Comparison of
myelomeningocele (MMC) dictates the level of orthotic treat-
different systems to
classify the
neurological level of
lesion in patients with
myelomeningocele
Åsa Bartonek* PT, Institute for Surgical Research, Section of
Orthopaedics, Karolinska Institute;
Helena Saraste MD PhD, Karolinska Hospital, Stockholm,
Sweden;
Loretta M Knutson PhD PT PCS, University of Central
Arkansas, Conway, AR, USA.
*Correspondence to first author at MotorikLab, Astrid
Lindgren Children’s Hospital, Karolinska Hospital, S-17176
Stockholm, Sweden.
E-mail: asa.bartonek@karo.ki.se
In 73 patients with myelomeningocele (mean 17.2 years,
range 5 to 40 years) the classification of level of lesion was
studied according to six commonly used classification systems
and to the ambulation groups of Hoffer. The distribution of
the patients into classes of thoracic-level lesions was the same
for four classification systems in 10 patients and for two
classification systems into categories of level L3 in 14
patients. For the other patients discrepancies occurred
between systems for lesions of level L3 and downward. None
of the patients was consistently categorized in the functional
ambulation groups of Hoffer using all classification systems.
The results show that it is not possible to compare
neurological lesion levels classified according to the different
systems described in this study and consequently that the
distribution into the functional ambulation groups of Hoffer
varies. To enhance communication and facilitate comparing
the results of treatment we suggest using some basic criteria
for patient documentation.
796 Developmental Medicine & Child Neurology 1999, 41: 796–805
The extent of neurological deficiencies in patients with
ment, physical therapy, and surgical treatment, and predicts
future ambulation. Comparisons of treatment outcomes are
difficult and confusing due to the different classification sys-
tems used to describe the motor function and neurological
level of lesion in patients with MMC (Samuelsson and Skoog
1988, Fraser et al. 1992, Alman 1996, Lindseth 1996).
The most frequently used classification systems of the neu-
rological level of lesion in the literature are those of Sharrard
(Sharrard 1964, Stark and Baker 1967, Stark 1971, Mazur and
Menelaus 1991, Abraham et al. 1996, Duffy et al. 1996a),
Hoffer (Hoffer et al. 1973, DeSouza and Carroll 1976, Feiwell
et al. 1978, Stillwell and Menelaus 1983, Mazur et al. 1986,
Katz et al. 1997), and Lindseth (Lindseth 1976, Asher and
Olson 1983, Rosenstein et al. 1987, Samuelsson and Skoog
1988, Fraser et al. 1992, Mannor et al. 1996, Stott et al. 1996).
In the 1990s two additional classification systems were
described. Broughton and coworkers (1993) presented a
modification to the neurosegmental levels of Sharrard
(Broughton et al. 1994, Fraser et al. 1995, Frawley et al. 1996,
Marshall et al. 1996). McDonald and coworkers (1991) pro-
posed that children with MMC should be grouped according
to a pattern of specific muscle strength, which they consid-
ered a better predictor of ambulatory status than the neu-
rosegmental patterns described by Sharrard.
Ferrari and colleagues (1985) described motor deficiency
at different neurological levels with the aim of finding com-
pensatory solutions for orthotic selection using practical
guidelines (Michael et al. 1990, Ferrari and Lodesani 1991,
Doll and Michael 1994).
The aim of our study was to analyse the classification of
neurological level of lesion in patients with MMC using six
commonly used classification systems, and to study the dis-
tribution of neurological level of lesion in each of the ambu-
lation groups according to Hoffer (Hoffer et al. 1973).
Method
SUBJECTS
Seventy-three patients with MMC, 38 males and 35 females,
were included in the analysis. The patients were recruited from
two different studies on ambulation in patients with MMC: a
cross-sectional study of children born between 1985 and 1992
who were treated at the Karolinska Hospital in Stockholm,
Sweden (unpublished data); and patients from a follow-up
study (Bartonek et al. 1999) who were treated at the Karolinska
Hospital and the Örebro Medical Center Hospital. All patients
from these two studies who were between 5 and 40 years of age
(mean 17.2 years, range 5 to 40 years) and who had some
motor deficits in the lower limbs were included. One patient
with spasticity in ankle, knee, and hip joints was excluded
because of difficulty isolating movements of the lower limbs.
Patients with a history of orthopaedic surgery, which could
decrease muscle strength, were also excluded. The status of
functional ambulation according to Hoffer (Hoffer et al. 1973,
Table I) was based on information given by patients and par-
ents at the clinical examination. Forty-one subjects were com-
munity ambulators (mean age 17.9 years, range 5 to 40 years),
14 were household ambulators (mean age 19.9 years, range 7
to 39 years), 11 were non-functional ambulators (mean age 8.9
years, range 5 to 15 years), and seven were non-ambulators
(mean age 20.7 years, range 5 to 39 years).

CLINICAL EXAMINATION
All patients were examined clinically at Karolinska
Hospital, Stockholm, and Örebro Medical Center by the
same trained physiotherapist (ÅB). Muscle strength was
assessed in both lower limbs according to a 0 to 5 graded
scale (Daniels et al. 1995). Additionally pelvic elevation
strength was examined according to the classification sys-
tem of Ferrari et al. (1985). Voluntary movement control
was judged to be present when the patient achieved grade
3 on manual muscle testing as suggested by Asher and
Olson’s (1983) interpretation of Lindseth’s (1976) recom-
mendation. Presence of spasticity in ankle, knee, and hip
joints was documented if the patients showed one or more
of the following signs: clonus, clasp-knife response
(catch), resistance throughout passive range of motion,
and static abnormal posturing with increased muscle ten-
sion despite full passive motion.
The results of the physical examination were used to
classify each patient according to each of the following six
systems.
DESCRIPTION OF THE CLASSIFICATION SYSTEMS
Sharrard (1964)
Sharrard’s classification was based on nerve root stimula-
tion in 41 infants with MMC who had normal or potentially
normal innervation of the lower limbs, performed at opera-
tion within 48 hours of birth. In addition, Sharrard per-
formed a clinical study of muscle strength in 253 children
with MMC. The resulting classification system defines a sin-
gle thoracic level, five lumbar, and two sacral levels.
Hoffer et al. (1973)
Hoffer and colleagues’ classification was based on studies of
56 patients with MMC aged between 5 and 42 years. The
resulting classification system has four categories: thoracic,
upper lumbar, lower lumbar, and sacral.
Lindseth (1976)
Lindseth’s description of a functional classification of motor
paralysis was based on voluntary joint control. In describing
the extent of neurological deficit, one thoracic level, five
lumbar levels, and one sacral level are given.
Ferrari et al. (1985)
Ferrari and coworkers developed their categories based on
clinical examinations of 200 patients between the ages of 1
and 20 years. Ferrari’s classification is part of a more elaborate
theoretical model for analysis of upright posture and walking
in patients with MMC. The motor levels are described as five
lumbar and three sacral levels. However, in the present study
we have modified Ferrari’s published classification to include
a thoracic level, consistent with his clinical practice in which
he uses thoracic level to describe a deficit of abdominal mus-
cles (personal communication).
McDonald et al. (1991)
McDonald and colleagues studied strength and mobility in
four groups of patients at the age of 5 years and after, with
mean and median ages of 174 months (14.5 years) and 171
months, respectively, at the last examination. The patients had
the following muscle patterns: (1) grade 0 to 1 iliopsoas and
quadriceps; (2) grade 2 to 3 iliopsoas; (3) grade 4 to 5 iliop-
Classification of Lesion Levels in Myelomeningocele Åsa Bartonek et al. 797
14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
soas and quadriceps, and 0 to 3 gluteus medius; and (4) grade
4 to 5 iliopsoas, quadriceps, and gluteus medius. McDonald
proposed specific patterns of lower-limb muscle strength to
predict future ambulation. These patterns included: partial or
complete reliance on a wheelchair, with muscle strength in
iliopsoas grade 0 to 3; community ambulation without com-
plete reliance on a wheelchair, with muscle strength in iliop-
soas and quadriceps grade 4 to 5; community ambulation
without aids or braces, with muscle strength in gluteal mus-
cles and tibialis anterior grade 4 to 5 (McDonald et al. 1991).
Broughton et al. (1993)
Broughton and colleagues’ neurosegmental levels represent
a modification of Sharrard’s classification. One thoracic level,
five lumbar levels, two sacral levels, and one category ‘no
loss’ are described.
Asymmetrical motor function
For those patients who showed asymmetrical motor function,
the neurological level of lesion was classified with respect to the
limb with least motor function according to all of the authors
of the present study.
A detailed summary of the classification systems is provid-
ed in Table II.
Results
The classification systems of Sharrard and Ferrari classified
21 and 23 patients (out of 73), respectively, with motor func-
tion covering two categories of neurosegmental level. These
patients were classified according to the lower function. The
motor function of all the patients corresponds to one of the
levels defined by Lindseth, Broughton, and Hoffer, as well as
one of the groups of McDonald
Spasticity in the lower limbs occurred in 23 patients. See
Table III for the distribution of patients with MMC according to
Sharrard, Hoffer, Lindseth, Broughton, and Ferrari. The
results according to McDonald et al. are presented in Table IV.
AGREEMENT
Agreement in the distribution of the patients into thoracic-
level categories was found between the classifications of
Sharrard, Lindseth, and Broughton in 10 patients. Hoffer’s
Table I: Functional ambulation classification according to
Hoffera
Community ambulators
Patients walk indoors and outdoors for most activities; may need
crutches, braces, or both. Wheelchair used only for long trips
out of community.
Household ambulators
Patients walk only indoors and with orthoses. Able to get in and
out of chair and bed with little, if any, assistance. May use
wheelchair for some indoor activities at home and school.
Wheelchair is used for all activities in community.
Non-functional ambulators
Patients walk during therapy session at home, in school, or in
hospital. Wheelchair used for all other transportation.
Non-ambulators
Patients are mobile only via a wheelchair but usually can transfer
from chair to bed.
 14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table II: Descriptions of the classification systems

Lesion Sharrard (1964) Hoffer et al. Lindseth Broughton et al. Ferrari et al. McDonald et al.
level (1973) (1976) (1993) (1985) (1991)

Thoracic Below 12th thoracic No sensation No motor function No active Deficit of

root, complete below hips and no in lower limbs. movement of abdominal
paralysis in all muscles power in muscles the hip. muscles.
in lower limb crossing hip joint
including hip flexors. or distal to it.
Lumbar L1: below 1st Upper lumbar: L1: voluntary L1: iliopsoas grade L1: insufficient Muscle strength
lumbar root, weak to one or more of control of hip 2 or better. quadratus in iliopsoas grade
moderate hip flexor following: some flexion (weak); lumborum. 0–3 (partial or
power. sensation below probable muscles complete reliance
hip joint, some functioning: on wheelchair).
power in hip iliopsoas.
L2: below 2nd lumbar adductors or L2: voluntary L2: iliopsoas, L2: partially
root, strong hip flexion, flexors, or in control of hip sartorius, and compromised:
moderately strong extensors of flexion (strong); adductus all grade pelvic elevators
hip adduction, and knees. probable muscles 3 or better. (quadratus
quadriceps (rectus functioning: lumborum) seriously
femoris moderately iliopsoas and compromised:
innervated). sartorius. knee flexors.
L3: below 3rd lumbar L3: voluntary L3: quadriceps L3: partially Muscle strength in
root, normal hip flexion control of knee grade 3 or better, compromised: hip iliopsoas and
and adduction. Almost extension; also meet criteria flexors (superficial quadriceps grade
normal power in knee probable muscles for L2. and deep) seriously 4–5 (no complete
extension (some functioning: compromised: reliance on
denervation in adductor quadriceps. knee extensors. wheelchair, most
magnus and quadriceps). become community
L4: below 4th lumbar Lower lumbar: L4: voluntary L4: medial L4: partially ambulators).
root, normal hip flexion, one or more of control of knee hamstrings or compromised: hip
adduction, and knee following: power flexion; probable tibialis anterior adductors (except
extension. Weak in flexors of knee, muscles grade 3 or better; for adductus
abduction (some or in dorsiflexors functioning: also meet criteria longus) quadriceps.
activity by tensor of ankle, or in medial hamstring for L3.
fasciae latae). Some abductors of hips. group.
weak action in medial
hamstrings. Strong
dorsiflexion and
inversion.
L5: below 5th lumbar L5: voluntary L5: lateral L5: seriously
root, normal hip flexion control of foot hamstrings or compromised: hip
and adduction. dorsiflexion and tibialis anterior extensors (including
Moderately strong eversion; probable grade 3 or better; gluteus minimus)
abduction (by tensor functioning also meet criteria hip abductors
fasciae latae, gluteus muscles: tibialis for L4 plus one of (including tensor
medius and minimus). anterior and following three: fascia lata) knee
No active hip extension. peroneals. gluteus medius flexors (sometimes
Normal knee extension grade 3 or better, traces of
and moderately strong peroneus tertius semitendinosus)
knee flexion by medial grade 4 or better, leg (calf) and foot
hamstring group). tibialis posterior muscles (sometimes
Normal power of grade 3 or better. partial tibialis anterior).
inversion by tibialis
anterior and tibialis
posterior. Moderate
power of eversion by
peroneal muscles.
Strong dorsiflexion. No
activity in plantarflexors
(except for some
palpable activity).
Sacral S1: below 1st sacral Sacral: (one or Sacral: voluntary S1: two of the S1: partially Muscle strength in
root, normal hip flexion, more of the control of foot following three: compromised: gluteal muscles and
adduction, and following): power plantarflexion and gastrocnemius/ gluteus medius tibialis anterior 4–5
abduction. Moderate in the plantarflexors eversion; probably soleus grade 2 or medial hamstrings (community
power of hip extension of ankle or toes, functioning better, gluteus foot dorsiflexors ambulators without

798 Developmental Medicine & Child Neurology 1999, 41: 796–805


classification agreed in nine of 10 patients; one patient had
some hip-flexor power and thus was classified as ‘upper lum-
bar level’. Ferrari’s classification agreed in three of 10 patients
categorized as having thoracic-level lesions with trunk flexion
muscles involved in the paresis. Of the remaining seven
patients, two were classified as L1 level without any lower-
limb or pelvic elevation muscle function, and five patients as
level L2 with pelvic elevation muscle function. Agreement was
also found between Lindseth and Broughton in 14 patients at
L3 level with knee extension strength grade 3 or better.
14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Among all other patients there was a lack of congruence in
classes L3 level and downward.
DISTRIBUTION ACCORDING TO HOFFER
Within each of the ambulation groups none of the patients
could be described as having the same neurological level of
lesion according to all six classification systems. One non-
functional ambulator and two non-ambulators were classi-
fied as thoracic level by five authors of the classification
system (Tables V to VIII).

Table II: (continued)

by gluteus maximus. or in extensors muscles: medius grade 3 or seriously aids or braces).

Strong knee flexion, of hips. gastrocnemius, better, gluteus compromised:
distinct weakness of soleus, and tibialis maximus grade 2 hip extensors
biceps femoris. Normal posterior. or better; also (gluteus maximus
dorsiflexion, inversion, meet criteria for L5. and medius) foot
and eversion. Moderate plantar flexors.
plantarflexion
(gastrocnemius and
soleus partially),
normal toe extension,
all toe flexors paralysed
except for flexor
hallucis longus.
Intrinsic muscles
completely paralysed
(except for abductor
hallucis and flexor
hallucis brevis)
S2: weakness of intrinsic S2: gastrocnemuis/ S2: partially
muscles of the foot. soleus grade 3 or compromised: hip
better and gluteus extensors (especially
medius and gluteus gluteus maximus)
maximus grade 4 hamstrings (especially
or better; also meet biceps femoris)
criteria for S1. foot plantarflexors
S3: deficit of intrinsic
muscles.
No loss All leg muscles have
normal strength.

Table III: Distribution of 73 patients with MMC according to

the classification of level of neurological lesion by Sharrard,
Hoffer, Lindseth, Broughton, and Ferrari Table IV: Distribution of 73 patients with MMC according to
McDonald et al. (1991)
Level Sharrard Hoffer Lindseth Broughton Ferrari
Patterna N
Thoracic 10 9 10 10 3
L1 0 0 0 2 1 13
L2 0 0 0 5 2 41
Upper lumbar 6 3 19
L3 3 14 14 1 a Pattern 1, 2, and 3 was not used by McDonald. This terminology
L4 20 21 15 2
has been adopted for ease of presentation.
L5 14 10 11 20
Pattern 1, muscle strength in iliopsoas grade 0–3, with partial or
Lower lumbar 14
complete reliance on wheelchair.
S1 12 9 15
Pattern 2, muscle strength in iliopsoas and quadriceps grade 4–5,
S2 14 1 11
with community ambulation without complete reliance on
S3 14
wheelchair.
Sacral 44 18
Pattern 3, muscle strength in gluteal muscles and tibialis anterior
Normal / no loss 13
4–5, with community ambulation without aids or braces.

Classification of Lesion Levels in Myelomeningocele Åsa Bartonek et al. 799


Discussion
MOTOR FUNCTION
Motor performance in patients with MMC is affected not only
by the level of motor paresis, but also by other factors such as
age, obesity, cognitive status, and motivation. Ambulation is
also influenced by other related medical problems such as
spasticity, upper motoneuron involvement, and orthopaedic
deformities. Lindseth (1996) reported problems using motor
function to describe the level of paralysis when CNS dysfunc-
tion is common. He recommended classifying function by
sensory level rather than by motor level to improve descrip-
tion of clinical status. Knutson and Clark (1991) suggested
using ‘functional motor level’ when a discrepancy occurred
between the child’s motor level and motor function.
14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Duckworth and Brown (1970) reported difficulty in dis-
tinguishing between voluntary and reflex activity. Mazur and
Menelaus (1986) stressed the importance of ‘recognising the
difference between a spastic muscle under no voluntary con-
trol and an overpowering muscle under voluntary control’
(p 217). Mazur and Menelaus (1991) suggested a patient
classification based on whether limbs were spastic or flaccid
in addition to the examination of muscle strength. We agree
with this because presence of spasticity might have a negative
influence on functional ambulation.
In the present study, 23 out of 73 patients had varying
degrees of spasticity in the lower limbs. Of those 23, only
one patient had spastic involvement in muscles crossing all
lower-limb joints. However, this female subject was not

Table V: Distribution of 41 community ambulators into classes of level of neurological

lesion according to Sharrard, Hoffer, Lindseth, Broughton, Ferrari, and according to
McDonald’s specific pattern of muscle strength

Subject Age Sex Sharrard Hoffer Lindseth Broughton Ferrari McDonalda

1 5 M S2 S S N S3 3
2 5 M L4 S L4 L4 L5 2
3 5 M L5 S L4 L5 S1 2
4 6 M S2 S S N S3 3
5 6 M L5 S L5 L4 S1 2
6 6 M L5 S L3 L3 S1 2
7 6 M L5 S L4 L5 S1 2
8 6 F L5 S L4 L4 S1 2
9 8 M S1 S S S1 S2 2
10 8 F L5 S L4 L5 S1 2
11 8 M L4 S L4 L4 L5 2
12 8 M S1 S L4 S1 S2 2
13 9 F S1 S L5 S1 S1 2
14 9 F S2 S S N S3 3
15 14 M S2 S S N S3 3
16 14 F L5 S L4 L5 S1 2
17 14 F L5 S L4 L5 S1 2
18 15 M S2 S S S1 S3 3
19 15 F S1 S L4 L5 S2 2
20 16 M S1 S S S1 S2 3
21 19 M S1 S S N S2 3
22 20 F L5 S L5 S1 S1 3
23 20 F S1 S L5 S1 S2 3
24 21 M S2 S S N S3 3
25 21 F L5 S L3 L3 S1 2
26 22 M S1 S L4 L5 S2 2
27 22 F L5 S L5 L5 S1 2
28 22 M L4 S L5 L4 S1 2
29 23 F S2 S S N S3 3
30 24 F S2 S S S2 S3 3
31 24 F S1 S L5 S1 S2 3
32 25 F S2 S S N S3 3
33 26 M L4 S L3 L3 L5 2
34 28 M L5 S L5 L5 S1 2
35 28 M S1 S S N S2 2
36 30 M S2 S S N S2 3
37 31 F S2 S S N S2 3
38 33 F S1 S S L5 S2 2
39 36 M S1 S L5 L4 S2 3
40 37 F S2 S S N S3 3
41 40 F S2 S S N S3 3

N, no loss; S, sacral.
a See Table III for descriptions of the numbers.

800 Developmental Medicine & Child Neurology 1999, 41: 796–805


excluded from the study because her motor function was at
the thoracic level.
The aim of this study was to compare classification
results based on the same clinical examination. The possi-
ble error caused by presence of spasticity was a systematic
error and remained the same when using each of the classi-
fication systems.
ASSESSMENT OF MUSCLE STRENGTH
All classification systems studied are based on the assess-
ment of muscle strength. When assessing muscle strength
in children, physical therapists are markedly more confi-
dent of the child’s performance if the child is aged 5 years
and older (McDonald et al. 1986). Therefore, we excluded
all patients younger than 5 years. To avoid the effect of
degenerative processes, all patients older than 40 years
were excluded.
14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Hoffer and colleagues (1973) included assessment of
presence of sensation at thoracic and upper-lumbar level.
However, they considered sensation as variable and difficult
to assess in lumbar and sacral areas. In the present study
assessment of sensation was not included because it was not
used in any of the other classification systems.
ASYMMETRICAL MOTOR FUNCTION
The classification systems handle the presence of asymmetry
differently. In some studies, the more caudal level of defect,
which corresponds to the best level of function, was used to
define the neurological lesion (Samuelsson and Skoog 1988,
Fraser et al. 1992). Hoffer and coworkers (1973) also record-
ed the better level of function when minor variations in motor
function occurred in the lower limbs. In the work of Sharrard
(1964), a right and left limb motor level was assigned.
Broughton and colleagues (1993) recorded muscle strength

Table VI: Classification of the neurological lesion level in 14 household ambulators according to
Sharrard, Hoffer, Lindseth, Broughton, Ferrari, and according to McDonald’s specific pattern of muscle
strength

Subject Age Sex Sharrard Hoffer Lindseth Broughton Ferrari McDonalda

1 7 M L4 S L3 L3 L5 2
2 7 F T T T T L2 1
3 9 M L4 LL L4 L4 L5 2
4 10 F L4 LL L4 L4 L5 2
5 10 F L4 LL L3 L3 L5 2
6 13 M L4 LL L3 L3 L5 2
7 15 M L5 S L4 L5 S1 2
8 20 M L4 LL L3 L3 L5 2
9 25 F T UL T T L2 1
10 25 F L4 LL L4 L4 L5 2
11 27 F L4 LL L4 L4 L5 2
12 35 F L4 LL L4 L4 L5 2
13 36 F L4 UL L3 L3 L5 2
14 39 F S1 S L5 S1 S2 2

S, sacral; LL, lower lumbar; UL, upper lumbar; T, thoracic.

a See Table III for descriptions of the numbers.

Table VII: Classification of the neurological lesion level in 11 non-functional ambulators according to
Sharrard, Hoffer, Lindseth, Broughton, Ferrari, and according to McDonald’s specific pattern of muscle
strength

Subject Age Sex Sharrard Hoffer Lindseth Broughton Ferrari McDonalda

1 5 M T T T T L1 1
2 7 F L4 UL L3 L3 L5 2
3 7 F L4 UL L3 L3 L5 2
4 8 M L4 LL L4 L4 L5 2
5 8 M T T T T L1 1
6 8 M T T T T L2 1
7 8 M L3 UL L3 L3 L4 1
8 9 M L4 LL L3 L3 L5 2
9 9 F T T T T T 1
10 14 F L4 LL L4 L4 L5 2
11 15 M L4 LL L4 L4 L5 2

LL, lower lumbar; UL, upper lumbar; T, thoracic.

a See Table III for descriptions of the numbers.

Classification of Lesion Levels in Myelomeningocele Åsa Bartonek et al. 801


and determined an ‘overall neurosegmental level’ for each
child for all muscles of the lower limbs, excluding children
with gross asymmetry. In the description of specific patterns
of lower-limb muscle strength, McDonald and coworkers
(1991) did not seem to address the management of asymme-
try. As Ferrari’s classification aims to examine the basis for
orthotic selection, the levels of both limbs are given (personal
communication).
To have the same basis for comparison of the different
classification systems we classified the patients according to
the limb with the least motor function.
VARIATION BETWEEN CLASSIFICATION SYSTEMS
There is a confusing variation in the definitions of classes of
lesion level and in the distribution of the patients into the
classes. The variation between the classes is not based on a
systematic error, but varies to the degree that it was impos-
sible to classify all subjects of this study consistently in all
systems, with some exceptions.
Slight differences occurred between the definitions of
thoracic class as defined by Sharrard, Hoffer, Lindseth,
Broughton, and Ferrari, but conversion between their classi-
fication systems is possible. As McDonald’s aim is to predict
future ambulation, this classification does not include tho-
racic level. At L1 level some agreement occurs between
Sharrard, Lindseth, and Broughton in the assessment of hip-
flexor strength. The definition of L3 level can be converted
between Lindseth and Broughton.
However, between all other classification systems the
definitions vary and it was not possible to convert the class-
es to one another when comparing results. As an example,
using the classifications of Lindseth and Broughton, the
same 11 patients are described as having L4 motor func-
tion. All patients had knee flexion grade 3 or better.
However, ankle dorsiflexion function varied. Broughton’s
L4 level is defined as knee flexion grade 3 or better, or foot
dorsiflexion grade 3 or better. According to Lindseth these
two functions are separated by a motor level with voluntary
control of ankle dorsiflexion representing the L5 level.
At sacral level Sharrard, Broughton, and Ferrari used two
and three levels respectively, whereas Hoffer used one defi-
nition of ‘power in the plantarflexors’. In the literature,
patients with sacral level lesions were described as having
gastrocnemius muscle strength greater than ‘poor’ or ‘no
strength’ (Park et al. 1997), whereas those authors using
Table VIII: Distribution of seven non-ambulating patients in classes of the neurological lesion level
according to Sharrard, Hoffer, Lindseth, Broughton, Ferrari, and according to McDonald’s specific
pattern of muscle strength
Subject Age Sex Sharrard Hoffer Lindseth
1 5 M T T T
2 6 F T T T
3 21 M T T T
4 21 F L3 UL L3
5 23 M L4 LL L3
6 30 F T T T
7 39 M L3 LL L4
LL, lower lumbar; UL, upper lumbar; T, thoracic.
a See Table III for descriptions of the numbers.
802 Developmental Medicine & Child Neurology 1999, 41: 796–805
14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Lindseth’s classification require the sacral level to have a
gastrocnemius muscle grade 3 or better on a 0 to 5 scale.
The results show that it is not possible to compare the
neurological lesion levels classified according to the differ-
ent classification systems.
DISTRIBUTION ACCORDING TO HOFFER
Due to the variation in the definitions of the classes of neuro-
logical lesion level, there were variations in the distribution
into the functional ambulation groups of Hoffer.
Community ambulators
None of the community ambulators was given the same
lesion level by all classification systems. According to
Hoffer’s classification system of neurological level of lesion,
all community ambulators were classified as sacral level.
All patients had quadriceps strength grade 4 to 5. This,
however, would not have been obvious with the results
according to only Lindseth or Broughton, because the defin-
ition of quadriceps strength as voluntary control of knee
extension of grade 3 or better is used by those two authors.
This definition of quadriceps strength remains throughout
all ambulatory levels.
Household ambulators
None of these patients was given the same lesion level by all
classification systems.
Among the household ambulators all but two patients had
quadriceps strength as described in the group of community
ambulators. One patient was classified as having a thoracic-
level lesion by Sharrard, Hoffer, Lindseth, and Broughton
and another was classified as having a thoracic-level lesion by
Sharrard, Lindseth, and Broughton, whereas Hoffer classi-
fied this patient with upper-lumbar level lesion. Both those
patients were classified by Ferrari as L2 level, which indicates
presence of pelvic elevator strength and might explain their
household ambulation.
Non-functional ambulators
One patient classified as having a lesion at thoracic level was
classified as a non-functional ambulator by five classification
systems.
Among the non-functional ambulators, i.e. walking only
during therapy session, six of 11 patients had quadriceps
strength grade 4 to 5. One patient had weakness in knee
Broughton Ferrari McDonalda
T T 1
T T 1
T L2 1
L3 L4 1
L3 L5 2
T L2 1
L4 L4 1

extensors, but still-functioning hip flexors. Four patients had
thoracic-level lesions and one of those had hip-elevation
function.
Non-ambulators
Two patients with thoracic-level lesions were classified as
non-ambulators by five classification systems.
Among the non-ambulators only one male subject had
knee-extension power of grade 4 to 5. He had lost his ambu-
latory status due to a stroke. Two patients had weakness in
knee extensors. Four patients were classified as having tho-
racic-level lesions by Sharrard, Hoffer, Lindseth, and
Broughton. According to Ferrari two were classified at tho-
racic-lesion level and two at L2 level.
Ambulatory goals according to McDonald
The goal of McDonald and colleagues (1991) is not to classify
motor level, but to predict ultimate walking ability and estab-
lish ambulatory goals. Although prognosis and goal setting
for ambulation was not needed for the adults in our study, we
used McDonald’s system of assessment in both children and
adults to compare with the results of the other authors and to
relate muscle strength to achieved ambulatory function.
Fifty-three of those 60 patients showing a muscle-
strength pattern of gluteal muscles and tibialis anterior
grade 4 to 5 (pattern 3), and iliopsoas and quadriceps grade
4 to 5 (pattern 2) had achieved community ambulation
without complete reliance on a wheelchair as correctly pre-
dicted by McDonald. Seven patients were non-functional
ambulators or non-ambulators and had not achieved the
predicted goal.
Of those 13 patients with muscle strength in iliopsoas
grade 0 to 3 (pattern 1), predicted to be partially or com-
pletely reliant on a wheelchair, all but two patients were non-
functional ambulators or non-ambulators. Two patients
were household ambulators, who were also partially reliant
on a wheelchair, although both of them had pelvic elevator
strength.
CRITICAL POINTS IN FUNCTIONAL ASSESSMENT OF MUSCLE
STRENGTH
Pelvic elevation
Normally the muscle strength of the lower-limb muscles is
assessed as the basis for defining the level of neurological
lesion. In this study we also assessed strength in pelvic eleva-
tor muscles because the classification used by Ferrari also
distinguishes pelvic elevation based on the assumption that
quadratus-lumborum strength is a determining factor in
ambulatory function for patients with high-lumbar lesions.
Among 10 patients in our study classified as having thoracic-
level lesions, five showed pelvic elevation strength grade 3 or
better, and thus were classified as having L2-level lesions
according to Ferrari.
Knee extension
Agreement was found in 14 patients with lesions at L3 level
who were assessed as having voluntary control of knee
extension and quadriceps muscle strength of grade 3 or bet-
ter by Lindseth and Broughton respectively. However, it was
not possible to find consensus between the six classification
systems concerning functional assessment of knee exten-
sion in an upright position. According to Lindseth, at the L3
Classification of Lesion Levels in Myelomeningocele Åsa Bartonek et al. 803
14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
level the quadriceps are ‘sufficiently well developed to be
functional during gait’ (Lindseth 1976), although sufficient
sensation and motor control of the knee is not present until
L4 to L5 levels (1996). Asher and Olson (1983) interpreted
Lindseth’s definition to mean manual muscle test grade 3 or
better (i.e. antigravity voluntary contraction). Broughton
and colleagues (1993, 1994) also defined the L3 level as
quadriceps strength grade 3 or better, i.e. with the possibili-
ty of some weakness. In their study they included patients
with L3-level lesions(quadriceps strength grade 3 or better)
in the high-level group. In the classification of Hoffer, the
upper-lumbar level is defined by ‘some power’ in the exten-
sors of the knees without giving a closer definition.
According to Sharrard (1964) knee-extension power is
defined as almost normal at L3 level and normal at L4 level.
Ferrari defines level L5 as being the normal strength of knee
extensors and L4 as partial. McDonald and coworkers
(1991) define group 2 of muscle-strength patterns as
quadriceps power grade 4 to 5.
A simplified classification was used by Smith and Smith
(1973) who divided the patients into a low- and high-lesion
group with the distinguishing feature being presence or
absence of an active quadriceps. Gaff and colleagues (1984)
used a high-lumbar group ‘without quadriceps’ and a lum-
bar group with quadriceps strength grade 3 or better
(according to British MRC scale). Fraser and coworkers
(1995) found it useful to expound upon Smith and Smith’s
(1973) description of quadriceps presence or absence
being key to defining low- or high-level of lesion. By also
defining the manual muscle test grade of the quadriceps
muscle, they described grade 3 or less as upper neurologi-
cal level (high-level lesions) and grade 4 to 5 as lower neu-
rological level (low-level lesions). We agree with this
presentation because there should be a functional differ-
ence between antigravity quadriceps strength grade 3 and
grade 4 to 5. Thus, it is important also to classify the
strength of the quadriceps muscle acting against gravity in a
weight-bearing position. This was confirmed by Mazur and
Menelaus (1991) who proposed that the quadriceps muscle
must have grade 4 or 5 strength for a patient to walk with
ankle–foot orthoses.
Hip abduction
Knowledge of gluteus-medius, gluteus-maximus, and ante-
rior-tibialis muscle strength helps in the prediction of ambu-
lation, with or without walking aids or braces (McDonald et
al. 1991). In the classification of Lindseth no assessment of
gluteus-maximus or -medius strength is included. However,
using the work of Asher and Olson (1983) who extended
Lindseth’s system, the clinician could expect hip abduction
power at manual muscle test grade 2 in patients with L4
lesions. Children with hip-abductor muscle weakness have
an energy-expensive lateral trunk sway (Duffy et al. 1996b).
Therefore, when planning future gait pattern as well as
orthopaedic and orthotic treatment, hip-abductor strength
should be a point of focus. In the classifications of Sharrard,
Broughton, Ferrari, and McDonald, this was distinguished.
Plantarflexion
Plantarflexor strength is defined by Hoffer as ‘power in the
plantarflexors’ (Hoffer et al. 1973, p 139), by Lindseth as ‘vol-
untary control of gastrocnemius and soleus’ (Lindseth 1976,

p 77), and by Broughton as S2 level with ‘gastrocnemius/
soleus strength grade 3 or better’ (Broughton et al. 1993, p
762). Plantarflexion strength is not explicitly mentioned by
McDonald, whereas discriminated as without deficiency by
Sharrard and Ferrari between the S2 and S1 levels respectively
S3 and S2 levels. To stabilize the ankle joint and control the
tibia’s forward tilt during stance, foot plantarflexor strength of
more than grade 3 is necessary.
Spasticity
Mazur and Menelaus (1991) reported that the orthopaedic
and orthotic management of patients with spasticity is more
difficult than of those with flaccid paresis, and that spasticity
causes more lower-limb contractures and deformities, and
orthotic fittings become difficult. Thus, recognition and
assessment of spasticity is important, both for detecting
changes of the paresis and for the planning of an habilitation
programme.
Functional motor level
On the basis of the diagnosis of level of neurological lesion, it
is not possible to identify functional motor ability using any
of the classification systems. In the case of a discrepancy
between muscle activity and the patient’s ability to translate
that activity into function, the functional ambulatory level
should be documented.
SUGGESTIONS TO ENHANCE COMMUNICATION
To enable comparison of the outcome of the different classifi-
cation systems we suggest the following: (1) the classifica-
tion of motor function based on muscle test according to the
less functional limb; (2) assessment of pelvic-
elevation function to distinguish patients with quadratus-
lumborum strength from patients with thoracic-level
lesions; (3) discrimination between knee-extensor strength
grade 3 or less and grade 4 to 5, with or without ability of
functional weight-bearing during stance; (4) assessment of
hip-abductor strength with regard to hip status, ambulatory
function, and orthotic treatment; (5) discrimination of foot
plantarflexion strength grade 3 or less versus grade 4 to 5,
and ability to control the tibia’s forward tilt during stance; (6)
assessment of flaccid or spastic paresis of the lower limbs; (7)
functional ambulation status in relation to the level of neuro-
logical lesion.
Accepted for publication 17th June 1999.
Acknowledgements
The authors would like to acknowledge those who provided the
funding to complete this work. In Sweden, support was provided to
Å Bartonek and H Saraste by Norrbacka Eugenia Foundation. In the
US, support was provided to L M Knutson by the National Institute
on Disability and Rehabilitation Research through a Mary Switzer
Distinguished Fellowship.
References
Abraham E, Lubicky JP, Songer MN, Millar EA. (1996) Supramalleolar
osteotomy for ankle valgus in myelomeningocele. Journal of
Pediatric Orthopaedics 16: 774–81.
Alman BA, Bhandarl M, Wright JG. (1996) Function of dislocated
hips in children with lower level spina bifida. Journal of Bone
and Joint Surgery 78-B: 294–8.
Asher M, Olson J. (1983) Factors affecting the ambulatory status of
patients with spina bifida cystica. Journal of Bone and Joint
Surgery 65-A: 350–6.
804 Developmental Medicine & Child Neurology 1999, 41: 796–805
14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Bartonek Å, Saraste H, Samuelsson L, Skoog M. (1999) Ambulation
in patients with myelomeningocele: a 12-year follow-up. Journal
of Pediatric Orthopaedics 19: 202–6.
Broughton NS, Malcolm BM, Cole WG, Shurtleff DB. (1993) The
natural history of hip deformity in myelomeningocele. Journal of
Bone and Joint Surgery 75-B: 760–3.
— Graham G, Menelaus MM. (1994) The high incidence of foot
deformity in patients with high-level spina bifida. Journal of Bone
and Joint Surgery 76-B: 548–50.
Daniels L, William M, Worthingham CWB. (1995) Muscle Testing.
Techniques of Manual Examination. Philadelphia: WB
Saunders.
DeSouza LJ, Carroll N. (1976) Ambulation of the braced
myelomeningocele patient. Journal of Bone and Joint Surgery
58: 1112–8.
Doll B, Michael T. (1994) Aktueller Stand der apparativen
Versorgung von Kindern mit Myelomeningocele. Zeitschrift für
Orthopädie 132: 201–6.
Duckworth T, Brown BH. (1970) Changes in muscle activity
following early closure in myelomeningocele. Developmental
Medicine & Child Neurology 12: 39–45.
Duffy CM, Hill AE, Cosgrove AP, Corry IS, Graham HK. (1996a) The
influence of abductor weakness on gait in spina bifida. Gait and
Posture 4: 34–8.
— — — — — (1996b) Energy consumption in children with spina
bifida and cerebral palsy: a comparative study. Developmental
Medicine & Child Neurology 38: 238–43.
Feiwell E, Sakai D, Blatt T. (1978) The effect of hip reduction on
function in patients with myelomeningocele. Journal of Bone
and Joint Surgery 60-A: 169–73.
Ferrari A, Boccardi S, Licari V. (1985) La stazione eretta ed il
cammino nella spina bifida. Paper published in the
proceedings from the 14 meetings of the Italian Society of
Physical Medicine and Rehabilitation. Ferrara, September
1985. p 167–204.
— Lodesani M. (1991) Il trattamento riabilitativo della spina
bifida. Urodinamico clinica in urologia pediatrica. Paper
published in the proceedings from the 14 meetings of the
Italian Society of Urodynamics. Rome, November 1990.
p 281–94.
Fraser RK, Hoffman EB, Sparks LT, Buccimazza SS. (1992) The
unstable hip and mid-lumbar myelomeningocele. Journal of
Bone and Joint Surgery 74-B: 143–6.
— Bourke HM, Broughton NS, Malcolm MB. (1995) Unilateral
dislocation of the hip in spina bifida. Journal of Bone and Joint
Surgery 77-B: 615–9.
Frawley PA, Broughton NS, Menelaus MB. (1996) Anterior release
for fixed deformity of the hip in spina bifida. Journal of Bone and
Joint Surgery 78-B: 299–302.
Gaff JE, Robinson JM, Parker PM. (1984) The walking ability of 14- to
17-year-old teenagers with spina bifida – a physiotherapy study.
Physiotherapy 70: 473–4.
Hoffer M, Feiwell E, Perry J, Bonnett C. (1973) Functional
ambulation in patients with myelomeningocele. Journal of Bone
and Joint Surgery 55-A: 137–48.
Katz DE, Haideri N, Song K, Wyrick P. (1997) Comparative study of
conventional hip-knee-ankle-foot orthoses versus reciprocating-
gait orthoses for children with high-level paraparesis. Journal of
Pediatric Orthopaedics 17: 377–86.
Knutson LM, Clark DE. (1991) Orthotic devices for ambulation in
children with cerebral palsy and myelomeningocele. Physical
Therapy 12: 947–60.
Lindseth RE. (1976) Treatment of the lower extremity in children
paralyzed by myelomeningocele (birth to 18 months). American
Academy of Orthopaedic Surgeons Instructional Course
Lectures 25: 76–82.
— (1996) Myelomeningocele. In: Morrissy RT, Weinstein SL,
editors. Lovell and Winter’s Pediatric Orthopaedics, fourth
edition. Philadelphia: Lippincott-Raven. p 503–36.
Mannor DA, Weinstein SL, Dietz FR. (1996) Long-term follow-up of
chiari pelvic osteotomy in myelomeningocele. Journal of
Pediatric Orthopaedics 16: 769–73.
Marshall PD, Broughton NS, Menelaus MB, Graham HK. (1996)
Surgical release of knee flexion contractures in myelomeningocele.
Journal of Bone and Joint Surgery 78-B: 912–6.
 14698749, 1999, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.1999.tb00545.x by CAPES, Wiley Online Library on [15/08/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Mazur JM, Stillwell A, Menelaus M. (1986) The significance of
spasticity in the upper and lower limbs in myelomeningocele.
Journal of Bone and Joint Surgery 68-B: 213–7.
— Menelaus MB. (1991) Neurologic status of spina bifida patients
and the orthopedic surgeon. Clinical Orthopaedics and Related
Research 264: 54–64.
McDonald CM, Jaffe KM, Shurtleff DB. (1986) Assessment of muscle
strength in children with myelomeningocele: accuracy and
stability of measurements over time. Archives of Physical
Medicine Rehabilitation 67: 855–61.
— — Mosca VS, Shurtleff DB. (1991) Ambulatory outcome of children
with myelomeningocele: effect of lower-extremity muscle
strength. Developmental Medicine & Child Neurology 33: 482–90.
Michael T, Doll B, von Moers A, Scheffner D. (1990) Dynamic
orthoses for children with myelomeningocele. In: Doll-Tepper,
Dahms, Doll, von Selzam, editors. Adapted Physical Activity.
Berlin, Heidelberg: Springer Verlag. p 475–81.
Park BK, Song HR, Vankoski SJ, Moore CA, Dias LS. (1997) Gait
electromyography in children with myelomeningocele at sacral
level. Archives of Physical Medicine Rehabilitation 78: 471–5.
Rosenstein BD, Greene WB, Herrington RT, Blum AS. (1987) Bone
density in myelomeningocele: the effects of ambulatory status
and other factors. Developmental Medicine & Child Neurology
29: 486–94.
Samuelsson L, Skoog M. (1988) Ambulation in patients with
myelomeningocele: a multivariate statistical analysis. Journal of
Pediatric Orthopaedics 8: 569–75.
Sharrard WJW. (1964) The segmental innervation of the lower limb
muscles in man. Annals of the Royal College of Surgeons
35: 106–22.
Smith GK, Smith ED. (1973) Selection for treatment in spina bifida
cystica. British Medical Journal 4: 189–97.
Stark GD. (1971) Neonatal assessment of the child with a
myelomeningocele. Archives of Disease in Childhood 46: 539–48.
— Baker CW. (1967) The neurological involvement of the lower
limbs in myelomeningocele. Developmental Medicine & Child
Neurology 9: 732–7.
Stillwell A, Menelaus MB. (1983) Walking ability in mature patients
with spina bifida cystica. Journal of Pediatric Orthopaedics
3: 184–90.
Stott NS, Zionts LE, Gronley JK, Perry J. (1996) Tibialis anterior
transfer for calcaneal deformity: a postoperative gait analysis.
Journal of Pediatric Orthopaedics 16: 792–8.

Classification of Lesion Levels in Myelomeningocele Åsa Bartonek et al. 805