the impact of orbital prefrontal cortex damage on ...narobert/he impact of orbital prefrontal...

Cognitive Affective amp Behavioral Neuroscience2004 4 (3) 307-316

The orbital region of the prefrontal cortex (the orbito-frontal cortex) has been implicated in the representationof emotional information and the regulation of emotionalprocesses The medial region of the orbitofrontal cortexin particular is thought to ldquoinfluence the affective lsquotonersquoof behaviorrdquo (Keay Clement amp Bandler 2000 p 327)Hypoactivation of the orbitofrontal cortex is associatedwith such emotional states and behaviors as apathy in-

difference poor planning and irritability (Horne 1993)Damage to the orbitofrontal cortex results in socio-emotional deficits including emotional outbursts impul-sivity risk taking difficulty with goal-directed behaviorand a failure to abide by social rules and norms (BecharaDamasio amp Damasio 2000 Hartikainen Ogawa ampKnight 2000 Kolb amp Whishaw 1990 Tucker Luu ampPribram 1995) In particular an increasing body of evi-dence suggests that the orbitofrontal cortex helps link in-ternal emotional cues such as physiological arousalwith environmental cues such as reinforcement or pun-ishment (Bechara Damasio amp Damasio 2000 Rolls2000) For example patients with damage to the ventro-medial region of the orbitofrontal cortex fail to recruit

307 Copyright 2004 Psychonomic Society Inc

This research was supported by NINDS Grants NS 21135 and PONS40813 to RTK and by NIA Grants AG19724 and AG17766 toRWL Correspondence concerning this article should be addressed toN A Roberts Department of Psychology University of WisconsinP O Box 413 Milwaukee WI 53201 (e-mail nrobertsuwmedu)

The impact of orbital prefrontal cortex damage onemotional activation to unanticipated and

anticipated acoustic startle stimuli

NICOLE A ROBERTSUniversity of Wisconsin Milwaukee Wisconsin

JENNIFER S BEERUniversity of California Davis California

KELLY H WERNER University of California Berkeley California

DONATELLA SCABINIUniversity of California Berkeley California

and Helen Wills Neuroscience Institute Berkeley California

SARA M LEVENSNew York University New York New York

ROBERT T KNIGHTUniversity of California Berkeley California

and Helen Wills Neuroscience Institute Berkeley California

and

ROBERT W LEVENSONUniversity of California Berkeley California

Damage to the orbital prefrontal cortex has been implicated in selectively diminishing electrodermalautonomic nervous system responses to anticipated punishing stimuli (eg losing money BecharaDamasio amp Damasio 2000) but not to unanticipated punishing stimuli (eg loud noises DamasioTranel amp Damasio 1990) We extended this research by examining the effects of orbitofrontal damageon emotional responses to unanticipated and anticipated acoustic startles and collecting a more ex-tensive set of physiological measures emotional facial behavior and self-reported emotional experi-ence Consistent with previous research patients showed intact physiology to an unanticipated startlebut failed to show appropriate anticipatory cardiovascular responses (patientsrsquo heart rates decreasedcontrolsrsquo increased) In addition patients displayed more surprise facial behavior and reported mar-ginally more fear than did controls in response to the unanticipated startle Thus orbitofrontal dam-age may compromise the ability to anticipate physiologically the onset of aversive stimuli despite in-tact or enhanced emotional responses when such stimuli occur unexpectedly

308 ROBERTS ET AL

anticipatory physiological responses that may guide de-cision making (Damasio 1994)

In the present study we examined the effects of ventro-medial orbitofrontal cortex damage on emotional activa-tion in response to powerful emotion-eliciting stimuli(acoustic startles) This study builds on previous re-search by examining multiple components of emotionincluding autonomic (electrodermal and cardiovascular)and somatic nervous system physiology emotional ex-pressive behavior and self-reported subjective emo-tional experience and by studying both unanticipatedand anticipated variants of the same stimulus Examin-ing multiple measures of emotion is necessary becausedifferent components of emotion and different measureswithin each component can reveal different patterns offindings (eg Boyce et al 1993 Gross amp Levenson1993 Lacey amp VanLehn 1952) In particular neurolog-ical disturbance as evidenced by neurological and clin-ical populations may be related to dissociations amongemotion components (Kring amp Neale 1996 Robertset al 2001)

The Role of the Orbitofrontal Cortex in EmotionSeveral theories of orbitofrontal functioning have been

advanced to explain the frequent observation of pro-nounced socio-emotional deficits (eg disinhibited so-cial behavior or detrimental decision-making strategies)among patients with damage to this region

Filtering irrelevant emotional information Shima-mura (2000) has proposed that the orbitofrontal cortexhelps regulate emotional behavior by gating or filteringout irrelevant sensory information that would otherwisebe translated into emotional information Through thistype of gating mechanism the orbitofrontal cortex fil-ters the influence of emotional responses on cognitionIn patients with orbitofrontal damage however the abil-ity to appropriately gate irrelevant information is inhib-ited Thus according to Shimamurarsquos theory orbito-frontal patients have difficulty identifying and attendingto relevant emotional information and therefore showheightened or inappropriate emotional reactions

Forming and reversing stimulusndashresponse associa-tions Rolls (2000) has posited that the orbitofrontal cor-tex plays a role in emotion by helping individuals learnand unlearn stimulusndashresponse associations Researchwith both animals and humans suggests that the orbito-frontal cortex assigns meaning to reinforcing cues in theenvironment (Rolls 2000 Schoenbaum Chiba amp Gal-lagher 2000 Schoenbaum amp Setlow 2001) For examplethe orbitofrontal cortex is activated when associationsbetween visual stimuli and taste rewards (or aversivetastes) are learned and then unlearned (ie reversedRolls 2000) Thus according to Rolls in situations inwhich reinforcement contingencies rapidly change suchas in social environments patients with orbitofrontaldamage may fail to adjust their behavior appropriately tothe demands of the situation

Generating anticipatory emotional signals Dama-sio (1994) has proposed that the orbitofrontal cortex is

involved in the generation of physiological reactions thatappropriately direct behavior In a series of studies Dama-sio and colleagues demonstrated that neurologically intactsubjects show greater electrodermal activity (ie skinconductance responses) when they are about to make arisky or potentially disadvantageous decision (eg se-lecting a card from a disadvantageous deck see BecharaDamasio amp Damasio 2000 for a review) Damasio sug-gests that these anticipatory physiological responses pro-vide cues or ldquohunchesrdquo that can be used to guide be-havior Patients with damage to the ventromedial regionof the orbitofrontal cortex however fail to show suchanticipatory physiological responses As a result pa-tients are more apt to engage in risky behavior and tomake poor decisions (Bechara Damasio amp Damasio2000 Rahman Sahakian Cardinal Rogers amp Robbins2001)

Orbitofrontal Cortex and Emotional ActivationDespite theoretical and empirical evidence suggesting

the role of the orbitofrontal cortex in emotion-relatedprocesses (eg filtering irrelevant information learningand unlearning stimulusndashresponse associations or antic-ipating stimuli with uncertain outcomes) less is knownabout the impact of orbitofrontal cortex damage on emo-tional activation (ie the patterned physiological ex-pressive and subjective responses activated to help theorganism respond efficiently to changes in the internaland external environment that have significance for goalsand well-being Levenson 2001) There is some evidenceto suggest that damage to the orbitofrontal cortex does notaffect the physiological aspect of emotional activation tosimple stimuli such as an unanticipated acoustic startle(Damasio 1994) that do not require extensive cognitiveprocessing Because previous studies have not systemati-cally investigated unanticipated and anticipated versions ofsuch simple stimuli we studied the impact of orbitofrontaldamage on emotional activation in response to acousticstartle stimuli that were either unanticipated or anticipatedTo ensure a more comprehensive assessment of emotionwe sampled from multiple domains including cardio-vascular electrodermal and somatic physiology expres-sive facial behavior and subjective emotional experience

The Use of Acoustic StartleResponses to acoustic startle stimuli are thought to

exist on the boundary between reflex (eg defensive re-actions) and emotion (Ekman Friesen amp Simons 1985)For this reason startle stimuli are useful for investigat-ing the role of the orbitofrontal cortex in emotion pro-viding a foundation for subsequent investigations ofmore complex emotional cognitive and social processes

We presented orbitofrontal patients and controls withbrief loud bursts of white noise of sufficient intensity(115 dB) to be unambiguously noxious These startlestimuli elicit stereotyped combinations of physiologicalbehavioral and subjective responses including primarydefensive behaviors (eg ducking onersquos head) and sec-ondary emotional responses (eg fear and surprise Ek-

ORBITOFRONTAL CORTEX AND EMOTION 309

man et al 1985 Landis amp Hunt 1939) As was notedearlier emotional reactions to these kinds of simplestimuli require much less elaborate cognitive processingjudgment and decision making than do other emotion-eliciting stimuli (eg card games or social scenarios)

In its most basic form the startle stimulus can be pre-sented without warning which gives subjects minimalopportunity for appraisal or preparation It can also bepresented with warning which gives subjects time to ap-praise anticipate and react to the impending stimulusWhen the startle is presented with warning (eg follow-ing a visible countdown) it is sufficiently aversive toproduce marked anticipatory emotional responses inneurologically intact individuals Exposing subjects toboth types of startles (ie unanticipated and anticipated)provides an opportunity to focus on anticipatory reac-tions in a situation with minimal cognitive demands andin which an aversive outcome is certain

Overview of the Present StudyThe aim of the present study was to examine the impact

of orbitofrontal cortex damage on emotional activationin response to startle stimuli that were either unantici-pated or anticipated Five male patients with bilateral le-sions to the ventromedial region of the orbital prefrontalcortex and 5 healthy male control subjects were pre-sented with two acoustic startles The first one occurredunexpectedly (unanticipated startle) and the second onefollowed a 20-sec countdown (anticipated startle) Thesubjectsrsquo physiological (cardiovascular electrodermaland somatic) behavioral (facial displays of fear and sur-prise) and subjective emotional (self-reports of fear andsurprise) reactions were examined during and immedi-ately following both startles as well as during the antic-ipatory period before the anticipated startle

HypothesesHypothesis 1 Reactions to unanticipated startle

We hypothesized that patients and controls would notdiffer significantly in their (a) physiological reactions(cardiovascular electrodermal and somatic) (b) displaysof emotional facial behavior (fear and surprise) or (c) self-reported subjective emotional experience (fear and sur-prise) in response to an unanticipated acoustic startle

This hypothesis was based on previous research indicat-ing that patients with orbitofrontal damage do not differsignificantly from control subjects in their electrodermalreactions to an unanticipated startle (Damasio 1994)

Hypothesis 2 Reactions to anticipated startle Wehypothesized that patients would (a) show diminishedphysiological reactions (cardiovascular electrodermaland somatic) (b) show fewer displays of emotional facialbehavior (fear and surprise) and (c) report less subjectiveemotional experience (fear and surprise) in response to ananticipated acoustic startle For the physiological and be-havioral measures we expected that the patientsrsquo dimin-ished responding also would be evident during the 20-seccountdown period (self-report data were not collected dur-ing this period so as not to interrupt the countdown)

This hypothesis was based in part on the finding thatorbitofrontal patients fail to show anticipatory electro-dermal responses prior to the onset of a punishing stim-ulus (Damasio 1994) and in part on the theory that theorbitofrontal cortex helps individuals learn stimulusndashresponse associations (Rolls 2000) Given that all thesubjects would experience the unanticipated startle firstwe expected the control subjects to retain their negativeemotional experience of the first startle and thus demon-strate anticipatory physiological increases and facial dis-plays in preparation for the second startle (Neurologicallyintact individuals typically show these kinds of anticipa-tory increases when an aversive stimulus is pending) Incontrast we expected that the patients would not retainthe negative emotional impact of the first startle and thuswould not show the same preparatory physiological andbehavioral increases as controls

Given our expectation that the patients would not per-ceive as aversive the fact that the startle noise was goingto occur at the end of the countdown we expected thatin contrast with the controls the patients would not be ina negative emotional state when the startle noise occurredPrevious research suggests that negative emotional statesare associated with heightened startle responses (egBradley amp Lang 2000) thus if the patients were to bein a less aroused state as compared with controls beforethe startle it follows that they would show smaller re-sponses when the startle occurred

METHOD

SubjectsFive male patients (mean age 524 years) with bilateral lesions

centered in the ventromedial region of the orbital prefrontal cortexand 5 healthy male control subjects (mean age 608 years) par-ticipated in the present study (A 6th patient participated in the pro-cedure but was excluded from data analyses because his lesion wasunilateral) The patients were recruited from the Northern CaliforniaVeterans Administration Health Care System Martinez Californiawhere they underwent neuropsychological and neuroanatomicalevaluations The patients had incurred focal injuries to the orbito-frontal cortex via head trauma (eg a motorcycle accident or fallingfrom a roof) Participation was voluntary and the patients received$10 per hour as compensation The control subjects were recruitedthrough newspaper advertisements and word of mouth The con-trols were not taking medication and did not have neurological orpsychiatric conditions Eleven control subjects participated in theprocedure because some of these subjects were recruited as com-parison subjects for a study involving an older cohort (ie patientswith dementia) data from 5 age-matched controls were analyzedfor comparison with the present sample of patients

Anatomy In Figure 1 individual axial slices are shown for the 5patients with lesions encompassing the bilateral orbital prefrontalcortex Lesion extent was obtained from MRI scans and by utilizinga template system (Rorden amp Brett 2000) Lesions for the groupoverlapped in the ventromedial portion of the orbital prefrontal cor-tex with extension into the inferior lateral prefrontal cortex in somepatients The patients were comparable in lesion localization to thosereported by the research group at the University of Iowa (egBechara Damasio amp Damasio 2000 Damasio et al 1990)

Acoustic Startle StimuliEach acoustic startle was a 115-dB 100-msec burst of white noise

administered through two loudspeakers located behind the subjectrsquos

310 ROBERTS ET AL

head The resulting noise was a highly salient and noxious stimulusthat simulated a gunshot It should be noted that this kind of acousticstartle is much louder and longer than the kind of background startleprobe stimulus used in contemporary studies of positive and negativeaffect (eg Bradley amp Lang 2000 Vrana Spence amp Lang 1988)

ApparatusPhysiological Three physiological measures were examined in

the present study The first was the cardiac interbeat interval Theelectrocardiogram was recorded using Beckman miniature elec-trodes filled with Redux paste and placed on opposite sides of thechest The length of the interval between successive R-waves wasmeasured in milliseconds The second physiological measure wasthe skin conductance level To measure the level of sweat gland activ-ity on the surface of the hand a small constant voltage was passedbetween Beckman regular electrodes attached to the palmar surfaceof the lower phalanges of the first and second fingers on the non-dominant hand The electrolyte used was sodium chloride in UnibaseThe third physiological measure was general somatic activity Tomeasure somatic activity or bodily movement an electromechani-cal transducer was attached to a platform under the subjectrsquos chairThe transducer generated an electrical signal proportional to theamount of movement in any direction Somatic activity was mea-sured in arbitrarily designated units In addition to the transducersused to record these measures a grounding clip filled with Reduxpaste was placed on the subjectrsquos left ear

These three physiological measures enabled continuous unob-trusive monitoring of cardiovascular electrodermal and somaticactivity all important aspects of emotional activation These mea-sures allowed us to expand on prior work with patients with orbito-frontal damage which primarily had examined electrodermal re-sponses (eg Damasio 1994)

Physiological data were collected continuously using a 12-channelGrass Model 7 polygraph and were processed by a microcomputer

system using software written by one of the authors (RWL) Second-by-second averages were derived for each measure

Audiovisual The subjects received instructions on a 48-cmcolor television monitor placed on a table 125 m from the subjectA remotely controlled high-resolution video camera partially con-cealed behind darkened glass and embedded in a bookshelf wasused to obtain a frontal view of the subjectrsquos face and upper torso inan unobtrusive manner

Self-reported emotion After each startle the subjects com-pleted an inventory (adapted from Ekman Friesen amp Ancoli 1980)that measured self-reported subjective emotional experience Thesubjects used 9-point Likert-type scales (anchored by 0 not atall 4 moderately and 8 very much) to rate how strongly theyfelt several specific emotions (including fear and surprise) at themoment when the loud noise occurred

ProcedureThe subjects completed one 4-h laboratory session On arrival

the subjects were seated in a comfortable chair in a well-lit 3 6 mroom Physiological sensors were attached and the subjects werepresented with 15 experimental trials (viewing 10 film clips pos-ing a silly face performing an isometric exercise performing abreathing exercise and being startled by two loud noises) Thefocus of the present investigation was on the last two conditions inthis session which each consisted of a single trial (ie one startlepresentation in each respective trial)1

Before each trial the experimenter verbally presented instruc-tions for the upcoming task and then these instructions were re-peated in written form and read aloud on the video monitor Base-line data were collected for 1 min before each trial and for 2 minafter each trial During these baseline periods the subjects wereasked to relax and stare at a letter ldquoXrdquo that appeared on the videomonitor2 The self-report emotion inventory was administered fol-lowing each posttrial baseline At the end of the experiment the ex-

Figure 1 Extent of lesions in orbitofrontal cortex patients (n 5) as reconstructed from MRI or CT scans Each row showsthe extent of damage in an individual patient as transcribed onto axial templates using 5-mm cuts (lesion marked in red) Thebottom row represents average extent of overlap across the 5 patients (percentage of overlap is indicated by color code)

Patient 1

Patient 2

Patient 3

Patient 4

Patient 5

Average

0 100


perimenter detached the physiological sensors paid and debriefedthe subject and answered any questions

Acoustic startle trials On the unanticipated startle trial theacoustic startle occurred without warning at the end of the 1-minbaseline period On the anticipated startle trial (which always fol-lowed the unanticipated startle trial) the subjects received the fol-lowing instructions ldquoIn this part of the experiment you will hear aloud noise You will know exactly when the loud noise will occurYou will see a countdown from 10 to 0 on the video screen Whenyou see ldquo0rdquo the loud noise will happen Before beginning the count-down I want you to relaxrdquo After the 1-min baseline period the fol-lowing instructions were repeated on the video monitor ldquoNow youwill see a countdown from 10 to 0 When the timer reaches 0 theloud noise will occurrdquo The countdown lasted for 20 sec with eachnumber presented on the screen for 2 sec

Data ReductionPhysiology For each physiological measure (cardiac interbeat

interval skin conductance level and somatic activity) second-by-second physiological data were averaged for (1) the 1-min baselineperiod before each startle (2) a 15-sec period starting with the star-tle onset and (3) for the anticipated startle the 20-sec countdownprior to the startle The use of a 15-sec time window for measuringthe responses to the startles was selected to account for differencesin the time course of the three physiological measures (eg skinconductance responds more slowly than heart rate and somatic ac-tivity) For each physiological measure and for each startle condi-tion change scores were computed by subtracting mean prestartlebaseline physiology from mean physiology during the 15-sec star-tle time window For the anticipated startle condition similar scoresalso were computed for the countdown period by subtracting meanprestartle baseline physiology from mean physiology during the20-sec countdown

Expressive behavior A team of trained research assistants codedthe subjectsrsquo expressive behavior from videotapes of each subjectrsquosface and upper torso using a modified version of Gross and Leven-sonrsquos (1993) Emotional Expressive Behavior coding system (inter-coder reliability [alpha] for the full set of codes was 80) Coderswere blind as to whether the subject was a patient or a control

In designing the coding system for the primary behavioral re-sponses to the startle we relied on previous research (Ekman et al1985 Landis amp Hunt 1939) that has indicated that the initial star-tle response typically includes seven behaviors hard eye closurelip stretch neck stretch head jerk shoulder raise forward lungeand torso raise Coders determined whether each of these seven be-haviors was present (coded as 1) or absent (coded as 0) A startle be-havior composite score was computed by averaging these codes

Startle stimuli also often generate secondary emotional expres-sionsmdashspecifically fear and surprise Coders rated each of theseemotions on a second-by-second basis during a period beginning5 sec before and ending 5 sec after each startle using a 4-point in-tensity scale (0 none 1 slight 2 moderate and 3 strong)

For each of the behavioral measures (startle behavior compositescore facial displays of fear and facial displays of surprise) second-by-second codes were averaged for (1) a 6-sec period starting withthe startle onset and (2) for the anticipated startle the last 5 sec ofthe countdown prior to the startle The use of a 6-sec time windowfor measuring the responses to the startles was selected to capturethe subjectsrsquo maximal display of expressive behavior

Self-reported emotion We examined the subjectsrsquo self-reportratings of fear and surprise in response to each of the startles

RESULTS

Data Analytic StrategyTo test our hypotheses we conducted a 2 2 (group

[controls or patients] startle type [unanticipated or an-

ticipated] ) analysis of variance (ANOVA) for each phys-iological behavioral and self-report measure Groupwas treated as a between-subjects factor and startle typewas treated as a within-subjects repeated measure Whenthere were significant interactions of startle type andgroup follow-up univariate comparisons were conductedfor each startle One-way analyses of variance were con-ducted to compare group differences in physiology andbehavior during the anticipatory countdown period priorto the anticipated startle (as was noted earlier there wereno self-report data collected during the countdown pe-riod) Bonferroni corrections were used for all compar-isons to minimize Type I error

Manipulation Check Effects of StartleCondition Across Participants

Before testing our hypotheses we examined the maineffects of startle condition to determine whether the twostartle conditions were successful in producing differentpatterns of responding We expected the unanticipated star-tle to produce larger responses than the anticipated startlebecause providing the opportunity to anticipatemdashand thuspsychologically prepare formdashthe startle stimulus has beenfound to lessen its effects on emotional responding (Ek-man et al 1985 Landis amp Hunt 1939) Although partici-pants occasionally have been found to habituate rapidly toan anticipated startle when it is presented after an unantic-ipated startle (Ekman et al 1985) previous work hasshown that anticipated startles typically generate smalleremotional responses than do unanticipated startles evenwhen startle conditions are counterbalanced (Landis ampHunt 1939 Soto Levenson amp Ebling in press)

Consistent with previous research indicating that anunanticipated startle produces larger emotional reactionsthan does an anticipated startle we found that the par-ticipants rated the unanticipated startle as more surpris-ing [F(18) 6955 p 01] and more fear inducing[F(18) 2293 p 01] than the anticipated startle Ascompared with the anticipated startle the unanticipatedstartle elicited more startle behavior [F(18) 764 p 05] more surprise behavior [F(18) 3600 p 01]and greater increases in somatic activity [F(18) 1016p 05] There were no differences between the startlesin terms of fear behavior elicited [F(18) 100 ns]cardiovascular activation [cardiac interbeat intervalF(18) 056 ns] or electrodermal activation [skinconductance F(17) 057 ns]

Group Differences in Reactions to Unanticipatedand Anticipated Acoustic Startle

Physiological activation Mean levels of physiologybefore during and after each startle and change scoresreflecting physiological activation in response to eachstartle (ie startle physiology minus baseline physiol-ogy) are displayed by group in Tables 1 and 2 for the tworespective startle conditions

To test the hypotheses that the patients and the controlswould show similar amounts of physiological activation(ie similar increases in physiology) to the unantici-

312 ROBERTS ET AL

pated startle (Hypothesis 1) but that the patients wouldshow less physiological activation than the controls tothe anticipated startle (Hypothesis 2) we examined theinteraction of group (patient or control) and startle condi-tion (unanticipated or anticipated) for each physiologicalmeasure There was a significant interaction of group andstartle condition for cardiac interbeat interval [F(18) 688 p 05] Follow-up comparisons revealed that aswas hypothesized the patients and the controls did notdiffer significantly in their cardiovascular responses tothe unanticipated startle [F(18) 010 ns] but thatthe patients showed diminished cardiovascular activa-tion as compared with the controls in response to theanticipated startle [F(18) 556 p 05] Specificallythe control subjects showed decreases in cardiac inter-beat interval (indicating greater cardiovascular activa-tion) in response to the anticipated startle whereas thepatients showed increases in cardiac interbeat interval(indicating less cardiovascular activation)

The interactions of group and startle condition werenot significant for skin conductance level [F(17) 010 ns] or somatic activity [F(18) 130 ns]

To assess differences between the patients and the con-trols during the countdown period prior to the anticipatedstartle one-way ANOVAs were conducted for each phys-iological measure These analyses revealed a significantdifference between the patients and the controls for cardiacinterbeat interval [F(18) 641 p 05] the controlsubjects showed decreases in cardiac interbeat interval(indicating greater cardiovascular activation) whereas thepatients showed increases in cardiac interbeat interval(indicating less cardiovascular activation) There wereno group differences in skin conductance level [F(17) 056 ns] or somatic activity [F(18) 100 ns] duringthe countdown

It should be noted that there were no group differencesin physiology during the 1-min baseline period prior toeither of the startles or during the 2-min recovery period

Table 1Mean Physiological Responses to Unanticipated Startle (With Standard Deviations)

Prestartle Startle Poststartle 1-Min Baseline Startle Mean Change From Baseline 2-Min Baseline

Measure M SD M SD M SD M SD

Controls

IBI 8187 1305 8097 1154 90 757 8489 1392SCL 22 13 30 16 09 04 26 12ACT 09 01 16 11 07 11 09 02

Patients

IBI 8583 565 8351 541 233 658 8862 358SCL 33 20 44 26 10 07 40 20ACT 15 12 23 13 07 14 17 15

NotemdashStartle values reflect a 15-sec time window beginning with the startle noise Change from baselinevalues reflect the mean of the time window of interest (ie the 15-sec startle window) minus the mean of thebaseline period For cardiac interbeat interval smaller values reflect greater physiological activation thusnegative change scores represent increases in cardiovascular activation and positive change scores representdecreases in cardiovascular activation IBI cardiac interbeat interval SCL skin conductance level ACT so-matic activity

Table 2Mean Physiological Responses to Anticipated Startle (With Standard Deviations)

Prestartle Countdown Startle Poststartle1-Min Countdown Change From Startle Change From 2-Min

Baseline Mean Baseline Mean Baseline Baseline

Measure M SD M SD M SD M SD M SD M SD

Controls

IBI 8394 1421 8120 1396 274 121 8039 1186 355 368 8363 1327SCL 21 12 24 11 03 05 29 12 08 06 25 11ACT 08 02 08 02 00 00 10 04 02 03 08 02

Patients

IBI 8645 670 8770 644 115 322 8889 605 244 434 8652 552SCL 32 16 33 14 01 04 41 19 08 06 35 16ACT 18 16 15 13 03 06 15 17 03 09 17 14

NotemdashStartle values reflect a 15-sec time window beginning with the startle noise The countdown is the 20-sec anticipa-tory period before the startle noise Change from baseline values reflect the mean of the time window of interest (ie thecountdown or the startle period) minus the mean of the baseline period For cardiac interbeat interval smaller values reflectgreater physiological activation thus negative change scores represent increases in cardiovascular activation and positivechange scores represent decreases in cardiovascular activation Asterisks indicate a significant group difference between thetwo means in that column of the table IBI cardiac interbeat interval SCL skin conductance level ACT somatic activ-ity p 05


following the unanticipated startle Thus group differ-ences in cardiovascular activation in response to thecountdown and in response to the anticipated startle arenot attributable simply to group differences in recoveryfrom the unanticipated startle

Expressive behavior The amount of startle behaviorfear and surprise displayed by the patients and the con-trols for each startle condition are presented in Table 3To test the hypothesis that the patients and the controlswould show similar amounts of surprise facial behaviorand fear facial behavior in response to the unanticipatedstartle (Hypothesis 1) but that in response to the antici-pated startle the patients would show fewer facial dis-plays of surprise and fear behavior than the controls (Hy-pothesis 2) we examined the interaction of group andstartle condition for each emotional behavior There was asignificant interaction of group and startle condition fordisplays of surprise [F(18) 1600 p 01] Contraryto our predictions the patients displayed more surprisethan did the controls in response to the unanticipatedstartle [F(18) 754 p 05] and the patients and thecontrols displayed similar amounts of surprise in response

to the anticipated startle [F(18) 018 ns] The inter-action of group and startle condition was not significantfor displays of fear behavior [F(18) 100 ns]

None of the subjects displayed surprise behavior orfear behavior during the countdown prior to the antici-pated startle therefore further comparisons between thepatients and the controls were not conducted

Self-reported emotional experience Mean ratingsof self-reported surprise and fear for the patients and thecontrols for each startle condition are presented in Table 4Although we expected that the patients and the controlswould not differ in their self-reports of surprise or fear inresponse to the unanticipated startle (Hypothesis 1) butthat the patients would report less surprise and fear thanthe controls in response to the anticipated startle (Hy-pothesis 2) we found that the interactions of group andstartle condition for self-reported surprise [F(18) 002 ns] and self-reported fear [F(18) 255 p 15]were not significant However because the F values forself-reported fear indicated a moderate effect for thissample size (partial η2 024) we conducted follow-upunivariate analyses for each startle These analyses re-vealed a trend toward more fear being reported by thepatients than by the controls in response to the unantici-pated startle [F(18) 457 p 07]

DISCUSSION

The aim of this study was to examine emotional activa-tion among patients with bilateral damage to the ventro-medial region of the orbital prefrontal cortex We com-pared the emotional reactions of patients and controlsubjects in response to relatively simple acoustic startlestimuli that were either unanticipated or anticipatedThis study extended previous research on the impact oforbitofrontal cortex damage on anticipatory responses byincluding a more comprehensive assessment of emotionalactivation (cardiovascular electrodermal and somaticphysiology facial expressions and subjective emotionalresponding) and by not confounding the presence or ab-sence of anticipation with the type of stimulus used (egcomparing unanticipated startle responses with antici-pated losses in a card game)

Reactions to Unanticipated Acoustic StartleIntact physiological subjective and primary star-

tle reactions The unanticipated startle was an equallypowerful stimulus for both the patients and the controlsubjects In response to the unanticipated startle the pa-tients and the controls did not differ in their self-reportedexperience of surprise primary startle behavior (eg hardeye closure head jerk or torso raise) somatic activity(eg how much they moved when the startle occurred)cardiovascular activation or electrodermal activationThese findings indicate that a sudden loud unexpectedacoustic startle is an effective stimulus for eliciting emo-tion even among patients with significant amounts oforbitofrontal damage

Table 3Mean Amount of Expressive Behavior Displayed in Response to

Acoustic Startles (With Standard Deviations)

Control Subjects Orbitofrontal Patients

Measure M SD M SD

Unanticipated Startle (6-sec Startle Time Window)

Startle behavior 20 19 26 17Surprise behavior 05 05 12 03Fear behavior 00 00 03 04

Anticipated Startle (6-sec Startle Time Window)


Anticipated Startle Countdown (5-sec Prestartle)

Surprise behavior 00 00 00 00Fear behavior 00 00 00 00

NotemdashStartle behaviors were coded as present (1) or absent (0) Fearand surprise were coded on an intensity scale ranging from 0 (none atall ) to 3 (strong displays of behavior) Asterisks indicate a significantdifference between the two means in that row of the table p 05

Table 4Mean Ratings of Self-Reported Emotion in Response to



Measure M SD M SD

Unanticipated Startle

Surprise 76 05 76 05Fear 30 20 62 27

Anticipated Startle

Surprise 14 17 16 26Fear 08 11 18 27

NotemdashSelf-report ratings were made on 9-point scales ranging from 0(not at all ) to 8 (very much) Asterisks indicate a significant differencebetween the two means in that row of the table p 07

314 ROBERTS ET AL

Previous findings have shown that electrodermal re-sponses to unanticipated startles remain intact despitefocal lesions in the orbitofrontal cortex (Damasio et al1990) With these new findings it appears that this preser-vation of responding extends to cardiovascular somaticprimary behavioral and subjective reactions as well

Larger emotional reactions Against this backdropof similarity of response the findings that the orbito-frontal patients showed more facial displays of surpriseand reported feeling more fear in response to the unan-ticipated startle than did the control subjects are partic-ularly interesting It is important to note that displayingsurprise and experiencing fear in response to an unan-ticipated startle are appropriate responses Thus this isnot an instance of patients showing a peculiar idiosyn-cratic response but rather a difference in amount Ourfindings of greater self-reported fear and greater sur-prise behavior for patients are consistent with the find-ings of Rule Shimamura and Knight (2002) whichshowed that patients with orbitofrontal damage showcentral nervous system disinhibition Our findings alsolend support to the notion that the orbitofrontal cortexmay act as a filtering system for sensory inputs (Shima-mura 2000) The patientsrsquo greater reports of fear and dis-plays of surprise behavior as compared with the controlsmay reflect a kind of behavioral disinhibition where thepatients showed more of an appropriate emotional re-sponse when confronted with a sudden powerful stimulusthat provided minimal opportunity to exert voluntary oreffortful control over their emotional reactions

Summary of reactions to unanticipated startleTaken together the patientsrsquo reactions to the unanticipatedstartle revealed that their basic emotional reactions wereintact However there is some indication of disinhibitionand a pattern of dissociation among the components ofemotion the patients showed intact physiological reac-tions and primary behavioral startle responses but showedgreater displays of surprise facial behavior than did con-trols and possibly experienced more fear

Reactions to Anticipated Acoustic StartleAfter presenting the participants with an unanticipated

acoustic startle they were informed that the loud noisewould occur again but that they would know exactly whenthe noise would occur (ie at the end of a 20-sec count-down) The startle stimulus is sufficiently intense thatwhen warned participants typically go to considerablelengths to prepare themselves and reduce the impact of thesecond occurrence Thus this paradigm provides an op-portunity to compare patientsrsquo and controlsrsquo anticipatoryreactions with a stimulus that makes relatively minimalcognitive demand and when the aversive outcome is cer-tain Although this study design presented a confound inthat the unanticipated startle was always presented firstthis consistent presentation order made it possible to ex-amine group differences in anticipation when the aversivenature of the stimulus was clearly established

Diminished cardiovascular activation We foundsignificant differences in the cardiovascular responses of

the patients and the control subjects during the count-down and in response to the anticipated startle Specifi-cally in both periods the control subjects showed in-creases in heart rate but the patients showed decreasesin heart rate The increases in heart rate in anticipationof and in reaction to aversive events shown by controlsare commensurate with defensive reactions associatedwith emotions such as fear (Cook amp Turpin 1997 LangBradley amp Cuthbert 1997) These physiological changesare often characterized as being part of the fight or flightresponse seen in reaction to aversive stimuli (eg theacoustic startle) The orbitofrontal patients showed a quitedifferent cardiovascular response consisting of decreasesin heart rate in anticipation of and in reaction to the an-ticipated startle These kinds of heart rate decreases are typ-ically associated with the orienting response (eg Cook ampTurpin 1997) Thus the cardiovascular responses of thecontrols were consistent with their mobilizing physio-logical resources to deal with an impending noxious stim-ulus whereas those of the patients were consistent withtheir attending to the stimulus This implies that orbito-frontal cortex damage may interfere with the normal mo-bilization of cardiovascular resources necessary to respondproperly to warnings that an aversive event is imminentOf note the patientsrsquo pattern of decreased cardiovascularactivation during the countdown may have influencedtheir cardiovascular responses to the startle itself Thissuggests that the process of anticipation and specifi-cally deficits in this process can influence respondingwhen the anticipated stimulus finally occurs

Although cardiovascular responses normally followclosely the demands created by activity of the large mus-cles (Obrist Webb Sutterer amp Howard 1970) we didnot find the same pattern of differences between the pa-tients and the controls for somatic activity as we did forheart rate Both the patients and the controls showed lit-tle change or a slight decrease in somatic activity duringthe countdown prior to the startle For the patients thismay have in fact reflected a lack of mobilization (con-sistent with their decreasing heart rate) whereas for thecontrols it may have reflected a momentary bracing againstthe impending noxious event prior to taking action (con-sistent with their increasing heart rate)

Why not diminished electrodermal activation Al-though our findings in the realm of cardiovascular activ-ity corroborate previous research that patients fail toshow anticipatory physiological responses to punishingstimuli (Damasio 1994) we did not replicate previousfindings that this deficit is found in electrodermal activ-ity In our sample the patients and the controls did notdiffer in their electrodermal reactions during the count-down or in response to the anticipated startle One pos-sible explanation for this difference in findings is the na-ture of the tasks or contexts in which the anticipatoryresponses occurred Bechara Tranel and Damasio (2000)found that orbitofrontal patients showed deficits in electro-dermal responding in cognitively complex situations in-volving a great deal of uncertainty (ie the potential lossof money in a gambling task) In contrast the anticipated


startle task used in the present study ensures a certainaversive outcome (ie that a loud noise will be soundedat the end of a countdown) In contrast with the complex-ities and uncertainties inherent in a gambling task ourstartle task more closely resembles a fear-conditioningparadigm which may be influenced more by the amyg-dala than by the orbitofrontal cortex (Bechara DamasioDamasio amp Lee 1999 Phelps et al 2001 Quirk Ar-mony amp LeDoux 1997) In addition whereas the gam-bling task involves a social or abstract consequence (iethreat of financial loss) the anticipated startle involves aphysical consequence (ie threat of bodily harm) Thusour findings suggest that the orbitofrontal cortex is notnecessarily implicated in generating arousal responsesmdashand specifically electrodermal activationmdashto physicallythreatening stimuli but rather may be implicated in gen-erating arousal responses to social stimuli This is con-sistent with findings suggesting that the orbitofrontal cor-tex is involved in facilitating socially appropriate behavior(Beer Heerey Keltner Scabini amp Knight 2003)

Dissociation among autonomic nervous systemmeasures The fact that in response to an anticipatedstartle orbitofrontal patients and controls did not differsignificantly in their electrodermal responses yet the pa-tients showed smaller cardiovascular responses than didthe controls (and specifically cardiovascular decreasesrather than increases) points to a dissociation among au-tonomic nervous system measures These kinds of dis-sociations are not uncommon among neurological aswell as clinical populations (eg Fowles 1980 Maschkeet al 2002) Fowles (1980) has proposed that cardio-vascular activation is associated with behavioral activa-tion (eg active avoidance) and electrodermal activationis associated with behavioral inhibition (eg passiveavoidance) Thus orbitofrontal patients may have someunderstanding that the pending startle stimulus is aver-sive and hence do not differ from controls in terms oftheir anticipatory electrodermal responses Howeverthey may not be mobilized for action to deal purpose-fully and effectively with the noxious event as is demon-strated by their lack of heart rate increases

Intact expressive behavior and subjective emo-tional experience We had expected that along with di-minished anticipatory physiological responses patientswould show fewer displays of emotion and report feelingless emotion than controls would in response to the an-ticipated startle However there were no group differ-ences in facial displays of fear or surprise prior to the an-ticipated startle (in fact neither the controls nor thepatients showed facial displays of fear or surprise duringthe anticipation period) nor were there group differ-ences in facial displays or self-reports of fear or surpriseduring and immediately after the anticipated startle Aswith the unanticipated startle it appears that these behav-ioral and subjective aspects of emotional activation to theanticipated startle remain intact in orbitofrontal patients

Summary of reactions to anticipated startle Aswas the case for the unanticipated startle the patientsrsquoemotional responses prior to as well as during the an-

ticipated startle revealed a pattern of dissociation acrosscomponents of emotion (ie physiology vs behavioralexpression and self-reported subjective experience) andeven within the physiological component of emotion(ie cardiovascular vs electrodermal and somatic acti-vation) In response to a simple forewarned noxiousstimulus most aspects of the emotional response re-mained intact in the orbitofrontal patients However thepatients showed a pattern of cardiovascular response thatsuggested they were merely orienting to the noxiousevent and not mobilizing the resources to deal with it inan efficient way Such an interpretation is consistentwith Rolls (2000) in that it may reflect a failure on thepart of the orbitofrontal patients to learn from the emo-tional impact of the first unanticipated startle and tomake associations that would prepare them for dealingwith the second anticipated startle

Conclusions About the Role of the OrbitofrontalCortex in Emotion

Our findings shed additional light on the role that theorbitofrontal cortex plays in shaping emotional responsesIt would be an overgeneralization to implicate the orbito-frontal cortex as a region responsible for emotion Evenwith bilateral orbitofrontal lesions the patients were ableto show emotional responses to powerful aversive stim-uli that had many of the same physiological behavioraland subjective features as the responses of the controlsNevertheless in the cases of both the unanticipated andthe anticipated startles subtle differences were revealedIn response to the unanticipated startle the patients showedmore facial displays of surprise and reported feelingmarginally more fear than did the controls suggestingsome disinhibition of response In response to the antic-ipated startle the patients showed a deceleration of heartrate consistent with orienting to the stimulus whereas thecontrols showed an acceleration of heart rate consistentwith mobilizing resources to deal with the aversive eventTo the extent that autonomic nervous system activationprovides cues that guide subsequent responses (Damasio1994) with the diminishing of these responses (cardio-vascular in our study electrodermal in Damasiorsquos studies)patients with orbitofrontal lesions may be left without im-portant physiological cues necessary to facilitate rapidand appropriate responses to threatening situations

Our findings also suggest that orbitofrontal patientshave a degree of dissociation among the physiological be-havioral and subjective aspects of emotion as comparedwith controls In the case of the unanticipated startle thepatientsrsquo physiological reactions were similar to those ofthe controls but their facial expressions and subjectiveexperiences were exaggerated In the case of the antici-pated startle the patients had similar facial displays ofemotion as controls but had an unconventional patternof cardiovascular activity (ie heart rate decreases ratherthan increases) With such deviations from the normalldquotuningrdquo of the emotional response it seems reasonableto expect orbitofrontal patients to demonstrate inappro-priate socio-emotional behavior that would be noticed by

316 ROBERTS ET AL

friends family clinicians and researchers Moreoverthese deviations in emotional response can be expectedto have an adverse impact on such cognitive processes asattention learning memory and decision making all ofwhich are profoundly influenced by emotional reactions

REFERENCES

Bechara A Damasio H amp Damasio A R (2000) Emotion decisionmaking and the orbitofrontal cortex Cerebral Cortex 10 295-307

Bechara A Damasio H Damasio A R amp Lee G P (1999) Dif-ferent contributions of the human amygdala and ventromedial pre-frontal cortex to decision-making Journal of Neuroscience 195473-5481

Bechara A Tranel D amp Damasio H (2000) Characterization ofthe decision-making deficit of patients with ventromedial prefrontalcortex lesions Brain 123 2189-2202

Beer J S Heerey E A Keltner D Scabini D amp Knight R T(2003) The regularity function of self-conscious emotion Insightsfrom patients with orbitofrontal damage Journal of Personality ampSocial Psychology 85 594-604

Boyce W T Chesterman E A Martin N Folkman S Cohen Famp Wara D (1993) Immunologic changes occurring at kindergartenentry predict respiratory illnesses after the Loma Prieta earthquakeJournal of Developmental amp Behavioral Pediatrics 14 296-303

Bradley M M amp Lang P J (2000) Affective reactions to acousticstimuli Psychophysiology 37 204-215

Cook E amp Turpin G (1997) Differentiating orienting startle anddefense responses The role of affect and its implications for psy-chopathology In P J Lang R F Simons amp M T Balaban (Eds) At-tention and orienting Sensory and motivational processes (pp 137-164) Mahwah NJ Erlbaum

Damasio A R (1994) Descartesrsquo error Emotion reason and thehuman brain New York Putnam

Damasio A R Tranel D amp Damasio H (1990) Individuals withsociopathic behavior caused by frontal damage fail to respond auto-nomically to social stimuli Behavioural Brain Research 41 81-94

Ekman P Friesen W V amp Ancoli S (1980) Facial signs of emotionalexperience Journal of Personality amp Social Psychology 39 1125-1134

Ekman P Friesen W V amp Simons R C (1985) Is the startle reac-tion an emotion Journal of Personality amp Social Psychology 491416-1426

Fowles D C (1980) The three arousal model Implications of Grayrsquostwo-factor learning theory for heart rate electrodermal activity andpsychopathy Psychophysiology 17 87-103

Gross J J amp Levenson R W (1993) Emotional suppression Phys-iology self-report and expressive behavior Journal of Personality ampSocial Psychology 64 970-986

Hartikainen K M Ogawa K H amp Knight R T (2000) Tran-sient interference of right hemispheric function due to automaticemotional processing Neuropsychologia 38 1576-1580

Horne J A (1993) Human sleep sleep loss and behavior Implica-tions for the prefrontal cortex and psychiatric disorder British Jour-nal of Psychiatry 162 413-419

Keay K A Clement C I amp Bandler R (2000) The neuroanatomyof cardiac nociceptive pathways In G J Ter Horst (Ed) The ner-vous system and the heart (pp 303-342) Totowa NJ Humana

Kolb B amp Whishaw I Q (1990) The frontal lobes In B Kolb amp IQWhishaw (Eds) Fundamentals of human neuropsychology (3rd edpp 463-501) New York Freeman

Kring A M amp Neale J M (1996) Do schizophrenic patients showa disjunctive relationship among expressive experiential and psy-chophysiological components of emotion Journal of Abnormal Psy-chology 105 249-257

Lacey J I amp VanLehn R (1952) Differential emphasis in somaticresponse to stress An experimental study Psychosomatic Medicine14 71-81

Landis C amp Hunt W A (1939) The startle pattern New York Far-rar amp Rinehart

Lang P J Bradley M M amp Cuthbert B N (1997) Motivated at-

tention Affect activation and action In P J Lang R F Simons ampM T Balaban (Eds) Attention and orienting Sensory and motiva-tional processes (pp 97-135) Mahwah NJ Erlbaum

Levenson R W (2001) Emotional functioning in frontotemporallobar degeneration Unpublished manuscript University of Califor-nia Berkeley Department of Psychology

Maschke M Schugens M Kindsvater K Drepper J KolbF P Diener H C Daum I amp Timmann D (2002) Fear condi-tioned changes of heart rate in patients with medial cerebellar lesionsJournal of Neurology Neurosurgery amp Psychiatry 72 116-118

Obrist P A Webb R A Sutterer J R amp Howard J L (1970)The cardiacndashsomatic relationship Some reformulations Psycho-physiology 6 569-587

Phelps E A OrsquoConnor K J Gatenby J C Gore J C Grillon Camp Davis M (2001) Activation of the left amygdala to a cognitiverepresentation of fear Nature Neuroscience 4 437-441

Quirk G J Armony J L amp LeDoux J E (1997) Fear condition-ing enhances different temporal components of tone-evoked spiketrains in auditory cortex and lateral amygdala Neuron 19 613-624

Rahman S Sahakian B J Cardinal R N Rogers R D amp Rob-bins T W (2001) Decision making and neuropsychiatry Trends inCognitive Sciences 5 271-277

Roberts N A Werner K H Beer J S Levenson R W RosenH J Perry R J amp Miller B L (2001 October) Cerebral later-alization of emotion in frontotemporal lobar dementia Poster pre-sented at the annual meeting of the Society for PsychophysiologicalResearch Montreal

Rolls E T (2000) The orbitofrontal cortex and reward Cerebral Cor-tex 10 284-294

Rorden C amp Brett M (2000) Stereotaxic display of brain lesionsBehavioural Neurology 12 191-200

Rule R R Shimamura A P amp Knight R T (2002) Orbitofrontalcortex and dynamic filtering of emotional stimuli Cognitive Affec-tive amp Behavioral Neuroscience 2 264-270

Schoenbaum G Chiba A A amp Gallagher M (2000) Changes infunctional connectivity in orbitofrontal cortex and basolateral amyg-dala during learning and reversal training Journal of Neuroscience20 5179-5189

Schoenbaum G amp Setlow B (2001) Integrating orbitofrontal cor-tex into prefrontal theory Common processing themes across speciesand subdivisions Learning amp Memory 8 134-147

Shimamura A (2000) The role of the prefrontal cortex in dynamic fil-tering Psychobiology 28 207-218

Soto J A Levenson R W amp Ebling R (in press) Cultures ofmoderation and expression Emotional experience behavior andphysiology in Chinese Americans and Mexican Americans Emotion

Tucker D M Luu P amp Pribram K H (1995) Social and emo-tional self-regulation In J Grafman K J Holyoak amp F Boller(Eds) Structure and function of the human prefrontal cortex (An-nals of the New York Academy of Sciences Vol 769 pp 213-239)New York New York Academy of Sciences

Vrana S R Spence E L amp Lang P J (1988) The startle probe re-sponse A new measure of emotion Journal of Abnormal Psychol-ogy 97 487-491

NOTES

1 Although the subjects may have experienced some fatigue by theend of the experiment the startle is a sufficiently powerful stimulus toelicit physiological behavioral and subjective emotional responses re-gardless of possible fatigue effects (Landis amp Hunt 1939) In additionwe do not have reason to suspect that the patients and the controls in thissample would be differentially fatigued by the experiment

2 Although physiological responses to an acoustic startle may becaptured by a relatively brief time window (ie 15 sec) 2 min of phys-iological data were collected after each startle to determine whetherthere were group differences in the time it takes for subjects to physio-logically recover (ie return to baseline) following the startle noise

(Manuscript received February 28 2003revision accepted for publication May 24 2004)

308 ROBERTS ET AL

anticipatory physiological responses that may guide de-cision making (Damasio 1994)

In the present study we examined the effects of ventro-medial orbitofrontal cortex damage on emotional activa-tion in response to powerful emotion-eliciting stimuli(acoustic startles) This study builds on previous re-search by examining multiple components of emotionincluding autonomic (electrodermal and cardiovascular)and somatic nervous system physiology emotional ex-pressive behavior and self-reported subjective emo-tional experience and by studying both unanticipatedand anticipated variants of the same stimulus Examin-ing multiple measures of emotion is necessary becausedifferent components of emotion and different measureswithin each component can reveal different patterns offindings (eg Boyce et al 1993 Gross amp Levenson1993 Lacey amp VanLehn 1952) In particular neurolog-ical disturbance as evidenced by neurological and clin-ical populations may be related to dissociations amongemotion components (Kring amp Neale 1996 Robertset al 2001)

The Role of the Orbitofrontal Cortex in EmotionSeveral theories of orbitofrontal functioning have been

advanced to explain the frequent observation of pro-nounced socio-emotional deficits (eg disinhibited so-cial behavior or detrimental decision-making strategies)among patients with damage to this region

Filtering irrelevant emotional information Shima-mura (2000) has proposed that the orbitofrontal cortexhelps regulate emotional behavior by gating or filteringout irrelevant sensory information that would otherwisebe translated into emotional information Through thistype of gating mechanism the orbitofrontal cortex fil-ters the influence of emotional responses on cognitionIn patients with orbitofrontal damage however the abil-ity to appropriately gate irrelevant information is inhib-ited Thus according to Shimamurarsquos theory orbito-frontal patients have difficulty identifying and attendingto relevant emotional information and therefore showheightened or inappropriate emotional reactions

Forming and reversing stimulusndashresponse associa-tions Rolls (2000) has posited that the orbitofrontal cor-tex plays a role in emotion by helping individuals learnand unlearn stimulusndashresponse associations Researchwith both animals and humans suggests that the orbito-frontal cortex assigns meaning to reinforcing cues in theenvironment (Rolls 2000 Schoenbaum Chiba amp Gal-lagher 2000 Schoenbaum amp Setlow 2001) For examplethe orbitofrontal cortex is activated when associationsbetween visual stimuli and taste rewards (or aversivetastes) are learned and then unlearned (ie reversedRolls 2000) Thus according to Rolls in situations inwhich reinforcement contingencies rapidly change suchas in social environments patients with orbitofrontaldamage may fail to adjust their behavior appropriately tothe demands of the situation

Generating anticipatory emotional signals Dama-sio (1994) has proposed that the orbitofrontal cortex is

involved in the generation of physiological reactions thatappropriately direct behavior In a series of studies Dama-sio and colleagues demonstrated that neurologically intactsubjects show greater electrodermal activity (ie skinconductance responses) when they are about to make arisky or potentially disadvantageous decision (eg se-lecting a card from a disadvantageous deck see BecharaDamasio amp Damasio 2000 for a review) Damasio sug-gests that these anticipatory physiological responses pro-vide cues or ldquohunchesrdquo that can be used to guide be-havior Patients with damage to the ventromedial regionof the orbitofrontal cortex however fail to show suchanticipatory physiological responses As a result pa-tients are more apt to engage in risky behavior and tomake poor decisions (Bechara Damasio amp Damasio2000 Rahman Sahakian Cardinal Rogers amp Robbins2001)

Orbitofrontal Cortex and Emotional ActivationDespite theoretical and empirical evidence suggesting

the role of the orbitofrontal cortex in emotion-relatedprocesses (eg filtering irrelevant information learningand unlearning stimulusndashresponse associations or antic-ipating stimuli with uncertain outcomes) less is knownabout the impact of orbitofrontal cortex damage on emo-tional activation (ie the patterned physiological ex-pressive and subjective responses activated to help theorganism respond efficiently to changes in the internaland external environment that have significance for goalsand well-being Levenson 2001) There is some evidenceto suggest that damage to the orbitofrontal cortex does notaffect the physiological aspect of emotional activation tosimple stimuli such as an unanticipated acoustic startle(Damasio 1994) that do not require extensive cognitiveprocessing Because previous studies have not systemati-cally investigated unanticipated and anticipated versions ofsuch simple stimuli we studied the impact of orbitofrontaldamage on emotional activation in response to acousticstartle stimuli that were either unanticipated or anticipatedTo ensure a more comprehensive assessment of emotionwe sampled from multiple domains including cardio-vascular electrodermal and somatic physiology expres-sive facial behavior and subjective emotional experience

The Use of Acoustic StartleResponses to acoustic startle stimuli are thought to

exist on the boundary between reflex (eg defensive re-actions) and emotion (Ekman Friesen amp Simons 1985)For this reason startle stimuli are useful for investigat-ing the role of the orbitofrontal cortex in emotion pro-viding a foundation for subsequent investigations ofmore complex emotional cognitive and social processes

We presented orbitofrontal patients and controls withbrief loud bursts of white noise of sufficient intensity(115 dB) to be unambiguously noxious These startlestimuli elicit stereotyped combinations of physiologicalbehavioral and subjective responses including primarydefensive behaviors (eg ducking onersquos head) and sec-ondary emotional responses (eg fear and surprise Ek-












METHOD






310 ROBERTS ET AL













Patient 1

Patient 2

Patient 3

Patient 4

Patient 5

Average

0 100











RESULTS










312 ROBERTS ET AL








Controls

IBI 8187 1305 8097 1154 90 757 8489 1392SCL 22 13 30 16 09 04 26 12ACT 09 01 16 11 07 11 09 02

Patients

IBI 8583 565 8351 541 233 658 8862 358SCL 33 20 44 26 10 07 40 20ACT 15 12 23 13 07 14 17 15






Controls

IBI 8394 1421 8120 1396 274 121 8039 1186 355 368 8363 1327SCL 21 12 24 11 03 05 29 12 08 06 25 11ACT 08 02 08 02 00 00 10 04 02 03 08 02

Patients

IBI 8645 670 8770 644 115 322 8889 605 244 434 8652 552SCL 32 16 33 14 01 04 41 19 08 06 35 16ACT 18 16 15 13 03 06 15 17 03 09 17 14








DISCUSSION







Measure M SD M SD











Measure M SD M SD


Surprise 76 05 76 05Fear 30 20 62 27

Anticipated Startle

Surprise 14 17 16 26Fear 08 11 18 27


314 ROBERTS ET AL



















316 ROBERTS ET AL


REFERENCES







































NOTES















METHOD






310 ROBERTS ET AL













Patient 1

Patient 2

Patient 3

Patient 4

Patient 5

Average

0 100











RESULTS










312 ROBERTS ET AL








Controls

IBI 8187 1305 8097 1154 90 757 8489 1392SCL 22 13 30 16 09 04 26 12ACT 09 01 16 11 07 11 09 02

Patients

IBI 8583 565 8351 541 233 658 8862 358SCL 33 20 44 26 10 07 40 20ACT 15 12 23 13 07 14 17 15






Controls

IBI 8394 1421 8120 1396 274 121 8039 1186 355 368 8363 1327SCL 21 12 24 11 03 05 29 12 08 06 25 11ACT 08 02 08 02 00 00 10 04 02 03 08 02

Patients

IBI 8645 670 8770 644 115 322 8889 605 244 434 8652 552SCL 32 16 33 14 01 04 41 19 08 06 35 16ACT 18 16 15 13 03 06 15 17 03 09 17 14








DISCUSSION







Measure M SD M SD











Measure M SD M SD


Surprise 76 05 76 05Fear 30 20 62 27

Anticipated Startle

Surprise 14 17 16 26Fear 08 11 18 27


314 ROBERTS ET AL



















316 ROBERTS ET AL


REFERENCES







































NOTES




310 ROBERTS ET AL













Patient 1

Patient 2

Patient 3

Patient 4

Patient 5

Average

0 100











RESULTS










312 ROBERTS ET AL








Controls

IBI 8187 1305 8097 1154 90 757 8489 1392SCL 22 13 30 16 09 04 26 12ACT 09 01 16 11 07 11 09 02

Patients

IBI 8583 565 8351 541 233 658 8862 358SCL 33 20 44 26 10 07 40 20ACT 15 12 23 13 07 14 17 15






Controls

IBI 8394 1421 8120 1396 274 121 8039 1186 355 368 8363 1327SCL 21 12 24 11 03 05 29 12 08 06 25 11ACT 08 02 08 02 00 00 10 04 02 03 08 02

Patients

IBI 8645 670 8770 644 115 322 8889 605 244 434 8652 552SCL 32 16 33 14 01 04 41 19 08 06 35 16ACT 18 16 15 13 03 06 15 17 03 09 17 14








DISCUSSION







Measure M SD M SD











Measure M SD M SD


Surprise 76 05 76 05Fear 30 20 62 27

Anticipated Startle

Surprise 14 17 16 26Fear 08 11 18 27


314 ROBERTS ET AL



















316 ROBERTS ET AL


REFERENCES







































NOTES














RESULTS










312 ROBERTS ET AL








Controls

IBI 8187 1305 8097 1154 90 757 8489 1392SCL 22 13 30 16 09 04 26 12ACT 09 01 16 11 07 11 09 02

Patients

IBI 8583 565 8351 541 233 658 8862 358SCL 33 20 44 26 10 07 40 20ACT 15 12 23 13 07 14 17 15






Controls

IBI 8394 1421 8120 1396 274 121 8039 1186 355 368 8363 1327SCL 21 12 24 11 03 05 29 12 08 06 25 11ACT 08 02 08 02 00 00 10 04 02 03 08 02

Patients

IBI 8645 670 8770 644 115 322 8889 605 244 434 8652 552SCL 32 16 33 14 01 04 41 19 08 06 35 16ACT 18 16 15 13 03 06 15 17 03 09 17 14








DISCUSSION







Measure M SD M SD











Measure M SD M SD


Surprise 76 05 76 05Fear 30 20 62 27

Anticipated Startle

Surprise 14 17 16 26Fear 08 11 18 27


314 ROBERTS ET AL



















316 ROBERTS ET AL


REFERENCES







































NOTES




312 ROBERTS ET AL








Controls

IBI 8187 1305 8097 1154 90 757 8489 1392SCL 22 13 30 16 09 04 26 12ACT 09 01 16 11 07 11 09 02

Patients

IBI 8583 565 8351 541 233 658 8862 358SCL 33 20 44 26 10 07 40 20ACT 15 12 23 13 07 14 17 15






Controls

IBI 8394 1421 8120 1396 274 121 8039 1186 355 368 8363 1327SCL 21 12 24 11 03 05 29 12 08 06 25 11ACT 08 02 08 02 00 00 10 04 02 03 08 02

Patients

IBI 8645 670 8770 644 115 322 8889 605 244 434 8652 552SCL 32 16 33 14 01 04 41 19 08 06 35 16ACT 18 16 15 13 03 06 15 17 03 09 17 14








DISCUSSION







Measure M SD M SD











Measure M SD M SD


Surprise 76 05 76 05Fear 30 20 62 27

Anticipated Startle

Surprise 14 17 16 26Fear 08 11 18 27


314 ROBERTS ET AL



















316 ROBERTS ET AL


REFERENCES







































NOTES










DISCUSSION







Measure M SD M SD











Measure M SD M SD


Surprise 76 05 76 05Fear 30 20 62 27

Anticipated Startle

Surprise 14 17 16 26Fear 08 11 18 27


314 ROBERTS ET AL



















316 ROBERTS ET AL


REFERENCES







































NOTES




314 ROBERTS ET AL



















316 ROBERTS ET AL


REFERENCES







































NOTES













316 ROBERTS ET AL


REFERENCES







































NOTES




the impact of orbital prefrontal cortex damage on ...narobert/he impact of orbital prefrontal...

Documents