Associated movements are normal
automatic postural adjustments which accompany
voluntary movements (1). By contrast, associated
reactions are abnormal reflex activities which
may occur in the absence of voluntary movements.
The incidence of these involuntary movements in
hemiplegic stroke patients and patients
interpretations of the phenomenon have not been
documented.
Associated reactions have been said to occur
when the uninvolved limbs are moved forcibly
(for example flexion of the normal arm
against resistance may produce a flexor response
in the hemiplegic arm (3) and in response to
artificial reflex stimulation. These two types
of stimuli (5) have been extensively studied and
form the basis of the Brunnstrom school of
physiotherapy (2) which aims to capitalise on
these movements in order to improve the mobility
and function of the affected arm. However,
little attention has been paid to the
involuntary reactions occurring in association
with quasi-automatic movements such as yawning,
sneezing, coughing and stretching. As the
hemiplepic patient begins to Yawn his
paralysed arm may flex involuntarily at the
elbow, the arm is drawn up in front of him and
it returns to its resting position as the yawn
ceases. During this movement the hand and
fingers are said to be held in extension (6,
9,10). Many of our stroke patients had these
movements. As they differed in pattern from
previous descriptions and as many patients had
false ideas about their nature and implications,
I decided to investigate the phenomenon.
Patients and methods : Forly
hemiplegic stroke patients who were attending
the out-patient physiotherapy department at
Nottingham General Hospital were studied. These
patients had been previousiy admitted under the
care of consultant physicians who had made a
diagnosis of stroke on clinical grounds. They
constituted all the stroke patients who were
regularly attending the physiotherapy
department. Each patient was questioned using a
standard questionnaire. Those patients who had
experienced associated reactions were asked to
describe the pattern of movement and to identify
activities which provoked the movement. They
were also asked about their interpretation of
the phenornenon. The patient's age and sex,
handedness, the side of the stroke and the
interval between stroke and onset of movements
were noted. The degree of arm movement was
gauged on a-3-point scale (no. minimal.. and
considerable arm movement) and a similar scale
was used for manipulative function (no. some and
very useful function of the hand).
Results : The ages of the patients
ranged between 40 and 82 vears. Four patients
had a moderate or severe speech defect: 3 of
these could answer 'Yes' or'No' appropriately,
and the fourth could not. The remainder of the
patients could give detailed answers to all
questions. Thirty-two patients had experienced
automatic movements of the hemiplegic arm. In 15
of them, the movements had begun within a month
of the stroke but in 4 a yaer or more had
elapsed between the stroke and the onset of
movements. In most of the patients. the movement
started within 6 months of the stroke.
Pattern of movement : In 28 of the 32
patients with associated reactions the elbow was
flexed during the movement; in 4 the elbow
became extended. Only 9 patients had the type of
movement hitherto described in which flexion was
accompanied by extension of the fingers. In 7
patients, the fist was clenched during each
episode. Of the remainder, one bad a fixed
flexion contracture of the hand: there was no
hand movement in 5 and 6 were not sure about the
presence or pattern of hand movement. ln 2
patients, the movement of the arm was so violent
that the hand hit their face.
Movements of the affected leg accompanying
automatic arm movements were mentioned
spontaneously by 8 patients. In 7 of these, the
leg only moved in concert with the arm when the
patient was lying in bed. In the eighth, the
knee flexed without any arm movement whenever
the patient coughed; however, when he yawned
only the elbow flexed. In 4 patients the knee
extended, in 3 it flexed and in one it moved
laterally.
Side of hemiplegia and associated
reactions : Twenty-three of the patients
studied had a leftsided hemiplegia, in 17 the
right side was affected. There was no
significant difference between the side of
paralysis and prevalence of involuntary
movements: of the 8 patients who did not have
associated reactions, 3 had sustained a left
hemiplegia, 5 a right hemiplegia. Of the 32
patients with associated reactions, 4 extended
their elbows whilst yawning. All 4 had a
left-sided hemiplegia. The other 28 flexed the
elbow: 16 of these had paralysis of the left
side of the body, 12 had right-sided paralysis.
There were no significant correlations between
finger flexion and extension and stroke
laterality.
Activities producing associated reactions
: In 31 patients, the arm moved
involuntarily in association with yawning.
The arm movement would begin at the onset of the
yawn and subsided as the yawn ended. Fifteen
patients had arm movements every time they
yawned. Several piatients found that the degree
of arm movement was proportional to the size of
the yawn. Others noted that yawn-induced arm
movements were more likely to occur when they
were lying in bed. In some cases, the arm fell
to the side quite abruptly at the end of the
yawn. One patient had to hold his paralysed arm
when he yawned to prevent wrenching of the
shoulder by sudden downward arm movement. Other
activities which triggered automatic movements
were sneezing (10 patients), coughing (8),
stretching (3), and laughing (1). In 4 patients
(including the only person in whom these
movements were not stimulated by yawning) the
associated reactions occurred whilst the patient
was lying in bed, in the apparent absence of
truncal movement.
Patients' interpretations of association
reactions : Twenty-three patients thought
that these movements were a sign of recovery. In
several patients the onset of associated
reactions engendered false optimism, they felt
that the arm was coming back to life and that
useful function would soon return. Four patients
thought that the involuntary arm movements were
a direct result of physiotherapy. One patient,
who was not referred to the physiotherapy
department until 17 months after a dense
left-sided hemiplegia had not had any abnormal
movements prior to treatment. Shortly after
starting therapy he began to have associated
reactions whenever he couched or yawned. Not
everyone greeted these movements with
enthusiasm. One man thought that they heralded
the onset of another stroke: another was
concerned that they might lead to flexion
contractures. Seven patients felt that the
movements were of no significance. One of these
had become weary of comments by therapists and
family about incipient improvement whenever the
arm moved automatically.
Patients with no associated reactions
: Eight patients had never had automatic
movements. ln all cases, at least 3 months had
elapsed between the date of the stroke and the
interview with the patient. A striking feature
of this group was the good movement and function
in the hemiplegic arm and hand: 7 had
coniderable voluntary arm movement and one had
minimal movement: 4 had very useful hand
function, and 3 had some use in the arm. This
contrasts with the arm movement and function in
those patients with associated reactions. No
other significant differences between the two
groups were found.
DISCUSSION Involuntary movements of
the affected arm are common in hemiplegia.
occurring in 80% of the patients studied. This
study does not cover an entirely representative
group of stroke patients (35 out of the 40 were
men, and all were weil enough to attend
outpatient physiotherapy), but it appears that
associated reactions are much more common than
would be expected from the brief descriptions
found in most contemporary neurological texis.
We do not know why involuntary arm movements
occur when hemiplegic patients yawn. One
attractive hypothesis (8) suggests an
evolutionary basis. Yawning may he considered as
a lower type of respiration than ordinary
respiration.Originally, respiratory movements in
animals were associated with limb movements. As
the limbs began to perform more specialized
movements, and respiratory movements became
increasingly automatic, limb and respiratory
movements became dissociated with the
development of inhibitory influences. After a
stroke, these inhibitory influences are removed
and the arm tends to resume its old association
with respiration.
Stejskal (6, 7) has studied the influence of
respiration on reflex neuromuscular activity. As
a general rule. inspiration stimulates
neuromuscular excitability, expiration promotes
relaxation. In spastic patients, there is an
increase in the activity of striated skeletal
muscles during inspiration. This activity
reaches a peak during the Valsalva manoeuvre,
when the patient inhales deeply and holds his
breath. When a hemiplegic patient performs the
Valsalva manoeuvre, certain muscles contract in
a stereotyped way: the arm becomes flexed,
adducted and pronated, with fingers flexed; the
leg assumes the extended posture with the foot
in plantarflexion and equinovarus. This
activation of antigravity muscle groups during
inspiration is greater if the patient is in the
supine rather than the prone position. Stejskai
suggests that in forced expiration
(Müller's manoeuvre) hemiplegic patients
very often spontaneously extend the fingers and
the elbow in the spastic upper limb. He states
that yawning is the most natural form of this
strong expiration and that the hemiplegic
patients who yawn have associated extension of
the affected arm.
De Jong (4) however states that yawning is a
prolonged inspiration. It is probable that
people yawn in different ways but a common
pattern comprises a prolonged inspiratory phase;
a phase when the breath is held. and an
expiratory phase. The patients whom I studied
noticed that the paretic arm began to move at
the onset of yawning; during the expiratory
phase the flexed arm would extend and return to
its resting position. Yawning cannot therefore
be considered to be simply prolonged inspiration
or forced expiration. Although yawning has been
considered to be of minor medical importance, it
is clearly not without interest and merits much
more detailed study.
There was no relationship between sidedness
of the stroke and prevalence of associated
reactions. However, all 4 patients who reported
elbow extension during yawning had left-sided
paralysis. No conclusion can be drawn from such
small numbers and larger-scale studies are
needed to determine whether there is indeed a
relationship between pattern of associated
reaction and the side of the stroke.
None of the patients with associated
reactions had received a satisfactory
explanation for these movements from their
doctor or physiotherapist and most patients had
mistaken notions about the movements. Many
hemiplegic stroke patients fall to regain useful
manipulative function in the paralysed arm and
the development of involuntary arm movements
frequently causes unwarranted optimism in the
patient and his or her family. The doctor
involved in the care of stroke patients must be
aware of the prevalence and implications of
these movements and must resist the temptation
to attach prognostic importance to them, since
this will lead to ultimate disappointment. It is
suggested that doctors should routinely ask
stroke patients about the presence of these
movements so that simple explanation and
reassurance can be given.
The commonest patterns of movement differed
from those previously described. Walshe (9, 10)
describes extension of the hands and digits
during the movement. In Walshe's patients the
extension was complete and was accompanied by
fanning of the fingers. In the present study
only 9 patients had extension of the fingers.
The reason for these differing patterns is
obscure but may be related to differences in
physiotherapy technique or to individual
variation in yawning patterns. This study shows
that the presence of voluntary arm control does
not exclude the appearance of typical associated
reactions: involuntary arm movements occurred in
many patients with useful hand function and
considerable arm movement. However, of the 8
patients who denied having automatic movements,
7 had considerable arm movement and useful
manipulative function. Fur ther work is
necessary to establish whether the absence of
these movements is a useful prognostic index in
determining arm recovery in hemiplegia.
REFERENCES
