-
Ann. N.Y. Acad. Sci. ISSN 0077-8923
ANNALS OF THE NEW YORK ACADEMY OF SCIENCESIssue: The Year in
Cognitive Neuroscience
Functional imaging studies of emotion regulation:a synthetic
review and evolving model of the cognitivecontrol of emotionKevin
N. Ochsner, Jennifer A. Silvers, and Jason T. BuhleDepartment of
Psychology, Columbia University, New York, New York
Address for Correspondence: Kevin Ochsner, Department of
Psychology, Columbia University, 369 Schermerhorn Hall,
1190Amsterdam Ave., New York, NY 10027.
[email protected]
This paper reviews and synthesizes functional imaging research
that over the past decade has begun to offer newinsights into the
brain mechanisms underlying emotion regulation. Toward that end,
the first section of the paperoutlines a model of the processes and
neural systems involved in emotion generation and regulation. The
secondsection surveys recent research supporting and elaborating
the model, focusing primarily on studies of the mostcommonly
investigated strategy, which is known as reappraisal. At its core,
the model specifies how prefrontaland cingulate control systems
modulate activity in perceptual, semantic, and affect systems as a
function of onesregulatory goals, tactics, and the nature of the
stimuli and emotions being regulated. This section also shows how
themodel can be generalized to understand the brainmechanisms
underlying other emotion regulation strategies as wellas a range of
other allied phenomena. The third and last section considers
directions for future research, includinghow basic models of
emotion regulation can be translated to understand changes in
emotion across the life span andin clinical disorders.
Keywords: amygdala; cognitive control; emotion; emotion
regulation; prefrontal cortex
. . .Thy fate is the common fate of all,Into each life some rain
must fall. . .
Henry Wadsworth LongfellowThe Rainy Day (1842)
. . .Every cloud, says the proverb, has a silverlining.
P. T. BarnumStruggles and Triumphs (1869)
It might be said that emotions are the weather ofour lives. Some
days, we experience the blue skies ofhappiness and the sunshineof
joy.Otherdays,we aredrenched by the rain clouds of sadness or
buffetedby the hot winds of anger. How we respond adap-tively to
our emotional weather patternsfindingthe silver lining in every
dark cloudhas impor-tant consequences for our physical andmental
well-being.17
Although we cannot control the weather outside,we are capable of
using myriad emotion regulationstrategies to take control of our
internal climates.8
Such strategies allow us to wholly or partially alterthe nature,
magnitude, and duration of our emo-tional responses, including
initiating new ones. Inrecent years, great strides have been taken
in usingneuroscience techniques to understand the mecha-nisms
underlying emotion regulation. In humans,this research has
primarily used functional imagingto examine our ability to control
affective responsesusing cognitive strategies. The overarching
goals ofthis paper are to review the progress made by suchresearch,
synthesize from it conclusions that suggestexpansion on and
elaborations of a model of thecognitive control of emotion (MCCE),
and showhow the model can make sense of a wide range ofemotion
regulatory abilities and allied phenomena.Toward these ends, the
remainder of the paper is
divided into three parts. In the first, we outline a
ba-sicMCCEwhose core elements have been describedpreviously.9,10 In
the second section, we review cur-rent imaging research suggesting
ways in which themodel can evolve to integrate new findings on
the
doi: 10.1111/j.1749-6632.2012.06751.xAnn. N.Y. Acad. Sci. 1251
(2012) E1E24 c 2012 New York Academy of Sciences. E1
-
Functional imaging studies of emotion regulation Ochsner et
al.
Figure 1. A multilevel approach to building a model of emo-tion
regulation. (A) In cognitive, affective, and social neuro-science
research, we seek to describe phenomena in terms ofrelationships
among three levels of analysis: experience andbehavior,
psychological processes, and neural systems. Thebidirectional
arrows between levels indicate that the relation-ships among them
are bidirectional. (B) Throughmeasurementand/or experimental
manipulation, neuroimaging research onemotion regulation can
observe phenomena at the behaviorallevel and the neural level and
use these observations to infer thenature of the intervening
cognitive and/or affective processes.The direction of the arrows
from the behavioral and neurallevels toward the process level
indicates the direction of causalinference (i.e., we cant observe
the operation of these processesdirectly, but infer
theiroperationbasedonbehavioral andneuralobservations).
brain bases of emotion regulation as well as be ap-plied to
account for other related phenomena, suchas affective learning,
affect-based decision making,and affective expectancies. Throughout
these firsttwo sections we focus primarily on one strategy
inparticularknown as reappraisalbecause it hasreceived the bulk of
empirical attention. In the thirdand last section, we summarize and
consider direc-tions for future basic and translational
research.
A model of the cognitive control of emotionAny model of emotion
regulation (or any otherphenomenon) is predicated on assumptions
abouthow different levels of analysis fit together. Our
as-sumptions follow those now commonplace in cog-nitive, affective,
and social neuroscience in whichresearchers seek to describe
phenomena in termsof the relationships among three levels of
analysis:behavior/experience, process, and neural systems(Refs.
1113; Fig. 1A). Neuroimaging research onemotion and its regulation
can observe phenomenaat the behavioral level (e.g., measures of
emotionalresponse and the specific regulatory strategies onemight
employ) and the neural level (e.g., fMRImea-sures of brain
activity) and use these observations toinfer the nature of the
intervening cognitive and/oraffective processes (Fig. 1B).
With this in mind, our review of current researchwill sometimes
be organized in terms of phenomenadescribed at the level of
behavior, including regula-tory goals, tactics, and target stimuli.
In other casesit will be organized in terms of issues concerning
theneural-level pathways on which the field has begunto make
progress. Taken together, the data reviewedin each section
constrains and influences ourMCCE(see Fig. 2).To understand how
emotion regulation works,
we must first have an idea of how emotions are gen-erated. As
such, our model has two main partsdescriptions of the mechanisms
supporting emo-tion generation on the one hand and the mecha-nisms
supporting emotion regulation on the other.For the sake of
simplicity, we present the psycho-logical and neural systems
involved in the gener-ation and regulation of emotion as being
distinct,yet it should be noted that there is evidence to sug-gest
that the underlying psychological14 and neu-ral mechanisms15,16 are
at least partially overlap-ping. Indeed, elsewhere we have noted
that thedistinction between emotion generation and reg-ulation is
blurry at best (e.g., Ref. 16), and whichterm one uses may reflect
their usefulness for ad-dressing a particular question more than
hard andfast differences in their mechanisms. Here, we treatthem
separately to make points about the ways inwhich putative control
and affect-triggering systemsinteract.
Mechanisms of emotion generationOur account of how emotions are
generated is mul-tileveled12 in its description of both the
processesand the neural systems that give rise to
emotionalresponses.
Processes involved in generating emotionThe black time line at
the bottomof Figure 2A showsa simple model of four steps involved
in generatingemotional responses.17 In the first step, a stimu-lus
is perceived in its current situational context.The stimulus could
be an internal thought, feel-ing, or sensation, or any number of
external cues,ranging from a facial expression or gesture to
anaction or event. At the second stage, one attendsto some of these
stimuli or their attributes. What-ever is in the focus of attention
is passed alongto subsequent emotion generative stages,
whereasignored or unattended stimuli may be either
E2 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
Figure 2. A model of the cognitive control of emotion (MCCE).
(A) Diagram of the processing steps involved in generating
anemotion and the ways in which cognitive control processes (blue
box) might be used to regulate them. As described in the text,the
effects of different emotion regulation strategies (the red arrows
descending from the cognitive control processes box) can
beunderstood in terms of the stages of the emotion generation
sequence that they influence. The pink box seen at the appraisal
stageis meant to indicate that neural systems involved in
generating emotion support this process. (B) Neural systems
involved in usingcognitive strategies, such as reappraisal, to
regulate emotion (left, blue boxes), systems involved in generating
those responses (left,pink boxes), and systems with an undefined or
intermediary role in reappraisal (left, yellow boxes).
excluded from these stages or receive diminishedsubsequent
processing. The third stage involves ap-praising the significance
of stimuli in terms of theirrelevance to ones current goals, wants,
or needs.This is the stage focused on by appraisal theoriesof
emotion, which describe the structure of dif-ferent appraisals that
lead to positive versus neg-ative reactions in general and to
specific types ofemotional responses in particular.18 Because
thecurrent neuroscience literature suggests that theremay not be
specific neural systems for different dis-crete emotions,19,20 for
present purposes, we simplydistinguish between basic
positive/appetitive ver-sus negative/aversive appraisals that have
been re-liably associated with specific neural systems thatare
described below. Finally, the fourth stage in-volves translating
these appraisals into changes inexperience, emotion-expressive
behavior, and au-tonomic physiology. Although these three
indica-tors of emotional response do not always correlatewith one
another for reasons that are not perfectlyunderstood,21 as noted
below, emotion regulation
strategies can affect changes in some or all of them,depending
on the strategy.
Neural systems involved in generatingemotionReviews and
meta-analyses of functional imagingstudies19,20 indicate that a
number of cortical andsubcortical brain systems may play key roles
in theappraisal and/or response stages of emotion gener-ation. For
present purposes, we focus on the fourthat have been most
frequently discussed in studiesof reappraisal in particular, and
emotion regulationstrategies more generally (see Fig. 2B; for
examplesof other emotion systems that may bemodulated byemotion
regulation, see Refs. 9 and 22).The first is the amygdala, which is
involved in the
perception and encoding of stimuli relevant to cur-rent or
chronic affective goals,23,24 ranging from re-wards or punishments
to facial expressions of emo-tion to aversive or pleasant images
and films.2527
Although the amygdala generally is sensitive todetecting and
triggering responses to arousing
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E3
-
Functional imaging studies of emotion regulation Ochsner et
al.
stimuli,28 it exhibits a bias toward detecting cues sig-naling
potential threats, like expressions of fear.2931
The second is the ventral striatum, which is in-volved in
learning which cues (ranging from so-cial signals, like smiling
faces, to actions, to ab-stract objects) predict rewarding or
reinforcingoutcomes.3234
The third is the ventromedial prefrontal cortex(vmPFC), which
integrates affective valuations ofspecific stimuli made by the
amygdala and ventralstriatum with inputs from other regions,
includingmedial temporal lobe systems that provide histor-ical
information about prior encounters with thestimuli as well as
inputs from brainstem motiva-tional and prefrontal control centers
that provideinformation about current behavioral goals.3543 Assuch,
vmPFC tracks the positive or negative valu-ation of stimuli in a
context and goal-dependentmanner.41,4446 Examples of this include
the findingthat vmPFC activity to an image of a healthy but
nottasty food depends on whether one has the goal toeat
healthily,47 and the findings that vmPFC lesionslead to
context-inappropriate affective responses inboth humans and
animals.39,48,49
The fourth brain system is the insula, which isthought to
represent a viscerotopic map of ascend-ing viscerosensory inputs
from the body50 and hasbeen implicated in negative affective
experience ingeneral.51,52 There appears to be
posterioranteriorfunctional gradient in the insula with posterior
re-gions associated with primary representations ofsensations from
the body and anterior regions asso-ciated with interoceptive
awareness of the body andinmotivational and affective states, like
disgust, thathave a strong visceral component.51,5356
Mechanisms of emotion regulationWith an understanding of how
emotions are gener-ated in the first place, we can turn to an
account ofthe processes and neural systems involved in regu-lating
them.
Processes involved in emotion regulationAlthoughmanybehaviors
can change our emotions,often these effects are unintended or
incidental (e.g.,your mood improves because you happen to havelunch
with a friend) and as such are not consideredto be examples of
emotion regulation, per se. In-stead, emotion regulation entails
the modificationof ongoingor the initiation of newemotional
responses through the active engagement of regu-latory
processes. That said, we can further distin-guish between cases
where emotion regulation isguided by regulatory goals that are
implicit or out-side awareness (e.g., Ref. 57) as compared with
ex-plicit and accessible to awareness. Although bothare interesting
and important, no neuroscience re-search has addressed the former
case and a greatdeal has addressed the latter case. Therefore, we
fo-cus here on thedeliberate deployment of an emotionregulation
strategy in the service of explicit goals tochange ones emotions.
To understand how such ex-plicit emotion regulation strategies
work, it is usefulto distinguish among five classes of strategies
whoseeffects on emotion can be understood in terms ofthe stage of
the emotion generation sequence onwhich they have an impact.58
It is important to note that the distinctions madebelow
originally were based on behavioral analy-ses of the aspects of
emotional responses targetedby different strategies.58 As such,
this analysis wasagnostic to the specific nature of the
regulatoryprocesses supporting each strategy, but tacitly as-sumed
that all strategies drew upon some combi-nation of cognitive
control processes (designatedby the blue box in Fig. 2A). In this
regard, func-tional imaging has made a substantial contributionto
our understanding of how emotion regulationworks because it
provides insight into the natureof the control processes supporting
emotion regu-lation that is not obtainable from behavioral
dataalone.9
As illustrated by the top portion of Figure 2A,the first two
strategies involve changing the natureof the stimulus inputs to the
emotion generationcycle. In situation selection, you keep yourself
awayfrom stimuli that elicit unwanted emotions and putyourself in
the presence of stimuli that elicit desiredemotions. An example is
staying away from a partywhere an old flame will be present if you
dont wantto feel pangs of sadness for having been dumped byher.
Situation modification is when you find yourselfin the presence of
a stimulus that elicits an unwantedemotion and change something
about the situationto alter its impact on you. In the old flame
example,youmight leave a party at which she is unexpectedlypresent
or leave the room in which she is havinga conversation. Although
these two strategies areundoubtedly effective (e.g., Ref. 59), they
can bedifficult to study neurally and have received little
E4 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
attention in imaging or using other neurosciencetechniques (see
below).The remaining three strategies are all amenable to,
andhave been studied, using imaging, albeit to vary-ing degrees.
Attentional deployment controls whatstimuli are gated into, or out
of, the emotion gen-eration process. The two most commonly
studiedexemplars10 are selective attention, which involvesshifting
the focus of attention toward or away fromstimuli or their
attributes, and distraction, whichinvolves limiting attention to an
external stimulusby focusing internally on information maintainedin
working memory. These types of strategies differfrom situation
selection in that they do not involvephysically altering ones
proximity or relationshipto an emotional stimulus, but rather they
manip-ulate attention so as to alter ones emotional re-sponse.
Cognitive change involves changing the wayone appraises the meaning
of a stimulus. It is oneof the most cognitively complex strategies
insofaras it draws on any of a number of different highercognitive
processes to support changes in stimulusmeaning, including language
and memory, as wellprocesses that also support other strategies,
such asattention and response selection. The most com-monly studied
exemplar is reappraisal, which in-volves reinterpreting the meaning
of a stimulus, in-cluding ones personal connection to it, to
changeones emotional response. Finally, response modu-lation
strategies target the systems for emotion-expressive behavior. The
most commonly studiedexemplar is expressive suppression,60 which
entailskeeping ones face still so that observers would notknow the
emotion you are experiencing.A great deal of behavioral and
psychophysiolog-
ical research has been devoted to comparing andcontrasting the
behavioral consequences of deploy-ing each of these strategies. For
example, its knownthat attentional deployment and reappraisal
canhave downstream effects on various components ofan emotional
response because they target the earlystages of the emotion
generation sequence.6065 Bycontrast, expressive suppression has an
impact ononly the behaviors it targets at the final responsestage
of emotion generation;whenkeeping your facestill, emotional
experiencemay subtly diminish,66,67
if at all, and your physiological arousal will increasefrom the
effort.60 There is also evidence that strate-gies differ in their
long-term effects. For example,reappraisal, but not distraction,
has been shown to
have long-lasting effects on ones tendency to havean emotional
response to a stimulus,65 presumablybecause only reappraisal
involves an active changein how one represents the affective
meaning of thatstimulus.
Neural systems involved in emotion regulationAs foreshadowed
previously, the use of functionalimaging has provided insight into
the nature of thecontrol systems that support regulatory
strategiesas well as the affect systems that these
strategiesmodulate to change an emotional response. Thissection
discusses core conclusions that can be drawnfrom reappraisal
studies and a model of emotionregulation that can be derived from
it.
Reappraisal as a paradigm case. Reappraisal is anappropriate
starting point for developing a MCCEfor three reasons. First,
because reappraisal is amongthe most cognitively complex
strategies, a model ofemotion regulation derived from reappraisal
workmay be generally applicable to other strategies andphenomena
that typically will be cognitively sim-pler. Second, the majority
of studies to date havefocused on reappraisal because (i) it can be
studiedeasily in an imaging environment and (ii) becauseit is the
strategy referenced by countless aphorismsthat advise us, [to] look
on the bright side. . ., [to]turn a sows ear into a silk purse. .
., When life givesyou lemons, make lemonade, and, [that] everydark
cloud has a silver lining. Third, in contrastto other areas of
emotion regulation research (re-viewed below) reappraisal studies
tend to be moremethodologically and conceptually similar to
oneanother and therefore provide a stronger base formechanistic
inferences. With these considerationsin mind, we now describe five
key insights into thebrain mechanisms supporting emotion
regulationthat have been derived from studies of reappraisal.9
Basic control systemaffect system relationships.When the first
fMRI studies of reappraisal were pub-lished approximately 10 years
ago, there were noimaging studies of any form of emotion
regulation.To develop hypotheses about how reappraisal mightwork,
an analogy was drawn between the use of cog-nition to control
emotion and the use of cognitionto control memory, attention, and
other thoughtprocesses.43 The simple idea was that prefrontal
andcingulate systems would support control processesthat modulate
activity in posterior and subcortical
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E5
-
Functional imaging studies of emotion regulation Ochsner et
al.
systems that generate emotional responses.10 Adecade and over 50
imaging studies later, this initialhypothesis has been strongly
supported.Figure 2B schematically illustrates the brain sys-
tems shown by current research to be involved inthe cognitive
control of emotion via reappraisal. Assuch, Figure 2B diagrams the
core elements of theMCCE. Three types of neural systems are
primarilyinvolved in generating and applying reappraisals.10
First, dorsolateral and posterior prefrontal cor-tex, along with
inferior parietal regions gener-ally implicated in selective
attention and workingmemory, may be used to direct attention
toreappraisal-relevant stimulus features and hold inmind
reappraisal goals as well as the content ofones reappraisal.6870
Second, dorsal regions ofthe anterior cingulate cortex implicated
in perfor-mance monitoring may help track the extent towhich ones
current reappraisals are changing emo-tional responses in the
intended way.71 Third, re-gions of ventrolateral prefrontal cortex
implicatedin selecting goal-appropriate (and inhibiting
goal-inappropriate) responses and information from se-mantic memory
may be used to deliberately se-lect a new stimulus-appropriate
reappraisal in favorof ones initial prepotent appraisal of that
stimu-lus.72,73 Finally, to the extent that ones
reappraisalinvolves focusing on and interpreting or reinter-preting
ones own emotional statesor those ofothersdorsomedial prefrontal
regions implicatedin attributing mental states also may be
active.74,75
With respect to the emotion-related regions thatare modulated by
reappraisal, the four regions de-scribed earlier in the section on
neural systemsfor emotion generation all have been implicatedalbeit
to differing extents. Far and away, the mostcommonly modulated
region is the amygdala, fol-lowed by the ventral striatum. The
insula and thevmPFC are the least commonly modulated re-gions9,10
(although see the section on pathways be-low for a potential role
of vmPFC in reappraisal asa modulator).Although we will discuss the
significance of the
differential modulation of these regions in moredetail later
(see later sections on valence speci-ficity and pathways), for now
we can highlight theconsistency with which they have been
observed.Figure 3 plots peak activation foci for 43 studies(see
Table 1) of reappraisal in healthy individuals asa function of
reappraisal goals (panel A), reappraisal
tactics (panel B), and the valence of the emotionbeing regulated
(panel C). Ignoring these distinc-tions for a moment, one can see
that the controlsystemaffect system relationships shown in Figure2B
and described previously have been observed re-liably across
numerous studies.
Moving beyond the basic modelWith the consistency of the core
controlaffect sys-tem relationship as a foundation, we are now ina
position to consider how this basic modelfirstproposed in 200243
and elaborated in 200510hasevolved. Belowwe discuss first new
conclusions thatcan be drawn about themodel from recent studies
ofreappraisal. In this section, we pay special attentionto two
emerging features of the model: (i) the po-tential intermediary
role of semantic/perceptual sys-tems in reappraisal, and (ii)
pathways linking con-trol and affect systems. Next, we discuss the
way inwhich the model can be applied to understandingregulatory
strategies other than reappraisal aswell asvarious allied phenomena
involving controlaffectsystem interactions.
Integrating new research on reappraisalRecent research provides
new insight into the distri-butionof emotion regulationrelated
activation focias a function of reappraisal goals (i.e., the
outcomeone hopes to achieve by regulating, for example,increasing
or decreasing an emotional experience),tactics (i.e., the specific
subtype of reappraisal oneimplements), and the emotional valence of
stim-uli (i.e., whether the stimulus evokes a positive ornegative
emotional response). Here, we consider theimplications of this work
for the evolving MCCE.
Goal specificityArguably, the most common goal when using
reap-praisal is to decrease negative emotion, as when weattempt to
make ourselves feel better about a dis-appointing paper rejection,
an argument, and thelike. It is not surprising then that this goal
has beenthe focus of the majority of reappraisal studies (seeTable
1). This is not the only goal that guides reap-praisal, however. In
some cases, as when we worry,ruminate, ormake ourselvesmore anxious
or fearfulby elaborating on themeaning of unpleasant events,we are
using reappraisal in service of the goal to in-crease emotion. A
small, but growing, number ofstudies have examined this reappraisal
goal as well.
E6 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
Figure 3. Plots of activation foci from the 43 studies of
reappraisal described in the text and Table 1. (A) Plots of foci as
afunction of the goals to decrease or increase emotion. (B) Plots
of foci as a function of the specific reappraisal tactics
usedeitherreinterpreting the meaning of events depicted in stimuli
or actively changing ones psychological distance from them. (C)
Plots offoci as a function of the valence of the stimuli eliciting
the emotions that participants attempted to regulate. Blue boxes
illustrateregions that are purported to support reappraisal
(increase> look and decrease> look contrasts). Pink boxes
illustrate regionsthat are purported to be modulated by reappraisal
(look > decrease and increase > look contrasts; for clarity,
only foci fallingwithin the boundaries of the amygdala and striatum
are shown).
Figure 3A plots peak activation foci for reap-praisal studies of
healthy individuals as a functionof decrease versus increase goals.
Perusal of this fig-ure highlights three findings. First, whereas
bothincrease and decrease goals recruit left prefrontalregions,
decrease goals recruit right prefrontal re-gions to a much greater
extent than do increasegoals. There are two interpretations of this
finding.First, it may be attributable to the fact that decreas-ing
an emotional response is more difficult than in-creasing one, and
therefore may require additionalcognitive control resources.76
Second, decreasingbut not increasingan emotional response
requiresinhibiting or limiting the expression of a
prepotentappraisal of a stimulus (e.g., as negative) in favor
ofselecting an alternative reappraisal (e.g., as neutralor
evenpositive). Research shows that right dorsaland especially
ventrolateralprefrontal cortex is in-volved in the selection and/or
inhibition of variouskinds of responses.7780
Second, there is some evidence that increasegoals differentially
involve anterior portions ofdorsomedial prefrontal cortex (dmPFC).
Of the12 studies directly comparing increasing emotionwith a
control condition in which participantsrespond naturally, six show
increases in anteriordmPFC.16,76,8183 Of the six studies that did
not,most showed activation in neighboring areas (suchas anterior
cingulate cortex).62,8488 Given the roleof dmPFC in making
judgments about mentalstates75,89,90 and that the majority of
reappraisalstudies use photographs of people as stimuli (seeTable
1), it is likely that these regions supportattention to and
elaboration of emotional states, in-tentions, and outcomes of the
individuals depictedin these photos.Third, whereas increase and
decrease goals both
seem to modulate the striatum (including both thecaudate and
putamen), they may differ in the waythey modulate the amygdala. On
the one hand,
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E7
-
Functional imaging studies of emotion regulation Ochsner et
al.
Table 1.Neuroimaging studies of reappraisal in healthy
individualsa
Design
Stimulus Timing of
Study Participants Goal Valence Tactic type reapp cue
Amygdala
Beauregard et al.204 HYA Dec Pos Dist Videos Early NoDomes et
al.81 HYA Both Neg Both Photos Late Yes
Eippert et al.84 HYA Both Neg Both Photos Late Yes
Erk et al.168 HC Dec Neg Dist Photos Early Yes
Goldin et al.66 HYA Dec Neg Reint Videos Early Yes
Harenski et al.205 HYA Dec Neg Both Photos Early Yes
Hayes et al.139 HYA Dec Neg Reint Photos Late Yes
Herwig et al.206 HYA Dec Both Reint Anticipate
photos
Early Yes
Hollmann et al.207 HYA Dec Pos (food) Reint Photos Early No
Ichikawa et al.82 HYA Both Neg
(errors)
Reint Task
errors
Early No
Kanske et al.127 HYA Dec Both Both Photos Late Yes
Kim et al.95 HYA Dec Both Reint Photos Early Inc pos
only
Kober et al.118 HYA
smokers
and non-
smokers
Dec Pos
(food/cigs)
Reint Photos Early Yes
Koenigsburg et al.199 HYA Dec Neg Dist Photos Early Yes
Krendl et al.208 HYA Dec Neg Unclear Photos Early Yes
Kross et al.112 HYA Dec Neg Reint Memories Late No
Lang et al.83 HC Both Neg Dist Scripts Early Inc only
Levesque et al.209 HYA Dec Neg Dist Videos Early NoMak et al.210
HYA Dec Both Unclear Photos Early NoMcRae et al.211 HYA Dec Neg
Reint Photos Early Yes
McRae et al.61 HYA Dec Neg Reint Photos Early Yes
McRae et al.180 Healthy aged
1022
Dec Neg Reint Photos Early No
McRae et al.15 HYA Dec Neg Reint Photos of
faces
Early Yes
Modinos et al.212 HYA Dec Neg Reint Photos Late Yes
New et al.85 HC Both Neg Reint Photos Late Yes
Ochsner et al.43 HYA Dec Neg Reint Photos Late Yes
Ochsner et al.76 HYA Both Neg Both Photos Early Yes
Ochnser et al.16 HYA Inc Neg Both Photos Early Yes
Ohira et al.115 HYA Dec Both Unclear Photos Early YesOpitz et
al.96 HYA and
HOA
Both Neg Reint Photos Late No
Phan et al.114 HYA Dec Neg Reint Photos Early YesPitskel et
al.86 Healthy aged
717
Both Neg Reint Photos Early Yes
Continued
E8 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
Table 1.Continued
Design
Stimulus Timing of
Study Participants Goal Valence Tactic type reapp cue
Amygdala
Schardt et al.213 HYA Dec Neg Dist Photos Early Yes
Schulze et al.87 HC Both Neg Both Photos Late No
Staudinger et al.214 HYA Dec Pos Dist Reward Early NoStaudinger
et al.215 HYA Dec Pos Dist Anticipate
reward
Early No
Urry et al.94 HOA Both Neg Reint Photos Late Inc only
Urry et al.216 HOA Both Neg Reint Photos Late Yes
van Reekum et al.88 HOA Both Neg Reint Photos Late Dec only
Vrticka et al.217 HYA Dec Both Reint Photos Early YesWager et
al.117 HYA Dec Neg Reint Photos Early Yes
Walter et al.218 HYA Dec Neg Dist Photos Early Yes
Winecoff et al.195 HYA and
HOA
Dec Both Dist Photos Late Yes
aStudies are ordered first by year and second by alphabetical
order.Only studies that reported contrasts (i.e., not only
functional connectivity or correlational analyses) for
psychologicallyhealthy individuals are included here. If a study
included a patient sample but still reported results for its
healthy adultcontrols separately, it was included.HYA, healthy
young adults, typically 1830 yrs old; HOA, healthy older adult,
typically aged 60 years or older; HC,healthy adult control
participants matched to patients; For design, Goal, goal pursued by
participants to increase ordecrease emotional responses; Valence,
Positive or negatively valenced emotional stimuli; Tactic, type of
reappraisalused, distancingor reinterpreting; StimType, stimulus
type; Timingof reapp cue, timingof instruction cue to
reappraiserelative to onset of stimulus, where early is just before
simulus onset and late is a few seconds after stimulus
onset;Amygdala, whether modulation of amygdala was reported.Note:
All studies used event-related designs (different types of trials
are presented in a randomized fashion so asto estimate responses on
a trial-by-trial basis) except the nine studies designated by in
the Timing of reapp cuecolumn, which indicates that they used a
block design (trials are blocked by type, such that many of one
type appearconsecutively). Also, for the stimulus-type column,
photo stimuli were drawn from the international affective
picturesystem111 unless otherwise specified. Goal: Dec, decrease;
Inc, increase; Both, both increase and decrease conditionswere
used; Valence: Neg, negative; Pos, positive; Both, both positive
and negative stimuli were used; Strategy: Both,both distancing and
reinterpreting were used (this only applies to Ref. 76), or
participants were given the choice ofdistancing or reappraising;
Dist, become more or less psychologically distant; Reint,
cognitively reinterpret; Unclear,unclear as to what tactic was
instructed.
decrease goals reliably modulate the amygdalasventral
(corresponding to the basal and lateralamygdala nuclei) and dorsal
portions (correspond-ing to the central nucleus) as well as the
sublentic-ular extended amygdala (SLEA30,91,92) that lies be-tween
the amygdala and the striatum. On the otherhand, increase goals may
modulate only the dorsalamygdala/SLEA. One speculative
interpretation ofthese data is that decrease goals influence
percep-tual and semantic inputs to the amygdala, whichcome through
the basolateral complex, whereas in-
crease goals influence the outputs of the amygdala,which flow
from the central nucleus.43,76 This hy-pothesis would fit with
anatomical data showingthat the basolateral complex has reciprocal
con-nections with ventrolateral PFC as well as tempo-ral and
parietal regions implicated in visuospatialand semantic
representation, whereas the centralnucleus receives inputs from
medial prefrontal re-gions and sends outputs to autonomic centers
thatimplement various components of an emotionalresponse.93
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E9
-
Functional imaging studies of emotion regulation Ochsner et
al.
The major caveat for all of these conclusions,however, is that
very few studies have examinedincrease goals, and as a consequence,
conclusionsabout the goal specificity of reappraisal-related
acti-vations must be considered tentative. That beingsaid, the
first study to directly compare increaseand decrease goals within
subjects obtained ex-actly the results described previously76both
in-creasing and decreasing negative affect recruited leftvlPFC and
dlPFC and modulated the dorsal amyg-dala/SLEA (increasing affect
increased amygdala ac-tivity, whereas decreasing negative affect
decreasedactivity), yet it was also revealed that increasing
neg-ative affect recruited the dmPFC to a greater degreethan did
decreasing negative affect and decreasingnegative affect recruited
right vlPFC andmodulatedventral amygdala to agreater extent thandid
increas-ing negative affect. At least two-thirds of
subsequentstudies comparing these goals have obtained resultsthat
are generally consistent with them76,84,87,9496
(other findings also have been reported, includ-ing increase
versus decrease differences only in theamygdala,81 striatal
modulation,76,88,94 and greaterright PFC activation for increasing
than decreas-ing84).
Tactic specificityIn the military, a distinction is commonly
made be-tween strategy and tactics. Strategy is the overallmeans by
which a goal (e.g., win the war) is to beachieved (e.g., divide and
conquer). Tactics are thespecific ways in which strategies are
implemented ina given circumstance (e.g., a quick infantry
advance,an airstrike, etc.). In the same way, one can distin-guish
between reappraisal as a strategy that involveschanging the meaning
of a stimulus and the tacticsused to implement that strategy.97
Two different reappraisal tactics have been stud-ied with
imaging.9,76 The first can be called reinter-pretation, which
involves changing ones interpre-tation of the elements of the
situation or stimulusthat elicits emotion. For example, if one is
presentedwith a photo of a sick man in the hospital that elic-its
feelings of sadness, one might reinterpret thisimage in a way that
decreases emotion by think-ing about the mans hearty constitution
and that hewill be healthy and well in the future. To
increaseemotion, one might instead think about how theman is in a
great deal of pain and may, in fact, getworse and even perish. The
second can be called
distancing , which involves changing ones personalconnection to,
or psychological distance from, thestimulus that elicits emotion.
In the example of thephoto of the sick man, one might decrease
emotionby viewing the image from the detached perspec-tive of an
objective, third person observer and/orimagining that the pictured
event took place a longtime ago or in a faraway location. One might
in-crease emotion by instead imagining that one is ex-periencing
pictured events in the present moment,from a first-person
perspective, which enables youto smell, hear, and directly observe
what is takingplace.As Table 1 shows, about twice as many
studies
have examined reinterpretation as have examineddistancing, with
a few allowing participants to en-gage in either tactic, and only a
single study directlycomparing them.76 Figure 3B plots peak
activationfoci for reappraisal studies of healthy individualsas a
function of reinterpretation versus distancingtactics.This figure
illustrates three conclusions that can
be drawn about reappraisal tactics. First, reinterpre-tation
seems todifferentially call uponventral lateralprefrontal regions
implicated in response selectionand inhibition.73,98,99 Presumably,
this reflects thefact that reinterpretation requires that one
mustlook up and select alternative meanings for stim-uli from
semantic memory to a greater extent thandoes distancing. Second,
distancing seems to recruitparietal regions implicated in spatial
attention andrepresentation to a greater extent, including
per-spective taking and the sense of agency.100103 Thismay reflect
the fact that distancing involves chang-ing the conceptual and
spatiotemporal perspectivefrom which stimuli are experienced.
Third, in gen-eral, the regions involved in reinterpretation
appearto be more strongly left lateralized in prefrontal
andtemporal cortices, whereas regions involved in dis-tancing
appear to be more strongly right lateral-ized in prefrontal cortex.
These patterns may reflectthe differential dependence of
reinterpretation anddistancing on linguistic and semantic processes
asopposed to spatial and attentional processes, whichgenerally
showa left versus right hemisphere patternof relative
specialization.76,104
Here again, however, because comparativelyfewer studies have
examined distancing, firmconclusions concerning the tactic
specificityof reappraisal-related activations await further
E10 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
research that directly test the conclusions drawnpreviously.
Valence specificityOn average, the impact of negative emotional
ex-periences seems to be greater than the impact ofpositive
emotional experiences, both in the shortand long term.105 Indeed,
problems with regulat-ing negative emotion are more often a
hallmark ofclinical disorders than are problems with
regulatingpositive emotion.106 As such, it is not surprising
thatTable 1 shows that the number of reappraisal studiesexamining
negative emotion outnumber those ex-amining positive emotion more
than three to one.That said, two conclusions can be drawn from
examining Figure 3C, which plots peak activationfoci for
reappraisal studies of healthy individualsas a function of the
negative versus positive va-lence of stimuli (and the emotions they
presumablyelicit). First, whereas reappraisal of both negativeand
positive stimuli depends upon left-hemisphereregions, reappraising
negative stimuli depends onright hemisphere regions as well. These
findingsmight reflect the fact that, to date, the majority
ofstudies of negative emotion involve decrease goals.As noted
earlier, decrease goals may require morecognitive resources than
increase goals, includingplacing greater demands on
selection/inhibitoryfunctions associated with right vlPFC.107109
Analternative explanation is that positive and neg-ative emotions
generally involve approach versusavoidance motivations, which have
been associ-ated with the left versus right prefrontal cortex.This
interpretation seems less likely, however, giventhat this
motivation-related prefrontal asymmetryis commonly observed in
EEG110 but not in fMRIstudies.19
Second, its apparent that reappraising negativestimuli typically
modulates activity in the amygdalaand less commonly activity in the
striatum. By con-trast, thehandful of studies examining reappraisal
ofpositive stimuli more commonly show modulationof the striatum,
including the ventral portions asso-ciated with reward and
reinforcement learning.33,34
These conclusions are again tentative, however,because so few
studies have examined reappraisalof positive stimuli and, in
general, studies of reap-praisal have focused overwhelmingly on
decreaserather than increase goals. As a consequence, it isnot yet
clear whether the patterns noted previously
are attributable to the pursuit of decrease versusincrease
goals, the use of negative versus positivestimuli, or both.
Stimulus specificityTodate, 33 out of the 43 reappraisal studies
shown inTable 1 have used photographic stimuli pulled fromthe
International Affective Picture System (IAPS).These stimuli have
been shown to reliably elicit ex-periential, physiological, and
facial expressive com-ponents of an emotional response in a
valence-specific manner.111 As such, they provide a
straight-forward means of eliciting affective reactions in
thescanner environment.That said, the emotions elicited by such
stim-
uli may or may not generalize to other contexts.For example,
IAPS photos are selected so as to benormatively positive or
negative.111 Although thisis suitable for many experimental
agendas, otherstimuli may be appropriate if one wants to exam-ine
the ability to reappraise specific emotions, theemotions elicited
by idiosyncratically self-relevantautobiographical
experiences,112,113 and so on.With these considerations in mind,
small num-
bers of studies have examined the ability to reap-praise the
specific emotions elicited by sad, sex-ual, or disgusting videos,
scripts that elicit partic-ular emotions, the recollection of
autobiographicalmemories, anticipation of reward or shock, or
thecommission of an error (see Table 1). Because so fewstudies have
used each of these stimuli, it is not use-ful at present to plot
activation foci for them or toattempt to draw conclusions about how
they mightdiffer as a function of stimulus type. It remains
forfuture research to directly address the question ofhow the
nature of the stimulus per se, as opposed tothe kind of emotion
elicited, influences the neuralsystems involved in reappraisal.
Pathways linking control and affectsystemsStudies of
reappraisaland more generally studiesof any form of emotion
regulationimplicitly orexplicitly assume that prefrontal regions
modulateemotional responses via their impact on affect sys-tems
like the amygdala and ventral striatum. Giventhe prevalence of this
assumption, it is somewhatsurprising that it has seldom been put to
a directtest. To be sure, a number of studies have
showncorrelations between prefrontal and amygdala
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E11
-
Functional imaging studies of emotion regulation Ochsner et
al.
Figure 4. Two kinds of mediation pathways involved in
reappraisal. (A) and (B) show pathways identified in two studies of
thedownregulation of negative emotion whereby dorsomedial or
ventrolateral prefrontal regions diminish amygdala responses
viatheir impact on ventromedial prefrontal cortex. These studies
did not report weights for the mediation paths between regions
ortest for full versus partial mediation. (C) and (D) show pathways
identified in two studies of the downregulation of negative
orpositive emotion whereby ventrolateral or dorsolateral prefrontal
regions diminish self-reports of negative affect or craving
viatheir impact on the amygdala or ventral striatum,
respectively.
activity76,95,114 or correlations between some mea-sure of
emotional response (typically self-report)and either
prefrontal76,114,115 or amygdala activ-ity.76,94,116 Only four
studies, however, have directlytested the mediation model implied
by the hypoth-esis that control systems influence emotional
re-sponse by influencing activity in affect systems (seeFig. 4).The
first two studies to use mediation exam-
ined the use of reappraisal to diminish responses tonegative
photos.94,116 Although both studies usedamygdala reactivity as
their measure of emotionalresponse, neither reported a main effect
of reap-praisal on diminishing amygdala activity.Motivatedby known
connections between the amygdala andvmPFC, both studies looked for
and found thatindividual differences in amygdala response
werecorrelated inversely with responses of vmPFC. Me-diation
analyses showed that vmPFC mediated a
relationship between either left dmPFC94 or leftvlPFC116 and the
amygdala (Fig. 4A and B) suchthat activity in these prefrontal
regions was posi-tively related with vmPFC activity, which, in
turn,was negatively related to amygdala responses. Thereare,
however, at least two qualifiers in interpret-ing these results.
First, the study94 identifying theleft dmPFC region did so in an
increase > attend(i.e., a no regulation baseline) > decrease
contrast,meaning that it generally is less active when de-creasing
negative emotion than when respondingnaturally in a baseline attend
condition. This sug-gests that to the extent one shows less
deactiva-tion when decreasing (relative to baseline), one willshow
greater activity in vmPFC, and, in turn, lessamygdala response. It
is not immediately clear howto interpret the reduced degree of
dmPFC deacti-vation in this context. Second, the study116
iden-tifying the vlPFCvmPFCamygdala pathway
E12 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
collapsed across activity in both amygdalae that wasextracted
from structural ROIs. As such, it is notclear whether the
prefrontal effects were strongerfor one amygdala or the other. That
said, whentaken together, these two studies suggest that effec-tive
reappraisal involves PFCvmPFCamygdalapathways.The second two
studies used similar analytic ap-
proaches to study either the use of reappraisal todiminish
responses to negative photos117 or forsmokers, the use of
reappraisal to diminish crav-ing elicited by photographs of
appetitive foodsor cigarettes.118 The study of negative
emotion117
showed that right vlPFC activity predicted dropsin self-reported
negative emotion, and that thisrelationship was independently
mediated by sep-arate pathways through the amygdala and the
ven-tral striatum (Fig. 4C). These two pathways weretaken to
reflect the use of reappraisal to mini-mize negative appraisals and
enhance positive reap-praisals, respectively (see also Table S1
from thatpaper, which shows left vlPFC involvement as well).The
study of craving118 showed that left dlPFCactivity predicted drops
in self-reported cravingvia modulation of activity in the ventral
stria-tum (Fig. 4D). Together, these two studies suggestthat
effective reappraisal involves a pathway linkingPFCsubcortexemotion
change, with the spe-cific elements of the pathway depending on the
na-ture of the stimulus and emotion involved.Why are there
differences between the results of
these pairs of studies? On one hand, because differ-ent
dependentmeasures of emotional responsewereused (amygdala response
vs. self-reported emotion),its possible that different reappraisal
pathways willemerge depending on the type of response. On theother
hand, its also possible that differences inmethodology may lead
participants to reappraisedifferently, and, in turn, recruit
different pathwaysfor effective emotion regulation. Here, two
differ-ences between the pairs of studies may be relevant.In the
first pair of studies that identified the
vmPFC-mediated pathway, both had participantsthat were up to 40
years older than the average par-ticipants in studies of
reappraisal in young adultsaged 6264 in one case94 and 1953
(average 33years) years in the other.116 Participants in the
sec-ond pair of studies were younger, as is the norm, av-eraging
22.3117 and 26.8118 years, respectively. Givenfindings that older
adults may be impaired in some
kinds of reappraisal; the lateral PFC thins, whereasvmPFC
thickens with age;119 and that even whennot told to regulate, older
adults can show greaterconnectivity between vmPFC and amygdala,120
it ispossible that the vmPFC plays a larger role in reap-praisal
for older compared with younger adults.Second, the first pair of
studies cued participants
to reappraise approximately 4 seconds into an ap-proximate
10-second presentation of an aversivephoto (a late cue), whereas
the secondpair presentedthe cue to reappraise just before onset of
aversivephotos (an early cue). The early cue method is in-tended to
provide participants with an opportunityto first have a
naturalistic emotional response to anaversive photo before they
begin to regulate and hasbeen used in 14 studies (see Table 1). The
late cuemethodmodels real-world situationswhere the goalto
reappraise comes online just as one encounters anemotionally
evocative stimulus and has been usedin 29 studies (see Table
1).Although the early cue method is analogous to
real-world situations where the goal to regulatecomes online
only after one already is having anemotional response, there is a
potential problemwith trying to model this in the lab. During the
ini-tial free viewing of an aversive photo, participantsmay try out
a few reappraisals just in case they areasked to subsequently
reappraise on that trial. If thiswere the case, then we might
expect one or both oftwo kinds of results.One possibility is that
the ability to detect an
effect of reappraisal on amygdala responses wouldbe diminished
for late cue studies, either becausethe amygdala responded early
and then habituated,or because once participants were asked to
reap-praise, the amygdalas response could have alreadydecreased
because the participants had begun gen-erating/practicing potential
reappraisals before theexplicit instruction cue appeared. Although
neitherof the mediation studies in question showed whole-brain
amygdala effects, and only one showed ef-fects using ROIs, weak
effects of reappraisal on theamygdala are probably related to other
factors (likeagesee above) given that roughly the same
ratio(roughly 2/3 to 3/4) of studies using the late andearly cue
methods show reappraisal-related amyg-dala modulation, especially
for studies using photos(see Table 1).A second possibility is that
the late versus early
timing of reappraisal cues changes the nature of
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E13
-
Functional imaging studies of emotion regulation Ochsner et
al.
ones reappraisals, even if, on average, they havesimilar effects
on amygdala responding. For exam-ple, in late cues studies, if
participants have had achance to view stimuli for a few seconds and
to thinkabout potential reappraisals before being explicitlytold to
reappraise, then vmPFC recruitment couldreflect decision processes
about which of a set ofprepared reappraisals they prefer and can
best usefor the stimulus at hand (see also section ondecisionmaking
below).To date, no imaging studies have com-
pared late and early cues. But one behavioral/psychophyiological
study comparing these cuesfound that the effects of increase goals
on somephysiologicalmeasures are greater for late than earlycues,
but that the effects of decrease goals weresimilar for each cue
type. Future work could fruit-fully illuminate these issues.All
this said, it is of course possible that both kinds
of pathways are important and that a
multistepvlPFCvmPFCamydala/striatumemotionresponse pathway may be
observed in futurestudies. To date, however, no published studies
haveexpressly tested for the existence of this complexpathway
underlying reappraisal success.
The role of perceptual and semanticsystemsA related issue is
whether and how reappraisal in-volves modulation not just of
systems involved inaffective appraisal and response, but of
systemsinvolved in representing the perceptual and se-mantic
properties of stimuli as well. As shown inFigure 3, activation of a
number of these systemsis often seen during reappraisal, including
regionsalong the middle and superior temporal sulci in-volved in
representing the visual properties of stim-uli, including nonverbal
social cues to emotion likemovements of lips and eyes;121123
temporal polarregions implicated in representing episodic and
se-mantic emotion knowledge;124 and regions near
thetemporalparietal junction involved in representa-tions of
beliefs, including false beliefs of the sortone generates when
considering alternative reap-praisals of stimuli.125,126
These data raise at least three questions. First,there is the
question of when activation of theseregions will be seen.
Certainly, cognitive changestrategies like reappraisal may involve
these regions,given that it involves an active reworking of the
meaning of a stimulus. Other strategies that do notfocus on
meaning may not involve these regions,however. Consistent with
this, two studies directlycomparing reappraisal and distraction
found thatreappraisal differentially recruited all three of
thetemporal regions listed above.61,127 Along theselines, it is
also likely that these regions will be moreinvolved in regulating
responses to visual stimuligiven the role of the temporal lobe in
the ventralvisual stream for representing information aboutobject
identity128130 (although this remains to betested directly). As
noted previously, there is not yetenough work using different kinds
of stimuli to saywhether reappraisal of stimuli in nonvisual
modal-ities (e.g., somatosensory or auditory) may involvemodulation
of corresponding modality-specific re-gions (e.g., somatosensory of
auditory cortices).Second, if these regions are more active
during
reappraisal, there is the question as to why this isthe case.
Does greater activity here reflect increasedattention to perceptual
and semantic aspects ofstimuli? Access to/retrieval of alternative
views ofreappraised stimuli? The process of actively restruc-turing
ones (visual) mental image of a stimulus?All three interpretations
are possible and could betested in future work.Third, there is the
question of whether these
temporal regions play a part in the regulationpathways described
earlierplaying an interme-diary role, for examplebetween prefrontal
con-trol systems and affective appraisal systems. Thispossibility
was raised in early reappraisal studies(e.g., Ref. 43) where it was
suggested that eventhough dorsolateral PFC regions do not have
directconnections to subcortical regions like the amyg-dala, they
may nevertheless modulate them viatheir impact on
perceptual/semantic systems. Inline with this view, PFC could
change ones men-tal representation of a stimuluss meaning fromthe
top down, and that representation of the reap-praised stimulus
would feed forward to the amyg-dala (and other structures that
trigger affective re-sponses). Because the amygdala now sees
thereappraised stimulus, its response changes. Al-though plausible,
this hypothesis has yet to bedirectly tested.
SummaryExtant data from functional imaging studies ofreappraisal
strongly support the MCCE depicted in
E14 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
Figure 2B. Although many questions remain to beaddressed about
how specific control systems mod-ulate specific affect systems as a
function of reap-praisal goals and tactics or various aspects of
stimuliand the emotions they elicit, a core controlaffectsystem
dynamic is now well established.
Generalizability to other forms of regulationGiven the
robustness of the MCCE (Fig. 2) inaccounting for reappraisal, the
question naturallyarises as to whether this model can be
generalizedto account for other types of emotion
regulationstrategies.As noted previously, the majority of
functional
imaging studies of emotion regulation have focusedon
reappraisal. That said, the other fourmain classesof emotion
regulation strategies diagrammed inFig-ure 2A have been targeted by
imaging studies tovarying degrees. Here, we consider each in
turn.
Situation selection and modificationThe two situation-focused
strategies, situation se-lection and situationmodification, have
received lit-tle attention thus far in human imaging research.
Asnoted earlier, this is at least partially attributable tothe
difficulty of devising appropriate lab paradigmsfor studying them.
The lone human imaging studyof situation selection builds on the
rodent literatureon avoidance conditioning. In a typical task, a
ratlearns to perform an action that allows it to avoidpresentation
of an aversive stimulus (e.g., Refs. 131and 132). In a human
analogue of this procedure,Delgado et al. found that avoidance
conditioning ac-tivates vlPFC and dlPFC control systems and
mod-ulates the amygdala.133 These findings provide aninitial
suggestion that situation selection may callsystems that maintain
regulatory goals and selectcontext-appropriate avoidance
responses.
Attentional deploymentBy contrast, studies of attentional
deployment havebeen relatively common, second in number only
tostudies of reappraisal. One set of these studies hasexamined the
use of selective attention to shift visualspatial attention away
from an affectively valencedstimulus or stimulus attribute and
toward a neutralone. Another set of these studies has examined
theuse of distraction to shift the focus of attention in-ward onto
some internally maintained mental rep-resentation (e.g., a relevant
working memory load,self generated stimulus-irrelevant thoughts, a
pleas-
ant mental image, and so on). As has been reviewedin detail
elsewhere,10,134 interpreting the findings ofboth of these kinds of
studies is clouded by threeissues. First, almost all of the studies
of selective at-tention, andmany studies of distraction, use
stimulithat donot elicit strong emotional responses, such asfacial
expressions of emotion. As such, these studiesare concernedwith the
regulationof evaluative judg-ment or perception rather than
affective respond-ing, per se. Second, when highly arousing and
affect-inducing stimuli are employed, they most often arestimuli
that cause physical pain. Although responsesto painful stimuli have
a strong negatively valencedaffective component, this componentmay
itself havea distinct neural signature because of its recruitmentof
dedicated pain-specific neural pathways.135,136 Assuch, it is an
empirical question whether the regu-lation of pain is similar to or
different from theregulation of negative affective responses more
gen-erally. Third, these studies are highly heterogenous,often
employing very different stimuli and meth-ods of controlling the
focus and level of attention,without a clear metric for assessing
the extent towhich attention has or has not been paid to a
givenaffective stimulus. Given these limitations, we referthe
reader to other reviews of this literature,137,138
although noting that they are generally consistentwith the model
depicted in Figure 2B, insofar asactivation of prefrontal systems
and modulation ofaffect systems (like the amygdala) is often (but
notalways) reported.
Response modulation. Finally, both behavioraland imaging studies
of response modulation have fo-cused on expressive
suppression,which is the abilityto hide behavioralmanifestations of
emotion.60 Thetwo imaging studies of expressive suppression
askedparticipants to suppress facial expressions of disgustelicited
by a film clip.66,139 Both found that expres-sive suppression not
only activated dorsolateral andventrolateral PFC regions associated
withmaintain-ing goals, response selection, and
inhibition,72,73,77
but also it increased activation of the insula, whichis involved
in triggering affective responses. Amyg-dala findings were more
mixed, however, with onestudy reporting increases66 and one
decreases139
in activity during suppression. Increases in in-sula and
amygdala fit with psychophysiologicalstudies, demonstrating that
expressive suppression
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E15
-
Functional imaging studies of emotion regulation Ochsner et
al.
boosts the autonomic component of emotionalresponding.60
In total, the available literature on emotion reg-ulation
strategies other than reappraisal is in somecases limited and in
other cases somewhat confus-ing, but in general supports the idea
that all emotionregulation strategies involve interactions
betweencognitive control and affect regions. Future neuro-maging
research must apply the same rigorous andthorough approach to these
other strategies that hasalready been applied to reappraisal.
Generalizability to other relatedphenomenaGiven the robustness
of the MCCE (Fig. 2B) in ac-counting for multiple forms of
regulation, a nat-ural next question is whether it can be
generallyapplied to other allied phenomena, such as
af-fective/emotional learning, decision making, andexpectancies.
These phenomena are typicallyconsidered in separate literatures,
but seemto involve related cognitiveaffective dynamics.Although
space limitationsprohibit an in-depthdis-cussion, here, we briefly
examine the broad applica-bility of the model in each of these
three cases.
Affective/emotional learning. At the outset of thispaper we made
a distinction between goal-directedforms of emotion regulation,
which are the focusof this review, and other behaviors that may
haveregulatory effects on emotion despite lacking a spe-cific goal
to do so. There are a number of formsof affective or emotional
learning that fit the lat-ter description. One of the most common
exam-ples is extinction of a conditioned fear response.In the
traditional fear conditioning paradigm,140
an animal learns that an ostensibly neutral stim-ulus, such as a
light (known as the conditionedstimulus or CS), predicts the
occurrence of an in-trinsically aversive stimulus, such as electric
shock(known as the unconditioned stimulus or UCS).Over time, the
repeated pairing of the light andshock lead the animal to respond
to the light itselfwith an anticipatory fear response. Elegant
animalstudies have shown that fear conditioning dependson
communication between input and output nu-clei of the
amygdala.140,141 Fear extinction involvesthe repeated presentation
of the CS in the absenceof the UCS.142 Over time, the organism
learns thatthe CS no longer predicts shock, ceases to have its
anticipatory fear response, and fear is said to be
ex-tinguished. Importantly, extinction is known to in-volve the
laying down of a new context-dependentmemory.143 In the current
temporal context, the CSdoes not predict shock, whereas in the past
temporalcontext it did. Rodent lesion studies have shown
thatwhereas the initial acquisition of extinction requiresonly the
amygdala, the ability to retain and expressmemory for extinction
depends on vmPFC.142 Inkeeping with this finding, studies in humans
haveshown thatboth themagnitudeof vmPFCactivationand vmPFC
thickness predict the speed of extinc-tion.144146
In the present model, phenomena like extinction(or
stimulusreward reversal learning, which alsodepends upon
vmPFC147,148), are somewhat hybridphenomena. On the one hand, they
can be viewedas an example of emotion generation, insofar as oneis
learning to express a new emotional response toa given stimulus. On
the other hand, they can beviewed as an implicit form of emotion
regulationwhere one does not have an explicit goal to regulate,but
the behaviors inwhichone engages directly, alterthe nature of ones
emotional response.Beyond this, there are a number of ways in
which
prefrontal control systems may have a regulatoryimpact on
affective learning. For example, as notedearlier, in some cases
reappraisal may involve in-teractions between PFC, vmPFC, and the
amyg-dala, when reappraisal paradigms give participants achance to
respond emotionally and potentially planreappraisals before
deciding whether to implementthem. Interactions of this sort also
have been ob-served in studies that use distraction to regulate
aconditioned response.149,150 In these studies, one isinitially
conditioned to expect either a painful shockor reward UCS following
a visual CS (e.g., a yellowtriangle). Later, one regulates the
conditioned re-sponse to the CS by thinking about a calm and
neu-tral scene unrelated to either the CS or the UCS. Inboth cases,
effective regulation involves activationof left dlPFC and
modulation of both the amygdalaand/or ventral striatum and the
vmPFC.
Affectivedecisionmaking. Affectivedecisionmak-ing involves
choosing among several stimuli thatonemay purchase, consume, or
own. In some cases,these choices are a simple matter of selecting
theoption that has the greatest value. Imaging re-search suggests
that activation in systems thought to
E16 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
represent affect and value like the ventral striatum,insula, and
vmPFC is sufficient to support and evenpredict such choices.151,152
But in other cases, thechoice options may be of similar value, or
the rea-sons for valuing themmay conflictwithone another.In the
model, such cases may draw on the controlsystems shown in Figure 2B
to modulate the valuesassociated with choice options, essentially
guiding atopdown revaluation of them to facilitate choice.Perhaps
the simplest example of this is where the
act of choice itself arouses conflict as one decideswhich
features of choice options they cant livewith-out and which
features of choice options they mustforgo. Classically, this
decision conflict is thought toarouse cognitive dissonance, which
the act of choos-ing reduces by placing a higher value on chosenand
a lesser value on unchosen stimuli.153 Imagingstudies show that
these choice-induced changes invalue involve control systems like
the anterior cingu-late cortex, which may signal the presence of
choiceconflict andmotivate value change, and systems likethe
ventral striatum, whichmay represent the reval-ued
stimuli.153156
Another type of choice that commonly requiresthe use of control
occurs when an individual mustdecide between options that fit
short-term versuslong-term goals. This is the dilemma faced by
adieter who must decide whether to eat a cupcakeor an apple.
Consuming the cupcake satisfies theshort-term goal of hedonic
pleasure, whereas eat-ing the apple satisfies the long-term goal of
liv-ing a healthy lifestyle. A recent imaging study47
of this choice dilemma showed that food choicesreflecting a
greater valuation of long-term healthover short-term tastiness
involve the modulation ofvmPFC by dlPFC. This is consistent with
the ideathat the cognitive control of choice involves interac-tions
between systems for maintaining choice goals(e.g., dlPFC) and
systems representing the value ofchoices with respect to those
goals (e.g., vmPFC).This same logic applies to studies of
intertem-
poral choice and delay of gratification,157 whereimaging158,159
and transcranial magnetic stimula-tion (TMS)160 studies suggest
that lateral PFC con-trol systems can be used to effortfully
represent thevalue of a larger delayed reward and guide selec-tion
of it over a smaller but immediately availablereward.More
generally, themodel can be applied to other
choices where control is needed to modulate the af-
fective valuations placed on choice options, rangingfrom risky
decisionmaking161 to interpersonal con-texts in which one must
decide whether to be fairtoward or punish others.162,163
Affective expectancies. In parallel to the growthand development
of imaging research on emotionregulation, there has been a
tremendous surge ofinterest in the brain mechanisms underlying the
in-fluence of expectancies on behavior.164 In imaging,expectancies
have been studied either by cueing par-ticipants that an upcoming
stimulus will have par-ticular properties (e.g., that it will or
will not bepainful, will be a neutral or aversive image, and soon)
or by inducing beliefs about the effects of aplacebo drug on their
experience (e.g., that an anal-gesic cream will reduce pain).In
themodel, these phenomena all involve the use
of prefrontal control systems to set and maintain
anexpectation,which, in turn, influences the responsesof affect
generating systems. For example, imagingstudies show that
expectancies and placebo beliefsabout pain activate lateral
prefrontal/parietal con-trol systems and/or medial prefrontal
systems164167
that may maintain expectations about upcomingevents. In turn,
these systems may influence theway one attends to and appraises the
meaning ofexpected stimuli, thereby increasing or
decreasingactivity in affect systems to be consistent with
thenature of ones expectations.
Summary and future directions for basicand translational
researchThe overarching goal of this paper has been to re-view and
synthesize current functional imaging re-search on emotion
regulation. Toward that end, weoutlined a basic model of the
processes and neuralsystems involved in emotion generation and
emo-tion regulation and surveyed various domains ofresearch that
support it. At its core, theMCCE spec-ifies how prefrontal and
cingulate control systemsmodulate activity in affect systems as a
function ofones regulatory goal, tactics, and the nature of
thestimuli and emotions being regulated. Although themodel was
built primarily from studies of one typeof cognitive change
strategy known as reappraisal,it is generally applicable to
understanding the brainmechanisms underlying the other emotion
regula-tion strategies depicted in Figure 2A as well as arange of
other allied phenomena.
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E17
-
Functional imaging studies of emotion regulation Ochsner et
al.
That said, there is much work yet to be done. Atvarious points
during the review weve highlightedthe limitations of current
knowledge and the short-comings of current methodologies. Future
work isneeded to clarify the mechanisms underlying all ofthe
emotion regulation strategies discussed here aswell as the roles
the brain systems supporting emo-tion regulation (Fig. 2B) play in
affective learning,affective decision making, and affective
expectan-cies. Progress is essential to refine our understand-ing
of the distinctionsmade here, but also to addressnewquestions about
howemotion regulationmech-anisms operate. For example, although it
is certainlyimportant that regulation strategies have
immediateeffects on emotional responses, it is also importantthat
their effects be long lasting. Indeed, whetherregulatory effects
last is critical both in everyday andclinical contexts inwhich one
could repeatedly reen-counter an emotionally evocative stimulus
(e.g., therisk of running into a girlfriend who dumped youbecause
you work for the same company). To date,this issue has been
addressed only twiceonce inan fMRI study168 reporting that the
effects of reap-praisal on diminishing amygdala responses may
en-dure for up to 40 minutes in healthy adults, but notthose with
major depression, and once in an ERPstudy169 showing that the
effects of reappraisal onarousal-related responses endure for up to
30 min-utes. Clearly, more work is needed here.In so doing, it will
be important for this work
to increasingly make use of techniques other thanfunctional
imaging (e.g., ERP,169177 TMS,160,178
and lesion methodologies179), as well as to in-tegrate insights
gained from human studies withthe large body of literature on
affective and reg-ulatory phenomena in nonhuman primates
androdents.39,142,144 Progress on all of these fronts isabsolutely
critical if we are to develop a model ofinteractions between
control and affect systems thatcan make sense not just of emotion
regulatory phe-nomena, but of all the other types of phenomenathat
recruit these systems as well.Another important direction for
future research
is the translation of basic research of emotion regu-lation to
understanding the full range of normal toabnormal differences in
emotional responding andregulatory ability. This is critical both
for under-standing the mechanisms underlying this variabil-ity and
for testing the boundaries of basic models ofemotion regulatory
mechanisms.
One domain in which this will prove importantis understanding
how and why our emotional livesevolve as we grow from childhood
through adoles-cence into adulthood and old age. On one hand,there
is growing evidence that childhood and ado-lescence are critical
times for the development ofthe emotion regulatory abilities needed
to adap-tively regulate affective impulses and the
deleterioushealth behaviors (e.g., obesity, substance use)
theycanpromote.A small but growingnumberof studieshave begun to
address this issue by asking how theneuralmechanisms of reappraisal
and emotional re-activity develop from adolescence into young
adult-hood. Some early results suggest that reappraisalability
increases linearly with age, whereas emo-tional reactivity remains
relatively constant180,181
(but see Ref. 182). On the other hand, althoughphysical health
and cognitive abilities tend todeclinewith age,183185 older adults
report more emotionalstability and a greater ratio of positive to
negativeexperiences in their daily life, with the extent of
pos-itive emotion predicting longevity.186188 Althoughmany have
hypothesized that this rosy glow of oldage is due in part to more
effective emotion regula-tion, to date there is little evidence
directly testingthis idea.187,189,190 One conundrum to resolve
herewill be the apparent dependence of emotion regula-tion on the
same kinds of prefrontal control systemsthat declinewith age. This
raises the question of howregulatory abilities improve as
theunderlyingneuralmachinery declines.191,192 Early results suggest
thatit may depend on the strategies older adults deploy,with spared
or greater regulatory ability shown forstrategies and tactics that
fit with long-term goalsand have become
habitual.9,96,190,193196
A second important goal for translational re-search will be to
understand how potential dysfunc-tion in the mechanisms of emotion
generation andregulation may underlie various forms of psychi-atric
and substance use disorders (for a more indepth discussion, please
see Ref. 197). This trans-lational direction is being pursued in
studies ofreappraisal across various disorders, ranging
fromdepression116,168,198 to borderline personality
disor-der,83,87,199 social anxiety disorder,200,201 phobia,202
posttraumatic stress disorder,83,85 cocaine users,203
and smokers.118 These studies can be useful in twoways. First,
theymay showdisorder-specificpatternsof altered function in control
and affect systems.For example, current data suggest that
depressed
E18 Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York
Academy of Sciences.
-
Ochsner et al. Functional imaging studies of emotion
regulation
individuals may show impaired recruitment ofvlPFC during
reappraisal,116 suggestive of an im-pairment of topdown control,
whereas border-line individuals may show heightened
amygdalaresponses coupled with diminished cingulate re-sponses,
suggestive of a failure to monitor para-doxical increases in
affective responding when at-tempting to decrease emotion.199
Second, imagingmethods for studying emotion regulation may beused
before and after treatment regimes as pre-dictors of and markers of
improvement. Althoughsuch studies are only beginning to emerge,
they holdgreat promise for understanding why some individ-uals
improve andwhether different treatments (e.g.,drugs versus
cognitive behavioral therapy) have dif-ferent mechanisms of
action.In the long run, the hope is that integrating ba-
sic and translational perspectives will help specifywhich
individuals are at greatest risk for maladap-tive health behaviors
and emotional outcomes, atwhat ages this risk is greatest, and
which regulatorymechanisms could be targeted in future
interven-tions during particular points in the life course.
Al-though realization of this dream is still a long wayaway,
current research provides a strong foundationfor getting there.
AcknowledgmentsPreparation of this paper was supported by
NIHGrants AG039279, MH076137, and DA022541awarded to Kevin N.
Ochsner, as well as fellowshipF31MH094056 awarded to Jennifer A.
Silvers.
Conflicts of interestThe authors declare no conflicts of
interest.
References1. Eisenberg, N. 2000. Emotion, regulation, and moral
devel-
opment. Annu. Rev. Psychol. 51: 665697.2. Fernandez, E. &
D.C. Turk. 1989. The utility of cogni-
tive coping strategies for altering pain perception: a
meta-analysis. Pain 38: 123135.
3. Diener, E. 1984. Subjective well-being. Psychol. Bull.
95:542575.
4. Wilson, B.J. & J.M. Gottman. 1996. Stress, Coping, and
Re-siliency in Children and Families. E.M. Hetherington &
E.A.Blechman, Eds.: 189228. Lawrence Erlbaum.Hillsdale, NJ.
5. Ayduk, O. et al. 2000. Regulating the interpersonal
self:strategic self-regulation for coping with rejection
sensitiv-ity. J. Pers. Soc. Psychol. 79: 776792.
6. Gross, J.J. & R.F. Munoz. 1995. Emotion regulation
andmental health. Clin. Psychol.: Sci. Prac. 2: 151164.
7. Mittal, D., R. Torres, A. Abashidze & N. Jimerson.
2001.Worsening of post-traumatic stress disorder symptomswith
cognitive decline: case series. J. Geriatr. PsychiatryNeu-rol. 14:
1720.
8. Gross, J.J. & R.A. Thompson. 2007. Handbook of
EmotionRegulation. Guilford Press. New York.
9. Ochsner, K.N. & J.J. Gross. 2008. Cognitive emotion
reg-ulation: insights from social cognitive and affective
neuro-science. Curr. Dir. Psychol. Sci. 17: 153158.
10. Ochsner, K.N. & J.J. Gross. 2005. The cognitive control
ofemotion. Trends Cogn. Sci. 9: 242249.
11. Marr, D. 1982. Vision: A Computational Investigation intothe
Human Representation and Processing of Visual Informa-tion. W.H.
Freeman. San Francisco.
12. Ochsner, K. 2007. Social Psychology: A Handbook ofBasic
Principles. A. Kruglanksi & E.T. Higgins, Eds.: 3966.Guilford
Press. New York.
13. Ochsner, K.N. & S.M. Kosslyn. 2012. The Oxford
Handbookof Cognitive Neuroscience. K.N. Ochsner & S.M.
Kosslyn,Eds. Oxford University Press. New York.
14. Gross, J.J. & L.F. Barrett. 2011. Emotion generation
andemotion regulation: one or two depends on your point ofview.
Emot. Rev. 3: 816.
15. McRae,K., S.Misra,A.K.Prasad, et al.
2012.Bottom-upandtop-down emotion generation: implications for
emotionregulation. Soc. Cogn. Affect Neurosci. 7: 253262.
16. Ochsner, K.N. et al. 2009. Bottom-up and top-down pro-cesses
in emotion generation: common and distinct neuralmechanisms.
Psychol. Sci. 20: 13221331.
17. Barrett, L.F., B. Mesquita, K.N. Ochsner & J.J. Gross.
2007.The experience of emotion. Annu. Rev. Psychol. 58: 373403.
18. Scherer,K.R., A. Schorr&T. Johnstone, Eds.
2001.AppraisalProcesses in Emotion: Theory, Methods, Research.
OxfordUniversity Press. New York.
19. Wager, T.D. et al. 2008.TheHandbook of Emotion.M.
Lewis,J.M.Haviland-Jones&L.F. Barrett, Eds.:
249271.Guilford.New York.
20. Kober, H. et al. 2008. Functional grouping and
cortical-subcortical interactions in emotion: ameta-analysis of
neu-roimaging studies. Neuroimage 42: 9981031.
21. Mauss, I.B., R.W. Levenson, L. McCarter, et al. 2005. Thetie
that binds? Coherence among emotion experience, be-havior, and
physiology. Emotion 5: 175190.
22. Kalisch, R. 2009. The functional neuroanatomy of
reap-praisal: time matters. Neurosci. Biobehav. Rev. 33:
12151226.
23. Cunningham,W.A., J.J. Van Bavel & I.R. Johnsen. 2008.
Af-fective flexibility: evaluative processing goals shape amyg-dala
activity. Psychol. Sci. 19: 15260.
24. Cunningham, W.A., N.L. Arbuckle, A. Jahn, et al.
2011.Aspects of neuroticism and the amygdala: chronic tuningfrom
motivational styles. Neuropsychologia 48: 33993404.
25. Phelps, E.A. 2006. Emotion and cognition: insights
fromstudies of the human amygdala. Annu. Rev. Psychol. 57:2753.
26. Hariri, A.R. & P.J. Whalen. 2011. The amygdala: inside
andout. F1000 Biol. Rep. 3: 2.
27. Davis, M. & P.J.Whalen. 2001. The amygdala: vigilance
andemotion.Mol. Psychiatry 6: 1334.
Ann. N.Y. Acad. Sci. 1251 (2012) E1E24 c 2012 New York Academy
of Sciences. E19
-
Functional imaging studies of emotion regulation Ochsner et
al.
28. Anderson, A.K. et al. 2003. Dissociated neural
represen-tations of intensity and valence in human olfaction.
Nat.Neurosci. 6: 196202.
29. Neta, M. & P.J. Whalen. 2011. The primacy of
negativeinterpretations when resolving the valence of
ambiguousfacial expressions. Psychol. Sci. 21: 901907.
30. Whalen, P.J. et al. 2004. Human amygdala responsivity
tomasked fearful eye whites. Science 306: 2061.
31. Vuilleumier, P. & G. Pourtois. 2007. Distributed and
inter-active brain mechanisms during emotion face
perception:evidence from functional neuroimaging.
Neuropsychologia45: 1741794.
32. Knutson, B. & J.C. Cooper. 2005. Functional magnetic
res-onance imaging of reward prediction. Curr. Opin. Neurol.18:
411417.
33. ODoherty, J.P. 2004. Reward representations and
reward-related learning in the human brain: insights from
neu-roimaging. Curr. Opin. Neurobiol. 14: 769776.
34. Schultz, W. 2007. Multiple dopamine functions at
differenttime courses. Annu. Rev. Neurosci. 30: 259288.
35. Cunningham, W.A., I.R. Johnsen & A.S. Waggoner.
2011.Orbitofrontal cortexprovides cross-modal valuationof
self-generated stimuli. Soc. Cogn. Affect Neurosci. 6: 286293.
36. Ongur, D., A.T. Ferry & J.L. Price. 2003. Architectonic
sub-division of the human orbital andmedial prefrontal cortex.J.
Comp. Neurol. 460: 425449.
37. Price, J.L. 1999. Prefrontal cortical networks related to
vis-ceral function andmood.Ann.N.Y.Acad. Sci.877:383396.
38. Davachi, L. 2006. Item, context and relational episodic
en-coding in humans. Curr. Opin. Neurobiol. 16: 693700.
39. Murray, E.A., J.P.ODoherty&G. Schoenbaum. 2007.Whatwe
know and do not know about the functions of the or-bitofrontal
cortex after 20 years of cross-species studies.J. Neurosci. 27:
81668169.
40. Fellows, L.K. 2011. Orbitofrontal contributions to
value-based decisionmaking: evidence from humans with frontallobe
damage. Ann. N. Y. Acad. Sci. 1239: 5158.
41. Schoenbaum,G., Y. Takahashi, T.L.
Liu&M.A.McDannald.2011. Does the orbitofrontal cortex signal
value? Ann. N. Y.Acad. Sci. 1239: 8799.
42. Rudebeck, P.H. & E.A. Murray. 2011. Balkanizing the
pri-mate orbitofrontal cortex: distinct subregions for compar-ing
and contrasting values.Ann. N. Y. Acad. Sci. 1239: 113.
43. Ochsner, K.N., S.A. Bunge, J.J. Gross & J.D. Gabrieli.
2002.Rethinking feelings: an FMRI study of the cognitive
regu-lation of emotion. J. Cogn. Neurosci. 14: 12151229.
44. Schoenbaum,G.,M.P. Saddoris&T.A. Stalnaker. 2007.
Rec-onciling the roles of orbitofrontal cortex in reversal
learningand the encoding of outcome expectancies.Ann. N. Y.
Acad.Sci. 1121: 320335.
45. Roy, M., D. Shohamy & T.D. Wager. 2012.
Ventromedialprefrontal-subcortical systems and the generation of
affec-tive meaning. Trends Cogn. Sci. 16: 147156.
46. Oya,H. et al. 2005. Electrophysiological correlates of
rewardprediction error recorded in the human prefrontal
cortex.Proc. Natl. Acad. Sci. U.S.A. 102: 83518356.
47. Hare, T.A., C.F. Camerer & A. Rangel. 2009. Self-control
indecision-making involves modulation of the vmPFC valu-ation
system. Science 324: 646648.
48. Beer, J.S., E.A.Heerey,D.Keltner, et al. 2003. The
regulatoryfunction of self-conscious emotion: insights from
patientswith orbitofrontal damage. J. Pers. Soc. Psychol. 85:
594604.
49. Damasio, A. 1994. Descartes Error: Emotion, Reason, andthe
Human Brain. G.P. Putnam. New York.
50. Mufson, E.J.&M.M.Mesulam. 1982. Insula of the
oldworldmonkey. II: Afferent cortical input and comments on
theclaustrum. J. Comp. Neurol. 212: 2337.
51. Craig, A.D. 2009.Howdo you feelnow?The anterior insulaand
human awareness. Nat. Rev. Neurosci. 10: 5970.
52. Wager, T.D. & L. Feldman Barrett. 2004. From af-fect to
control: functional specialization of the insulain motivation and
regulation. URL www.affective-science.org/pubs/2004/Wager Edfest
submitted copy.pdf[accessed 20 August 2012].
53. Calder, A.J., A.D. Lawrence&A.W. Young. 2001.
Neuropsy-chology of fear and loathing.Nat. Rev. Neurosci. 2:
352363.
54. Augustine, J.R. 1996. Circuitry and functional aspects of
theinsular lobe in primates including humans. Brain Res. Rev.22:
229244.
55. Nitschke, J.B., I. Sarinopoulos, K.L.Mackiewicz, et al.
2006.Functional neuroanatomy of aversion and its
anticipation.Neuroimage 29: 106116.
56. Critchley, H.D., S. Wiens, P. Rotshtein, et al. 2004.
Neuralsystems supporting interoceptive awareness. Nat. Neurosci.7:
189195.
57. Williams, L.E., J.A. Bargh, C.C. Nocera & J.R. Gray.
2009.The unconscious regulation of emotion: nonconsciousreappraisal
goals modulate emotional reactivity. Emotion9: 847854.
58. Gross, J.J. 1998. The emerging field of emotion
regulation:an integrative review. Rev. Gen. Psychol. 2: 271299.
59. Davis, J.I., J.J. Gross & K.N. Ochsner. 2009.
Psychologicaldistance and emotional experience: what you see is
whatyou get. Emotion 11: 438444.
60. Gross, J.J. 1998.Antecedent- and response-focused
emotionregulation: divergent consequences for experience,
expres-sion, and physiology. J. Pers. Soc. Psychol. 74: 224237.
61. McRae, K. et al. 2010. The neural bases of distraction
andreappraisal. J. Cogn. Neurosci. 22: 248262.
62. Urry, H.L. 2009. Using reappraisal to regulate
unpleasantemotional episodes: goals and timing matter. Emotion
9:782797.
63. Jackson, D.C., J.R. Malmstadt, C.L. Larson & R.J.
David-son. 2000. Suppression and enhancement of emotional
re-sponses to unpleasant pictures. Psychophysiology 37: 515522.
64. Ayduk, O. & E. Kross. 2008. Enhancing the pace of
recov-ery: self-distanced analysis of negative experiences
reducesblood pressure reactivity. Psychol. Sci. 19: 229231.
65. Kross, E. & O. Ayduk. 2008. Facilitating adaptive
emotionalanalysis: distinguishing distanced-analysis of depressive
ex-periences from immersed-analysis and distraction. Pers.Soc.
Psychol. Bull. 34: 924938.
66. Goldin, P.R., K. McRae, W. Ramel & J.J. Gross. 2008.
Theneura