Revista Chilena de Semiótica N°1, octubre de 1996 Facultad de Ciencias Sociales Departamento de Ciencias y Técnicas de la Comunicación Asociación Chilena de Semiótica HTML, diagramación y gráficos: Oscar Aguilera F. ([email protected]) Programa de Informática, Facultad de Ciencias Sociales, © 1996.
Ucp.ptCopyright 2007 by the American Psychological Association 2007, Vol. 7, No. 4, 745–754 Functional Magnetic Resonance Imaging Responses Relate to Differences in Real-World Social Experience Naomi I. Eisenberger, Shelly L. Gable, and Matthew D. Lieberman University of California, Los Angeles Although neuroimaging techniques have proven powerful in assessing neural responses, little is knownabout whether scanner-based neural activity relates to real-world psychological experience. A jointfunctional magnetic resonance imaging (fMRI)/experience-sampling study investigated whether individ-ual differences in neurocognitive reactivity to scanner-based social rejection related to: (a) moment-to-moment feelings of social rejection during real-world social interactions ("momentary social distress")and (b) the extent to which these momentary feelings corresponded with end-of-day global assessmentsof social disconnection ("end-of-day social disconnection"). Individuals who showed greater activity inregions associated with affective and pain processing (dorsal anterior cingulate cortex, amygdala,periaqueductal gray) during scanner-based social rejection reported feeling greater momentary socialdistress during their daily social interactions. In contrast, individuals who showed greater activity inregions associated with memory and self-referential memory encoding (hippocampus, medial prefrontalcortex) showed a stronger correspondence between momentary social distress and end-of-day socialdisconnection, such that greater momentary social distress was associated with greater end-of-day socialdisconnection. These findings complement previous work showing a dissociation between momentaryand retrospective reports of affect and suggest that these processes rely on dissociable neural systems.
Keywords: fMRI, experience-sampling, social rejection, neural, real-world social experience, momentaryversus retrospective reports of affect In the past decade, there has been a surge in the number of Previous work has shown that neural activation during an epi- studies that have used neuroimaging techniques to elucidate the sode of social rejection in the scanner is strongly correlated with neural correlates of psychological experience. Despite the fact that self-reports of social distress taken immediately after the rejection the goal of this work is presumably to understand how the human episode (e.g., "I felt rejected," "I felt invisible"). Individuals who mind functions in everyday life, relatively little is known about showed greater activity in the dorsal anterior cingulate cortex how neural activity during scanner-based tasks relates to real- (dACC) during a social rejection episode reported feeling more world experience. Although it may be reasonable to assume that distressed by the rejection episode (r ! 0.88; Eisenberger, neural activity during certain, basic psychological processes gen- Lieberman, & Williams, 2003). However, it is not clear whether eralizes from the functional magnetic resonance imaging (fMRI) neural responses to an experimental episode of social rejection scanner to the real world, it is less clear whether these same relate to real-world social experience. In other words, does dACC assumptions can be made about the processing of social and activity in response to an experimental episode of social rejection emotional information. The real-world instantiation of these pro- in the scanner relate to an individual's tendency to feel socially cesses, embedded in complex social dynamics and ongoing social rejected or accepted during real-world social interactions? More- relationships, bears little resemblance to the performance of simple over, does dACC activity in response to an experimental episode tasks in the confined and controlled context of an fMRI experi- of social rejection relate to the extent to which momentary feelings ment. With the increasing interest in the neural underpinnings of of social rejection or acceptance are incorporated into more global social and emotional processes, it is important to assess whether beliefs about one's social standing? Because it is not yet possible neural activity to specific social or emotional experiences in the to directly assess whole-brain neural activity during naturalistic, scanner has meaningful correlates in the real world.
real-world social encounters, the present study investigatedwhether neural responses during an experimental episode of socialrejection within the fMRI scanner correlated with real-world ex- Naomi I. Eisenberger, Shelly L. Gable, and Matthew D. Lieberman, De- periences during ongoing social interactions.
partment of Psychology, Franz Hall, University of California, Los Angeles.
The first goal of the present study was to examine whether This research was funded by a postdoctoral research fellowship from the neural activity in response to social rejection in the scanner related National Institutes of Mental Health to N. I. Eisenberger (T32 MH-019925) to moment-to-moment feelings of social rejection in daily social and by grants from the National Institutes of Mental Health to M. D.
interactions. To investigate neural activity in response to social Lieberman (R21MH66709-01; R21MH071521-01). We thank the staff of rejection, participants were scanned while they were excluded the UCLA Brain Mapping Center for their assistance.
during a virtual ball-tossing game, allegedly with two other indi- Correspondence concerning this article should be addressed to Naomi I.
Eisenberger, Department of Psychology, 300 Medical Plaza, University of viduals (as has been done previously; Eisenberger et al., 2003). To California, Los Angeles 90095–7076. E-mail: [email protected] assess moment-to-moment feelings of social distress, participants EISENBERGER, GABLE, AND LIEBERMAN completed a 10-day experience-sampling study in which they were might expect that individuals who show greater activity in the randomly signaled at different times during the day and asked to hippocampus and other medial temporal lobe (MTL) regions, report on their feelings of social distress in their most recent social shown to be associated with memory encoding processes (Brewer, interaction ("momentary social distress;" e.g., "I felt accepted/ Zhao, Desmond, Glover, & Gabrieli, 1998; Wagner et al., 1998), rejected by my interaction partner"). We hypothesized that indi- during social rejection would show a stronger correspondence viduals who showed greater dACC activity in response to an between momentary social distress and end-of-day social discon- experimental episode of social rejection in the scanner would also nection. Greater activity in these regions during real-world expe- report feeling higher levels of momentary social distress during riences of social rejection would increase the likelihood that these their daily social interactions. In other words, individuals who are experiences are encoded into long-term memory and then more more sensitive to scanner-based social rejection, as indexed by easily retrieved later on when asked to reflect on global feelings of greater dACC activity, should also be more sensitive to rejection- related cues or the possibility of rejection in their daily social Finally, we examined how self-reported trait measures, known interactions (even if these interactions do not involve explicit to predict feelings of social distress (rejection sensitivity, social rejection like that seen in the social exclusion task) and thus report anxiety, neuroticism), related to real-world social experience as greater momentary social distress.
well. These measures were included to examine how neural as- A second goal of the present study was to investigate whether sessments compared with self-report assessments in relating to neural activity in response to scanner-based social rejection both momentary social distress and the correspondence between related to the extent to which momentary social distress during momentary social distress and end-of-day social disconnection.
daily social interactions corresponded with end-of-day global assess-ments of social disconnection. Previous work has shown that real-time experience and retrospective reports of that experience do not neces-sarily correspond (Fredrickson & Kahneman, 1993; Kahneman, Participants and Design Overview Fredrickson, Schreiber, & Redelmeier, 1993; Redelmeier &Kahneman, 1996; Updegraff, Gable, & Taylor, 2004). For exam- Forty-two healthy participants (22 female; mean age ! 21.12 ple, two individuals may experience similar levels of positive and years, SD ! 3.94), all right-handed, provided written informed negative affect over the course of a day, but then retrospectively consent to participate in this study. Experimental procedures were report different levels of well-being at the end of the day, depend- approved by the Human Subjects Protection Committee at the ing on which experiences are retrieved when making that reflective University of California, Los Angeles.
assessment. Accordingly, a recent study demonstrated that indi- Participants came in at two different times to complete the viduals who scored higher in trait-level approach motivation, com- study. At Time 1, participants completed self-report measures pared with those who scored lower, were found to give greater related to the tendency to experience social distress (neuroticism, weight to momentary positive experiences when making end-of- social anxiety, rejection sensitivity) and then began the 10-day day judgments of life satisfaction (Updegraff et al., 2004). In other experience-sampling study in which they were randomly signaled words, for those higher in trait approach motivation, there was a at different times during the day to report on their most recent stronger correspondence between momentary positive affect and social interaction. At Time 2 (which took place 1 to 2 weeks end-of-day life satisfaction.
following the completion of the experience-sampling study), par- In the present study, we investigated whether neural responses ticipants completed a neuroimaging component in which they were to scanner-based social rejection provided a meaningful index for socially excluded in the fMRI scanner.
how momentary feelings of social distress during daily social The daily experience-sampling procedure was conducted prior interactions corresponded with end-of-day assessments of social to the neuroimaging component to ensure that participants had a disconnection. To examine this, participants provided a global reasonable number of daily entries prior to being scanned. Since assessment of social disconnection at the end of each of the 10 each neuroimaging session is quite expensive, only participants days ("end-of-day social disconnection;" e.g., "Today, I generally who had enough daily entries to allow for the examination of how felt accepted by others: strongly agree/strongly disagree"), and these daily assessments related to neural activity participated in the correlations were computed between momentary social distress scanning phase of the study. Of the 42 participants who began the and end-of-day social disconnection ratings across the 10-day daily experience study, 33 completed the neuroimaging component period. This correlation provided an index of the extent to which (20 females). Reasons for not completing the neuroimaging com- momentary social distress corresponded with end-of-day assess- ponent included: not completing the experience-sampling compo- ments of social disconnection. We then investigated how neuralactivity during social rejection in the scanner related to this cor-respondence measure. This analysis allowed us to examine 1 Both of our main analyses focused on momentary social distress in whether certain kinds of neural activity during an episode of daily life, either: (a) on its own or (b) as it corresponds with end-of-day rejection increased the likelihood of momentary social distress self-reports of social disconnection. We focused on these assessments corresponding with end-of-day reports.1 because neural responses to social rejection in the scanner are most likelyto serve as a proxy for momentary experiences in social interactions. We If a strong correspondence between momentary social distress did not examine the direct relation between neural responses to scanner- and end-of-day social disconnection is a function of memory based social rejection and end-of-day self-reports because social rejection encoding processes (such that individuals who encode these expe- in the scanner is not a meaningful proxy for these retrospective judgments, riences more deeply are more likely to show end-of-day reports and the rejection experience in the scanner is not the target of these that correspond more closely with momentary experiences), we FUNCTIONAL MAGNETIC RESONANCE IMAGING RESPONSES nent (n ! 2), not having enough entries to be considered for go off at eight different times during the day (regardless of when scanning (n ! 5), data loss (n ! 1), and claustrophobia (n ! 1).
the volume was on or off). Similar techniques have been validated In addition, three participants were excluded from the neuroimag- and used widely (Reis & Gable, 2002).
ing analyses: one due to excessive motion, one due to being an To assess momentary social distress, participants rated how so- outlier on neural data (greater than 3 SDs below the mean for the cially disconnected or rejected they felt during their most recent social sample), and one due to prior experience with the social exclusion interaction using two scale items: "I felt connected to/distant from my task. The final sample consisted of 30 healthy participants (18 interaction partner" and "I felt accepted/rejected by my interaction women; mean age ! 20.73, SD ! 3.23).
partner" (modified from the Need-Threat scale; Williams, Cheung, &Choi, 2000). These items were rated on a scale ranging from (1) very Individual Difference Measures accepted (or connected to) to (7) very rejected (or distant from). Tocompute momentary social distress, average scores of these two items Before completing the daily experience-sampling assessment, par- were computed for each subject across each day. The reliability of this ticipants completed several measures assessing traits known to predict measure was strong (" ! 0.83).
feelings of social distress during social interactions. Specifically, At the end of each day, participants provided end-of-day retro- participants completed measures of rejection sensitivity (e.g., "I some- spective reports of social disconnection through a brief question- times take criticism too hard;" Mehrabian, 1976), social anxiety (e.g., naire that was emailed to them each evening. To assess end-of-day "I feel anxious when I speak in front of a group;" Fenigstein, Scheier, feelings of social disconnection, participants rated the extent to & Buss, 1975), and neuroticism (e.g., "Are your feelings easily hurt?" which they agreed with two statements: "Today, I generally felt "Would you call yourself a nervous person?" Eysenck & Eysenck, connected to others" and "Today, I generally felt accepted by 1975). The alpha reliabilities of these scales were 0.56, 0.76, and 0.88, others" (modified from the Need-Threat scale; Williams et al., respectively. These measures were assessed to determine how neural 2000). These items were rated on a scale ranging from (1) strongly assessments compared to self-reported trait measures in relating to agree to (7) strongly disagree.2 The reliability of this measure was social experiences.
strong (" ! 0.85).
The extent to which momentary social distress correlated with Daily Experience Assessment end-of-day assessments of social disconnection was used as anindex of the degree to which these social experiences were inte- To assess momentary social distress during daily social inter- grated into global assessments of social disconnection. Thus, in- actions, participants were loaned a PalmPilot device running the dividuals with a large, positive correlation coefficient had a stron- Experience Sampling Program (Barrett & Barrett, 2001), which ger correspondence between momentary and end-of-day measures, administered the relevant questions. Over the course of 10 days, such that higher levels of social distress during the day were participants were randomly signaled at different times during the associated with higher levels of social disconnection at the end of day and, once signaled, answered questions on the PalmPilot each day. Individuals with a small correlation coefficient (either related to their most recent social interaction. Social interactions positive or negative) showed little correspondence between mo- were defined as any interaction with one or more individuals that mentary and end-of-day measures. Finally, individuals with a lasted for 5 minutes or longer, but not including email or web- large, negative correlation coefficient showed a correspondence based interactions.
between momentary and end-of-day measures in the unexpected Participants were given a limited amount of time to respond to direction, such that higher levels of social distress during the day each signal, and if they did not respond during that time, the were associated with lower levels of self-reported social discon- PalmPilot shut off. If participants responded to a signal within the nection at the end of the day. The average correlation, for each time window, they were first asked whether they were able to participant, between momentary social distress and end-of-day complete an interaction entry; if the participant answered no, the social disconnection across the 10-day assessment period was PalmPilot shut off. If they answered yes, they were then asked if medium-sized (average r ! 0.31; Rosenthal & Rosnow, 1991).
they had had a new interaction since the last signal. If the partic-ipant answered no, the questionnaire ended; if the participant Cyberball Social Exclusion Task answered yes, the participant completed the questionnaire andcould not return to previously answered questions. Again, if par- To assess neurocognitive reactivity to social rejection, participants ticipants did not respond to a question item within a certain time were scanned while completing the Cyberball social exclusion task, in window, the PalmPilot shut off, preventing participants from re- a manner similar to previous work (Eisenberger et al., 2003; Williams turning to a questionnaire at a later point in time.
et al., 2000). Participants were told that they would be playing a Participants were informed that they could turn the volume off virtual ball-tossing game with two other individuals who were also in on the PalmPilot when they could not be disturbed, such as while fMRI scanners. In reality, however, there were no other players; driving, during class, or in important meetings. In addition, the participants were playing with a preset computer program. Each game PalmPilot was set so that it would only signal participants during began with a still picture of the two virtual players in the upper corners hours when they reported they would typically be awake. Becauseparticipants were told that they could turn off the volume on the 2 We refer to this scale as a measure of social disconnection rather than PalmPilot when they could not be disturbed, participants were a measure of social connection only for ease of interpretation, so that both signaled more frequently to ensure that enough assessments were momentary social distress and end-of-day social disconnection are simi- collected when the PalmPilot's volume was on. In order to obtain larly valenced. Ideally, future investigations would use items worded in the approximately four daily assessments, the PalmPilots were set to same way for both the momentary and end-of-day reports.
EISENBERGER, GABLE, AND LIEBERMAN of the screen and a hand, representing the participant, in the lower- distress and end-of-day social disconnection, correlations between center portion of the screen. The participant's name was displayed momentary social distress and end-of-day social disconnection below the hand while two other names were displayed below each of across the 10-day period were calculated for each participant. Two the two virtual players' animated cartoon representations. After 9 participants were excluded from this analysis because they did not seconds, the cartoon player in the upper left-hand corner started the have enough momentary social interaction entries and end-of-day game by throwing the ball to either the other cartoon player or the entries completed on the same days to compute a correlation score.
participant. The participant could return the ball to one of the players The correlations between momentary social distress and end-of- by pressing one of two keys on a button box. The Cyberball program day social disconnection were then entered separately as regressors was set for 60 throws per game, with the computer players waiting 0.5 into a random effects, whole-brain group analysis, comparing to 3.0 seconds (determined randomly) before making a throw to activations for the exclusion compared to the inclusion episode.
heighten the sense that the participant was actually playing with other All analyses were thresholded using an uncorrected p value of .005 combined with a cluster size threshold of 10 voxels (Forman et al., During the task, participants completed two scans. In the first 1995). All coordinates are reported in Montreal Neurological In- scan (inclusion), participants played with the two other players for stitute (MNI) format.
the entire scanning period, with each virtual player throwing theball to the participant on approximately 50% of the throws. In the second scan (exclusion), participants only received the ball for atotal of seven throws and were then excluded for the rest of the scan when the two players stopped throwing the ball to the par-ticipant (60 –90 seconds).3 Immediately following the scanning Participants responded to an average of four (SD ! 1.24) signals session, participants completed a measure of self-reported social per day. Of these four signals, participants reported that they could distress, in which they were asked to rate how socially distressed not complete approximately one of the signals (M ! 0.7; SD ! they felt during the final ball-tossing game (e.g., "I felt rejected," 0.76) and that they had had no new interaction on approximately "I felt invisible;" Williams et al., 2000).
one signal (M ! 0.7; SD ! 0.69). On average, they completedapproximately 2.5 assessments of momentary social distress perday (SD ! 1.21). Neither overall compliance (responding to the fMRI Data Acquisition and Data Analysis PalmPilot device) nor the number of momentary social distress Data were acquired on a Siemens Allegra 3T scanner. Head move- ratings made per day was significantly associated with self- ments were restrained with foam padding and surgical tape placed reported levels of momentary social distress or with any of the across each participant's forehead. For each participant, a high- self-report measures that might relate to social distress.
resolution structural T2-weighted echo-planar imaging volume (spin- With regard to the end-of-day assessments, participants com- echo; time for repetition (TR) ! 5000 ms; time for echo (TE) ! 33 pleted an average of 9.6 end-of-day entries (SD ! 1.10) across the ms; matrix size 128 # 128; 36 axial slices; field of view (FOV) ! 20 10-day period. Compliance in completing the end-of-day assess- cm; 3-mm thick, skip 1-mm) was acquired coplanar with the func- ments was also not associated with end-of-day reports of social tional scans. Two functional scans were acquired (echo planar T2*- disconnection or with any of the self-report measures that might weighted gradient-echo, TR ! 3000 ms, TE ! 25 ms, flip angle ! relate to social distress.
90°, matrix size 64 # 64, 36 axial slices, FOV ! 20-cm; 3-mm thick,skip 1-mm), each lasting 2 minutes and 30 seconds.
Does Neural Activity During Scanner-Based Social The imaging data were analyzed using statistical parametric Rejection Task Relate to Momentary Social Distress in mapping (SPM'99; Wellcome Department of Cognitive Neurol- Real-World Social Interactions? ogy, Institute of Neurology, London, United Kingdom). Images foreach participant were realigned to correct for head motion, nor- We first examined whether neural activity during the exclu- malized into a standard stereotactic space as defined by the Mon- sion episode, relative to the inclusion episode, correlated with treal Neurological Institute, and smoothed with an 8 mm Gaussian momentary social distress assessed during daily social interac- kernel, full width at half maximum. For each participant, periods tions across the 10-day assessment period. To do this, we of inclusion and exclusion were modeled as epochs based on the regressed average social distress scores across the 10-day pe- length of that participant's inclusion and exclusion episodes, which riod into a whole-brain group analysis ( p $ .005, 10-voxel were individually timed for each participant (these varied slightly extent threshold). Results revealed that, in response to the between participants due to the random delay assigned to the scanner-based social rejection episode, individuals who showed virtual players when throwing the ball). After the task was mod- greater activity in the left dACC reported greater levels of eled for each participant, planned comparisons were computed as momentary social distress during their social interactions linear contrasts to investigate neural activity during the exclusion (%12,32,38; r ! 0.53, p $ .005; see Figure 1A). Moreover, compared to the inclusion episode.
To assess correlations between momentary social distress and 3 Although it would have been ideal to counterbalance the order of the neural activity, the measure of momentary social distress was inclusion and exclusion scans across participants, having the exclusion entered as a regressor into a random effects, whole-brain group scan come before the inclusion scan would likely change the meaning of analysis, comparing activations for the exclusion episode to acti- the inclusion scan for participants. Thus, participants who were first vations for the inclusion episode. Similarly, to assess how neural excluded might subsequently worry about being excluded again or antic- reactivity moderated the relationship between momentary social ipate that another exclusion episode was possible.
FUNCTIONAL MAGNETIC RESONANCE IMAGING RESPONSES Predicted values for combined dACC/amygdala/PAG activity Neural activity during exclusion, relative to inclusion, that correlated positively with momentary social distress during daily social interactions in the (A) dorsal anterior cingulate cortex (dACC), (B) amygdala,and (C) periaqueductal gray (PAG). (D) Scatterplot showing the regression of dACC, amygdala, and PAGactivity onto momentary social distress scores. The x-axis represents the predicted values from including dACC,amygdala, and PAG activity in a regression analysis estimating momentary social distress. The predicted valuesare an average of the activity in these three regions weighted according to the weights in the regression equation.
The y-axis represents average momentary social distress scores across the 10-day assessment period. Each pointrepresents the data from a single participant.
greater activity in this region of the dACC was significantly relating to real-world social experiences. Together, the neural associated with greater self-reported social distress in response activity in the dACC, amygdala, and PAG during a single social to the Cyberball game (r ! 0.43, p $ .05). We also found that rejection episode in the fMRI scanner accounted for 45% of the individuals who showed greater activity in the left amygdala between-subjects variance in momentary social distress, aver- and left periaqueductal gray (PAG), regions associated with aged across the 10-day period, F(3, 28) ! 6.90, p $ .005; see affective and pain processing (Davis & Whalen, 2001; Peyron, Figure 1D. In contrast, self-reported rejection sensitivity, social Laurent, & Garcia-Larrea, 2000; Rainville, Duncan, Price, anxiety, and neuroticism together accounted for 28% of the Carrier, & Bushnell, 1997), also reported significantly greater variance in momentary social distress, F(3, 31) ! 3.53, p $ .05.
levels of momentary social distress during their social interac- Moreover, activity in the dACC, amygdala, and PAG was still tions (r ! 0.57 and 0.55, respectively; see Figure 1, B and C; significantly associated with momentary social distress after see Table 1a for a complete list of activations).4 These two controlling for these three self-report measures, F(3, 22) ! regions, however, were not significantly correlated with self- 5.84, p $ .005; however, the self-report measures were no reported social distress in response to the Cyberball game, longer significantly associated with momentary social distress consistent with our previous findings (Eisenberger et al., 2003).
after controlling for the neural measures, F(3, 22) ! 1.97, p ! There were no significant negative correlations between neural .15, ns. Thus, neural activity in response to a single social activity and momentary social distress.
We also assessed several traits known to predict feelings of social distress during social interactions (rejection sensitivity, 4 The PAG activation should be interpreted with caution due to the small social anxiety, neuroticism) to determine how these self- size of this brainstem nucleus and the lack of spatial resolution afforded by reported trait measures compared with the neural assessments in fMRI to accurately identify neural regions of this size.
EISENBERGER, GABLE, AND LIEBERMAN Table 1Neural Activity Related to (A) Momentary Social Distress and (B) the Correspondence Between Momentary Social Distress and End-of-Day Social Disconnection Montreal Neurological Institute (A) Correlations between neural activity and momentary social distress during10-day assessment period Medial parietal cortex (B) Correlations between neural activity and the correspondence between momentarysocial distress and end-of-day socialdisconnection Medial prefrontal cortex Note. dACC ! dorsal anterior cingulate cortex; PAG ! periaqueductal gray. Neural activity is always taken from the contrast of exclusion minus inclusion conditions during the Cyberball game. All coordinates are in MNI coordinate space. Significance was determined using p $ .005 with a 10-voxel extent threshold.
rejection episode in the scanner provided a strong index of the Activity in the dACC, amygdala, and PAG in response to tendency to experience momentary social distress in daily life.
experimental social rejection did not significantly relate to thecorrespondence between momentary social distress and end-of-daysocial disconnection in whole-brain analyses (see Table 2 for Does Neural Activity During Scanner-Based Social intercorrelations among study variables); however, activity in the Rejection Task Relate to the Extent to Which Momentary left hippocampus and medial prefrontal cortex (MPFC) did (see Social Distress Corresponded With End-of-Day Figure 2, A and B; see Table 1b for a complete list of activations).
Assessments of Social Disconnection? In response to an experimental episode of social rejection, indi- We also examined which neural regions were involved in the viduals who produced greater activity in the hippocampus, a neural extent to which momentary social distress during daily social region that has been shown to be associated with episodic memory interactions corresponded with end-of-day assessments of social encoding (Brewer et al., 1998; Wagner et al., 1998), and in the disconnection across the 10-day assessment period. To create an MPFC, a neural region that has been shown to be associated with index of the extent to which momentary social distress corre- self-referential or autobiographical memory encoding (Cabeza et sponded with end-of-day social disconnection, we computed a al., 2004; Macrae, Moran, Heatherton, Banfield, & Kelley, 2004) correlation, for each subject, between momentary social distress among other functions (Gallagher & Frith, 2003; Gusnard, and end-of-day social disconnection across the 10-day assessment Akbudak, Shulman, & Raichle, 2001; Wager, Phan, Liberzon, & period. The resulting correlation coefficients (M ! 0.30; SD ! Taylor, 2003), evidenced a stronger correspondence between mo- 0.37) ranged from r ! .92 (greater momentary social distress was mentary social distress and end-of-day reports of social discon- associated with greater end-of-day social disconnection) to r ! nection, such that greater momentary social distress during the day %0.55 (greater momentary social distress was associated with was associated with greater end-of-day assessments of social dis- lower end-of-day social disconnection). After computing the cor- connection. Neither of these activations was significantly corre- relation coefficient for each subject, we then regressed these cor- lated with self-reported social distress in response to the Cyberball relation coefficients into a whole-brain group analysis ( p $ .005, game. These findings suggest that those who demonstrate a strong 10-voxel extent threshold).
correspondence between momentary and end-of-day measures FUNCTIONAL MAGNETIC RESONANCE IMAGING RESPONSES experiences of social connection or rejection, such that those Correlations Between the Daily Experience Assessments and who are the most sensitive to an experimental episode of social Neural Activity in Response to the Cyberball Game rejection in the scanner are also most sensitive to these types ofexperiences in their everyday lives. It should be noted, how- ever, that it is not yet clear why certain neural regions (e.g., amygdala, PAG) that related to real-world experiences of social Correspondence measure distress did not relate to self-reported social distress following the Cyberball task itself, whereas other neural regions (e.g., Amygdala (%24,%8,%18) dACC) related to both. Future work will be needed to examine the role that these regions play in the specific types of feelings Hippocampus (%30,%8,%18) or cognitions that occur in response to the Cyberball game.
We also found that individuals who showed greater hip- Note. dACC ! dorsal anterior cingulate cortex; PAG ! periaqueductal pocampal and MPFC activity during an experimental episode of gray; MPFC ! medial prefrontal cortex.
social rejection demonstrated a greater correspondence between a Correspondence measure refers to the correlation between momentary momentary social distress and end-of-day social disconnection.
social distress and end-of-day social disconnection for each participant. Apositive correspondence measure indicates that greater social distress is Notably, the neural regions associated with a greater correspon- associated with greater end-of-day social disconnection.
dence between momentary social distress and retrospective *** p $ .005; all others, p & .12.
reports of social disconnection are similar to those found inneuroimaging studies of memory encoding (hippocampus:Brewer et al., 1998; Wagner et al., 1998) as well as neuroim- may show greater activity in neural regions associated with mem- aging studies of self-referential processing (MPFC: Gusnard et ory and self-referential memory encoding processes, and thus may al., 2001), including self-referential and autobiographical mem- encode these events more deeply, making them easier to retrieve ory encoding (MPFC: Cabeza et al., 2004; Macrae et al., 2004).
when making end-of-day global assessments of social disconnec- In these memory-encoding studies, individuals who demon- strated greater activity in the hippocampus when viewing pre- In addition, whereas hippocampal and MPFC activity together sented stimuli or in the MPFC when viewing self-referential accounted for 37% of the variance in the relationship between stimuli were more likely to remember those stimuli in a subse- momentary social distress and end-of-day social disconnection, quent memory test. In a similar fashion, the present data sug- F(2, 26) ! 6.89, p $ .005; Figure 2C, the self-reported trait gests that social experiences that are more deeply encoded when measures (rejection sensitivity, neuroticism, and social anxiety) they occur may then be more easily retrieved when making accounted for only 12% of the variance in this relationship, which global assessments of social disconnection at the end of the day.
was not a statistically significant amount, F(3, 28) ! 1.10, p ! .37, These findings have several important implications. First, they ns. Moreover, when controlling for the three self-report measures, demonstrate that certain types of scanner-based neural activity hippocampal and MPFC activity was still significantly associated have meaningful correlates in real-world experience. Second, they with the correspondence between momentary social distress and point to a double dissociation in the neural systems associated with end-of-day social disconnection, F(2, 21) ! 4.63, p $ .05.
momentary and retrospective reports of social disconnection(Lieberman, 2007). As revealed here, the brain regions associated with momentary social distress (dACC, amygdala, PAG) were notsignificantly associated with the correspondence between momen- The present study found that neural activity assessed within the tary social distress and end-of-day assessments of social discon- fMRI scanner has a meaningful relationship with real-world social nection, and the brain regions associated with this correspondence and emotional experience. Activity in separate neural systems in measure (MPFC, hippocampus) were not significantly associated response to an experimental episode of social rejection in the with momentary social distress (see Table 2). These findings map scanner provided a meaningful index for how an individual expe- nicely onto previous behavioral work showing that moment-to- riences social relationships in daily life as well as the extent to moment and retrospective reports of affect do not necessarily corre- which an individual may encode those experiences into more spond (Fredrickson & Kahneman, 1993; Kahneman et al., 1993; global assessments of social disconnection.
Redelmeier & Kahneman, 1996; Updegraff et al., 2004) and suggest We found that individuals who generated greater dACC, that part of the reason for this may be due to the fact that these amygdala, and PAG responses to a single episode of social processes rely on the computational substrates of two separate neural rejection in the scanner also reported experiencing greater lev- els of momentary social distress in their daily social interac- This dissociation may shed new light on ways to investigate tions. This is a notable finding given that this neural activity altered emotional experience in certain clinical populations. For was assessed during a single episode of social rejection that is example, it has been shown that patients with amygdala damage probably quite unlike what most individuals experience in their show no differences from normal controls in the magnitude or daily lives (most real-world experiences involving social rejec- frequency of self-reported positive or negative affect (Anderson tion are likely to be more subtle than what occurs during & Phelps, 2002), suggesting that the amygdala may not be Cyberball). However, the strong correlation between neural necessary for the generation of affective experience. However, responses to social rejection and self-reports of social distress in this study, reports of affect were assessed at the end of the during real-world interactions suggests a core sensitivity to day in a retrospective manner ("surveying the feelings and EISENBERGER, GABLE, AND LIEBERMAN nemoM Correspondence ss and end-of-day istre disconnection Regression estimates for combined Neural activity during exclusion, relative to inclusion, that correlated positively with the corre- spondence between momentary social distress and end-of-day social disconnection in the (A) hippocampus and(B) medial prefrontal cortex (MPFC). (C) Scatterplot showing the regression of hippocampal and MPFC activityonto correspondence scores between momentary social distress and end-of-day social disconnection. The x-axisrepresents the predicted values from including hippocampal and MPFC activity in a regression analysisestimating the correspondence between momentary social distress and end-of-day social disconnection. Thepredicted values are an average of the activity in these two regions weighted according to the weights in theregression equation. The y-axis represents the correspondence between momentary social distress and end-of-day social disconnection. Positive correlations represent a higher correspondence between momentary andend-of-day measures, such that greater feelings of social distress during the day are associated with greaterreports of social disconnection at the end of the day. Each point represents the data from a single participant.
emotions that they had experienced over the course of that More generally, this functional dissociation may have impor- day") and thus may reflect spared memory encoding processes tant implications for understanding how neural reactivity to rather than intact affective experience. Thus, it is possible that certain tasks predicts future behaviors and decisions. Because amygdala-damaged patients could still have altered moment-to- decisions are often based on memories of what was liked or moment experiences of affect, but show normal retrospective disliked, decisions may be more dependent on what has been reports of affect. To determine whether the amygdala is in- encoded into memory and integrated into global self-views volved in moment-to-moment affective experience, it is neces- rather than what was actually experienced in the moment sary to see if amygdala-damaged patients report moment-to- (Morewedge, Gilbert, & Wilson, 2005). Thus, even though moment experiences of affect (rather than retrospective certain neural regions, such as the dACC, may be critical experiences of affect) that are different from those of healthy determinants of momentary experiences of distress during a rejection episode, activity in these regions may not exclusively FUNCTIONAL MAGNETIC RESONANCE IMAGING RESPONSES determine how an individual remembers the rejection episode, (1998). Making memories: Brain activity that predicts how well visual integrates it into self-views, and then uses that information to experience will be remembered. Science, 281, 1185–1187.
make future decisions. Rather, hippocampal and MPFC activity Cabeza, R., Prince, S. E., Daselaar, S. M., Greenberg, D. L., Budde, M., during a rejection episode may relate more strongly to the Dolcos, F., et al. (2004). Brain activity during episodic retrieval of extent to which an experience is encoded and then later re- autobiographical and laboratory events: An fMRI study using a novel trieved for making judgments or decisions. Thus, two people photo paradigm. Journal of Cognitive Neuroscience, 16, 1583–1594.
could have similarly strong reactions to social rejection while it Davis, M., & Whalen, P. J. (2001). The amygdala: Vigilance and emotion.
Molecular Psychiatry, 6, 13–34.
is occurring, but the individual who encodes this experience Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does into memory may be more likely to retrieve this memory and rejection hurt? An fMRI study of social exclusion. Science, 302, 290 – use it when making reflective social or self-judgments or when making explicit decisions regarding future social encounters.
Eysenck, S. B. G., & Eysenck, H. J. (1975). Manual of the Eysenck This study also has a number of limitations. First, although Personality Questionnaire. London: Hodder & Stoughton.
one of the main goals of the study was to investigate whether Fenigstein, A., Scheier, M. F., & Buss, A. H. (1975). Public and private neural responses to scanner-based social rejection related to self-consciousness: Assessment and theory. Journal of Consulting and real-world feelings of rejection, the task used to assess neural Clinical Psychology, 43, 522–527.
responses to social rejection (Cyberball) is most likely not all Forman, S. D., Cohen, J. D., Fitzgerald, M., Eddy, W. F., Mintun, M. A., that similar to what individuals typically experience in their & Noll, D. C. (1995). Improved assessment of significant activation in daily social interactions. Presumably, many daily social inter- functional magnetic resonance imaging (fMRI): Use of a cluster-size actions do not involve social rejection and those that do prob- threshold. Magnetic Resonance in Medicine, 33, 636 – 647.
ably involve forms of rejection that are much more subtle and Fredrickson, B. L., & Kahneman, D. (1993). Duration neglect in retrospec- tacit than the overt social exclusion episode that occurred tive evaluations of affective episodes. Journal of Personality and SocialPsychology, 65, 45–55.
during the Cyberball game. Nonetheless, there were still strong Gallagher, H. L., & Frith, C. D. (2003). Functional imaging of "theory of relationships between neural activity to this experimental epi- mind." Trends in Cognitive Sciences, 7, 77– 83.
sode of social rejection and real-world feelings of rejection, Gusnard, D. A., Akbudak, E., Shulman, G. L., & Raichle, M. E. (2001).
indicating that these neural responses still provide a meaningful Medial prefrontal cortex and self-referential mental activity: Relation to proxy for real-world social experience.
a default mode of brain function. Proceedings of the National Academy An additional limitation is that the present study only examined of Sciences, USA, 98, 4259 – 4264.
neural activity when social rejection was occurring, but not when Kahneman, D., Fredrickson, B. L., Schreiber, C. A., & Redelmeier, D. A.
individuals were reflecting on past rejection experiences. Thus, we (1993). When more pain is preferred to less: Adding a better end.
could only assess which neural regions were involved in the Psychological Science, 4, 401– 405.
possible encoding of these rejection experiences but not which Lieberman, M. D. (2007). Social cognitive neuroscience: A review of core neural regions were involved in the later retrieval of these rejection processes. Annual Review of Psychology, 58, 259 –289.
memories. Future studies would benefit from examining the neural Macrae, C. N., Moran, J. M., Heatherton, T. F., Banfield, J. F., & Kelley, regions associated with reflecting on social experience in addition W. M. (2004). Medial prefrontal activity predicts memory for self.
to the neural regions involved in the initial encoding of these Cerebral Cortex, 14, 647– 654.
Mehrabian, A. (1976). Questionnaire measures of affiliative tendency and sensitivity to rejection. Psychological Reports, 38, 199 –209.
In sum, this study utilized a novel technique to assess Morewedge, C. K., Gilbert, D. T., & Wilson, T. D. (2005). The least likely whether neural reactivity to a single experimental episode of of times: How remembering the past biases forecasts of the future.
social rejection in the scanner related to real-world social Psychological Science, 16, 626 – 630.
experience. By using daily experience sampling in conjunction Peyron, R., Laurent, B., & Garcia-Larrea, L. (2000). Functional imaging of with neuroimaging techniques, it is possible to investigate the brain responses to pain. A review and meta-analysis. Clinical Neuro- ways in which neural reactivity to simple tasks relates to or physiology, 30, 263–288.
predicts behavior and experience in more naturalistic situations.
Rainville, P., Duncan, G. H., Price, D. D., Carrier, B., & Bushnell, M. C.
Examining the relationships between neural responses within (1997). Pain affect encoded in human anterior cingulate but not somato- the fMRI scanner and real-world experiences may provide sensory cortex. Science, 277, 968 –971.
important information regarding how individuals experience Redelmeier, D. A., & Kahneman, D. (1996). Patients memories of painful their social worlds and the neurocognitive processes that un- medial treatments: Real-time and retrospective evaluations of two min- derlie these experiences.
imally invasive procedures. Pain, 66, 3– 8.
Reis, H. T., & Gable, S. L. (2002). Event-sampling and other methods for studying everyday experience (pp. 190 –222). In H. T. Reis and C. M.
Judd (Eds.), Handbook of Research Methods in Social and PersonalityPsychology. New York: Cambridge University Press.
Anderson, A. K., & Phelps, E. A. (2002). Is the human amygdala critical Rosenthal, R., & Rosnow, R. L. (1991). Essentials of behavioral research: for the subjective experience of emotion? Evidence of intact disposi- Methods and data analysis (2nd ed.). New York: McGraw-Hill.
tional affect in patients with amygdala lesions. Journal of Cognitive Updegraff, J. A., Gable, S. L., & Taylor, S. E. (2004). What makes Neuroscience, 14, 709 –720.
experiences satisfying? The interaction of approach-avoidance motiva- Barrett, L. F., & Barrett, D. J. (2001). An introduction to computerized tions and emotions in well-being. Journal of Personality and Social experience sampling in psychology. Social Science Computer Review, Psychology, 86, 496 –504.
Wager, T. D., Phan, K. L., Liberzon, I., & Taylor, S. F. (2003). Valence, Brewer, J. B., Zhao, Z., Desmond, J. E., Glover, G. H., & Gabrieli, J. D.
gender, and lateralization of functional brain anatomy in emotion: A EISENBERGER, GABLE, AND LIEBERMAN meta-analysis of findings from neuroimaging. Neuroimage, 19, 513– Effects of being ignored over the Internet. Journal of Personality and Social Psychology, 79, 748 –762.
Wagner, A. D., Schacter, D. L., Rotte, M., Koutstaal, W., Maril, A., Dale, A. M., et al. (1998). Building memories: Remembering and forgetting ofverbal experiences as predicted by brain activity. Science, 281, 1188 – Received August 30, 2006 Revision received February 15, 2007 Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). Cyberostracism: Accepted February 15, 2007 ! Correction to Nielsen and Kaszniak (2006)
In the article "Awareness of Subtle Emotional Feelings: A Comparison of Long-Term Meditatorsand Nonmeditators," by Lisbeth Nielsen and Alfred W. Kaszniak (Emotion, 2006, Vol. 6, No. 3, pp.
392-405), the copyright attribution is incorrect. The article is in the public domain.
Hôpital de La Chaux-de-Fonds Service des Urgences N°31 , novembre 2006 Auteur : Dr G. John Responsable : Dr C.Sénéchaud l'haloperidol (Haldol) peut être utile. Bien que les neuroleptiques atypiques comme l'olanzapine (Zyprexa) ou la risperidone (Risperdal) devraient entraîner moins de problèmes cardiaques et de syndromes extrapyramidaux, il n'existe pas encore d'étude clinique prouvant leur supériorité dans cette indication. Si l'état