Case of the Week 10/4/2000: What people said...

Dear Chris,
I would just like to revisit the exploration of arm swing. Take a look
at this patient. As his left foot is in swing, and before heel contact,
it is obvious that there is an internal rotaion of the left femur, which
most likely is happening from one of the hip flexors. As the knee
reaches full internal rotation, the right arm has reached its maximum
swing in adduction, which is a much different arm swing than the
contralateral side. Could this type of arm swing be a type of
compensatory motion for the the abnormal left internal rotaion of the
left leg? How is that for a run on sentence? I know I didn't go into all
of this patients problems (which are many) but I would like to know your

Henry Slomowitz

Certainly that left side has shown some of the deleterious effects of a TAL
with loss of ankle power and the development of crouching. He has on
kinematics reduced swing phase knee flexion with diminished velocity into
flexion after toe-off. he has increased ankle AM1 (moment) on the left an
increased hip flexion in swing on the left. Clinically he has increased
anteversion and increaed internal rotation in gait and on kinematics. His 5
degrees of ETT probably is contributing to his pronated valgus feet.
He probably needs: Bilateral Fem Derotation Ost., Left rectus transfer with
hamstring release, Rt. hamstring release, left Gastroc recession and
possible lateral column lengthening possily with inter-tib osteotomy.
Bilateral psoas rescessions at brim.

>From Shriners Gait Lab Vancouver BC.

Richard Beauchamp

What are the problems?

Do you think the previous surgery has made a positive or negative contribution?

          I think the surgery has made a negative contribution on ankle power, affecting progression and unstability
diuring stance

What further surgery would you advocate?

          I would like to try first with an orthosis UCBL type
          and gait retraining using FES on pretibials before thinking on another surgery

What results do you anticipate?

    Better kinematics and kinetics of  left ankle

Best Regards

Dr. Raúl Díaz González Santibáñez
Laboratorio de Análisis de Movimiento
Centro de Rehabilitación Infantil Teletón
Av. Gustavo Baz #219 Col. San Pedro Barrientos
Tlalnepantla, Estado de México
Tel: (52)-5-3212246

Hi Chris,

It was great to finally meet you last week. I just took a look at the
comparison between Orthotrak and BodyBuilder.  Thanks for doing this, it is
pretty interesting.  I have a feeling that some of the differences might be
attributed to filtering differences (decreased peaks and phase shifs).  Do
you have any insight into the differences after the meeting?  Any thoughts
on the curves that were only offset in magnitude (eg. ankle

Edi Cramp

I'm back at my desk after a very intensive and stimulating week in
Rochester. It was nice to meet so many CGA subscribers at the GCMAS meeting,
by the way!

For those of you who weren't there, I might describe here the process used
on the last day, in which several cases were reviewed by an expert panel,
culminating with a wireless survey of the audience responses.

I wonder if we might implement a similar strategy for Case of the Week.
We have actually tried this before (Case 25-9-97 - see:
/cgi-bin/cga_form.cgi?logFile=25-9-97) for
observational analysis. I wonder if we might broaden the concept to include
the whole assessment. I've made a (non-functional) first draft at
/archives/10-4-00/survey.html and I'll see
if Ray Smith in Australia can convert this to a CGI script. Please let me
have your comments about what needs to be included.

Now, back to the Case in hand. I will try to tackle it in the methodical
manner that was demonstrated at the conference. Note, BTW, that for this
case the right side is denoted by the red curve and left is blue.

Observational - toeing-in bilat, perhaps worse on left; with left patella
internally rotated. Trendelenberg, with right hemipelvis dropping in swing.
Pelvis rotated with right side forward. Right arm thrust at ipsilateral
contact. Left forefoot contact with flexed knee; right knee slightly flexed
at right contact, but heel-strike preserved.

Kinematics - right hemipelvis internally rotated, with (surprising small
compared to video) Trendelenberg fall during swing. Bilateral anterior tilt,
with no double bump (in fect, rather the opposite). Hip flexion normal
bearing in mind the pelvic offset; right side abducted during stance.
Loading response is lost, especially at the left knee, which is also a
little stiff in swing. Left ankle is about 10 degrees equinus and fails to
dorsiflex in stance, while the right collapses into excessive dorsiflexion
and fails to plantarflex in push-off.

As Jim Gage demonstrated, the hip rotations can be seen to be compensations
for the pelvic rotation (see my diagram at:
/archives/10-4-00/rotation.gif), which
leaves only about 20 degrees (8 deg int + 10 deg ext foot progression)
unaccounted for bilaterally, which is presumably tibial torsion. This
doesn't fit with the clinical exam, so I'm a bit confused - perhaps someone
can correct me?

I was going to go on to look at the kinetics, but perhaps this would be a
good place to stop and hand the baton to someone else?

Dr. Chris Kirtley MD PhD
Associate Professor
HomeCare Technologies for the 21st Century (Whitaker Foundation)
NIDRR Rehabilitation Engineering Research Center on TeleRehabilitation
Dept. of Biomedical Engineering, Pangborn 131 (office 105b)
Catholic University of America

Since nobody has replied about my comments yesterday, I thought I might try
to explain what I was saying again a bit more clearly.

The pelvic rotation graph shows that the patient walks with the right pelvis
held forward (internally-rotated), confirming the video findings.

A common reason for this, according to Jim Gage, is to permit the hip
adductor on that side to supplement the hip flexor for pull-off. I've
checked the curves, however, and the kinetics don't seem to bear this out -
I expected to see an adductor moment on the right at toe-off, but in fact,
it's close to the normal abductor pattern. So, I wonder if anyone else can
suggest a reason for the pelvic rotation?

Anyway, because of this pelvic rotation, the right hip would have to
externally rotate and left hip internally rotate by the same amount (about
15 degrees) in order for both femurs to continue to point down the plane of
progression. The hip rotation curve, in fact, confirms this to be the case,
although the right (red curve) may be a little under-compensated, implying
that there is very little femoral anteversion.

Since the foot progression on each side is about 8 degrees internal (whereas
it should be about 15 degrees external), I am saying that this must
therefore mean that there is about (15+8) 23 degrees of tibial torsion to
explain the discrepancy (see the diagram at:

However, the clinical exam reported anteversion of 60 degrees and about 5
degrees of external tibial torsion bilaterally. I must admit that the video
seems to back the exam rather than the motion analysis findings, but it's
difficult to be sure.

Am I correct in my reasoning, or have I missed something? Perhaps Orthotrak
incorporates the clinical anteversion value in its output? I stand prepared
to be corrected!

Dr. Chris Kirtley MD PhD
Associate Professor
HomeCare Technologies for the 21st Century (Whitaker Foundation)
NIDRR Rehabilitation Engineering Research Center on TeleRehabilitation
Dept. of Biomedical Engineering, Pangborn 131 (office 105b)
Catholic University of America

Dear readers,

Your calculations above suggest that you are quite confident about what
you measure on the graphs. Is the VICON that accurate in the transverse
plane? (I am not a VICON user.) This question first occurred to me when
I read the paper in JPO from Stefko et al (1998) who measured the effect
of distal tibial derot. osteotomy with VICON.

You then conclude that the video backs the exam rather than what you
measure. Does this mean that your final decision will be based mostly on
the video and phys. exam?

The generalised question is a recurring one: Do we trust our data, or do
we prefer the video/exam?


Dr Gabor Barton MD  
Clinical Scientist
Gait Analysis Laboratory             tel: +44 (0)151 252 5949
Alder Hey Children's Hospital        fax1: +44 (0)870 052 1935
Eaton Road, Liverpool, L12 2AP, UK   fax2: +44 (0)151 252 5846

Dear Gabor, and others interested,

I need to clarify a few more things:

1. This case was recorded using a Motion Analysis Corp. (not Vicon)
system. However, it is also the first case in which I have run a
parallel analysis using Vicon BodyBuilder (see:
/archives/10-4-00/Rkinem.gif and
/archives/10-4-00/Lkinem.gif). While I
am not yet completely confident that I am running the BodyBuilder
Cleveland model, which may account for a few discrepancies, the
differences are actually quite minor.

2. Of course, we all know that frontal and transverse plane measurements
are particularly unreliable - although  my own values (ICCs taken from
the maxima and minima) are not too bad:

Pelvic Rotation    0.72 and 0.7
Hip Rotation        0.86 and 0.92

.. see: /faq/reliability

3. All measurements in this field are subject to error - including the
video and clinical exam (not necessarily the examiner, but sometimes in
the transcribing of the chart), so I think one has to be constantly
cross-checking each source of information against the others - that it
what makes this field such an interesting challenge!

So, I didn't mean to suggest that the video evidence overrides the
motion analysis data. I am trying to understand where the discrepancy
between the rotations is arising - is it:

a. from me (i.e. have I made a boo-boo?)
b. fom the motion analysis (and therefore a technical problem that needs
discussing) or,
c. from the clinical exam

Dr. Chris Kirtley MD PhD
Associate Professor
HomeCare Technologies for the 21st Century (Whitaker Foundation)
NIDRR Rehabilitation Engineering Research Center on TeleRehabilitation
Dept. of Biomedical Engineering, Pangborn 131 (office 105b)
Catholic University of America


I did some checking on the OrthoTrak calculations.  Do you know if Andreas
is using the new version of OT (PC or Mac based) or the old version (Sun
based)?  Here are the details on the new version.  I know there are
difference with the old version, but I don't know what they are.

The static trial is done with three markers on the Pelvis, triads on the
thighs and shanks, medial and lateral knee and ankle markers, and a
forefoot marker.  For the dynamic trials, the medial and lateral knee and
ankle markers are removed and a heel marker is added (at the same height as
the forefoot marker).  I am not sure what method of calculating hip centers
is used in OT.  Note that there may be two static trials done: one for each

I hope this helps.

Ed Quigley

Dear All Analyzers,

I'm on my travels in Europe, and Andreas in Vienna tells me that he's
ready to disclose the post-op results.

Our discussion of this case seemed to stall when I pointed out that the
clinical exam findings didn't fit with the rotations as recorded by the
gait analysis. Andreas has since confirmed that the gait findings are
correct, so here's a nice example of validation (or rather invalidation)
of the clincal exam. There are so many possible sources of errors and
inaccuracies, even in the clinical exam, that it's vital to constantly
cross-check abnormal findings. This is, of course, also true in other
branches of medicine and science.

Anyway, I wondered whether anyone had any more comments before Andreas
posts the post-op curves and videos?

mailto:[n/a] with your comments.

Dr. Chris Kirtley MD PhD
Associate Professor
HomeCare Technologies for the 21st Century (Whitaker Foundation)
NIDRR Rehabilitation Engineering Research Center on TeleRehabilitation
Dept. of Biomedical Engineering, Pangborn 105B
Catholic University of America

I feel that surgery had a negative effect on the gait due to weakening
the calf muscles.  Simon declares the calf muscles to be the major
postural muscle of the body and both Sutherland and Simon have found
loss of calf muscle strength to be a major detriment to gait efficiency,
stability and step length.  Perry has explain how an ipsilateral pelvis
drop and posterior pelvic rotation may be due to a weak soleus on the
stance limb which leads to a loss of heel rise and relative lengthening
of the limb.

I would suggest that strengthening of the calf muscles, gastrocnemius
and soleus as well as the gluteus maximus muscles as well as others
testing weak.  Perhaps the peroneal longus muscles also need
strengthening.  With electrical stimulation (NMES/FES) and the use of a
remote switch it is possible to stimulate weak muscles close to the
timing that would normally occur in gait.  When strengthening muscles
while during a specific task the strengthening will be more likely to
continue with a carry over in the task, gait, that if strengthened while
sitting.  It has been shown that stimulation of calf muscles in children
with cerebral palsy result in a planargrade gait.  I do not agree with
the recommendation to stimulate the anterior tibials muscles.  Often
children with cerebral palsy are using the anterior tibialis to bring
the leg forward.  It is also my experience over time strengthening of
the calf can result in barefoot heel strike. It has also been shown in a
statistical study with 14 children with cerebral palsy that stimulating
the tibialis anterior muscle had no effect. (Comeaux P, et al Task
Specific Approach with NMES to Calf Muscles.  Pediatric Phy Ther,

Judy Carmick

Dear Judith and others,

Many thanks for your comments. I have been thinking for some time that
the foot and ankle are key to the management of many of these cases that
we've seen. However, it does seem to be also most controversial area of
treatment (as anyone who witnessed the debate between Casey Kerrigan and
Jim Gage at the recent North American Gait meeting will, I'm sure,
agree!). Basically, we have the following possibilities, I suppose:

Have I missed any more???

Such a bewildering array of treatments in medicine normally means that
none is entirely satisfactory.

Concerning your points about the calf muscles being the major postural
muscle of the body, I would agree, but I wonder what you think of
Skillsfield and Strayer's distinction between monoarticular (soleus) and
polyarticular (gastrocnemius) muscles in antigravity activity. This view
(which has been popularized by Gage) sees the biarticular muscles as the
primary cause of the spasticity and the soleus as the primary
antigravity component. Hence the invention of the Strayer operation,
which lengthens gastroc while leaving soleus intact.

I confess that I have been a faithful follower of this philosophy without asking what
evidence this idea is based on! I wonder if anyone can point to any?

It seems to me that we need a much more rigorous scientific approach to
ankle spasticity if we are to prescribe the above treatments

Dr. Chris Kirtley MD PhD
Associate Professor
HomeCare Technologies for the 21st Century (Whitaker Foundation)
NIDRR Rehabilitation Engineering Research Center on TeleRehabilitation
Dept. of Biomedical Engineering, Pangborn 105B
Catholic University of America

In response to Dr. Kirtleys list of possible procedures, focusing only on
the knee, you could make the argument to consider a left rectus transfer, as
the peak is delayed and diminished, the transition from late stance to swing
is gradual (reduced slope), and there is definite rectus activity in swing.
This being said, however, I don't think that our lab would recommend this at
this time, as foot position is so shifted into plantarflexion, it is
difficult to predict how much the knee curve will correct if the foot
position is corrected. If I am interpreting the physical exam data
correctly, the gastroc appears to be tight muscle with only minimal soleus
involvement as there is a 35 degree difference between max dorsiflection
with the knee at 0 vs. 90.  The surgical preference in our hospital would be
a Baker type gastroc lengthening. Based on previous experience we would
anticipate that this would correct his foot position from mid-stance on,
however, he would continue to land with his foot in a plantarflexed position
(although better than pre-op) with his knee flexed greater than normal.  His
ankle moment curve would be nearly normal (a slight bump at loading) and his
push-off power would improve because the improved foot position prior to
push-off. There would still be some small absorption during loading from the
initially plantarflexed foot position. I would be interested in hearing if
others would also anticipate this and also if anyone has ideas on how to get
these patients to achieve heel strike -  strengthening the anterior tib?

Bruce Macwilliams, Ph.D.
Co-Director, Movement Analysis Laboratory
Shriners Hospitals for Children Intermountain Unit
Fairfax Rd. at Virginia St.
Salt Lake City, UT 84103
Phone:  801-536-3800
Fax:      801-536-3782

Dear all
Thanks Andreas.

What's he got?

Possible femoral internal rotation on both sides, with functional
consequences on the left (evidence: internal hip rotation while walking on
the left and compensatory external rotation of the left hemipelvis {Chris,
surely the pelvis compensates for the hip - not the other way round?} ;
increased femoral anteversion on both sides {notorious measure}). Does
Andreas have internal/external hip rotation ranges? These would provide
additional surety.

Inappropriate rectus femoris activity on both sides with functional
consequences on the left (evidence:reduced and broadened peak knee flexion
wave on the left with a slow rate of rise in knee flexion; rectus femoris
activity in mid-swing on both sides; poor clearance in swing on the left;
positive for rectus spasticity on both sides)

Left Gastroc fixed shortness (evidence: extraordinary plantarflexion
contracture with knee extended; absence of dorsiflexion while walking,
probably worse than the kinematics indicate because of foot deformities on
this side (though I cannot see the video that clearly); poor ankle power at

Mild Right gastroc fixed shortness (evidence: mild plantarflexion
contracture with the knee extended; right knee a little more flexed than one
would like in terminal stance)

Foot Deformities on the right (evidence: video, but dodgy - needs clinical
examination of the foot)

Note: need more info on hamstring fixed shortness, I assume below that
evidence is provided.

What would you advise?

Further thoughts:
I agree with Judy about stimulation to the calf muscles, but exactly how you
apply the stimulation I think is still up for grabs. Personally, I think FES
of the calf muscles probably has limited potential, but as a therapeutic
measure electrical stimulation could exercise and consequently increase the
bulk and the length (! I can feel people saying "what is he on about?") of
the plantarflexors when that exercise is in the correct range. Incidentally,
the emg activity of the this guy's tib ants look ferocious - is that right
Andreas? - if so, I expect FES of thes muscles would prove fruitless.

Adam Shortland

Chris - great job at summarizing the various possibilities.  Does show the
challenges.  Perhaps add Intrathecal Baclofen infusion pump therapy, and
other proactive drug-based approaches?  Also assistive technologies such as
walkers?  As I look down your list, all make sense except one (yes, my
personal bias): Dorsal Rhizotomy, which I'd certainly never consider if I
had a child of my own with CP.  (Also, certain ones, like serial casting,
beg for new technology and approaches.)

By the way, the Workshop on Pediatric Gait, this past Saturday in
Chicago, was very good (and well attended), and the book tied to the
Workshop has been published, and is available - Jerry Harris and Peter Smith
(co-organizers) did a great job.

Jack Winters

Dear Analyzers
Here are my answers to the case of the week:

1.-What are the problems?

Kinetics EMG Power:

2.- I think that the previous surgery was premature because we see same problems but to being fare left foot kinematics and kinetics is good, so it wasn´t so bad at all

3.- I would not recomend surgery at this moment I would try inhibitory orthosis on the right ankle UCBL type, Botox therapy on rectus femoris, medial gastrocnemius and long head of biceps femoris on the right. And biofeedback on right tibialis anterioris and with FES for gait retraining

4.- With another surgery would anticipate poor activity of triceps surae during stance with high dorsiflexión, poor exentric contraction of triceps surae during stance with worse ankle imbalance

Thank you very much

Dr. Raúl Díaz González Santibáñez
Laboratorio de Análisis de Movimiento
Centro de Rehabilitación Infantil Teletón
Av. Gustavo Baz #219 Col. San Pedro Barrientos
Tlalnepantla, Estado de México

Tel: (52)-5-3212246

Dear Raul,

I agree with you that another surgery would anticipate poor activity of the
triceps surae.  The child already demonstrates a poor push off. We have seen
children with cerebral palsy, especially adolescents, with spindly legs and
no calf muscle bulk. One may wonder how the child will continue to walk when
full weight and height have been achieved.

The gait of children with cerebral palsy is known to deteriorate over
time.(1,2)  Johnson(2)  found that this deterioration began as early as six
years of age and more so in those who had had surgeries.   Riewald(3)  found
that rectus femorus transfers did not change the timing of the muscle as it
continued to act as a knee extensor not a flexor. We need to strengthen the
short head of the biceps femoris in order to achieve improved knee flexion at
heel rise.

Isn't it time to think of strengthening the needed muscles and helping the
child to use them rather than weaken them or move them to other places?  No
harm will come as it has been shown that strengthening spastic muscles leads
to increased ROM, strength, balance, decreased spasticity and improved

1.  Norlin R, Odenrick P.  Development of gait in spastic children with
cerebral palsy.  Journal of Pediatric Orthopedics 1986;6:674-680.
 2.  Johnson DL, Damiano DI, Abel M. The evaluation of gait in childhood and
adolescent cerebral palsy.  J Ped Orthopedics 1997;17:392-396.
3.  Riewald SA, Delp SL. The action of the rectus femorus muscle following
distal tendon transfer: does it generate knee flexion moment?  Dev Med Child
Neurol. 1997;39:99-105.

Judy Carmick

Dear Gabor,

When I was mentioning that the tibialis anterior might help to advance the
leg I was thinking of helping it during swing.  The hip, knee and ankle
flexion synergy is initially used in pre-walkers and supported walkers as
well as new walkers.  In order to assist limb advancement the tibialis
anterior can act more strongly to help lift the foot and may assist bringing
the leg forward in swing.  Immature children and children with cerebral palsy use a
coactivation pattern of the tibialis anterior and the calf muscles during
stance. With the development of free walking the pattern becomes more
reciprocal in the child without cerebral palsy.  Those with cerebral palsy
usually continue in the immature pattern as the child in the case appears to
be doing.

Berger has pointed out that the magnitude of the tibialis anterior EMG did
not change after 2 years of age with the maturation of gait but the
final magnitude of the gastrocnemius was not established until around 4 to 5
years of age in children without disability who had the ability to continue
to strengthen the calf muscles.  The children with cerebral palsy do not have
this ability to strengthen the calf. This is even more so when solid AFOs,
Botox injections, selective dorsal rhizotomy or surgical lengthenings are

Marsh and Sale have shown that plantarflexion is the best position for the
tibialis anterior muscle to function.  They found that the tibialis anterior
is strongest in 30 degrees of plantarflexion and the plantarflexors are
strongest in 10 degrees of dorsiflexion.  That may be why the young child
walks in plantarflexion until the calf muscles have developed enough to be
stronger than the tibialis anterior muscle. Adam pointed out that the child
in the CGA case study already has a "ferocious" tibialis anterior muscle and
a weak push off or calf muscle.  Normally the calf should be at least three
times stronger than the tibialis anterior muscle; they are not normally of
equal strength.

1.  Leonard, Charles T, Hirschfield, Helga, and Forssberg, Hans, The
development of independent walking in children with cerebral palsy, Dev Med
Child Neurol. 1991;33:567-575.
 2. Berger, W, Altenmueller E. and V. Dietz, Normal and Impaired Development
of Children's Gait, Human Neurobiology, 1984;3:163-170

Judy Carmick

Dear Bruce,

I feel that in order to get heel strike one needs to strengthen the calf
muscles not the tibialis anterior muscles.  We know that the child in this
study already has a "ferocious" [Thanks Adam,] tibialis anterior muscle and a
weak calf.  We know that the calf is a major postural muscle of the body and
that its development is very important and that loss or weakness can be very
detrimental.  Normally the calf doubles in size from the age of two to four
or five.  However, children with cerebral palsy are often not able to
contract and relax the muscles and thus are not able to develop the muscle.
If the ankle is blocked via braces or weakened through other therapeutic
means the calf development is even less.  NMES may make calf development and
function possible if stimulated during a specific task.

One can stimulate the calf muscles for muscle contraction and see immediate
changes in the child's gait and stance.  In young children (ages 18 months to
5 years) strengthening the calf with NMES will take about 30 seconds to see
the child move from plantarflexed standing to plantargrade.  In older
children it may take 15 minutes.  Often the older children (age 10 up) will
say that walking is much easier.  One child in the Comeaux article(1)
mentioned he no longer felt like he was walking in sand.  Older children who
are very weak may also need NMES to the gluteus maximus muscles as well as
the gastrocsoleus muscles in order to see the dramatic changes that are seen
in the younger less heavy children.

Adam mentioned that knowing how to apply the stimulation was up for grabs.
This appears to be true as many continue to put them on the tibialis anterior
muscles.  The use of the electrodes was first presented in a paper.(2)  I
often use 1 and 1/4 inches  X 2 inch electrodes on the muscle bulk of the
calf.  The more active electrode, the cathode, is placed at or near the motor
point of the medial calf.  The anode, dispersive electrode, is placed on the
motor point of the lateral calf.  There is about 3/4 to one inch separation
between the electordes.  All amplitude is to the child's tolerance.  An
asymmetrical wave form is used.  When standing one calf receives stimulation
to the electrodes for 10 to 15 seconds and then the other calf receives
stimulation for 10 or fifteen seconds.  When the child is walking, it is
desirable to stimulate the muscles, as they would normally do during gait.
The anterior tibialis mainly functions in the mature gait during swing, and
the calf during stance.  So a remote switch is used at heel strikes or
initial contact of the stimulated calf throughout stance and then released at
heel strike or initial contact as it may be throughout stance on the other
leg.  The NMES unit is set to alternate channels as the remote switch is
pressed.  So that when pressed one leg is stimulated and when released the
other leg is stimulated.

Maybe not so incidentally I have found that stimulating the tibialis anterior
muscle during swing of a child under five or one who had not first developed
the calf muscles led to more toe walking and not less. Both Hazelton(3) et al
and Comeaux et al(4)  used NMES/FES to the tibialis anterior and demonstrated
that NMES to the tibialis anterior muscles did not improve function in gait.
I find it is helpful only if the child has a true drop foot and is older than
five years old.(5)

I hope that this will interest you enough to have your physical therapist try
it.  I would be happy to answer her questions should she or he have any.

 1. Comeaux P, Patterson N, Rubin M, Meiner R.   Effect of neuromuscular
electrical stimulation during gait in children with cerebral palsy.  Pediatr
Phys Ther. 1997;9:103-109.
 2. Carmick J.  Clinical use of neuromuscular electrical stimulation for
children with cerebral palsy, part 1: lower extremity.  Phys Ther.
 3. Hazelwood ME, Brown JK, Rowe PJ, Salter PM.  The use of therapeutic
electrical stimulation in the treatment of hemiplegic cerebral palsy.   Dev
Med Child Neurol. 1994, 36:661-673.
 4. Comeaux P, Patterson N, Rubin M, Meiner R.   Effect of neuromuscular
electrical stimulation during gait in children with cerebral palsy.  Pediatr
Phys Ther. 1997;9:103-109.
 5 . Carmick J. (Letter) NMES during Gait. Pediatr Phys. Ther. 1998;10:92.

Judy Carmick

Dear Chris,

It was most interesting to hear the comments about my suggestion to
strengthen the calf muscles instead of the traditional weakening.  It was
delightful to read those that agreed.

When you use NMES to the calf and see immediate plantarflexion you will think
that perhaps strengthening of the calf should be given a chance before
resorting to surgery, which will weaken the muscles lengthened.  The Strayer
may not touch the soleus but the gastrocnemius muscle does flex the knee and
helps prevent hyperextension of the knee so why weaken it?  Houtez (1) found
that in adults without disability when asked to hyperextend the knee they
relaxed the gastrocnemius.

If you do not believe what can happen with NMES to the calf muscles you may
want to look at some photographs(2) which show immediate changes in the
posture of a 4-year-old child with moderate spastic diplegia receiving NMES
to the calf.  He moved immediately from plantarflexion to plantargrade the
first time it was used.  A second article (3) concerns the same child and
also shows photographs which show how important it is to allow the calf to
function even slightly when wearing hinged AFOs. Two rows of photographs show
this same child before NMES when walking in solid ankle and hinged braces.
In the solid ankle braces, 4 surgeries were recommended as the child walked
in plantarflexion with a high arm guard.  When the calf was allowed to work
in hinged AFOs the child immediately moved from the arms in high guard to
medium guard and walking with a heel strike, increased step length and no
longer internally rotating the hips.  The reasons for the surgery, except for
hip flexion, were then questionable.  The problems appeared to be more
biomechanical brought on by braces that blocked movement, than neurological
caused by the cerebral palsy.  By ten years of age, the child in that case
study progressed to being able to walk barefoot with a heel strike, to ride a
two wheel bicycle without training wheels, to jump up 12 inches and to pass
all the activities on the GMFM although he at one time was diagnosed with
moderate spastic diplegia and had been recommended to have four surgeries.
His spasticity and clonus dropped significantly.

The child in the case study of this list CGA had surgery at four years of age
and still has a heelcord contracture and a weak push-off.  One cannot imagine
that he can stand on one foot, hop or jump forward with two feet, or upwards
for a significant distance.

I do not have experience with the Strayer as the children with whom I work do
not need calf muscle surgery after using NMES although it may have been
recommended at first.  I wonder if anyone has children who have had surgery
for the calf muscles have then seen the children progress to a heel strike,
or pass the GMFM test, or jump forward 36 inches or up even 10 inches.  What
is happening to those who had surgeries that weakened the calf muscles when
they have reached their adult weight and height?  Rang et (4) all said that
"Lengthening a contracted calf will eliminate equinus but the child will not
be able to walk on tip toes." If he cannot do that how can he do all the
things needed in the GMFM as well as the extra fun things that the children
who have had the calf muscle strengthen been able to do?

It seems clear to me which approach to use.  If NMES is given a fair chance
and no changes are used one could also return to surgery or Botox. NMES does
no harm if done correctly, as we now know that strengthening spastic muscles
does not increase spasticity as once believed.

Judy Carmick

1.  Houtz SJ, Walsh FP.  Electromyographic analysis of the function of the
muscles acting on the ankle during weight bearing with special reference to
the triceps surae.  J Bone Joint Surg. 1959;41A:1469-1481.
2.    Carmick J.  Managing equinus in children with cerebral palsy:
electrical stimulation to strengthen the triceps surae muscle. Dev Med Child
Neurol. 1995;37:965-975.
3.  Carmick J.  Managing equinus in children with cerebral palsy: merits of
hinged ankle-foot orthoses.  Dev Med Child Neurol. 1995;37:1006-1010.
4.  Rang M., Silver, R., de la Garza, J.  Cerebral palsy.  In Lovell WW,
Winter RB, (Eds.) Pediatric Orthopaedics, vol. 2, ed. 5.  Philadelphia,
Pennsylvania: J. B. Lippencott,1986  pp. 345-390.

Dear Adam,

Thank you for your comments about positive stimulation to the calf muscles
and fruitless stimulation to the tibialis anterior muscles.  I hope you will
read my suggestions on how to do it in the answer to Bruce.  If you see NMES
used to the calf on young children you may not think that the potential is

NMES can help knee flexion at heel rise if the short head of the biceps
femorus muscle is stimulated at heel rise.  I have used this often and the
children all have said that they prefer stimulation to the short head of the
biceps femoris over stimulating to the quadriceps muscle when walking.  They
say it is easier with the short head of the biceps femoris.  Over time the
child can flex the knee independently of NMES and the Ashworth test has
resulted in some children to a 1 or a 0 grade to fast knee extension.  Of
course these children also had NMES to the gastrocsoleus and gluteus maximus

Judy Carmick

Dear Chris,
I think the Strayer operation increases the the overall gastroc
musculo-tendinous length by a substantial increase in tendon length. I
believe it shortens the gastrocnemius belly- and fibre- length, reducing the
peak force produced by this muscle, its active range (hence no walking on
toes) and its maximal velocity of shortening. Consequently, power is
reduced. The up side is that position is improved. I expect you do something
injurious to the Soleus muscle belly as well since the operation (so I
understand) is on the internal tendon of the Soleus (Gastroc external
I echo what you write about evidence for the relative spasticity of
bi-articular muscles. I have a vague notion that Perry's lab measured emg in
rectus femoris and hip flexors and showed that the Duncan-Ely test (rectus
stretch) does not result in rectus activity only. Could someone fill in the

Dear Judy,
A short gastrocnemius can result in hyperextension of the knee in toe
walkers because of the large external ankle moment due to the abnormal foot
position. There was an article in ASME Biomench Eng. a few years ago (Brown,
I think) demonstrating that in certain positions in stance the hamstrings
may act as knee extensors - this is probably also true of the gastrocnemii
 a good experiment for your site Chris!).
Also, I am intrigued by your electrical stimulation technique. Long-term
electrical stimulation, one can imagine, alters the structure (mechanical
potential) of a muscle. What mechanism do you think changes an equinus to a
plantargrade position in a few seconds? Are these individuals very dynamic?
Perhaps the muscle spindles are being electrically reset and strong tonic
reflexes are being blocked by your stimulation.

I can't imagine how the technique would fair on individuals with severe
fixed contractures.

All the best,

Adam Shortland PhD.,
One Small Step Gait Laboratory,
Guy's Hospital,

Dear Sarah,

I have used NMES on a few adults who had cerebral palsy.  One was in her 50's
and also is  an occupational therapist.  She had hemiplegia.  We worked
mostly on the shoulder and hand but also did the lower extremity.  When NMES
was applied to the ulnar nerve she had improved sensation in the hand. I had
not started there but on other muscles located in the forearm.  After feeling
the sensation in the hand she asked me to give her a stereognosis test.  This
was done and she knew everything I gave her.  Unfortunately I had not done
this first, but she said that all through OT school she was not able to feel
anything.  She commented that she didn't know if she had better feeling
because it was better or if it was that by that time she was better able to
manipulate the item in her hand.  Needless to say, now I do a stereognosis
test before working with anyone with NMES on the hand.

We also worked with the calf muscles during gait while using a remote switch.
 Her main problem was tripping and not shifting weight to the involved side.
After six 15 minute sessions of walking with NMES to the calf, she no longer
tripped.  And she no longer had pain at night on the uninvolved right leg
from overworking the right leg.  She must have been weight shifting more
consistently to the involved side.  After 6 months of no NMES the pain and
tripping returned.  Then NMES was begun again and she improved again as

On the return visit she asked to have NMES to the anterior tibialis instead
of the calf.  We had use the active/cathode on the medial gastrocnemius and
the anode on the left gastrocnemius.  We had also stimulated the soleus.  So
I did as she asked.  After walking about 15 feet she asked to have it removed
as she felt it was throwing her off balance.  She tended to have some
supination to the foot during swing so more stimulus to the anterior tibialis
was not a good idea.

I find NMES every exciting.  It is easier to see immediate changes with the
very young 18 months of age and up.  The older children or adults are also
interesting and show changes some of which may be how they feel about the
movement or sensation.  They may tell you it is so much easier to walk with
it.  So it is most important to ask them how it feels.

Judy Carmick

Dear Adam,

Yes, I agree that a short gastrocnemius as well as a weak gastrocnemius can
cause hyperextension of the knee.  It is better to have strong muscles than a
long tendon and a weak muscle, which would result from the surgery. Muscle
needs to be used to grow, so strengthening the muscle would be a good way to
go.  It is very difficult to strengthen muscles in children with cerebral
palsy, as they do not know how to "turn it on."  NMES helps them do that.

I do not know what mechanism is working when changing a child from
plantarflexion to plantargrade but my mind has been running wild on this idea
at times.  So I am glad that you asked.  We have been told that the spinal
cord does not work correctly in children with cerebral palsy.  So one idea I
have is that the afferent input from the muscles contracting via NMES may
have something to do with stimulating unused circuits of the spinal cord that
may help its organization.  It would be wonderful to hear from the experts on
this.  Sahrmann and Norton  found that the primary impairment of movement is
not due to antagonist stretch reflexes, but rather to limited and prolonged
recruitment of agonist contraction and delayed cessation of agonist
contraction at the termination of movement. I wonder if NMES has some effect
that assists the body to use reciprocal contractions or to assist the child
at not contracting the muscle at unwanted times.

NMES gives the child some sensory awareness of the muscle and how it is
over-reacting or in constant contraction. This sensory input appears also to
help with some children who are trying too hard to reach or to walk and end
up using cocontraction.  For example children who adduct the legs in sitting
perhaps thinking that they need to do this in order to sit erect.  I tried an
experiment on 4 children who had good communication skills. Three were
moderate or severely involved used wheelchairs and walked only when held, one
of which could not walk or even take steps when held at that time.  The
fourth child was about 11 years old with spastic diplegia and could walk
independently with a walker or quad canes.  Two of the entire group were with
mixed spastic athetosis, ages 12 and 6.  One was 8 years old with triplegia
and 4 years post rhizotomy.   All used excessive hip adduction and ROM was
limited or very tight.  NMES was placed on the adductor muscles, alternating
from right to left side for 15 seconds at a time.

After NMES was used for a few minutes, the child was asked if the legs felt
the same or  did the legs feel that they were going together or going apart.
(The children were all seen individually.)  They all said that the legs felt
that they were going apart.  In fact they did move apart.  Hip adduction was
much increased by several inches with hip and knee flexion or extension.
External rotation of the legs passively was increased when doing hip
abduction with knee extension.  I felt that the children had been
involuntarily pulling the legs together before NMES and the NMES gave them
some sensation of what they were doing and so they stopped pulling so much.
I have found this to be true in many muscles and so usually stimulate the
spastic muscle rather than the antagonist.

I have used NMES long term on a child whose orthopedist felt that he had a
true contraction and needed heelcord lengthenings.  The doctor was willing to
let me continue with NMES and see what could happen.   The child was 6 years
old and had gone through growth spurts and should have had braces earlier but
he was doing so well it was put off.  So now with the loss of ROM after the
growth spurt, he was given hinged AFOs and continued with NMES to the
gastrocsoleus, gluteus maximus, lateral hamstrings and hip adductor muscles.
After a year his functional level improved so that he can pass all the GMFM
tests, except one task.  The test requires the child to jump up 12 inches but
the child could only jump up 10 inches.  He could also walk while barefoot
slowly with heel strike, jump forward 38 inches and not fall. He can hop on
either foot while bouncing a ball.  These are wonderful achievements for a
child with spastic diplegia.  I hope to write a case report on him soon.

Thank you for your interest and comments,

Judy Carmick

Dear Judy & others,

All this discussion of NMES (FES) is interesting. I actually did my PhD in
this area in the late '80s, when FES was much in vogue for paraplegic

I have to inject a firm dose of scepticism, I must say, partly based on my
experiences with paraplegia - although I can well imagine that the means of
action may be quite different in CP. I can imagine, for example, that you
would postulate such phenomena as carry-over and neural plasticity in CP,
which are not associated with FES for paraplegia.

With due respect, what we really need is some evidence. I wonder if you have
any cases, Judy, where you have documented the improvement by video? Such
cases would make excellent future Cases of the Week!


Hi Chris,

Yes, I do have videos of children, as well as my published papers.  I am
speaking only about cerebral palsy and feel that spasticity from a spinal
cord injury would be very different from the spasticity of an upper motor
neuron lesion (UMN).

There has been much research recently discussing spasticity in UMN lesion and
how it seems to be different that we once thought.  Two interesting studies
discuss the problems of having short muscles and the body's compensations.
Ada and O'Dwyer's findings suggests that instead of spasticity causing
contracture, contracture may actually potentiate spasticity in some patients.
 They felt that the amount of attention directed to reflex hyperexcitability
associated with spasticity is out of proportion with its effects and that
reduction of spasticity by itself may be of limited value

1. O'Dwyer NJ, Ada L. Reflex hyperexcitability and muscle contracture in
relation to spastic hypertonia.  Current opinion in neurology 1996;9:451-455.

2. Ada D, Vattanasilp W, O'Dwyer NJ, Crosbie J.  Does spasticity contribute
to walking dysfunction after stroke? J. Neurol Neurosurg Psychiatry

Judy Carmick


My name is Spiro Boulos, 26 y\o, from Israel, I am a physiotherapist (BPT)
making my master degree. I saw the videos of the 14 years old diplegic cp boy.

I think that achilles tendon release is not enough. In observation you can see real differences between the left side and the
right side.

1) in both sides there are muscle contraction.

                               right                   left

iliopsoas muscle         +                      ++
& rectus femoris

hamstrings               +                      ++

gastrocnemius &      +                      ++

2) he has dynamic spasticity in the adductor muscles which make him  internal
rotate and adduct his lt leg at the swing phase .at the same time he has
lack of control and weeknes in the  lt.gluteus medius muscle
( add.& ir at swing phase , pelvic drop at stance phase)

3) there is not full extension at both knees , at any phase  lt > rt

4) there is no dorsiflexion at his lt ankle which make him over flex his
hip and litle hicking at his pelvic while swinging his lt leg

5)  lt. initial contact = ball contact
6) pelvic in  a big  APT , and he has no counter rotation at walking.

I think he needs IP & RF release and achilles too ,then he should be treated
by physiotherapist for strengthing his pelpic muscles and keeping his ROM
functional as much as he can .
I wouldn't make him hamstings release at this stage cause from my experience
(not big experience yet) it is hard for a child to rehabilitate from large
amount of surgeries at the same time.

I hope that you write me back for more discussion , because it is interesting  me


Spiro Boulos

Back to Case