Neuroscience and Biobehavioral Reviews, Vol. 22, No. 4, pp.

571578, 1998
1998 Published by Elsevier Science Ltd. All rights reserved
Printed in Great Britain
0149-7634/98 $32.00 + .00

Pergamon

PII: S0149-7634(97)00046-8

Spasticity and Spastic Gait in Children with

Cerebral Palsy
PAOLO CRENNA1
Institute of Human Physiology, Faculty of Medicine, University of Genoa, Vle. Benedetto XV, I-16132 Genoa, Italy

CRENNA, P. Spasticity and spastic gait in children with cerebral palsy. NEUROSCI BIOBEHAV REV 22(4)571578, 1998.The
current notion of spasticity as a velocity-dependent increase of muscle response to imposed stretch was mainly derived from studies
performed under stationary experimental conditions. To address the issue of a spastic muscle behaviour under dynamic conditions, we
conceived a novel approach, aimed at quantitatively assessing motor output over the lengthening periods which take place during
unperturbed functional movements. Application to the analysis of overground walking in children with spastic cerebral palsy (CP)
revealed that, for representative lower limb muscles, the relationship between EMG levels and estimated muscle lengthening rate
displays either increased gain or reduced velocity threshold. Parallel measurement of gait kinetics frequently showed congruent increase
of the mechanical resistance to joint rotation. Abnormalities preferentially targeted the lengthening contractions occurring around the
ground contact period of the stride. The pathophysiological profile of what is clinically defined as spastic gait in CP children did not
only consist of dynamic spasticity, as described above. Most often it resulted from the simultaneous contribution of other factors,
including paresis, co-contraction, immature and non-neural components. 1998 Published by Elsevier Science Ltd. All rights reserved.
Spasticity

Locomotion

Muscle kinematics

Joint mechanics

Comparative and follow up studies have also shown that

tendon jerk briskness and tonic responses to stretch are not
well correlated in magnitude, topography and evolution in
time after brain injury (24,29,31), which suggests that the
two signs are not necessarily an expression of the same
pathophysiological phenomenon.
Proposed spinal mechanisms accounting for disturbed
stretch reflex transmission in spastic patients include
reduced effectiveness of presynaptic inhibition acting on
spindle primary afferents (e.g. (23,25,30)) and reduced or
absent autogenic inhibition on alpha motoneurons (MNs)
(26). Reciprocal inhibition from Ia fibres appeared either
depressed (2,24,63) or enhanced (3,9), whereas indirect
testing of the fusimotor system failed to reveal signs of
hyperactivity (37), although further analysis is needed.

THE CURRENT CONCEPT OF SPASTICITY

IN THE history of neurological semeiology, the word spasticity has been used with various meanings, most often to
indicate excess of muscular action. In 1979, the participants
of an international workshop convened in Scottsdale, Arizona, agreed upon a narrow consensus notion, which is
widely referred to in the current topical literature. Accordingly, spasticity is a motor disorder characterised by (i)
exaggerated tendon jerks (hyper-reflexia) and (ii) increased
muscle response to applied stretch, positively correlated
with the lengthening rate (velocity-dependent hypertonia)
(43). Dysfunction is ascribed to hyperexcitability of the
myotatic reflex arc(s).
Spastic phenomena, as defined above, were quantitatively
investigated in humans by means of a classic experimental
paradigm, whereby EMG activity of the tested muscle is
recorded during passive lengthening produced at various
rates
by
controlled
mechanical
perturbations
(10,13,46,48,62). This approach demonstrated that in clinically-spastic adult patients the relationship between the
amplitude of the reflex response and the velocity of the
lengthening stimulus displays either an increased gain
(36,59) or a reduced velocity threshold (48,56). Detailed
analysis of stretch-induced responses revealed the contribution of short-latency, phasic components (related to
clinical hyperreflexia) and late, sustained activity (preferentially connected with clinical hypertonia) (41,46,62).
1

Cerebral palsy

LIMITS OF THE STATIC NOTION

As pointed out by Lance in 1990 (44), the consensus

notion of spasticity does not include impaired movement
and an abnormal posture. Rather, it applies to observations
obtained under static experimental conditions in which
muscles are relaxed or tonically activated. Such a conceptual limitation might probably reflect the reported poor
correlation between pathological responsiveness to stretch
measured at rest and motor deficits assessed during natural
actions (17,39,60). It might also be related to the evidence
that amelioration of hyperactive reflexes induced by drugs
or physical therapy does not necessarily lead to improved
functional movements (15,45,51).

Tel.: 39-10-3538183; fax: 39-10-3538194.

571

572

CRENNA

Differences in the mechanical context of the muscle

tendon system during movements or during their absence
might partly explain the discrepancy between observations
in static and dynamic conditions. Changes in the lever arm
and in the position within the forcelengthvelocity planes
which correspond to different levels of activity attained in
the two conditions would obviously influence the effectiveness of muscles in generating joint torques, and consequently their actual contribution to the motor performance
(55). Changes in the central set might also account for the
predictive inadequacy of static analysis. In fact, the excitability of relevant systems involved in segmental reflex
transmission (alpha MNs, dynamic and static gamma
MNs, presynaptic inhibition on primary afferents) is
known to differ significantly when tested under passive
conditions and during intentional movements (e.g. (57)).
Consistent modulation has also been described depending
on the motor task (12,50,53,61) and even on the phase of a
given (rhythmic) activity (16,18,35). In addition, the afferent input will also change with the above conditions. Hence,
it is not surprising that stationary measurements can only in
part detect abnormalities which will be possibly expressed
during voluntary movements, specific classes of movement
or selected phases of a movement.
The arguments discussed above emphasise the difficulty
of extrapolating and generalising observations obtained
under simplified experimental conditions. Indeed, for the
concept of spasticity to be endowed with functional and
operational significance, pathological muscle activation
upon lengthening should be objectively confirmed, quantitatively assessed and topographically characterised over a
wider spectrum of motor tasks, and particularly during daily
natural movements.
RATIONALE FOR A NEW DYNAMIC APPROACH

Probably because of methodological limits, motor output

during stretching has rarely been investigated under
dynamic conditions, the few available reports mainly dealing with locomotor function. Rapid stretching of calf muscles during treadmill walking was obtained by mechanical
devices fixed to the subjects leg (1,49), or indirectly by
abrupt changes in the belt speed (6). By means of the latter
technique, enhanced short-latency responses, ascribed to
myotatic reflexes, were recorded in spastic patients (7).
Unfortunately, the wide use of these methods and their
transfer to the clinical setting is thwarted by significant
drawbacks, such as major interference with movement
execution and applicability to only one or two muscle
groups, most often over limited time windows during movement. Difficulties become even greater for paediatric
patients.
In the early 1990s, owing also to the availability of
improved technologies for non-invasive analysis of human
motion, we readdressed the question of spastic muscle behaviour during natural movements. However, we sought to
adopt a new strategy, alternative to the one based on externally-applied mechanical disturbances. The central point of
the novel approach was a quantitative study of muscle activity during the lengthening phases which take place during
unperturbed motor actions. The rationale for this choice
comes from recording muscle afferents in awake behaving
animals, and from indirect evidence in humans, indicating

that self-induced lengthening phases are accompanied by

augmented inflow from muscle spindles, which are sometimes higher than during passive stretching of similar amplitude and velocity (11,58). Such an afferent barrage could
potentially be responsible for the activation of the homonymous MNs, depending on the current excitability state of
the stretch reflex system. As a consequence, when velocity
sensitivity of the stretch reflex is pathologically up-regulated (which is supposed to occur in spastic subjects), the
lengthening periods, and particularly those exhibiting peak
lengthening rates, become selectively prone to the interference of autogenetic excitation.
Albeit limited, a number of observations in humans
apparently support the above contention. Passive lengthening of limb muscles produced during voluntary contraction
of their antagonist, in the absence of any external perturbations, was shown to be hindered by abnormal EMG recruitment in spastic patients, both in the upper and lower limb
segments ((14), cerebral palsy (CP); (52), adult hemiparesis). With reference to rhythmic locomotor movements,
stretching of calf muscles in the post-contact stride period is
known to be frequently accompanied by synchronous bursts
in triceps surae, both in spastic adults and in CP children
(8,40,47). Again, the phases of the pedalling cycle in which
leg muscles are supposed to be stretched were reported to be
parallelled by increased muscle activation in spastic paraparetics (5).
Although no attempts were made at quantitatively estimating muscle kinematics, the above findings suggest that
the phases of a voluntary movement in which muscles are
being lengthened might indeed by susceptible to the interference of enhanced motor output in spastic individuals. In
particular, for a given muscle and a given motor task, over
the lengthening phases characterised by EMG silence during normal movement execution, the excessive dynamic
sensitivity to stretch would become apparent as an extraactivity, time-locked to the peak lengthening rate. Instead,
over the lengthening periods normally marked by motor unit
recruitment (lengthening contractions), enhanced reflex
contribution would be integrated into the basic muscle
activity. In both instances, our working hypothesis was
that the higher the effectiveness of velocity-sensitive excitatory reflex components, the tighter would be the correlation between motor output and velocity of muscle
lengthening.
APPLICATION TO THE ANALYSIS OF GAIT

To verify whether the above concepts might be used for

dynamic assessment of spasticity during walking, a new
processing sequence was devised, aimed at quantitatively
characterising the lengthening phases during the stride
cycle in representative leg muscles, and at correlating
motor output to muscle lengthening rate over each period
(20,21). Muscle kinematics were estimated by geometrical
models (34). On the basis of anthropometric parameters
(measured in the individual subject) and relative joint angles
at the hip, knee and ankle (obtained by a motion analyser)
these algorithms compute the overall length changes of the
muscletendon complex, and, as first derivative with
respect to time, the velocity of length changes. For each
isolated lengthening phase, the correlation between motor
output and muscle kinematics is then determined by plotting

GAIT ANALYSIS IN SPASTICITY

the level of current EMG activity (rectified, integrated

and normalised to the maximum amplitude), as a function
of the muscle lengthening velocity which is measured
simultaneously.
PATHOLOGICAL PATTERNS IN CHILDREN WITH CEREBRAL
PALSY

The upper panel in Fig. 1 is a look-up chart of the lengthening phases occurring in representative lower limb muscles

573

(quadriceps (Quad), medial hamstrings (mHam), soleus

(Sol) and tibialis anterior (T. Ant.)), during walking at natural speed in ten healthy children aged 510 years. It
appears that Quad is lengthened twice over the stride
cycle: in the early stance phase, during knee yielding
associated with body weight loading (L1 Quad), and during
the toe off flexion synergy (L2 Quad). Hamstrings exhibit one
lengthening period in the midlate swing phase, concurrent
with hip flexion and knee extension (L Ham). Calf muscles
undergo two lengthening phases: in the stance phase, along
with the forward rotation of the shank (L1 Sol), and in swing,
during the foot-clearance ankle dorsiflexion (L2 Sol). The
pretibial group is lengthened at touch down during ankle
plantar flexion (L1 TAnt) and again around the stance-toswing transition (L2 TAnt).
Analysis of ten age-matched diplegic children walking
unsupported at a speed largely overlapping with control
values, revealed that the main lengthening phases are most
often retained, despite slight changes in timing (e.g. anticipation of L1 Sol). However, changes in the global gait pattern
(e.g. foot-flat or forefoot ground contact) led to the disappearance of specific lengthening periods (e.g. L1 Quad and
L1 TAnt), and/or appearance of new non-physiological ones
(e.g. Ham in post contact phase, Lextra Ham) (Fig. 1(lower)).
In keeping with the working hypothesis, correlating
motor output to muscle kinematics disclosed distinct
abnormalities in the CP children, which can be summarised
in two main patterns. The first pattern consisted of increased
gain in the relationship between EMG and lengthening velocity. This was typically observed in the calf muscles during
early stance (L1 Sol), where the increase of EMG output per
unitary muscle lengthening velocity was significantly higher
than in normal controls (Fig. 2(upper)). The second pattern
of disturbance was characterised by reduced lengthening
velocity threshold for motor unit recruitment. A similar
phenomenon occurred in posterior thigh muscles in the
second half of the swing phase (L Ham), when the locomotor
activity braking forward leg swing was turned on at abnormally low lengthening rates (Fig. 2(lower)).
ASSOCIATED CHANGES IN JOINT MECHANICS

FIG. 1. Look-up charts of the stride periods characterised by absolute

lengthening of the muscletendon complex in four lower limb muscles
of normal and diplegic children walking at comparable speed (range 85
105% height/sec). Each box represents the average stride time (with SDs)
occupied by a single lengthening phase. Vertical bars across the boxes mark
the mean times of the peak lengthening velocities. Upper panel: reference
data from ten normal children, 510 years of age. Lower panel: data from
ten age-matched diplegic children. Filled boxes indicate lengthening phases
homologous to the normal templates (reported underneath as dashed lines);
open boxes indicate lengthening periods which are frequently lacking in
diplegic children; the hatched box (Lextra mHam) is a lengthening period
never observed in normal controls. Time is normalised as a percentage of
the stride cycle (interval between two consecutive ground contacts of the
same foot). Quad, quadriceps (vastus medialis); mHam, medial hamstrings
(semimembranosussemitendinosus complex); Sol, soleus; T.Ant., tibialis
anterior.

To obtain clues about the possible mechanical effects of

the described recruitment abnormalities, we sought to monitor the mechanical properties of the joints over the critical
lengthening phases. Accordingly, an ad hoc method was
devised, based on the analysis of the relationship between
external joint moments and relative joint angles, computed
in the sagittal plane, over single cycles (21,32,33). In the
resulting momentangle diagram, which can be related to
the forcelength graph characterising the stiffness of a
spring, the slope of the curve can be used to infer the
resistance offered by the joint during rotation.
In children exhibiting enhanced EMG recruitment gain
and in those with reduced velocity threshold, our findings
demonstrated that the momentangle relationship was most
often marked by congruent increase in the slope during the
relevant time window. This suggests that abnormalities in
matching motor output and muscle kinematics in the tested
muscles might yield significant mechanical effects,
although the overall joint resistance to rotation will be
obviously affected by activity in all muscles crossing the
joint and by passive properties of tissues (22,32).

574

CRENNA

FIG. 2. Patterns of correlation between EMG levels and muscle lengthening velocity during walking in normal and spastic diplegic children. Upper graphs:
soleus muscle tested during the stance phase lengthening period (L1 Sol). Four consecutive strides are represented for a normal child aged 7 years (NOR)
(mean walking speed 130% height/sec) and an age-matched diplegic child (DIP), (mean walking speed 114% height/sec). Loops follow a counterclockwise
path and data are sampled every 20 ms. Note the clearcut increase in the slopes (gain) in the spastic patient. Lower graphs: medial hamstrings tested during
the midlate swing lengthening phase (L mHam) in an 8-year-old normal and an age-matched diplegic child (mean walking speed 65% and 73% height/sec,
respectively). Same conventions as for Sol. Note the reduced velocity threshold for EMG recruitment in the diplegic child, and the shortlasting inversion of
the curve in the rising phase, which indicates a synchronous burst of activity.

Figure 3 reports examples of momentangle curves for

the ankle joint, obtained from one normal (A) and three
spastic diplegic children ((B)(D)). The left diagram of
each pair refers to the loop computed for the entire stride
cycle, while the flaming diagram on the right focuses on
the earlymid stance phase, when the first Sol lengthening
period (L1 sol) takes place. In this second graph, information
about levels of muscle activation during locomotion was
superimposed. In particular, for each point of the curve,
EMG activity was represented as a line with a length proportional to the amplitude of the instantaneous rectified
myoelectric signal and an orientation corresponding to the
mechanical effect of muscle activation (i.e. towards plantarflexion and increased external dorsiflexor moment (left, up)
for calf muscles; towards dorsiflexion and reduced external
dorsiflexor moment (right, down) for the pretibial group).
Analysis of the flaming diagrams reveals that, in the normal child, the slope is low after ground contact and is progressively increased during mid-stance, until maximum foot
dorsiflexion is attained (Fig. 3(A)). Such a figure, which
indicates progressive increase of mechanical resistance to
ankle joint dorsiflexion, is most probably related to the
recruitment in the plantarflexor muscles, as suggested by

the progressively higher levels of activity in the Sol EMG

(flames pointing left and up).
In diplegic children, abnormal high-level activation of calf
muscles in earlymid stance was frequently associated with
global increase in the slope of the curve (Fig. 3(B)), or with
earlier occurrence of the steep mid-stance portion (Fig. 3(C)),
which indicates a pathological increase of the mechanical
resistance offered by the ankle joint to dorsiflexion. Cases
were also observed in which low-level EMG activity, or
even EMG silence in triceps surae during L1 Sol was parallelled
by high-slope momentangle curves (Fig. 3(D)). This pattern
might be accounted for by the enhanced contribution of
muscles which were not recorded (e.g. tibialis posterior).
Alternatively, this same pattern might be explained by non
neural mechanisms, related to changes in the intrinsic mechanical properties of the muscle effectors (20,28). Reduced
excursion at the relevant joint during passive mobilisation
in the relaxed subject would support the latter explanation.
EVIDENCE OF CRITICAL LENGTHENING PERIODS

Systematic mapping of the pathological patterns in a

group of 30 spastic diplegic and hemiplegic children

GAIT ANALYSIS IN SPASTICITY

575

FIG. 3. Momentangle plots about the ankle joint in one normal (A) and three spastic diplegic children ((B)(D)). Each panel reports two graphs: the left-side
loop is computed for the entire stride cycle, while the flaming diagram on the right focuses on the earlymid stance phase, when the first soleus lengthening
period (L1 Sol) takes place. In this second graph, information about levels of locomotor muscle activation was added in the form of lines pointing up and left
(Sol), and down and right (T.Ant.). According to the conventions used (see arrows), the slopes of the curves in the upper right quadrant of each diagram
provide information about the resistance of the ankle joint to dorsiflexion during the corresponding stride phase. Note the higher slopes in spastic diplegic
children. Moments of force in the vertical axes are normalised with respect to body weight (Nm*Nbw) and, for convenience multiplied by 100. Angles in the
horizontal axes are expressed in degrees. Walking speeds: (A) 130% height/sec; (B) 110% height/sec; (C) 71% height/sec; (D) 60% height/sec.

walking unsupported revealed a rather unexpected picture.

In fact, abnormal motor output and the associated mechanical changes were not confined to the muscles exhibiting
maximal lengthening velocities (or excursions). Nor, for a
given muscle, were they limited to the stride periods characterised by peak lengthening rates, as implicitly predicted
by the classical notion of spasticity. In contrast, they particularly occurred during specific stride phases which proved
to be selectively prone to abnormal muscle activation upon
lengthening.
In mildly disabled children, early stance Sol (L1 Sol)
and midlate swing Ham (L Ham) lengthening phases
were hot spots for abnormal recruitment patterns. In a
few instances, Lextra Ham was also affected. In one subject
with marked clinical involvement, mid-swing L2 Sol was
disturbed by synchronous EMG activity, time-locked to
peak lengthening velocity. Interestingly, the Quad was
never hyperactive around toe off (L2 Quad), despite fast
stretching of anterior thigh muscles. This was true also
for spastic children exhibiting clonic knee jerks at bedside examination, indicating that exaggerated responsiveness to stretch, assessed in resting (or tonically active)
muscles is not necessarily carried over during dynamic
function.
From the above results, one can conclude that different
control mechanisms are responsible for the integration of
central locomotor commands and peripheral (stretch reflex)
contributions, during lengthening periods occurring while
walking in CP children.

HYPOTHESES ABOUT NEURAL MECHANISMS

An obvious question now is why some lengthening

phases are preferentially prone to the interference of pathological motor output, whereas others appear to be more
resistant. Clues to understanding the possible mechanisms
come from both animal and human studies, and take into
account the neural systems controlling autogenic excitation
during locomotion.
Pre-synaptic gating on primary afferents was reported to
be active both in animal and human stride cycle, although
most probably with phase-dependent differences in intensity
(4,18,53). Defective pre-synaptic inhibition might play a
role in the abnormal gain observed in calf muscles in early
stance (L1 Sol), as suggested by studies employing applied
perturbations (27) and the Hoffmann reflex (18,19).
Post-synaptic inhibition of alpha MNs was demonstrated
in animals during the phases of inactivity of their rhythmic
locomotor discharge, which most often correspond to a firing burst in the functionally antagonistic motor nuclei (e.g.
(42)). The phylogenetic coherence of the basic locomotor
mechanisms would argue in favour of a similar control
strategy during the homologous periods of the human stride
cycle. In this connection, powerful post-synaptic inhibitory
effects might subserve the strong depression of the H-reflex
observed by several authors in the Sol muscle during the
period of EMG silence, in the swing-phase (12,18,35). Presynaptic gating of Ia input might also be strengthened
during the stride periods characterised by locomotor
muscle inactivity (see (18,65)). It is likely that one of the

576

above mechanisms, or even both, operate during L2 Sol and

L2 Quad, in which muscle lengthening occur in the absence of
motor output. The consequence would be consistent reflex
suppression and safer protection of ankle and knee extensor
motor nuclei against undesired peripheral activation during
foot clearance. This figure would explain the observed
lower susceptibility of the swing phase lengthening periods
to abnormal motor output. Interestingly, in spastic adults
exhibiting pathological calf muscle activation at L2 Sol, Hreflex testing frequently (but not always) revealed marked
depression or disappearance of the typical inhibitory phase
(19,64).
Activation of static fusimotor neurons which reduce
dynamic sensitivity of muscle spindles is an additional
mechanism available for lowering stretch reflex responsiveness during locomotion (38,54,57). It can then be hypothesised that abnormal modulation of the static gamma drive
might also play a role in the disturbed stretch reflex
transmission, particularly during lengthening contractions
(when alpha and gamma MNs are likely to be coactivated)
and in cases characterised by enhanced gain of the EMG
lengthening velocity relationship.
As mentioned above, the activity of several control systems (including pre-synaptic inhibition, post-synaptic inhibition and static gamma drive) undergoes changes in
intensity over the stride cycle during normal locomotion.
Such a modulation is thought to result from central commands and peripheral reafferent signals. It is reasonable to
propose that cerebral lesions in spastic diplegic children
affect supraspinal regulation over these mechanisms in a
non uniform fashion, preferentially disrupting descending
control on the gating systems acting around ground contact
(e.g. during L1 Sol and possibly L Ham, Lextra Ham), and interfering much less with the more robust inhibitory effects
which operate during the flexion phase of the stride (e.g.
during L2 Sol and L2 Quad). As a corollary conclusion, control
mechanisms working in the ground contact phase would be
more dependent on supraspinal regulation than those operating around toe off and early swing.
MULTI-COMPONENT PATHOPHYSIOLOGICAL PROFILE OF THE
SPASTIC GAIT

The adoption of complementary protocols for noninvasive analysis of movement (17,20,21) has shown
that abnormalities of motor output during lengthening
only in a few instances dominate the pathophysiological
picture of what is clinically described as spastic gait in
children with cerebral palsy. Most often, dynamic spasticity
is just one component of a multiple profile, which results
from simultaneous contribution of several additional factors. The latter include deficient recruitment of motor units
(paretic component), unselective activation of functionally
antagonist muscles (co-contraction component), and
changes in the passive muscle properties (non-neural component) (see (20,28)). Our current research is aimed at
quantitatively assessing the relative contribution of each

CRENNA

of these components, to the production of a pathophysiological profile of the individual gait pattern (see (20)). This
is expected to provide a valuable tool for planning patienttailored therapeutic and (re)habilitation procedures.
CONCLUSIONS

Dynamic analysis of the spastic component of spastic

gait, as described above, has several advantages which
include minimal interference with the subjects motor performance, the possibility of testing complex natural movements and the application to virtually any muscle groups
amenable to modelisation. By means of this method, substantial abnormalities in the patterns of muscle activation
upon lengthening have been detected in clinically-spastic
CP children, and congruent mechanical changes were
observed at the relevant joints. These appeared more frequently during lengthening contractions around ground contact, probably because inhibitory mechanisms acting on
velocity-sensitive afferents during such a critical stride period are more heavily dependent on supraspinal control.
However, it is worth stressing that although data are compatible with an enhanced contribution of myotatic reflex
components, analysis was performed on the overall locomotor EMG output. The latter results from the integration of
peripheral contributions and centrally generated motor commands, which cannot be reliably separated. As such, the
correlation between EMG levels and lengthening velocity
is not sufficient to provide evidence for increased dynamic
sensitivity of the stretch reflex component on its own, as
should be required by the classical definition of spasticity.
In this framework, the concept of a dynamic spasticity needs
to be broadened with respect to the current static notion. A
tentative working definition could be: a movement disorder
characterised by increased coupling between the levels of
muscle activation and muscle lengthening velocity (either in
terms of increased gain or reduced threshold), associated
with enhanced mechanical resistance to rotation at the relevant joint(s). The hypothetic underlying mechanism is a
disturbed modulation of velocity-sensitive reflex contribution to the motor output. However, abnormalities in the
central pattern generation should also be taken into account,
particularly when the velocity threshold for locomotor muscle activation appears to be lowered.
ACKNOWLEDGEMENTS

This work was supported by the Italian Ministry of

University and Scientific Research (MURST, 60%). I am
grateful to Ing. Carlo Frigo and colleagues at the Department of Bioengineering, Fnd. Don Gnocchi (Milan), where
recordings were performed, for discussing biomechanical
aspects of the present work and for the invaluable work in
the development of computer programs. Thanks are also due
to Dr Fedrizzi and colleagues at the Department of Child
Neurology, Neurological Institute C. Besta (Milan) for
allowing us access to patients under their care.

GAIT ANALYSIS IN SPASTICITY

577

REFERENCES
1. Andersen, J.B. and Sinkjaer, T., Stretch reflex variations during gait.
In Neuroprosthetics, ed. A. Pedotti, M. Ferrarin, J. Quintern and R.
Riener. Springer, Berlin, 1996, pp. 4550.
2. Artieda, J., Quesada, P. and Obeso, J., Reciprocal inhibition between
forearm muscles in spastic hemiplegia. Neurology, 1991, 41, 286
289.
3. Ashby, P. and Wiens, M., Reciprocal inhibition following lesions of
the spinal cord in man. Journal of Physiology (London), 1989, 414,
145157.
4. Baev, K. and Shimansky, Y. P., Principles of organization of neural
systems controlling automatic movements in animals. Progress in
Neurobiology, 1992, 39, 45112.
5. Benecke, R., Conrad, B., Meinck, M. and Hohne, J., EMG analysis of
bicycling on an ergometer for evaluation of spasticity of lower limbs in
man. In Motor Control Mechanisms in Health and Disease, ed. J. E.
Desmedt. Raven Press, New York, 1983, pp. 10351046.
6. Berger, W., Dietz, V. and Quintern, J., Corrective reactions to
stumbling in man: neuronal coordination of bilateral leg muscle
activity during gait. Journal of Physiology (London), 1984, 357,
109125.
7. Berger, W., Hortsmann, G. and Dietz, V., Spastic paresis: impaired
spinal reflexesintact motor programs. Journal of Neurology, Neurosurgery and Psychiatry, 1988, 51, 568571.
8. Berger, W., Quintern, J. and Dietz, V., Pathophysiology of gait in
children with cerebral palsy. Electroencephalography and Clinical
Neurophysiology, 1982, 53, 538548.
9. Boorman, G., Hulliger, M., Lee, R. G., Tako, K. and Tanaka, R.,
Reciprocal Ia inhibition in patients with spinal spasticity. Neuroscience Letters, 1991, 127, 5760.
10. Burke, D., Gillies, J. D. and Lance, J. W., Quadriceps stretch reflex in
human spasticity. Journal of Neurology, Neurosurgery and Psychiatry, 1970, 33, 216223.
11. Burke, D., Hagbarth, K. E. and Lofstedt, L., Muscle spindle activity in
man during shortening and lengthening contractions. Journal of
Physiology (London), 1978, 277, 131142.
12. Capaday, C. and Stein, R. B., Difference in the amplitude of the human
soleus H reflex during walking and running. Journal of Physiology,
1989, 392, 513522.
13. Cody, F. W. L., Richardson, H. C., MacDermott, N. and Ferguson, I.
T., Stretch and vibration reflexes of wrist flexor muscles in spasticity.
Brain, 1987, 110, 433450.
14. Corcos, D., Gottlieb, G., Penn, R., Myklebust, B. and Agarwal, G.,
Movement deficits caused by hyperexcitable stretch reflexes in spastic
humans. Brain, 1986, 109, 10431058.
15. Corston, R. N., Johnson, F. and Godwin-Ausen, R. B., The assessment
of drug treatment of spastic gait. Journal of Neurology, Neurosurgery
and Psychiatry, 1981, 44, 10351039.
16. Crenna, P. and Frigo, C., Evidence of phase-dependent nociceptive
reflexes during locomotion in man. Experimental Neurology, 1984, 85,
336345.
17. Crenna, P. and Frigo, C., Monitoring gait with a vector diagram
technique in spastic patients. In Clinical Neurophysiology of Spasticity, ed. P. J. Delwaide and R. R. Young. Elsevier, Amsterdam, 1985,
pp. 109124.
18. Crenna, P. and Frigo, C., Excitability of the soleus H-reflex arc during
walking and stepping in man. Experimental Brain Research, 1987, 66,
4960.
19. Crenna, P. and Frigo, C., Excitability of soleus H-reflex arc during
stepping in spastic patients. Proceedings of the XIV Meeting of the
European Neuroscience Association, Cambridge, U.K., 1991, p. 315.
20. Crenna, P., Inverno, M., Frigo, C., Palmieri, R. and Fedrizzi, E.,
Pathophysiological profile of gait in children with cerebral palsy. In
Movement disorders in children. Medicine and Sport Sciences, Vol.
36, ed. H. Forrssberg and H. Hirschfeld. Karger, Basel, 1992, pp. 186
198.
21. Crenna, P. and Inverno, M., Objective detection of pathophysiological
factors contributing to gait disturbance. In Motor Development in
Children, ed. E. Fedrizzi, G. Avanzini and P. Crenna. John Libbey,
London, 1994, pp. 103118.
22. Davis, R. B. and DeLuca, P. A., Gait characterization via dynamic
joint stiffness. Gait and Posture, 1996, 4, 224231.
23. Delwaide, P. J., Etude experimentale de lhyperreflexie tendineuse en
clinque neurologique. Arscia Editions, Bruxelles, 1971.
24. Delwaide, P. J., Electrophysiological testing of spastic patients: its
potential usefulness and limitations. In Clinical Neurophysiology of

25.

26.
27.
28.

29.
30.

31.

32.
33.

34.

35.
36.
37.
38.

39.

40.
41.

42.

43.
44.
45.
46.

Spasticity, ed. P. J. Delwaide and R. R. Young. Elsevier, Amsterdam,

1985, pp. 185203.
Delwaide, P. J., Martinelli, P. and Crenna, P., Clinical neurophysiological measurement of spinal reflex activity. In Spasticity: Disordered
Motor Control, ed. R. G. Feldmann, R. R. Young and W. P. Koella.
Year Book Medical Publishers, Chicago, 1980, pp. 345371.
Delwaide, P. J. and Olivier, E., Short-latency autogenic inhibition (Ib
inhibition) in human spasticity. Journal of Neurology, Neurosurgery
and Psychiatry, 1988, 51, 15461550.
Dietz, V., Quintern, J. and Berger, W., Afferent control of human
stance and gait: evidence for blocking of group I afferents during gait.
Experimental Brain Research, 1985, 61, 153163.
Dietz, V., Quintern, J. and Berger, W., Electrophysiological studies of
gait in spasticity and rigidity: evidence that altered mechanical
properties of muscle contribute to hypertonia. Brain, 1981, 104,
431449.
Dohrman, G. J. and Nowak, K. W., Relationship between various
clinical signs in lesions of the descending motor system (a statistical
study). Diseases of the Nervous System, 1974, 35, 375377.
Faist, M. D., Mazevet, D., Dietz, V. and Pierrot Deseilligny, E. A.,
Quantitative assessment of presynaptic inhibition of Ia afferents in
spastics. Differences in hemiplegics and paraplegics. Brain, 1994,
117, 14491455.
Fellows, S. J., Ross, H. F. and Thilmann, A. F., The limitations of the
tendon jerk as a marker of pathological stretch reflex activity in human
spasticity. Journal of Neurology, Neurosurgery and Psychiatry, 1993,
56, 531537.
Frigo, C., Crenna, P. and Jensen, L. M., Momentangle relationship at
lower limb joints during human walking at different velocities.
Journal of Electromyography and Kinesiology, 1996, 6, 177190.
Frigo, C. and Jensen, L. M., Gait feature characterization through
moment angle relationship. In: Proceedings of the 10th Congress of
the International Society of Electrophysiology and Kinesiology,
Charleston, SC, 1994, pp. 7475.
Frigo, C. and Pedotti, A., A model of the mechanisms controlling
neuromuscular activation during ambulation based on feedback control hypothesis. Tower International Technomedical Journal of Life
Sciences, 1978, 8, 2935.
Garret, M., Ireland, A. and Luckwill, R. G., Changes in excitability of
the Hoffmann reflex during walking in man. Journal of Physiology
(London), 1984, 355, 23P.
Gottlieb, G. L., Agarwal, G. C. and Penn, R., Sinusoidal oscillation of
the ankle as a means of evaluating the spastic patient. Journal of
Neurology, Neurosurgery and Psychiatry, 1978, 41, 3239.
Hagbarth, K. E., Wallin, G. and Lofstedt, L., Muscle spindle response
to stretch in normal and spastic subjects. Scandinavian Journal of
Rehabilitation Medicine, 1973, 5, 156159.
Hulliger, M., Durmuller, N., Prochazka, A. and Trend, P., Flexible
fusimotor control of muscle spindle feedback during a variety of
natural movements. In J. H. J. Allum and M. Hulliger, eds.. Progress
in Brain Research, 1989, 80, 87101.
Knuttson, E. and Martensson, A., Dynamic motor capacity in spastic
paresis and its relation to prime motor disfunction, spastic reflexes and
antagonist co-activation. Scandinavian Journal of Rehabilitation
Medicine, 1980, 138, 114.
Knuttsson, E. and Richards, C., Different types of disturbed
motor control in gait of hemiparetic patients. Brain, 1979, 102,
405430.
Ibrahim, I. K., Berger, W., Trippel, M. and Dietz, V., Stretch induced
electromyographic activity and torque in spastic elbow muscles.
Differential modulation of reflex activity in passive and active
motor tasks. Brain, 1993, 116, 971989.
Jordan, L., Brainstem and spinal cord mechanisms for the initiation of
locomotion. In Neurobiological Basis of Human Locomotion, ed. M.
Shimamura, S. Grillner and V. R. Edgerton, Japan Scientific Societies
Press, Tokyo, 1991, pp. 320.
Lance, J., Symposium synopsis. In Spasticity: Disordered Motor
Control, ed. R. G. Feldmann, R. R. Young and W. P. Koella. Year
Book Medical Publishers, Chicago, 1980, pp. 485494.
Lance, J., What is spasticity. Lancet, 1990, 335(1), 60610.
Lapierre, Y., Bouchard, S. and Tansey, S., Treatment of spasticity with
tizanidine in multiple sclerosis. Canadian Journal of Neurological
Science, 1988, 14, 513517.
Lee, R. G. and Tatton, W. G., Motor responses to sudden limb
displacements in primates with specific CNS lesions and in human

578

47.
48.
49.
50.
51.
52.
53.

54.
55.

CRENNA
patients with motor system disorders. Canadian Journal of Neurological Science, 1975, 2, 285293.
Leonard, C. T., Hirschfeld, H. and Forssberg, H., The development of
independent walking in children with cerebral palsy. Developmental
Medicine and Child Neurology, 1991, 33, 567577.
Levin, M. F. and Feldmann, A. G., The role of the stretch reflex
threshold regulation in normal and impaired motor control. Brain
Research, 1994, 65, 723730.
Llewellyn, M., Prochazka, A. and Vincent, S., Transmission of human
tendon jerk reflexes during stance and gait. Journal of Physiology
(London), 1986, 382, 82.
Llewellyn, M., Yang, J. F. and Prochazka, A., Human H-reflexes are
smaller in difficult beam walking than in normal treadmill walking.
Experimental Brain Research, 1990, 83, 2228.
McLellan, D. L., Co-contraction and stretch reflexes in spasticity
during treatment with baclofen. Journal of Neurology, Neurosurgery
and Psychiatry, 1977, 40, 130.
Mizrahi, E. M. and Angel, R. W., Impairment of voluntary movement
by spasticity. Annals of Neurology, 1979, 5, 594595.
Morin, C., Katz, R., Mazieres, L. and Pierrot-Deseilligny, E., Comparison of Sol H-reflex facilitation at the onset of Soleus contraction
produced voluntarily and during the stance phase of human gait.
Neuroscience Letters, 1982, 33, 4753.
Murphy, P. R. and Martin, H. A., Fusimotor discharge patterns during
rhythmic movements. Trends in Neuroscience, 1993, 16, 273278.
Partridge, L. D., Muscle properties: a problem for the motor controller
physiologist. In Posture and Movement, ed. R. E. Talbot and D. R.
Humphrey Raven Press, New York, 1981, pp. 189229.

56. Powers, R. K., Marder-Meyer, O. T. and Rymer, W. Z., Quantitative

relations between hypertonia and stretch reflex threshold in spastic
hemiparesis. Annals of Neurology, 1988, 23, 115124.
57. Prochazka, A., Sensorimotor gain control: a basic strategy of motor
systems?. Progress in Neurobiology, 1989, 33, 281307.
58. Prochazka, A., Trend, P., Hulliger, M. and Vincent, S., Ensemble
proprioceptive activity in the cat step cycle: towards a representative
look-up chart. In J. H. J. Allum and M. Hulliger, eds.. Progress in
Brain Research, 1989, 80, 6174.
59. Rack, P. M. H., Ross, H. F. and Thilmann, A. F., The ankle stretch
reflexes in normal and spastic subjects. Brain, 1984, 107, 637654.
60. Sahrmann, S. A. and Norton, B. J., The relationship of voluntary
movement to spasticity in the upper motor neuron syndrome. Annals of
Neurology, 1977, 2, 460465.
61. Soechting, J. F., Dufresne, J. R. and Laquaniti, F., Time-varying
properties of myotatic responses in man during some simple motor
tasks. Journal of Neurophysiology, 1981, 46, 12261243.
62. Thillmann, A. F., Fellows, S. J. and Garms, E., The mechanisms of
spastic muscle hypertonus. Variation in reflex gain over the time
course of spasticity. Brain, 1991, 114, 233244.
63. Yanagishawa, N., Tanaka, R. and Ito, Z., Reciprocal Ia inhibition in
spastic hemiplegia of man. Brain, 1976, 99, 555579.
64. Yang, J. F., Fung, J., Edamura, M., Blunt, R., Stein, R. B. and Bargeau,
H., H-reflex modulation during walking in spastic paretic subjects.
Canadian Journal of Neurological Science, 1991, 18, 443452.
65. Yang, J. F. and Whelan, P. J., Neural mechanisms contributing to
cyclical modulation of the Soleus H-reflex in walking in humans.
Experimental Brain Research, 1993, 95, 547556.