and variability
Werner J. Z’Graggen a,b, Rebekka
aDepartment of Neurosurgery, Inselspital, Bern Universi
bDepartment of Neurology, Inselspital, Bern University H
c Sobell Department of Motor Neuroscience and Moveme
a r t i c l e i n f o
Article history:
Accepted 13 April 2011
Available online 8 May 2011
Keywords:
Muscle velocity recovery cycle
Relative refractory period
Early supernormality
Late supernormality
Measuring multi-fiber velocity recovery cycles (VRC) of muscle
action potentials is a method to study muscle membrane proper-
ties (Z’Graggen and Bostock, 2009). This technique is adapted from
measuring recovery cycles of nerves, but uses direct muscle stim-
ulation and recording to measure the velocity of muscle fiber
action potentials (Troni et al., 1983; Mihelin et al., 1991). For
cycles are exquisitely sensitive to changes in membrane potential.
Two recent studies using multi-fiber VRC recordings in healthy
subjects and in patients with chronic renal failure have indicated
that muscle VRCs also are very sensitive to changes in membrane
potential (Z’Graggen and Bostock, 2009; Z’Graggen et al., 2010).
VRC recordings are based on the principle that the action potential
is followed by a depolarizing afterpotential. This represents the
charge left on the capacitance of the muscle fiber membrane, as
the inward and outward charge shifts during the action potential
are not equal. The depolarizing charge left on the capacitance of
the muscle membrane decays over about a second. Unique for
⇑ Corresponding author at: Sobell Department, Institute of Neurology, Queen
Square, London WC1N 3BG, UK. Tel.: +44 20 7837 3611; fax: +44 20 7813 3107.
Clinical Neurophysiology 122 (2011) 2294–2299
Contents lists availab
ro
.e lE-mail address: H.Bostock@ion.ucl.ac.uk (H. Bostock).or apparent velocity. There were also no significant differences between the recordings from men and
women, or between the recordings from older (mean 44.9 y) and younger (26.5 y) subjects.
Conclusions: VRC measures are sufficiently consistent to be suitable for comparing muscle membrane
function both within subjects and between groups. Early supernormality measurements benefit most
from within subject comparisons.
Significance: These normative data sets provide a firm basis for planning clinical studies.
 2011 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights
reserved.
1. Introduction nerves, the C-fiber VRCs, like the A-fiber excitability recoveryMuscle membrane potential1388-2457/$36.00  2011 International Federation o
doi:10.1016/j.clinph.2011.04.010Troller b, Karin A. Ackermann b, Andrea M. Hummb, Hugh Bostock b,c,⇑
ty Hospital and University of Bern, Switzerland
ospital and University of Bern, Switzerland
nt Disorders, Institute of Neurology, University College London, London, UK
h i g h l i g h t s
 Measurement of velocity recovery cycles (VRCs) may be useful for the detection of abnormal muscle
membrane properties in myopathies.
 Repeatability and variability of VRC measures were tested in normal subjects.
 The reliability of early muscle supernormality was found to be excellent, although refractory period
may be more sensitive to membrane potential.
a b s t r a c t
Objective: Velocity recovery cycles (VRCs) of human muscle action potentials have been proposed as a
new technique for assessing muscle membrane function in myopathies. This study was undertaken to
determine the variability and repeatability of VRC measures such as supernormality, to help guide future
clinical use of the method.
Methods: To assess repeatability, VRCs with one and two conditioning stimuli were recorded from bra-
chioradialis muscle by direct muscle stimulation in 20 normal volunteers, and the measurements
repeated 1 week later. To further assess variability and dependence on electrode separation, age and
sex, recordings from an additional 20 normal volunteers were added.
Results: There was a high intraclass correlation between repeated recordings of early supernormality,
indicating excellent reliability of this VRC measure. However, relative refractory period had a smaller
coefficient of repeatability in relation to the changes previously described during ischemia. We found
no evidence that any of the excitability measures depended on electrode separation, conduction timeVelocity recovery cycles of human muscle action potentials: RepeatabilityClinical Neu
journal homepage: wwwf Clinical Neurophysiology. Publishle at ScienceDirect
physiology
sevier .com/locate /c l inphed by Elsevier Ireland Ltd. All rights reserved.

ments have provided evidence that muscle membrane depolariza-
surface electrode was used as anode (Red Dot, 3 M Health Care,
rophD-46325 Borken, Germany) and was placed about 1 cm lateral to
the cathode. Rectangular current pulses of 0.05 ms duration, gener-
ated by a computer and converted to current with an isolated lin-
ear bipolar constant-current stimulator (DS5, Digitimer Ltd.,
Welwyn Garden City, Hertfordshire, UK) were used for stimulation.
2.2.2. Recording
Recordings were made with concentric 30 G EMG electrodes
(Medtronic, Skovlunde, Denmark). This electrode was inserted into
the brachioradialis muscle about 15–20 mm further proximal to
the stimulating electrodes. The needle was repositioned until a sta-
ble monophasic electric response could be recorded with stimulus
intensities of less than 10 mA. The signal was amplified (gain 1000,
bandwidth 1.6–2 kHz) and digitized (National Instruments NI
DAQCARD-6062E, National Instruments Europe Corp., Debrecen,
Hungary) with a sampling rate of 20 kHz. Stimulation and record-2.2. Multi-fiber velocity recovery cycles
2.2.1. Stimulation
Subjects were comfortably seated in an armchair. The left fore-
arm was positioned on cushions, with the angle of the elbow set
close to 90. Skin temperature was monitored with a temperature
probe and kept in the range of 31.5–33 C. Studies were performed
using a recently described protocol (Z’Graggen and Bostock, 2009).
In brief, a monopolar insulated needle electrode served as cathode
(TECA, VIASYS Healthcare, Madison, Wisconsin, USA). This needle
was inserted perpendicularly into the brachioradialis muscle at
about 25% of the distance from the lateral epicondyle of the
humerus to the styloid process of the radius. The non-polarizabletion is caused by hyperkalemia (Z’Graggen et al., 2010).
Here, we present an assessment of the repeatability of multi-
fiber VRC measurements in successive examinations. Such an
assessment is needed if muscle VRCs are to be used as a tool for fol-
low-up investigations of muscle membrane properties.
2. Methods
2.1. Subjects and ethical approval
A total of 40 healthy subjects (23 women and 17 men) partici-
pated in this study, aged between 22 and 74 y (mean 36.1 y). A
subset of 20 of these subjects (15 women and 5 men), within a nar-
rower age range (24–39 y, mean 30.9 y), were recorded on 2 occa-
sions, approximately 1 week apart, to assess repeatability. Ethical
approval was obtained from the local ethics committee (Kantonale
Ethikkommission Bern, Switzerland) and conformed to the Decla-
ration of Helsinki. Subjects provided written informed consent to
participate in the study.muscle membranes is that there is a second phase of enhanced
afterpotential around 100 ms. This so called late depolarizing
afterpotential is attributed to potassium accumulation in the
t-tubule system (Z’Graggen and Bostock, 2009).
We recently used multi-fiber VRCs to study the effect of ische-
mia on muscle membrane potential in healthy subjects. We could
demonstrate a progressive increase in relative refractory period
(RRP) and reduction of early supernormality as indicators of the
ongoing membrane depolarization (Z’Graggen and Bostock,
2009). In patients with chronic renal failure muscle VRC measure-
W.J. Z’Graggen et al. / Clinical Neuing were controlled by QTRAC software (written by H. Bostock,
copyright Institute of Neurology, London, UK), using the menu-
driven recording protocol 1200RCM.QRP.Multi-fiber VRCs with test stimuli alone, single conditioning
stimuli, and paired conditioning stimuli 10 ms apart, were
recorded from 20 normal subjects on two occasions 7–8 days apart.
The test stimuli were delivered every 2 s and the interval between
single or paired conditioning stimuli and the test stimulus was var-
ied in 34 steps between 1 s and 2 ms in an approximately geomet-
ric series (Z’Graggen and Bostock, 2009). Since muscle fiber
conduction velocity is sensitive to temperature (Blijham et al.,
2008), special attention was given to skin temperature, which
was not allowed to deviate by more than 0.5 C between the two
measurements. Temperature was monitored using a surface tem-
perature probe, which was placed adjacent to the EMG needle. If
temperature differed by more than 0.5 C at the time of the second
examination, the forearm was covered by warm or cold blankets
depending on the deviation of skin temperature compared to the
first examination. Both recordings in one subject were always
carried out by the same investigator to rule out an additional in-
ter-investigator effect on repeatability.
To help understand the sources of variability in the recordings,
the new data (first recording only) was combined with recordings
from 20 additional healthy subjects made with the same tech-
nique, some of which were included in previously published stud-
ies (Z’Graggen and Bostock, 2009; Z’Graggen et al., 2010).
2.3. Data analysis
The recovery cycle data was acquired and analyzed with QTRAC,
using the features previously described in detail (Z’Graggen and
Bostock, 2009). Stimuli were presented sequentially on QTRAC
channels 1–3, with the test stimulus at 1100 ms preceded by 0–2
conditioning stimuli, respectively. Latencies of elicited responses
were measured from the start of the test stimulus to the negative
peak of the muscle action potential. The effect of one and two pre-
ceding conditioning stimuli on the latency of the test response
were calculated as percentage differences compared to the imme-
diately preceding trial with test stimulus alone. Latency changes as
a function of interstimulus interval were converted to slowing as a
function of interspike interval (ISI) to take in account the changes
in ISI with conduction distance. This conversion has no effect on
relative refractory period, or on early supernormality (Z’Graggen
and Bostock, 2009; Bostock et al., 2003).
The following measures were derived from VRC recordings with
one and two conditioning stimuli: From VRCs with one condition-
ing stimulus: (1) relative refractory period in ms (RRP), i.e. the
interpolated ISI at which velocity first reached its unconditioned
value, (2) early supernormality (ESN), measured as the peak per-
centage increase in velocity at ISIs shorter than 15 ms, and (3) late
supernormality (LSN), measured as the average percentage in-
crease in velocity at ISIs between 50 and 150 ms. From recordings
with two conditioning stimuli: (4) extra late supernormality due to
the second conditioning stimulus (XLSN), expressed as the mean
percentage increase in velocity at ISIs from 50 to 150 ms.
The QTRAC data analysis software was used for statistical anal-
ysis. Refractory periods were log transformed to normalize the
distributions before calculating correlation coefficients or applying
statistical tests. However, for simplicity the means and SDs in
Table 1 are not log transformed. As a measure of repeatability,
the British Standards Institution’s ‘coefficient of repeatability’
(Bland and Altman, 1986; Humm et al., 2004) was calculated as
twice the standard deviation of the differences between the two
recordings. Given one measurement, 95% of further measurements
are expected to fall within the range of the first ± this coefficient.
Hopkins (2000) has argued the merits of ‘typical percentage error’
ysiology 122 (2011) 2294–2299 2295(which equates to the standard deviation of the differences divided
by the mean  100/
p
2) for comparing the reliability of different
measures, and these values are also included in Table 1. Finally, a