ArticlePDF AvailableLiterature Review

Autonomic Nervous System Activity in Emotion: A Review


Abstract and Figures

Autonomic nervous system (ANS) activity is viewed as a major component of the emotion response in many recent theories of emotion. Positions on the degree of specificity of ANS activation in emotion, however, greatly diverge, ranging from undifferentiated arousal, over acknowledgment of strong response idiosyncrasies, to highly specific predictions of autonomic response patterns for certain emotions. A review of 134 publications that report experimental investigations of emotional effects on peripheral physiological responding in healthy individuals suggests considerable ANS response specificity in emotion when considering subtypes of distinct emotions. The importance of sound terminology of investigated affective states as well as of choice of physiological measures in assessing ANS reactivity is discussed.
Content may be subject to copyright.
Biological Psychology 84 (2010) 394–421
Contents lists available at ScienceDirect
Biological Psychology
journal homepage:
Autonomic nervous system activity in emotion: A review
Sylvia D. Kreibig
Department of Psychology, University of Geneva and Swiss Center for Affective Sciences, Geneva, Switzerland
article info
Article history:
Received 26 June 2009
Accepted 10 March 2010
Available online 4 April 2010
Autonomic nervous system
Emotional response specificity
Autonomic response organization
Cardiovascular system
Respiratory system
Electrodermal system
Autonomic nervous system (ANS) activity is viewed as a major component of the emotion response
in many recent theories of emotion. Positions on the degree of specificity of ANS activation in emo-
tion, however, greatly diverge, ranging from undifferentiated arousal, over acknowledgment of strong
response idiosyncrasies, to highly specific predictions of autonomic response patterns for certain emo-
tions. A review of 134 publications that report experimental investigations of emotional effects on
peripheral physiological responding in healthy individuals suggests considerable ANS response speci-
ficity in emotion when considering subtypes of distinct emotions. The importance of sound terminology
of investigated affective states as well as of choice of physiological measures in assessing ANS reactivity
is discussed.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction ......................................................................................................................................... 395
1.1. Current positions on autonomic responding in emotion ................................................................................... 395
1.2. Physical components of autonomic responding in emotion ................................................................................ 396
1.3. Conceptual levels of autonomic response organization in emotion ........................................................................ 396
2. Empirical findings of ANS activity in emotion ...................................................................................................... 397
2.1. The negative emotions ...................................................................................................................... 400
2.1.1. Anger ............................................................................................................................... 400
2.1.2. Anxiety ............................................................................................................................. 403
2.1.3. Disgust ............................................................................................................................. 403
2.1.4. Embarrassment .................................................................................................................... 404
2.1.5. Fear................................................................................................................................ . 404
2.1.6. Sadness............................................................................................................................. 405
2.2. The positive emotions ....................................................................................................................... 406
2.2.1. Affection ........................................................................................................................... 406
2.2.2. Amusement ........................................................................................................................ 406
2.2.3. Contentment ....................................................................................................................... 406
2.2.4. Happiness .......................................................................................................................... 406
2.2.5. Joy.................................................................................................................................. 407
2.2.6. Pleasure, anticipatory .............................................................................................................. 407
2.2.7. Pride................................................................................................................................ 407
2.2.8. Relief ............................................................................................................................... 408
2.3. Emotions without clear valence connotation ............................................................................................... 408
2.3.1. Surprise ............................................................................................................................ 408
2.3.2. Suspense ........................................................................................................................... 408
Present address: Department of Psychology, 450 Serra Mall, Bldg 420, Stanford, CA 94305, United States.
E-mail addresses:,
0301-0511/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 395
3. Discussion ........................................................................................................................................... 408
3.1. Autonomic responding in emotion .......................................................................................................... 408
3.1.1. Summary of empirical emotion effects and their relation to models of autonomic response organization ..................... 408
3.1.2. Measures of autonomic activation components .................................................................................. 409
3.1.3. Emotion terminology .............................................................................................................. 410
3.2. Boundary conditions......................................................................................................................... 410
3.2.1. Feeling changes without concomitant autonomic changes ....................................................................... 410
3.2.2. Autonomic changes without concomitant feeling changes ....................................................................... 410
3.2.3. Decoupling of subsystems in emotion............................................................................................. 411
Acknowledgements ................................................................................................................................ . 411
Appendix A. Overview of reviewed studies...................................................................................................... 411
References........................................................................................................................................... 416
1. Introduction
Autonomic responding in emotion has been an active research
topic since, almost a century ago, Walter Cannon (1915) first stud-
ied the physiology of emotion (Brown and Fee, 2002; Dale, 1947).
Still, there is no scientific consensus on whether there exists a
relation between emotion and the organization of autonomic ner-
vous system (ANS) activity and, if so, in what form. The various
positions, which contemporary researchers hold on this topic,
are first addressed in this article, before turning to the physical
components—or the hardware—of autonomic responding in emo-
tion. Next, a brief overview of the various theories and models
that have been suggested to explain and identify mechanisms of
autonomic response organization in emotion is given. The cen-
ter part of this article consists of a review of the empirical basis
for the postulate of emotion-specific ANS activity, considering 134
experimental studies on ANS activity in emotion. The next sec-
tion summarizes and discusses how empirical emotion effects
relate to models of autonomic response organization, points to the
importance of choosing adequate measures of autonomic activa-
tion components, and addresses the issue of emotion terminology.
A final section considers boundary conditions of the definition of
emotion employed in the present article and its implications for
identifying emotion-specific ANS activation.
1.1. Current positions on autonomic responding in emotion
Contemporary researchers in the field of emotion hold con-
trary positions on the topic of ANS activation in emotion. At one
extreme, Feldman-Barrett (2006, p. 41), for example, stated that “it
is not possible to confidently claim that there are kinds of emo-
tion with unique and invariant autonomic signatures,” but rather
that configurations follow general conditions of threat and chal-
lenge and positive versus negative affect. Feldman-Barrett named
three points of critique regarding the evidence for autonomic differ-
ences between emotions: first, the high heterogeneity of effects in
meta-analytical studies (e.g., Cacioppo et al., 2000) is interpreted
to suggest the presence of moderator variables in the relation of
emotion and ANS activity; second, autonomic differences that do
emerge between specific emotions are viewed to be along lines of
dimensional differentiation; and third, ANS activity is said to be
“mobilized in response to the metabolic demands associated with
actual behavior [...] or expected behavior” (p. 41) and because dif-
ferent behaviors have been shown neither to be emotion-specific
nor to be context-invariant (e.g., Lang et al., 1990), Feldman-
Barrett views emotion-specific autonomic patterns as a priori
An intermediate position is suggested by meta-analyses of phys-
iological responding in emotion (Cacioppo et al., 1997, 2000) that
report some degree of autonomic emotion specificity. Besides cer-
tain reliable differences between specific emotions, Cacioppo et
al. also noted context-specific effects of ANS activity in emo-
tion (i.e., according to different induction paradigms). Moreover,
valence-specific patterning was found to be more consistent than
emotion-specific patterning: negative emotions were associated
with stronger autonomic responses than positive emotions (cf.
Taylor, 1991). However, only one positive emotion, happiness,
which subsumed joy, was used in the meta-analysis. This unequal
representation of merely one positive as contrasted to a sample
of five negative emotions may significantly bias the kind of dis-
tinction discerned. Due to a limited number of studies considered,
a restricted range of physiological variables (only cardiovascular
and electrodermal, but no respiratory measures), and the univari-
ate nature of the meta-analytic approach, such results give only an
imperfect answer to the question of autonomic patterning in emo-
tion. Authors of review articles thus typically acknowledge that
discrete emotions may still differ in autonomic patterns even if
they do not differ in single variables (Larsen et al., 2008; Mauss
and Robinson, 2009).
Diametrically opposed to Feldman-Barrett’s (2006) position,
Stemmler (2009) argued why the ANS should not convey specific
activation patterns for emotions, if those have specific functions for
human adaptation. Stemmler (2004, 2009) reasoned that emotions
have distinct goals and therefore require differentiated autonomic
activity for body protection and behavior preparation. Autonomic
activity for behavior preparation is physiological activation that
occurs before any behavior has been initiated that itself engages
the ANS according to behavioral demands. Such autonomic activ-
ity has even been reported in experimentally paralyzed animals
(Bandler et al., 2000), underlining that it is not merely overt behav-
ior that causes this activity. This also resonates with Brener’s (1987)
notion of “preparation for energy mobilization,” which contrasts to
Obrist’s (1981) view of ANS activity as a component of the motoric
Stemmler (2004) reported on a meta-analysis on autonomic
responding in fear and anger—two emotions that are believed
to share similar valence and arousal characteristics—in which he
found considerable specificity between the two. Taking a functional
approach to autonomic responding in emotion, Stemmler (2003,
2004) stressed the importance of studying autonomic regulation
patterns in emotion rather than single response measures. Accord-
ing to the view that the central nervous system (CNS) is organized to
produce integrated responses rather than single, isolated changes
(Hilton, 1975), any variable which can be described or measured
independently is constituent of several such patterns. Only when
considering comprehensive arrays of physiological measures can
such regulation patterns be discerned. Stemmler (2009) stressed
that this should include variables that indicate both specific and
unspecific effects of emotion. Unspecific emotion effects distin-
guish between control and emotion conditions, but not between
emotions, whereas specific emotion effects distinguish between
emotions. The pool, from which indicators of independent auto-
nomic activation components can be drawn, is considered in the
subsequent section.
396 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
1.2. Physical components of autonomic responding in emotion
Although physiologists at the beginning of the last century char-
acterized the ANS as too slow and undifferentiated to quickly
produce highly organized response patterns in emotion (Cannon,
1927), contemporary physiologists see considerable room for such
organization (Bandler et al., 2000; Folkow, 2000; Jänig and Häbler,
2000; Jänig, 2003; see also Levenson, 1988). Research over the
past 50 years has invalidated the view that the sympathetic
devision of the ANS functions in an ‘all-or-none’ fashion with-
out distinction between different effector organs (Cannon, 1939).
Rather, each organ and tissue is innervated by distinct sympathetic
and parasympathetic pathways, with very little or no cross-talk
between them (Jänig and McLachlan, 1992b,a; Jänig and Häbler,
2000). Pools of sympathetic neurons can be selectively engaged,
such that individual systemic circuits or other effector units are
independently activated (Folkow, 2000).
The originally assumed functional unity of the sympatho-
adreno-medullary system is now known to consist of two
separately controlled system parts—a direct-nervous and an
adrenomedullary hormonal one—that under most situations have
different functional roles (Folkow, 2000). Whereas the former exe-
cutes precise, rapid, and often highly differentiated adjustments,
the latter independently modifies important metabolic functions.
In some emergency situations, where massive and generalized
sympatho-adrenal system activation can occur, the two parts may
also mutually support each other.
The inclusion of respiratory measures under autonomic mea-
sures also deserves some comment here. Respiratory activity
evidences effects of autonomic control as well as significant inde-
pendent contribution of peripheral and central chemoreceptors
sensitive to CO2(Wilhelm et al., 2005). Measures of respiratory
activity may thus yield additional information on ANS functioning
in emotion to that indicated by cardiovascular and electroder-
mal measures. There moreover exist important interactions of
the respiratory system with the cardiovascular system, as, for
example, attested by the phenomenon of respiratory sinus arrhyth-
mia (Grossman and Taylor, 2007). Here, respiratory measures
are important in the interpretation of effects of ANS functioning
indicated by cardiovascular measures, which are modulated by res-
piratory effects. Finally, the cardiorespiratory control system can
be viewed as one functional unit as it pursues the common aim
of providing the tissues with oxygen, nutrients, protective agents,
and a means of removing waste by-products (e.g., Feldman and
Ellenberger, 1988; Poon and Siniaia, 2000; Taylor et al., 1999). Thus,
comprehensive assessment of cardiovascular, electrodermal, and
respiratory measures can provide complementary information on
ANS functioning in emotion.
Central coordination of autonomic activity represents a corner-
stone of current views of integrated nervous system functioning (cf.
central autonomic network, CAN; Benarroch, 1993, 1999; see also
Damasio, 1998; Thayer and Lane, 2000). Unlike the original concep-
tualization of the ANS as functioning independently of the rest of
the nervous system (e.g., involuntary, automatic, and autonomous
control), close interactions between the central and autonomic
nervous systems exist in various ways. Thus, like the somatic ner-
vous system, the ANS is integrated at all levels of nervous activity.
Whereas segmental autonomic reflexes are coordinated by the
spinal cord, suprasegmental integration higher in the brain stem
is required for regulation of functions such as respiration, blood
pressure, swallowing, and pupillary movement. More complex
integrating systems in the hypothalamus influence the brain stem
autonomic subsystems. Many of the activities of the hypothala-
mus are, in turn, governed by certain cortical areas, particularly the
insular, anterior cingulate, and ventromedial prefrontal cortices as
well as the central nucleus of the amygdala, that process inputs
from the external environment. Thus, fundamental adjustments
of the organism to its environment can only be attained by the
concerted coordination and integration of somatic and autonomic
activities from the highest level of neurological activity in the cor-
tex down to the spinal cord and peripheral nervous system. This
high degree of specificity in ANS organization is needed for precise
neural regulation of homeostatic and protective body functioning
during different adaptive challenges in a continuously changing
environment. In this context, emotions may provide quick and reli-
able responses to recurrent life challenges. But still, the question
remains how autonomic response organization in emotion might
be achieved.
1.3. Conceptual levels of autonomic response organization in
William James is often credited for originating the idea of
peripheral physiological response specificity in emotion (e.g.,
Ellsworth, 1994; Fehr and Stern, 1970; see also Friedman, this
issue, for a historical overview). James’s (1884) proposal that the
feeling component of emotion derives from bodily sensations, i.e.,
the perceived pattern of somatovisceral activation, reversed the
causality of emotion and bodily responding. Acknowledging a high
degree of idiosyncrasy in emotion, James stated “that the symp-
toms of the same emotion vary from one man to another, and
yet [...] the emotion has them for its cause” (1894, p. 520). Even
more so, James believed that the physiological responses were
“almost infinitely numerous and subtle” (1884, p. 250), reflecting
the infinitely nuanced nature of emotional life. Still, James rec-
ognized limits to bodily variations in emotion: “the symptoms of
the angers and of the fears of different men still preserve enough
functional resemblance, to say the very least, in the midst of their
diversity to lead us to call them by identical names” (1894, p. 520,
emphasis in original). James thus strongly argued for “a deduc-
tive or generative principle” (James, 1890, p. 448) that may explain
physiological response specificity in emotion.
James’ claims associated with his peripheral perception theory
of emotion were met with differentiated reactions—they insti-
gated critique (most prominently the five-point rebuttal by Cannon,
1927), support (e.g., Angell, 1916), as well as various propositions
for general organizing principles of autonomic responding in emo-
tion. Although a number of different models have been proposed
since then, these co-exist in a rather disjunct fashion, without clear
empirical rejection of one or the other. As detailed in Kreibig (in
press), the various models of autonomic responding in emotion can
be organized by recognizing that these models address different
conceptual levels, on which an organizing principle of autonomic
responding in emotion may operate. Table 1 shows how the various
theories map onto different conceptual levels that span from the
physiological over the behavioral to the psychological level. A first
class of models is identified, which draw on a basic physiological
systems level; these are models that see the organizing principle of
autonomic responding in emotion in the structure and functioning
of the ANS or in the functioning of transmitter substances. A second
class of models is based on brain–behavior interactions and views
the organizing principle of autonomic responding in emotion in
the functioning of brain–behavioral systems and refined behavioral
modes. A third class of models centers on psychological processes
of meaning assessment and memory retrieval; these models place
particular emphasis on the functioning of psychological appraisal
modules and associative networks as a general organizing principle
of autonomic responding in emotion. A detailed discussion of the
various models on each level can be found in Kreibig (in press).
It is of note that from a component-view of emotion (Scherer,
2009), models on the same conceptual level rival each other. In
contrast, models on different levels have complementary value, as
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 397
Table 1
Conceptual levels of autonomic response organization in emotion (Kreibig, in press).
Psychological level
Functioning of appraisal modules
Componential process model Ellsworth (1994); Ellsworth and Scherer (2003); Scherer (1984, 1987, 2001, 2009)
Specific cardiovascular appraisal hypotheses Blascovich and Katkin (1993); Blascovich et al. (2003); Gendolla (2004); Gendolla and Wright (2005);
Wright (1996, 1998); Wright and Kirby (2001)
Functioning of Associative Networks
Bio-informational theory of emotional imagery Lang (1979, 1993); Miltner et al. (1986); Sartory (1993)
Brain–behavioral level
Functioning of brain–behavioral systems
Behavioral coping Obrist (1981); Schneiderman and McCabe (1989)
Dual-system models Bradley and Lang (2000); Cloninger (1987); Davidson (1998); Lang and Bradley (this issue); Lang et al.
Polyvagal theory Porges (1995); Porges et al. (1996); Porges (2001, 2007)
Reinforcement sensitivity theory Beauchaine (2001); Corr (2008); Fowles (1980); Gray (1982, 1987); Gray and McNaughton (2000)
Functioning of behavioral modes
Basic modes of defensive coping Folkow (2000); Stemmler (2009)
Modes of defensive coping and environmental demands Bandler and Shipley (1994); Bandler et al. (2000); Bernard and Bandler (1998); Keay and Bandler
(2001, 2002)
Predator imminence model Bradley and Lang (2000); Craske (1999); Fanselow (1994); Lang et al. (1997)
Peripheral physiological level
Functioning of autonomic systems
Undifferentiated sympathetic activation Cannon (1915, 1927)
Parasympathetic activation Kling (1933); Vingerhoets (1985); Vingerhoets et al. (2000)
Sympathetic versus parasympathetic response dominance Gellhorn (1964, 1965, 1970); Hess (1957)
Autonomic space Berntson et al. (1991)
Functioning of transmitter substances
Catecholamine hypothesis Ax (1953); Funkenstein et al. (1954)
Receptor-types hypothesis Stemmler (2003, 2004, 2009)
they address different levels of response organization (cf. Mausfeld,
2003). It will be seen in the discussion section how these models
fit with the empirical findings that are presented next.
2. Empirical findings of ANS activity in emotion
To what extent are postulated differences between emotion
reflected in empirical data on ANS functioning? To address this
question, a qualitative review of research findings was carried out,
focusing on effects of experimentally manipulated emotions on
ANS responding in healthy individuals. To cover both the psycho-
logical and medical literature, an exhaustive literature search using
the databases PsycINFO, PsycARTICLES, and PubMed was conducted
with the following search terms:
[emotion] and [autonomic nervous system or cardiovascular or
cardiac or heart or respiration or respiratory or electrodermal or
skin conductance]
References of such identified publications were additionally
screened for further research reports falling under the specified
criteria. Because the present review aimed at surveying the extent
to which autonomic effects of emotion are reported in research
studies, an inclusive approach was chosen, applying only basic
validity and reliability criteria to study selection. Publications were
included in the final selection if data from an original experi-
ment were reported, in which emotions were manipulated and
ANS measures were assessed during emotional responding. Emo-
tion, for this purpose, was broadly defined, covering definitions
of dimensional models of emotion (Bradley and Lang, 2000; Lang,
1994; Russell, 2003), discrete emotion theory (Ekman, 1999; Izard,
1992), as well as appraisal models of emotion (Scherer, 2001; Smith
and Kirby, 2004). Emotion was thus conceptualized as a multi-
component response to an emotionally potent antecedent event,
causing changes in subjective feeling quality, expressive behavior,
and physiological activation. Terms such as mood or affect were
considered synonymous with emotion, if the experimental manip-
ulation targeted a stimulus- or event-related change of subjective
feeling (see the concluding section for boundary conditions for such
a conceptualization of emotion). Experiments involving patient
groups and/or emotion regulation were excluded; control groups
of these studies were, however, included (i.e., healthy individuals
or unregulated responding, respectively). Publications were also
excluded if no specific emotion label was provided or if no specific
emotion contrasts were tested (e.g., if only reporting valence and/or
arousal contrasts or only coding according to positive/negative
affectivity). Publications were moreover excluded if not measuring
physiological activity during the period of emotional responding,
not reporting data from an original experiment, or not reporting
analyses pertinent to the present review (e.g., regression or pattern
classification were not considered). Articles were also excluded if,
instead of individual physiological variables, a composite score was
formed and only this measure was reported.
This literature search resulted in the identification of 134 pub-
lications. A detailed account of the studies included in the present
review can be found in Table A.1 (Appendix). To summarize this
information, tag clouds were created. A tag cloud is a visualiza-
tion of word frequency in a given database as a weighted list. For
the present purpose, coding labels in Table A.1 were used as tags
(drawn from individual columns). The absolute frequency of tag
occurrence is visualized with font size. Tag clouds were created
with the web application (
Fig. 1 presents an illustration of the relative number of studies that
investigated different emotions (Fig. 1a), using different kinds of
emotion induction paradigms (Fig. 1b), and quantified physiologi-
cal variables according to different averaging durations (Fig. 1c). It
can be seen from these illustrations, that the emotions most often
investigated are anger, fear, sadness, disgust, and happiness. Exper-
imental manipulations most often utilize film clips for emotion
induction, followed by personalized recall, real-life manipulations,
picture viewing, and standardized imagery. Response measures are
most often averaged over 60- or 30-s intervals; other common
averaging intervals are 1/2- or 10-s intervals and 120-, 180-, or
300-s intervals. It is noted that studies were coded for averaging
period because it was hypothesized that this factor might influ-
398 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
Fig. 1. Illustration of relative frequency of investigated emotions (a), emotion induction methods (b), and averaging duration for physiological variables. (c). Figures were
simplified by omitting low-frequency words.
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 399
ence the reported pattern of physiological responses. This effect
was, however, not observed in the present data and is thus not
further considered here.
Fig. 2 provides an illustration of the relative number of stud-
ies that have used different cardiovascular (Fig. 2a), respiratory
(Fig. 2b), or electrodermal (Fig. 2c) measures as well as their
overall usage (Fig. 2d). These figures show that heart rate is the
cardiovascular response variable most often reported; other pop-
ular cardiovascular measures include systolic and diastolic blood
pressure, heart rate variability, and finger temperature. For respi-
ratory measures, respiration rate is the most often reported index
together with respiratory period and respiratory depth as well as
tidal volume, duty cycle, and respiratory variability. For electro-
dermal measures, skin conductance level is the response variable
most often reported, followed by skin conductance response rate
and skin conductance response amplitude. Over all autonomic mea-
sures, heart rate is the indicator most often reported, followed by
skin conductance level and other cardiovascular variables.
Reports of physiological responses in emotions were coded
according to the emotion label provided by the authors and sub-
sequently grouped together based on synonymous expressions
drawn from Merriam-Webster Online Dictionary (2009). Thus, six
negative and eight positive emotion groups, and two emotion
groups without clear valence connotation were identified (labels
subsequently listed in parentheses were considered synonymous).
For the negative emotions, these were:
(a) anger (approach-oriented anger, withdrawal-oriented anger,
anger in defense of other, anger in self-defense, indignation);
(b) anxiety (dental anxiety, performance anxiety, agitation);
(c) disgust (disease-related disgust, food-related disgust);
(d) embarrassment (social anxiety, shame, social rejection);
(e) fear (threat);
(f) sadness (achievement failure, dejection, depression).
For the positive emotions, these were:
(a) affection (love, tenderness, sympathy);
(b) amusement (humor, mirth, happiness in response to slapstick
Fig. 2. Illustration of relative frequency of use of ANS measures as indicated by relative font size. Figures were simplified by omitting low-frequency words.
400 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
Fig. 2. (Continued ).
(c) contentment (pleasure, serenity, calmness, peacefulness, relax-
(d) happiness (except happiness in response to slapstick comedy);
(e) joy (elation);
(f) anticipatory pleasure (appetite, sexual arousal);
(g) pride;
(h) relief (safety).
For the emotions without clear valence connotation, these were:
(a) surprise (wonder);
(b) suspense.
Subsequent sections present a summary of findings of auto-
nomic emotion responses reported in studies described in Table A.1
(numbers in brackets refer to the study number in Table A.1). Direc-
tion of change in ANS activity was coded as change from baseline
or, if present, from a neutral comparison condition. Table 3 gives
abbreviations, full names, and near-synonymous expressions of
autonomic measures used in the following. It should be stressed
that the current review is of qualitative nature; thus, the results of
different studies were not integrated using a weighing procedure
that considers sample size, mean, and standard deviation, and thus
power of a study. Rather, to organize and integrate the different
findings reported in the various studies, a modal response pattern
was defined as the response direction reported by the majority
of studies (unweighted), with at least three studies indicating the
same response direction. Modal response patterns for each emotion
are summarized in Table 2.
2.1. The negative emotions
2.1.1. Anger
Physiological responding in anger-eliciting contexts of harass-
ment or personalized recall describe a modal response pattern of
reciprocal sympathetic activation and increased respiratory activ-
ity, particularly faster breathing.
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 401
Table 2
Overview of modal* ANS responses found for reviewed emotions.
Note. *Modal responses were defined as the response direction reported by the majority of studies (unweighted), with at least three studies indicating the same response direction.
Arrows indicate increased (), decreased (), or no change in activation from baseline (), or both increases and decreases between studies (↓↑). Arrows in parentheses indicate
tentative response direction, based on fewer than three studies. Abbreviations: pause – respiratory pause time; depth – respiratory depth; exp – respiratory expiration time; insp
– respiratory inspiration time; var – respiratory variability. For abbreviations of other physiological measures, see Table 3.
402 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
Table 3
Abbreviations, full names, and synonymous expressions of autonomic measures used in studies on emotion.
Abbreviation Full name Near-synonymous expression
Cardiovascular measures
CO Cardiac output Cardiac output * BSA (CI)
DBP Diastolic blood pressure
EPA Ear pulse amplitude
EPTT Ear pulse transit time
FPA Finger pulse amplitude
FPTT Finger pulse transit time
FT Finger temperature
HI Heather index
HR Heart rate 1/Interbeat interval (IBI)
HRV Heart rate variability
CVT Coefficient of temporal variability
HF High frequency spectral HRV (RSA)
LF Low frequency spectral HRV
LF/HF Low frequency/high frequency ratio
MF Mid-frequency spectral HRV
MSD Mean difference between successive RR intervals
MSSD Mean square of successive RR interval differences
pNN50 Percentage of successive normal sinus RR intervals >50 ms
RMSSD Root-mean-square of successive normal sinus RR interval differences
RSA Respiratory sinus arrhythmia
SDNN Standard deviation of the normal-to-normal intervals
SDSD Standard deviation of successive differences
VLF Very low frequency spectral HRV
HT Forehead temperature
LVET Left ventricular ejection time
MAP Mean arterial pressure
PEP Preejection period
PWA P-wave amplitude
SBF Skin blood flow
SBP Systolic blood pressure
SV Stroke volume Stroke index * BSA (SI)
TPR Total peripheral resistance
TWA T-wave amplitude
Respiratory measures
FRC Functional residual capacity
I/E ratio Inspiratory/expiratory ratio
HV Hyperventilation
pCO2End-tidal carbon dioxide partial pressure End-tidal fractional CO2concentration (FETCO2)
PePost-expiratory pause time
PiPost-inspiratory pause time
RC/VtPercentage of rib cage contribution to Vt
RD/Ttot Amount of respiratory work (depth divided by breath cycle duration)
Ros Oscillatory resistance
RR Respiration rate 1/Total respiratory cycle duration (Ttot)
SaO2Transcutaneous oxygen saturation
TeExpiratory time
TiInspiratory time
Ti/Ttot Inspiratory duty cycle
VeExpiratory volume
Ve/TeExpiratory flow rate or expiratory drive
ViInspiratory volume
Vi/TiInspiratory flow rate or inspiratory drive
VmMinute ventilation
VtTidal volume Respiration depth (RD), typically uncalibrated ribcage
measurements in arbitrary units
Vt/TiMean inspiratory flow rate
VtVTidal volume variability
Electrodermal measures
nSRR Nonspecific skin conductance response rate
OPD Ohmic Perturbation Duration index
SCL Skin conductance level
SCR Skin conductance response (amplitude, evoked)
SYDER SYDER skin potential forms
SRA Skin conductance response amplitude (spontaneous)
In particular, the anger response is characterized by - and
-adrenergically mediated cardiovascular effects: increased HR,
increased SBP and DBP, and increased TPR, accompanied either
by increased SV and CO [51, 104], decreased SV and increased
CO [88, real-life 111], decreased SV and unchanged CO [83, 89],
or decreased SV and CO (“anger out”, i.e., anger directed outward
away from the self) [40, 54]. Increased SBP, DBP, CO, and TPR, but no
increase in HR and SV (stressful interview) [2] as well as increased
HR, SBP, DBP, SV, CO, and unchanged TPR (personalized recall)
[106] have also been reported. Other studies, that did not assess all
indices, produce partial replications [7, 14, 25, 29, 35, 36, 37, 55, 63,
80, 87, 90, 96, 119, imagery-task 105, 107, 113, 123, 128, 130, 131,
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 403
134]. This response pattern is further characterized as an - and
-adrenergically mediated response by measures indicating short-
ened PEP [54, 81, 83, 87, 106, 111] and LVET [81, 106, 111], lower
TWA [110, 111], increased HI [81, 110, 111], and increased R–Z time
[110]. Moreover, decreased FPA [29, 110, 111, 123] or unchanged
FPA [75], and shortened FPTT [75, 111, 123], decreased HT [7, 104]
and FT [7, 98, 107], increased HT [109, 111], or unchanged FT [89]
point to vasoconstrictive effects in the periphery and local increases
of circulation in the face.
Cardiac parasympathetic inhibition is indicated by decreased
HRV (MSD [21]; spectral RSA [77]; RMSSD [87, 110]); others have
found unchanged HRV (peak-valley and spectral RSA, RMSSD, MSD,
SDNN [90]; SDNN [113]). Reports furthermore indicate increased
electrodermal activity (increased SCR [29]; increased nSRR [7, 65,
87, 111]; increased SCL [7, 21, 35, 37, 77, 80, 93, 98, 107, 109, 111,
115]), additionally implicating sympathetic effects at the eccrine
sweat glands, an effect which is cholinergically mediated.
For respiratory variables, findings indicate increased respiratory
activity, particularly faster breathing. Specifically, unchanged [14]
or increased RR [7, 34, 75, 80, 90, 93], shortened Tiand Te, increased
Pi[15], shortened Teand decreased I/E-ratio [80], increased [34],
unchanged [75], or decreased [15] respiratory depth, and increased
FRC, increased Ros [93], and increased variability of respiratory
amplitude [90] have been found.
Two exceptions to the modal response pattern of reciprocal
sympathetic activation in anger are noteworthy: first, responding
to material that features expressions of anger differs from respond-
ing to harassing material. Specifically, physiological responding to
picture viewing of facial emotional expressions of anger diverges
such that HR decelerates instead of an acceleration, SCL decreases
instead of an increase, and HRV (spectral RSA) increases instead
of a decrease or no change [28, 59, 129]. Because emotional
responses to anger expressions that signal threat have been related
to fear, this response pattern may be taken as suggestive of
a fear response rather than an anger response (see discussion
of fear responses associated with decreased HR, below). Sim-
ilarly, film viewing for anger elicitation differs in resulting in
decreased HR in the presence of decreased HRV (MSD [21]), point-
ing to sympathetic–parasympathetic cardiac deactivation that may
rather indicate passive sensory intake (Obrist, 1981; Schneiderman
and McCabe, 1989). Along these lines, Stemmler and colleagues
(2007) demonstrated that approach-oriented anger was char-
acterized by unchanged HR, while withdrawal-oriented anger
showed decreased HR [110]. This finding may point to the fact
that motivational direction influences the heart rate response in
A second deviation from the modal response pattern in anger
is evident in the absence of -adrenergic vasoconstrictive effects
in several studies: directed facial action (DFA) of anger is charac-
terized by increased, instead of decreased, FT [32, 74, 75] (although
decreased FT has also been reported for anger in DFA [73]), an effect
that reflects -adrenergically mediated vasodilation in contrast to
-adrenergically mediated vasoconstriction (Cohen and Coffman,
1981; Rowell, 1986). TPR decreased in association with increased
HR, LVET, SV, CO, HI, SBP, DBP, and MAP and shortened PEP in a film
study of anger [81]. Similarly, a response pattern labeled “anger in”
(i.e., anger directed toward the self) is characterized by increased
HR, SV, and CO, unchanged SBP and DBP, and decreased TPR [2,
40]. Increased HR, SBP, DBP, SV, CO, and forearm blood flow, but
decreased levels of TPR have also been reported under conditions
of experimenter harassment in accompaniment of a friend [71].
Finally, increased HR and SBP, but decreased DBP and MAP was
found in the context of emotional step walking [105]. These find-
ings suggest that various subforms of anger may exist, which are
differentiated by motivational direction that appears to influence
the heart rate and -adrenergic response.
2.1.2. Anxiety
Using predominantly experimental paradigms that incorpo-
rate an anticipatory component (e.g., threat of shock [12, 13,
17, 20, 127]; speech preparation [82, 118]), anxiety has been
almost unanimously characterized by sympathetic activation and
vagal deactivation, a pattern of reciprocal inhibition, together with
faster and shallower breathing. Apparent overlaps with the above-
reviewed anger response on certain response variables will have to
be addressed in future research that will need to fill gaps of mea-
sures that are either predominantly assessed in anger research (e.g.,
- and -adrenergically affected measures of sympathetic func-
tioning, such as PEP, LVET, MAP, and TPR) or in anxiety research
(e.g., respiratory measures of sighing or carbon-dioxide blood lev-
In particular, reports on anxiety indicate increased HR [2, 31,
82, 97, 118, 121], decreased HRV (spectral RSA [82]; peak-valley
RSA [84]) as well as increased LF and LF/HF [82], increased SBP
[2, 118], increased DBP [84, 118] or unchanged DBP and TPR [2],
unchanged SV [2, 84] and increased CO [2], decreased FPA [12, 13,
118] as well as decreased FPTT and EPTT [118], decreased FT [91,
97], and increased HT [91]. Reports include moreover increased
electrodermal activity (increased SCR and nSRR [12] and increased
SCL [12, 20, 82, 93]). Respiratory variables indicate increased RR
due to decreased Tiand Te[12, 30, 84, 121], as well as decreased Vt
[12, 121], increased sigh frequency and Vtvariability [12] (however,
higher sigh frequency during relief than tension has also been found
[127]), increased Ros [93], decreased end-tidal pCO2[30, 121], and
increased oxygen consumption [30].
A striking exception to this otherwise classic pattern of recip-
rocal sympathetic activation and parasympathetic deactivation for
anxiety constitutes a study of picture viewing (e.g., pictures of a
snake, shark, tornado, knife, or attack [94]): this study reports HR
deceleration, accompanied by increased HRV (peak-valley RSA),
and a trend of increased Ttot associated with increased Teand
decreased Ti, decreased Vm, and an unspecific small increase in
Ros. Thus, this study suggests a pattern of reciprocal parasympa-
thetic activation and decreased respiratory activity for anxiety.
Other exceptions that do not fully support a pattern of reciprocal
sympathetic activation for anxiety include results from a threat-
of-shock context, where unchanged HR [13] or decreased HR and
increased SCR [17] has been reported. HR deceleration, accompa-
nied by increased PEP and LVET, has also been found in the context
of music-induced agitation [84]. All these response patterns point
to response fractionation across organ systems (Lacey, 1967).
2.1.3. Disgust
Disgust-related autonomic responding falls into two partially
overlapping patterns: (a) disgust elicited in relation to contam-
ination and pollution (e.g., pictures of dirty toilets, cockroaches,
maggots on food, foul smells, facial expressions of expelling food),
characterized by sympathetic–parasympathetic co-activation and
faster breathing, particularly decreased inspiration (cf. physio-
logical response associated with vomiting; Sherwood, 2008); (b)
disgust elicited in relation to mutilation, injury, and blood (e.g.,
injections, mutilation scenes, bloody injuries), characterized by
a pattern of sympathetic cardiac deactivation, increased electro-
dermal activity, unchanged vagal activation, and faster breathing.
Increased HRV sets contamination-related disgust apart from most
other negative emotions, which typically show decreased HRV.
Similarly, decreased CO distinguishes disgust in general from the
other negative emotions, which show increased CO, as is typical for
mobilization for action (Obrist, 1981).
Specifically, contamination-related disgust is associated with
HR acceleration [3, 14, 49, 73, 128] or no change from baseline
[32, 74, 75, 99]. HR acceleration has also been reported in the con-
text of personalized recall [73, 89] or films [63] where disgust-type
404 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
remained unspecified. This response pattern is furthermore char-
acterized by increased HRV (SDNN [63], RMSSD [99], peak-valley
RSA [94]), increased TPR, and decreased SV [89, 99], suggesting
sympathetic–parasympathetic coactivation. As a notable excep-
tion, unchanged or even decreased skin conductance has been
reported in response to contamination pictures [22] and no change
in nSRR has been reported in response to film clips depicting
contamination-related material [66].
Mutilation-related disgust, on the other hand, was character-
ized by HR deceleration [9, 18, 21, 23, 44, 46, 62, 85, 99, 108, 133]
or a depressed phasic HR response [70]. Palomba et al. (2000) note
that HR reduction occurred between the first and the last interval
of a 132-s film, indicating a slow late deceleration [85]. Similarly, in
response to picture viewing, Winton et al. (1984) describe a tripha-
sic response pattern of HR change that was characterized by an
early deceleration, a brief and dampened acceleration, followed by
an early onset of a second deceleration [133]. This response pattern
is furthermore characterized by no change in HRV (RMSSD [99];
peak-valley RSA [85]; Porges’ RSA [9]; however, increased HRV
(spectral RSA) and decreased LF/HF has also been reported [108]).
Increased TWA [85] and no change in SV and TPR [99] have also
been found for mutilation-related disgust, suggesting decreased
cardiac and increased electrodermal sympathetic control together
with unchanged vagal influence (increased SCR for mutilation- ver-
sus contamination-related disgust has also been reported [16, 22]).
Still, one study [22] reported non-differential HR deceleration for
both contamination and mutilation pictures that was largest com-
pared to all other affective categories.
Both response patterns, i.e., mutilation- and contamination-
related disgust, were non-differentially accompanied by increased
SBP, DBP, and MAP [21, 69, 89] or no change in blood pressure
[99], decreased PEP, LVET, CO [99], or no effect on CO and FT [89],
increased FT [32, 74, 75], decreased FT [24, 44, 46, 73], decreased
FPA [44, 46, 69, 75], increased FPTT [46, 69], and decreased FPTT
[75], no change in EPTT [46], and decreased facial blood flow and
velocity [108]. Responses in these variables do not seem to fall into
a coherent pattern.
Across paradigms (e.g., picture viewing, film clips, DFA, and
personalized recall), disgust is consistently reported to be non-
differentially associated with increased electrodermal activity, as
indicated by increased SCR [18, 60, 62, 70, 133], increased nSRR
[60, 65, 108], and increased SCL [21, 23, 26, 32, 44, 46, 49, 69, 74, 75,
85, 99, 108, 115, 126]. Electrodermal activity is furthermore char-
acterized by long-duration SCR [3] in response to disgust-eliciting
odorants, whereas picture viewing of disgust-expressing faces has
been reported to elicit relatively short OPD, small SCR, positive skin
potentials of rapid increase and slow decrease [24] or a delayed SCR
of medium response size and slow rise time [132].
There is a general effect of increased RR in disgust [15, 24, 34,
46, 69, 75, 85], although increased respiratory duration [94] or no
change [108] have also been reported. Notably, contamination-
related disgust has been characterized by decreased Tiand
increased or no change in Te[14, 15, 94], that may contribute to
decreased Ti/Ttot and Vt/Ti[15], decreased respiratory volume (e.g.,
decreased Vt,Vm[14, vomiting clip 15, 24, 75, 94]), and increased Ros
[94], as well as larger variability in Te,Vt,Vm,Vt/Ti[vomiting clip 15].
Other than decreased Vt[69] for mutilation-related disgust, gener-
ally no change in respiratory timing [9, torture clip 15] or volume
parameters [46] is reported (see also [34]). In summary, the dis-
tinction between contamination versus injury disgust appears to be
important in determining the specific type of disgust response and
will need to be more systematically investigated in future research.
2.1.4. Embarrassment
Inducing embarrassment by experimenter humiliation, watch-
ing a video of oneself singing, or imagery, studies consistently
indicate broad sympathetic activation and vagal withdrawal, a pat-
tern of reciprocal inhibition. Whereas this response pattern largely
overlaps with those of anger and anxiety reviewed above, the rel-
atively small number of studies as well as the limited number of
response variables assessed highlights the importance for future
research to test specific physiological differences between negative
emotions, such as facial blushing in embarrassment.
Studies inducing embarrassment in particular report increased
HR [4, 52, 54, 56], accompanied by decreased PEP, no effect on CO,
and increased TPR [54], increased SBP and DBP [52], decreased HRV
(peak-valley RSA), and increased SCL [56]. Harris (2001) reports
that HR rose significantly during the first minute of watching an
embarrassing film of oneself singing, but returned to baseline levels
during the second minute, a pattern that replicated in a second
study [52]. As the empirical basis for the physiological response
pattern of embarrassment is scant, much remains to be done in
future research.
2.1.5. Fear
Laboratory fear inductions typically use presentation of threat-
ening pictures, film clips, or music, standardized imagery or
personalized recall, and real-life manipulations (e.g., imminent
threat of electric short circuit). One of the earliest attempts to
induce fear in the laboratory, used a sudden backward-tilting chair
[11]. Due to the nature of the manipulation, it is, however, not clear
whether in fact fear, or rather surprise, was induced. Moreover,
because confounds caused by the change in body posture compli-
cate interpretation of results, this study is not considered here.
Overall, studies on fear point to broad sympathetic activation,
including cardiac acceleration, increased myocardial contractility,
vasoconstriction, and increased electrodermal activity. In distinc-
tion to the physiological response to anger, peripheral resistance
typically decreased in fear, whereas it increased in anger. This
response is accompanied by decreased cardiac vagal influence and
increased respiratory activity, particularly faster breathing based
on decreased expiratory time, resulting in decreased carbon dioxide
blood levels.
Various of the studies investigating fear report increased HR [5,
8] or increased electrodermal activity in single measures (increased
SCR [24]; increased nSRR [65]; increased SCL [132]) or in co-
assessment (nSRR [35, 114]; SCL [48, 74, 79, 80, 114]; although
increased HR and unchanged SCL [124] and unchanged nSRR [66]
have also been reported), indicating a general arousal response.
More complete patterned responses are derived from studies
that assessed combinations of cardiovascular and/or cardiorespira-
tory parameters. A number of studies report increased HR together
with indicators of increased vasoconstriction: decreased FT [32, 64,
73, 89, 107, 109] (see, however [74] for a report of increased FT);
decreased FPA [67, 75, 109, 110, 111]; decreased FPTT [75, 111];
and decreased EPTT [64] (see, however [67] for a report of increased
FPTT and EPTT). Increased HR and increased blood pressure have
also been variously reported: increased SBP and DBP [7, 64, 81,
87, 89, 96, 104, 107, 111], as well as increased MAP [21, 67, 72,
imagery 105, 130]; some have reported unchanged DBP [exercise
105, 106, 119] and decreased MAP [exercise 105]. Reports on vascu-
lar resistance indicate either increased TPR [81, 89] or, more often,
decreased TPR [87, 104, 106, 111]. Furthermore, HR increase co-
occurs with increased myocardial contractility: increased ejection
speed [111], shortened PEP [64, 81, 87, 106, 110, 111], decreased
[106, 110, 111] or unchanged LVET [81], and increased HI [110, 111]
(however, see [81] for a report of decreased HI). These are associ-
ated with consequent changes in cardiac pump function: increased
[7, 104] or decreased SV [64, 81, 89, 106, 110, 111], and increased
[104, 106, 111], unchanged [89], or decreased CO [81] have been
reported. Increased sympathetic cardiac control is furthermore
indicated by increased PWA and decreased TWA [85, 110, real-life
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 405
111]. Vagal withdrawal is evidenced in decreased HRV (MSD [21];
MSSD [111]; RMSSD [42, 43, 87, 110]; peak-valley RSA [90]; spectral
RSA [90, 126]), although some report unchanged HRV (peak-valley
RSA [67, 85]; spectral RSA [64]) and unchanged LF [126].
Further studies report HR increases together with increased
respiratory activity, including measures of breathing rhythm:
increased RR [7, 34, 64, 67, 75, 80, 85, 86, 90, 111, 122], and
either both decreased Tiand Te[14, 120, 122], or predominantly
decreased Teand unchanged Ti, as also indicated by increased Ti/Ttot
and I/E-ratio [33, 34, 64], and increased Pi[14]. Volumetric mea-
sures moreover indicate increased respiratory volume [34, 75] or
decreased respiratory volume [14, 67, 120, 122], and increased Vm
[64]. Gas exchange analysis indicates decreased pCO2[64, 120, 122].
Furthermore, increased variability of respiratory parameters has
been noted, such as increased variability of respiratory amplitude
[90] or increased variability in pCO2and Vi/Ti[121].
The already above-mentioned increase in electrodermal activ-
ity was also found in numerous of these multi-measure studies
(increased SCR [7, 111]; increased nSRR [87, 111, 119]; increased
SCL [21, 42, 43, 64, 80, 85, 107, 111, 126, 130]).
Only a few studies report HR deceleration in the context of lab-
oratory fear elicitation: decreased HR along with signs of increased
vasoconstriction (decreased FPA and EPTT, unchanged FPTT) has
been found in response to a film clip eliciting fear of falling [39];
decreased HR and unchanged HRV (SDNN) has been reported in
children watching a film clip that portrayed Snow White running
through a dark haunted forest [113]; decreased HR and increased
SCR was reported in response to picture presentation of snakes
and spiders [27] or other threatening material (e.g., angry face,
aimed gun, attack [10, 22]); decreased HR, decreased SCL, and
increased HT has been found in a real-life induction context (radio
play, announcement of uncontrollable event, and sudden outage
of light [109]); decreased SCL has been similarly found for fear
induced by music excerpts [67]. It is possible that these latter fear
paradigms elicited a stronger degree of self-involvement, leading to
higher imminence of threat (Bradley and Lang, 2000; Craske, 1999;
Fanselow, 1994; Lang et al., 1997), such that participants were fur-
ther along the “fear continuum,” characterized by immobilization
rather than an active coping response that leads to sympathetic
inhibition (see also the above discussion of outliers for anger and
anxiety). However, such findings will need to be contrasted with
such intense fear responses as found, for example, in phobias, which
constitute a good model to study the type of fear with high imme-
diate threat characteristics (e.g., Wilhelm and Roth, 1998).
2.1.6. Sadness
Inspecting the activation components reported for sadness
reveals a heterogeneous pattern of sympathetic–parasympathetic
coactivation. Only a few studies considered mediating vari-
ables, such as cry-status [45, 102, 103]. These studies associate
uncoupled sympathetic activation with crying sadness, whereas
sympathetic–parasympathetic withdrawal appears to be charac-
teristic of non-crying sadness.
Parsing reports of physiological response patterns of sadness
that were not analyzed according to cry-status suggests two
broad classes of physiological activity in sadness—an activating
response and a deactivating response. The activating sadness
response, which partially overlaps with the physiological response
of crying sadness, is characterized by increased cardiovascular sym-
pathetic control and changed respiratory activity, predominantly
reported in studies using DFA, personalized recall, and some stud-
ies using film material. On the other hand, the deactivating sadness
response, which partially overlaps with the physiological response
of non-crying sadness, is characterized by sympathetic withdrawal,
reported in the majority of studies using film material, as well
as music excerpts, and standardized imagery. A distinct charac-
teristic of deactivating/non-crying sadness to all other negative
emotions is the decrease in electrodermal activity. In contrast, the
activating/crying sadness response largely overlaps with that of, for
example, anxiety—a point that will be returned to below.
Specifically, for participants who cried in response to a sadness-
inducing film clip, studies unanimously report increased HR,
associated with increased SCL, decreased FPA, FT, smaller increases
in RR, and non-differentially increased RD [45], increased nSRR and
unchanged SCL [101], or increased RR, unchanged HRV (spectral
RSA) and Vt[103]. In contrast, sad participants who did not cry
while watching the film clip, exhibited decreased HR, associated
with decreased electrodermal activity (decreased SCL and smaller
nSRR [45, 101]), increased respiratory activity (increased RR and
RD [45, 103]), increased [103] or decreased respiratory depth [45],
decreased HRV (spectral RSA [103]), and decreased FPA and FT [45].
With respect to the activating response in sadness, which par-
tially overlaps with the physiological response of crying sadness,
DFA has been found to consistently prompt increased HR [14, 32,
73, 74, 75]. In some studies, shortened FPTT and increased FPA [75],
increased SCL [74], increased [73], unchanged [74], or decreased
FT [32], and increased RR and respiratory depth [75] or decreased
RR, Ti,Te, and Vt, and increased Piand FRC [14] is reported. Sim-
ilarly, sadness elicited by personalized recall is characterized by
increased HR associated with increased [32, 98, 115] or unchanged
[77] SCL as well as increased SBP, DBP, and TPR [51, 83, 89, 106],
unchanged [51, 89] or decreased SV [83, 106], increased [51] or
unchanged CO [83, 89, 106], and increased [83] or decreased PEP
and LVET [106]. FT has been reported to remain unchanged [89]
or to decrease [98]. For HRV, decreases (MSD, SDNN [90]; spectral
RSA [77]), no change (spectral RSA [98]; a correlation of increased
HRV with increased sadness intensity is, however, also reported), or
increases (peak-valley RSA [90]) were found. Respiration was char-
acterized by increased respiration period and increased variability
in respiration period [90]. Only small increases in HR and SBP and
unchanged DBP have also been reported [105]. Allen et al. (1996),
examining social rejection and achievement failure, characterize
the emotion they investigated as high-arousal sadness and report
increased HR [4]. Some studies using films for sadness induction
report increased HR [63, 68, 76], increased electrodermal activity
(nSRR [65, 100]; SCL [68, 93, 117, 126]; although no effect on nSRR
has also been reported [66]), and increased RR [68, 100], associated
with decreased FPA, FPTT, FT [68] or unchanged HR and FPTT [100],
decreased HRV (spectral RSA) and unchanged LF [126] or increased
HRV (SDNN [63]), unchanged SBP, DBP [76], and increased Ros [93].
The activating response contrasts with a deactivating sadness
response, which partially overlaps with the physiological response
to non-crying sadness. This response pattern is found in the large
majority of studies using film clips for sadness induction, which
report a pattern of decreased cardiac activation and decreased elec-
trodermal activity: decreased HR [6, 18, 21, 31, 47, 49, 64, 86, 113,
114, 116] (although see [39] for report of unchanged HR), longer PEP
[64, 78], increased HRV (MSD [21]; spectral RSA [78]) or unchanged
HRV (SDNN [113]; RMSSD [49]), unchanged [64] or decreased DBP
and MAP [21], increased EPTT and FPTT, associated with decreased
EPA, FPA, and FT [39, 64], decreased electrodermal activity (SCR
[18]; SCL [21, 47, 78, 112, 116]; however, see [114] for increased SCL
and unchanged nSRR, and [64] for increased nSRR). Some studies
report decreased respiratory activity [47], as indicated by decreased
RR and increased pCO2[64], while others report increased RR [86,
116]. Averill (1969) also reported decreased HR and SCL, however,
together with increased SBP, DBP, FPA, unchanged FT, increased
nSRR, and unchanged RR and respiratory irregularity, as elicited by
a film clip on the aftermath of the assassination of John F. Kennedy
[6], showing the funeral and burial of the US President—material
that might have elicited nostalgia or mixed emotions of both sad-
ness and anger.
406 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
Such cardiovascular deactivation has also been found in an
exercise paradigm for emotion induction [exercise 105], in which
sadness was the only emotion that evidenced decreases in HR,
SBP, DBP, and MAP compared to a neutral comparison condition.
Music-induced sadness is similarly reported to be characterized
by decreased HR associated with decreased RR and increased Te
[33], decreased RR and RD [67], unchanged [61] or increased RR,
associated with decreased Te,Ti, and Pe[84].
Sadness elicited in the context of standardized imagery is sim-
ilarly reported to be characterized by decreased HR [41, 122] or
only small HR increases [124], unchanged SCL [41, 124], increased
Tiand Te, resulting in decreased RR, and increased pCO2[120, 122].
Another study [30] also reports of decreased ventilation, decreased
oxygen consumption, and increased pCO2in the context of hypno-
sis, as well as decreased [120] or unchanged Vt[122]. Increased HR
and decreased nSRR has also been reported [35]. Similarly, in an
emotion self-generation task, unchanged HR and decreased SCL for
sadness has been reported [55].
Picture viewing for sadness induction has been reported to lead
to increased HR and Ros, unchanged HRV (peak-valley RSA) and
ventilation (depressing picture content, such as hospital patients,
scenes of catastrophe, soldiers in action, or dead animals [92]),
decreased HR, Ti,Vt, increased Ttot ,Te, HRV (peak-valley RSA),
unchanged Ros and SCR (depressing picture content, such as ceme-
tery, plane crash, war victim, or a duck in oil [94]), or moderately
increased RR, decreased FT, smallest SCR, and positive SP (pictures
of sad facial expressions [24]).
Contrasting contents related to the activating and deactivating
sadness responses suggests a differentiation according to immi-
nence of loss, with the activating pattern occurring predominantly
in response to film clips that depict scenes related to impending
loss, such as individuals coping with cancer or Alzheimer’s, a hus-
band waiting for the result of his wife’s operation, or a man talking
to his dying sister (cf. helplessness; Seligman, 1975). On the other
hand, the deactivating pattern occurs predominantly in response
to film clips that depict scenes related to a loss that has occurred,
such as a mother at her daughter’s funeral, a young boy crying over
his father’s death, or the death of Bambi’s mother. It may be that
such distinctions as anticipatory sadness (i.e., worry or anticipa-
tion of loss) as contrasted to acute sadness in the experience of
loss or grieving in the aftermath of a loss play a role in addition
to cry-status in differentiating physiological responses in sadness
(Barr-Zisowitz, 2000; Kreibig, 2004). To allow a clearer picture of
the type of autonomic activation associated with sadness, it will be
important for future research to consider cry-status in analyzing
physiological responses. Moreover, care should be taken to distin-
guish between anticipatory and acute sadness.
2.2. The positive emotions
2.2.1. Affection
Love, tenderness, or sympathy evoked by film clips [15, 31]
or personalized recall [115], have been reported to be associ-
ated with decreased HR (similar to sadness [31]), an unspecific
increase in SCL [115], and increased Ti/Ttot , increased variation in
Te, and decreased variation in Vt,Vm, and Vt/Ti[15]. Because of
the few studies that have investigated physiological responding in
affection-related emotions, no conclusive statement on the type of
response pattern can be made.
2.2.2. Amusement
Laboratory elicitation of amusement has almost exclusively
employed film clips; only two studies used alternative paradigms
(picture viewing [62] or personalized recall [37]; see also [38]).
Although all film clips depicted comedic material, several response
components emerge. Overall, response variables point to increased
cardiac vagal control, vascular -adrenergic, respiratory, and elec-
trodermal activity, together with sympathetic cardiac -adrenergic
deactivation in amusement.
HR is the most variable response component, with reports of
deceleration [18, 21, 26, 58, 62, 112], no change [47, 50, 53, 57],
or acceleration [6, 37, 63, 116]. More consistently, increased HRV
(SDNN [63]; MSD [21]; spectral RSA [26]), unchanged LF/HF [26],
and increased PEP and decreased CO [53] are reported. Blood
pressure remains unchanged (SBP [6, 53]; DBP [6]; MAP [50]) or
increases (SBP, DBP, MAP [21]). Increased vasoconstriction is indi-
cated by decreased FPA, FPTT, EPTT, and FT [47, 50]; increased
TPR [53], and decreased FPA and unchanged FT [6] have also
been reported. Respiratory activity is increased, as evidenced in
increased RR [6, 47, 57, 93, 116], increased RD [47], increased
respiratory irregularity [6], increased Ros [93], decreased Ti,Vt,
Ti/Ttot , and increased Piand variability of Te,Vt,Vm, and Vt/Ti[15].
Increased electrodermal activity is shown in increased SCR [18],
increased nSRR [6, 57, 65], and increased SCL [37, 47, 57, 62, 93,
116, 117]; still, some have reported unchanged SRA [50] and nSRR
[66] or even decreased electrodermal activity (SCL and nSRR [6, 58,
2.2.3. Contentment
Studies on psychophysiological effects of contentment or plea-
sure have particularly relied on film clips displaying nature scenes
[21, 85, 94], standardized imagery (e.g., wood fire, book reading,
soft music [83, 120, 122, 128]) or personalized recall [25, 105].
Taken together, decreased cardiovascular, respiratory, and electro-
dermal activation is suggestive of decreased -, -adrenergically,
and cholinergically mediated sympathetic activation and mild car-
diac vagal activation. Compared to the physiological response to
amusement, the physiological response to contentment appears to
have a stronger sympathetically deactivating component, whereas
both share cardiac vagal activation. Further studies are, however,
needed to clarify the exact nature of autonomic and respiratory
activity in contentment.
Studies on the physiological response of contentment indicate
HR deceleration [21, 55, 84, 85, 94, 105, 122] or unchanged HR
[25, 79], increased TWA, unchanged HRV (peak-valley RSA), and
increased RR [85], or decreased HRV (MSD [21]), decreased SBP,
DBP, MAP [21, 105], and decreased SCL [21, 55, 85] or unchanged
SCL [79]. Decreased RR has been reported together with increased
HRV (peak-valley RSA [94]), increased Ti,Te[94, 120] or unchanged
Tiand Te[122], decreased Vt[94, 122] or increased Vt[120], and
increased pCO2[120, 122] as well as unchanged Ros, SCR, and Vm
[94]. Unchanged I/E ratio and moderately increased respiratory
work, depth, and rate has also been reported [34]. Using music
excerpts for emotion induction [84], increased LVET and unchanged
FPA, together with increased RR, and decreased HRV (peak-valley
RSA), Ti,Te, and Pihas been found. Moderate increases in HR, SBP,
DBP, PEP, TPR, unchanged CO, and decreased SV has been reported
for relaxation imagery [83]. As this overview shows, the physio-
logical response pattern of contentment is similar to a relaxation
response. Still, inconsistencies of the response pattern noted by
various studies will have to be addressed in future research.
2.2.4. Happiness
Happiness has been induced with various emotion elicitation
paradigms, including DFA [14, 73, 74, 75], personalized recall
[77, 89, 90, 105, 115], standardized imagery [41], film clips [100,
113, 126], music [33, 61, 67, 84], or pictures [59]. The autonomic
response pattern of happiness is characterized by increased cardiac
activity due to vagal withdrawal, vasodilation, increased electro-
dermal activity, and increased respiratory activity. This response
pattern points to a differentiated sympathetic activation state of
decreased - and -adrenergically mediated influences, while at
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 407
the same time cholinergically-mediated effects are increased. Hap-
piness shares with various negative emotions a central cardiac
activation component due to vagal withdrawal, whereas it is dis-
tinguished from these by peripheral vasodilation.
In particular, the physiological response to happiness includes
increased HR [14, 41, 55, 59, 61, 74, 75, 77, 79, 89, 90, 92, 105,
113] or unchanged HR [33, 84, recall visualizing 131] (although
decreased HR has been reported in [67]), unchanged HRV (SDNN
[113]; peak-valley RSA [92]) or decreased HRV (spectral RSA [59,
77, 90, 126]; peak-valley [67, 84, 90]), and unchanged LF [126]. Fur-
thermore, reports indicate increased blood pressure (increased SBP,
DBP, MAP [67, imagery 105]; increased SBP, DBP [61, 89, recall visu-
alizing 131]; increased SBP, decreased DBP, MAP [exercise 105];
unchanged SBP and DBP [84]). Increased PEP and unchanged LVET
and SV has been furthermore found [84]. Increased TPR, decreased
SV, and unchanged CO have also been reported [89]. Vasodilation is
moreover reported, including increased FT [74, 75, 109] (however,
unchanged or decreased FT have been reported in [89] and [67],
respectively), increased [109], unchanged [75], or decreased FPA
[67, 84], and lesser shortening [75] or increase of FPTT and EPTT
[67, 84]. Increased electrodermal activity is shown in increased
SCL [74, 109, 115, 126] and increased nSRR [61, 100]. Some studies
also reported unchanged SCL [41, 55, 75, 77] or decreased SCL [67].
Increased respiratory activity is evidenced in increased RR [14, 33,
61, 67, 75, 84, 90, 100] or unchanged RR [59], decreased Tiand Te[14,
84], decreased Pe[84], increased Piand FRC [14], or unchanged Ti,
decreased Te, and unchanged respiratory variability [33], increased
depth [75] or decreased depth [14, 67], decreased respiratory vari-
ability of period and amplitude [90], increased Vt/Ti, unchanged
FRC, and increased Ros [92].
A few exceptions are of note that occurred in happiness induc-
tion with visual material, such as pictures [28, 94] or film clips [49,
131]: instead of the typical increase in HR, these studies report
decreased or unchanged HR. Decreased HR and increased SCL have
been found in response to pictures of happy faces [28]. Decreased
cardiac activity (decreased HR and slightly increased HRV, i.e., peak-
valley RSA) and decreased respiratory activity (decreased RR, Vt,
Ros, and increased Tiand Te) have been reported in [94] for happi-
ness elicited with pictures from the International Affective Picture
System (e.g., family, sky divers, happy teens, roller coaster, water
slide; Lang et al., 2005). Decreased HR and SCL have been found
in children in response to a happy scene in the film Bambi [112].
Decreased HR has also been found in response to a film depicting a
figure skater winning an Olympic gold medal [49]. Decreased car-
diovascular activity as expressed in decreased HR and unchanged
SBP and DBP have been reported in response to a film clip depict-
ing a joyful mother–daughter interaction [131]. This variance may
point to the fact that a relatively wide range of positive emotions
is commonly subsumed under the umbrella term ‘happiness.’ For
certain of the above cases, a label such as admiration, contentment,
excitement, joy, or pride may be a more appropriate descriptor.
Certain emotional stimuli may also derive special meaning from
the context in which they occur, such as pictures of smiling faces
in the event of winning or losing a game (Vrticka et al., 2009).
2.2.5. Joy
Laboratory joy elicitation has particularly relied on standard-
ized imagery [35, 124, 128, 130, 134] and personalized recall [83,
106] for emotion induction. Some studies have also used picture
viewing (e.g., faces [129]), real-life manipulations (e.g., expres-
sion of appreciation and reward by experimenter [119]), or the
Velten method [19]. Taken together, an autonomic response pat-
tern of increased cardiac vagal control, decreased -adrenergic,
increased -adrenergic, and increased cholinergically mediated
sympathetic influence as well as increased respiratory activity may
be concluded, however, awaiting confirmation by further investi-
gations. Whereas all other positive emotions are characterized by
decreased -adrenergic sympathetic influence, joy appears to be
characterized by increased -adrenergic sympathetic activation,
an autonomic response component that has been associated with
increased motivational engagement (Wright, 1996), co-occurring
with increased vagal activation in the response pattern of joy.
Specifically, the physiological response pattern of joy was gen-
erally characterized by increased HR, accompanied by reports of
either unchanged SCL [124, 128, 130] or increased SCL [129] as
well as increased nSRR [35, 119]. The physiological response pat-
tern of joy was further characterized by increased HRV (SDNN [63]),
decreased PEP and LVET, and unchanged CO and TPR [106], or
increased PEP and TPR, decreased SV, and unchanged CO [83], as
well as increased SBP, DBP, and MAP [83, 134], or increased SBP
and unchanged DBP or MAP [106, 119, 130]. Effects on respira-
tory activity show increased RR [119]. Using the Velten method for
joy induction [19], no change in HR, SBP, DBP, and MAP has been
reported. For an emotion amalgam of joy and pride elicited in the
context of a computer game [125], mildly increased SCR, decreased
HR in anticipation of the event, and increased HR after onset of the
event, an initial deceleration, followed by an increase, and a second
decrease in FPTT, as well as faster rise in FT at low difficulty levels,
as contrasted to stronger decrease in FT at high difficulty levels has
been reported.
2.2.6. Pleasure, anticipatory
The emotion complex “anticipatory pleasure” here considers
both appetite [18] and sexual arousal [1, 23, 35, 70, 94, 120, 122,
133]. Physiological responses of anticipatory pleasure appear to be
grouped according to type of task, indicating physiological deacti-
vation when emotionally evocative material is visually presented
(e.g., picture viewing [18, 70, 94] or film clips [1, 23]) and physio-
logical activation when emotionally evocative material is imagined
(e.g., standardized imagery [35, 120, 122]). Overall, these studies
suggest that visual material that relates to anticipatory pleasure
elicits increased cardiac vagal control, increased electrodermal
activity, and respiratory deactivation. On the other hand, imag-
ined material that relates to anticipatory pleasure elicits increased
cardiac activation (either via increased sympathetic or decreased
parasympathetic influence) and increased respiratory activity.
Looking at material that relates to anticipatory pleasure is asso-
ciated with decreased HR [10, 18, 22, 23, 94] and increased SCR [10,
22, 70] (although small or unchanged SCR have also been reported
[18, 94]) and increased SCL [23] together with increased FT [18]
and increased HRV (peak-valley RSA), Ti,Te, decreased RR, Vt,Vm,
and unchanged Ros [94]. Imagining material that relates to anticipa-
tory pleasure, in contrast, is associated with increased HR [35, 122],
increased nSRR [35], and increased RR together with decreased
pCO2,Ti,Te, and Vt[120, 122]. As an exception, increased HR and
increased SCR has been reported in the context of presenting erotic
pictures [133] and increased HR, HRV, SBP, DBP, SCR, SCL, decreased
FT, and unchanged HT, RR, and respiratory variability has been
reported in the context of presenting an erotic film clip [1]—notably,
both studies included only male participants.
2.2.7. Pride
Laboratory induction of pride has used film clips [49], per-
sonalized recall [115], or real-life manipulations of experimenter
praise [54]. These studies report decreased HR and unchanged HRV
(RMSSD [49]), increased SCL [49, 115], and a small increase in HR
together with unchanged PEP, CO, and TPR [54]. These results may
suggest an activation pattern of decreased -adrenergic cardio-
vascular activity, increased cholinergic sympathetic influence, and
unchanged vagal control in pride. However, due to the small num-
ber of studies that investigated pride, further research is strongly
408 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
2.2.8. Relief
Conceptualizing the absence of danger in a threat-of-shock
paradigm as relief (e.g., Vlemincx et al., 2009), such studies charac-
terize the physiological response to relief by decreased sympathetic
vascular and electrodermal influence and decreased respiratory
activity. As is true for the largest part of physiological responding
in positive emotion, only further research will allow firm conclu-
sions. Similar to sadness, the physiological response to relief shows
decreased electrodermal and respiratory activation, which is a dis-
tinguishing characteristic of relief to all other positive emotions.
In particular, the physiological response to relief is marked by
moderate cardiovascular changes (mild HR acceleration [17]; or
unchanged HR [13]; and increased FPA [12, 13]). There is moreover
a decrease in respiratory activity (decreased RR, associated with
increased Ti,Te, increased Vt, and decreased Vtvariability as well
as decreased sigh frequency [12]; or increased Viincluding sighs,
unchanged Viexcluding sighs, and increased sigh frequency [127]).
Notably, increased sigh frequency has also been reported for con-
ditions of relief in animal experiments (Soltysik and Jelen, 2005).
Finally, decreased electrodermal reactivity is typically reported
(decreased SCR reactivity [12, 17]; decreased nSRR [12]; decreased
SCL [12, 20]).
2.3. Emotions without clear valence connotation
2.3.1. Surprise
Surprise has been reported to be associated with short-duration
SCR [3] of medium response size and characterized by rapid
increase and rapid return [24], increased SCL [74], increased HR [14,
32, 74] , decreased [32] or increased FT [24, 74], unchanged respi-
ratory timing and volume parameters [14], or decreased RR and
increased respiratory depth [34]. Feleky (1916, p. 230) pointed out
a “decided inspiratory pause” of the characteristic breathing curve
of wonder, that—albeit its overall similarity to that of fear—makes
it distinct. No uniform response pattern can be derived due to the
limited number of studies investigating surprise. Including the lit-
erature on unexpected stimulus presentation (Epstein et al., 1975;
Niepel, 2001; Qiyuan et al., 1985) and the orienting reflex (Siddle
and Heron, 1976; Siddle et al., 1983; Siddle, 1985, 1991; Sokolov,
1990) may prove more conclusive.
2.3.2. Suspense
Suspense, induced in the context of film clips, has been found to
be associated with decreased HR, increased nSRR and SCL [57, 58] as
well as increased RR, decreased Te,Pe,Vt/Ti, and variability of Te, and
increased Ti/Ttot [15]. While the physiological response to suspense
clearly differs from that to surprise by cardiorespiratory measures,
further research will have to address whether suspense constitutes
a separate emotion class or whether it may be subsumed under
anxiety (see Nomikos et al., 1968).
3. Discussion
ANS activity is viewed as a major component of the emotion
response in many recent theories of emotion (see Table 1). Different
levels, on which an organizing principle of autonomic responding in
emotion might be located, were identified in the introduction and
the complementary nature of these approaches was pointed out.
The empirical review compiled a large database that can be drawn
on to evaluate such statements. What is the empirical evidence for
positions of various degrees of ANS specificity in emotion?
3.1. Autonomic responding in emotion
With the chosen approach, both specificity and similarity of
autonomic activity in emotion was shown. Table 2 presents a sum-
mary of the modal response pattern found for each emotion. The
large scope of this review necessitated a considerable degree of
abstraction; thus, only direction, but not magnitude of response,
was coded (cf. Folkow, 2000). This choice was made because quan-
tification of response magnitude ultimately depends on the type of
baseline or comparison condition used, operationalization of which
varied greatly across studies (see Kreibig et al., 2005; Levenson,
1988, for issues of physiological response quantification in emo-
tion in relation to baseline choice). Also, a number of assumptions
had to be made in order to code and classify the large variety of
studies. Moreover, numerous conclusions remain tentative at best,
as the number of studies reporting effects on certain parameters
remains limited. In that way, Table 2 may serve as an instructive
guide for future research of specific emotion contrasts and auto-
nomic parameters that demand further empirical study.
3.1.1. Summary of empirical emotion effects and their relation to
models of autonomic response organization
A number of notable differences between emotions emerged:
HR was increased for negative (anger, anxiety, contamination-
related disgust, embarrassment, fear, crying sadness) and pos-
itive emotions (imagined anticipatory pleasure, happiness, joy)
as well as for surprise. HR decreased in mutilation-related
disgust, imminent-threat fear, non-crying sadness, acute sad-
ness, affection, contentment, visual anticipatory pleasure, and
suspense—emotions that all involve an element of passivity, and
may be taken to suggest vagal mediation (cf. Porges, 1995, 2001;
Vingerhoets, 1985). Contamination-related disgust was, however,
the only negative emotion with conclusive data on increased
cardiac vagal influence, as indicated by increased HRV (see also
predictions of PNS involvement in disgust, Woody and Teachman,
2000). Acute sadness may be characterized by increased cardiac
vagal influence as well, an assumption that remains to be clari-
fied in future research. For positive emotions, increased HRV was
present in amusement and joy, whereas HRV was decreased in
happiness and visual anticipatory pleasure. This pattern of results
supports previous statements that PNS activity may play a role in
both pleasant and unpleasant emotions (e.g., Gellhorn, 1970; Kling,
TWA, an index of sympathetic influence on the heart (Furedy et
al., 1992; but see Contrada, 1992), was found to be decreased in both
anger and fear, whereas it was increased for mutilation-related
disgust and contentment, indicating decreased cardiac sympa-
thetic influence in the latter. Decreased HR in mutilation-related
disgust and contentment may thus be caused by sympathetic with-
drawal rather than parasympathetic influences (see also decreased
LF/HF in mutilation disgust). In line with this, contentment was
the only emotion that evidenced increased LVET, pointing to
decreased left ventricular contractility that indicates decreased
-adrenergic sympathetic activation. Likewise, decreases in car-
diac contractility were present in acute sadness, amusement, and
happiness, as indicated by increased PEP. Notably, these emotions
have all been related to approach motivation—with either suc-
cessful (amusement, happiness) or unsuccessful outcome (acute
sadness)—whereas emotions that are related to increased cardiac
contractility (anger, disgust, embarrassment, and fear) may be
summarized as an active coping response to aversive situations
(Obrist, 1981; Schneiderman and McCabe, 1989) or be located on a
dimension of avoidance, with the exception of anger that has been
suggested to be associated with approach motivation (Carver, 2001;
Harmon-Jones et al., this issue; but see the distinction of ‘moving
against’ and ‘moving toward’; Roseman, 2001). Effects of decreased
-adrenergic activation in certain approach-related emotions are
also evident in peripheral cardiovascular measures. Decreased acti-
vation was found for acute sadness, with decreased blood pressure
(SBP, DBP, MAP) and increased pulse transit time. Decreased blood
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 409
pressure moreover occurred in contentment, and lengthening of
pulse transit time in happiness. Larger pulse amplitude was present
for anticipated sadness as well as for relief, although skin temper-
ature generally decreased for different types of sadness.
Fear and anger were similar in a number of parameters, though
differed—as predicted by the catecholamine hypothesis (Ax, 1953;
Funkenstein et al., 1954; Stemmler, 2003, 2009)—regarding TPR,
which increased in anger, whereas it decreased in fear. Remarkably,
fear was the only emotion in the present review that evidenced a
decrease in TPR. All other emotions were characterized either by
increased (anger, contamination-related disgust, embarrassment,
anticipatory sadness, amusement, happiness) or unchanged TPR
(mutilation-related disgust, joy, pride).
Emotional activation was moreover shown to be related to
notable differences in respiratory activity. For contamination-
related disgust, respiratory timing parameters indicated faster
breathing with increased expiratory and decreased inspiratory
duration. This expiratory shift is also indicated in decreased Ti/Ttot ,
and may function to expel foul smell and related agents that the
organism might have inhaled, as would be postulated by a basic
coping strategies approach (compare to the physiological response
pattern of vomiting; Sherwood, 2008). Tiand Ti/Ttot were also
decreased in amusement, possibly reflecting effects of laughing
on respiration, which notably occurs during the expiratory part
of breathing. Of note, whereas amusement and contamination-
related disgust were remarkably similar with respect to changes
indicated by respiratory variables as well as vagal indicators, the
two differed on -adrenergic cardiac activation, with decreased PEP
in contamination-related disgust, and increased PEP in amusement.
In contrast, both increased Tiand increased Te, resulting in a general
slowing of breathing, occurred in contentment, visual anticipatory
pleasure, and relief. A marked inspiratory pause was present in
anger, fear, and surprise, together with increased breathing fre-
quency and increased Ti/Ttot . Fast deep breathing has been found
for non-crying sadness that may function as an expressive emo-
tion regulation strategy to actively suppress crying—a hypothesis
that needs to be addressed in future research. Slow deep breath-
ing has been found for relief, whereas shallow breathing occurs in
anxiety, disgust, certain types of sadness, as well as anticipatory
pleasure. Decreased pCO2, indicating hyperventilation, was more-
over reported for anxiety, fear, and imagined anticipatory pleasure,
whereas increased pCO2was reported for acute sadness and con-
tentment. These constellations may suggest variations according
to basic motivational features such as valence and arousal (Bradley
and Lang, 2000; Lang et al., 1993) or shared core processes (see
Berridge, 1999, for a discussion of commonalities between anxiety,
fear, and anticipatory pleasure, viz. desire).
Decreases in electrodermal activity were present but in a few
emotions, namely non-crying sadness, acute sadness, contentment,
and relief. All other emotions were accompanied by increased
electrodermal activity, which has been proposed to reflect cog-
nitively or emotionally mediated motor preparation (Fredrikson
et al., 1998), consistent with the notion of emotion causing an
increase in action tendency (Brehm, 1999; Frijda, 1986). The
decrease in electrodermal activity may in turn be taken as indica-
tive of a decrease of motor preparation in the former emotions:
sadness is typically experienced under conditions when a loss
has occurred that cannot be undone, relief is experienced after
a threat has passed, and contentment is experienced when one
has attained a satisfactory outcome. As Brehm (1999, p. 7) pointed
out, “the outcome has already occurred and there is nothing more
to be done about it.” Hence, neither emotion is characterized by
an urge for action; rather, passivity is the shared motivational
Across response systems, psychophysiological responses in
sadness-inducing contexts were characterized by decreased FPA,
increased pulse transit time, and decreased electrodermal activity.
As an exception, anticipatory sadness showed a reversed response
pattern that was remarkably similar to that of anxiety in a number
of measures. This may point to a shared dimension of anticipation of
harm or loss, as discussed in more detail below. Differential asso-
ciation of sadness or grief with either predominant SNS (Averill,
1968) or PNS activation (Gellhorn, 1964, 1970) might have been
the result of having such different types of sadness as crying ver-
sus noncrying sadness or anticipatory versus acute sadness in
It may be asked whether such positive emotions as amuse-
ment, happiness, and joy differ physiologically. The present review
suggests that, whereas in amusement and joy HRV increases,
it decreases in happiness. Amusement and happiness share a
lengthening of PEP that is less clear in joy. All three emo-
tions are characterized by increased electrodermal activity and
faster breathing, which is deeper in amusement, but shallower in
happiness. Similarly nuanced physiological response differences
between interest, joy, pride, and surprise have been reported by
Kreibig et al. (this issue).
3.1.2. Measures of autonomic activation components
Scientific investigation should not stop at the question of
whether emotions differ physiologically, but rather ask whether
and in which way emotions differ in terms of activation compo-
nents of the ANS (e.g., Berntson et al., 1991, 1993; Stemmler et al.,
1991; Stemmler, 1993). Investigations of ANS responding in emo-
tion have long been impeded by the exclusive use of “convenience
measures,” such as HR and electrodermal activity, as sole indicators
of the activation state of the organism (notably 23 of the publi-
cations included in the present review). However, as far back as
William James (1884, 1894), complex emotion syndromes of highly
specific and regionally organized regulation patterns have been
described that include various quantifiable cardiovascular, eccrine,
and respiratory responses. Because the heart is dually innervated by
the SNS and PNS that speed or slow HR either in coupled (reciprocal,
coactivated, or coinhibited) or uncoupled modes, HR is not informa-
tive of the respective branch’s influence upon cardiac functioning
(Berntson et al., 1991, 1993). Measures such as PEP and RSA that
have been shown to be indicative of -adrenergic sympathetic and
vagal influence on the heart, respectively, are more informative and
should thus be preferred. Moreover, skin conductance cannot func-
tion as the sole indicator of sympathetic activity since directional
fractionation between response systems, such as the cardiovascu-
lar and electrodermal, is known to exist (Lacey, 1967). In addition,
Berntson et al. (1991, p. 483) pointed out that “even chronotropic
and inotropic influences on the heart ... are mediated by sepa-
rate efferent pathways that may be subject to differential central
control. Consequently, indices should optimally be derived from
the same functional dimension of the target organ.” Thus, as the
physiological adjustments that are elicited by emotion consist of
an integrated pattern of responses, it is important to judiciously
select a sufficient number of response measures to allow for the
response pattern and its variations to be identified (Hilton, 1975;
Schneiderman and McCabe, 1989; Stemmler, 2004).
Current models of autonomic control may moreover serve as a
guide for interpreting findings of autonomic measures, in particular
within replication studies of emotions (Berntson et al., 1991). Low
replicability of autonomic response patterns of certain emotions
may indicate low directional stability (i.e., nonmonotonic response
functions), a restricted dynamic range, and low response lability
(i.e., small rate of change) that is characteristic of nonreciprocal
modes of activation. In contrast, high replicability of autonomic
response patterns would speak for high directional stability, a wide
dynamic range, and high response lability that is characteristic of
reciprocal modes of activation.
410 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
3.1.3. Emotion terminology
In measuring autonomic responding in emotion, it is moreover
important to place expected or observed effects on a sound concep-
tual basis. In this context, the importance of a clear and generally
agreed upon terminology for labeling emotions cannot be stressed
enough. Part of noted inconsistencies can be attributed to a lax and
indistinct use of emotion labels for describing investigated emo-
tions. For example, it is important to distinguish between such
emotions as fear and anxiety, although they are both related to
threat appraisals, but differ on the dimension of threat imminence
(Barlow, 1991; Craske, 1999) or may be altogether based on two dis-
tinct behavioral systems (e.g., Gray, 1982; Gray and McNaughton,
2000). Similarly, amusement and happiness are both emotions
related to a pleasurable experience. Amusement, however, refers
to appealing to the sense of humor and should be reserved to such
emotion inductions as those using slapstick comedy, whereas hap-
piness refers to feelings of well-being or a pleasurable or satisfying
experience, often caused by a deed of good fortune external to one’s
proper control (Aristotle, 1893; Veenhoven, 1991). Another impor-
tant differentiation that could not be given due account in the above
review of research findings is the distinction of shame and embar-
rassment (Lewis and Granic, 2000; Tangney et al., 1996; Teroni
and Deonna, 2008). Whereas shame is typically instigated by per-
sonal failure, embarrassment is more related to social exposure.
On the other hand, the low-arousal positive emotions, here sub-
sumed under the label of contentment, appear under a number of
different names, such as pleasure, serenity, calmness, peacefulness,
and relaxation. Moreover, anticipatory states of fear (anxiety) and
sadness (anticipatory sadness), that were here distinguished from
other forms of fear and sadness, respectively, might be regrouped
into a category of worry or mental distress resulting from concern
for an impending or anticipated painful experience of harm or loss,
cutting across linguistically-defined boundaries (cf. Barr-Zisowitz,
2000, for a discussion of types of sadness). Both share an uncer-
tainty about the kind of harm and what can be done to prevent a
fatal outcome (cf. helplessness; Seligman, 1975). Appraisal mod-
els that present prescriptive appraisal–emotion mappings (e.g.,
Roseman, 1984; Roseman et al., 1994; Scherer, 1982, 2001; Smith
and Ellsworth, 1985) may serve as a general guide of how to label
different experimental emotion conditions.
Apparent inconsistencies previously noted regarding auto-
nomic activity in emotion (e.g., Feldman-Barrett, 2006) may thus
be accounted for by conceptualizing “modal emotions” (Scherer,
1994, 2001) or “emotion families” (Ekman, 1997, 1999) as umbrella
terms, under which different subtypes of that emotion exist, related
to small but important differences in appraisal outcomes. In that
sense, emotions might be grouped together in functional com-
plexes under an abstract theme (cf. core relational themes; Lazarus,
1991) with its various specific, i.e., condition-sensitive, implemen-
3.2. Boundary conditions
The present review focused on the relation between emotion
and ANS activity. Emotion was defined as a multi-component
response to an emotionally potent antecedent event, causing
changes in subjective feeling quality, expressive behavior, and
physiological activation. However, there is no one-to-one relation-
ship between emotion and changes in autonomic activation: feeling
changes may occur without concomitant autonomic changes, just
as autonomic changes may occur without concomitant feeling
changes. Moreover, the present review assumed that study par-
ticipants can faithfully report on their emotional state. However,
decoupling of subsystems may occur, such as in emotion elicita-
tion by subliminal stimulus presentation, unconscious emotions
(presence of physiological effects, but absence of conscious feel-
ings), or low response system coherence due to some intervening
process, such as emotion regulation. To conclude, boundary condi-
tions of the relation between emotion and autonomic activity and
their implications for our understanding of emotion, feeling, and
autonomic changes are discussed.
3.2.1. Feeling changes without concomitant autonomic changes
A large body of literature reports on feeling changes in the
absence of effects on autonomic responding. Typically, the type of
affect manipulated within the context of such studies is labeled
‘mood,’ referring to a diffuse and long-lasting affective state that is
not object-related, i.e., not experienced in simultaneous awareness
of its causes (Frijda, 1993; Gendolla, 2000; Schwarz and Clore, 1988;
however, see also the concept of the ‘as-if body loop,’ Damasio,
1999). Unlike emotions that are associated with specific motiva-
tional functions, e.g., motivating to remove the object of anger or to
escape from the object of fear, moods do not have specific and stable
motivational functions, but only informational function. Although
moods have thus no direct impact on behavior, they do influence
effort investment in subsequent behavior, such as performing a
Thus, whereas moods have immediate effects on subjective
feeling state and facial expression, autonomic effects are typi-
cally absent during mood induction. No change from baseline
activation of systolic and diastolic blood pressure, heart rate, and
skin conductance level or spontaneous response rate has been
found in the context of disguised mood manipulations, ranging
between eight and ten minutes, with film excerpts (e.g., Silvestrini
and Gendolla, 2007), music excerpts (e.g., Gendolla and Krüsken,
2001), autobiographic recall (e.g., Gendolla and Krüsken, 2002),
or odors (Kiecolt-Glaser et al., 2008). Still, autonomic activation
in subsequent task performance is moderated by mood, with the
direction of effect depending on perceived difficulty level of the task
(Gendolla, 2003; Gendolla and Brinkmann, 2005). When addressing
affective effects on ANS activity, it is therefore of utmost importance
to distinguish mood from emotion in order to know when to expect
autonomic effects and when not.
3.2.2. Autonomic changes without concomitant feeling changes
Reviewed results of effects of emotion on autonomic activ-
ity necessarily underly a specific measurement model. The ANS
is not exclusively servant to emotion. Non-emotional physical,
behavioral, and psychological factors affect physiological activation
before, during, and after emotion, producing a complex amalgam
of effects on physiological activity. Emotions are typically assumed
to influence the ANS during a relatively brief period of time in the
range of seconds to only a few minutes (Ekman, 1984, 1994). Once a
behavioral reaction has been initiated, the physiological activity is
in the service of that behavior and no longer reflects predominantly
effects of emotion (Levenson, 2003; Stemmler, 2004).
To disentangle the potential confounding context effects from
emotional effects on physiological activation, three major fac-
tors have been recognized that influence physiological responding
(Stemmler et al., 2001; Stemmler, 2004): (a) effects of the non-
emotional context include posture, ambient temperature, ongoing
motor activity, or cognitive demands, that are not in the ser-
vice of emotion, constraining the physiological effects that the
other components may exert; (b) effects of the emotional context
include organismic, behavioral, and mental demands of enacting
the emotion, given the specific momentary situational allowances
and constraints on the emotional behavioral response, represent-
ing context-dependent effects of emotion that may be variable
across situations; (c) effects of the emotion proper reflect spe-
cific physiological adaptations with the function to protect the
organism through autonomic reflexes and to prepare the organism
for consequent behavior, representing context-independent effects
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 411
of emotion, which are expected to be stable across situations.
Only the third component of the model, the emotion signature
proper, is expected to allow statistical identification of specific,
non-overlapping emotion responses (Stemmler et al., 2001).
3.2.3. Decoupling of subsystems in emotion
To demarcate emotion from other physical and psychologi-
cal influences on ANS activity, subsystem synchronization has
been proposed as a distinctive feature of emotion (Scherer, 2001).
Coherence constraints between response systems of emotion have,
however, been noted in some studies (e.g., Mauss et al., 2005;
Reisenzein, 2000; Ruch, 1995). Such dissociation among different
measures of emotion may be relatively normal rather than reflect-
ing aberrant functioning. Emotion regulation, as one prominent
process in this regard, may influence subsystem coherence in vari-
ous ways, such as with respect to awareness of emotional responses
(Koole, 2009). Emotions can, moreover, be elicited by subliminally
presented stimuli that do not enter conscious awareness (e.g., Flykt,
Esteves, & Öhman, 2007; Öhman, Carlsson, Lundqvist, & Ingvar,
2007; Wiens et al., 2008). Thus, although feelings are often and typ-
ically conscious, conditions may arise, under which people do not
report and/or are not aware of an emotional experience, although
other subsystems, such as facial expression, physiological activa-
tion, and behavioral tendency indicate occurrence of emotion (cf.
unconscious emotions; Wilson, 2002; Winkielman and Berridge,
2003, 2004).
Collecting valid data on autonomic responding in emotion has
been and remains to be a challenge to emotion research (e.g.,
Levenson, 1988; Stemmler, 2003). For progress in the understand-
ing of the functional organization of ANS activity in emotion, future
researchers will have to closely scrutinize and, if possible, verify
the specific type of emotion elicited as well as individual variations
when analyzing autonomic parameters that need to be selected
such that they allow differentiation of the various activation com-
ponents of the ANS. Only if the hypothesis of autonomic response
organization is properly tested, can valid inferences be drawn. It is
hoped that this will pave the road to arriving at James’ (1890) call
for a generative principle that can summarize and account for the
varieties of emotion.
Sylvia D. Kreibig is now at the Department of Psychology, Stan-
ford University, Stanford, CA 94305. I thank Guido Gendolla, Klaus
Scherer, and Tom Cochrane as well as two anonymous reviewers
and special issue editor Bruce H. Friedman for helpful comments
on earlier versions of this manuscript. Thanks also to research
assistant Nora Meier for assisting in the literature search. This
research was supported by the Swiss National Science Foundation
(PBGEP1-125914) as well as by the National Center of Compe-
tence in Research (NCCR) Affective Sciences financed by the Swiss
National Science Foundation (51NF40-104897) and hosted by the
University of Geneva.
Appendix A. Overview of reviewed studies
Table A.1 provides an overview of the studies considered in
the present review. Emotions were coded according to the emo-
tion labels provided by the authors. The table moreover indicates
the type of emotion induction method as well as assessed phys-
iological measures (grouped into cardiovascular, respiratory, and
electrodermal). Averaging period is the time segment over which
averages for physiological variables were calculated; in case of dif-
ferent averages for different physiological variables, more than one
number is indicated; in case of varying averaging periods due to dif-
ferent stimulus presentation lengths, the mean averaging duration
rounded to the next full minute is indicated.
This table can be downloaded as a text file from Data presented in this
table were also used to generate the tag clouds.
Table A.1
Overview of studies on effects of emotion on autonomic nervous system activity.
No. Authors Year NEmotion labels Experimental
Cardiovascular Respiratory Electrodermal Averaging
period (in s)
1Adamson et al. 1972 10 Sexual arousal Film clips HR, HRV (CVT),
RR nSRR 30, 120, 240
2Adsett et al. 1962 30 Anger, anxiety,
Stress-interview HR, SBP, DBP,
3Alaoui-Ismaïli et
1997 44 Anger, disgust, fear,
happiness, sadness,
odorants HR, palm temp.,
4Allen et al. 1996 100 Achievement failure,
social rejection
HR 30
5Aue et al. 2007 42 Goal conduciveness,
relevance, threat
Picture viewing
HR, FT, arm
1, 5
6Averill 1969 54 Mirth, sadness Film clips HR, SBP, DBP,
FPA, FT, face
RR, respiratory
nSRR 15, 360
7Ax 1953 43 Anger, fear Real-life
(harassment, threat
of short-circuit)
DBP, FT, face
RR, Ti/Ttot , RD nSRR, SCR 6
8Baldaro et al. 1996 30 Fear Film clips HR 120
9Baldaro et al. 2001 42 Disgust Film clips HR, HRV (RSA
RR 600
10 Bernat et al. 2006 48 Sexual arousal,
Picture viewing
11 Blatz 1925 18 Fear Real-life (sudden
12 Blechert et al. 2006 42 Anxiety Threat of shock HR, TWA, HRV
(Porges), LF/HF,
RR, Ti,Te,Pi,Pe,
Ti/Ttot ,Vt,Vm,
pCO2, sigh
frequency, sigh
Vt, % thoracic Vt
412 S.D. Kreibig / Biological Psychology 84 (2010) 394–421
Table A.1 (Continued)
No. Authors Year NEmotion labels Experimental
Cardiovascular Respiratory Electrodermal Averaging
period (in s)
13 Bloom and
1977 64 Anxiety Threat of shock HR, FPA 30
14 Boiten 1996 16 Anger, disgust, fear,
happiness, sadness,
Directed facial
HR Ttot ,Ti,Te,Pi,Vt,
10, 30
15 Boiten 1998 27 Amusement, disgust,
fear, suspense,
Film clips Ttot,Ti,Te,Pi,Pe,
Ti/Ttot ,Vt,Vm,
Vt/Ti, RC/Vt,
16 Bradley et al. 2001 95 Disgust Picture viewing
HR SCR 0.5
17 Bradley et al. 2008 49 (control) Dental anxiety Threat of shock HR SCR 20
18 Britton et al. 2006 40 Appetite, disgust,
amusement, sadness
Film clips HR SCR 30, 90
19 Brown et al. 1993 16 Elation, sadness Velten method HR, SBP, DBP,
20 Chan and
1996 23 Threat Threat of shock SCL 40
22 (Control)
21 Christie and
2004 34 Amusement, anger,
disgust, fear, sadness
Film clips IBI, HRV (MSD),
SCL 60
22 Codispoti and
De Cesarei
2007 50 Disgust, sexual
arousal, threat
Picture viewing
HR SCR 0.5
23 Codispoti et al. 2008 55 Disgust, sexual
film clips HR, HRV (RSA
SCL 60
24 Collet et al. 1997 30 Anger, disgust, fear,
happiness, sadness,
Picture viewing
SBF, palm temp. RR OPD, SYDER,
SCR, duration
25 Davidson and
1976 20 Anger, relaxation Personalized recall HR 120
26 Demaree et al. 2004 26 (control) Amusement, disgust Film clips IBI, HRV (RSA
(HF), LF/HF)
SCL 120
27 Dimberg 1986 28 Fear Picture viewing HR SCR 1
28 Dimberg and
2007 28 (control) Anger, happiness Picture viewing
HR SCR 1, 5
29 Drummond 1999 19 (control) Anger Real-life
FPA, forehead
SCR 15
30 Dudley 1964 10 Anger, anxiety,
Hypnosis RR, Vm, pCO2
31 Eisenberg et al. 1988 82 Anxiety, sadness,
Film clips HR 0.5, 3.5
32 Ekman et al. 1983 16 Anger, disgust, fear,
sadness, surprise
Directed facial
action, personalized
HR, FT SCL 10, 30
33 Etzel et al. 2006 13 (18) Fear, happiness,
Musical excerpts HR, HRV (SDNN,
Ttot ,Ti,Te,
1, 5, 65
34 Feleky 1916 6 Anger, disgust, fear,
hatred, laughter,
pleasure, wonder
Personalized recall (RR), I/E ratio,
RD, RD/Ttot
35 Fiorito and
1994 31 (control) Anger, contentment,
fear, joy, sadness,
sexual arousal
personalized recall
HR nSRR 20
36 Foster et al. 1999 36 Anger Real-life,
personalized recall
HR SCL 30?
37 Foster and
2001 10 Anger, mirth personalized recall HR SCL 30
38 Foster et al. 2003 23 Mirth Real-life,
personalized recall
HR SCL 30?
39 Fredrickson and
1998 60 72 Fear, sadness Film clips HR, FPTT, EPTT,
40 Funkenstein et
1954 69 Anger Real-life
41 Gehricke and
2002 20 Happiness, sadness Standardized
42 Gilissen et al. 2008 78 Fear Film clips HRV (RMSSD) SCL 60
43 Gilissen et al. 2007 78 Fear Film clips HRV (RMSSD) SCL 60
44 Gross 1998 120 Disgust Film clips IBI, FPA, FT SCL 1, 60
45 Gross et al. 1994 150 Sadness Film clips HR, FPA, FPTT,
RP, RD SCL 100
46 Gross and
1993 43 Disgust Film clips HR, FPA, FPTT,
RP, RD SCL 1, 60
S.D. Kreibig / Biological Psychology 84 (2010) 394–421 413
Table A.1 (Continued)
No. Authors Year NEmotion labels Experimental
Cardiovascular Respiratory Electrodermal Averaging
period (in s)
47 Gross and
1997 180 Amusement,
Film clips IBI, FPA, FPTT,
RP, RD SCL 210
48 Grossberg and
1968 18 Fear Adapted
10 (control)
49 Gruber et al. 2008 54 (control) Disgust, happiness,
pride, sadness