JOURNAL OF NEUROTRAUMA

Volume 20, Number 8, 2003

Mary Ann Liebert, Inc.

The Effect of Treadmill Training on Motor Recovery after a

Partial Spinal Cord Compression-Injury in the Adult Rat

SYLVIE MULTON,* RACHELLE FRANZEN,* ANNE-LISE POIRRIER, FELIX SCHOLTES,

and JEAN SCHOENEN

ABSTRACT

Locomotor training on a treadmill is a therapeutic strategy used for several years in human para-
plegics in whom it was shown to improve functional recovery mainly after incomplete spinal cord
lesions. The precise mechanisms underlying its effects are not known. Experimental studies in adult
animals were chiefly performed after complete spinal transections. The objective of this experiment
was to assess the effects of early treadmill training on recovery of spontaneous walking capacity af-
ter a partial spinal cord lesion in adult rats. Following a compression-injury by a subdurally in-
flated microballoon, seven rats were trained daily on a treadmill with a body weight support sys-
tem, whereas six other animals were used as controls and only handled. Spontaneous walking ability
in an open field was compared weekly between both groups by two blinded observers, using the
Basso, Beattie and Bresnahan (BBB) locomotor rating scale. Mean BBB score during 12 weeks was
globally significantly greater in the treadmill-trained animals than in the control group, the bene-
fit of training appearing as early as the 2nd week. At week 7, locomotor recovery reached a plateau
in both animal groups, but remained superior in trained rats. Daily treadmill training started early
after a partial spinal cord lesion in adult rats, which accelerates recovery of locomotion and pro-
duces a long-term benefit. These findings in an animal model mimicking the closed spinal cord in-
jury occurring in most human paraplegics are useful for future studies of optimal locomotor train-
ing programs, their neurobiologic mechanisms, and their combination with other treatment
strategies.

Key words: BBB locomotor rating scale; locomotor recovery; spinal cord injury; treadmill training

INTRODUCTION over a treadmill using a modulable body weight-support

system. Although Wernig et al. (1995) demonstrated that

L OCOMOTOR TRAINING ON A TREADMILL has been uti-

lized for several years in human paraplegia due to
spinal cord injury (SCI) as a therapeutic strategy to im-
prove motor recovery (Dietz et al., 1994, 1995; Dobkin
et al., 1995; Wernig and Mller, 1992). Training is gen-
erally performed 5 days/week in the chronic phase of SCI
by suspending paraplegic patients in an upright position
a higher level of locomotor restoration can be achieved
by such treadmill training than by conventional therapies,
the underlying mechanisms of action are not well under-
stood.
Experimental studies have shown that locomotor train-
ing improves the ability to perform full weight-bearing
stepping on a treadmill in cats after a complete spinal

Research Center for Cellular and Molecular Neurobiology, Neuroanatomy Laboratory, University of Liege, Liege, Belgium.
*These authors contributed equally to this work.

699
 MULTON ET AL.
cord transection (Barbeau and Rossignol, 1987; de Leon
et al., 1998a; Edgerton et al., 1997; Lovely et al., 1986),
which has supported the presence in the lower spinal cord
of a central pattern generator (CPG) able to generate
rhythmic motor activities in the absence of supraspinal
descending inputs (Grillner, 1985; Grillner and Wallen,
1985). How the CPG can shape its functional properties
in response to the training is not clear, but it is thought
that reinforcement by sensory afferents of existing sen-
sorimotor pathways, rather than generation of new con-
nections, might be responsible for the beneficial effects
(de Leon et al., 1998a, 1999, 2001).
Clinical studies in human paraplegics indicate, how-
ever, that the efficiency of treadmill training depends
largely on the severity of the lesion. Patients with com-
plete spinal cord lesions never improve up to the stage
of being able to walk without the assistance of the weight-
support system and consequently are not able to step on
a static floor (Van de Crommert et al., 1998). The in-
ability of patients with complete transections to achieve
unassisted walking, unlike the fully spinalized cat, sug-
gests that the greater improvement observed in subjects
with incomplete lesions may not solely be attributable to
spinal mechanisms, since generation of stepping is prob-
ably more dependent on supraspinal and/or propriocep-
tive inputs in humans than in cat (Edgerton et al., 2001;
Van de Crommert et al., 1998). Considering these clini-
cal and functional data and knowing that the ratio of in-
complete versus complete spinal cord lesions is becom-
ing increasing known in the population of paraplegics
(Tator et al., 1993), experimental studies using animals
with incomplete spinal cord injuries appear clinically and
pathophysiologically relevant. However, such studies are
scarce and have provided contradictory results. Fouad et
al. (2000) found no beneficial effect on free locomotor
activity after a very partial lesion of the cord, whereas
Thota et al. (2001) reported on the opposite improved
functional recovery after treadmill training in rats with
partial spinal cord lesions, although motor activity was
assessed on a treadmill, and not in an open field. The pur-
pose of this study was to use a previously developed
model of a partial closed injury of the spinal cord in adult
rat (Martin et al., 1992) and to assess in this model the
effects on recovery of free walking capacity of daily lo-
comotor training on a treadmill performed as in human
studies with body weight support.
MATERIALS AND METHODS
Animals
The study was performed on a total of 13 adult female
Wistar rats (200250 g). The animals were housed indi-
700
vidually in a colony room under a normal 12:12 h
light/dark cycle and temperature control at 22 6 2C with
food and water provided ad libitum. All experiments con-
formed to the rules and regulations of the Belgian Na-
tional Funds for Scientific Research concerning the use
and care of laboratory animals and were approved by the
local ethics committee.
Surgical Procedure
Animals were anaesthetized by intraperitoneal (i.p.) in-
jection of xylazine (10 mg/kg) and ketamine (75 mg/kg).
The depth of anesthesia was adjusted for absence of
corneal reflex. The back of the rat was shaved and dis-
infected. The eyes were protected from drying with arti-
ficial tears. The dorsal surface of the spinal cord was ex-
posed by a laminectomy at the T9 thoracic level. An
inflatable microballoon (model GV15; CathNet Science,
Paris, France) mounted on a catheter was introduced into
the subarachnoid space through a small hole in the dura
mater and moved rostrally for two metameric levels. The
balloon was rapidly inflated with 20 mL of sterile water
and left in place for 5 min, after which it was deflated
and removed. We have shown previously that this vol-
ume-duration combination produces a necrotic lesion of
the dorsal half of the spinal cord, including the corti-
cospinal tracts and a partially reversible paraplegia (Mar-
tin et al., 1992). The surgical wound was closed in two
layers, and animals were allowed to awake in large cages.
Their bladder was expressed once a day until spontaneous
micturition was recovered. All rats received daily i.p. in-
jection of antibiotics (2.7 mg/kg, Augmentin ; Smith
Kline Beecham) during 15 days to prevent urinary tract
infection.
Treadmill Training
Three days before the spinal cord compression-injury
was performed, all rats were handled and trained daily to
walk on a customized treadmill. After the injury, seven
rats were trained daily on a treadmill and seven others,
used as controls, were taken out of their cages and only
handled daily. Both procedures were started as soon as
the animals reached a score of 1 on the BBB scale. The
time to reach this score ranged from 2 to 4 days post-in-
jury. Allocation to the treadmill-trained or untrained
group was made in a balanced way such as to obtain com-
parable post-injury delays to score 1 in both groups.
Taking into account published data (Edgerton et al.,
1983; Fouad et al., 2000) as well as feasibility, the fol-
lowing treadmill training protocol was chosen: speed of
5.8 cm/sec; 30 min/day (three sessions of 10 min sepa-
rated by 5-min pauses); 5 days/week for 12 weeks. For
the training session, animals were placed in an adjusted
 TREADMILL TRAINING EFFECT ON MOTOR RECOVERY
hanging harness, which allows free movements of all
limbs (quadrupedal stepping) and partially supports body
weight. During the first week of training, when all rats
had a profound flaccid paraplegia, the hind paws were
manually placed in a plantar stepping position. Follow-
ing this period, all treadmill-trained animals were stimu-
lated by pinching the tail or rubbing the skin around the
hip, when they stopped stepping during the training ses-
sion.
One animal in the handled control group died during
the 1st week after SCI and was not included in the analy-
sis.
Behavioral Study
Two experimenters blinded to the treatment performed
all behavioral tests. The motor behavior was observed
during 4 min in each animal every week for 12 weeks.
For this purpose, the animals were placed in an open field
(95 cm of diameter) and the BBB locomotor rating scale
was used. In this scale, which has been specifically de-
veloped to measure motor recovery in spinal cordinjured
rats, scores range from 0 to 21 depending on the level of
complexity of motor activities in the hindlimbs (Basso et
al., 1995). Behavioral assessments were always made in
the morning and thus before the training sessions in the
treadmill group, which were performed in the afternoon.
Histological Analysis
After a survival time of 12 weeks, rats were deeply
anaesthetized and perfused-fixed with 500 mL of 3%
paraformaldehyde in 0.1 M phosphate buffer saline (PBS,
pH 7.4). The spinal cords were rapidly dissected and post-
fixed in the same fixative for 24 h at 4C. Tissue blocks
were taken for histological analysis from the lesion site.
After osmification and embedding in resin (Epon 812),
semi-thin transverse sections (1 mm thick) from the epi-
center of the lesion were stained with toluidine blue. Us-
ing a light microscope (Reichert, 340 magnification), an
investigator, blinded to the experimental group, identi-
fied the area of normal appearing white matter on a sec-
tion from the lesion epicenter in each lesioned animal
and, for comparison, on a section from a corresponding
spinal segment (T7) of a normal, unlesioned rat. He then
outlined these white matter areas using a camera lucida
drawing at the 34 magnification and colored them with
black ink. The drawings were converted on an image an-
alyzer into black and white digital images with black pix-
els attributed to spared white matter and white pixels to
lesioned tissue. The area occupied by spared white mat-
ter in lesioned animals was expressed as a percentage of
the white matter outlined in the T7 section of a normal
unlesioned rat.
701
Statistical Analyses
Results were expressed as means 6 SEM. A repeated-
measures analysis of variance (ANOVA) was used to as-
sess the general effect of the treatment on motor recov-
ery and to compare BBB scores between trained and
non-trained animals during the entire 12-week observa-
tion period. Week by week comparisons between the two
groups were performed with a planned comparisons
test. The Mann-Whitney U test was used to compare
spared white matter between the two groups of animals.
The level of significance was set to p , 0.05.
RESULTS
BBB Locomotor Scale
The mean post-injury time to attain score 1 on the BBB
scale was not significantly different between the tread-
mill-trained (3.5 6 0.53 days) and the untrained control
groups (3 6 0.22 days). Thereafter BBB scores improved
significantly over the 12-week observation period in all
animals (F(1,11) 5 2.9; p , 0.0001). However, during
the 12 weeks of treadmill training, trained rats (n 5 7)
globally achieved a significantly higher level of post-in-
jury motor recovery than non-trained rats (n 5 6;
F(1,11) 5 8.19; p 5 0.015). The difference in BBB
scores between the two groups became significant from
week 2 on (F(1,11) 5 5.53; p 5 0.04; Fig. 1).
In treadmill-trained animals, the highest BBB score
(10) was reached after 5 weeks and there was no sup-
plementary benefit at later time points. The score of 10
on the BBB scale indicates that the animals are able to
support their body weight and to perform occasional steps
but without hindlimb/forelimb coordination. The highest
BBB score obtained by non-trained rats was 8 and was
reached only after 7 weeks. The BBB score of 8 corre-
sponds to the capacity to move all three joints of the hind
legs extensively but without any body weight support.
Histology
Gross inspection of the dissected spinal cords showed
severe, irregular shrinking at the level of the compres-
sion-injury. Under the light microscope there was a pro-
found tissue remodeling at the lesion site. Massive ner-
vous tissue loss predominated in the dorsal two-thirds of
the lesioned cord (Fig. 2). Microcavitations, lipid-loaden
macrophages, Schwann cell and arachnoid cell infiltra-
tion, regenerating dorsal root afferents, neo-vascularisa-
tion and fibro-glial scarring were observed to varying de-
grees in all animals. There was also in all sections some
sparing of white matter tracts in ventral and ventro-lat-
eral funiculi. After comparison of the area occupied by
 MULTON ET AL.
FIG. 1. Mean weekly BBB scores over the 12-week observation period in control non-trained and treadmill-trained rats after
a partial compression-injury of thoracic spinal cord at T67. *p , 0.05.
this spared tissue to the total area of white matter in a
corresponding segment of normal spinal cord, percent-
age areas of spared white matter were on average com-
parable between treadmill-trained (12.17 6 2.20%) and
untrained animals (11.42 6 3.08%; Fig. 3).
DISCUSSION
Our data suggest that treadmill training initiated early
after an incomplete closed spinal cord compression-in-
jury can improve locomotor recovery in adult rats. Com-
pared to non-trained animals, the benefit appears as soon
as the second week of training and is maximal at 4 weeks.
Between weeks 7 and 12 post-injury, motor performance
as assessed by the BBB rating scale reaches a plateau,
during which the difference between the two animal
groups remains stable. At the end of the 12-week obser-
vation period, the gain in mean BBB score in treadmill-
trained rats, although consisting of only two points, is
functionally significant, as it indicates that they are able
to support their body weight, while non-trained rats are
not. It is unlikely that the difference in motor outcome
between untrained and treadmill-trained animals is due
to a less severe spinal cord injury in the latter. There was
indeed no difference in the amount of spared ventral
white matter between the two groups. Also, after the com-
pression-lesion, treadmill-trained rats reached a BBB
score of 1 slightly, though not significantly, later (mean:
3.5 days) than untrained animals (mean: 3 days).
The results presented here are in accordance with data
on beneficial effects of treadmill training on stepping re-
702
covery in human paraplegics (Dietz et al., 1994, 1995;
Dobkin et al., 1995; Wernig and Mller, 1992; Wernig
et al., 1995) and in experimental studies in cats (Barbeau
and Rossignol, 1987; de Leon et al., 1998a; Lovely et al.,
1986).
However, our study differs from previous ones on the
effects of treadmill training in experimental paraplegia
by the type and severity of the spinal cord injury (SCI)
and the motor activities that were assessed. SCIs in the
cat model consist in total transection of the cord. The
spinal cord being isolated from the brain, the improve-
ment in functional recovery has been interpreted on the
basis of training-induced biochemical and structural
changes in the lumbar central pattern generator (CPG; de
Leon et al., 2001; Rossignol, 2000). By contrast, with the
compression protocol used in our model, the lesion de-
stroys mainly the gray matter and dorsal tracts (Martin
et al., 1992), which are involved in the fine control of
movements (Fouad et al., 2000; Metz et al., 1998), par-
tially sparing the ventral and ventro-lateral tracts, which
are more implied in the initiation of walking and coor-
dination between fore- and hindlimbs (Brustein and
Rossignol, 1998; Gorska et al., 1996; Jordan, 1998). As
a first probable consequence of the incompleteness of the
spinal lesions performed in our study, non-trained con-
trol animals were able to recover spontaneous locomotor
activity. Such partial restoration of voluntary stepping has
also been observed after incomplete lesions in cat
(Brustein and Rossignol, 1998; Rossignol et al, 1999).
The rapid motor recovery over weeks might be attributed
to the partial sparing of descending pathways, sufficient
to provide the necessary drive for the initiation of loco-
 TREADMILL TRAINING EFFECT ON MOTOR RECOVERY

A B

C D

FIG. 2. Toluidine bluestained (A,B) and corresponding camera lucida drawings (C,D) of transverse semi-thin (1 mm) sections
at the epicenter of the lesion from a treadmill-trained (A,C) and an untrained control animal (B,D). Dark-shaded areas represent
spared white matter and are similar in both animals.

motion as well as some postural control. To reinforce this
hypothesis, in our model of spinal compression-injury
with a subdural balloon, more severe lesions produced
by inflating the balloon with 40 mL for 5 min were never
followed by any recovery of spontaneous walking move-
ments in an open field (Martin et al., 1992 and unpub-
lished data). Another important consequence of an in-
complete spinal lesion is that it leaves open the possibility
that the significant improvement of locomotor recovery
by treadmill training might not be solely due to plastic
changes in the caudal segment containing the CPG, but
also to the progressive reinforcement of partially spared
descending connections through sprouting and/or synap-
tic plasticity. In spinalized vertebrates, motor recovery
after such a training can indeed only be explained by
changes of caudal spinal networks. It has been shown,
for instance, that hindlimb motor training (de Leon et al.,
1999) is able to attenuate the increase of inhibitory trans-
mitters that occurs in the caudal spinal cord after a
complete injury (Robinson and Goldberger, 1986;
Tillakaratne et al, 2000). Most studies which have ex-
amined this, came to the conclusion that the contribution
of muscle trophicity or reflex activity in the beneficial
effects of locomotor training was negligible (de Leon et
al., 2001; Edgerton et al., 2001; Van de Crommert et al.,
1998).
Another difference between previous studies and
ours pertains to the method used to evaluate motor re-
covery. In cats, locomotor recovery is assessed by ex-
amining induced stepping on a treadmill, as adult ani-
mals have no spontaneous walking activity after
spinalization (Edgerton et al., 1983). In fact, in spinal-

703
 MULTON ET AL.
FIG. 3. Histogram of the percentages (mean 6 SEM) of
spared white matter at the lesion site in treadmill-trained (dark
bar) and untrained control animals (white bar) compared to a
corresponding segmental level of a normal rat spinal cord.
There is no significant difference between the two groups.
ized adult animals, the type of motor recovery depends
directly on the type of training they receive (Edgerton
et al., 2001); that is, treadmill training will improve
their ability to walk on a treadmill, but they can also
be selectively trained to stand, and not to step (de Leon
et al., 1998b; Edgerton et al., 1997; Hodgson et al.,
1994). This could also be a bias in Thota et al.s study
(2001), which assessed improvement by a kinematic
analysis of walking on a treadmill, but not in an open
space. By contrast, in the present study, we evaluated
the rats ability to walk spontaneously in an open field.
We chose the BBB locomotor rating scale (Basso et al.,
1995), because it is well validated and adapted to re-
veal early changes in animals with such a severe loco-
motor deficit as ours; it is particularly suited to detect
a shift from a no body weight support situation to oc-
casional stepping. Our results show that, as in humans
with incomplete SCI, the treadmill training can im-
prove a motor activity other than the one practiced and
that this spontaneous locomotor activity does not need
to be stimulated by a treadmill nor by other sensory
stimuli.
704
Finally, the difference between our results and those
reported by Fouad et al. (2000), who also examined
open field walking capacity but did not find any mo-
tor recovery improvement with treadmill training, may
have several explanations. In their study, the spinal
cord lesion was clearly less severe than in ours and
they propose as a possible explanation for the absence
of benefit the fact that their animals were able to move
around quite well after SCI, reaching a BBB score of
9 after 1 week, and that they were capable of train-
ing themselves in their cages, hence limiting the po-
tential beneficial effect of the treadmill practice. In our
study the mean BBB score 1 week post-injury was 2,
which means that motor activity in hindlimbs was rudi-
mentary and that the animals were unlikely to be ca-
pable of effective self-training during this time period.
The more severe initial paraplegia may also be the rea-
son why, in our study, we found a window span-
ning the first post-injury month during which tread-
mill training produced an improvement of motor
recovery. Interestingly, the benefit that appeared dur-
ing the first month after injury, and which did not in-
crease further afterwards, persisted until the end of the
12-week observation period, suggesting that locomo-
tor training ought to be commenced as early as possi-
ble after SCI. However, its potential effects at later
stages have not yet been studied in animals with in-
complete lesions, nor have the effects of its interrup-
tion after the first month.
Besides its possible implications for clinical train-
ing programs in spinal cordinjured patients, the par-
adigm of locomotor training we have used here has an-
other potential interest. From a therapeutic and basic
perspective, it can be combined with pharmacologic
interventions or transplantation experiments targeting
aminergic systems in the deafferented caudal seg-
ments, neuroprotective mechanisms, or axonal regen-
eration.
To conclude, the objective of this study was to im-
prove the residual voluntary locomotor function by tread-
mill training in rats with a partial thoracic spinal cord le-
sion. The results indicate that such a locomotor training
can accelerate and enhance durably recovery of sponta-
neous walking in an open field. It exerts its effect during
the first month after injury and is followed by a persis-
tent, though non-increasing, benefit over an observation
period of 3 months. This may be relevant for the design
of treadmill training programs in human paraplegics and
offers a template to which other experimental interven-
tions targeting neuronal protection and regeneration can
be combined to promote functional recovery after spinal
cord injury.
 TREADMILL TRAINING EFFECT ON MOTOR RECOVERY
ACKNOWLEDGEMENTS
This study was supported by the National Fund for Sci-
entific Research (Belgium) and by a grant from the
French Community of Belgium (ARC 99/04-241). We
are grateful to G. Brook (Department of Neurology, Uni-
versity of Aachen, Germany) for helpful suggestions dur-
ing the course of the study.
REFERENCES
BARBEAU, H., and ROSSIGNOL, S. (1987). Recovery of lo-
comotion after chronic spinalization in the adult cat. Brain
Res. 412, 8495.
BASSO, D.M., BEATTIE, M.S., and BRESNAHAN, J.C.
(1995). A sensitive and reliable locomotor rating scale for
open field testing in rats. J. Neurotrauma 12, 121.
BRUSTEIN, E., and ROSSIGNOL, S. (1998). Recovery of lo-
comotion after ventral and ventrolateral spinal lesions in the
cat. I. Deficits and adaptive mechanisms. J. Neurophysiol.
80, 12451267.
de LEON, R.D., HODGSON, J.A., ROY, R.R., et al. (1998a).
Locomotor capacity attributable to step training versus spon-
taneous recovery after spinalization in adult cats. J. Neuro-
physiol. 79, 13291340.
de LEON, R.D., HODGSON, J.A., ROY, R.R., et al. (1998b).
Full weight-bearing hindlimb standing following stand train-
ing in the adult spinal cat. J. Neurophysiol. 80, 8391.
de LEON, R.D., TAMAKI, H., HODGSON, J.A., et al. (1999).
Hindlimb locomotor and postural training modulates glycin-
ergic inhibition in the spinal cord of the adult spinal cat. J.
Neurophysiol. 82, 359369.
de LEON, R.D., ROY, R.R., and EDGERTON, V.R. (2001). Is
the recovery of stepping following spinal cord injury medi-
ated by modifying existing neural pathways or by generat-
ing new pathways? A perspective. Phys. Ther. 81, 1904
1911.
DIETZ, V., COLOMBO, G., and JENSEN, L. (1994). Loco-
motor activity in spinal man. Lancet 344, 12601263.
DIETZ, V., COLOMBO, G., JENSEN, L., et al. (1995). Loco-
motor capacity of spinal cord in paraplegic patients. Ann.
Neurol. 37, 574582.
DOBKIN, B.H., HARKEMA, S., REQUEJO, P., et al. (1995).
Modulation of locomotor-like EMG activity in subjects with
complete and incomplete spinal cord injury. J. Neurol. Re-
habil. 9, 183190.
EDGERTON, V.R., DUKE, J., JOHNSON, L.A., et al. (1983).
Effects of treadmill exercises on hind muscles of the spinal
cat, in: Spinal Cord Reconstruction. C.C. Kao, R.P. Bunge,
and P.J. Reier (eds), Raven Press: New York, pps. 435443.
705
EDGERTON, V.R., DE LEON, R.D., TILLAKARATNE, N.,
et al. (1997). User-dependent plasticity in spinal stepping and
standing, in: Advances in Neurology: Neuronal Regenera-
tion, Reorganisation, and Repair. LippincottRaven Pub-
lishers: Philadelphia, pps. 233247.
EDGERTON, V.R., DE LEON, R.D., HARKEMA, S.J., et al..
(2001). Retraining the injured spinal cord. J. Physiol. 533,
1522.
FOUAD, K., METZ, G.A., MERKLER, D., et al. (2000).Tread-
mill training in incomplete spinal cord injured rats. Behav.
Brain Res. 115, 107113.
GORSKA, T., BEM, T., MAJCZYNSKI, H., et al. (1996). Dif-
ferent forms of impairment of the fore-hindlimb coordina-
tion after partial spinal lesions in cats. Acta Neurobiol. Exp.
56, 177188.
GRILLNER, S. (1985). Neurobiological bases of rhythmic mo-
tor acts in vertebrates. Science 228, 143149.
GRILLNER, S., and WALLN, P. (1985). Central pattern gen-
erators for locomotion, with special reference to vertebrates.
Annu. Rev. Neurosci. 8, 233261.
HODGSON, J.A., ROY, R.R., DE LEON, R., et al. (1994). Can
the mammalian lumbar spinal cord learn a motor task? Med.
Sci. Sports Exerc. 26, 14911497.
JORDAN, L.M. (1998). Initiation of locomotion in mammals.
Ann. N.Y. Acad. Sci. 860, 8393.
LOVELY, R.G., GREGOR, R.J., ROY, R.R., et al. (1986). Ef-
fects of training on the recovery of full-weight-bearing step-
ping in the adult spinal cat. Exp. Neurol. 92, 421435.
MARTIN, D., SCHOENEN, J., DELRE, P., et al. (1992).
Experimental acute traumatic injury of the adult rat spinal
cord by a subdural inflatable balloon: methodology, be-
havioral analysis, and histopathology. J. Neurosci. Res. 32,
539550.
METZ, G.A., DIETZ, V., SCHWAB, M.E., et al. (1998). The
effects of unilateral pyramidal tract section on hindlimb mo-
tor performance in the rat. Behav. Brain Res. 96, 3746.
ROBINSON, G.A., and GOLDBERGER, M.E. (1986). The de-
velopment and recovery of motor function in spinal cats. I.
The infant lesion effect. Exp. Brain Res. 62, 373386.
ROSSIGNOL, S. (2000). Locomotion and its recovery after
spinal injury. Curr. Opin. Neurobiol. 10, 708716.
ROSSIGNOL, S., DREW, T., BRUSTEIN, E., et al. (1999).
Locomotor performance and adaptation after partial or com-
plete spinal cord lesions in the cat. Prog. Brain Res. 123,
349365.
TATOR, C.H., DUNCAN, E.G., EDMONDS, V.E., et al.
(1993). Changes in epidemiology of acute spinal cord injury
from 1947 to 1981. Surg. Neurol. 40, 207215.
THOTA, A., CARLSON, S., and JUNG, R. (2001). Recovery
 MULTON ET AL.

of locomotor function after treadmill training of incomplete WERNIG, A., MLLER, S., NANASSY, A., et al. (1995).
spinal cord injured rats. Biomed. Sci. Instrum. 37, 6367. Laufband therapy based on rules of spinal locomotion is ef-
fective in spinal cord injured persons. Eur. J. Neurosci. 7,
TILLAKARATNE, N.J., MOURIA, M., ZIV, N.B., et al.
823829.
(2000). Increased expression of glutamate decarboxylase
(GAD(67)) in feline lumbar spinal cord after complete tho- Address reprint requests to:
racic spinal cord transection. J. Neurosci. Res. 60,
Jean Schoenen, M.D., Ph.D.
219230.
Research Center for Cellular and
VAN DE CROMMERT, H., MULDER, T., and DUYSENS, J. Molecular Neurobiology
(1998). Neural control of locomotion: sensory control of the Neuroanatomy Laboratory
central pattern generator and its relation to treadmill train- University of Liege
ing. Gait Posture 7, 251263. Rue de Pitteurs, 20 (L3)
WERNIG, A., and MLLER, S. (1992). Laufband locomotion 4020 Liege, Belgium
with body weight support improved walking in persons with
severe spinal cord injuries. Paraplegia 30, 229238. E-mail: jschoenen@ulg.ac.be

706