re
n o
ris
gy, U
all, O
rm 3
cle activity ismonitored using an electromyogram (EMG) Rhudy et al., 2005).
Despite the popularity of theNFRparadigm in clinical
and experimental pain research, there is no consensus on
howbest to defineNFR threshold. Following a procedure
first proposed by Willer (1977), recent reviewers have
ain. Published by Elsevier B.V. All rights reserved.
* Corresponding author. Tel.: +1 740 593 1079; fax: +1 740 593
0579.
E-mail address: France@ohio.edu (C.R. France).
Pain 128 (2007) 20304-3959/$32.00  2006 International Association for the Study of P1. Introduction
The nociceptive flexion reflex (NFR) paradigm iswide-
ly used in pain research to investigate pharmacological
modulation of nociception, spinal and supraspinal influ-
ences on nociception, and individual differences in noci-
ceptive processing in participants with and without pain
disorders (Skljarevski and Ramadan, 2002; Sandrini
et al., 2005). TheNFR is a polysynaptic spinal withdrawal
reflex that is elicited following activation of nociceptive
A-delta afferents. To assess theNFR, biceps femoris mus-
during the application of varying intensities of electrocu-
taneous stimulation to the ipsilateral sural nerve. Based
on the observed EMG response, intensity of stimulation
required to elicit the NFR is used as an objective index
of nociceptive threshold (Skljarevski and Ramadan,
2002; Sandrini et al., 2005). Prior research has demon-
strated thatNFR threshold is often highly correlatedwith
subjective pain threshold (Willer, 1977;Willer et al., 1979;
Chan andDallaire, 1989; Guieu et al., 1992) and themag-
nitude of the reflex response is related to intensity of per-
ceived pain (Willer et al., 1979; Chan and Dallaire, 1989;Abstract
Despite the widespread use of the nociceptive flexion reflex (NFR) paradigm in clinical and experimental pain research, there is
currently no consensus on how best to define NFR threshold. Accordingly, the present studies were designed to assess the accuracy
and reliability of different NFR threshold scoring criteria. Study 1 compared 13 scoring criteria in their accuracy for identifying the
presence of the NFR, then generated empirically derived cut-points for the best criteria, and examined the test–retest reliability of
NFR thresholds derived from these cut-points. Study 2 evaluated the replicability of these findings in an independent sample.
Results from the two studies suggested that standardized peak (NFR Interval Peak z score) and mean (NFR Interval z score) biceps
femoris electromyogram (EMG) activity were accurate and reliable criteria for defining NFR threshold. Acknowledging that cut-
points may need to be adjusted for different research designs, graphs depicting sensitivity and specificity across a range of cut-points
have been provided to facilitate researcher’s decision-making. It is hoped that the results of these studies will promote a standard
NFR threshold assessment methodology, and further encourage the application of the NFR paradigm in the investigation of mech-
anisms and characteristics of both painful and non-painful diseases.
 2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
Keywords: Nociceptive flexion reflex; RIII; Threshold; Measurement; Receiver operating characteristicsDefining the nociceptive flexion
participants: A compariso
Jamie L. Rhudy a, Ch
a Department of Psycholo
b Department of Psychology, 245 Porter H
Received 31 May 2006; received in revised fodoi:10.1016/j.pain.2006.09.024flex (NFR) threshold in human
f different scoring criteria
topher R. France b,*
niversity of Tulsa, USA
hio University, Athens, OH 45701, USA
1 August 2006; accepted 14 September 2006
www.elsevier.com/locate/pain
44–253

e / Panoted the NFR threshold is defined as the intensity of
stimulation required to elicit a stable response at a rate
of 60–90% in a series of approximately 20 stimuli (Sklja-
revski and Ramadan, 2002; Sandrini et al., 2005).
Although this definition provides some direction, it does
not provide the detail required to ensure accurate replica-
tion. As a result, various researchers have developed
unique definitions that vary in terms of minimum EMG
criteria, required duration of EMG activation, and
post-stimulation response window (e.g., Boureau et al.,
1991; Andersen et al., 2000; Desmeules et al., 2003; Banic
et al., 2004; Terkelsen et al., 2004). Other researchers (e.g.,
Edwards et al., 2001; France et al., 2002a; France et al.,
2002b; France et al., 2005; Rhudy et al., 2005) have used
standardized criteria that stipulate specific within-subject
EMG change in order to address the array of factors that
can influence absolute EMG activity such as electrode
design, quality of electrode application and surface prep-
aration, and participant muscle mass and subcutaneous
adipose tissue. Whereas most investigators have been
careful to outline their uniqueNFR threshold definitions,
it is often not possible to directly compareNFR threshold
findings across studies due to varying definitions.
Because the lack of standardized scoring procedures
may limit the adoption of NFR assessment methodolo-
gy in clinical and research settings, we compared the
accuracy and reliability of different scoring criteria in
an effort to develop a potential NFR threshold stan-
dard. Specifically, existing nociceptive flexion reflex data
gathered from healthy young adults were analyzed using
diverse criteria to define reflex threshold (Study 1). The
best performing indices were then re-examined in a sec-
ond set of data to establish consistency of performance
across samples (Study 2).
Study 1: Comparing the accuracy and reliability of
scoring criteria for the NFR
2. Methods
2.1. Participants
Participants included 40 healthy young adults (20
men and 20 women) who had their nociceptive flexion
reflex thresholds measured repeatedly as part of an
investigation of consistency and individual differences
in nociceptive responding (French et al., 2005). All par-
ticipants were undergraduate students and all received
course credit in return for their participation.
2.2. Laboratory testing procedure
Participants were scheduled for two laboratory ses-
sions, approximately 24 h apart (±1 h). They were
instructed to refrain from caffeine, nicotine, alcohol,
J.L. Rhudy, C.R. Francand strenuous exercise for at least four hours and fromanalgesic medication for 24 h prior to testing. At each
testing session, participants completed two assessments
of their nociceptive flexion reflex (NFR) threshold.
To begin the session, the electrocutaneous stimula-
tion and electromyogram (EMG) recording sites were
prepared and the required electrodes were attached.
All electrode sites were cleaned and abraded with Omni
Prep electrode paste and an impedance of less than
10 KX, verified using a UFI Checktrode (model MKII),
was achieved before proceeding. To record NFR activi-
ty, a differential electromyographic (EMG) electrode
was secured over the biceps femoris muscle of the left
leg, 10 cm superior to the popliteal fossa, and a reference
(common ground) electrode attached over the lateral
epicondyle of the femur. EMG activity was amplified
using a DelSys, Bagnoli-2 differential amplifier and the
signal was recorded and processed using a CED
Micro1401 analog-to-digital converter and Spike2 soft-
ware. To elicit the NFR, electrocutaneous stimulation
generated by a Digitimer, DS7A constant-current stim-
ulator was applied over the retromalleolar pathway of
the sural nerve of the left leg using a Nicolet bar elec-
trode (a bipolar stimulating electrode with a fixed
30 mm interelectrode distance) that was applied with
anode inferior.
Participants were then seated in a Hi-Seat rehabilita-
tion chair (model 2000) with the left leg rest adjusted to
maintain knee flexion at 60 ± 5 from horizontal. Once
the leg rest was adjusted, participants received a series of
four electrocutaneous stimulations of increasing intensi-
ty (0, 2, 4, and 6 mA) to acclimatize them to the stimu-
lation. NFR threshold was then determined twice, with
a 10-min rest period between each assessment. As in pre-
vious studies conducted in our laboratory (Page and
France, 1997; France and Suchowiecki, 2001; France
et al., 2002a; France et al., 2002b), NFR occurrence
was defined as a mean EMG response in the 90 to
150 ms post-stimulation interval that exceeded mean
EMG activity during the 60 ms pre-stimulation baseline
(65 to 5 ms) interval by at least 1.5 standard devia-
tions (i.e., NFR Interval z score = 1.5, see Section
2.4.2. below). Although a range of intervals (from 80
to 180 ms post-stimulation) have been used in prior
investigations to identify the RIII reflex, a more narrow
90–150 ms window avoids possible contamination of the
RIII reflex by the low-threshold cutaneous flexor reflex
(RII) which can precede 90 ms and by startle reactions
and voluntary movements that can begin as early as
150 ms post-stimulation (Dowman, 1991, 1992).
To elicit NFR activity, repeated electrocutaneous
stimulation was applied over the sural nerve according
to a variable interval schedule of 6 s (range 4–8 s) to
decrease the likelihood of stimulus predictability. Each
stimulation trial consisted of a volley of five 1 ms rectan-
gular pulses with a 3 ms interpulse interval (total dura-
in 128 (2007) 244–253 245tion = 17 ms; 250 Hz). Using an up-down staircase