system to rate levels of evidence that had been described for
randomised clinical trials (van Tulder et al., 2003) was used to
summarize the results. For the purpose of this review the trials
were replaced with studies. The level of evidence was dened as
strong (consistent ndings from multiple high-quality studies),
moderate (consistent ndings from one high-quality study and one
or more moderate to low-quality studies, or multiple moderate to
low-quality studies), limited (one low quality study), conicting
(inconsistent ndings among multiple studies) or no evidence (no
studies) (van Tulder et al., 2003).

were found (three in German) that fullled the eligibility criteria

(Fig. 1).
3.2. Methodological quality
Corresponding author of four studies were contacted due to the
lack of exact gures for the dependent variables and two provided
rl et al., 2011; Haik et al., 2013). Scores
the requested details (Mo
were discussed for 17 individual questions to reach consensus (7%
of all questions). The average score for the methodological quality
assessment was 64% (range 42e93%). Four studies (Lephart et al.,
1994; Anderson and Wee, 2011; Hung and Darling, 2012; Haik
et al., 2013) were assessed as high quality, and seven (Smith and
Brunolli, 1989; Jerosch and Thorwesten, 1998; Aydin et al., 2000;
Chu et al., 2002; Machner et al., 2003; Zuckerman et al., 2003;
Barden et al., 2004) as moderate quality (Table 2). Only two
studies reported power (Anderson and Wee, 2011; Hung and
Darling, 2012) and three studies (Barden et al., 2004; Anderson
and Wee, 2011; Haik et al., 2013) provided sufcient information
regarding recruitment of participants, thereby indicating higher
external validity. The other studies recruited participants from a
single hospital or clinic, or did not report their recruitment method.
Confounding may have affected the outcomes as the recruitment of
control subjects as being from a similar population group as those
from patients was not always clear (Q4 and Q12, Table 2). Most
studies did not report whether or not symptoms were experienced
by patients during testing although these may have confounded the
results (Q13).
3.3. Overview of included studies
The studies had a median of 15 patients (range 6e55) (Table 3).
One study (N 55) (Rokito et al., 2010) investigated proprioception
in two groups of patients with post-traumatic glenohumeral
instability of which one group (n 30) had been included in an
earlier study (Zuckerman et al., 2003). To avoid bias, only one of the
groups (n 25) (Rokito et al., 2010) was included in the metaanalysis. Thirteen studies evaluated proprioception in shoulders
with post-traumatic glenohumeral instability (Smith and Brunolli,
1989; Lephart et al., 1994; Forwell and Carnahan, 1996; Jerosch
and Thorwesten, 1998; Machner et al., 1998; Slobounov et al.,
1999; Aydin et al., 2000; Chu et al., 2002; Zuckerman et al., 2003;
Potzl et al., 2004; Rokito et al., 2010; Hung and Darling, 2012);
one with multi-directional glenohumeral instability (Barden et al.,
2004); two with shoulder impingement syndrome (SIS) (Machner
et al., 2003; Haik et al., 2013); one with patients with unspecic
rl et al., 2011), and one with chronic rotator cuff
shoulder pain (Mo
pain (CRCP) (Anderson and Wee, 2011). Three studies only
compared the injured to the contralateral uninjured shoulder
(Machner et al., 2003; Zuckerman et al., 2003; Rokito et al., 2010).
Equipment and methods used for the assessment of proprioceptive variables were diverse (Table 1). Ten studies measured AJPS
in abduction, exion, internal and/or external rotation, with target
angles ranging from 40 to 150 . All six studies measuring PJPS used
external rotation, four also evaluated internal rotation, and abduction and exion were measured in two studies (Table 3). TTDPM was
assessed in six studies. The included studies used a large variation of
number of trials per target, from 3 to 10 trials per target, with total
trials up to 72 trials per arm (Hung and Darling, 2012).

3. Results
3.1. Search result

3.4. Post-traumatic glenohumeral joint instability and multidirectional instability

The search yielded 378 studies and after exclusion of duplicates,

irrelevant titles and screening abstracts and full text, 17 articles

A meta-analysis was performed for data for participants with

post-traumatic glenohumeral instabilities. For the between-group

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

31

Fig. 1. Flowchart for the search strategy.

Table 2
Results of the methodological quality assessment.
Authors (Date)

Categories and questions

Reporting

Smith and Brunolli (1989)

Lephart et al. (1994)
Forwell and Carnahan (1996)
Jerosch and Thorwesten (1998)
Machner et al. (1998)
Slobounov et al. (1999)
Aydin et al. (2000)
Chu et al. (2002)
Machner et al. (2003)
Zuckerman et al. (2003)
Barden et al. (2004)
Potzl et al. (2004)
Rokito et al. (2010)
Anderson and Wee (2011)
rl et al. (2011)
Mo
Hung and Darling (2012)
Haik et al. (2013)
Total

EV

Internal validity

10

11

12

13

Power
14

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
17

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
17

1
1
1
1
1
1
1
1
1
0
1
0
1
1
0
1
1
14

1
2
1
1
1
1
1
1
NA
NA
1
0
NA
1
1
2
2
Score 1: 10
Score 2: 3 NA: 3

1
1
1
1
1
1
1
0
1
1
0
1
1
1
1
1
1
15

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
17

1
1
0
0
0
0
0
1
0
1
1
1
0
1
0
1
1
8

0
0
0
0
0
0
0
0
0
0
1
0
0
1
0
0
1
3

1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
16

1
1
0
1
1
0
1
1
1
1
1
1
0
1
1
1
1
14

0
1
0
1
0
0
1
1
0
1
0
0
1
1
1
1
1
10

0
1
0
0
0
0
0
0
1
0
0
0
0
1
0
1
1
5

0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
2

0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
2

EV: External Validity; NA: not applicable; U: unable to determine; H: High quality (75%), M: moderate quality (60e74%); L: low quality(<60%).

Total
(%)

Quality
classication

64
80
43
60
53
47
60
60
61
62
60
50
54
93
53
93
87

M
H
L
M
L
L
M
M
M
M
M
L
L
H
L
H
H

32

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

Table 3
Summary of the included studies.
Authors

Subjects characteristics
(age, height, body mass)

Control group
(age, height, body mass)

Traumatic recurrent shoulder instabilities and multi-directional instability

18 men 2 women
Aydin et al. (2000)
24 men 2 women with traumatic
28.8 years (SD 4.7)
recurrent anterior GHJ instability
23.2 years (SD 4.7)

Measuring device

Task and measured variable

Cybex II Dynamometer
and Onclinometer

AJPS test
Starting position: Seated,
shoulder in 90 abd, 0
and 20 ER.
Target angles: 10 from
starting position for IR
and ER.
Trials: 3/target
AJPS test
Starting position: Seated
Target angles: a
self- selected position.
Trials: 10/target
AJPS test,
Starting position: Supine,
shoulder in 90 abd.,
elbow in 90 exion and
forearm in pronation.
Target angles: 10 from
full ER, 30 ER and 30 IR.
Trials: 3/target
Manual aiming and
pointing
Starting position: Seated
with arm rested on
a waist height table.
Target angles: movement
accuracy at approximately
100 , 120 and 150 ex,
performed with (1) full
vision, (2) blinded, and
(3) blinded and with
vibration to opposite arm.
Variables: movement time,
nger pointing accuracy,
peak nger velocity.
Trials: 9/target
AJPS and PJPS test
Starting position: Seated
with 0 of shoulder abd.,
neutral rot. and 90 ex .
Target angles:
Active 45 , 90 , 135 abd.;
45 IR; 45 or 90 ER.
Passive 45 IR; 45
and 75 ER.
Speed: NR
Trials: 8/target
AJPS test
Starting position:
Target angles: Various
positions in ex, abd
and rot.
(in 90 abd), with and
without vision.
Trials: 4/target, in total 36
PJPS test and TTDPM
Starting position: Supine,
90 elbow ex, 90 shoulder
abd. and neutral or 30 ER.
PJPS target angles: 10 from
starting point in both IR
and ER (within-group)
TTDPM movement direction:
IR and ER (within- and
between-group)
Speed: 0.5 /s
Trials: 3/target
TTDPM
Starting point: seated,
shoulder abd between

Barden et al. (2004)

9 men 3 women with

multidirectional instability
27.0 year (SD 8.3)

9 men 3 women
26.3 years (SD 5.4)

4 Vicon Motion System

video cameras in
3D: 9 markers

Chu et al. (2002)

12 men 8 women with one

or more anterior GHJ
dislocations (athletes)
21.8 years (SD 3.7)

10 men 10 women
21.9 years (SD 2.5)

Cybex II isokinetic
dynamometer
Electrogoniometer

Forwell and
Carnahan (1996)

3 men 3 women with

recurrent anterior GHJ
instability
23.3 years (SD 5.2)

4 men 3 women
22.8 years (SD 4.3)

Waterloo Spatial Motion

Analysis and Recording
Technique system
(WATSMART)

Hung and Darling (2012)

7 men 3 women with

anterior GHJ dislocation
(19e37 years)

11 men 4 women
(20e39 years)

Electromagnetic
tracking system
(The Ascension
Technology mini-BIRD)
Custom-made Plexiglas
manipulandum consisting
of a movable arm.
Goniometr

Jerosch and
Thorwesten (1998)

13 men 15 women with

anterior shoulder dislocation,
due to undergo arthroscopic
stabilization
26.8 years (range 21e35); average
recurrence 8.5 (range 4e32)

18 men 12 women,
27.5 years (range 22e36),

Movement analysis
system (TOPOMED,
Orthodata, Ldenscheid),
markers on olecranon
and distal ulna

Lephart et al. (1994)

30 men with chronic, recurrent,

traumatic
anterior GHJ dislocation or
subluxation (athletes)
24.3 years (SD 6.5)

26 men 14 women
19.4 years (SD 1.2)

Proprioception testing
device

Machner et al. (1998)

6 men 4 women with traumatic

anterior GHJ dislocations, history
of 5.4 mean recurrences

9 men 6 women,
27 years (no SD provided)

Custom built chair

and CPM equipment

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

33

Table 3 (continued )
Authors

Subjects characteristics
(age, height, body mass)

Control group
(age, height, body mass)

(range 3e17); on list for surgical

stabilization
29 years (no SD provided).

Measuring device

Task and measured variable

(Ortho-med Co.,Lautertal,
Germany)

40 and 60 .
Movement direction: abd
Speed: 1.3 /s
Trials: 5/direction
AJPS test
Starting position: neutral
position, IR starting
from 90 abd.
Target angles: 50 , 100 ,
150 abd.; 50 , 100 , 150
ex.; 0 , 45 ER; 45 IR.
Trials: 4/target, in total
36 trials
Presented combined
results only.
PJPS test and TTDPM
Starting position: seated
with back support and
footrest; ER was tested
standing.
Target angles/movement:
ex., abd. and ER. No angles
available.
Speed: 2 /s
Trials: 3/target
AJPS test
Starting position: Standing
and arm resting by their side.
Target angles: 40 and 80 abd.
Trials: 5/target

Potzl et al. (2004)

7 men 7 women with traumatic

recurrent anterior GHJ instability,
due to undergo stabilizing surgery
28 years (16e52)

15 volunteers
28 years (19e33)

Contact free motion analysis

system (Kinemetrix 3.0,
Orthodata, Germany):
Change-coupled device,
camera,
water level and goniometer

Rokito et al. (2010)

55 patients with unilateral

recurrent anterior GHJ instability,
due to undergo surgery
30 years (16e51). One group
(n 25) is included in the
meta-analysis.

Compared to contralateral
uninjured shoulder

Modied dynamometer
testing apparatus,
Electronic goniometer

Slobounov et al. (1999)

4 men 2 women with

traumatic GHJ instability
21.8 years (SD 2.4)

Real-time motion- tracking

system

Smith and Brunolli (1989)

6 men 2 women with

anterior GHJ dislocation
34.0 years (SD 14.3)

Gr 1 (healthy)
3 men 3 women
21.6 years (SD 2.6)
Gr 2 (control)
4 men 6 women
21.3 years (SD 3.1)
5 men 5 women
22.9 years (SD 1.9)

Zuckerman et al. (2003)

30 patients with unilateral

traumatic recurrent anterior
GHJ instability due to
undergo surgery
29.7 years (17e51)

Compared to contralateral
uninjured shoulder

Electronic dynamometer
and goniometer

15 men 11 women with

CRCP, including impingement
symptoms;
56.4 years (SD 10.8)

17 men 13 women
55.5 years (SD 4.5)

6 Vicon M Series cameras

with 20 mm lens: 7 markers

Haik et al. (2013)

15 women with SIS

35.5 years (SD 5.8),
Occupation: assembly line
workers

15 women
34.4 years (SD 5.5),
Occupation: assembly
line workers
(Second control group:
not assembly workers,
not included in this review)

Isokinetic dynamometer
(Biodex Multi Joint
System 3)

Machner et al. (2003)

9 men 6 women with

unilateral SIS stage II
46 years (42e59)

Compared to contralateral
uninjured shoulder

Custom built chair

(Ortho-med Co.,
Lauteral, Germany)

Other disorders
Anderson and Wee (2011)

Shoulder-wheel
apparatus

PJPS test and TTDPM

Starting position: Supine,
90 abd. and 90 elbow ex.
PJPS target angles: random
angle between 5 and 30
in ER and IR.
TTDPM movement direction:
ER and IR.
Speed: 1e2 /s
Trials: 3/target
PJPS test and TTDPM
Starting position: seated
with back support and
footrest in ex. and abd.,
standing, ER.
Target angles: 40 ex. 40 ,
abd. 20 , ER.
Speed: 2 /s
Trials: 3/target
AJPS test
Starting position: Seated,
shoulder position not reported.
Target angle: 40 and 100 abd.
Trials: 5/target
AJPS and PJPS test
Starting position: Seated;
90 abd. in scapular plane,
90 IR or from neutral.
Target angles:
45 , 75 ER.
Speed:
AJPS: 5 /s
PJPS: 2 /s
Trials: 3/target
TTDPM
Starting position: seated,
abd. 60 .
Movement direction: abd.
Speed: 1.3 /s
Trials: NR
(continued on next page)

34

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

Table 3 (continued )
Authors

Subjects characteristics
(age, height, body mass)

Control group
(age, height, body mass)

Measuring device

Task and measured variable

rl et al. (2011)
Mo

Gr 1: 10 men 2 women with

unspecic shoulder pain
45.7 years (SD 10.8)
Gr 2: 6 men 2 women with
unspecic shoulder pain
47.9 years (SD 7.4)

2 men 6 women
28 years (SD 4.3)

Zebris CMS-HS 3D motion

measuring system
(Zebris Medical GmbH,
Isny, Germany):
7 markers

AJPS test
Starting position: arm
hanging relaxed by the
side of the body.
Target angles:
60 , 90 , 120 ex.; 60 ,
90 , 120 abd.; 0
ER and 0 IR.
Trials: 3/target

Abd: abduction; AJPS: active joint position sense; CPM: continuous passive motion; CRCP: Chronic rotator cuff pain syndrome; ER: external rotation; ex.: exion; GHJ:
Glenohumeral joint; IR: internal rotation; NR: Not reported; PJPS: passive joint position sense; Rot.: rotation; SIS: Subacromial impingement syndrome; TTDPM: threshold to
detection of passive motion.

comparison (involved shoulder of patients compared to controls)

no statistically signicant difference was found for the pooled,
multiple directional AJPS (P 0.06, Supplementary File 2) and PJPS
(P 0.19, Supplementary File 3). However, when considering the
individual movement directions, signicant between-group differences were found for AJPS for exion (42 injured participants;
SMD, 95%CI: 0.62 , 0.19e1.05 ; P 0.005) and external rotation (84
injured participants, SMD, 95%CI: 1.00 , 0.25e1.74 , P 0.009). A
statistically signicant difference was found for the pooled TTDPM
(SMD, 95% CI, 2.27, 0.94e3.59 , P < 0.001) and for all individual
directions (Supplementary File 4) indicating impairment for this
domain for the involved side following traumatic shoulder
dislocations.
Only one study reported within-group comparisons (involved
compared to uninvolved shoulders) for AJPS for patients with
traumatic GHJ instability, with an SMD (95%CI) of 0.68 (0.12e1.24 ,
P 0.02) for internal rotation and 0.69 (0.13e1.25 , P 0.02) for
external rotation (Aydin et al., 2000). Statistically signicant differences were found for within-group comparisons for the pooled
PJPS (SMD, 95%CI, 3.05 , 2.04e4.06 , P < 0.001, Supplementary File
5) and TTDPM (SMD, 95%CI, 2.37, 1.56e3.18 , P < 0.001,
Supplementary File 6).
Sensitivity tests were performed for the meta-analyses by
excluding low quality studies included in the between-group analyses for AJPS (Slobounov et al., 1999; Potzl et al., 2004), PJPS
(Machner et al., 1998) and TTDPM (Rokito et al., 2010), and for
within-group analyses for PJPS (Rokito et al., 2010) and TTDPM
(Machner et al., 1998; Rokito et al., 2010). While this decreased the
number of analyses possible for individual movement directions,
the overall pooled ndings for the domains for between- and

within-group comparisons were largely unchanged, that is, the

conclusion of statistically signicant differences did not change.
Funnel plots for AJPS were symmetrical (Supplementary File 7),
indicating no bias due to selective reporting was present within
these studies. Those for PJPS and TTDPM were asymmetrical due to
the low number of studies, thus reporting bias may be contributory.
Two studies exploring proprioception for participants with GHJ
instability were excluded from the meta-analysis as different
methods of assessing proprioception were used. Forwell and
Carnahan (1996) used a pointing task to assess proprioception,
assessing nger pointing accuracy, movement time and peak nger
velocity, with and without vision. We calculated mean differences
(95%CI) for the variables, measured from gures, for the no-vision
conditions. No signicant differences were found between the
involved and uninvolved shoulders for this variable. Barden et al.
(2004) evaluated proprioception of patients with multidirectional
instability using a motion analysis system. AJPS was performed at a
self-selected position and errors were measured in millimetres,
nding a mean difference (95% CI) of 4.8 mm (0.6e13.2) when
compared to a control group.
Based on the van Tulder rating system, the meta-analysis for
shoulders with post-traumatic instabilities provides moderate
evidence that, overall, AJPS (P 0.06) and PJPS (P 0.19) did not
differ signicantly from control participants. Compared to the
contralateral uninjured shoulders, there was limited evidence for
decits for AJPS (P < 0.05) and moderate evidence for decits for
PJPS (P < 0.001). Similarly, moderate evidence exists for decits
for motion sense, reected by increased TTDPM, for the involved
shoulder for patients compared to controls (P < 0.001) and
compared to the contralateral uninvolved shoulder (P < 0.001).

Table 4
Mean differences for proprioceptive variables for patients with rotator cuff and sub-acromial impingements syndromes and unspecied shoulder pain, compared to control
groups (between-group comparisons) and contralateral uninvolved shoulders (within-group comparisons).
Movement

Author (Year)

Active joint repositioning sense

Abduction
Anderson and Wee, 2011
rl et al., 2011
Mo
rl et al., 2011
Flexion
Mo
Internal rotation
Haik et al., 2013
rl et al., 2011
Mo
External rotation
Haik et al., 2013
rl et al., 2011
Mo
Passive joint repositioning sense
Internal rotation
Haik et al., 2013
External rotation
Haik et al., 2013
Threshold to detection of passive motion
Abduction
Machner et al., 2003

26
20
20
15
20
15
20

Standardized mean differences (95% CI)

patients;
patients;
patients;
patients;
patients;
patients;
patients;

30 control
8 control
8 control
15 control
8 control
15 control
8 control

15 patients; 15 control
15 patients; 15 control
15 patients

Between-group analysis

Within-group analysis

0.85 (0.30 to 1.40)

1.07 (0.20 to 1.95)
0.73 (1.58 to 0.12)
0.04 (0.86 to 0.78)
0.64 (1.38 to 0.09)
0.01 (0.73 to 0.70)
0.20 (1.02 to 0.62)

0.83 (0.26 to 1.40)

0.30 (0.42 to 1.03)

0.15 (0.57 to 0.86)

0.10 (0.62 to 0.81)

0.03 (0.69, 0.74)

0.04 (0.68 to 0.75)

0.06 (0.66 to 0.78)

3.05 (1.96 to 4.15)

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

3.5. Chronic rotator cuff pathology, sub-acromial impingement

syndrome and unspecied shoulder pain
Table 4 provides SMD (95% CI) for between- and within-group
comparisons for studies exploring other shoulder disorders,
including chronic rotator cuff pain syndrome (CRCP) (Anderson and
Wee, 2011), sub-acromial impingement syndrome (Machner et al.,
rl et al.,
2003; Haik et al., 2013) and unspecied shoulder pain (Mo
2011). In a high quality study, Anderson and Wee (2011) provided
evidence for decreased active JPS for shoulders with CRCP
compared to controls and the uninjured contralateral shoulder.
rl et al. (2011) showed signicantly decreased AJPS for
Similarly, Mo
participants with unspecied shoulder pain compared to controls
for abduction. However, no signicant differences were found for
rl
AJPS exion and rotations (external and internal rotation) (Mo
et al., 2011; Haik et al., 2013), and for passive rotation JPS (Haik
et al., 2013). Increased TTDPM were found for shoulders with unspecied pain compared to the opposite side (Machner et al., 2003).
There was, thus, limited evidence from individual trials for
signicantly decreased AJPS in abduction for patients with CRCP
rl et al.,
(Anderson and Wee, 2011) and with unspecied pain (Mo
2011) compared to controls.
4. Discussion
This systematic review found moderate evidence that patients
with shoulder injuries exhibited increased TTDPM compared to
uninjured controls and when comparing the involved to the uninvolved shoulders, thus indicating decits for motion sense. The
ndings were less consistent for joint position sense: while no
statistically signicant differences were found for pooled AJPS or
PJPS between patients and controls, moderate to limited evidence
was found for decreased joint position sense for the involved
compared to uninvolved contralateral shoulders. Overall, ndings
are limited due to a small number of studies included, with a median of 15 patients across all studies.
For patients with post-traumatic glenohumeral instability, all
movement directions were combined for each domain in the metaanalysis to be more representative of functional proprioceptive
sensibility. It could be argued that the active tests have greater
validity for functional movements compared to the passive ones
and AJPS was, indeed, the most common variable used for betweengroup comparisons. While the overall results indicate no signicant
difference between groups for AJPS (P 0.06, Supplementary File
2), signicant differences were found for abduction and external
rotation as individual movement directions. The high quality study
by Anderson and Wee (2011) also found signicant between- and
within-group differences for patients with CRCP compared to
controls for AJPS in abduction. These movement directions could
thus be considered for future research investigating effects of
shoulder injury compared to controls or the contralateral uninjured
shoulder, rather than exion and internal rotation.
It is currently unknown whether relationships exist between the
domains, for example between decits of AJPS and TTDPM, or
another proprioceptive domain. Similarly, it is still unclear whether
there is a relationship between proprioceptive decits of the
shoulder and athletic performance (such as throwing accuracy),
functional capacity or other measures of impairment (such as
muscle weakness). There is evidence that proprioceptive acuity
improves following surgery for recurrent anterior glenohumeral
dislocations (for example, Rokito et al., 2010). Fortier et al. (2010)
showed that fatigue following isometric, concentric and eccentric
exercise in uninjured participants leads to reduced AJPS, measured
with potentiometers and a strain gauge. However, a more recent
study indicated no differences in proprioceptive acuity after

35

fatiguing throwing exercises (Freeston et al., 2014). Investigators

have also explored differences in proprioceptive acuity between
overhand athletes and non-athletes, with conicting ndings.
Dover et al. (2003) showed that overhand athletes had decreased
external rotation AJPS (determined with an inclinometer)
compared to non-athletes, and for external rotation and extension
AJPS for the throwing compared to non-throwing shoulders. More
recently, however, increased mean proprioceptive acuity of three
joints, the shoulder, ankle and spine, was found to be associated
with level of performance of elite athletes (Han et al., 2013). The
relationship between proprioceptive changes and recovery
following shoulder injury, functional capacity and athletic performance needs further investigation.
This systematic review included a wide range of musculoskeletal disorders, namely instabilities, dislocations, rotator cuff syndrome, SIS and unspecied shoulder pain. Participants were also
recruited from diverse population groups, including sporting
groups and occupationally injured groups. Furthermore, proprioception was measured with various technological devices,
including custom-built devices in the earlier studies, inclinometers,
motion analysis systems and isokinetic dynamometry, using
different starting positions and target angles. There appears to be
no standardized assessment procedure, which adds towards the
clinical heterogeneity of the studies. Clinical heterogeneity was
evident in different ranges of motions that were tested which is
also likely to inuence the results (Janwantanakul et al., 2003; Yang
et al., 2008; King et al., 2013).
There was no consistent reporting of current symptoms experienced by participants at the time of data collection. Two of the
rl et al., 2011) indicated that
studies (Anderson and Wee, 2011; Mo
the presence of pain could have inuenced results. Barden et al.
(2004) indicated that 16 of the 24 shoulder with multidirectional
instability were symptomatic without specifying whether the
symptoms included pain and/or apprehension. It is also unclear
whether those participants experienced symptoms while performing the pointing task investigated in that study. Pain may affect
the proprioception sensibility, thus reporting of possible symptoms
during the tasks would be important (Matre et al., 2002).
Clinics usually do not have the specialized equipment, such as
motion analysis systems or isokinetic dynamometers, thus
methods using inclinometers, pointing or tracking accuracy tests
may be more relevant for clinical practice. A recent study showed
that throwing accuracy is sensitive to fatiguing exercises while
these exercises did not affect proprioception, as assessed by AJPS
with a custom-built device (Freeston et al., 2014). The clinical
relevance of the proprioceptive tests for AJPS, PJPS and TTDPM may
need to be explored, particularly as they are often performed while
constraining the shoulder to a specic movement direction in one
plane. Using assessments that include all domains of the sensorimotor system and the entire kinetic chain, such as accuracy of
throwing, may have more potential to differentiate between
injured and uninjured people, and potentially as an indicator for
risk of injury.
While eleven studies were scored as having moderate or high
quality on the modied Blacks and Downs checklist, six were
considered to have low quality. The sub-category of reporting was
generally well represented by most studies, however, there were
limitations in terms of external validity, internal validity and power.
For example, very few studies clearly presented their recruitment
of patient and control participants. It was common that the patients
were recruited only from one clinic or hospital, thus generalizability of results may be limited. Furthermore, demographic data
for control participants, with regards to their sports or occupation
factors related to overhand activities were unclear. It is recommended that future studies should consider matching control

36

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

participants more clearly with the injured participants, and that

sample sizes should be more carefully dened in order to give
greater condence in the ndings. The ndings of this review may
assist in performing a priori power analyses for such future studies.
4.1. Limitations
This review was limited to joint and movement sense, thus
other domains, such as sense of effort or force, need further
investigation. The original Downs and Black (1998) checklist for the
quality assessment and the rating of the level of evidence (van
Tulder et al., 2003) were modied for the purpose of this study.
The clinical heterogeneity of the included studies can be viewed as
a limitation of this review. We standardized extraction of data by
including the data for assessments with the angle closer to end
range of motion. We also acknowledge the heterogeneity of the
participants with shoulder injuries and understand that it might
have inuenced the results. However, it is known that specic
impairments are not always associated with specic injuries, and
also that tissue injury is not directly related with symptoms (Lewis,
2011). It is currently unclear whether specic injuries, such as
traumatic instabilities, are more likely to have proprioceptive deficits compared, for example, to those with CRCP.
5. Conclusion
Exercises are often prescribed for patients with musculoskeletal
shoulder disorders with the aim of improving proprioception. This
systematic review and meta-analysis investigated whether there
were differences in proprioception in patients with musculoskeletal disorders compared to a control group, or to the contralateral
uninjured shoulder. Overall, there is moderate to limited evidence
suggesting signicant decits for AJPS and PJPS for participants
with traumatic GHJ instabilities comparing the involved to the
contralateral uninvolved shoulders but no signicant differences
when comparing patient to control groups. For movement sense,
signicant decits were found, evident by increased TTDPM, for the
injured shoulders when compared to the contralateral uninjured
shoulder and to controls. There is limited evidence that AJPS is
decreased in patients with CRCP compared to controls and when
comparing the symptomatic to contralateral shoulder for
abduction.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.math.2014.08.006.
References
Anderson VB, Wee E. Impaired joint proprioception at higher shoulder elevations in
chronic rotator cuff pathology. Arch Phys Med Rehabil 2011;92:1146e51.
Aydin I, Yildiz Y, Ozgurbuz C, Yamur H, Genc U, Kalyon TA. Proprioceptive function
of healthy and unstable shoulder joints (German) Dtsch Z Sportmed 2000;51:
336e40.
Barden JM, Balyk R, Raso VJ, Moreau M, Bagnall K. Dynamic upper limb proprioception in multidirectional shoulder instability. Clin Orthop Relat Res
2004;420:181e9.
Chu JC, Kane EJ, Arnold BL, Gansneder BM. The effect of a neoprene shoulder stabilizer on active joint-reposition sense in subjects with stable and unstable
shoulders. J Athl Train 2002;37:141e5.
Cuomo F, Birdzell MG, Zuckerman JD. The effect of degenerative arthritis and
prosthetic arthroplasty on shoulder proprioception. J Shoulder Elb Surg
2005;14:345e8.
Dover GC, Kaminski TW, Meister K, Powers ME, Korodyski MB. Assessment of
shoulder proprioception in the female softball athlete. Am J Sports Med
2003;31:431e7.

Downs S, Black N. The feasibility of creating a checklist for the assessment of the
methodological quality both of randomised and non-randomised studies of
health care interventions. J Epidemiol Community Health 1998;52:377e84.
Fortier S, Basset FA. The effects of exercise on limb proprioceptive signals.
J Electromyogr Kinesiol 2012;22:795e802.
Fortier S, Basset FA, Billaut F, Behm D, Teasdale N. Which type of repetitive muscle
contractions induces a greater acute impairment of position sense?
J Electromyogr Kinesiol 2010;20:298e304.
Forwell LA, Carnahan H. Proprioception during manual aiming in individuals with
shoulder instability and controls. J Orthop Sports Phys Ther 1996;23:111e9.
Freeston J, Adams R, Ferdinands RE, Rooney K. Indicators of throwing arm fatigue in
elite adolescent male baseball players: a randomized crossover trial. J Strength
Cond Res 2014;28:2115e20.
Haik MN, Camargo PR, Zanca GG, Alburquerque-Sendin F, Salvini TF, MattielloRosa SM. Joint position sense is not altered during shoulder medial and lateral
rotations in female assembly line workers with shoulder impingement syndrome. Physiother Theory Pract 2013;29:41e50.
Han J, Waddington G, Anson J, Adams R. Level of competitive success achieved by
elite athletes and multi-joint proprioceptive ability. J Sci Med Sport 2013. http://
dx.doi.org/10.1016/j.jsams.2013.11.013 [Epub ahead of print].
Hoch MC, McKeon PO. Peroneal reaction time after ankle sprain: a systematic review and meta-analysis. Med Sci Sports Exerc 2014;46:546e56.
Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median,
range, and the size of a sample. BMC Med Res Methodol 2005;5.
Hung YJ, Darling WG. Shoulder position sense during passive matching and active
positioning tasks in individuals with anterior shoulder instability. Phys Ther
2012;92:563e73.
Israel H, Richter RR. A guide to understanding meta-analysis. J Orthop Sports Phys
Ther 2011;41:496e504.
Janwantanakul P, Magareya ME, Jonesa MA, Grimmera KA, Miles TS. The effect
of body orientation on shoulder proprioception. Phys Ther Sport 2003;4:
67e73.
Jerosch J, Thorwesten L. Proprioceptive capability of the glenohumeral complex in
patients with posttraumatic shoulder instability (German) Z Orthop Ihre
Grenzgeb 1998;136:230e7.
King J, Harding E, Karduna A. The shoulder and elbow joints and right and left sides
demonstrate similar joint position sense. J Mot Behav 2013;45:479e86.
Lee HY, Teng CC, Chai HM, Wang SF. Testeretest reliability of cervicocephalic
kinesthetic sensibility in three cardinal planes. Man Ther 2006;11:61e8.
Lephart SM, Warner JJ, Borsa PA, Fu FH. Proprioception of the shoulder joint in
healthy, unstable, and surgically repaired shoulders. J Shoulder Elb Surg 1994;3:
371e80.
Lewis JS. Subacromial impingement syndrome: a musculoskeletal condition or a
clinical illusion? Phys Ther Rev 2011;16:388e98.
Machner A, Merk H, Becker R, Rohkohl K, Wissel H, Pap G. Kinesthetic sense of the
shoulder in patients with impingement syndrome. Acta Orthop Scand 2003;74:
85e8.
Machner A, Wissel H, Heitmann D, Pap G. Changes in proprioceptive capacities of
the shoulder joint in ventral shoulder instability. A comparative study before
and after arthroscopic labrum rexation (German) Sportverletz Sportschaden
1998;12:138e41.
Marzetti E, Rabini A, Piccinini G, Piazzini DB, Vulpiani MC, Vetrano M, et al. Neurocognitive therapeutic exercise improves pain and function in patients with
shoulder impingement syndrome: a single-blind randomized controlled clinical
trial. Eur J Phys Rehabil Med 2014;50:255e64.
Matre D, Arendt-Nielsen L, Knardahl S. Effects of localization and intensity of
experimental muscle pain on ankle joint proprioception. Eur J Pain 2002;6:
245e60.
Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for
systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339:
b2535.
rl F, Matkey A, Bretschneider S, Bernsdorf A, Bradl I. Pain relief due to physioMo
therapy doesn't change the motor function of the shoulder. J Bodyw Mov Ther
2011;15:309e18.
Munn J, Sullivan SJ, Schneiders AG. Evidence of sensorimotor decits in functional
ankle instability: a systematic review with meta-analysis. J Sci Med Sport
2010;13:2e12.
Myers JB, Oyama S. Sensorimotor factors affecting outcome following shoulder
injury. Clin Sports Med 2008;27:481e90.
O'Sullivan K, Verschueren S, Van Hoof W, Ertanir F, Martens L, Dankaerts W. Lumbar
repositioning error in sitting: healthy controls versus people with sittingrelated non-specic chronic low back pain (exion pattern). Man Ther
2013;18:526e32.
Potzl W, Thorwesten L, Gotze C, Garmann S, Steinbeck J. Proprioception of the
shoulder joint after surgical repair for Instability: a long-term follow-up study.
Am J Sports Med 2004;32:425e30.
Proske U. Kinesthesia: the role of muscle receptors. Muscle Nerve 2006;34:
545e58.
Ratcliffe E, Pickering S, McLean S, Lewis J. Is there a relationship between subacromial impingement syndrome and scapular orientation? A systematic review. Br J Sports Med 2014;48:1251e6.
Relph N, Herrington L, Tyson S. The effects of ACL injury on knee proprioception: a
meta-analysis. Physiotherapy 2014;100:187e95.
Riemann BL, Myers JB, Lephart SM. Sensorimotor system measurement techniques.
J Athl Train 2002;37:85e98.

C. Fyhr et al. / Manual Therapy 20 (2015) 28e37

Rokito AS, Birdzell MG, Cuomo F, Di Paola MJ, Zuckerman JD. Recovery of shoulder
strength and proprioception after open surgery for recurrent anterior instability: a comparison of two surgical techniques. J Shoulder Elb Surg 2010;19:
564e9.
Slobounov SM, Poole ST, Simon RF, Slobounov ES, Bush JA, Sebastianelli W, et al. The
efcacy of modern technology to improve healthy and injured shoulder joint
position sense. J Sport Rehabil 1999;8:10e23.
Smith RL, Brunolli J. Shoulder kinesthesia after anterior glenohumeral joint dislocation. Phys Ther 1989;69:106e12.

37

van Tulder M, Furlan A, Bombardier C, Bouter L. Updated method guidelines for

systematic reviews in the Cochrane Collaboration back review group. Spine
2003;28:1290e9.
Yang JL, Chen S, Jan MH, Lin YF, Lin JJ. Proprioception assessment in subjects with
idiopathic loss of shoulder range of motion: joint position sense and a novel
proprioceptive feedback index. J Orthop Res 2008;26:1218e24.
Zuckerman JD, Gallagher MA, Cuomo F, Rokito A. The effect of instability and
subsequent anterior shoulder repair on proprioceptive ability. J Shoulder Elb
Surg 2003;12:105e9.