Physiotherapy 91 (2005) 94–100

Mechanosensitivity of the median nerve and mechanically produced

motor responses during Upper Limb Neurodynamic Test 1
Shapour Jaberzadeha,∗ , Sheila Scutterb , Homer Nazeranc
a
Department of Physiology, Adelaide University, Adelaide, SA 5005, Australia
b School of Physiotherapy, University of South Australia, Adelaide, SA, Australia
c School of Informatics and Engineering, Flinders University, Adelaide, SA, Australia

Abstract

Objectives In spite of the widespread use and recognised importance of the Upper Limb Neurodynamic Test 1 (ULNT1) in clinical practice,
controversy remains about the neurophysiological basis for sensory and motor responses to the test. The aims of this study were to determine the
effects of two limb positions (neutral and ULNT1) on mechanosensitivity of the median nerve, and to investigate the mechanisms underlying
muscle stiffness and loss of range during the elbow extension component of the ULNT1.
Design In both limb positions, a KIN-COM® dynamometer controlled passive elbow extension at 3◦ /second, and recorded both elbow
extension range of movement and elbow flexor resistive torque.
Setting Electrophysiology laboratory, School of Physiotherapy, University of South Australia.
Participants Twenty-six asymptomatic subjects.
Main outcome measures Range of elbow extension and elbow flexor resistive torque. Surface electrodes placed over 10 upper limb muscles
recorded electromyographic (EMG) activity. Using a hand-held micro-switch, participants indicated occurrence of pain onset and pain that
limited further movement.
Results The median nerve was more sensitive to mechanical longitudinal stresses during passive elbow extension in the ULNT1 position than
in the neutral position, as demonstrated by increased EMG activity and increased mean elbow flexor resistive torque (neutral position, 0.8 Nm,
95% confidence interval 0.7–0.9 Nm; ULNT1 position, 3.9 Nm, 95% confidence interval 3.8–4.0 Nm). Pain onset and pain limit occurred
earlier in range in the ULNT1 position (pain onset, 45◦ , 95% confidence interval 40–50◦ ) than in the neutral position (3◦ , 95% confidence
interval 0–6◦ ). In the ULNT1 position, EMG activity was greater in muscles responsible for an antalgic posture in the upper limb, although
some EMG activity was evident prior to pain onset.
Conclusions The mechanosensitivity of the median nerve was greater during elbow extension in the ULNT1 position than in the neutral
position. The increased EMG activity that occurred primarily in muscles contributing to an antalgic posture with concurrent changes in
resistive torque and range of movement may be explained by a flexor withdrawal response to pain that acts to reduce the stretch on the median
nerve. Muscle activity prior to pain onset is probably due to mechanoreceptor activation following preferential stretch of the median nerve.
© 2004 Chartered Society of Physiotherapy. Published by Elsevier Ltd. All rights reserved.

Keywords: Upper limb tension test; Neurodynamic tension test; Mechanosensitivity; Median nerve; Flexor withdrawal reflex

Introduction
Over the last two decades, physiotherapists have devel-
oped examination techniques known as ‘neural tension tests’
to assess the mechanosensitivity of the major nerve trunks
and their central connections [1–4]. The techniques for the
∗ Corresponding author. Tel.: +61 8 8303 6433; fax: +61 8 8303 3356.
E-mail address: shapour.jaberzadeh@adelaide.edu.au (S. Jaberzadeh).
upper limb are based on the premise that upper limb move-
ments require neural tissues to move, adapt and extend in
length [5], and use precise and sequential positioning of the
limb to apply tension to selected neural components. For ex-
ample, to tension the median nerve using the Upper Limb
Neurodynamic Test 1 (ULNT1), the sequence and positions
described by Butler [3] are as follows: shoulder girdle de-
pression, glenohumeral abduction to 110◦ , forearm supina-
tion, wrist and finger extension, and glenohumeral external

0031-9406/$ – see front matter © 2004 Chartered Society of Physiotherapy. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.physio.2004.09.021
 S. Jaberzadeh et al. / Physiotherapy 91 (2005) 94–100
rotation, with elbow extension being the final component of
the test. It is theorised that the ULNT1 will move and ex-
ert longitudinal stress on the median nerve and its proximal
nerve roots and the cervicobrachial plexus [6,7], with asso-
ciated sensory and motor responses providing information
about the mechanosensitivity of the median nerve. Anatom-
ical studies have shown the anatomical validity of this test
[8–11].
Despite the widespread use of neural tensioning tech-
niques, little is known about whether neural tissues be-
come more sensitive to loading in the ULNT1 position
compared with a neutral position (arm resting in 30◦ ab-
duction, forearm supinated, wrist and hand in neutral, and
elbow flexion at 90◦ ). It is, therefore, important to ex-
plore the responses by comparing mechanosensitivity in
neutral and ULNT1 positions using the final component
of ULNT1, passive elbow extension. Mechanosensitivity
can be assessed by monitoring changes in elbow exten-
sion range and elbow flexor resistive torque during elbow
extension, and the corresponding mean electromyographic
(EMG) responses of selected shoulder and arm muscles
[12].
The aims of this study were to determine the effects of
two limb positions (neutral and ULNT1) on mechanosensi-
tivity of the median nerve, and to investigate the mechanisms
underlying motor responses during the elbow extension com-
ponent of the ULNT1. Greater mechanosensitivity would be
indicated by increased EMG activity and elbow flexor resis-
tive torque, decreased elbow extension range and pain that
occurred earlier in range.
Method
Twenty-six asymptomatic normal subjects (13 males, 13
females), aged 21–72 years (mean: 39.6 years; S.D.: 15.5),
participated in the study. All subjects were right handed and
gave written consent.
A KIN-COM® dynamometer (125 AP, Chattex Corpora-
tion, Chattanooga, TN, USA) was used for measurement of
elbow extension range and resistive torque. An integrated
multi-channel AMLAB® -based system (AMLAB Interna-
tional, Sydney, Australia) was developed to simultaneously
record multi-channel EMG activity and to integrate resistive
torque and range of movement [13].
Bipolar surface electrodes (Duo-Trode® Myo-Tronics
Inc., USA) were secured over 10 shoulder and arm muscles
referenced to anatomical landmarks [14]. The raw EMG sig-
nals were differentially amplified in the AMLAB® system
with a common-mode rejection ratio of 120 dB. Each ampli-
fier was individually isolated and had a high input impedance
(500 M) and low noise (<1 ␮V). The EMG signals were
band-passed filtered between 10 and 500 Hz. These signals
were amplified with a gain of 3000 and sampled at 1000 Hz.
The digitised data were then imported into an Excel spread-
sheet. The EMG signals were full wave rectified and the mean
95
EMG for 3 seconds before and after movement onset, pain
onset and pain limit were calculated.
The muscles were classified into two groups according to
their function: agonists (experimental muscles) and antago-
nists (control muscles) for an upper quarter antalgic posture.
The antalgic posture of shoulder girdle elevation, shoulder ad-
duction with medial rotation and some degrees of elbow and
wrist flexion [15] may reduce upper limb pain by shortening
the anatomic course of the median nerve and relieving ten-
sile stress. Based on this premise, the experimental muscles
selected were upper and middle fibres of trapezius, pectoralis
major, biceps brachii, flexor carpi radialis and brachialis. The
control muscles were triceps, deltoid, infraspinatus and the
lower fibres of trapezius. The reference electrode was a lip
clip [16].
Subject positioning and stabilisation
Stabilisation and control of the subjects’ position dur-
ing testing was achieved using the following apparatus: (1)
subject stabilisation baseboard; (2) range-of-motion-control
device; and (3) adjustable seat/plinth assembly of the KIN-
COM dynamometer. Subjects were positioned supine on the
subject stabilisation baseboard with a chinstrap to prevent
upper cervical extension. The finger and wrist splint of the
range-of-motion-control device was then placed on the poste-
rior surface of the wrist and hand with the hinge overlying the
subject’s axis of wrist flexion/extension. The KIN-COM lever
arm axis was then adjusted to the range-of-motion-control de-
vice elbow axis. The final position of the subject’s arm prior to
testing in the neutral position was with the arm resting in 30◦
of abduction, 90◦ of elbow flexion, the forearm supinated, and
the wrist and hand in neutral. The final position of the subject
prior to testing in the ULNT1 position was with the arm in
110◦ of abduction and full lateral rotation, the elbow in 90◦
flexion, the forearm completely supinated, and the wrist and
hand in full extension. In this position, the length of brachialis
and flexor carpi radialis are almost identical compared with
the neutral position. There is some degree of stretch in the
other four experimental muscles.
Passive elbow extension
Once in the test position (neutral or ULNT1), passive el-
bow extension was controlled at 3◦ /second by the KIN-COM
dynamometer to minimise the effects of the myotatic stretch
reflex [17]. In the neutral position, a single familiarisation
trial of elbow extension was taken to the point of pain on-
set. One further trial to the point of pain limit was conducted
for data collection. During this trial, passive elbow exten-
sion, elbow extension range, elbow flexor resistive torque and
EMG activity of shoulder and arm muscles were recorded.
Once resting symptoms were recorded, the arm was placed
in the ULNT1 position in the following sequence [3]: (1)
application of a constant shoulder girdle depression force of
40 mmHg [18]; (2) abduction to 110◦ ; (3) forearm supination;
96 S. Jaberzadeh et al. / Physiotherapy 91 (2005) 94–100
(4) wrist and finger extension; and (5) glenohumeral exter-
nal rotation. Subjects practiced using the hand-held micro-
switch to indicate pain during a familiarisation test, and were
instructed to activate the switch ‘as soon as the tightening
sensation changes to pain’ (pain onset) and ‘as soon as the
pain changes to the maximum level of pain a physiotherapist
is prepared to provoke’ (pain limit). This point is well within,
and quite different from, a level representing intolerable pain
for the subject [19].
Each component was performed manually to the onset
of pain. The stretch was slightly reduced and the range was
held by tightening the locking screws of the range-of-motion-
control device. This kept subjects pain free before adding the
subsequent components to complete the ULNT1 position.
Subjects could stop the test at any point in range by using a
separate hand-held switch. To avoid a sequencing bias, the
order of testing in neutral and ULNT1 positions was ran-
domised.
Results
Data were retrieved offline. Elbow extension range and
elbow flexor resistive torque at movement onset, pain onset
and pain limit in the neutral and ULNT1 positions are shown
in Fig. 1.
In both the neutral and ULNT1 positions, elbow flexor
resistive torques at pain onset and pain limit were significantly
larger than at movement onset (two-way repeated measure
ANOVA, P < 0.001).
Fig. 1. Elbow extension range of motion (A) and elbow flexor resistive
torque (B) at movement onset, pain onset and pain limit in the neutral and
ULNT1 positions.
Fig. 2 shows resistive torque and EMG responses in the
ULNT1 position in one representative subject.
There was increased EMG activity in all experimental
muscles, except flexor carpi radialis, before the onset of
pain that further increased after the onset of pain. With the
exception of triceps, activity in the control muscles remained
very low.
To allow comparison of data at equivalent ranges in the
neutral and ULNT1 positions, 3-second mean EMG in the
neutral position was determined at the ranges that corre-
sponded to movement onset, pain onset and pain limit in the
ULNT1 position (Fig. 3).
The EMG activity in the neutral position remained at base-
line levels with minimal changes at ranges that corresponded
with pain onset and pain limit in the ULNT1 position. While
most of the progressive increase in activity before pain onset
to pain limit occurred in the experimental muscles, particu-
larly the upper fibres of trapezius, a similar pattern of activity
was also evident in triceps. Note that activity in all muscles
declined once the stretch was removed after pain limit (two-
way repeated measure ANOVA, P < 0.001). Pairwise com-
parison of mean differences for 3-second mean EMG before
and after movement onset, pain onset and pain limit during
passive elbow extension in different shoulder and arm mus-
cles using Tukey’s Honestly Significantly Different (HSD)
test supports the significant changes in experimental muscles
and triceps after movement onset (Fig. 3).
Discussion
Effects of shoulder and arm position on
mechanosensitivity of the median nerve
It was hypothesised that the median nerve in the ULNT1
position would be more mechanically sensitive than in the
neutral position of arm and shoulder joints. This study
demonstrated that the elbow extension range in the ULNT1
position is halted by pain in the middle range of joint excur-
sion, and is associated with an increase in elbow flexor resis-
tive torque and EMG activity of the experimental muscles.
Therefore, the median nerve is more mechanically sensitive
in the ULNT1 position. Elbow extension stresses the median
nerve by elongating the nerve itself, the proximal nerve roots
and the brachial plexus [8–10,20]. Clinically, elbow extension
in the ULNT1 position has been shown to be limited when
the median nerve is involved in carpal tunnel syndrome [21].
Specific analysis of the EMG responses during elbow ex-
tension in the ULNT1 position revealed a complex interaction
in which the muscle activity may be involved to protect the
nervous system from tensile forces. Before movement onset,
the EMG activity of shoulder and arm muscles reflected re-
laxed muscles. Mean EMG activity after pain onset increased
in all experimental muscles and in the triceps brachii. This
indicated that submaximal pain and/or tensioning of contin-
uous structures, such as the median nerve, elevated the EMG
activity in the experimental muscles.
 S. Jaberzadeh et al. / Physiotherapy 91 (2005) 94–100 97

Fig. 2. Resistive torque and EMG responses in the ULNT1 position in one representative subject.
98 S. Jaberzadeh et al. / Physiotherapy 91 (2005) 94–100

Fig. 3. EMG group data (mean ± S.E.) in muscles during passive elbow extension in the ULNT1 and neutral positions. b, a, m, o, p and l stand for before, after,
movement, onset, pain and limit, respectively. Significant differences (P < 0.05) are indicated by asterisks.

In the neutral position, during passive elbow extension,
elbow flexor resistive torque and EMG activity of the experi-
mental muscles remained unchanged, whereas in the ULNT1
position, there were significant increases at pain onset and
pain limit. This suggests that continuous structures, such
as the median nerve and/or the vascular system and mus-
cular fascias [22], may be more mechanically sensitive in
the ULNT1 position. This sensitivity is likely to predispose
recruitment of muscles that are responsible for the antalgic
posture of the shoulder and arm.
Elbow extension in the neutral and ULNT1 positions was
performed very slowly (3◦ /second) to prevent reflex activa-
tion of the elbow flexor muscles. Therefore, through-range or
end-range resistance during passive elbow extension in the
ULNT1 position may be a combination of visco-elastic prop-
erties of lengthening tissues such as elbow flexor muscles, the
median nerve trunk [8–10] or the subclavian artery [23].
The results support a relationship between elevated per-
ception of pain and motor responses such as elbow flexor
resistive torque and EMG activity of shoulder and arm mus-
cles. EMG activity increased prominently in the experimental
muscles, which are responsible for the antalgic posture of the
shoulder and arm. The cause of the through-range or end-
range increased resistive torque and muscle activity during
ULNT1 may be a withdrawal response to pain that acts to
indirectly protect the nerve by preventing further tensioning,
in agreement with Hall et al. [24] and Elvey [15]. The lack
of significant changes in EMG activity in the majority of an-
tagonistic control muscles during the experimental task also
supports the concept of a withdrawal reflex. This can be ex-
plained by reciprocal inhibition, which allows the agonists,
the experimental muscles, to act unopposed. Therefore, even
where some mechanical limitation of neural tissue extensibil-
ity and mobility might be possible, reflex muscle contraction
limits range of movement [23,25–28].
The pattern of EMG activity in the triceps muscle was
different to the other control muscles. This muscle showed a
slight increase in EMG activity, especially at pain limit. This
 S. Jaberzadeh et al. / Physiotherapy 91 (2005) 94–100
finding may be explained by the occurrence of an inverse
myotatic reflex. In this reflex, with increasing elbow exten-
sion range, the resistance in elbow flexor muscles increases,
stimulating the golgi tendon organs. In the spinal cord, golgi
tendon organ input via Ib afferent fibres is transmitted to in-
terneurones that turn off the elbow flexor muscles and turn on
the triceps muscle. This may serve to increase the mechanical
stiffness [29] and may provide a protective function for the
elbow joint [30].
In the experimental muscles, the highest and lowest EMG
changes occurred in the upper fibres of trapezius and flexor
carpi radialis, respectively. Sunderland [5,31] reported that
shoulder girdle depression produced greater strain on the
C4–C7 nerve roots and roots of the brachial plexus com-
pared with the strain produced on these nerves by shoul-
der abduction, external rotation or elbow extension. Since
shoulder depression is a key movement to increase the
anatomic course traversed by the brachial plexus, the up-
per fibres of trapezius work to elevate the shoulder dur-
ing elbow extension in the ULNT1 position. This action
may decrease the anatomic course of the median nerve
and reduce tensile stress on the affected nerve trunk and
roots.
An explanation for low EMG activity of the flexor carpi
radialis muscle may be that, unlike other experimental mus-
cles which are generally prime movers of their designated
movements, this muscle is not a prime mover for wrist flex-
ion. Full wrist extension increases tension significantly in the
middle and distal part of the median nerve [8,9]. Therefore,
it seems that wrist flexion has less effect on relieving stress
in the proximal part of the median nerve and affected nerve
root during the test.
The brachialis muscles showed greater EMG activity than
biceps brachii during passive elbow extension in the ULNT1
position. Basmajian and Latif [32] determined that brachialis
is the primary flexor of the elbow. Compared with biceps,
which can potentially act as a shoulder/elbow flexor and fore-
arm supinator, the brachialis muscle is a pure elbow flexor
regardless of forearm position. Second, brachialis is a one-
joint muscle and as such is usually recruited first [33]. Third,
brachialis is generally active during most elbow joint move-
ments and postures [33], whereas biceps is more active during
quick elbow flexion with an added load [30].
The relative increase of EMG activity from movement
onset to 3 seconds before pain onset suggests that the
increase in EMG activity of the shoulder and arm muscles
during elbow extension at the ULNT1 position is not solely
the result of pain. This finding is of considerable importance
since it suggests that a number of mechanisms may contribute
to the EMG responses of muscles during elbow extension in
the ULNT1 position. Other mechanisms linking the muscle
activity with neural tissue tension include mechanical
sensitivity of ganglion cell bodies [34], and existence of
mechanoreceptors in the thoracolumbar spinal cord [35] and
peripheral nervous system [36]. Parke and Whalen [35] and
Theophilidis and Kiartzis [36] contend that the somatic ner-
99
vous system has normal mechanisms that protect the nerve
against tensile stresses. Therefore, muscle-pain interaction
via the nociceptive-mediated flexor withdrawal reflex may
not be the only mechanism responsible for EMG activity of
shoulder and arm muscles during passive elbow extension in
the ULNT1 position. This finding supports Balster and Jull
[28] who concluded that mechanosensitivity of peripheral
nerves in asymptomatic subjects is not solely mediated by
pain.
The findings of this study have implications beyond neu-
rodynamic testing of the upper limb. They provide a better
understanding of the underlying mechanisms for pain and its
related muscular activity as two key parameters during assess-
ment and treatment of neural structures. Understanding of
these mechanisms during neural tensioning manoeuvres has
implications for the management of patients with neck and
upper limb pain and dysfunctions such as peripheral nerve or
brachial plexus injury, whiplash injury and cervical spondy-
losis. These findings do not support using techniques, such
as nerve (and muscle) stretching, which only address the me-
chanics of the nervous system. Techniques that respect the
mechanosensitive state of the nervous system would be more
appropriate. Mobilisation of peripheral nerves by gentle slid-
ing techniques [37] short of reflex muscle contraction [1]
could be advocated. These gentler techniques may prevent
re-inforcement of pathophysiological changes and responses
in the central nervous system processing in symptomatic sub-
jects.
Key messages
• The increase in mechanosensitivity of the median
nerve in the ULNT1 position supports the premise
that neural tissues are tensioned during ‘neural ten-
sion’ tests. Increased muscle activity evoked during
the ULNT1 may be a flexor withdrawal response to
pain that acts to indirectly protect the median nerve by
preventing further tensioning via an antalgic posture.
• Activation of muscles prior to pain onset may indi-
cate a peripheral nerve mechanoreceptor response to
preferential stretching of the median nerve during the
ULNT1.
Ethical approval: Human research ethics committee,
University of South Australia.
Funding: Faculty of Health and Biomechanical Sciences,
University of South Australia.
References
[1] Elvey RL. Treatment of arm pain associated with abnormal brachial
plexus tension. Aust J Physiother 1986;32:225–30.
100 S. Jaberzadeh et al. / Physiotherapy 91 (2005) 94–100
[2] Kenneally M, Rubenach H, Elvey R. The upper limb tension test.
The straight leg raise test of the arm. In: Grant ER, editor. Physical
therapy of the cervical and thoracic spine. Melbourne: Churchill
Livingstone; 1988. p. 167–94.
[3] Butler DS. Mobilisation of nervous system. 1st ed. Edinburgh:
Churchill Livingstone; 1991.
[4] Butler DS. The sensitive nervous system. Adelaide: Noigroup Pub-
lications; 2000.
[5] Sunderland S. Nerves and nerve injuries. Edinburgh: Churchill Liv-
ingstone; 1978.
[6] McLellan D, Swash M. Longitudinal sliding of the median nerve
during movements of the upper limb. J Neurol Neurosurg Psychiatr
1976;39:556–70.
[7] Elvey RL. Brachial plexus tension tests and the pathoanatomical
origin of arm pain. In: Idczak R, editor. Aspects of manipulative
therapy. Melbourne: Lincoln Institute of Health Sciences; 1979.
[8] Kleinrensink G, Stoeckart R, Vleeming A, Snijders C, Mulder P. Me-
chanical tension in the medial nerve. The effects of joint positions.
Clin Biomech 1995;10:240–4.
[9] Kleinrensink G, Stoeckart R, Vleeming A, Snijders C, Mulder P,
Van Wingerden J. Peripheral nerve tension due to joint motion. A
comparison between embalmed and unembalmed human bodies. Clin
Biomech 1995;10:235–9.
[10] Kleinrensink GJ, Stoeckart R, Mulder PGH, et al. Upper limb tension
tests as tools in the diagnosis of nerve and plexus lesions: anatomical
and biomechanical aspects. Clin Biomech 2000;15:9–14.
[11] Byl C, Puttlitz C, Byl N, Lotz J, Topp K. Strain in the median
and ulnar nerves during upper-extremity positioning. J Hand Surg
2002;27:1032–40.
[12] Shacklock MO. Neurodynamics. Physiotherapy 1995;81:9–16.
[13] Jaberzadeh S, Nazeran H, Scutter S, Warden-Flood A. An integrated
AMLAB® -based system for acquisition, processing and analysis of
evoked EMG and mechanical responses of upper limb muscles. Aust
Phys Eng Sci Med 2003;26:70–8.
[14] Perotto AO. Anatomical guide for the electromyographer: the limb
and trunk. Springfield: Charles C Thomas Publisher; 1994.
[15] Elvey RL. Physical evaluation of the peripheral nervous sys-
tem in disorders of pain and dysfunction. J Hand Ther 1997;10:
122–9.
[16] Turker KS, Miles TS, Lee TH. The lip clip: a simple, low-electrical
resistance ground electrode for use in human electrophysiology.
Brain Res Bull 1988;21:139–41.
[17] Birmingham TB, Chesworth BM, Hartsell HD, Stevenson AL,
Lapenskie GL, Vandervoort AA. Peak passive resitive torque at max-
imum inversion range of motion in subjects with recurrent ankle
inversion sprain. J Orthop Sports Phys Ther 1997;25:342–8.
[18] Edgar D, Jull G, Sutton S. Relationship between upper trapezius
muscle length and upper quadrant neural tissue extensibility. Aust J
Physiother 1994;40:99–103.
[19] Maitland G, Hengeveld E, Banks K, English K. Maitland vertebral
manipulation. 6th ed. Oxford: Butterworth Heinemann; 2001.
[20] Millesi H. The nerve gap. Theory and clinical practice. Hand Clin
1986;2:651–63.
[21] Coveney B, Trott P, Grimmer KA, Bell A. The upper limb ten-
sion test in a group of subjects with a clinical presentation of
carpal tunnel syndrome. In: Proceedings of the 10th Biennial Con-
ference of Manipulative Physiotherapists Association of Australia,
1997.
[22] Ginn K. An investigation of tension development in upper limb
soft tissues during the upper limb tension test. In: Proceedings of
the International Federation of Orthopaedic Manipulative Therapists
Congress, 1989.
[23] Wilson S, Selvaratnam P, Briggs C. Strain at the subclavian
artery during the upper limb tension tests. Aust J Physiother
1994;40:243–8.
[24] Hall TM, Pyne EA, Hamer P. Limiting factors of the straight leg
raise test. In: Proceedings of the 8th Biennial Manipulative Physio-
therapists Association of Australia Conference, 1993.
[25] Goeken LN, Hof L. Instrumental straight leg raising: results in
healthy subjects. Arch Phys Med Rehabil 1993;74:194–203.
[26] Wright A, Thurnwald P, O’Callaghan J, Smith J, Vicenzino
B. Hyperalgesia in tennis elbow patients. J Musculoskelet Pain
1994;2:83–97.
[27] Hall TM, Zusman M, Elvey RL. Manually detected impediments
during the straight leg raise test. In: Proceedings of the 9th Biennial
MTAA Conference, 1995. p. 48–53.
[28] Balster SM, Jull GA. Upper trapezius muscle activity during the
brachial plexus tension test in asymptomatic subjects. Man Ther
1997;2:144–9.
[29] Rothwell JC. Control of human voluntary movement. 2nd ed. Lon-
don: Chapman and Hall; 1994.
[30] Basmajian JV, De Luca CJ. Muscle alive: their function revealed by
electromyography. 5th ed. Baltimore: William and Wilkins; 1985.
[31] Sunderland S. The anatomy and physiology of nerve injury. Muscle
Nerve 1990;13:771–84.
[32] Basmajian JV, Latif A. Integrated actions and functions of the chief
flexors of the elbow: a detailed electromyographic analysis. J Bone
Joint Surg 1957;39A:1106–18.
[33] Smith LK, Weiss EL, Lehmkuhl LD. Brunnstrom’s clinical kinesi-
ology. 5th ed. Philadelphia: F.A. Davis Company; 1996.
[34] Devor M, Rappaport ZH. Pain and the pathophysiology of damaged
nerves. In: Fields H, editor. Pain syndromes in neurology. London:
Butterworth; 1990.
[35] Parke WW, Whalen JL. The pial ligaments of the anterior spinal
artery and their stretch receptors. Spine 1993;18:1542–9.
[36] Theophilidis G, Kiartzis K. Stretching of an isolated paret of
the sciatic nerve of the frog (Ranarindibunda) causes excita-
tion of tis sensory nerve fibres. Neurosci Lett 1996;215:99–
102.
[37] Elvey RL, Hall T. Neural tissue evaluation and treatment. In: Do-
natelli RA, editor. Physical therapy of shoulder. 3rd ed. New York:
Churchill Livingstone; 1997.