Spinal Cord (1999) 37, 402 ± 409

ã 1999 International Medical Society of Paraplegia All rights reserved 1362 ± 4393/99 $12.00
http://www.stockton-press.co.uk/sc

The pattern of re¯ex recovery during spinal shock

H-Y Ko1, JF Ditunno Jr*,2, V Graziani2 and JW Little3
1
Department of Rehabilitation Medicine, Pusan National University Hospital, Pusan National University College of
Medicine, Pusan, Korea; 2Department of Rehabilitation Medicine, Je€erson Medical College of Thomas Je€erson
University, Philadelphia, Pennsylvania, USA; 3Department of Rehabilitation Medicine, University of Washington,
Seattle, Washington, USA

Study Design: A prospective descriptive study of the course of recovery of re¯exes following
acute spinal cord injury (SCI).
Objectives: The purpose of the study was to observe the pattern of re¯ex recovery following
acute SCI in order to determine the prognostic signi®cance of re¯exes for ambulation and
their relationship to spinal shock.
Setting: A regional spinal cord injury center in Philadelphia, Pennsylvania, USA.
Methods: Fifty subjects admitted consecutively over a 9 month period and on the day of
injury were observed for the following re¯exes; bulbo-cavernosis (BC), delayed plantar
response (DPR), cremasteric (CRM), ankle jerk (AJ), knee jerk (KJ), and normal plantar
response for 5 ± 7 days a week and 6 ± 8 weeks duration. The 50 subjects were assessed for
ambulation of 200 feet at time of discharge. MRI studies were reviewed on 13/28 complete
(ASIA A) injuries.
Results: Thirty-®ve subjects (28 ASIA A, 4 ASIA B, 3 ASIA C) had a DPR of 2 days or
longer duration and these subjects were not ambulatory. The fourteen subjects (12 ASIA D
and 2 ASIA C), who were ambulatory, either had no DPR (11/14) or had a DPR of only 1
days duration (3/14). One subject (ASIA B) was not ambulatory and had a DRP of 1 days
duration. The DPR was the ®rst re¯ex to recover most often, followed by the BC, CRM in the
®rst few days and later followed by the deep tendon re¯exes (AJ & KJ) by 1 ± 2 weeks
respectively. Less than 8% of subjects had no re¯exes on the day of injury and the re¯exes did
not follow a caudal-rostral pattern of recovery.
Conclusions: Prognosis for ambulation based on re¯exes early after SCI should not be linked
to current descriptions of spinal shock. In fact, the view of spinal shock, based on the absence
of re¯exes and the recovery of re¯exes in a caudal to rostral sequence, is of limited clinical
utility and should be discarded. The evolution of re¯exes over several days following injury
may be more relevant to prognosis than the use of the term spinal shock and the presence or
absence of re¯exes on the day of injury.

Keywords: spinal cord injury; deep tendon re¯ex; plantar re¯ex; ambulation; outcome; spinal
shock

Introduction
The term spinal shock, although in use since 1850,1 Its utility as a clinical term has been reduced by
refers to the absence or depression of segmental spinal statements that one can not prognosticate until the
re¯exes that accompany a rostral spinal cord injury. Its patient emerges from spinal shock.7 According to
cause, however, remains obscure.2,3 Based on animal Holdsworth,8 as segmental re¯exes recover (ie as
studies, spinal shock is due to loss of normal spinal shock resolves) from rostral spinal cord injury,
facilitation and/or inhibition onto cord interneurons and no accompanying recovery of voluntary move-
and motorneurons from corticospinal, rubrospinal, ment is observed, then the prognosis for ultimate
vestibulospinal and reticulospinal pathway.4 ± 6 functional recovery is extremely poor. These reports
assume a de®ned period when spinal shock ceases. As
originally described, spinal shock is associated with
severe injury or transection of the spinal cord in which
there is sensory loss, motor paralysis and a gradual
*Correspondence: JF Ditunno, Jr., MD, Department of
Rehabilitation Medicine, Thomas Je€erson University Hospital, 111
recovery of re¯exes following their complete aboli-
S. 11th Street, Suite 375 Main Building, Philadelphia, Pennsylvania tion.1,9 Does spinal shock's initial phase end, when
19107, USA re¯exes begin to recover? Which re¯exes does this refer
 Reflex recovery during spinal shock
H-Y Ko et al
403

to; cutaneous, which are polysynaptic, deep tendon, performed in the acute spinal cord unit following
which are monosynaptic or pathological re¯exes, such admission from the emergency room and continued
as the Babinski sign, which is multisegmental? after transfer to the rehabilitation unit in 2 ± 3 weeks.
Although Guttmann10 stated that the recovery of The ASIA impairment grade was determined at the
re¯exes was quite variable in man, he indicated they time of the initial neurological examination in the
followed a pattern and the ®rst re¯exes to return were emergency room.
the bulbocavernosis (BC) and the anal cutaneous. The normal plantar response and the Babinski sign
Although he reported on only ®ve cases, there had were elicited by a lightly applied stimulus to the lateral
been detailed descriptions of re¯ex recovery by sole of the foot in the area of the S1 dermatome.
Riddock11 and Guillian12 from World War I. The Flexion of the toes (normal plantar response) and
clinical ®ndings in these complete injuries were often extension of the toes (Babinski sign) are brisk
con®rmed by evidence of cord transection at post- reactions and the response to stimuli are less than
mortem examination due to the high mortality13 at 90 ms.26 The delayed plantar response, however
this time. Guttmann10 suggested, however, there was a requires an unusually strong stimulus, in contrast to
need to study the re¯ex recovery serially and system- the Babinski sign or normal plantar response, and is
atically in order to better understand the pattern of applied by stroking with a blunt instrument upward
recovery from spinal shock and its prognostic from the heel toward the toes along the lateral sole of
signi®cance. In response to this suggestion and the foot continued medially across the volar aspect of
reports by ourselves14 and others15 of the presence the metatarsal heads. Following the stimulus the toes
and prognostic signi®cance of a pathological re¯ex ¯ex and relax in a delayed sequence, which at times is
known as the delayed plantar response (DPR), which robust and at other times is slight.14 The duration of
occurs early after spinal cord injury (SCI), it appeared the stimulus to the foot and the response of the ¯exor
reasonable to examine more thoroughly re¯ex recovery digitorum brevis muscle was recorded in a representa-
during the immediate period following SCI. In tive patient. The sole of the foot was stimulated
addition, due to the increased emphasis on relating mechanically by a sharp object picked up by a surface
initial neurological ®ndings alone to functional electrode. The EMG tracing was recorded from a
recovery16 ± 18 or combined with imaging studies19,20 needle electrode placed in the ¯exor digitorum
electrophysiological studies21,22 and histopathol- muscles. The recordings were from dual channels and
ogy,23,24 the following study was undertaken. superimposed one on the other (Figure 6).
The purpose of the study was to observe the pattern MRI studies were performed on 13 of the 28
of re¯ex recovery following acute SCI in order to patients with complete injuries, but 10 of 28 subjects
determine the prognostic signi®cance of the re¯exes were not studied because of a gunshot wound. The
and the relationship to spinal shock. We hypothesized remaining ®ve subjects did not have studies due to
that the presence and duration of the delayed plantar invasive metal, which precluded a MRI study. The
response correlated with the severity of the injury to neuroradiologist, who interpreted the MRI studies was
the spinal cord and the prognosis for ambulation. blinded to the clinical features of the subjects.
Outcome data for ambulation was determined at
Method time of discharge from the rehabilitation facility.
Functional ambulators were de®ned as a person who
Seventy patients with a diagnosis of acute SCI were were able to walk at least 200 feet with or without
admitted during a 9 month period (4/96 ± 1/97) to a orthoses or walking aids at the time of rehabilitation
regional SCI center within 24 h of injury. The majority discharge.
of patients received methylprednisolone except those
sustaining a gunshot wound. Seventeen patients with
SCI and associated lower extremity fracture, preexist-
ing neurological de®cit or lower motor neuron injury
Table 1 Onset of re¯exes within 2 ± 3 days of injury in
or disease were excluded. Of the remaining 53 patients subjects with complete injuries (ASIA A) compared to
the mechanisms of injury included motor vehicle subjects with incomplete injuries (ASIA B & C and ASIA
accidents (n=17), falls (n=17), violence including D). The Fisher Exact Test was calculated to determine
gun shot wounds (n=15) and sports (n=4). Neurolo- signi®cance (P50.05)
gical testing, both motor and sensory, was performed
on each subject, according to the International DPR BC CRM BS AJ KJ
Standards for Neurological and Functional Classifica- ASIA A 31 26 11 2 5 2
tion of Spinal Cord Injury,25 to determine the (n=31)
neurological level of injury and ASIA Impairment ASIA B & C 8 10 5 4 8 7
Scale. The following re¯exes were recorded in the (n=10)
emergency room and 5 of 7 days serially for 6 ± 8 weeks ASIA D 1 12 9 10 12 12
by one of the authors (H-Y Ko); bulbocavernosis, (n=12)
delayed plantar, cremasteric, Babinski, ankle jerk, knee Total N 53 53 43 53 53 53
P value 0.001 NS 0.003 0.001 0.001 0.001
jerk and normal plantar. These examinations were
 Reflex recovery during spinal shock
H-Y Ko et al
404

Data on the comparison of patients with complete for a period of time until the DPR stopped. The
and incomplete injuries (Table 1) were analyzed by cessation of the DPR, therefore, showed an inverse
Fisher's Exact Tests. This analysis was used because of relationship to the beginning of the Babinski sign (see
the number of cells containing low frequencies. Data Figure 1). In several subjects in which the stimulus was
on the pattern of recovery by age of subjects (Table 2) applied to the sole of the foot and was unusually
were analyzed by Wilcoxon Signed Rank Tests. This strong in order to elicit the DPR, there was an
analysis was a conservative measure, because of the immediate dorsi¯exion of the great toe (Babinski sign)
small number of data points, which may not have been followed by the DPR suggesting that the same area of
normally distributed. stimulation elicited two di€erent re¯exes temporarily
distinct from each other. This was observed in subjects
Results with complete injuries.
The cremasteric re¯ex (L1-2) occurred in 11/27
Fifty-three of the 70 subjects examined during the ®rst males within 1 day, 17/27 within 3 days, 22/27 within 2
2 days following injury had clinical evidence of an weeks in patients with complete injuries (1/28 subjects
upper motor neuron injury manifested by the presence was female). The pattern of recovery of the other
of re¯exes such as the DPR, BC and/or CRM in initial cutaneous re¯exes (BC and DPR) in complete
complete injuries and additional re¯exes in incomplete subjects is illustrated in Figure 2.
injuries. Three of these subjects (all ASIA A) died The ankle jerks (AJ) occurred in 10/28 and the knee
within 14 ± 28 days but were included in the analysis of jerks (KJ) in 7/28 complete injuries (ASIA A) by the
the initial 2 week recovery period. The remaining 50 third day following injury (see Figure 3). The onset of
subjects, whose mean age was 37.5 years (median age the deep tendon re¯exes (AJ and KJ) was delayed in
43 years), were followed to the time of discharge. complete injuries and the average onset of the AJ was
Twenty-eight subjects had complete injuries (ASIA 7.7 days and for the KJ 15.2 days following the injury.
A) and remained complete during the period of their
follow-up examinations. Twenty-two subjects had
incomplete injuries (ASIA B n=5, ASIA C n=5,
ASIA D n=12) and they remained incomplete during
the period of follow-up. Forty-two subjects had
cervical injuries (Complete=22, Incomplete=20) and
11 subjects had thoracic injuries (Complete=9,
Incomplete=2). The three subjects, who died, had
cervical injuries.

Re¯ex changes after complete SCI

The DPR was present in all subjects with complete
injuries (31/31). It was the ®rst re¯ex observed in the
emergency room (ER) in 11/31 subjects, who had
complete injuries (ASIA A), but it followed the BC by
1 ± 2 days in 4/31 and occurred at the same time as BC
in 16 subjects. The DPR, once present, persisted for an
average duration of 14.3 days in subjects with complete
injuries.
The Babinski sign never occured before the DPR Figure 1 Recovery of the Delayed Plantar Response (DPR-
and usually started at some time after the DPR ceased, C) versus the recovery of the Babinski Sign (BS-C) in 28
although, in several subjects it occurred simultaneously complete injuries (ASIA A)

Table 2 Onset of the Ankle Jerk (AJ), Babinski Sign (BS) and the duration of the Delayed Plantar Response (DPR) in 28
subjects with complete injuries (ASIA A) are compared based on the age of two groups. Group 1 (n=9) are the older subjects
(48 years median) and are compared to Group II (n=19), the younger subjects (28 years median). Signi®cance was calculated by
the Wilcoxon Signed Rank Test
Age Onset AJ Onset BS Duration DPR
Group I 48 years median 1 days median 7 days median 6 days median
n=9 49 years mean 1.5 days mean 10 days mean 11 days mean
Group II 28 years median 14 days median 13 days median 15 days median
n=19 33 years mean 14 days mean 14 days mean 15 days mean
P50.05 P50.001 P50.01 P50.05
 Reflex recovery during spinal shock
H-Y Ko et al
405

Within the 28 complete subjects, two patterns of incomplete injuries (ASIA C n=2, ASIA D n=11).
recovery occurred based on the onset of the AJ, BS, The duration of the DPR was 2 days median and
duration of the DPR and the age of the patients (see 3.8 days mean in subjects with incomplete injuries.
Table 2). In Group I (n=9), the median age 48 years, In only one case with an incomplete injury the DPR
the onset of the AJ was one day (median) as compared persisted beyond 7 days, whereas the DPR lasted a
to Group II, median age 28 years, the onset of the AJ week or longer in the majority of subjects with a
was 13 days (median). complete injury (ASIA A's 24/31). The Babinski sign
appeared within 10 days in all nine patients with
incomplete injuries, who had a DPR, but followed
Re¯ex changes in complete versus incomplete SCI no reciprocal pattern as seen in complete injuries. In
The DPR was observed in all subjects with complete the subjects with complete injuries only one half had
injuries, but it was present in less than one half of Babinski signs by 10 days and one half still had a
subjects with incomplete injuries (9/22) and these DPR present (Figure 1). The Babinski sign in the
subjects included 5/5 ASIA B, 3/5 ASIA C and 1/12 incomplete subjects recovered signi®cantly earlier
ASIA D. It was not present in 13 subjects with than in the complete subjects (Figure 4 and Table
1).
The cutaneous re¯exes including the BC and the
cremasteric in 8/11 males, together with the deep
tendon re¯exes were all present in the emergency room
in all eleven ASIA D subjects. Deep tendon re¯exes
were seen in 5/5 ASIA C subjects in the ER, but all
®ve ASIA B subjects had a delay of several days or
more of onset of DTR's. The Babinski sign lagged
slightly behind the AJ and KJ in the incomplete
subjects (Figure 5) but all recovered signi®cantly
earlier than in the complete injuries (Figure 2 and
Table 1).

MRI and ambulation outcome

MRI studies were performed on less than one half of
the subjects (13/28) with complete injuries, because of
the large number of gunshot wounds or invasive metal
present. Little of statistical signi®cance was possible
Figure 2 Recovery of the cutaneous re¯exes, which includes with such limited numbers, related to prognosis of
the Bulbo-cavernosis (BC), the Delayed Plantar Response motor recovery and ambulation based on the severity
(DPR) in 28 complete injuries and the Cremasteric (CRM) in of SCI found on MRI. The only trend of early re¯ex
20 male complete injuries (ASIA A) recovery occurred in subjects with evidence of

Figure 3 Recovery of the deep tendon re¯exes (Ankle Jerk ± Figure 4 Recovery of the Babinski Sign (BS) in 28 complete
AJ, Knee Jerk ± KJ) and the Babinski Sign in 28 complete injuries (ASIA A) versus 21 incomplete injuries (ASIA B, C
injuries (ASIA A) and D)
 Reflex recovery during spinal shock
H-Y Ko et al
406

In this ®gure, the duration of the stimulus is 500 ms

and the onset of the response of the ¯exor digitorum
brevis following the stimulus is 500 ms or a full second
following the initiation of the stimulus.

Discussion
The pattern of re¯ex recovery after traumatic SCI
di€ers for those with complete SCI (ASIA A) from
those with incomplete SCI (ASIA B, C & D). The most
distinguishing re¯ex was the DPR; it was more often
present and lasted longer in those without recovery of
voluntary movement. Thus, the continued presence of
the DPR has poor prognostic signi®cance for
neurological recovery.14,15,27 These prognostic ®ndings
are consistent with a previous publication14 from our
center in which the DPR was not performed serially
Figure 5 Recovery of the deep tendon re¯exes (Ankle Jerk ± but only at the time of admission and it was not
AJ, Knee Jerk ± KJ) in 22 incomplete injuries and Babinski compared to other re¯exes. This study, in contrast to
Sign (BS) in incomplete injuries (ASIA B, C and D) our previous report, shows that the DPR occurs more
frequently in patients with complete and incomplete
injuries, when they are examined serially rather than
only at the time of admission. The study also revealed
that the duration of the DPR is of greater signi®cance
for prognosis of ambulation than its mere presence or
absence at the time of admission.
The DPR is a most unusual re¯ex; its presence and
duration correlates with the severity of upper
motorneuron damage. It is apparently unmasked for
a period of time then disappears, even in complete
SCI, as other hyperre¯exia develops (eg Babinski sign,
tendon hyperre¯exia, ¯exor withdrawal hyperre¯exia).
It shows a reciprocal relationship with the Babinski
Figure 6 Latency of the Delayed Plantar response. A sign (Figure 1) and the DPR is gradually replaced by
represents duration of plantar stimulation (duration 500 ms) the Babinski sign as observed by Riddock11 and
and B represents initiation of EMG response of ¯exor Landau26 in most patients. Several patients for a
digitorum brevis (500 ms post stimulation) short time manifested a Babinski sign followed by a
DPR in response to the same area of stimulation. It
would appear that the same local signature of
stimulation of the S1 dermatome, applied strongly,
persistent compression of the spinal cord by MRI. In produces dorsi¯exion of the great toe immediately and
eight subjects with evidence of cord compression the is then followed by the delayed plantar ¯exion of the
AJ re¯ex recovered in 6.6 days as compared to the toes, suggesting separate pathways and responses from
remaining 5/13 in which the AJ re¯ex recovered in 12.1 the same area of stimulation. Van Gijn28 in a review of
days. There did not appear to be a signi®cant the pathophysiology of the Babinski sign, makes a
correlation of the presence of cord stenosis or distinction between the di€erences in the receptive
spondylosis and early return of re¯exes. ®elds to produce the normal plantar response and the
Thirteen (13/50) subjects were ambulatory at the Babinski sign and also refers to the paradoxical
time of discharge from the rehabilitation facility. Two downward toe response (DPR). He states `the normal
of these subjects (ASIA C n=1, ASIA D n=1), had a plantar response is a skin re¯ex in the strict sense, a
DPR of 1 days duration. Of the remaining 37 subjects, unisegmental re¯ex with a limited receptive ®eld, . . .
who were not ambulatory at the time of discharge, all analogous to the abdominal and cremasteric re¯exes.
but one subject had a DPR of 2 days or longer. This The receptive zone for re¯ex withdrawal of the leg,
included all ASIA A n=28, all ASIA B n=5 and 3/5 including dorsi¯exion of the hallux, extends beyond
ASIC C. The one subject, who did not ambulate and the sole of the foot . . . and the Babinski toe
had a DPR of only 1 days duration was ASIA B on phenomena is part of the primitive but complex
admission and ASIA C at discharge. medullary automatism, whereas the normal, down-
The DPR has been previously described as having a ward response of the toes constitutes a separate re¯ex
delayed response to an unusually strong stimulus to with a limited receptive ®eld.' Although he refers to
the sole of the foot and this is illustrated in Figure 6. the DPR as `an initial downward response of the toes
 Reflex recovery during spinal shock
H-Y Ko et al
407

after a cord injury' he does not o€er an explanation involvement at a segmental level, but it may require
for the `long latency and duration'. The physiological several days of observation before this is evident. The
basis of the DPR is unclear given its markedly immediate presence of the DPR, likewise should be
prolonged latency of 500 ± 1000 ms (see Figure 6). It tested for several days following the injury. A more
is likely that considerable temporal summation of accurate description of spinal shock should character-
a€erent input is required and the a€erent conduction ize this state as a period of altered appearance of
may be due to very slowly conducting a€erents (eg cutaneous and deep tendon re¯exes and the emergence
unmyelinated C-®bers).26 Because it is only transiently and at times disappearance of pathologic re¯exes over
present for days to several weeks after SCI, it has been days and weeks. The evolution of spinal re¯exes over
only minimally studied clinically and neurophysiologi- several days, therefore, appears to be more relevant to
cally.26 the prognosis for ambulation than the mere presence
Few complete SCI patients (ASIA A within 24 h at or absence of re¯exes on the day of injury.
injury and at discharge from rehabilitation) had no Another observation of re¯ex recovery in complete
re¯exes on initial examination (2 of 28 subjects) and SCI was the presence of AJ and KJ in one-third of
the majority had two or more re¯exes (the DPR and/ patients with complete injuries within the ®rst 3 ± 6
or BC and cremasteric re¯exes). The deep tendon days as compared to an average recovery of 1 ± 2
re¯exes were absent in all but one of the complete weeks respectively for the entire 29 subjects. The older
injuries. The presence of spinal shock, de®ned as the age of those that recovered DTRs by 3 ± 6 days
absence of all re¯exes on the day of injury, is unusual suggests that spinal stenosis with preexisting subclini-
in current clinical practice. In fact, almost a third of cal myelopathy such as the solid cord syndrome31 may
subjects were examined within 1 ± 3 h of injury and all have contributed to faster recovery of re¯exes. The
had re¯exes present. Thus, even though some subjects limited number of MRI studies, however, did not
had high blood alcohol levels, some may have had con®rm the correlation of cervical spondylosis with the
residual ephaptic e€ects29 given they were within hours early recovery of deep tendon re¯exes. Animal studies
of cord injury, some were hypotensive, a few had have clearly shown faster recovery of tendon re¯exes
associated head injury, most were receiving narcotic in those with prior upper motoneuron lesions (eg
pain medication, and most were receiving methylpred- motor cortex, vestibular nucleus or cord hemisection;32
nisolone, some spinal re¯exes were elicitable in 93% of and absence of spinal shock is well known in patients
all subjects. with slowly progressive myelopathy.11 Older patients,
Although some10 say after complete SCI, the BC who sustain incomplete SCI, have a di€erent clinical
re¯ex recovers ®rst, followed by the deep tendon course and the majority do not ambulate as compared
re¯exes in a caudal to rostral direction; we were to a younger group of which 90% are ambulatory.33
unable to con®rm this by our observations. The DPR This di€erence in clinical course may also be due to
(S1-2) was the ®rst re¯ex observed in 20 subjects, the pre-existing damage to the cord in the older subjects,
BC (S2-5) in 16 and the cremasteric (L2) was seen in but the MRI studies were inconclusive.
®ve subjects in the emergency room. In only one Further study is needed to determine the relative
subject were the AJ and KJ present and the average value of the known prognostic signs such as sacral
recovery of these re¯exes was 7.7 and 15.2 days sparing of pinprick,16,34 return of voluntary movement
respectively. In our study, the ®rst elicitable re¯exes by 72 h,35 return of spinal re¯exes in the absence of
were the cutaneous, polysynaptic re¯exes (DPR, BC, recovering voluntary movement10 and a DPR, in
cremasteric) rather than the monosynaptic tendon establishing the prognosis for functional recovery at
re¯exes and this distinction was more apparent than the earliest possible date. An accurate prediction of
a caudal-to-rostral distinction. It appears that the recovery is essential to target medications and
cutaneous, polysynaptic re¯exes receive less supra- therapies appropriately toward ambulation or toward
spinal facilitation and/or recover faster than the wheelchair use, as well as to advise patient and family.
monosynaptic deep tendon re¯exes or synaptic zones Preliminary observations in animals and humans
are less disrupted because descending pathways make suggest that there may be a critical period during
less total contribution.30 acute rehabilitation when CNS stimulants,36 aggressive
In view of the above, it is dicult to support the active exercise,37 electrical stimulation38 and training,
proposition that one cannot prognosticate during such as body weight support39 ± 43 can enhance
spinal shock, when spinal shock is de®ned as the recovery of voluntary movement. As these new
absence of all re¯exes. Since all re¯exes are rarely rehabilitation interventions develop, failure by rehabi-
absent, we should discard this view of spinal shock. litation physicians to recognize these favorable
We should also discard the view of spinal shock that prognostic signs and failure to initiate these new
re¯ex return occurs in a caudal to rostral sequence, therapies may result in suboptimal functional out-
since the cutaneous re¯exes do not follow this rule. comes. Current studies of prognosis should be
Furthermore, the BC re¯ex is not typically the ®rst undertaken to incorporate new factors such as the
re¯ex to recover in patients with complete injuries. DPR and the severity of the damage to the spinal cord
Presence or absence of the BC re¯ex may aid in by MRI studies.19 These ®ndings could be used20 to
distinguishing between upper or lower motoneuron develop formulas that integrate all of the prognostic
 Reflex recovery during spinal shock
H-Y Ko et al
408

factors and take into account the new acute 13 Cushing H. Organization and activities of the neurological
service American expeditionary forces. In: Ireland MW, ed. The
interventions such as improved emergency medical medical department of the United States Army in the world war.
technician (EMT) services and neuroprotective Government Printing Oce: Washington, DC, 1927; pp 749 ±
agents.44 ± 46 758.
The DPR has prognostic signi®cance, when it 14 Weinstein DE, Ko HY, Graziani V, Ditunno Jr JF. Prognostic
persists for 48 h or longer in regard to the severity signi®cance of the delayed plantar re¯ex following spinal cord
injury. Spinal Cord Med 1997; 20: 207 ± 211.
of the injury (ASIA grade) and the potential for 15 Braakman R, Penning L. Injuries to the cervical spine. In: Vinken
ambulation at the time of discharge. The absence of all PJ, Bruyn GW (eds). Handbook of clinical neurology. North-
re¯exes was observed in less than 8% of cases on the Holland: Amsterdam 1976; 227 ± 380.
day of injury and it was usually two or more 16 Crozier KS, Graziani V, Ditunno Jr JF, Herbison GJ. Spinal
cutaneous re¯exes that were present. These cutaneous cord injury: prognosis for ambulation based on sensory
examination in patients who are initially motor complete. Arch
re¯exes are less depressed than the deep tendon Phys Med Rehabil 1991; 72: 119 ± 121.
re¯exes. The cremasteric re¯ex (L2) often precedes 17 Ditunno Jr JF, Stover SL, Freed MM, Ahn JH. Motor recovery
the AJ (L5, S1) and the DPR (S1) frequently precedes of the upper extremities in traumatic quadriplegia: a multicenter
the BC (S3-5), which is not consistent with a caudal- study. Arch Phys Med Rehabil 1992; 73: 431 ± 436.
18 Burns SP et al. Recovery of ambulation in motor-incomplete
rostral recovery pattern of re¯exes. The DPR, which tetraplegia. Arch Phys Med Rehabil 1997; 78: 1169 ± 1172.
often precedes or is seen at the same time as the BC, 19 Flanders AE et al. Acute cervical spine trauma: correlation of
gradually disappears and is followed by the Babinski MR imaging ®ndings with degree of neurologic de®cit. Radiology
and shortly thereafter by the AJ and KJ. 1990; 177: 25 ± 33.
These ®ndings warrant a reconsideration of the term 20 Flanders AE et al. Forecasting motor recovery after cervical
spinal cord injury: value of MR imaging. Radiology 1996; 201:
spinal shock, and assumptions based on previous 649 ± 655.
descriptions and a limited examination of re¯exes. 21 Dimitrijevic MR, Nathan PW. Studies of spasticicy in man. 3.
The evolution of spinal re¯exes over several days Analysis of re¯ex activity evoked by noxious cutaneous stimuli.
appears to be more relevant to prognosis for Brain 1968; 91: 349 ± 368.
22 Calancie B et al. Evidence that alterations in presynaptic
ambulation than the presence or absence of re¯exes inhibition contribute to segmental hypo- and hyperexcitability
on the day of injury. after spinal cord injury in man. Electroencephalogr Clin
Neurophysiol 1993; 89: 177 ± 186.
23 Croul SE, Flanders AE. Neuropathology of human spinal cord
References injury. Adv Neurol 1997; 72: 317 ± 323.
24 Bunge RP, Puckett WR, Hiester ED. Observations on the
pathology of several types of human spinal cord injury, with
1 Hall M. Synopsis of the diastaltic nervous system: or the system of
emphasis on the astrocyte response to penetrating injuries. Adv
the spinal marrow, and its re¯ex arcs; as the nervous agent in all the
Neurol 1997; 72: 305 ± 315.
functions of ingestion and of egestion in the animal oeconomy.
25 Ditunno Jr JF, Young W, Donovan WH, Creasey G. The
London: Mallett J., 1850;
international standards booklet for neurological and functional
2 Denny-Brown D. Motor mechanisms. In: Field J (ed). Handbook
classi®cation of spinal cord injury. American Spinal Injury
of physiology, section 1. Neurophysiology, part II. American
Association. Paraplegia 1994; 32: 70 ± 80.
Physiological Society: Washington, DC 1960, pp 781 ± 796.
26 Landau WM, Clare MH. The plantar re¯ex in man, with special
3 Illis LS. Clinical evaluation and pathophysiology of the spinal
reference to some conditions where the extensor response is
cord in the chronic phase. In: Illis LS (ed). Spinal Cord
unexpectedly absent. Brain 1959; 82: 321 ± 355.
Dysfunction. Oxford University Press: New York 1988; p107 ±
27 Anonymous. The sign of Babinski. Lancet 1960; 321 ± 322.
128.
28 van Gijn J. Pathophysiology. In: The Babinski sign ± a centenary.
4 Barnes CD, Schadt JC. Release of function in the spinal cord.
University Utrecht: Uttrecht 1996, pp 89 ± 117.
Prog Neurobiol 1979; 12: 1 ± 13.
29 Cavallari P, Pettersson LG. Tonic suppression of re¯ex
5 Mendell LM. Physiological aspects of synaptic plasticity: the Ia/
transmission in low spinal cats. Exp Brain Res 1989; 77: 201 ± 212.
motoneuron connection as a model. Adv Neurol 1988; 47: 337 ±
30 Illis LS. The motor neuron surface and spinal shock. Mod Trends
360.
Neurol 1967; 4: 53 ± 68.
6 Leis AA et al. Spinal motoneuron excitability after acute spinal
31 Bunge RP, Puckett WR, Becerra JL. Observations on the
cord injury in humans. Neurology 1966; 47: 231 ± 237.
pathology of human spinal cord injury: A review and
7 Stau€er ES. Rehabilitation of posttraumatic cervical spinal cord
classi®cation of 22 new cases with details from a case of chronic
quadriplegia and pentaplegia. In: Cervical Spine Research
cord compression with extensive focal demyelination. In: Seil FJ
Society ES (ed). The cervical spine. Lippincott: Philadelphia;
(ed). Advances in Neurology. Raven Press Ltd: New York 1993,
1983; p317 ± 322.
pp 75 ± 89.
8 Holdsworth F. Fractures, dislocations, and fracture-dislocations
32 Liu C, Chambers W. Interspinal sprouting of dorsal root axons.
of the spine. J Bone Joint Surg Am 1970; 52: 1534 ± 1551.
Arch Neurol 1958; 79: 46 ± 61.
9 Sherrington CS. The integrative action of the nervous system.
33 Penrod LE, Hegde SK, Ditunno Jr JF. Age e€ect on prognosis
Constable: London, 1906;
for functional recovery in acute, traumatic central cord
10 Guttmann L. Studies on re¯ex activity of the isolated cord in
syndrome. Arch Phys Med Rehabil 1990; 71: 963 ± 968.
spinal man. J Nerv Ment Dis 1952; 116: 957 ± 972.
34 Folman Y, el Masri W. Spinal cord injury: prognostic indicators.
11 Riddock G. The re¯ex functions of the completely divided spinal
Injury 1989; 20: 92 ± 93.
cord in man, compared with those associated with less severe
35 Brown PJ, Marino RJ, Herbison GJ, Ditunno Jr JF. The 72-hour
lesions. Brain 1917; 40: 264 ± 401.
examination as a predictor of recovery in motor complete
12 Guillain G, Barre JA. Etude anatomo-clinique de quinze cas de
quadriplegia. Arch Phys Med Rehabil 1991; 72: 546 ± 548.
section totalle de la moelle. Annales de Medicine 1917; 4: 178 ±
222.
 Reflex recovery during spinal shock
H-Y Ko et al
409

36 Walker-Batson D et al. Amphetamine paired with physical 42 Edgerton VR et al. A physiological basis for the development of
therapy accelerates motor recovery after stroke. Further rehabilitative strategies for spinally injured patients. J Am
evidence. Stroke 1995; 26: 2254 ± 2259. Paraplegia Soc 1991; 14: 150 ± 157.
37 Little JW, Harris RM, Lerner SJ. Immobilization impairs 43 Wernig A, Muller S. Laufband locomotion with body weight
recovery after spinal cord injury. Arch Phys Med Rehabil 1991; support improved walking in persons with severe spinal cord
72: 408 ± 412. injuries. Paraplegia 1992; 30: 229 ± 238.
38 Muir GD, Steeves JD. Sensorimotor stimulation to improve 44 Bracken MB et al. A Randomized, Controlled Trial of
locomotor recovery after spinal cord injury. Trends Neurosci Methylprednisolone or Naloxone in the Treatment of Acute
1997; 20: 72 ± 77. Spinal Cord Injury. New Eng J Med 1990; 322: 1405 ± 1411.
39 Barbeau H, Danakas M, Arsenault B. The e€ects of locomotor 45 Bracken MB et al. Administration of methylprednisolone for 24
training in spinal cord injured subjects: a preliminary study. or 48 hours or tirilazad mesylate for 48 hours in the treatment of
Restorative Neurology and Neuroscience 1994; p81 ± 84. acute spinal cord injury. Results of the Third National Acute
40 Fung J, Stewart JE, Barbeau H. The combined e€ects of Spinal Cord Injury Randomized Controlled Trial. National
clonidine and cyproheptadine with interactive training on the Acute Spinal Cord Injury Study. JAMA 1997; 277: 1597 ± 1604.
modulation of locomotion in spinal cord injured subjects. J 46 Geisler FH, Dorsey FC, Coleman WP. Recovery of motor
Neurol Sci 1990; 100: 85 ± 93. function after spinal-cord injury ± a randomized, placebo-
41 Curt A, Keck ME, Dietz V. Clinical value of F-wave recordings controlled trial with GM-1 ganglioside. New Eng J Med 1991;
in traumatic cervical spinal cord injury. Electroencephalogr Clin 324: 1829 ± 1838.
Neurophysiol 1997; 105: 189 ± 193.