Pain 136 (2008) 250–261

www.elsevier.com/locate/pain

Emotional control of nociceptive reactions (ECON):

Do affective valence and arousal play a role?
Jamie L. Rhudy *, Amy E. Williams, Klanci M. McCabe,
Jennifer L. Russell, Lauren J. Maynard
Department of Psychology, The University of Tulsa, 600 South College Avenue, Tulsa, OK 74104, USA

Received 30 November 2006; received in revised form 6 June 2007; accepted 25 June 2007

Abstract

Prior research suggests emotional picture-viewing modulates motoric (nociceptive flexion reflex), autonomic (skin conductance
response, heart rate acceleration), and subjective (pain rating) reactions to noxious electrodermal stimulation. The present study
sought to determine whether emotional valence and arousal contribute to nociception modulation. To do so, pictures varying in
emotional content (erotica, food, neutral, loss, attack) were chosen to manipulate emotional valence (pleasant = erotic and food;
unpleasant = loss and attack) and arousal (low = food and loss; moderate = erotica and attack). Pictures were presented in pseu-
dorandom order to elicit emotional processing while noxious electric stimulations were delivered to the sural nerve. Nociceptive flex-
ion reflex (NFR) magnitude, skin conductance response (SCR), heart rate (HR) acceleration, and subjective pain ratings to each
stimulation were measured, standardized, averaged by picture content, and analyzed. Results suggested that picture-viewing
explained 52% of the variance in the multivariate combination of the nociceptive reactions and modulated them in parallel. Pleasant
pictures inhibited reactions, whereas unpleasant pictures enhanced them. However, only erotica and attack pictures elicited signif-
icant modulation relative to neutral pictures, suggesting arousal also contributed. An exploratory multilevel analysis also supported
this conclusion. Together, these data suggest emotional control of nociceptive reactions (ECON) is associated with a valence-by-
arousal interaction. Implications of these findings for how emotional picture-viewing can be used to study supraspinal modulation
are discussed.
 2007 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

Keywords: Affect; Nociceptive flexion reflex; Heart rate; Skin conductance; Pain perception; Pain modulation

1. Introduction 2007b). Unfortunately, little research has been con-

Emotional processing influences subjective pain (e.g.,
de Wied and Verbaten, 2001; Keefe et al., 2001; Mea-
gher et al., 2001; Villemure et al., 2003) and physiologi-
cal reactions to noxious stimulation (nociceptive flexion
reflex, skin conductance response, heart rate accelera-
tion, cortical potentials) (Kenntner-Mabiala and Pauli,
2005; Rainville et al., 2005; Rhudy et al., 2005, 2006,
*
Corresponding author. Tel.: +1 918 631 2839; fax: +1 918 631
2833.
E-mail address: jamie-rhudy@utulsa.edu (J.L. Rhudy).
ducted to elucidate factors that determine the emo-
tion–pain relationship.
We have used motivational priming theory (MPT;
Lang, 1995) as a conceptual framework to make predic-
tions about the influence of emotional processing on
pain (Rhudy and Meagher, 2001). MPT purports two
central opponent motive systems are responsible for
emotional processing (Lang and Davis, 2006). The appe-
titive system is activated by appetitive stimuli (e.g., sex,
food) and results in positively valenced emotions. The
defensive system is activated by harmful or potentially
harmful stimuli (e.g., threat, noxious stimuli) and results
in negatively valenced emotion. Thus, measures of

0304-3959/$34.00  2007 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.pain.2007.06.031
 J.L. Rhudy et al. / Pain 136 (2008) 250–261
valence/pleasure have been used to assess which motive
system is activated.
The imminence of appetitive consummation (appeti-
tive system) or threat (defensive system) determines the
strength/intensity of motive system activation and pro-
motes action via metabolic arousal (Bradley, 2000).
Thus, measures of arousal are used to assess the degree
of system activation. Not surprisingly, stimuli with
greater motivational-relevance, such as sexual objects
and threatening stimuli, lead to strong appetitive and
defensive system activation (and arousal), respectively
(Bradley et al., 2001).
MPT predicts that if a motive system is activated
(primed), future responses emanating from that system
will be facilitated, whereas responses from the opposing
system will be inhibited (Lang, 1995). For example, defen-
sive system activation facilitates subsequent defensive
responses, while inhibiting appetitive responses. Further,
MPT predicts the degree of system activation determines
the degree of facilitation/inhibition (Cuthbert et al.,
1996): greater activation = greater modulation. There-
fore, defensive reactions to nociceptive input should be
inhibited by appetitive activation (positive emotions)
and facilitated by defensive activation (negative emo-
tions). Moreover, nociception modulation should be
greater when systems are strongly activated (indirectly
assessed from higher arousal). No single study, however,
has tested this hypothesis by independently manipulating
the system activated (valence) and the degree of activation
(arousal). Based on MPT, it is predicted that the motive
system (emotional valence) determines the direction,
and emotional activation (arousal) determines the degree,
of nociception modulation.
The present study used a picture-viewing paradigm to
study the emotion–pain relationship. Pictures varying in
content (attack, loss/grief, neutral objects, food, erotica)
were used to manipulate motive system activation
(valence: unpleasant, neutral, pleasant) and activation
level (arousal: low vs. moderate). During picture-view-
ing, noxious electric stimulations were delivered to the
sural nerve and nociceptive reactions were assessed
(nociceptive flexion reflex, skin conductance response,
heart rate acceleration, subjective pain). It was predicted
that unpleasant pictures would facilitate nociceptive
reactions and pleasant pictures would inhibit them
(Rhudy et al., 2005, 2006, 2007b). However, modulation
was expected to be greatest during erotic and attack pic-
tures because they elicit the greatest activation/arousal.
2. Method
2.1. Participants
All procedures were fully approved by The University of
Tulsa Ethics Review Board and participants recruited from
the psychology department gave informed consent after details
251
of the experiment were described to them. Participants were
excluded for: <18 yrs old, cardiovascular, neurological, and/
or circulatory problems, recent use of analgesic, anxiolytic,
or antidepressant medication, psychological trauma as defined
by the DSM-IV (APA, 2000), specific phobia of snakes or spi-
ders, or Raynaud’s disease. Twenty-five healthy participants
(10 male, 15 female) completed the study and were included
in the analyses; however, an additional male participant was
excluded for equipment failure and three others (2 female, 1
male) were excluded because a nociceptive flexion reflex could
not be obtained (see Section 2.6.1). Of the total 29 participants,
most were female (59%), White non-Hispanic (55.2%), single
(89.7%), and unemployed (55.6%), with an average age of
21.81 yrs (SD = 5.50).
2.2. Apparatus
A computer running LabVIEW software (National Instru-
ments, Austin, TX) equipped with dual monitors and A/D
board (National Instruments, PCI-6036E) controlled all stim-
uli, questionnaire presentation, and data acquisition, and
was used for offline data reduction. One 1700 flat panel monitor
was used by the experimenter to monitor physiological signals
and experimental timing. The other 1700 flat panel monitor
(positioned .5 m from the participant) presented visual stimuli
and questionnaires. Electrodermal stimulations were delivered
by a Grass Instruments stimulator (Model S88, West Warwick,
RI), stimulus isolation unit (Model SIU8T), constant current
unit (Model CCU1), and bipolar stimulating electrode (Nico-
let, 019-401400, Madison, WI). The stimulating electrode
was attached to the left ankle over the retromalleolar pathway
of the sural nerve. The onset/offset of the stimulator was
controlled by computer, and a computer-controlled voltage
regulator varied the current to the participant (max cur-
rent = 40 mA). All psychophysiological signals were sampled
at 1000 Hz and collected/filtered using a Grass Instruments
Model 15LT Bipolar Amplifier with Quad AC (15A54) and
Dual DC (15A12) modules. Skin conductance response
(SCR) was measured using an adaptor (Grass, Model SCA1)
for the 15A12 amplifier and electrodes filled with isotonic paste
(EC33, Grass Instruments).
To apply all EMG, ECG, and stimulating electrodes, first
the skin was degreased with alcohol, slightly abraded using
NuPrep gel to achieve impedances below 10 KX, and then con-
ductive gel (EC60, Grass Instruments) was applied. All record-
ing electrodes were Ag–AgCl. Psychophysiological signals
were sampled in 11 s trials: 3 s before picture presentation
and 8 s after picture onset. Throughout testing, participants
were seated in a recliner with a small pillow positioned under
their left ankle to facilitate relaxation. Sound attenuating
headphones and a video camera allowed the experimenter to
communicate with and monitor the participant from an adja-
cent room.
2.3. Emotion-induction: picture-viewing
Emotion elicited by picture-viewing has been shown to reli-
ably alter pain and nociception (de Wied and Verbaten, 2001;
Meagher et al., 2001; Wunsch et al., 2003; Kenntner-Mabiala
and Pauli, 2005; Rhudy et al., 2005, 2006, 2007b). Therefore,
252 J.L. Rhudy et al. / Pain 136 (2008) 250–261
60 digital pictures1 from five picture contents (12 pictures per
content) were chosen from the International Affective Picture
System (CSEA, 1999; Lang et al., 1999) that should manipu-
late emotional valence and arousal. Valence was manipulated
by two pleasant picture contents (couples in erotic poses,
food), two unpleasant picture contents (depictions of loss/
grieving, animal and human attack scenes), and neutral pic-
tures (household objects, mushrooms). Erotic and attack pic-
ture contents were chosen to elicit moderate motive system
activation and arousal, whereas food and loss contents were
chosen to elicit low motive system activation and arousal.
Mean normative valence/pleasure and arousal ratings for each
picture content were: erotic couples (valence: M = 6.66, arou-
sal: M = 6.10), food (valence: M = 6.88, arousal: M = 4.82),
neutral (valence: M = 4.93, arousal: M = 2.53), loss (valence:
M = 2.65, arousal: M = 4.56), and human and animal attack
(valence: M = 3.00, arousal: M = 6.35) (See section entitled
‘‘Subjective reactions to picture-viewing’’ below for description
of rating scales) (Lang et al., 1999). Thus, pictures containing
erotic and food contents were roughly equivalent in normative
pleasure/valence ratings, as were attack and loss contents.
Moreover, erotic and attack contents were roughly equivalent
in normative arousal ratings, but higher than loss and food
contents which were roughly equivalent. Therefore, pictures
containing erotica, food, loss, and attack contents were chosen
to independently manipulate affective valence and arousal, and
neutral pictures served as controls.2 Pictures were presented by
computer in a dimly lit room. The order of presentation was
randomized across participants with the limitation that not
more than two pictures of similar content were shown consec-
utively. Picture duration was also manipulated. Half of the pic-
tures were presented for 500 ms and the other half 6 s
(duration balanced across content) because these durations
have been shown to effectively modulate the defensive acoustic
startle reflex (Codispoti et al., 2001). Picture duration did not
have a significant influence on nociceptive reactivity; thus, the
picture duration IV was removed from the final statistical
models predicting nociceptive outcomes.
To examine the effect of picture-viewing on nociceptive
reactions, electric stimulations at 120% NFR threshold (see
Section 2.6.1) were delivered randomly during 50% of pictures,
with the limitation that stimulations were equally distributed
1
Image numbers were: erotic couples (4607, 4611, 4651, 4666, 4669,
4670, 4690, 4810, 4606, 4608, 4623, 4653), food (7230, 7270, 7283,
7320, 7330, 7350, 7390, 7400, 7430, 7450, 7470, 7475), neutral (7000,
7004, 7010, 7025, 7030, 7040, 7080, 7090, 7150, 7175, 7235, 5531), loss
(2141, 2205, 2276, 2455, 2700, 2800, 2810, 2900, 9000, 9001, 9220,
9421), and human and animal attack scenes (1052, 1111, 1205, 1301,
1525, 1932, 3500, 6230, 6242, 6250, 6550, 6560).
2
Although these picture contents were chosen to independently
manipulate emotional valence and arousal, it must be pointed out that
valence and arousal manipulations are by definition confounded with
picture content. Thus, it is possible that modulatory differences noted
between attack and loss, or erotic and food contents may reflect
differences due to content rather than arousal/intensity. Indeed, a
recent study attempted to manipulate valence and arousal independent
of content and found that picture content, emotional valence, and
emotional arousal all had independent modulatory effects on the
acoustic startle reflex (Bernat et al., 2006). Specifically, erotic and
threat pictures had the greatest modulatory influences that appeared
independent of reported arousal.
across picture content and picture duration. Therefore, six
stimulations were delivered during each picture content. To
minimize conditioning to the pictures, 10 stimulations were
also delivered during randomly chosen inter-picture intervals,
for a total of 40 stimulations during the picture-viewing
procedure.
2.4. Subjective reactions to picture-viewing
Subjective emotional evaluations of pictures were assessed
using the Self-Assessment Manikin (SAM; Bradley and Lang,
1994). The SAM is a two item questionnaire that consists of
two sets of five pictographs depicting affective valence/pleasure
(unpleasant–pleasant) and arousal (calm-excited). A comput-
erized version of the SAM was used to rate affective valence
and arousal following each picture (Rhudy et al., 2005). To
respond, participants dragged an indicator on or between
any of the five pictographs for each scale and submitted their
answers by computer mouse. This yielded ratings between 1
and 9 for each dimension, with higher scores indicating greater
pleasure or arousal, depending on the scale. Pictures were
expected to vary in emotional valence, with erotic and food
pictures eliciting the greatest pleasure, neutral pictures inter-
mediate, and loss and attack pictures the least pleasure (Lang
et al., 1999). Erotic and food pictures were expected to elicit
equivalent pleasure, whereas loss and attack were expected
to elicit equivalent displeasure (low pleasure). Arousal was
expected to be highest in response to erotic and attack pictures,
intermediate for food and loss, and lowest for neutral (Lang
et al., 1999).
2.5. Psychophysiological reactions to picture-viewing
In paradigms using picture-viewing to evoke emotions, psy-
chophysiological reactions are typically used in conjunction
with subjective reactions to verify emotion-induction. SCR is
used as a physiologic correlate of subjective arousal and HR
as a correlate of subjective valence (Lang et al., 1993; Bradley
et al., 2001). However, we have failed to observe these picture-
evoked ANS responses when shocks are delivered throughout
picture-viewing, even from pictures not receiving a shock
(Rhudy et al., 2005, 2006). We argue this is likely due to the
masking of subtle picture-evoked ANS reactions when pictures
are viewed in a threatening context (i.e., during randomly
delivered shocks) (Rhudy et al., 2005, 2006). Nonetheless,
these measures were recorded and analyzed to facilitate com-
parisons with other picture-viewing studies. However, we do
not predict SCR or HR will covary significantly with subjective
arousal or valence, respectively.
SCR was measured by placing electrodes filled with isotonic
paste (EC33, Grass Instruments) on the volar surface of the
middle and ring fingers after participants washed and dried
their hands (Venables and Christie, 1980). Electrocardiogram
(ECG) was measured using Ag–AgCl electrodes that were
filled with conductive gel (EC60, Grass Instruments) and
applied to the left and right forearms. ECG was converted off-
line to heart rate (HR) in beats-per-minute in half-second bins
by determining the interbeat interval in milliseconds using
LabVIEW software. Picture-evoked reactions in SCR and
HR were calculated by subtracting mean activity in the 1 s
 J.L. Rhudy et al. / Pain 136 (2008) 250–261
prior to picture presentation from 12 half-second bins follow-
ing picture onset (Bradley et al., 2001). Picture-evoked SCR
was determined by finding the maximum change that occurred
1–4 s after picture onset. Picture-evoked HR was calculated in
two ways: (1) maximum deceleration from baseline in the first
3 s of picture-viewing and (2) peak acceleration during the last
3 s of picture-viewing (Bradley et al., 2001). To eliminate stim-
ulus artifact due to shocks, picture-evoked SCR and HR reac-
tions were only calculated for pictures in which a shock was
not presented.
2.6. Nociceptive reactions
2.6.1. Nociceptive flexion reflex (NFR)
The NFR is a protective withdrawal reflex in response to a
noxious stimulus (Skljarevski and Ramadan, 2002; Sandrini
et al., 2005). Experimentally, the NFR is typically elicited by
electrodermal stimulation over the sural nerve that activates
primary nociceptors (Ad and C fibers). Nociceptor activation
results in a nociceptive signal to the spinal cord that subse-
quently elicits a withdrawal response in the leg that can be
quantified using electromyography (EMG) (Wiesenfeld-Hallin
et al., 1984; Sandrini et al., 2005). The NFR is a spinal reflex
because supraspinal regions are not necessary for its elicitation
(e.g., Sandrini et al., 1999). The stimulation intensity necessary
to reliably elicit the NFR (i.e., NFR threshold) correlates with
subjective report of pain threshold, and the magnitude of the
NFR correlates with subjective pain intensity (Chan and Dal-
laire, 1989; Guieu et al., 1992; Rhudy et al., 2005). Thus, the
NFR is used as an objective, indirect measure of spinal
nociception.
Two active recording electrodes were placed over the left
biceps femoris muscle 10 cm superior to the popliteal fossa
to assess EMG activity associated with the NFR. A common
ground electrode was placed over the left lateral epicondyle
of the femur. Biceps femoris EMG activity 90–150 ms post-
stimulation was used to define the NFR. Prior research has
shown that using the 90–150 ms timeframe avoids potential
contamination by the non-nociceptive RII reflex, startle
responses, and/or voluntary movements (Dowman, 1992;
France and Suchowiecki, 2001). The raw biceps femoris signal
was amplified by 20,000, bandpass filtered (10–300 Hz), and
rectified.
The intensity of electrodermal stimulation delivered
throughout picture-viewing was set at 120% NFR threshold.
To determine NFR threshold, procedures were adapted from
France and colleagues (France et al., 2002). During threshold
assessment, NFR was defined as a mean EMG response in the
90–150 ms post-stimulus interval that exceeded mean EMG
activity during the 60 ms pre-electrical stimulus baseline inter-
val by at least 1.0 standard deviation. Stimulation of the sural
nerve for threshold determination proceeded using an up–
down staircase method with a variable interval of 8–12 s while
the participant sat comfortably in a recliner with their leg out-
stretched (knee angle approximately 180; e.g., Chan and
Tsang, 1985). Each stimulation consisted of five rectangular
wave pulses of 1 ms duration and 3 ms interstimulus interval
(250 Hz). The threshold assessment procedure started at
0 mA (current) and increased in 1.5 mA steps until a NFR
was detected. Stimulus intensity was then decreased in
253
.75 mA steps until a reflex was no longer detected. This up–
down procedure was repeated two more times with .5 mA
steps. The average stimulus intensity of the last two peaks
and troughs was increased 20% to generate the intensity of
stimulation used during picture-viewing (120% NFR thresh-
old). This intensity minimizes ceiling and floor effects (Danzi-
ger et al., 1998). Participants for which a NFR could not be
reliably obtained before tolerance was achieved were given
the choice of repeating the threshold assessment procedure
or discontinuing the experiment (three discontinued as noted
in the Section 2.1).
To examine the influence of picture-viewing on NFR,
within-subject changes in NFR magnitude were examined.
NFR magnitude was calculated for stimulations delivered dur-
ing picture-viewing by subtracting mean activity in the 60 ms
prior to electric stimulation from mean activity in the 90–
150 ms post-stimulation window.
2.6.2. Subjective pain ratings
To rate each electric stimulation, a computer-presented rat-
ing scale oriented vertically was used (France et al., 2002;
Rhudy et al., 2005). Bottom-to-top, the scale was labeled: 0
(no sensation), 1 (just noticeable), 25 (uncomfortable), 50
(painful), 75 (very painful), and 100 (maximum tolerable).
To respond, participants moved an indicator anywhere along
the scale and submitted their answers by computer mouse.
These ratings were used as a measure of subjective pain.
2.6.3. Autonomic reactions
Research from our laboratory has found that peak auto-
nomic reactions (skin conductance response [SCR]; peak heart
rate [HR] acceleration) to noxious stimulation are modulated
by emotional processing (Rhudy et al., 2007b). Specifically,
peak SCR and HR acceleration following noxious shock are
diminished by appetitive activation and enhanced by defensive
activation. Therefore, we chose to focus only on these auto-
nomic measures.3 SC activity was first averaged by .5 s epochs,
and then the mean activity in the 1 s pre-shock interval was
subtracted from each .5 s post-shock epoch (Bradley et al.,
2001). SCR was defined as the maximum skin conductance
increase in the 1–4 s post-shock interval (Dawson et al.,
3
These methods of scoring SCR and HR acceleration have been
used in previous emotion research to determine picture-evoked
reactions (Bradley et al., 2001). However, these scoring methods may
lead to artificially high values, because they base responses on peak
values in a narrow scoring window. To determine whether the
emotional modulation effects noted in this study were artifacts of
our scoring methods we also calculated HR as the mean change in the
1–5 s post-stimulus window and SC as the mean change in the 1–4
post-stimulus window. These change scores were then standardized
and compared to the standardized versions of our original HR
acceleration and SCR variables. The two scoring methods were highly
correlated with each other (SCR, r > .90; HR, r = .59; ps < .001).
Analyses of the effect of picture-viewing on these new variables
resulted in nearly identical results as those reported using the HR
acceleration and SCR variables. Picture content explained 36% of the
variance in SC change and 23% in HR change. The pattern of SC
change means and HR change means for each picture content were the
same as those noted for SCR and HR acceleration, thus we chose to
keep our original variables in the manuscript to be consistent with our
previous study (Rhudy et al., 2007b).
254 J.L. Rhudy et al. / Pain 136 (2008) 250–261
2000). Therefore, the .5 s post-shock epoch (in the 1–4 s post-
shock interval) with the greatest positive change was used as
the response definition (Bradley et al., 2001). ECG was con-
verted offline to HR in beats-per-minute from interbeat inter-
vals and visually inspected for errors. Like SC, the HR
waveform was first averaged by .5 s epochs and the mean activ-
ity in the 1 s pre-shock interval was subtracted from each .5 s
post-shock epoch. Peak HR acceleration was defined as the
.5 s post-shock epoch (in the 1–5 s post-shock interval) with
the greatest positive change.
2.7. Procedure
Health status of all participants was determined using a
brief health questionnaire and interview to determine eligibil-
ity. If eligible, electrodes were applied, the participant was
seated comfortably in a reclining chair, and then familiarized
with the rating scales used throughout testing (SAM, pain
rating scale). Ratings were made using a computer mouse
positioned on a lap desk. Participants were told that there
were two phases to the experiment. Phase 1 (NFR threshold
assessment) involved sending electric pulses to the ankle to
determine the stimulus intensity necessary to elicit the
NFR. During this phase, the participant viewed the pain rat-
ing scale positioned next to a digital green light (both were
continuously displayed on the computer monitor). The light
indicated when the participant was to make a rating (follow-
ing every stimulus). The participant was told to make a rating
whenever the light came on, regardless of whether the stimu-
lus was perceived. During phase 2, the participant was
instructed to view every picture presented on the computer
monitor, and told that electric stimuli would be delivered ran-
domly during and in between pictures. Picture duration was
varied in a pseudorandom manner, with half of the pictures
having 500 ms duration and the other half 6 s duration. Pic-
ture durations were equally distributed across picture con-
tents. Inter-picture intervals varied randomly from 12 to
22 s. Electric stimuli (at 120% NFR threshold) were delivered
during 50% of the pictures (six for each picture content, bal-
anced across picture duration) and 10 inter-trial intervals.
Thus, a total of 40 stimulations were delivered during phase
2. During phase 2, electric stimuli were randomly delivered 3–
5 s after picture onset and 11–21 s after inter-picture interval
onset to reduce predictability (shock onset intervals were
always an integer). The 3–5 s post-picture-offset interval was
chosen because it has been shown to produce the largest emo-
tional modulation effects on the acoustic startle reflex (Brad-
ley et al., 1993; Codispoti et al., 2001). Following each
picture, the SAM was administered to assess emotional
responses to the pictures. The pain rating scale was adminis-
tered following pictures and inter-picture intervals in which
an electric stimulus was delivered. At the end of the experi-
ment, the participant was debriefed and thanked for their
participation.
2.8. Data reduction and analyses
All self-report and psychophysiological variables were aver-
aged across pictures with similar content (attack, loss, neutral,
food, erotic). Before averaging, SCR, HR acceleration, NFR,
and pain ratings were standardized within-individuals by con-
verting to z scores.4 Standardizing has the effect of eliminating
between-subject variability (by setting each participant’s mean
response to zero), while retaining within-subject variability due
to picture-viewing, and is common in psychophysiological
studies (e.g., Blumenthal et al., 2005). Therefore, individual
differences in the absolute magnitude of each variable are
removed and the effect of picture-viewing (a within-subject
effect) is retained. Furthermore, standardizing has the benefi-
cial effect of placing all nociceptive reactions in a common met-
ric (standard deviation units) thus facilitating the
interpretation of analyses that collapse across reaction type
(analyzing all reactions simultaneously, see below).
Although analyses were conducted to examine the indepen-
dent influence of affective valence and arousal, a single inde-
pendent variable (picture content) was used to model the
influence of valence and arousal rather than separate indepen-
dent variables. This allowed statistical models to make com-
parisons to neutral pictures (which could not be done if
valence and arousal were separate independent variables).
Analyses on manipulation checks for emotion-induction were
conducted using a 5 (Picture Content) · 2 (Picture Dura-
tion) · 2 (Participant Sex) repeated measures MANOVA. To
examine the influence of picture-viewing on nociceptive reac-
tions, a 4 (Reaction Type) · 5 (Picture Content) repeated mea-
sures MANOVA was conducted. Initially, participant sex and
picture duration were also included as independent variables in
the model predicting nociceptive reactions, but these effects
were not significant (Fs < 2.10, ps > .16). Thus, for parsimony,
they were dropped from the final model. Follow-up compari-
sons were conducted with Bonferroni adjusted means tests to
control for Type I error rate, except a priori hypotheses of
nociceptive outcomes were one-tailed Fisher LSD tests. Unless
noted, analyses were conducted using the GLM procedure
within SPSS 14.0. Partial eta-squared (g2) was used as the effect
size for F tests and Cohen’s d was used for mean comparisons.
Cohen (1977) provides guidelines for interpreting g2
(small = .01, medium = .06, large = .14) and d (small = .2,
medium = .5, large = .8). Wilks’ k are the multivariate statis-
tics interpreted but are not presented given the redundancy
with partial eta-squared.
2.8.1. Hypotheses
It was predicted that nociceptive reactions would be modu-
lated in parallel and influenced by emotional valence and arou-
sal (valence-by-arousal interaction). Specifically, nociceptive
reactions were expected to be facilitated by unpleasant pictures
and inhibited by pleasant pictures. However, pictures eliciting
greater arousal were expected to show stronger modulation –
attack pictures were expected to elicit the greatest facilitation
and erotic pictures the greatest inhibition. Therefore, only a
main effect of picture content was predicted to be significant
in the analysis of nociceptive reactions. Reactions during
4
It is important to note that within-subject z transformations can
have the beneficial effect of minimizing the impact of outlier values,
because means are centered at zero and the within-subject variability is
put on a common metric (in this case standard deviation units).
Therefore, the standardized means will ‘‘rein in’’ the outlier. At times,
this can cause a mean to change its magnitude relative to other means.
This is most obvious when comparing unstandardized means to
standardized means.
 J.L. Rhudy et al. / Pain 136 (2008) 250–261 255

attack and erotic pictures were expected to be significantly dif- For unpleasant contents, loss (p = .001, d = .89) and
ferent (higher and lower, respectively) than reactions during attack pictures (p < .001, d = 1.74) were more arousing
neutral pictures, whereas reactions during loss and food pic- than neutral, and attack pictures were more arousing
tures were not expected to differ significantly from reactions than loss (p < .001, d = .92). For pleasant contents, food
during neutral pictures.
(p = .001, d = .93) and erotic (p < .001, d = 1.46) pic-
tures were more arousing than neutral, but the compar-
3. Results ison between erotic and food pictures eluded significance
(p = .10, d = .57). The main effect of picture duration
3.1. Subjective reactions to picture-viewing suggested 6 s pictures elicited greater arousal than
500 ms pictures, albeit a small difference in subjective
Table 1 depicts subjective valence and arousal ratings arousal. Together, these data suggest that valence and
by picture content, picture duration, and participant sex. arousal were independently manipulated, although ero-
Analysis of valence/pleasure ratings found a main effect tic pictures evoked less subjective arousal than expected.
of picture content, F(4, 20) = 35.13, p < .001, g2 = .88.
Erotic and food pictures elicited similar pleasure 3.2. Psychophysiological reactions to picture-viewing
(p = 1.00, d = .13) that was greater than neutral pictures
(ps < .001, ds > 2.24). Attack and loss pictures were sim- Table 2 depicts psychophysiological reactions to pic-
ilar in pleasure (p = 1.00, d = .11) and lower than neu- ture-viewing by picture content and picture duration. As
tral pictures (ps < .001, ds > 1.41). However, this main predicted, picture contents did not lead to differential
effect was qualified by significant interactions of Picture psychophysiological reactions. For picture-evoked
Content · Participant Sex [F(4, 20) = 3.45, p = .027, SCR, only the main effect of picture duration was signif-
g2 = .41] and Picture Content · Picture Duration icant [F(4, 23) = 7.07, p = .01, g2 = .24], but not the
[F(4, 20) = 8.01, p = .001, g2 = .62]. To decompose the main effect of picture content (p = .42) or the Picture
Picture Content · Participant Sex interaction, the simple Content · Picture Duration interaction (p = .23). Brief
effect of sex was examined. Only the simple effect of sex pictures led to slightly greater SCR (0.13 lS vs.
for the neutral pictures was significant. Men rated neu- 0.18 lS). For picture-evoked HR deceleration, the main
tral pictures as slightly less pleasurable than women effects of picture content (p = .14) and picture duration
(Ms = 4.54 vs. 4.88; p = .036, d = .63). To decompose (p = .63) and the Picture Content · Picture Duration
the Picture Content · Picture Duration interaction, the interaction (p = .23) were all non-significant. For pic-
simple effect of picture duration was examined for each ture-evoked HR acceleration, the main effects of picture
picture content. The only effect of duration noted was content (p = .10) and picture duration (p = .06), and the
for ratings of loss content, with 6 s pictures eliciting Picture Content · Picture Duration interaction (p = .10)
greater displeasure than 500 ms pictures (Ms = 2.45 vs. were all marginally significant. HR acceleration tended
3.13; p < .001, d = .81). to be higher during brief pictures. However, exploratory
Analysis of arousal ratings revealed main effects of Bonferroni comparisons for the main effect of picture
picture content [F(4, 20) = 12.99, p < .001, g2 = .72] content and the simple effects of picture duration were
and picture duration [F(1, 23) = 7.18, p = .01, g2 = .24]. all non-significant (ps > .05). Thus, these data are con-
Table 1
Means and standard deviations of picture-evoked subjective reactions by picture content, picture duration, and participant sex
Attack Loss Neutral Food Erotica
Mean SD Mean SD Mean SD Mean SD Mean SD
Valence ratings
500 ms Male 3.08 0.73 3.18 0.55 4.57 0.25 6.03 0.73 6.43 1.36
Female 2.44 0.96 3.09 0.94 4.84 0.34 6.10 1.53 5.50 1.28
Total 2.70 0.92 3.13 0.80 4.73 0.33 6.07 1.25 5.87 1.37
6s Male 2.88 0.91 2.37 0.67 4.52 0.64 6.33 1.20 6.58 1.25
Female 2.37 1.09 2.51 1.01 4.92 0.46 6.28 1.34 5.58 1.28
Total 2.57 1.03 2.45 0.88 4.76 0.57 6.30 1.26 5.98 1.34

Arousal ratings
500 ms Male 5.60 1.11 3.87 0.95 3.24 1.57 4.89 1.80 5.79 1.32
Female 6.16 1.96 5.11 1.57 3.22 1.42 4.31 0.86 4.95 1.32
Total 5.93 1.67 4.61 1.47 3.23 1.45 4.54 1.31 5.29 1.36
6s Male 6.05 0.76 4.60 1.21 3.70 1.80 5.03 1.58 5.83 1.07
Female 6.51 1.95 5.08 1.85 3.22 1.39 4.40 1.26 4.96 1.43
Total 6.33 1.58 4.89 1.61 3.41 1.55 4.65 1.40 5.31 1.34
256 J.L. Rhudy et al. / Pain 136 (2008) 250–261

Table 2
Means and standard deviations of picture-evoked psychophysiological reactions by picture content and picture duration
Attack Loss Neutral Food Erotica
Mean SD Mean SD Mean SD Mean SD Mean SD
SCR (DlS)
500 ms 0.20 0.32 0.17 0.28 0.16 0.21 0.17 0.32 0.22 0.27
6s 0.16 0.23 0.13 0.17 0.18 0.32 0.10 0.18 0.13 0.17
Total 0.18 0.23 0.15 0.18 0.17 0.21 0.13 0.20 0.17 0.19
HR deceleration (Dbpm)
500 ms 5.84 4.27 4.18 2.80 3.83 2.90 4.13 2.90 4.94 4.19
6s 5.20 3.44 4.02 2.40 4.87 4.37 4.16 2.97 5.47 3.09
Total 5.52 3.01 4.10 1.92 4.35 3.06 4.14 2.20 5.21 2.64
HR acceleration (Dbpm)
500 ms 4.00 5.40 2.27 2.23 4.07 3.31 4.69 3.63 3.70 4.19
6s 4.37 6.52 2.45 2.54 3.06 2.55 3.30 2.90 2.39 2.81
Total 4.19 5.54 2.36 1.90 3.57 2.48 3.99 2.70 3.05 3.07
SCR, skin conductance response; HR, heart rate; bpm, beats-per-minute.

sistent with our previous research suggesting subtle
physiological responses to pictures are not observed
when shocks are randomly delivered during picture-
viewing (Rhudy et al., 2005, 2006).
3.3. The effect of picture-viewing on nociceptive reactions
As predicted, only a significant main effect of picture
content was found that explained 52% of the variance in
the multivariate combination of the standardized
reactions [F(4, 21) = 5.76, p = .003, g2 = .52]. Fig. 1
depicts standardized nociceptive reactions by picture
content. Standardized reactions during attack pictures
were significantly larger than during neutral (p = .032,
d = .59), but not loss pictures (p = .10, d = .38). Alter-
natively, standardized reactions during erotic pictures
were significantly smaller than all other picture contents
(ps < .02, ds = .68–1.55). Although reactions during loss
0.4
0.3
Response Magnitude
and food pictures were in the predicted directions, these
pictures did not lead to significant modulation relative
to neutral pictures (ps > .16, ds < .30). Importantly, the
Reaction Type · Picture Content interaction was not
significant (F = .45, p = .91). This suggests reactions
were modulated in parallel by picture contents. None-
theless, to provide a more stringent test of parallel mod-
ulation, the simple effect of reaction type was examined
at each level of picture content using exploratory Bon-
ferroni comparisons. Reactions were not significantly
different during attack (ps > .12), loss (ps > .70), neutral
(ps > .35), food (ps > .30), or erotic (ps > .26) pictures.
To determine the variance explained by picture content
in each separate reaction, partial g2 was calculated from
the simple effect of picture content. Picture content
explained 15% of the variance in NFR, 42% of the var-
iance in pain ratings, 37% of the variance in SCR, and
32% of the variance in HR acceleration. For reference
purposes, Table 3 presents unstandardized means for
each nociceptive reaction by picture content. However,
NFR
it is important to point out that standardization of vari-
Pain ables can cause changes in the relative magnitude of the

0.2
SCR means and are thus not directly comparable to unstan-
HR
dardized variables.
(z score)

0.1
0.0
-0.1 3.4. Exploratory analyses
-0.2
-0.3 3.4.1. Potential confounds
-0.4
Attack Loss Neutral Food Erotic Given that NFR, SCR, and HR reactions are all cal-
culated as a change from baseline, there is the potential
Fig. 1. The influence of picture content on nociceptive flexion reflex
(NFR), pain ratings (pain), skin conductance response (SCR), and
for picture-viewing to alter baseline physiological activ-
heart rate acceleration (HR) following noxious stimulation of the sural ity, and thus confound the interpretation of these vari-
nerve. Generally, unpleasant pictures enhanced nociceptive reactions, ables (even though prior analyses suggested pictures
whereas pleasant pictures inhibited them. However, the degree of did not lead to differential psychophysiological
modulation was determined by the emotional intensity (arousal) responses). For example, an incorrect inference of
elicited by the pictures – only the most arousing pictures (attack,
erotic) elicited modulation that was significantly different from the
SCR inhibition could result if erotic pictures, but not
neutral pictures. This suggests that valence and arousal both contrib- attack pictures, significantly increased baseline skin con-
ute to the modulation of nociception and pain. Error bars are SEM. ductance level, causing the response magnitude to
 J.L. Rhudy et al. / Pain 136 (2008) 250–261
Table 3
Means and standard deviations of unstandardized nociceptive reactions by picture content
ITI Attack Loss
Mean SD Mean SD Mean
NFR magnitude (DlV) 5.49 3.24 7.16 8.09 6.59
Pain rating (0–100) 43.49 23.46 47.45 24.73 47.29
SCR (DlS) 1.26 1.20 1.32 1.32 1.21
HR acceleration (Dbpm) 7.90 3.82 8.41 4.04 7.88
Note: NFR, nociceptive flexion reflex; SCR, skin conductance response; HR, heart rate; bpm, beats-per-minute.
appear smaller after the baseline activity is subtracted
(although the actual response magnitude might not have
changed). To ensure that this was not the case, the aver-
age local baselines used to calculate the change scores
for each physiological reaction were analyzed using 5
(Picture Content) · 2 (Picture Duration) ANOVAs.
Fig. 2 depicts local baselines averaged by picture content
for each of the physiological reactions as well as post-
shock waveforms of each reaction averaged by picture
content. No significant effects were found, suggesting
baselines were not significantly altered by the content
or duration of pictures. Therefore, change scores are a
valid method of quantifying the magnitude of these
responses.
3.4.2. The relationship between self-reported emotion and
nociceptive reactions
Although it is assumed that subjective emotion
(valence, arousal, Valence · Arousal interaction) at least
partially mediates the relationship between picture con-
tent and nociceptive reactivity, MANOVA models do
not directly test the relationship between subjective emo-
tion and nociception. To examine this relationship, a
multilevel model (hierarchical) correlational approach
was conducted using the SPSS 14.0 MIXED procedure
to examine covariation between trial-by-trial valence
and arousal ratings and trial-by-trial nociceptive reac-
tions. Thus, a benefit of this approach is that it takes
into account inter- and intra-individual variability in
reactions to each picture and each shock. A custom
model with REML estimation was built that included
reaction type, valence, arousal, Valence · Arousal inter-
action, and a Valence · Arousal · Reaction Type inter-
action as fixed effects and subject number as a random
effect. Significant effects of arousal [F(1, 2967)=11.48,
p = .001] and a Valence · Arousal interaction [F(1,
2967)=9.76, p = .002] were found. The effect of valence
just eluded significance [F(1, 2967)=3.47, p = .063]. The
main effect of reaction type [F(3, 2967)=.93, p = .42]
and the interaction of Valence · Arousal · Reaction
Type [F(1, 2967)=1.27, p = .28] were non-significant.
This suggests that subjective emotion does covary with
nociception modulation. Further, the analysis provides
additional evidence that a valence-by-arousal interac-
tion is associated with nociception modulation, and that
reactions were modulated in parallel.
257
Neutral Food Erotic
SD Mean SD Mean SD Mean SD
6.22 6.88 7.53 6.30 5.40 5.65 4.58
25.41 45.81 23.06 45.17 24.93 43.51 24.38
1.07 1.20 1.16 1.14 1.23 1.06 1.09
4.43 7.22 4.45 7.30 4.63 6.20 4.95
4. Discussion
The present study used picture-viewing to study
supraspinal modulation of objective (NFR magnitude,
SCR, HR acceleration) and subjective (pain ratings)
reactions to noxious sural nerve stimulation – a para-
digm referred to as emotional control of nociceptive
reactions (ECON). Extending prior research, this study
used different picture contents to vary valence and arou-
sal. Moreover, this was the first study to evaluate the
influence of picture-viewing on four nociceptive reac-
tions simultaneously. Pictures effectively manipulated
valence, with erotic and food pictures eliciting the great-
est pleasure, attack and loss pictures the lowest pleasure,
and neutral intermediate. Pictures also manipulated
emotional arousal, with erotic and attack pictures elicit-
ing the greatest arousal, loss and food pictures interme-
diate arousal, and neutral pictures the lowest arousal.
Surprisingly, erotic pictures did not elicit significantly
greater arousal than food pictures, as was predicted
from normative ratings. It is tempting to speculate that
sexual arousal may have been underreported in our sam-
ple due to geographic social norms. However, there is no
way to verify this given that objective picture-evoked
psychophysiological reactions are masked when pictures
are viewed in a threatening context (Rhudy et al., 2005,
2006).
Analyses suggested valence and arousal contributed
to modulation of nociceptive reactions. Reactions were
larger during unpleasant pictures and smaller during
pleasant pictures. But, only reactions elicited during ero-
tic and attack pictures were significantly modulated rel-
ative to neutral pictures. This was supported by an
exploratory multilevel analysis suggesting self-reported
arousal and the valence-by-arousal interaction were sig-
nificant predictors of nociceptive reactions.
Consistent with prior research (Rhudy et al., 2005,
2006, 2007b), the present study found nociceptive reac-
tions were modulated in parallel. To test this, all reac-
tions were placed on a common metric and analyzed
using a Reaction Type · Picture Content interaction
within MANOVA (minimizing Type I error). The inter-
action effect, exploratory simple effects tests of reaction
type, as well as the multilevel analysis all suggested reac-
tions were modulated in parallel. Importantly, picture
content explained 52% of the variance in the multivari-
258 J.L. Rhudy et al. / Pain 136 (2008) 250–261

6 NFR 13 NFR
12
Attack
Neutral

Change in Biceps Femoris EMG

11
Baseline Biceps Femoris EMG 5
Erotica
10
9
4
(micro Volts)

(micro Volts)
8
7
3
6
5
2
4
3
1 Nociceptive
2
Flexion
1 Reflex Window
0 0
Att Loss Neu Food Ero 30 60 90 120 150 180 210 240
Time from Shock (ms)

1.75 Attack
10 SCR SCR
Loss
1.50 Neutral
Skin Conductance Change

9
Baseline Skin Conductance

Food
1.25
(microSiemens)

Erotic
(micro Siemens)

8
1.00
7
0.75

6
0.50

5 0.25

4 0.00
Att Loss Neu Food Ero 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Time from Shock (s)

74 12 Attack
HR HR
Loss
72 10 Neutral
Food
Baseline Heart Rate

Heart Rate Change

70 8 Erotic
(bpm)

(bpm)

68 6

66 4

64 2

62 0
Att Loss Neu Food Ero 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

-2
Time from Shock (s)

Fig. 2. Local baselines used to calculate physiological nociceptive reactions and the time course of physiological nociceptive reactions following
noxious sural nerve stimulation. Panels on the left depict local baselines averaged by picture content for each of the physiological reactions (Att,
attack; Neu, neutral; Ero, erotic). Panels on the right are the post-shock waveforms of each reaction averaged by picture content (NFR waveforms
for loss and food pictures are not depicted in the top right panel to improve readability). The top panels are the nociceptive flexion reflex (NFR), the
middle panels are skin conductance response (SCR), and the bottom panels are heart rate (HR). Erotic pictures reduced physiological reactivity and
attack pictures enhanced physiological reactivity. (Note: Individual trials depicting NFR waveforms are characterized by a phasic response in the 90–
150 ms post-stimulation interval. In the present figure, however, the waveforms in the upper right panel appear ‘‘plateau-like,’’ with activity
beginning in the 90–150 ms interval and extending beyond it. This is due to grand averaging across individuals and trials that obscures the ‘‘silent
period’’ that typically occurs after the NFR (RIII component), but before startle and voluntary movements in individual trials. In the present graph,
this averaging artifact is particularly prominent, because there were participants who demonstrated large absolute EMG responses after the 90–
150 ms interval, and grand averages (means) are heavily influenced by outliers. This large individual variability in response magnitude is an important
reason for standardizing biceps femoris EMG (NFR magnitude) when the effect of interest is a within-subject effect, i.e., emotional modulation.)
 J.L. Rhudy et al. / Pain 136 (2008) 250–261
ate combination of the reactions, and considerable var-
iance in individual reactions (15% NFR, 42% pain rat-
ings, 37% SCR, 32% HR acceleration).
As noted here, picture content generally explains
more variance in pain report than other nociceptive
reactions (Rhudy et al., 2005, 2006, 2007b). One poten-
tial explanation could be report bias, because pain rat-
ings and picture (emotion) ratings are made together.
Contrary to this hypothesis, the picture content IV
explains greater variance in pain than emotion ratings
(Rhudy et al., 2005, 2006). Alternatively, different noci-
ceptive responses may be modulated by different pro-
cesses. Supporting this, emotional modulation of the
NFR is abolished when noxious shocks are predictable,
despite intact modulation of subjective pain (Rhudy
et al., 2006). Thus cortico-cortical circuits may continue
to modulate subjective pain even when brain-to-spinal
cord circuits are disengaged. This would explain why
some cognitive-emotional factors (catastrophizing, self-
efficacy) modulate subjective pain but not NFR (France
et al., 2002; Brant et al., 2007; Rhudy et al., 2007a).
Although compelling, the reason for modulatory differ-
ences across responses cannot be determined at this time.
4.1. Relation to prior research
This is the fourth study from our laboratory to exam-
ine the influence of emotional picture-viewing on noci-
ceptive reactions. Rhudy and colleagues (2005) were
the first to demonstrate emotional processes can activate
descending circuits that influence spinal nociception. In
that study, erotic, neutral, and attack pictures modu-
lated the spinally-mediated NFR and subjective pain.
NFR magnitudes and pain ratings were greater during
attack pictures and smaller during erotic pictures. As a
result, we argued ECON procedures represent novel,
non-invasive methods for engaging descending modula-
tion in humans.
Rhudy et al. (2006) examined whether reactions to
predictable noxious shocks were also modulated by pic-
ture-viewing. Hypothetically, if nociceptive reactions to
predictable shocks were also modulated, then employing
predictable shocks could reduce participant stress given
the preference for noxious events to be predictable
(Miller, 1981; Mineka and Henderson, 1985). In that
study, half of the participants received cued (predict-
able) shocks during picture-viewing, and the other half
received uncued (unpredictable) shocks. Data from the
uncued shock group replicated the first study. However,
pain ratings, but not NFR, were modulated in the cued
shock group. Thus, it was concluded that unpredictable
shocks should be used to observe NFR modulation (and
ensure brain-to-spinal cord circuitry is activated).
The third study combined data from the first two
studies to examine the impact of picture-viewing on
autonomic reactions to noxious shocks (Rhudy et al.,
259
2007b). Skin conductance and heart rate had been
recorded in the prior experiments as a physiological
manipulation check for emotion-induction. These sig-
nals were rescored to calculate shock-evoked change
(nociceptive reaction), rather than picture-evoked (emo-
tion-induction). As predicted, autonomic nociceptive
reactions were also modulated by emotion. Shock-
evoked SCR and HR acceleration were smaller during
pleasant pictures than during unpleasant pictures. Thus,
converging evidence suggests ECON can be used to
study modulation of nociceptive reactions assessed from
multiple response systems (motoric, autonomic, subjec-
tive), mediated by circuits at spinal (NFR) and supraspi-
nal (SCR, HR) levels.
4.2. The emotion–pain relationship
This study provides preliminary support for the
hypothesis that the emotion–pain relationship is deter-
mined by a valence-by-arousal interaction. Specifically,
valence determines the direction of the modulation and
arousal the magnitude. This hypothesis may not extend
to highly arousing emotions, however. Animal and
human studies (e.g., Fanselow, 1994; Rhudy and Mea-
gher, 2000) suggest nociceptive reactions are inhibited
by events that elicit intense defensive activation (highly
arousing negative emotion) – a reversal of the direction
of modulation compared to negative emotions with low-
to-moderate arousal. However, the direction of modula-
tion does not appear to reverse during highly arousing
positive emotions. The current study suggests positive
emotions with low-to-moderate arousal inhibit pain,
whereas other studies suggest intense appetitive activa-
tion (e.g., genital stimulation) that elicits highly arousing
positive emotion (e.g., orgasmic bliss) leads to profound
pain inhibition (Komisaruk and Whipple, 1986). When
all data are considered, a valence-by-arousal interaction
may still characterize the emotion–pain relationship
(Fig. 3). Picture-viewing would not be expected to
induce negative emotions capable of producing pain
inhibition, however. Inhibition is expected only when
negative emotions accompany imminent threat (Fanse-
low, 1994) and pictures are only likely to represent an
imminent threat for certain clinical populations (e.g.,
specific phobia of snakes).
4.3. Limitations
The present study was limited by a small sample. For-
tunately, picture-viewing exerts a large statistical effect on
nociceptive reactivity. However, other predictors in the
model (participant sex, picture duration) may have
achieved significance with a larger sample. Although the
Reaction Type · Picture Content interaction did not
achieve significance, exploratory simple effects tests pro-
vided some confidence low power was not responsible.
260 J.L. Rhudy et al. / Pain 136 (2008) 250–261

Defensive Activation (negatively valenced emotions)

independent influence of arousal. Indeed, a recent study
Appetitive Activation (positively valenced emotions) found that valence, arousal, and picture content all con-
tributed uniquely to emotional modulation of the acous-
Nociception Modulation
Facilitation

tic startle reflex (Bernat et al., 2006). Despite our

confounding between picture content and arousal,
exploratory multilevel analysis of subjective emotion
ratings suggested that a valence-by-arousal interaction
did significantly predict nociceptive reactions.
Inhibition

4.4. Implications
Arousal (Emotional Intensity)
This represents the third independent test of ECON
Low Arousal Moderate High Arousal
Arousal procedures. Nociception modulation has been reliably
elicited across all three samples, and picture content
Range of arousal elicited by
picture-viewing consistently explained large proportions of variance in
nociceptive outcomes. Thus, ECON appears to be a
Fig. 3. A hypothetical figure depicting the influence of emotional
valence and arousal on nociception. Negative emotions with low-to-
valid method of engaging supraspinal modulation.
moderate arousal facilitate nociception, with the degree of facilitation Although it is important that future studies determine
determined by arousal level. However, highly arousing negative whether individual differences (e.g., participant sex)
emotions inhibit nociception. In contrast, positive emotions always influence which pictures elicit greatest modulation, these
inhibit nociception, with the degree of inhibition determined by results suggest the greatest modulation is evoked by ero-
arousal level. Thus, an interaction of emotional valence and arousal
characterizes the emotion–pain relationship. Emotion-induction by
tic and attack pictures. Thus, researchers should be cau-
picture-viewing would not be expected to elicit intense, highly arousing tious when averaging reactions across pictures of similar
negative emotions indicative of active defense. Therefore, negative valence (but that vary in content/arousal) as this may
emotions would not be expected to inhibit nociceptive reactions in the attenuate the modulation observed.
current study. (This figure is adapted from Rhudy and Williams (2005),
reprinted by permission of the publisher; Rhudy and Williams: Gender
differences in pain: do emotions play a role? Gender Medicine;2:208– 4.5. Summary
26. Copyright 2005 by Excerpta Medica Inc.).
This study found emotional valence and arousal con-
tribute to modulation of nociceptive reactions (NFR,
It is noteworthy that modulation of subjective pain
subjective pain, SCR, and HR acceleration), with mod-
was a clinically small effect, despite being a statistically
ulation being associated with a valence-by-arousal inter-
large effect. Thus, emotional picture-viewing may prove
action. The combined influence of valence and arousal
better for studying nociceptive processing than for cop-
had a coordinating effect, with picture content explain-
ing with chronic pain (Seminowicz and Davis, 2007).
ing 52% of the variance in the four reactions. Together,
Before this conclusion can be drawn though, the effects
these data suggest the ECON paradigm is a reliable tool
of picture-viewing on clinical pain must be established
for studying supraspinal modulation.
(c.f., Montoya et al., 2005).
In the present study respiration rate was not measured.
Thus, picture-evoked breathing changes may have modu-
Acknowledgements
lated shock-evoked HR (respiratory sinus arrhythmia)
rather than pain modulatory circuits. However, given that
The first author would like to acknowledge the Henry
ANS reactions may contribute to pain perception (Chap-
Kendall College of Arts and Sciences and the Office of
man et al., 2002), respiration may represent another
Research at The University of Tulsa for partial financial
mechanism by which emotion can indirectly influence
support of this research through a faculty development
pain and nociceptive reactions. Nonetheless, our HR out-
summer fellowship. The authors thank Carl P. Latti-
comes should be interpreted with caution.
more III for help with data processing, as well as Emily
A final limitation is due to the method of manipulat-
Bartley and two anonymous reviewers for their com-
ing arousal. Given the arousal manipulation is con-
ments on earlier drafts of this paper.
founded with picture content, we are unable to
determine whether quantitative differences in activa-
tion/arousal are responsible for our results, or whether
References
something qualitatively different about attack and erotic
pictures enabled them to modulate nociception. Addi- APA. Diagnostic and statistical manual of mental disorders. 4th ed.
tional research is needed to manipulate arousal within Text Revision. Arlington, VA: American Psychiatric Association;
a picture content to provide additional support for the 2000.
 J.L. Rhudy et al. / Pain 136 (2008) 250–261
Bernat E, Patrick CJ, Benning SD, Tellegen A. Effects of picture
content and intensity on affective physiological response. Psycho-
physiology 2006;43:93.
Blumenthal TD, Cuthbert BN, Filion DL, Hackley S, Lipp OV, van
Boxtel A. Committee report: guidelines for human startle eyeblink
electromyographic studies. Psychophysiology 2005;42:1–15.
Bradley MM. Emotion and motivation. In: Cacioppo JT, Tassinary
LG, Berntson GG, editors. Handbook of psychophysiology
2000. New York: Cambridge University Press; 2000. p. 602–42.
Bradley MM, Codispoti M, Cuthbert BN, Lang PJ. Emotion and
motivation I: defensive and appetitive reactions in picture process-
ing. Emotion 2001;1:276–98.
Bradley MM, Cuthbert BN, Lang PJ. Pictures as prepulse: attention
and emotion in startle modification. Psychophysiology
1993;30:541–5.
Bradley MM, Lang PJ. Measuring emotion: The self-assessment
manikin and the semantic differential. J Behav Ther Exp Psychiatry
1994;25:49–59.
Brant AX, Rhudy JL, Williams AE. Perceived self-efficacy and
experimental pain: descending modulation is not engaged. Wash-
ington, DC: American Pain Society; 2007.
Chan CW, Dallaire M. Subjective pain sensation is linearly correlated
with the flexion reflex in man. Brain Res 1989;479:145–50.
Chan CW, Tsang HHY. A quantitative study of flexion reflex in man:
relevance to pain research. Adv Pain Res Ther 1985;9:361–9.
Chapman CR, Donaldson GW, Nakamura Y, Jacoson RC, Bradshaw
DH, Gavrin J. A psychophysiological causal model of pain report
validity. J Pain 2002;3:143–55.
Codispoti M, Bradley MM, Lang PJ. Affective reactions to briefly
presented pictures. Psychophysiology 2001;38:474–8.
Cohen J. Statistical power analysis for the behavioral sciences (rev.
ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.; 1977.
Cuthbert BN, Bradley MM, Lang PJ. Probing picture perception:
activation and emotion. Psychophysiology 1996;33:103–11.
Danziger N, Fournier E, Bouhassira D, Michaud D, De Broucker T,
Santarcangelo E, et al. Different strategies of modulation can be
operative during hypnotic analgesia: a neurophysiological study.
Pain 1998;75:85–92.
Dawson ME, Schell AM, Filion DL. The electrodermal system. In:
Cacioppo JT, Tassinary LG, Berntson GG, editors. Handbook of
psychophysiology 2000. Cambridge: Cambridge University Press;
2000. p. 200–23.
de Wied M, Verbaten MN. Affective pictures processing, attention,
and pain tolerance. Pain 2001;90:163–72.
Dowman R. Possible startle response contamination of the spinal
nociceptive withdrawal reflex. Pain 1992;49:187–97.
Fanselow MS. Neural organization of the defensive behavior system
responsible for fear. Psychonomic Bulletin & Review
1994;1:429–38.
France CR, France JL, al’Absi M, Ring C, McIntyre D. Catastro-
phizing is related to pain ratings, but not nociceptive flexion reflex
threshold. Pain 2002;99:459–63.
France CR, Suchowiecki S. Assessing supraspinal modulation of pain
perception in individuals at risk for hypertension. Psychophysiol-
ogy 2001;38:107–13.
Guieu R, Blin P, Pouget J, Serratrice G. High level sportsmen and
nociceptive flexion reflex of the lower limb. Can J Neurosci
1992;19:69–71.
Keefe FJ, Lumley M, Anderson T, Lynch T, Carson KL. Pain and
emotion: new research directions. J Clin Psychol 2001;57:587–607.
Kenntner-Mabiala R, Pauli P. Affective modulation of brain potentials
to painful and nonpainful stimuli. Psychophysiology
2005;42:559–67.
Komisaruk BR, Whipple B. Vaginal stimulation-produced analgesia in
rats and women. In: Kelly DD, editor. Stress-induced analgesia
1986;Vol. 467. New York, NY: Annals of the New York Acad-
emy of Sciences; 1986. p. 30–9.
261
Lang PJ. The emotion probe: studies of motivation and attention. Am
Psychol 1995;50:372–85.
Lang PJ, Bradley MM, Cuthbert BN. International affective picture
system (IAPS): instruction manual and affective ratings. Tech Rep
1999;4.
Lang PJ, Davis M. Emotion, motivation, and the brain: reflex
foundations in animal and human research. Prog Brain Res
2006;156:3–29. Chapter 1.
Lang PJ, Greenwald MK, Bradley MM, Hamm AO. Looking at
pictures: affective, facial, visceral, and behavioral reactions.
Psychophysiology 1993;30:261–73.
Meagher MW, Arnau RC, Rhudy JL. Pain and emotion: effects of
affective picture modulation. Psychosom Med 2001;63:79–90.
Miller SM. Predictability and human stress: toward a clarification of
evidence and theory. In: Berkowitz L, editor. Advances in
experimental social psychology 1981;Vol. 14. New York: Aca-
demic Press; 1981. p. 203–56.
Mineka S, Henderson RW. Controllability and predictability in
acquired motivation. Annu Rev Psychol 1985;36:495–529.
Montoya P, Sitges C, Garcia-Herrera M, Izquierdo R, Truyols M,
Blay N, et al. Abnormal affective modulation of somatosensory
brain processing among patients with fibromyalgia. Psychosom
Med 2005;67:957.
Rainville P, Bao QVH, Chretien P. Pain-related emotions modulate
experimental pain perception and autonomic responses. Pain
2005;118:306.
Rhudy JL, Maynard LJ, Russell JL. Does in-vivo catastrophizing
engage descending modulation of spinal nociception? J Pain
2007a;8:325–33.
Rhudy JL, McCabe KM, Williams AE. Affective modulation of
autonomic reactions to noxious stimulation. Int J Psychophysiol
2007b;63:105–9.
Rhudy JL, Meagher MW. Fear and anxiety: divergent effects on
human pain thresholds. Pain 2000;84:65–75.
Rhudy JL, Meagher MW. The role of emotion in pain modulation.
Curr Opin Psychiatry 2001;14:241–5.
Rhudy JL, Williams AE. Gender differences in pain: do emotions play
a role? Gender Med 2005;2:208–26.
Rhudy JL, Williams AE, McCabe K, Nguyen MA, Rambo P. Affective
modulation of nociception at spinal and supraspinal levels.
Psychophysiology 2005;42:579–87.
Rhudy JL, Williams AE, McCabe KM, Rambo PL, Russell JL.
Emotional modulation of spinal nociception and pain: the impact
of predictable noxious stimulation. Pain 2006;126:221–33.
Sandrini G, Milanov I, Willer JC, Alfonsi E, Moglia A, Nappi G.
Different effect of high doses of naloxone on spinal reflexes in
normal subjects and chronic paraplegic patients. Neurosci Lett
1999;261:5–8.
Sandrini G, Serrao M, Rossi P, Romaniello A, Cruccu G, Willer JC.
The lower limb flexion reflex in humans. Prog Neurobiol
2005;77:353–95.
Seminowicz DA, Davis KD. A re-examination of pain–cognition
interactions: implications for neuroimaging. Pain 2007;130:8–13.
Skljarevski V, Ramadan NM. The nociceptive flexion reflex in humans
– review article. Pain 2002;96:3–8.
Venables PH, Christie MJ. Electrodermal activity. In: Martin I,
Venebles PH, editors. Techniques in psychophysiology 1980. New
York: Wiley; 1980. p. 4–67.
Villemure C, Slotnick BM, Bushnell MC. Effects of odors on pain
perception: deciphering the roles of emotion and attention. Pain
2003;106:101–8.
Wiesenfeld-Hallin Z, Hallin RG, Persson A. Do large diameter
cutaneous afferents have a role in the transmission of nociceptive
messages? Brain Res 1984;311:375–9.
Wunsch A, Philippot P, Plaghki L. Affective associative learning
modifies the sensory perception of nociceptive stimuli without
participant’s awareness. Pain 2003;102:27–38.