Clinical Neurophysiology 115 (2004) 779–789

www.elsevier.com/locate/clinph

Group II spindle fibres and afferent control of stance.

Clues from diabetic neuropathy
Antonio Nardonea,*, Marco Schieppatib
a
Posture and Movement Laboratory, Division of Physical Therapy and Rehabilitation, Fondazione Salvatore Maugeri (IRCCS),
Scientific Institute of Veruno, I-28010 Veruno (Novara), Italy
b
Department of Experimental Medicine, Section of Human Physiology, University of Pavia and Human Movement Laboratory (CSAM),
Fondazione Salvatore Maugeri (IRCCS), Scientific Institute of Pavia, I-27100 Pavia, Italy
Accepted 10 November 2003

Abstract
Objective: Since patients with large-fibre neuropathy do not show abnormal body sway during stance, the hypothesis was tested that
postural control is not impaired until myelinated fibres of medium size are affected.
Methods: In 22 diabetic neuropathic patients and 13 normals, we recorded: (1) body sway area (SA), (2) stretch responses of soleus (Sol)
and flexor digitorum brevis (FDB) to toe-up rotation of a platform, (3) Sol and FDB H reflex and FDB F wave, (4) conduction velocity (CV)
of tibial, deep peroneal and sural nerve. In patients, detection thresholds for vibration, cooling (CDT), warming and heat-pain (HPDT) were
assessed.
Results: Body SA was increased in patients with respect to normals. Toe-up rotation elicited short- (SLR) and medium-latency (MLR)
responses in Sol and FDB in all normals. In patients, SLR was absent in FDB and reduced in Sol, and MLR was delayed in both muscles; the
FDB H reflex was absent. The CV of tibial nerve group II afferent fibres, as estimated from the afferent time of FDB MLR, was reduced in
patients. All sensory detection thresholds were increased. Stepwise multiple regression showed that increased SA was explained by increased
latency of MLR, decreased CV of group II fibres and augmented CDT and HPDT.
Conclusions: Unsteadiness in diabetic neuropathy is related to alterations in medium-size myelinated afferent fibres, possibly originating
from spindle secondary terminations.
q 2004 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: Sway; Diabetes; Neuropathy; Stretch reflex; Group II fibre

1. Introduction The above notion has been tackled by a recent

Lower limb muscle afferent fibres play a major role in the
reflex control of quiet upright stance (Fitzpatrick and
McCloskey, 1994; Peterka and Benolken, 1995). Accord-
ingly, increased body sway is found in many peripheral
neuropathies (Bergin et al., 1995; Schieppati et al., 1999),
particularly when patients stand upright with their eyes
closed. Loss or damage of spindle afferent fibres of large
diameter (group Ia) is considered responsible for the
instability (Weiss and White, 1986; Griffin et al., 1990;
Knazan et al., 1990; Kuwabara et al., 1999; Quintern et al.,
1999).
* Corresponding author. Tel.: þ 39-0322-884906; fax: þ 39-0322-
830294.
E-mail address: anardone@fsm.it (A. Nardone).
1388-2457/$30.00 q 2004 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.clinph.2003.11.007
investigation, which showed that patients affected by
Charcot-Marie-Tooth type 1A neuropathy (CMT1A) exhib-
ited good postural stability during quiet stance (Nardone
et al., 2000). These patients have demyelination and loss of
the largest motor and sensory fibres, including spindle Ia
fibres, while myelinated fibres of smaller calibre are much
less affected (Dyck et al., 1993a,b). Consistent with their
nerve pathology, these CMT1A patients show negligible
decrease in conduction velocity of the spindle group II fibres
(Nardone et al., 2000; Schieppati et al., 1995). Conversely,
spastic patients with hyperexcitability of the Ia-mediated
monosynaptic reflex do not show abnormal postural findings
(Nardone et al., 2001a,b). Therefore, selective impairment
in the structure or function of the feedback control based on
the Ia afferent input does not seem to be necessary for proper
stance control. Indeed, loss of the velocity-sensitive input
780 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789

from the spindle primaries would hardly affect stance, in
agreement with the predominantly low-velocity changes in
the leg muscle length occurring during quiet stance,
mirrored in the low-frequency displacement of the body
centre of mass (Gurfinkel, 1973; Nardone et al., 1997).
However, whether the proprioceptive information travelling
along fibres of smaller calibre than Ia (i.e. the group II
afferent fibres from the length-sensitive spindle secondaries)
plays the major role in postural control remains an open
question.
To investigate this possibility, we studied patients
affected by diabetic polyneuropathy, who show loss of
both large and small diameter myelinated fibres (Behse
et al., 1977; Le Quesne et al., 1990; Agostino et al., 2000;
Perkins and Bril, 2003). These patients do suffer from
impaired balance during quiet stance (Simoneau et al.,
1994; Uccioli et al., 1997). We hypothesized that, contrary
to CMT1A, diabetic patients would exhibit increased body
sway and decreased CV of group II fibres.
Therefore, we assessed the conduction velocity of these
fibres in a group of diabetic patients with variable degrees of
balance impairment by means of a method based on the
analysis of the latency of the stretch responses of postural
muscles in standing subjects (Schieppati et al., 1995). Other
factors possibly co-responsible for the poor postural control
were also tested, i.e. clinical and electrophysiological
variables related to the peripheral nervous system function.
Preliminary findings have been presented (Nardone et al.,
1998, 2001a).
2. Methods
2.1. Subjects
Twenty-two diabetic patients (12 men and 10 women;
age range 47 –83 years, mean 65.1 ^ 9.6 standard deviation
(SD); height 164.8 ^ 9.8 cm) were studied. Thirteen
unimpaired subjects constituted the control group (3 men
and 10 women; age range 43– 82 years, mean 65.3 ^ 10.9;
height 159.9 ^ 7.8 cm) (see Table 1). All subjects gave
informed consent to the study; the experiments were
conducted according to the Declaration of Helsinki, and
had been approved by the local ethics committee.
2.2. Clinical data
Seventeen diabetic patients were of adult onset, 5 of
juvenile. All patients were on medication: 10 assumed oral
antidiabetics, 10 insulin, and two both types of drug. On
average, the duration of diabetes (time from initial
diagnosis) was 13.4 ^ 8.7 years. All patients presented
with symmetrical distal sensorimotor polyneuropathy and
none had clinical signs of involvement of the central

Table 1
Clinical findings in the diabetic polyneuropathic patients

Patient ID Sex/Age NDS Muscle weakness Tendon reflex Sensory examination

(years)
Total Iliopsoas Glutei Ham Quad D-F P-F Achilles Quad Touch press Prick pain Vibrat. Joint pos.

1 A.G. M/77 22 0 0 0 0 0 0 2 2 1 1 1 1
2 A.G. F/70 25 2 2 1 1 0 0 2 0 0.5 0 1 1
3 B.A. F/63 11 0 0 0 0 0 0 1.5 1 0 0 0 0
4 B.W. M/43 17 0 0 0 0 0 0 2 1 0 0 2 1
5 B.V. M/63 17 0 0 0 0 0 0 2 2 0 0 0 0
6 C.L. M/50 40 1 1 1 1 0 0 2 2 1 2 2 1
7 D.M. F/54 27 2 3 3 3 2 0 2 1.5 0 1.5 0 0
8 D.R. M/63 65 1 1 1.5 1 2 2 1.5 2 0 0 2 1
9 F.G. M/71 58 1 1 1 1 0 1 2 2 1 1 2 1
10 G.C. M/77 33 1 0 0 1 0 0 2 2 1.5 1 2 1.5
11 M.A. F/75 20 0.5 0 0 1 0 0 2 1 0 0 0.5 0
12 M.C. M/54 7 0 0 0 0 0 0 2 1.5 0 1.5 0 0
13 M.E. M/46 18 0 0 0 1 0.5 0 2 0 0 0 0.5 0
14 M.V. M/69 16 0 0 0 0 0 0 0 1 0 1 2 0
15 N.M. M/73 14 0 0 0 0 0 0 2 2 1 1 2 1
15 N.C. F/79 13 0 0 1 1 1 1 0 0 0 0 2 0.5
17 P.B. F/72 6 0 1 0 0 0 0 2 0 0 0 0 0
18 Q.R. F/82 14 0 0 0 0 0 0 2 2 0 0 0 0
19 R.M. F/58 30 1.5 0 1 2.5 2.5 1 0.5 0.5 0.5 0 0 0
20 S.G. M/61 11 0 0 0 0 0 0 2 1 0 0.5 0 0
21 S.C. F/66 24 0 0 0 0 0.5 0 2 2 0 1 1.5 0
22 V.M. F/71 37 1 1 0 0 2 0 2 1 2 2 1.5 1.5

NDS, Neurological Disability Score (Dyck et al., 1980); Ham, hamstrings; Quad, quadriceps; D-F, dorsiflexors. P-F, plantarflexors. Touch press, touch-
pressure sense; Prick pain, pricking pain sense; Vibrat., vibration sense; Joint Pos., joint position sense. The scores of the clinical variables are the mean scores
from the two lower limbs. Scoring for muscle weakness: 0, no deficit; 1, mild deficit; 2, moderate deficit; 3, severe deficit; 4, complete absence of function or
extremely severe deficit. Scoring for reflexes and sensation: 0, normal; 1, reduced; 2, absent.
 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789
nervous system. A standard neurological examination was
performed and quantified according to the Neurological
Disability Score (NDS) (Dyck et al., 1980): lower limb
muscle strength (distal and proximal muscle groups), touch-
pressure, vibration (128 Hz tuning fork), joint position,
pricking pain, and quadriceps and Achilles tendon reflexes
were evaluated.
2.3. Electrophysiological study
M waves evoked by electrical stimulation of the parent
nerves were recorded from the flexor digitorum brevis
(FDB), extensor digitorum brevis (EDB), soleus (Sol) and
tibialis anterior (TA) muscles of all patients in order to
assess the degree of axonal damage (American Diabetes
Association, 1988). FDB and Sol H reflexes were evoked by
electrical stimulation of the tibial nerve at the internal
malleolus and popliteal fossa respectively, at rest or with
voluntary activation of the relevant muscle. Conduction
velocity (CV) and amplitude of the sensory action potential
(SAP) of the sural nerve and CV of motor fibres to the FDB
and EDB were measured in the leg. In particular, the CV of
the motor fibres of the tibial nerve directed to the FDB
muscle was obtained from the latencies of the FDB M wave
evoked by stimulating at the popliteal fossa and internal
malleolus. In order to estimate the CV of the entire nerve,
the F wave from the FDB muscle was also searched in 17
patients (Panayiotopoulos and Chroni, 1996); from the F-
wave latency, whenever it was elicited, the motor CV was
estimated by means of a formula described by Kimura et al.
(1975). To grade the electrophysiological severity of the
polyneuropathy, a composite ‘neuropathy score’ (NS) for
the lower limb was calculated (Bergin et al., 1995). This was
obtained by adding together the scores based on the M-wave
amplitudes of EDB and FDB, the amplitude of sural SAP,
and the CV scores for the 3 nerves.
2.4. Reflex responses to stance perturbation
Subjects stood with eyes open and feet spaced 10 cm
apart on a movable platform. Perturbations consisted in
20 –30 toe-up rotations of 38 amplitude and 50 deg/s. The
interval between trials varied between 8 and 10 s, to allow
full recovery to the initial quiet stance posture. The surface
EMG was recorded from the flexor digitorum brevis
(FDB), the soleus (Sol) and the tibialis anterior (TA)
muscle. The FDB muscle had been included in the
recording set, because the method for measuring the
conduction velocity of both the motor fibres and of
the sensory group II fibres can be reliably applied
(Schieppati and Nardone, 1999). The electrodes were
fixed at an inter-electrode distance of about 2 cm on the
sole of the foot midway between heel and metatarsal heads
(FDB), on the lower third of the leg posteriorly (Sol), and
on the upper third anterolaterally (TA). EMG was
amplified ( £ 10 000), bandpass filtered (100 Hz to 3 kHz,
781
2 6 dB/octave), converted analogue-to-digital at a
sampling rate of 1 kHz, and fed into a personal computer
together with the platform position signal. The acquisition
period was 500 ms. EMG signals were full-wave rectified
and averaged. Two subsequent EMG bursts (the short-
latency, SLR, and medium-latency response, MLR) are
always recorded in FDB and Sol muscles of normal
subjects in response to the perturbation. A late response in
the TA muscle, called antagonist reaction (AR) occurs after
about 140 ms (Nardone et al., 1990). The latency of the
responses (SLR, MLR and AR) was the time interval
elapsing between the onset of platform movement and the
beginning of the relevant burst of EMG activity, as
identified on the superimposed traces of the rectified and
averaged signals recorded from the muscles of the two
legs. For each of two homonymous muscles, the onset of
the responses was set at the time when the EMG increased
beyond two SDs of the mean value of the background,
prestimulus EMG activity (Schieppati and Nardone, 1997).
When the EMG did not diminish below 2 SDs between
the SLR and the MLR, the onset of the MLR was taken as
the lowermost point in the EMG trace. Uncertainties in the
measurements were resolved by comparing the super-
imposed traces from both legs. The latency values used in
the following analysis were the means of the response
latencies from the two legs. The offset of the signal was
the time when the EMG burst decreased below two SDs of
the mean value of the background EMG activity. The size
of the responses was calculated as the area subtended by
the rectified and averaged EMG trace. The limits within
which the area was computed were the onset and
termination of the EMG bursts (Schieppati et al., 1995).
The value of the response size used for further analysis was
the mean of the sizes measured in the two legs.
2.5. Estimation of CV of group II fibres
The latency of the FDB-MLR was used to calculate the
peripheral afferent time of the fibres belonging to group
II, which conduct the afferent volley responsible for the
MLR. We subtracted from the response latency the
efferent time of the motor fibres. To obtain the efferent
time, the distance from FDB muscle to the T12-L1 inter-
vertebral space was measured with a tape and divided by
the CV of the FDB motor fibres of the distal part of the
nerve (see Section 2.3.), after check that the same CV was
common to both distal and proximal part of the nerve (see
above and Fig. 1). We then subtracted the neuromuscular
delay (1 ms), the peripheral lag due to platform inertia
(4 ms) and the central delay of the MLR (6.7 ms)
(Nardone and Schieppati, 1998; Schieppati and Nardone,
1999). From the afferent time and the distance between
the FDB and the corresponding spinal neuromere, the CV
of the afferent fibres mediating the MLR was estimated
(Nardone et al., 2000).
782 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789
Fig. 1. (Upper graph) Motor conduction velocity (CV) of tibial nerve to
FDB, obtained by stimulating the nerve at the popliteal fossa and internal
malleolus, is slower in diabetic patients than normal subjects
(***P , 0:0005). (Lower graph) The individual motor CV values
(abscissa), from which the mean value in the right bar of the upper graph
was obtained, are plotted versus the CV of the same nerve obtained from the
F-wave latency (ordinate). The association between the two sets of data
indicates that the distal part and the whole length of the motor fibres were
similarly affected in the patient population. The equation of the best-fit line
is FWCV ¼ CV £ 1:7 2 21:8 (P , 0:001; r 2 ¼ 0:70).
2.6. Instrumentally-assessed sensory detection threshold
Using a fully automated testing system (Computer
Assisted Sensory Examination, CASE IV), the detection
thresholds for vibration (VDT), cooling (CDT), warming
(WDT), and heat-pain (HPDT) were quantified in the
patients. Although CDT, WDT and HPDT should be not
directly relevant to postural control in upright stance, these
detection thresholds have been evaluated to try to assess the
whole spectrum of sensory fibre impairment in patients, and
to further support the hypothesis that impairment of small
afferent fibres is related to postural instability. CASE IV uses
a discrete set of 25 standardized stimulation levels for patient
testing and analysis. These levels are termed ‘just noticeable
differences’ or JNDs. The JNDs have a baseline level: for
vibration stimuli, the baseline is 0 mm of displacement; for
cooling, 30 8C; for heat-pain, 34 8C (Dyck et al., 1993b). VDT
was assessed by administering the vibratory stimuli to the
dorsal aspect of the big toe. The stimulus was delivered
through a 9 mm-diameter probe vibrating at a frequency of
125 Hz. The pre-load force of the probe acting on the toe was
30 g. During the test, each patient was given 15 stimuli of
different amplitudes (between 0 and 576 mm), preceded by a
warning signal. The CDT, WDT and HPDT were assessed by
administering a series of thermal stimuli to the dorsal surface
of the foot through a probe. According to the 4, 2, 1 stepping
algorithm (Dyck et al., 1993b), during the testing process 5
null stimuli were placed randomly among a total of 20
stimulus trials to prevent false results. Testing began at
a middle JND level. Patients were asked to press a ‘yes’ or
‘no’ button with the index finger in response to each stimulus
whether detected or not. When the test was complete, the
system calculated the detection threshold of the above
sensory modalities.
2.7. Stabilometry
Subjects were required to stand still for 51 s on a
dynamometric platform (Kistler, type 9281B, Switzerland).
In a first series of trials, the feet were spaced 10 cm apart
(FA); in a second series, feet were together (FT). These
postures were selected because they are frequently used in
the literature, and for the sake of comparison with previous
results in neuropathic patients affected by Charcot-Marie-
Tooth disease (Nardone et al., 2000). Between the two series
subjects were allowed to sit. Each trial within a series was
repeated twice, alternatively with eyes open (EO), when
subjects gazed at a target placed at 50 cm at eye height, and
eyes closed (EC). The 3 components of the force acting on
the platform were sampled at 10 Hz. At the end of each trial,
a program calculated the instantaneous changes in position
of the resultant vector, from which the average centre of foot
pressure (CFP) was obtained. The following variables were
then measured: (1) The sway area (SA), i.e. the area of the
surface swept by the shifts of the line joining the position of
the average CFP to the successive positions of the
instantaneous centre of pressure (Mauritz et al., 1979;
Diener et al., 1984a); (2) mean CFP-AP and CFP-ML, the
average position of CFP along the anterior-posterior and
medial-lateral axes, respectively. To normalize all subjects’
data, CFP-AP was expressed as a percentage of the total foot
length, and CFP-ML as a percentage of body height.
2.8. Statistical analysis
When not differently stated, values are expressed as
mean ^ SD. The findings obtained in the stabilometric tests
were analysed separately according to the postural and
visual conditions (FA EO, FT EO, FA EC, FT EC). For all
subjects, two trials for each condition were considered for
analysis. Since sway values were not normally distributed,
logarithmic transformation was made. For stabilometric
findings, 3-way (2 groups £ 2 postural £ 2 visual con-
ditions) analysis of variance (ANOVA) was used to assess
differences between patients and controls. When the result
of the ANOVA was significant (P , 0:05), the Newman-
Keuls post hoc test was run. For EMG and postural
responses data, the unpaired t test between normal subjects
and patients was performed. Test of slopes and parallelism
between regression lines fitted through the data points was
performed through the ANOVA first, followed by the t test
for the slopes, and through the analysis of co-variance
(ANCOVA) followed by the t test for the intercepts
(Armitage, 1971). The relationship between clinical and
electrophysiological data as independent variables and body
 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789 783

sway recorded under the various postural and visual
conditions as dependent variables was analysed through
stepwise multiple regression analysis. In order to assess
which of the independent variables contributed most to the
prediction of body sway, the regression coefficients were
examined. The contribution of any independent variable
was considered significant when its regression coefficient
was significant (P , 0:05). The software Statistica (Stat-
Soft, Tulsa, OK) was used for data analysis.
3. Results
3.1. Clinical data
Most diabetic patients showed absent knee and ankle
jerks in response to tendon percussion, and decreased
vibration and touch-pressure sense (Table 1). On average,
severity of neuropathy (assessed through the neurological
disability score, NDS) was relatively minor. In detail,
muscle weakness was minimal (below 1 score of the NDS
scale, on the average), and there was no difference between
the tested muscles regardless of their proximal or distal
position and function. Decrease of reflexes was scored
1.7 ^ 0.6 and 1.3 ^ 0.8 (paired t test, P ¼ 0:01) for Sol and
quadriceps, respectively. Decrease of sensation was scored
below 1 score of the NDS scale, on the average, for touch-
pressure, pricking pain and joint position; it was 1 score for
vibration. Vibration sense proved to be significantly
(P , 0:001) more impaired than the other sensations.
3.2. Electrophysiological study
Sural nerve sensory action potential was not obtained in
18 patients; it was reduced in two and normal in two others
(Table 2). Motor CV of the tibial nerve to FDB (the mean
values are reported in Fig. 1, top graph) and the deep peronal
nerve to EDB was decreased in 13 and 12 patients,
respectively, compared to normal values. M-wave ampli-
tude of FDB, EDB and Sol was below the laboratory normal
value in 19, 14 and 6 patients, respectively. The F wave in
the FDB muscle was tested in all but 3 patients, and could
not be elicited in two of them. In the others, the motor
conduction velocity was calculated from the F-wave latency
(FWCV), according to the formula of Kimura et al. (1975);
its value was very similar (and not significantly different,
paired t test) to that calculated on the distal part of the nerve
(Fig. 1, bottom graph), broadly indicating that both the distal
and proximal parts of the nerve were similarly affected. The
soleus H reflex could not be elicited in 17 patients. In the
remaining 5 patients, it occurred at a mean latency of
34.8 ^ 3.0 ms, significantly (P , 0:01) longer than that of
normal subjects (29.1 ^ 2.5 ms). On average, Hmax/Mmax
ratio was 8.5 ^ 18.4%, significantly smaller (P , 0:02)
than in normal subjects (32.0 ^ 17.2%). Only in 8 patients
was the FDB H reflex sought, but in none could it be evoked.
3.3. Reflex responses to stance perturbation
During the 100 ms period before perturbation, back-
ground EMG activity was present in both groups of subjects.

Table 2
Results of electrophysiological study and vibratory detection threshold assessment in the diabetic polyneuropathic patients

Patient ID Sex/Age M wave M wave M wave Sural nerve Tibial nerve Deep peroneal F wave FWCV Tibial NS FDB MLR Sol MLR
(years) FDB (mV) EDB (mV) Sol (mV) SAP (mV) CV (m/s) nerve CV (m/s) FDB (ms) nerve (m/s) (ms) (ms)

1 A.G. M/77 0.4 1.0 12.3 0 32 34 71 32 10 119 111

2 A.G. F/70 3.6 2.9 n.t. n.t. 32 36 59 33 7 127 91
3 B.A. F/63 2.8 4.2 1.0 0 32 36 59 36 6 115 83
4 B.W. M/43 9.0 2.0 15.2 4.5 35 38 55 39 6 111 85
5 B.V. M/63 11.0 11.5 14 0 34 37 62 38 4 103 99
6 C.L. M/50 1.8 0.3 10.6 0 37 34 57 40 13 112 88
7 D.M. F/54 2.1 0.7 n.t. 0 31 38 77 28 10 119 82
8 D.R. M/63 0.8 0.5 n.t. 6 34 22 63 35 13 n.t. n.t.
9 F.G. M/71 17.0 10.0 18 0 40 41 49 46 2 113 87
10 G.C. M/77 0.3 0.1 10.0 0 32 30 n.t. n.t. 10 135 116
11 M.A. F/75 1.0 6.2 12.0 0 36 40 63 32 7 106 79
12 M.C. M/54 10.0 5.5 8.3 2 36 38 58 39 2 99 84
13 M.E. M/46 0.4 0 n.t. 0 30 n.m. n.t. n.t. 13 128 92
14 M.V. M/69 1.0 7.7 0.5 0 31 30 70 34 8 126 99
15 N.M. M/73 1.0 1.6 6.7 0 33 35 66 31 9 122 88
15 N.C. F/79 5.5 10.0 2.5 3 36 44 53 42 2 100 76
17 P.B. F/72 1.8 1.3 8.2 0 42 40 n.t. n.t. 5 n.t. n.t.
18 Q.R. F/82 2.0 2.0 3.0 2 31 37 n.t. n.t. 6 138 113
19 R.M. F/58 4.3 0 7.6 0 34 n.m. 60 36 7 121 76
20 S.G. M/61 6.7 6.7 10.3 0 36 36 61 36 3 134 85
21 S.C. F/66 5.0 0 11.6 0 39 n.m. n.t. n.t. 7 104 87
22 V.M. F/71 1.4 1.0 1.7 0 29 29 83 23 9 125 121

FWCV, tibial nerve CV calculated from latency of FDB F wave; NS, neuropathy score (Bergin et al., 1995) (0 corresponds to normal); FDB and Sol MLR,
onset of medium-latency response of FDB and Sol, respectively; n.m., not measurable; n.t., not tested.
784 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789
The EMG area of FDB and Sol background activity in
normal subjects was 0.65 ^ 0.29 and 1.0 ^ 0.48 mV·ms,
respectively, whilst in diabetic patients it was 0.74 ^ 0.45
and 1.62 ^ 0.77 mV·ms. The background activity of Sol
was therefore just (P , 0:02) larger in diabetic patients than
normal subjects.
Typical averaged EMG responses of FDB and Sol to toe
up perturbation, obtained from a normal subject and a
patient, are reported in Fig. 2 (upper and middle panels,
respectively). In this subject, the perturbation evoked in
both muscles a SLR followed by a MLR. In the patient, SLR
was absent in the FDB and inconstant in the Sol, but a full-
size MLR was present in both muscles.
In 14 patients out of 22, no SLR was evoked in the FDB.
In the other 8 patients, the area of FDB SLR was slightly
smaller (0.38 ^ 0.23 mV·ms) than that of normal subjects
(0.43 ^ 0.32 mV·ms). In 5 patients, the Sol SLR was
bilaterally absent, while in the other 17 patients, its area was
0.79 ^ 0.40 mV·ms, i.e. smaller than that of normal subjects
(1.26 ^ 0.96 mV·ms), though not significantly so.
In all patients, including those with no or small SLR in
either or both the Sol and FDB muscle, a clear-cut MLR was
present in both muscles, albeit significantly (P , 0:001)
delayed in latency. The delay was about 20 ms for FDB and
18 ms for Sol, on average, with respect to normal subjects
(Fig. 2 (lower panel) and Table 2). The duration of both Sol
and FDB MLR was similar in patients and normal subjects,
being respectively 57.7 ^ 17.0 and 53.7 ^ 20.1 ms for Sol
MLR, and 64.0 ^ 15.5 and 67.6 ^ 20.2 ms for FDB MLR.
Fig. 2. Superimposed traces of flexor digitorum brevis (FDB) (left column)
and soleus (right column) EMG responses to toe up perturbation, obtained
from a normal subject (upper panel) and a diabetic patient (middle panel).
Average of 30 rectified and filtered (time constant ¼ 1 ms) EMG traces. In
the normal subject, a short-latency (SLR) and a medium-latency response
(MLR) were evoked in both muscles. In the patient, the SLR was evoked
only in the soleus, and was smaller than in the normal subject; the MLR
could be evoked in both muscles, though at a delayed latency. Lower panel.
The latency of onset of FDB and soleus MLRs was significantly delayed in
the patients. Mean latency ^ standard deviation (SD) from all normal
subjects and patients. **P , 0:005:
Fig. 3. Flexor digitorum brevis (FDB), soleus and tibialis anterior (TA)
EMG responses to toe up perturbation, obtained from all normal subjects
(left column) and diabetic patients (right column) (average of the rectified
and filtered EMG traces from all subjects). In the normal subjects, a short-
latency response (SLR) and a medium-latency response (MLR) were
evoked in both FDB and Sol. In the patients, a small SLR was present only
in the Sol, and was smaller than in normal subjects. The MLR was evoked
in both muscles of diabetic patients at a delayed latency with respect to
normals. The amplitude of Sol MLR is somewhat larger in the patient
average traces, with respect to normal subjects, due to the later occurrence
of the antagonist reaction (AR) in the patients’ TA.
In the patients, the area of Sol was 3.26 ^ 1.46 mV·ms,
significantly (P , 0:001) larger than in normal subjects, in
whom it was 1.44 ^ 0.56 mV·ms. Conversely, area of FDB
MLR was 3.28 ^ 2.39 mV·ms, a value slightly larger than
that of normal subjects (2.10 ^ 1.14 mV·ms) (P not
significant). After 150 ms an antagonist reaction (AR)
occurred in the TA in normal subjects (Fig. 3) (Nardone
et al., 1990). This response reciprocally inhibited the Sol
MLR. On the contrary, in patients the area of the Sol MLR
was not aborted by the reciprocal inhibitory effect induced
by the AR. In fact, the AR was also delayed in patients,
occurring at about 170 ms latency, i.e. after the completion
of the Sol MLR.
3.4. Conduction velocity of group II fibres
In the FDB muscle, the delayed latency of the MLR
in diabetic patients was in part due to the slowed CV of
its motor fibres, which was 34.3 ^ 3.3 m/s, significantly
(P , 0:001) slower than that of the normal subjects
(Fig. 1, upper graph). The efferent time of the MLR in
patients was 34.4 ^ 3.2 ms, significantly (P , 0:001)
longer than in the normal subjects (Fig. 4, upper graph).
The afferent time of the MLR was then obtained as the
difference between the total latency of the response and
the efferent time plus the correction factors for the
central and peripheral delays as described in Section 2
(Fig. 4, middle graph). The mean afferent time proved
to be 67.3 ^ 9.7 ms, significantly (P , 0:01) longer in
 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789 785

Fig. 5. Detection thresholds of patients for vibration, warming, cooling, and

heat-pain, expressed as Just Noticeable Difference (JND). The data points
in the upper row belong to patients who were insensitive to the stimuli.

3.6. Stabilometry

Fig. 6 (upper graph) shows an example of body sway area

recorded by means of the stabilometric platform in a normal
subject and in a representative diabetic patient during stance
with feet apart under both visual conditions. Fig. 6 (middle

Fig. 4. (Upper graph) The efferent time of FDB MLR, calculated on the
basis of the CV of tibial nerve motor fibres, is significantly longer in
diabetic patients than in normal subjects. (Middle graph) The afferent time
of MLR, calculated as the difference between latency and efferent time of
MLR (plus the correction factors for the central and peripheral delays, see
Section 2), is longer in the patients. (Lower graph) The conduction velocity
of FDB group II afferent fibres is slower in the patients. Mean values ^
standard deviation (SD) from all normal subjects and patients.
***P , 0:0005; **P , 0:005; *P , 0:05:

patients than in controls, indicating that the delay of the

MLR observed in patients was explained not only by
the decreased CV of motor fibres but also by the slowed
CV of the afferent pathway. The estimated CV of the
afferent fibres responsible for the MLR was 17.7 ^ 2.9
m/s in the patients, on average, significantly (P , 0:05)
slower than that of the normal subjects (Fig. 4, lower
graph).

3.5. Instrumentally-assessed sensory detection threshold

Fig. 5 shows the average values of the detection

thresholds for vibration (VDT), cooling (CDT), warming
(WDT), and heat-pain (HPDT) obtained in the patients. In 4,
6, 9 and 4 patients no just noticeable difference (JND) could
be assessed for VDT, CDT, WDT and HPDT, respectively.
The increased JNDs are in keeping with the loss of the
relevant afferent fibres. The large changes in WDT, CDT
and VDT indicate that the whole spectrum of afferent fibres
is affected by the disease, and that small as well as large
afferent fibres are severely affected.
Fig. 6. (Upper graph) Sway of centre of foot pressure (CFP), recorded by the
stabilometric platform in a normal subject and a representative diabetic
patient during stance, under eyes open (EO) and eyes closed (EC)
conditions with feet apart. Note the increased sway and the forward shift of
CFP in the patient. Sway area (middle graph) and distance between heel and
CFP along the anterior-posterior axis (lower graph) are shown for all
postural (feet apart, FA, and feet together, FT) and visual conditions.
Average values (^ SD) from all normal subjects and patients.
***P , 0:0005; **P , 0:005; *P , 0:05:
786 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789
and lower graph) summarizes the mean values of body sway
area (SA) (middle graph) and average position of CFP along
the anterior-posterior (CFP-AP) axis (lower graph) recorded
under all postural and visual conditions in normal subjects
and patients. ANOVA showed that the SA of patients was
significantly (F ¼ 36:4; df ¼ 1; 248; P , 0:001) increased
with respect to normal subjects. There was also a significant
effect of postural (F ¼ 112:2; df ¼ 1; 248; P , 0:001) and
visual (F ¼ 97:7; df ¼ 1; 248; P , 0:001) conditions, SA
increasing under either eyes closed or feet together
conditions. However, the interaction between groups and
conditions was not significant, suggesting that the disease’s
effect on the performance was similar under the various
postural and visual conditions. The post hoc test showed
significant (P , 0:01) differences between normal subjects
and patients in all the corresponding conditions. ANOVA
showed that the CFP-AP was significantly shifted forward in
diabetic patients with respect to normal subjects (F ¼ 43:0;
df ¼ 1; 236; P , 0:001). On average, this forward shift was
about 7% of foot length (i.e. about 1.8 cm), and was
unaffected by either visual or postural condition. No
difference in CFP-ML position was found between patients
and normal subjects (not shown).
3.7. Relationship between sway area and clinical variables
For all patients, SA values obtained under each visual
and postural condition were plotted against the neurological
disability score (NDS) (Table 1). No significant relationship
was found between SA and the NDS under any condition.
The association of clinical values obtained from the single
NDS scores (strength of Sol and TA muscles, Achilles’ and
quadriceps tendon reflexes, touch-pressure, vibration, joint
position and pricking pain sense) with SA values obtained
under each visual and postural condition was assessed
through multiple linear regression. No significant regression
was found between any clinical variable and SA values.
3.8. Relationship between sway area and peripheral nerve
fibre function
Across all patients, no relationship was found between
the composite neuropathy score (NS) (Table 2) and body SA
values. NS is a score derived from the electrophysiological
assessment of large motor and sensory nerve fibres of lower
limb muscles and gives no information about the involve-
ment of small-diameter myelinated afferent fibres. Then, to
try to correlate small fibre damage with sway in diabetic
patients, we calculated the multiple regression between SA
as the dependent variable and the following independent
variables: CV of motor fibres and of group II afferent fibres,
latency of Sol MLR, and detection thresholds for vibration
(VDT), cold (CDT), warm (WDT) and heat pain (HPDT).
The body sway values obtained in diabetic patients under all
postural and visual conditions entered the stepwise
regression model. There was a different strength in
Fig. 7. Summary of the stepwise multiple regression analysis between sway
area under feet apart (FA) condition and peripheral nerve fibre functional
items. With EO, increase of sway area was significantly explained by the
increased latency of the Sol MLR and increased detection thresholds for
cooling and heat-pain. With EC, in addition to the previous variables, the
CV of group II fibres was significantly and inversely related to sway area.
Note that CV of motor fibres was positively related to SA, speaking against
a role of motor fibre impairment in instability. ***P , 0:0005;
**P , 0:005; *P , 0:05:
the association between variables: under FA conditions
(EO and EC) the coefficient of determination (r 2) was
greater than 0.6, while under FT conditions (EO and EC) it
was smaller than 0.3. Therefore, the effect of the single
independent variables on body sway was evaluated only for
the FA condition. The variable which explained most of the
regression (beta coefficient) under FA EO and FA EC
conditions was the latency of Sol MLR (P , 0:001). HPDT
and CDT were also significantly related to the increase of
sway area (P , 0:001 and , 0:05; respectively). Under FA
EC condition only, the CV of both group II and motor fibres
significantly (both P , 0:05) contributed to the regression.
In particular, the slope of the relationship between CV of
group II fibres and SA was negative, indicating that slowing
of these fibres was indeed associated to increased SA. On
the contrary, the slope of the relationship between CV of
motor fibres and SA was positive, pointing to a minor role, if
any, of motor fibre impairment in increasing SA. A
summary of the beta regression coefficients for the above
condition is shown in Fig. 7.
4. Discussion
Diabetic patients oscillate during upright stance signifi-
cantly more than normal subjects (Simoneau et al., 1994;
Boucher et al., 1995; Uccioli et al., 1997; Di Nardo et al.,
1999). In our patients, who were on medication and
generally well compensated, an abnormal control of stance
could also be detected. Most likely, sway increase was
related to the impairment of sensory, rather than motor
fibres. This was not unexpected, since the significant
impairment of the motor fibres, as witnessed by their
decreased conduction velocity and muscle force, exhibits no
connection with sway, under any postural or visual
condition (Bergin et al., 1995). In our sample, motor CV
was not severely decreased in most patients, and those
patients with decreased lower limb muscle strength and
 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789
M-response amplitude were among the more stable ones.
These findings are an indication of the minor role of the
efferent pathway and leg muscle force output in stance
control during quiet stance. In passing, these findings
confirm the notion that a surprisingly minimal force is
necessary for the maintenance of quiet upright posture, and
emphasize the role of timely and accurate nervous control of
stance rather than the need for a constant stiffness of the
postural muscles (Schieppati et al., 1994; Morasso and
Schieppati, 1999). Interestingly, also in CMT1A polyneuro-
pathic patients, motor fibres and muscle units are damaged,
even to a larger extent than in diabetic patients, but balance
control is barely degraded compared to both the diabetic
patients studied here (Nardone et al., 1998) and to normal
subjects (Nardone et al., 2000). The greater abnormalities in
sway control during upright stance in diabetic compared to
CMT1A patients must therefore be based on a damage to
sensory fibres. Morphometric studies (Le Quesne et al.,
1990; Dyck et al., 1993a) have shown that both diseases
share a loss of large myelinated motor and sensory nerve
fibres. Accordingly, T and H reflexes in the lower limb
muscles clinically and instrumentally assessed vibration
perception, and movement sense in the distal leg were
similarly decreased in both CMT1A (Nardone et al., 2000b)
and in the diabetic patients studied here. However, whilst
CMT1A patients show demyelination mainly of the largest
fibres, with smaller myelinated fibres relatively spared
(Dyck et al., 1993a), diabetic patients show loss of both
large and small nerve fibres, with the small ones usually
more affected (Le Quesne et al., 1990). This is in keeping
with our finding that 11 out of 22 diabetic but only 3 out of
15 CMT1A patients (Nardone et al., 2000b) had impairment
in the pricking pain sensation, which is mediated by small-
diameter myelinated fibres. These considerations lead us to
put forward the notion that in diabetic patients a pattern of
sensory neuropathy different from that in CMT1A patients
accounts for the increased body sway observed during quiet
stance. In particular, the critical factor may be related to loss
or impairment in afferent myelinated fibres of smaller
diameter than the large Ia fibres in diabetic patients.
The findings from the perturbation protocol give further
indications that a prominent role in reducing stance control
must be played by impairment of the group II fibres. The
short-latency responses to muscle stretch were absent or
very small in diabetic patients, as also expected from the
absence of the H reflex, and in keeping with loss of large
afferent fibres. However, the medium-latency responses
(MLRs) were delayed with respect to normal subjects
(Inglis et al., 1994), in keeping with impairment of group II
fibres. Admittedly, these responses were not diminished in
size, but could be even larger than in normal subjects. Two
considerations can be made in connection with this finding:
first, the background Sol and FDB EMG was larger in
patients, due to their slightly forward inclined posture (see
also below), and this is known to facilitate the appearance of
the MLR (Schieppati et al., 1995). Second, the final part of
787
the Sol MLR in patients was not aborted by the reciprocal
inhibitory effect induced by the ‘antagonist reaction’ (AR),
or the activation of the TA muscle normally present at this
time (Nardone et al., 1990): in fact, the AR was also delayed
in patients, occurring at about 170 ms latency, i.e. after the
completion of the Sol MLR.
The report of increased body sway during quiet upright
stance in diabetic patients is not new (Simoneau et al.,
1994; Boucher et al., 1995; Uccioli et al., 1997; Schieppati
and Nardone, 1999). However, it is shown here that
postural unsteadiness is typical for a neuropathy affecting
both large and smaller-diameter afferent fibres, and is not
typical of all peripheral polyneuropathies, in which the
largest fibres are almost selectively affected, as in CMT1A
disease (Nardone et al., 2000b). On the basis of the
previous data on CMT1A and of the present findings in
diabetic patients it appears therefore safe to propose that:
(a) the impairment of the largest motor and sensory fibres,
notably the a-motoneurones and group Ia fibres, is by itself
not detrimental to the control of upright posture, and (b)
the impairment of both large and small afferent fibres is
detrimental to postural control. This proposition is based
on the finding that not only the SLR was absent and
reduced in amplitude in diabetic patients (in addition to
being delayed when present), but also that the MLR was
disproportionately delayed. In fact, the disease must have
severely affected also the group II spindle afferent fibres,
the CV of which for the FDB muscle could be as slow as
12 m/s, while it is about 20 m/s on average in normal
subjects (Schieppati et al., 1995; Nardone et al., 1996;
Nardone and Schieppati, 1998). The notion that group II
afferent fibres play a specific role in controlling body sway
during quiet stance is further supported by the finding of a
significant negative relationship between sway area (under
eyes closed condition) and the conduction velocity of the
group II afferent fibres (Fig. 7).
The results of the multiple regression analysis are not at
odds with the above notion. Increased sway appeared to be
significantly explained by the latency of the Sol MLR,
cooling and heating pain thresholds. This corresponds to the
progressive impairment in smaller diameter fibres and sway
control in diabetic patients. Notably, the standardized
regression coefficient for Sol MLR latency was much larger
than that for cooling and heating pain, indicating a major
role in sway control for changes in proprioceptive input
mediated by medium-size fibres rather than for cutaneous
sensation from small fibres. Note that the relation between
increased latency of Sol MLR and sway depends only on the
lower CV of the sensory than of motor fibres, since the
contribution to the regression of the motor fibre CV was
opposite to what one would have been predicted. Further,
under eyes-closed condition, the estimated CV of the group
II fibres appeared also to be inversely related to sway area.
The increased latency of Sol MLR and the diminished CV of
group II fibres from the FDB muscle can therefore be
considered two aspects of the same deficit. Taking into
788 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789
consideration that (a) selective sensory blockade (Dietz
et al., 1980; Diener et al., 1984b; Magnusson et al., 1990),
(b) loss of large-diameter spindle fibres (Nardone et al.,
2000b) may not affect balance, and that (c) impairment of
conduction velocity of motor fibres plays no role, these
findings point strongly to the role of the lower limb
secondary spindle afferent fibres in the control of sway.
We do not deny that large fibres other than Ia can play a
role in postural tasks, as for instance those originating from
skin receptors of the foot sole. In fact, the increased threshold
to cutaneous vibration found with CASE IV would suggest an
impairment of large cutaneous fibres. Cutaneous input has
been reported to play both a minor (Mauritz and Dietz, 1980;
Diener et al., 1984b) and a sizeable role in balance control
during quiet stance (Magnusson et al., 1990). In the present
study, diabetic patients exhibited no significant relationship
between impairment of clinically tested touch-pressure sense
and sway. This is at variance with the notion that residual
cutaneous afferent fibres could play a critical role in stance
control.
The association, observed under eyes-closed condition,
between diminished CV of group II fibres and increased
sway suggests that when vision is denied the central nervous
system has to rely on the input from the myelinated
medium-size afferent fibres. In this context, the finding of a
forward shift of the CFP under all postural and visual
conditions in diabetic patients with respect to normal
subjects may be interpreted as an adaptive behaviour to
obtain a ‘safer’ body position during stance, diminishing the
risk of falling backwards (Schieppati et al., 1994). Another,
not mutually exclusive interpretation for this change of
posture may be that the minor but not negligible lengthening
of the triceps surae muscle leads to a ‘compensatory’
increase of static sensory input from the spindles secondary
terminations.
A mild impairment in central sensorimotor conduction
has been shown to exist in diabetic patients (Abbruzzese
et al., 1993); however, any role of such impairment in
deteriorating sway in diabetic patients has been discarded
(Uccioli et al., 1997). The lower limb polyneuropathy must
then be the real cause of the impaired stance control in
diabetics (Cavanagh et al., 1992).
As a general consideration, it is conceivable that the
prevalent role played by the large- and medium-size afferent
fibres may depend on the ongoing postural task. Given the
velocity sensitivity of primary muscle spindle terminals, one
would predict a key role of Ia fibres under conditions in
which subjects have to correct rapid perturbations of the
body (Dietz et al., 1980; Nardone et al., 1990; Corna et al.,
1999; Schieppati et al., 2002). Conversely, where very slow
movements are concerned, as during maintenance of quiet
stance, the signals from the length-sensitive spindle
secondaries may be largely sufficient to provide the input
to the central nervous system needed to detect the low-
frequency changes of toes and ankle joints (Gurfinkel, 1973;
Nardone et al., 1997). Interestingly, under quiet upright
stance the afferent input from Ia afferent fibres is gated by
the presynaptic inhibition (Nardone et al., 1986; Katz et al.,
1988; Koceja et al., 1993), further lessening the potential
functional meaning of proprioceptive Ia input under this
postural condition.
Acknowledgements
This work was supported in part by a ‘Ricerca
Finalizzata’ 2001 Grant (no. 174) from the Italian Ministry
of Health. We thank Dr Stefano Corna for patient referrals,
Mrs Margherita Grasso for technical assistance, and Dr
Rosemary Allpress for scrutinizing the English.
References
Abbruzzese G, Schenone A, Scramuzza G, Caponnetto C, Gasparetto B,
Adezati L, Abbruzzese M, Viviani GL. Impairment of central motor
conduction in diabetic patients. Electroenceph clin Neurophysiol 1993;
89:335–40.
Agostino R, Cruccu G, Romaniello A, Innocenti P, Inghilleri M, Manfredi
M. Dysfunction of small myelinated afferents in diabetic polyneuro-
pathy, as assessed by laser evoked potentials. Clin Neurophysiol 2000;
111:270– 6.
Armitage P. Statistical methods in medical research. Oxford: Blackwell
1971;279– 301.
American Diabetes Association. Report and recommendations of the San
Antonio conference on diabetic neuropathy. Neurology 1988;38:
1161–5.
Behse F, Buchthal F, Carlsen F. Nerve biopsy and conduction studies in
diabetic neuropathy. J Neurol Neurosurg Psychiatry 1977;40:1072 –82.
Bergin PS, Bronstein AM, Murray NMF, Sancovic S, Zeppenfeld K. Body
sway and vibration perception thresholds in normal aging and in
patients with polyneuropathy. J Neurol Neurosurg Psychiatry 1995;58:
335 –40.
Boucher P, Teasdale N, Courtemanche R, Bard C, Fleury M. Postural
stability in diabetic polyneuropathy. Diabetes Care 1995;5:638– 45.
Cavanagh PR, Derr JA, Ulbrecht JS, Maser RE, Orchard TJ. Problems with
gait and posture in neuropathic patients with insulin-dependent diabetes
mellitus. Diabetic Med 1992;9:469–74.
Corna S, Tarantola J, Nardone A, Giordano A, Schieppati M. Standing on a
continuously moving platform: is body inertia counteracted or
exploited? Exp Brain Res 1999;124:331–41.
Di Nardo W, Ghirlanda G, Cercone S, Pitocco D, Soponara C, Cosenza A,
Paludetti G, Di Leo MAS, Galli I. The use of dynamic posturography to
detect neurosensorial disorder in IDDM without clinical neuropathy.
J Diabetes Complications 1999;13:79–85.
Diener HC, Dichgans J, Bacher M, Gompf B. Quantification of postural
sway in normals and patients with cerebellar disease. Electroenceph
clin Neurophysiol 1984a;57:134 –42.
Diener HC, Dichgans J, Guschlbauer B, Mao H. The significance of
proprioception on postural stabilization as assessed by ischemia. Brain
Res 1984b;296:103 –9.
Dietz V, Mauritz K-H, Dichgans J. Body oscillations in balancing due to
segmental stretch reflex activity. Exp Brain Res 1980;40:89–95.
Dyck PJ, Sherman WR, Hallcher LM, Service FJ, O’Brien PC, Grina LA,
Palumbo PJ, Swanson CJ. Human diabetic endoneurial sorbitol,
fructose, and myo-inositol related to sural nerve morphometry. Ann
Neurol 1980;8:590–6.
Dyck PJ, Chance P, Lebo R, Carney JA. Hereditary motor and sensory
neuropathies. In: Dyck PJ, Thomas PK, Griffin JW, et al., editors.
 A. Nardone, M. Schieppati / Clinical Neurophysiology 115 (2004) 779–789
Peripheral neuropathy. Philadelphia, PA: Saunders; 1993a. p.
1094–136.
Dyck PJ, O’Brien PC, Kosanke JL, Gillen DA, Karnes JL. A 4, 2 and 1
stepping algorithm for quick and accurate estimation of cutaneous
sensation threshold. Neurology 1993b;43:1508–12.
Fitzpatrick R, McCloskey DI. Proprioceptive, visual and vestibular
thresholds for the perception of sway during standing in humans.
J Physiol (Lond) 1994;478:173 –86.
Griffin JW, Cornblath DR, Alexander E, Campbell J, Low PA, Bird S,
Feldman EL. Ataxic sensory neuropathy and dorsal root ganglionitis
associated with Sjögren’s syndrome. Ann Neurol 1990;27:304–15.
Gurfinkel EV. Physical foundations of stabilography. Agressol 1973;14:
9–13.
Inglis JT, Horak FB, Shupert CL, Jones-Rycewicz C. The importance of
somatosensory information in triggering and scaling automatic postural
responses in humans. Exp Brain Res 1994;101:159 –64.
Katz R, Meunier S, Pierrot-Deseilligny E. Changes in presynaptic
inhibition of Ia fibres in man while standing. Brain 1988;111:
417–37.
Kimura J, Bosch P, Lindsay GM. F-wave conduction velocity in the central
segment of the peroneal and tibial nerves. Arch Phys Med Rehabil
1975;56:492 –7.
Knazan M, Bohlega S, Berry K, Eisen A. Acute sensory neuronopathy with
preserved SEPs and long-latency reflexes. Muscle Nerve 1990;13:
381–4.
Koceja DM, Trimble MH, Earles DR. Inhibition of the soleus H reflex in
standing man. Brain Res 1993;629:155 –8.
Kuwabara S, Asahina M, Nakajima M, Mori M, Fukutake T, Hattori T,
Yuki N. Special sensory ataxia in Miller Fisher syndrome detected by
postural body sway analysis. Ann Neurol 1999;45:533–6.
Le Quesne PM, Fowler CJ, Parkhouse N. Peripheral neuropathy profile in
various groups of diabetics. J Neurol Neurosurg Psychiatry 1990;53:
558–63.
Magnusson M, Enbom H, Johansson R, Pyykkö I. Significance of pressor
input from the human feet in anterior-posterior postural control. Acta
Otolaryngol (Stockh) 1990;110:182–8.
Mauritz K-H, Dietz V. Characteristics of postural instability induced by
ischemic blocking of leg afferents. Exp Brain Res 1980;38:117–9.
Mauritz K-H, Dichgans J, Hufschmidt A. Quantification of postural sway in
late cortical cerebellar atrophy of the anterior lobe and other forms of
cerebellar ataxia. Brain 1979;102:461– 82.
Morasso P, Schieppati M. Can muscle stiffness alone stabilise upright
standing? J Neurophysiol 1999;83:1622 –6.
Nardone A, Schieppati M. Medium-latency response to muscle stretch in
human lower limb: estimation of conduction velocity of group II fibres
and central delay. Neurosci Lett 1998;249:29–32.
Nardone A, Di Francesco G, Schieppati M. Reflex excitability of the soleus
muscle in standing man, at rest or prior to a voluntary triceps
contraction. Neurosci Lett 1986;26:S165.
Nardone A, Giordano A, Corrà T, Schieppati M. Responses of leg muscles
in humans displaced while standing. Effects of types of perturbation and
of postural set. Brain 1990;113:65 –84.
Nardone A, Grasso M, Giordano A, Schieppati M. Different effect of height
on latency of leg and foot short- and medium-latency EMG responses to
perturbation of stance in humans. Neurosci Lett 1996;206:89–92.
789
Nardone A, Tarantola J, Giordano A, Schieppati M. Fatigue effects on body
balance. Electroenceph clin Neurophysiol 1997;105:309–20.
Nardone A, Miscio G, Pisano F, Schieppati M. Sensory ataxia in
polyneuropathy is connected to dysfunction of group II afferent fibers.
J Periph Nerv Syst 1998;3:151.
Nardone A, Tarantola J, Miscio G, Pisano F, Schenone A, Schieppati M.
Loss of large-diameter spindle afferent fibres is not detrimental to the
control of body sway during upright stance: evidence from neuropathy.
Exp Brain Res 2000;135:155 –62.
Nardone A, Corna S, Schieppati M. Group II afferent fibres in balance
control: evidence from neurological disease. In: Gantchev N, editor.
From basic motor control to functional recovery II. Sofia: Academic
Publishing House; 2001a. p. 331 –8.
Nardone A, Galante M, Lucas B, Schieppati M. Stance control is not
affected by paresis and reflex hyperexcitability: the case of spastic
patients. J Neurol Neurosurg Psychiatry 2001b;70:635–43.
Panayiotopoulos CP, Chroni E. F-waves in clinical neurophysiology: a
review, methodological issues and overall value in peripheral
neuropathies. Electroenceph clin Neurophysiol 1996;101:365 –74.
Perkins BA, Bril V. Diabetic neuropathy: a review emphasizing diagnostic
methods. Clin Neurophysiol 2003;114:1167–75.
Peterka RJ, Benolken MS. Role of somatosensory and vestibular cues in
attenuating visually induced human postural sway. Exp Brain Res 1995;
105:101–10.
Quintern J, Immisch I, Albrecht H, Pollmann W, Glasauer S, Straube A.
Influence of visual and proprioceptive afferences on upper limb ataxia
in patients with multiple sclerosis. J Neurol Sci 1999;163:61 –9.
Schieppati M, Nardone A. Medium-latency stretch reflexes of foot and leg
muscles analysed by cooling the lower limb in standing humans.
J Physiol (Lond) 1997;503:691–8.
Schieppati M, Nardone A. Group II spindle afferent fibers in humans: their
possible role in the reflex control of stance. In: Binder MD, editor.
Progress in brain research, vol. 123. Amsterdam: Elsevier; 1999. p.
461– 72.
Schieppati M, Hugon M, Grasso M, Nardone A, Galante M. The limits of
equilibrium in young and elderly normal subjects and in parkinsonians.
Electroenceph clin Neurophysiol 1994;93:286–98.
Schieppati M, Nardone A, Siliotto R, Grasso M. Early and late stretch
responses of human foot muscles induced by perturbation of stance.
Exp Brain Res 1995;105:411 –22.
Schieppati M, Tacchini E, Nardone A, Tarantola J, Corna S. Subjective
perception of body sway. J Neurol Neurosurg Psychiatry 1999;66:
313– 22.
Schieppati M, Giordano A, Nardone A. Variability in a dynamic postural
task attests ample flexibility in balance control mechanisms. Exp Brain
Res 2002;144:200– 10.
Simoneau GG, Ulbrecht JS, Derr JA, Becker MB, Cavanagh PR. Postural
instability in patients with diabetic sensory neuropathy. Diabetes Care
1994;17:1411–21.
Uccioli L, Giacomini PG, Pasqualetti P, Di Girolamo S, Ferrigno P,
Monticone G, Bruno E, Boccasena P, Magrini A, Parisi L, Menzinger
G, Rossini PM. Contribution of central neuropathy to postural
instability in IDDM patients with peripheral neuropathy. Diabetes
Care 1997;20:929–34.
Weiss JA, White JC. Correlation of 1A afferent conduction with the ataxia
of Fisher syndrome. Muscle Nerve 1986;9:327–32.