Determinants of endogenous analgesia magnitude in a diffuse noxious inhibitory control (DNIC) paradigm: Do conditioning stimulus painfulness, gender and personality variables matter?
Descending modulation of pain can be demonstrated psychophysically by dual pain stimulation. This study evaluates in 31 healthy subjects the association between parameters of the conditioning stimulus, gender and personality, and the endogenous analgesia (EA) extent assessed by diffuse noxious inhibitory control (DNIC) paradigm. Contact heat pain was applied as the test stimulus to the non-dominant forearm, with stimulation temperature at a psychophysical intensity score of 60 on a 0-100 numerical pain scale. The conditioning stimulus was a 60s immersion of the dominant hand in cold (12, 15, 18 degrees C), hot (44 and 46.5 degrees C), or skin temperature (33 degrees C) water. The test stimulus was repeated on the non-dominant hand during the last 30s of the conditioning immersion. EA extent was calculated as the difference between pain scores of the two test stimuli. State and trait anxiety and pain catastrophizing scores were assessed prior to stimulation. EA was induced only for the pain-generating conditioning stimuli at 46.5 degrees C (p=0.011) and 12 degrees C (p=0.003). EA was independent of conditioning pain modality, or personality, but a significant gender effect was found, with greater EA response in males. Importantly, pain scores of the conditioning stimuli were not correlated with EA extent. The latter is based on both our study population, and on additional 82 patients, who participated in another study, in which EA was induced by immersion at 46.5 degrees C. DNIC testing, thus, seems to be relatively independent of the stimulation conditions, making it an easy to apply tool, suitable for wide range applications in pain psychophysics.
Determinants of endogenous analgesia magnitude in a diﬀuse noxious
inhibitory control (DNIC) paradigm: Do conditioning stimulus
painfulness, gender and personality variables matter?
, Irit Weissman-Fogel
, Yonathan Crispel
, Dorit Pud
, Elliot Sprecher
, David Yarnitsky
Faculty of Social Welfare and Health Studies, University of Haifa, Israel
Department of Neurology, Rambam Medical Center, and The Laboratory of Clinical Neurophysiology, Technion Faculty of Medicine, Haifa, Israel
Received 21 November 2006; received in revised form 21 June 2007; accepted 22 June 2007
Descending modulation of pain can be demonstrated psychophysically by dual pain stimulation. This study evaluates in 31
healthy subjects the association between parameters of the conditioning stimulus, gender and personality, and the endogenous anal-
gesia (EA) extent assessed by diﬀuse noxious inhibitory control (DNIC) paradigm. Contact heat pain was applied as the test stim-
ulus to the non-dominant forearm, with stimulation temperature at a psychophysical intensity score of 60 on a 0–100 numerical pain
scale. The conditioning stimulus was a 60 s immersion of the dominant hand in cold (12, 15, 18 C), hot (44 and 46.5 C), or skin
temperature (33 C) water. The test stimulus was repeated on the non-dominant hand during the last 30 s of the conditioning immer-
sion. EA extent was calculated as the diﬀerence between pain scores of the two test stimuli. State and trait anxiety and pain catas-
trophizing scores were assessed prior to stimulation. EA was induced only for the pain-generating conditioning stimuli at 46.5 C
(p = 0.011) and 12 C(p = 0.003). EA was independent of conditioning pain modality, or personality, but a signiﬁcant gender eﬀect
was found, with greater EA response in males. Importantly, pain scores of the conditioning stimuli were not correlated with EA
extent. The latter is based on both our study population, and on additional 82 patients, who participated in another study, in which
EA was induced by immersion at 46.5 C. DNIC testing, thus, seems to be relatively independent of the stimulation conditions,
making it an easy to apply tool, suitable for wide range applications in pain psychophysics.
2007 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
Keywords: Endogenous analgesia; DNIC; Pain modulation; QST; Gender; Personality
Pain messages ascending in the spinal cord generate,
upon arrival at the brainstem, endogenous descending
inhibitory and facilitatory messages, which reach the
dorsal horn and modulate the perception of noxious
stimuli. One aspect of the inhibitory mechanism that
modulates pain processing at the spinal cord level is
termed ‘diﬀuse noxious inhibitory control’ (DNIC), in
which the activity of pain-signaling neurons in the spinal
dorsal horn and in trigeminal nuclei is attenuated in
response to noxious stimuli applied to a remote area
of the body (Le Bars et al., 1979; Schouenborg and
Dickenson, 1985; Le Bars and Willer, 2002). The term
heterotopic noxious conditioning stimulation (HNCS)
is often used to describe eﬀect of endogenous analgesia
(EA) in humans that can be assessed by DNIC para-
digm. In recent years, DNIC has been identiﬁed as an
advanced psychophysical measure, with high clinical rel-
evancy, in the characterization of one’s capability to
0304-3959/$34.00 2007 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
Corresponding author. Tel.: +972 48542605; fax: +972 48542944.
E-mail address: firstname.lastname@example.org (D. Yarnitsky).
Pain 136 (2008) 142–149
modulate pain and consequently one’s susceptibility to
The eﬀect of EA can be measured experimentally
using various noxious stimulation modalities and a
range of pain testing paradigms, including pain thresh-
olds, tolerance, supra-threshold painful stimulation, or
temporal summation (Grill an d Coghill, 2002; Lautenb-
acher et al., 2002; Marchand and Arsenault, 2002;
Bouhassira et al., 2003; Edwards et al., 2003a; Staud
et al., 2003; Tuveson et al., 2006). Several variables seem
to aﬀect the extent of the EA response, including site,
surface area, duration, and intensity of conditioning
and test stimuli, as well as gender and age (Le Bars,
2002; Edwards et al., 2003a; Staud et al., 2003 ).
It is logical to assume that stronger conditioning stim-
uli will induce greater EA. Nevertheless, the literature is
not clear on this issue; some previous studies showed posi-
tive relationships between the intensity of the condition-
ing stimulus and the magnitude of EA response (Le
Bars et al., 1995; Villanueva and Le Bars, 1995; Fujii
et al., 2006). However, Pud et al. (2005) and Baad -Hansen
et al. (2005) found no correlation between the pain scores
of the conditioning stimulus and EA extent. Furthermore,
even the question of whether the conditioning stimulus
needs to be painful in order to induce EA is still open .
Le Bars (2002) found that only painful stimuli induced
DNIC, whereas Lautenbacher and Rollman (1997) and
Lautenbacher et al. (2002) found that non-painful stimuli
also induced EA.
The possibilities that EA eﬀect expressed by DNIC
extent (i) can be independent of the reported intensity
of the conditioning stimulus and (ii) can be evoked by
non-painful conditioning stimuli raise the hypothesis
that the modulation processes expressed by the DNIC
response are subconscious. As such, ascending activity
in the spinal pain tracts is suﬃcient, upon arrival in
the brainstem, to activate the descending modulation
response, regardless of whether the ﬁna l cortical experi-
ence induced by that barrage is painful or not.
The present study explored several variables related to
the EA response in healthy subjects. In addition to the
eﬀects of stimulation modality, gender and psychological
status, we focused on the eﬀect of painfulness of the con-
ditioning stimulus – whether perceived as painful or not,
and its reported magnitude – on the extent of EA.
The major study population was composed of 31 paid
healthy volunteers (21 men and 10 women), aged 24.6 ± 4.6
(mean ± SD), with a range of 20–39 years. The volunteers were
recruited by advertisement, mainly from the student body of
the University of Haifa in northern Israel. All tests were con-
ducted during the morning hours under the same environmen-
tal conditions (e.g., room temperature, acoustics) by the same
investigator at the Neurophysiology Laboratory of the Tech-
nion in Haifa, Israel (YC). The study was approved by the
local Ethics Committee, and informed consent was obtained
from all participants before the beginning of the experiment.
Participants were enrolled in the study after meeting the fol-
lowing criteria: (1) absence of chronic pain history; (2) no med-
ication use on a regular basis (except for oral contraceptives);
and (3) ability to communicate and understand the purpose
and instructions of the study.
For test stimulation, we used contact heat pain produced by
a Thermal Sensory Analyzer (TSA) 2001 system (Medoc,
Ramat-Yishai, Israel) with a 30 · 30 mm Peltier surface stimu-
lator applied to the volar part of the non-dominant forearm.
Baseline temperature was 32.0 C, with an increasing tempera-
ture rate of 1.0 C/s and a back to baseline temperature rate of
8.0 C/s. For conditioning stimulation, we used a water bath
apparatus (Heto CBN 8-30 Lab equipment, Allerod, Den-
mark), which is a temperature-controlled water bath with a
maximum temperature variance of ±0.5 C and continuous
stirring action to ensure the maintenance of an even tempera-
ture throughout the bath. Temperatures were chosen accord-
ing to these criteria: (i) painful and non-painful; (ii)
representation of two modalities – heat and cold; and (iii) suit-
ability for induction of DNIC in clinical practice. After some
preliminary experimentation among middle-aged subjects
who were not familiar with the procedure, we chose to apply
conditioning temperatures of 12, 15, 18, 44, and 46.5 C and
the natural stimulus of 33 C for control.
2.3. Study design
2.3.1. Training phase
Subjects were given a short training in order to familiarize
them with the devices, the perceived sensations, and the task.
First, they were exposed to three short heat stimuli (43, 45,
and 47 C), applied to their dominant forearm by the contact
thermode, each lasting for 7 s starting from the time the stim-
ulus intensity reached the destination temperature. Subjects
were asked to rate the level of pain intensity using a numerical
pain scale (NPS) ranging from 0 = ‘‘no pain’’ to 100 = ‘‘the
worst pain imaginable’’. Thereafter, subjects were asked to
immerse their non-dominant hand twice into the water bath
for 60 s each time, with a 10-min interval in between. One
immersion was to hot water (46.5 C) and the other was to cold
water (12 C), provided in random order. While their hand was
immersed in the bath, subjects were trained to rate their per-
ceived pain intensity seven times (at the start and also at 10,
20, 30, 40, 50, and 60 s), using the NPS.
2.3.2. Determination of test stimuli intensity (pain-60)
Following completion of the above-described training
phase, we determined the test stimulation temperature that
induced pain-60. The individually determined psychophysical
parameter of pain-60 is the temperature that induces pain expe-
rience at a magnitude of 60 on a 0–100 NPS (for more details,
see Granot et al., 2006). The ﬂow chart of determining pain-60
M. Granot et al. / Pain 136 (2008) 142–149 143
is depicted in Fig. 1. In general, subjects were exposed to a ser-
ies of hot stimuli of 7 s duration. The ﬁrst series consisted of
45, 46, and 47 C stimulations with a 1-min inter-stimulus
interval. After each stimulus, subjects were asked to verbally
report the level of pain. If one of these stimuli induced pain-
60, that temperature was chosen as the test stimuli for the rest
of the experiment; if not, additional steps for determination of
pain-60 were applied (see Fig. 1). In order to reconﬁrm the
pain-60 temperature, an additional stimulus at the same inten-
sity was given and scored at the end of the process. This spe-
ciﬁc temperature served as the test stimulus for the rest of
the study. It should be noted that despite the use of a ﬁxed
stimulus intensity in accordance with the determination of
pain-60, reported pain scores usually vary over the course of
the test sessions in relation to various factors, such as stimulus
duration and stimulation sequence.
2.3.3. Administration of the test stimuli
The test stimuli were applied for 30 s each at the intensity
previously determined for pain-60. The thermal probe was
strapped in place on the forearm during each individual stim-
ulus period. For each stimulus, subjects rated the level of pain
intensity three times: at 10, 20, and 30 s. The mean scores of
the three pain ratings were calculated. Each test stimulus was
given twice, both before the conditioning stimulation and dur-
ing exposure to the conditioning stimuli.
2.3.4. Conditioning stimuli
Five minutes after delivering the ﬁrst test stimuli, subjects
were asked to place their non-dominant hand in the water
bath, in a still position with the ﬁngers wide apart for 60 s.
The DNIC assessment was performed using six diﬀerent condi-
tioning stimulus temperatures (12, 15, 18, 33, 44, and 46.5 C)
in two sessions, conducted at least one week apart. Each ses-
sion consisted of three DNIC tests with randomized condition-
ing temperatures. In order to avoid bias of residual eﬀect
induced by one noxious stimulus to the next, special attention
was given to maintaining an interval of 20 min between tests.
Subjects were asked to rate the level of pain intensity three
times: immediately after immersion of the hand in water (time
0), after 10 s, and after 20 s following immersion. The mean
scores of the three pain ratings were calculated for each stim-
ulus. Subsequent to the third pain rating for the conditioning
stimulus, when the test stimulus was applied again, subjects
were asked to shift their focus to the contact heat pain and
to rate its intensity while their non-dominant hand was still
in the bath.
2.4. Personality questionnaires
Anxiety level was assessed by Spielberger’s State-Trait Anx-
iety Inventory (Spielberger et al., 1969), using the validated
Hebrew version (Teichman and Malineck, 1978). The ﬁrst part
Heat stimulations of 45˚C, 46˚C and 47˚C
7 sec each
at least in one of
at all temps.
of 48˚C and 49˚C
at all temps.
of 43˚C and 44˚C
at all temps.
at least in one of
at all temps.
at least in one of
Fig. 1. Flow chart showing the determination process of pain-60. Note: All subjects met the criteria for this stage and none of them was excluded.
144 M. Granot et al. / Pain 136 (2008) 142–149
of the questionnaire assesses the level of state anxiety, and the
second part assesses the level of trait anxiety. Each part
includes 20 statements describing emotional conditions, and
the subjects were asked to rate their feelings about each state-
ment on a four-point scale prior to starting the pain tests.
Pain catastrophizing level was assessed by the Pain Catas-
trophizing Scale (PCS) ( Sullivan et al., 1995). The instrument
includes 13 items representing the three components of pain
catastrophizing: rumination (e.g., ‘‘I can’t seem to keep it
out of my mind’’); magniﬁcation (e.g., ‘‘I wonder whether
something serious may happen’’); and helplessness (e.g.,
‘‘There is nothing I can do to reduce the intensity of pain’’).
This scale has been validated in Hebrew (Granot and Gold-
stein Ferber, 2005). Subjects completed the questionnaire prior
to starting the pain tests and were asked to do so in reference
to a previous pain event.
2.5. Data analysis
Statistical analyses were performed with SPSS 12 (SPSS
Inc., Chicago, IL, USA) and JMP (SAS Institute, Cary, NC,
USA), and power analyses were performed using G*Power 3
gpower3). EA extent measured by DNIC paradigm was deter-
mined by the diﬀerence between the pain scores of the ﬁrst test
stimuli and those of the test stimuli applied during the condi-
tioning stimulation. Bonferroni-corrected paired t-tests were
used to examine diﬀerences in the test stimuli ratings. The rela-
tionships between EA extent, the conditioning pain scores, and
the personality variables were assessed using Pearson correla-
tions. The comparisons between women and men for pain-
related personality variables were performed by t-test. The
eﬀects of gender, type of pain modality (hot or cold), age
and personality were evaluated using repeated measures
ANOVA. The data are presented as means ± standard devia-
tion. Statistical signiﬁcance was deﬁned as p 6 0.05.
The pain scores obtained for the six DNIC tests
induced by various conditioning temperatures are pre-
sented in Fig. 2. As shown, only the exposure to 46.5
and 12 C induced a signiﬁcant decrease in pain scores
(from 46.9 ± 14.9 to 33.7 ± 19.4, p = 0.011; and from
41.5 ± 17.0 to 25.5 ± 19.9, p = 0.003, respectively, as
calculated by Bonferroni-corrected paired t-tests). No
signiﬁcant reduction in pain perception was noted for
44 C (from 43.7 ± 15.8 to 37.6 ± 18.6, p = 0.176),
for 18 C (36.3 ± 15.3 vs. 31.7 ± 17.8, p = 0.249), for
15 C (34.1 ± 15.7 to 28.7 ± 17.7, p = 0.139), or for
the neutral water temperature of 33 C (44.8 ± 18.4 vs.
43.3 ± 16.5, p = 0.363). The ﬁndings that a signiﬁcant
EA was induced only by conditioning stimuli whose
mean scores were higher than 20 on a 0–100 NPS and
that pain scores under 20 can thus be considered as only
a mild pain experience lead us to suggest that a painful
conditioning experience is apparently a ‘sine qua non’
circumstance for the induction of EA. Therefore, further
analysis included only the two conditioning stimuli of
heat pain at 46.5 C and cold pain at 12 C. Data for
these tests, separat ed for males and females, are pre-
sented in Table 1.
A signiﬁcantly positive correlation was noted
between the DNIC extent induced by 46.5 Candby
12 C(p = 0.0248, r = .402), suggesting that individuals
who show greater EA with heat also show greater EA
with cold. Thus, it appears that the EA extent is related
to the individual characteristics of pain modulation
response, rather than to the conditioning stimulus
modality. No correlation was found between the tem-
peratures inducing pain-60 for the test stimulus and
the DNIC extent for the tests at 46.5 Corat12C.
3.1. Painfulness of the conditioning stimulus and EA
No correlation was found between the conditioning
stimuli pain scores and the magnitude of EA responses
within each temperature (r = 0.145, p = 0.426 for the
conditioning stimulus of 12 C and r = 0.187,
p = 0.313 for the conditioning stimulus of 46.5 C). In
addition, no eﬀect of age was found to possibly con-
found the relationship between pain scores for the con-
ditioning stimuli and the magnitude of EA.
In order to examine the possibility that gender diﬀer-
ences could be confounded by diﬀerences in perceived
painfulness of the conditioning stimuli, separate mixed
12 15 18 33 44 46.5
Pain scores of test
First test stimulus
Test stimulus during conditioning
12 15 18 33 44 46.5
Pain scores of conditioning
Fig. 2. Pain scores reported for the test stimulus, before and during the
exposure to the conditioning stimulus (upper graph). Pain scores in
response to the conditioning stimulation (lower graph).
M. Granot et al. / Pain 136 (2008) 142–149 145
model ANOV As were performed (for 12 C and for
46.5 C), employing gender, pain scores of the condi-
tioning stimulus, and the interaction between them, with
subject as a rando m factor and with DNIC as the depen-
dent variable. The interaction eﬀect of each analysis is a
direct test of the inﬂuence of gender on the relationships
between conditioning stimulus pain scores and DNIC.
In both cases, there was no signiﬁcant interaction
(p = 0.631 for 12 Candp = 0.931 for 46.5 C), indicat-
ing that the gender diﬀerence was not confounded by
diﬀerences in painfulness of the condition ing stimuli.
Since the results showing lack of correlation between
the magnitude of the conditioning stimuli and EA extent
were obtained from a relatively small sample, we tried to
strengthen the statistical power of these results by adding
data from another study performed in our laboratory in
which the EA was assessed using the same protocol, but
only for conditioning stimuli of 46.5 C. In this study,
pre-surgical psychophysical assessments were performed
in 82 pain-free subjects (47 men and 35 women), aged
61.7 ± 13.2. Similar to our ﬁndings from the 31 subjects
in the present study, pain scores in these subjects signif-
icantly decreased (p < 0.001) in response to the condi-
tioning stimulus from 61.0 ± 16.5 to 46.8 ± 19.6 (mean
DNIC eﬀect 14.2 ± 19.2). Mean pain scores of the condi-
tioning stimulus itself were 67.9 ± 23.2 (range 13.3–100
in NPS). Although higher pain scores of the conditioning
stimulus were reported by these subjects (as compared to
the group of 31 subjects), no correlation was found
between the conditioning stimuli pain scores and the
magnitude of EA responses (r = 0.081, p = 0.510).
Age was not associated with EA or with pain scores for
the conditioning pain. Accordingly, the results of the
combined data from both groups (n = 113) are in line
with those obtained from each group alone, showing
no correlation between the extent of EA and the pain
scores reported for the conditioning stimuli (r =
0.092, p = 0.332). A power analysis of the enlarged
group of 113 subjects, assuming at least a medium-eﬀect
size (Cohen, 1988), that is, r = 0.3 or better, indicates
that the statistical power of this analysis exceeds 0.85,
thereby supporting our assertion of a lack of correlation
between these two factors (see Fig. 3 ).
3.2. Pain-related personality variables
Greater EA responses induced by the cold pain
(12 C) stimulus were found to be associated with higher
levels of pain catastrophizing (r = .395, p = 0.029) and
state anxiet y (r = .357, p = 0.048), but not with trait
Descriptive data for psychophysical measures for males and females
DNIC test Males
Temperature of pain-60 46.9 C ± 0.23
Test stimulus at baseline Test stimulus during conditioning Conditioning
12 C water bath 46.6 ± 3.9 25.3 ± 4.4 24.1 ± 3.6
46 C water bath 46.8 ± 3.3 30.2 ± 4.1 47.2 ± 5.2
Temperature of pain-60 46.2 C ± 0.3
Test stimulus at baseline Test stimulus during conditioning Conditioning
12 C water bath 30.9 ± 4.9 25.8 ± 6.4 21.8 ± 5.2
46 C water bath 47.3 ± 3.4 41.3 ± 5.9 41.5 ± 6.3
r = 0.092
0 10 50 100
Conditioning pain scores
Fig. 3. Combined data from 113 pain-free subjects revealed no association between DNIC extent and pain scores of the conditioning stimulus.
146 M. Granot et al. / Pain 136 (2008) 142–149
anxiety. No co rrelations between DNIC induced by the
46.5 C stimulus or by any of the assessed personality
scores were noted.
In a separate analysis of gender diﬀerence s in relation
to personality variable scores, diﬀerences were noted
only for state anxiety, with higher scores for women
(36.8 ± 2.1 vs. 30.2 ± 1.5, p = 0.018). No diﬀerences
were found for trait anxiety or pa in catastrophizing
scores (34.2 ± 1.91 vs. 32.7 ± 13, and 17.3 ± 3.2 vs.
16.5 ± 2.2, respectively).
3.3. Overall eﬀect of modality, gender and personality
Inordertoassess theeﬀects of conditioningpain modal-
ity (hot or cold), gender, anxiety, and pain catas trophizing
on EA, we applied a mixed model ANOVA, employing
gender, pain modality, the interaction of gender and
modality, state anxiety score, and PCS score as indepen-
dent variables, with subject as a random factor and with
DNIC as the dependent variable. Results of this analysis
demonstrated a signiﬁcant gender eﬀect, with a greater
EA response in males (F(1,27) = 4.573, p = 0.047). No
eﬀects of the conditioning modality (cold or hot pain) or
personality variables on EA were found. For both hot
and cold stimuli, males exhibited a greater EA response
as compared with females (16.7 ± 18.5 and 6.0 ± 8.5,
respectively, for 46.5 C; 21.3 ± 17.5 and 5.02 ± 13.1,
respectively, for 12 C). Becau se of the low levels of EA
in women, separate paired t -tests between test stimulus
pain at the ﬁrst application and during conditioning stim-
ulation werealso conducted onfemales alone for eachtem-
perature. These tests indicated that there was only a trend
for signiﬁcant DNIC (p = 0.053) for the temperature
46.5 C, while there was no signiﬁcant DNIC eﬀect
(p = 0.256) at all for the 12 C condition.
A review of the literature on EA reveals that diﬀerent
laboratories conduct diﬀerent approaches to evoke
DNIC. Similar to the case for other psychophysical mea-
sures, there are no accepted uniﬁed criteria or gold stan-
dards for performance and calculation of EA eﬀect.
Indeed, various types of conditioning and test stimuli
have been proposed in relation to area of targeting,
modality, stimulus duration, and time of pain measure-
ment. One of the unresolved issues is the characterization
of the conditioning stimulus in the DNIC paradigm –
whether it should be painful or not and how much its pain-
fulness determines the extent of this modulation process.
The present study shows that the eﬀectiveness of EA
is independent of the pain magnitude reported for the
conditioning stimulus, as long as the latter is perceived
as painful and regardless of the conditioning stimulus
modality. In addition, this study further emphasizes
the eﬀect of gender on EA, showing better eﬀectiveness
in males, again regardless of the conditioning stimulus
modality or its painfulness.
Several studies have suggested a relationship between
the conditioning stimulus intensity and DNIC magni-
tude (Le Bars et al., 1992; Villanueva and Le Bars,
1995; Le Bars and Willer, 2002). However, other studies
have demonstrated that DNI C is independent of the
level of subjective experience of the conditioning stimu-
lus. Lautenbacher and Rollman (1997) and Lautenbacher
et al. (2002) reported that DNIC can be induced by a
non-painful conditioning stimulus. Similarly, Bouhass-
ira et al. (1994, 1998) demonstrated inhibition of the
RIII reﬂex in response to non-painful visceral stimula-
tion. This dispute raises two possible mechanisms for
the EA eﬀect – it is generated either by a consciously
perceived painful experience or by activity in the ascend-
ing pain pathways that does not induce an overt pain
experience. The latter mechanism may explain Lautenb-
acher, Rollman and Bouhassira ﬁndings, based on ﬁring
in the spino-reticular tract reaching only as high as the
lower brainstem level.
Yet, Lautenbacher et al. (2002), who reported of EA
induction by non-painful conditioning stimulation,
actually used stimuli of 42 C, a temperature that is on
the verge of being painful. We therefore propose, in line
with our ﬁndings, that the perception of an apparently
painful experience is required in order to induce EA.
Bouhassira et al. (1994, 1998), who seem to be at odds
with our assertion claiming the induction of EA by
non-painful conditioning stimulation, explored visceral
sensations where diﬀerent degrees of spatial and tempo-
ral summation are required.
Our ﬁnding that the reported magnitude of the condi-
tioning pain is not associated with the EA extent suggests
that EA generation is a rapidly saturating phenomenon,
such that even a low amount of conditioning pain is suﬃ-
cient to induce EA. Thus, this modulating system does not
have a direct proportional relationship with the magni-
tude of percept ion of the conditioning pain, which high-
lights the eﬃciency of this pain inhibiting system. The
EA extent, therefore, seems to be a consequence of the
individual eﬃciency of pain modulation rather than the
parameters of the conditioning pain.
Applying this understanding to clinical psychophys-
ics strengthens the potential of dynamic psychophysical
testing. Most psychophysical measures depend heavily
on the speciﬁcations of the administration of the stimuli,
such that minor changes in the testing protocol are suf-
ﬁcient to generate major diﬀerences in the results. The
fact that in the process of generat ing EA, pain magni-
tude or modality of the conditioning stimulus does not
determine the outcome, positions this paradigm in the
ﬁrst line of practical psychophysical tests.
In contrast to previous DNIC studies, which applied
test stimuli of ﬁxed intensity or intensity anchored to indi-
vidual pain thresholds, we applied the pain-60 measure to
M. Granot et al. / Pain 136 (2008) 142–149 147
determine the temperature of stimulus intensity. We have
recently shown, in the context of measuring temporal
summation (Granot et al., 2006), the advantage of apply-
ing psychophysical stimuli that are tailored to individual
pain perception at the supra-threshold range and are not
derived from pain threshold. Tuveson et al. (2006) also
used a similar approach in a study of heterotopic ischemic
noxious conditioning stimulation by setting the test stim-
ulus intensity at a score of 7 on a 1–10 numerical pain
scale. Using the pain-60 parameter, we applied test stimuli
of various temperatures and therefore were able to inves-
tigate the relations between the physical intensity of the
test stimulus and EA response. We found that the speciﬁc
temperatures used to generate pain were not associated
with EA response. Although this study used only one psy-
chophysical level of test stimulation, it seems that DNIC
testing, as performed in this study, allows the individual
characteristics of pain modulation to be revealed in a
way that is minimally aﬀected by the characteristics of
the psychophysical paradigm.
The role of gender in the variability of pain perception
in experimental and clinical settings ha s been well
explored. Enhanced pain response in females has been
attributed to physiological and psychological variables,
including mechanisms of endogenous inhibition, capabil-
ity to endure pain, genetic factors, pain expectation and
personality traits (Keefe et al., 2000; Wiesenfeld-Hallin,
2005). We found that state anxiety was higher in females
than in male s. Accordingly, brain imaging demonstrated
greater activation in females in the contralateral insula,
prefrontal cortex, and thalamus – regions assumed to be
involved in the aﬀective aspects of pain proce ssing (Paul-
son et al., 1998; Mayer et al., 2004).
The amygdala is speciﬁcally activated in negative emo-
tional responses, such as stress, fear, and anxiety. It was
previously established that gender diﬀerences in the mag-
nitude and direction of response of the l-receptor system
in the amygdala’s ventro-basal nuc lei and anterior thala-
mus showed higher activation of the opioid receptors in
males (Zubieta et al., 2002). Since the amygdal a is directly
connected to the brainstem structures responsible for
descending modulation, and since mechanisms of pain
modulation are attributed to endogenous opioid activity
(Le Bars et al., 1992), it can be suggested that this reduced
opioid activity will lead to an attenuated EA response in
females. Although there is greater activation of the amyg-
dala in females following an anxiety state, it may be insuf-
ﬁcient to activate descending inhibition as a result of
deﬁciency in endogeno us opioid activity.
The results of the present study, despite the unbal-
anced gender sample size, strengthen the ﬁndings of pre-
vious psychophysical reports among healthy subjects
indicating that EA is less eﬀective in females (Staud
et al., 2003; Serrao et al., 2004). In these studies, gender
diﬀerences in favor of male s were noted for the EA eﬀect
when the test stimulus was temporally summed. Simi-
larly, women were found to be less prone to maintain
EA eﬀect over time than men (Ge et al., 2004). In con-
trast, other studies found no gender diﬀerences in EA
(France and Suchowiecki, 1999; Baad-Hansen et al.,
2005; Pud et al., 2005). In our study EA extent was
related to gender, but independently of modality or per-
ception of the conditioning pain scores, thus supporting
our hypothesis that EA represents the individual capa-
bility to modulate pain.
Edwards et al. (2003a,b) found no association
between the extent of EA tested by DNIC and psycho-
logical parameters, including proﬁle of mood states,
locus of control, level of vigilance, and stress. We found
an association between pain catastrophizing, trait anxi-
ety, and DNIC induced by immersion a t 12 C. This
association may point to possible cortical inﬂuences on
EA that may be aﬀected by stimulus modality. Never-
theless, the results of the mixed ANOVA model revealed
no signiﬁcant eﬀects of personality variables on EA
response. Seminowicz and Davis (2006) have shown,
using fMRI, that pain catastrophizing associ ated with
activity in cortical regions, including the dorsolateral
prefrontal, insula, and anterior cingulate, is associated
with aﬀective and attention aspects of pain. It has
recently been demonstrated that activation of the pre-
frontal cortex under pain-expectation condition s modu-
lates the strength of spinal nociceptive responses in
humans via EA mechanism (Goﬀaux et al., 2007). Thus,
brainstem-based reﬂex-like EA phenomenon is probably
aﬀected by cortical areas, including those inﬂuenced by
pain-related personality variables. Understating the role
of pe rsonality on DNIC requires more elaborate studies.
Augmented pain modulation, related to both sensitiza-
tion of ascending pathways and deﬁcienci es in pain
descending inhibitory systems, has been suggested as pos-
sible mechanism of several idiopathic chronic pain disor-
ders (Olesen, 1991; Sigurdsson and Maixner, 1994;
Lautenbacher and Rollman, 1997; Edwards et al., 2005;
Julien et al., 2005). Atten uated EA evaluated by DNIC
test was found in TMD (Maixner et al., 1995); ﬁbromyal-
gia (Kosek and Hansson, 1997; Lautenbacher and Roll-
man, 1997); tension-type headache (Pielsticker et al.,
2005); and migraine (Sandrini et al., 2006). Attenuated
EA was also found to be associated with greater pain his-
tory in healthy subjects (Edwards et al., 2003a). Measure-
ment of DNIC extent may therefore prove to be valuable
in representing the endogenous mod ulation system, and
as such in predicting the development of chronic pain con-
ditions (Edwards et al., 2003a; Edwards, 2005).
In conclusion, this study suggests that EA tested by
DNIC paradigm is activated in response to painful con-
ditioning stimuli, regardless of the magnitude of pain
induced by them. DNIC test is also relatively free of
individual variability in the magni tude of the test stimu-
lus. Thus, DNIC seems to be a relevant testing paradigm
for wide use in the clinical pain setting.
148 M. Granot et al. / Pain 136 (2008) 142–149
Baad-Hansen L, Poulsen HF, Jensen HM, Svensson P. Lack of sex
diﬀerences in modulation of experimental intraoral pain by diﬀuse
noxious inhibitory controls (DNIC). Pain 2005;116:359–65.
Bouhassira D, Chollet R, Coﬃn B, Lemann M, Le Bars D, Willer JC,
et al. Inhibition of a somatic nociceptive reﬂex by gastric disten-
tion in humans. Gastroenterology 1994;107:985–92.
Bouhassira D, Sabate JM, Coﬃn B, Le Bars D, Willer JC, Jian R.
Eﬀects of rectal distensions on nociceptive ﬂexion reﬂexes in
humans. Am J Physiol 1998;275:G410–7.
Bouhassira D, Danziger N, Attal N, Guirimand F. Comparison of the
pain suppressive eﬀects of clinical and experimental painful
conditioning stimuli. Brain 2003;126:1068–78.
Cohen J. Statistical power analysis for the behavioral sciences. 2nd
ed. Hillsdale, NJ: Erlbaum; 1988.
Edwards RR, Ness TJ, Weigent DA, Fillingim RB. Individual
diﬀerences in diﬀuse noxious inhibitory controls (DNIC): associ-
ation with clinical variables. Pain 2003a;106:427–37.
Edwards RR, Fillingim RB, Ness TJ. Age related diﬀerences in endoge-
nous pain modulation: a comparison of diﬀuse noxious inhibitory
controls in healthy older and younger adults. Pain 2003b;101:155–65.
Edwards RR. Individual diﬀerences in endogenous pain modulation as
a risk factor for chronic pain. Neurology 2005;65:437–43.
Edwards RR, Sarlani E, Wesselmann U, Fillingim RB. Quantitative
assessment of experimental pain perception: multiple domains of
clinical relevance. Pain 2005;114:315–9.
France CR, Suchowiecki S. A comparison of diﬀuse noxious inhibitory
controls in men and women. Pain 1999;81:77–84.
Fujii K, Motohashi K, Umino M. Heterotopic ischemic pain atten-
uates somatosensory evoked potentials induced by electrical tooth
stimulation: diﬀuse noxious inhibitory controls in the trigeminal
nerve territory. Eur J Pain 2006;10:495–504.
Ge HY, Madeleine P, Arendt-Nielsen L. Sex diﬀerences in temporal
summation characteristics of descending inhibition control: an
evaluation using repeated bilateral experimental induction of
muscle pain. Pain 2004;110:72–8.
Goﬀaux P, Redmond WJ, Rainville P, Marchand S. Descending
analgesia – when the spine echoes what the brain expects. Pain
Granot M, Granovsky Y, Sprecher E, Nir RR, Yarnitsky D. Contact
heat-evoked temporal summation: tonic versus repetitive-phasic
stimulation. Pain 2006;122:295–305.
Granot M, Goldstein Ferber S. Prediction of post-operative pain
intensity by assessment of pain catastrophizing scale and anxiety.
Clin J Pain 2005;21:439–45.
Grill JD, Coghill RC. Transient analgesia evoked by noxious stimulus
oﬀset. Neurophysiology 2002;87:2205–8.
Julien N, Goﬀaux P, Arsenault P, Marchand S. Widespread pain in
ﬁbromyalgia is related to a deﬁcit of endogenous pain inhibition.
Keefe FJ, Lefebvre JC, Egert JR, Aﬄeck G, Sullivan MJ, Caldwell DS.
The relationship of gender to pain, pain behavior, and disability of
osteoarthritis. Pain 2000;87:325–34.
from vibration and heterotopic noxious conditioning stimulation
(HNCS) in ﬁbromyalgia and healthy subjects. Pain 1997;70:41–51.
Lautenbacher S, Rollman GB. Possible deﬁciencies of pain modulation
in ﬁbromyalgia. Clin J Pain 1997;13:189–95.
Lautenbacher S, Roscher S, Strian F. Inhibitory eﬀects do not depend
on the subjective experience of pain during heterotopic noxious
conditioning stimulation (HNCS): a contribution to the psycho-
physics of pain inhibition. Eur J Pain 2002;6:365–74.
Le Bars D, Dickenson AH, Besson JM. Diﬀuse noxious inhibitory
controls (DNIC). I. Eﬀects on dorsal horn convergent neurons in
the rat. Pain 1979;6:283–304.
Le Bars D, Villanueva L, Bouhassira D, Willer JC. Diﬀuse noxious
inhibitory controls (DNIC) in animals and in man. Patol Fiziol
Eksp Ter 1992;4:55–65.
Le Bars D, Bouhassira D, Villanuera L. Opioids and diﬀuse noxious
inhibitory controls in the rat. In: Bromm B, Desmendt JE, editors. Pain
and the brain: from nociceptor to cortical activity’, advances in pain
researchandtherapy1995;22. NewYork: RavenPress;1995.p. 517–39.
Le Bars D. The whole body receptive ﬁeld of multireceptive neurones.
Prog Brain Res 2002;40:29–44.
Le Bars D, Willer JC. Pain modulation triggered by high-intensity
stimulation: implication for acupuncture analgesia. In: Satoh A,
editor. International Congress series 1238. Amsterdam: Elsevier; 2002.
Marchand S, Arsenault P. Spatial summation for pain perception:
interaction of inhibitory and excitatory mechanisms. Pain
Mayer EA, Berman S, Chang L, Naliboﬀ BD. Sex based diﬀerences in
gastrointestinal pain. Eur J Pain 2004;8:451–63.
Maixner W, Fillingim R, Booker D, Sigurdsson A. Sensitivity of
patients with painful temporomandibular disorders to experimen-
tally evoked pain. Pain 1995;63:341–51.
Olesen J. Clinical and pathophysiological observations in migraine and
tension-type headache explained by integration of vascular, supra-
spinal and myofascial inputs. Pain 1991;46:125–32.
Paulson PE, Minoshima S, Morrow TJ, Casey KL. Gender diﬀerences
in pain perception and patterns of cerebral activation during
noxious heat stimulation in human. Pain 1998;76:223–9.
Pielsticker A, Haag G, Zaudig M, Lautenbacher S. Impairment of pain
inhibition in chronic tension-type headache. Pain 2005;118:215–23.
Pud D, Sprecher E, Yarnitsky D. Homotopic and heterotopic eﬀects of
endogenous analgesia in healthy volunteers. Neurosci Lett
Sandrini G, Rossi P, Milanov I, Serrao M, Cecchini AP, Nappi G.
Abnormal modulatory inﬂuence of diﬀuse noxious inhibitory
controls in migraine and chronic tension-type headache patients.
Schouenborg J, Dickenson A. Eﬀects of a distant noxious stimulation
on A and C ﬁbre-evoked ﬂexion reﬂexes and neuronal activity in
the dorsal horn of the rat. Brain Res 1985;25:23–32.
Serrao M, Rossi P, Sandrini G, Parisi L, Amabile GA, Nappi G, et al.
Eﬀect of diﬀuse noxious inhibitory controls on temporal summa-
tion of the RIII reﬂex in humans. Pain 2004;112:353–60.
Seminowicz DA, Davis KD. Cortical responses to pain in healthy
individuals depends on pain catastrophizing. Pain 2006;120:297–306.
Spielberger CD, Gorsuch RL, Lushene R. The State-Trait Anxiety
Inventory. Palo Alto: Consulting Psychologists Press; 1969.
Staud R, Robinson ME, Vierck Jr CJ, Price DD. Diﬀuse noxious
inhibitory controls (DNIC) attenuate temporal summation of
second pain in normal males but not in normal females or
ﬁbromyalgia patients. Pain 2003;101:167–74.
Sigurdsson A, Maixner W. Eﬀects of experimental and clinical noxious
counterirritants on pain perception. Pain 1994;57:265–75.
Sullivan MJL, Bishop S, Pivic J. The pain catastrophizing scale:
development and validation. Psychol Assess 1995;7:524–32.
Tuveson B, Leﬄer AS, Hansson P. Time dependent diﬀerences in pain
sensitivity during unilateral ischemic pain provocation in healthy
volunteers. Eur J Pain 2006;10:225–32.
Teichman Y, Malineck F. Manual for the Hebrew State-Trait Anxiety
Inventory. Tel-Aviv Israel: Tel-Aviv University Press; 1978.
Villanueva L, Le Bars D. The activation of bulbo-spinal controls by
peripheral nociceptive inputs: diﬀuse noxious inhibitory controls.
Biol Res 1995;28:113–25.
Wiesenfeld-Hallin Z. Sex diﬀerences in pain perception. Gend Med
Zubieta JK, Smith YR, Bueller JA, Xu Y, Kimbourn MR, Jewett DM,
et al. Mu-opioid receptor-mediated antinociceptive response diﬀer
in men and women. J Neurosci 2002;22:5100–7.
M. Granot et al. / Pain 136 (2008) 142–149 149