- Email: [email protected]
Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference
J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
Contents lists available at ScienceDirect
Journal of Behavior Therapy and Experimental Psychiatry journal homepage: www.elsevier.com/locate/jbtep
Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference Todd A. Girard a, *, Louis Lakatos b, Mahesh Menon c a
Ryerson University, Toronto, Ontario, Canada Laurentian University, Sudbury, Ontario, Canada c University of British Columbia, Vancouver, British Columbia, Canada b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 April 2016 Received in revised form 15 November 2016 Accepted 16 November 2016 Available online xxx
Background and objectives: Delusions of reference are thought to reflect abnormally heightened attributions of salience to mundane events or stimuli that lead to convictions that they are personally significant or directed at the observer. Recent findings highlight abnormal recruitment of brain regions associated with self-referential processes among patients with referential delusions. Given the inherent overlap of emotion, incentive salience, and self-relevance, as well as with aberrant thought processes in psychosis, this study investigated the implicit relations between participants’ perception of the emotional valence of stimuli on neural correlates of self-referent judgments among schizophreniaspectrum patients with referential delusions. Methods: During fMRI scanning, participants indicated whether sentences describing personal characteristics seemed to refer specifically to them. Subsequently, participants rated their perceived emotional valence of each statement. Results: Regression analyses revealed differential relations between groups across regions associated with self-referential processing, including prefrontal regions, anterior cingulate, insula, precuneus, and dorsal striatum. Within these regions, greater activation related to sentences rated as more positive among healthy comparison participants and more negative among patients. Limitations: The current results warrant replication and extension with larger and longitudinal samples to assess potential moderating relations of clinical and demographic individual differences. Conclusions: These findings support aberrant brain activation associated with emotional and salience brain networks in schizophrenia and highlight the importance of considering specific emotional attributes (valence) in discrete domains of delusional thought (self-referential communication). © 2016 Elsevier Ltd. All rights reserved.
Keywords: Referential delusions Emotional valence Self-reference fMRI Psychosis Schizophrenia-spectrum disorders
1. Introduction Schizophrenia has been characterized as a disorder of the self (e.g., Sass & Parnas, 2003) that is often accompanied by delusions (Breier & Berg, 1999). Although different types of delusions have long been defined, recent work highlights the potential importance of attending to their distinctions. Delusions of reference are among the most common type of delusions, and refer to instances where mundane events, stimuli, or other persons hold strong personal significance to, or directed towards, the observer (Startup, Bucci, & Langdon, 2009) and fit well with the view that delusions generally
* Corresponding author. Department of Psychology, Ryerson University, 350 Victoria St., Toronto, Ontario, M5B 2K3, Canada. E-mail address: [email protected] (T.A. Girard).
reflect exaggerated or abnormal attribution of salience (Kapur, 2003). David and colleagues (Aleman & David, 2006; Gibbs & David, 2003) have also promoted need to investigate neurocognitive mechanisms underlying such symptoms of psychosis in relation to emotional processes. These features have sparked interest in better understanding the appraisal of self-relevance in psychosis and potential relations to emotional content (e.g., Blackwood et al., 2004; Menon et al., 2011). Here we investigate the emotional modulation of brain activation during self-referential processing. In this context, selfreferential processing refers to cognitive mechanisms through which individuals deem the relevance of stimuli to their ‘self’ in a personal and environmental context (Northoff et al., 2006; van der Meer, Costafreda, Aleman, & David, 2010). Cortical midline structures in the brain play key roles in self-referential processing
http://dx.doi.org/10.1016/j.jbtep.2016.11.007 0005-7916/© 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
2
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
(Menon et al., 2011; Northoff & Bermpohl, 2004; Northoff et al., 2006; van der Meer et al., 2010). Core regions of this network include medial and inferior prefrontal cortices (PFC), anterior (ACC) and posterior cingulate cortices, medial-parietal cortex, insula, as well as connected sub-cortical regions. Despite overall convergence in supporting self-referential processing, these regions may form different subsystems with unique roles. For instance, a recent meta-analysis of self-referential processing involving conscious self-reflective decisions generally corroborates differential involvement of cortical-midline structures, and suggests that the dorsal-medial PFC may play a more general role in reflection and decision processes (van der Meer et al., 2010). Menon et al. (2011) observed expected widespread cortical-midline and limbic activation that was greater in response to items endorsed than those not deemed self-referent by a healthy sample. However, a schizophrenia sample with prominent delusions failed to show this decision-related distinction in activation within the dorsalmedial PFC, insula, ventral and dorsal striatum; i.e., unlike healthy individuals, these regions revealed comparable activation to items endorsed and non-endorsed as self-referent. Interestingly, the right ventral striatum and insula were among regions specifically revealing a positive correlation between endorsed items and the intensity of referential delusions in the patient sample. Overall, these findings are consistent with aberrant attribution of salience to stimuli accompanying schizophrenia, in combination with reduced subsequent reflection (Menon et al., 2011). Similar to the dorsal and ventral systems reviewed above in relation to self-referential processing, Phillips and Seidman (2008) differentiate dual brain systems involved in emotional processes. These systems complement each other and enable efficient perception, expression and experience of emotion. A ventral stream includes the amygdala, anterior insula, ventral ACC, orbital and ventral-lateral PFC. These structures are thought to be engaged in processes supporting both attribution of salience and emotional feeling. The deliberate mediation of these emotional processes is associated with more dorsal and archicortical structures including the dorsal ACC, dorsolateral PFC, and the hippocampus (Christensen & Bilder, 2000; Gerber et al., 2008; Phillips & Seidman, 2008). Importantly, however, emotional valence modulates the engagement of these systems. For instance, positive emotional stimuli may engage brain regions involved in more holistic and self-referential processing to a greater degree; whereas negative stimuli evoke regions linked to more detail-oriented sensory processing (e.g., Mickley & Kensinger, 2008). Moreover, despite similar response profiles among sub-cortical regions, Northoff et al. (2009) observed differential modulation of activation in the dorsal-medial PFC and ventral striatum by reported selfrelatedness, positive emotional valence, and emotional intensity. In contrast, Moran, Macrae, Heatherton, Wyland, and Kelley (2006) observed valence-related modulation of self-referential processing in the ventral ACC, with greater decreases in activation to negative stimuli. Despite some inconsistencies that may relate to specific task demands, it appears that emotional processing plays inherent but also unique roles in the phenomenological experience and neural systems involved in forms of self-referential processing. Here we assess the relation between emotional valence and brain activation during self-referent judgments among patients with referential delusions. The current study builds upon that by Menon et al. (2011) in which participants reflected on the extent to which descriptive sentences (e.g., “he likes to drink coffee”) referred specifically to them. More specifically, we investigate correlations between participants’ own valence ratings of these stimuli and their patterns of brain activation, as measured by fMRI (functional magnetic resonance imaging) during this self-reflective task. We hypothesized that we would observe aberrant modulation
of brain activation by affective valence across cortical-midline structures during self-referential processing among patients with prominent delusions of reference, compared to a healthy sample. 2. Methods 2.1. Participants The patient sample comprised 14 participants with a DSM-IV diagnosis of schizophrenia or schizoaffective disorder and prominent delusions of reference (score 4 on this item of the Schedule for Assessment of Positive Symptoms, SAPS; Andreasen, 1984). Additional inclusion criteria included fluency in English and the ability to provide voluntary consent (confirmed using the MacArthur Test of Competence; Appelbaum & Grisso, 1995). Exclusion criteria included serious unstable medical illness, concomitant major medical or neurological illness, history of head trauma that resulted in >30 min of unconsciousness, acute suicidal or homicidal ideation, formal thought disorder rating >2 on the SAPS, DSM-IV substance dependence (except caffeine and nicotine) within one month prior study entry, and MRI-related exclusions (e.g., metal implants, claustrophobia, worked with metal, pregnancy). Inclusion/exclusion criteria for healthy comparison participants (n ¼ 15) were consistent with the above with the addition of no history of any Axis I conditions as determined using the Mini International Neuropsychiatric Interview (MINI; Sheehan et al., 1998), and no reported use of illegal psychotropic drugs within the past two years. The patient group was recruited through the Centre for Addiction and Mental Health (Toronto, ON) and the non-psychiatric comparison group was recruited through advertisements in the surrounding community. The study was approved by the institutional REB and all participants provided written consent for voluntary participation. The patient and comparison samples were matched on age (Patient, M ± SD ¼ 40.6 ± 12.8; Comparison, 35.9 ± 6.9) and gender (10 males/group). Consistent with the nature of schizophrenia, the patient group had fewer years of education (Patient, M ± SD ¼ 12.9 ± 2.5; Comparison, 16.9 ± 2.1), t(27) ¼ 4.75, p < 0.001, d ¼ 1.76, and lower estimated premorbid IQ (Patient, M ± SD ¼ 97.7 ± 15.3; Comparison, 107.8 ± 8.9), t(27) ¼ 2.19, p < 0.05, d ¼ 0.81, as estimated by the WRAT-Reading test (Wide Range Achievement Test; Wilkinson & Jastak, 1993). Nonetheless, these means reflect that the patient sample had at least high-school education and were in the Average range of intelligence. All patients were on atypical antipsychotic medication (Mean chlorpromazine equivalent ¼ 413 mg; Woods, 2003). The mean SAPS score for delusions of reference was 4.3 (SD ¼ 0.9). See Menon et al. (2011) for further details. 2.2. Self-reference paradigm The experimental paradigm is described in detail by Menon et al. (2011). Sixty sentence stimuli were presented using E-Prime software (Psychology Software Tools, Pittsburgh, Pennsylvania). Each statement was presented for 5 s, with a variable inter-stimulus interval (ISI) of 1.5e3 s during which participants saw a fixation cross. The sentences described personal characteristics in the third person with the pronoun matched to participant sex, yet ambiguous in reference (e.g., “She is lazy”). Participants completed a detailed instruction phase and three practice trials to confirm that they were to make a binary decision regarding whether they judged each sentence gave rise to the feeling that it was written specifically about them, not just whether it was self-descriptive or true of them. The sentences were presented twice each during scanning (across 2 runs) in order to increase power and reliability of
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
the imaging data. Subsequently (“off-line”), participants separately rated the emotional valence of each statement in the set using an 11-point visual analogue scale ranging from very negative (0) to very positive (10). This paradigm design affords the ability to assess the relation between participants' independent ratings of emotional valence and their brain activation during self-referent processing. In order to assess the potential impact of rating strength independent of valence per se, we also derived a measure of “affective salience” by taking the absolute deviation of the ratings relative to neutral (i.e., the absolute difference score from ‘5’which was neutral. Thus, ‘very negative’ and ‘very positive’ were both coded as ‘5’, and neutral scored as 0). 2.3. Image acquisition MR images were acquired using a GE Signa 1.5 T scanner (General Electric, Waukesha, WI) equipped with a standard head coil. For localization of fMRI activation, we acquired IR-prepped 3D FSPGR recalled T1-weighted anatomic images: 120 contiguous axial 1.1-mm thick slices; 12-ms TR; 5.4-ms TE; 20 flip angle; 256 256 matrix; 200 200 mm2 FOV. The self-referential ideation task was administered across two scan sessions, each with 60 trials (30 statements, each presented twice/scan). Participants viewed backprojected images on a screen placed at the foot of the scanner bed via an adjustable mirror located above their eyes. In each fMRI session, 152 whole-brain volumes were acquired using a T2*sensitive spiral sequence: 28 contiguous axial 4.4-mm thick slices; 2300-ms TR; 40-ms TE; 85 flip; 64 64 matrix; 200 200 mm2 FOV. The first three volumes per session were discarded to allow for T1 equilibrium effects. 2.4. Image processing and analyses Consistent with the original pre-processing of the data (Menon et al., 2011), analyses were conducted using SPM5 (The Wellcome Department of Cognitive Neurology, London, UK). In order to analyze data across time within and across participants, the functional images needed to be spatially aligned. First, all functional images for a given participant were realigned to their first volume using a six-parameter rigid body transformation. Following qualitychecking of the data, those from one patient were distorted following normalization and were excluded from further processing (corresponding data were also omitted from behavioural analyses). A mean functional image was created for each participant as a representative image of their brain in order to determine parameters required to spatially normalize their brain into standard Montreal Neurological Institute (MNI) space using the EPI brain template. These computed transformation parameters were then applied to all functional images, which were also interpolated to 2mm3 isotropic voxels. These normalized images were smoothed using an 8-mm FWHM isotropic Gaussian kernel to further address anatomical variability and meet statistical assumptions. Primary whole-brain analyses assessed the correlations between BOLD (blood-oxygen level dependent) responses to items during the self-referent processing task and participants' subjective valence ratings of the stimuli. At the first-level (single person), two regression analyses provided statistical parametric correlation maps revealing brain regions where activation during the selfreferential task correlated 1) positively or 2) negatively with each participants’ independent valence ratings (on the above-described 0e10 negative-positive valence scale). That is, clusters showing negative correlations reflect greater regional brain activation in response to stimuli deemed emotionally more negative/less positive in valence by a given participant; whereas, positively correlated activation indicates regions with greater BOLD activation during
3
the decision task when presented items that were later rated as emotionally more positive/less negative. Subsequent second-level (group) t-tests were used to assess group differences in these brain correlation maps. Two contrasts were defined to assess brain regions where the healthy group correlations were greater than for patients, as these were redundant with the converse (i.e., regions where the healthy group shows a greater positive/less negative correlation equates to where patients show a greater negative/less positive correlation). Results were evaluated using a whole-brain false discovery rate (FDR) of q < 0.05 (Genovese, Lazar, & Nichols, 2002) and an extent threshold of 190 voxels corresponding to a cluster-wise p < 0.05 for this dataset. Secondary analyses investigated the extent to which the overall affective salience ratings correlated with regional brain activation. Thus, whereas our original paper (Menon et al., 2011) had examined the effect of selfreference per se, in the current analysis we examine the effects of affective valence on regional brain activation in the context of selfreferential processing. 3. Results Groups did not differ significantly in their mean valence ratings, t (adjusted df ¼ 14.33) ¼ 1.58, p ¼ 0.132. Overall, the mean ratings from both groups were near neutral (Patient M ¼ 4.35; Comparison M ¼ 4.98), but the patient sample's ratings were more variable (SD ¼ 1.37, range from 3.00 to 7.89) than the healthy comparison sample (SD ¼ 0.46, range from 4.13 to 5.58), Levene's test, F ¼ 7.02, p ¼ 0.014. Whole-brain analysis revealed greater positive correlations between BOLD signal and valence ratings among the healthy comparison group relative to the patient sample across a wide range of brain regions with most peaks in the left hemisphere, but extending bilaterally, including inferior, middle, medial, and precentral frontal regions, anterior cingulate, insula, precuneus, and caudate nucleus clusters (See Fig. 1 and Table 1). These significant clusters reflect brain regions where greater activation during self-
Fig. 1. Brain regions more positively correlated with valence ratings among healthy comparison than patient participants (voxel-wise FDR q ¼ 0.05; cluster threshold ¼ 1520 mm3, p < 0.05). Regions shown include supplementary motor area, anterior cingulate, caudate nucleus, insula, and precuneus (see Table 1 for further details). The colour bar reflects SPM T-scores. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
4
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
Table 1 Brain regions more positively correlated with valence ratings within the comparison than patient sample. Region
Cluster size (mm3)
Medial-lateral frontal Supplementary Motor Area Middle frontal g. Anterior cingulate
21,944
L Frontal-insular cortex Pars orbitalis Insula Pars triangularis
5000
L Precentral g.
MNI Coordinate of peak voxel (x, y, z)
z value
8 22 8
18 50 38
46 28 10
4.20 4.14 3.92
48 30 40
16 26 24
8 2 2
4.11 4.00 3.95
3576
38 40 26
4 6 12
54 34 56
3.98 3.55 3.49
L Precuneus
1560
6
70
38
4.38
L Caudate
1688
6
6
6
3.89 3
Voxel-wise and cluster-extent thresholds (190 voxels ¼ 1520 mm ) set at FDR ¼ 0.05.
referent judgements was correlated with more positive/less negative valence ratings of the corresponding sentences by healthy participants. Conversely, activation in these regions was related to less positive/more negative valence ratings among patients. Consistent with the group-level data at the peak voxels within these clusters illustrated in Fig. 2, further interrogation of the data indicated that the comparison group revealed robust positively correlated activation with valence across these regions; whereas within-group analysis of brain activation in the patient group demonstrated negative correlations that failed to reach significance at the FDR < 0.05 threshold. Analyses of activation correlated with affective salience failed to survive FDR correction. 4. Discussion Results highlight differential relations between participants'ratings of emotional valence and brain activity during appraisal of
Fig. 2. Plot of contrasts for activation correlated with participants' valence ratings by group corresponding to the regions of peak activation in Table 1. Activation in the healthy comparison (HC) group (rightward bars) was positively correlated with participants' valence ratings (i.e., greater activation during self-referential processing is associated with stimuli they rated more as more positive). In contrast, patient (SCZ) activation (leftward bars) tended to be minimally or negatively correlated with their valence ratings (i.e., correlations were less robust and in the direction of greater activation for more negatively valenced stimuli). The plotted data are the mean (þSE) contrast estimates extracted from the peak voxels in each cluster as reported in Table 1. SMA ¼ Supplementary motor area.
self-reference among patients with prominent delusions compared to a healthy comparison sample. More specifically, the healthy comparison group revealed greater positive correlations between their emotional valence ratings and fMRI BOLD signal than the patient sample across several core regions associated with selfreferential processing (Moran et al., 2006; Northoff & Bermpohl, 2004; Northoff et al., 2006; van der Meer et al., 2010), including clusters in medial and inferior PFC (extending to ACC and insula), precuneus, and dorsal striatum (caudate) (See Fig. 1 and Table 1). The current analysis examines the implicit effects of affective valence during self-referential judgments. Thus, these regions were associated with greater activation among the healthy comparison group than patients for sentences they later rated more positively. In contrast, activation in these regions was less, and if anything negatively correlated with valence ratings in the patient group (See Fig. 2). These different brain-behaviour correlations are not simply reflections of basic behavioural differences (both groups had comparable mean valence ratings) or an artefact of range restrictions. The frontal, limbic, and striatal regions differentially modulated by valence between groups are also consistent with aberrant brain activation associated with emotional and salience networks in schizophrenia (Menon, 2011; Phillips & Seidman, 2008; Schmitz & Johnson, 2007). The results are also partially consistent with observations of Holt et al. (2009) of differential modulation of activation among cortical-midline structures in schizophrenia. They observed greater activation to positive stimuli among healthy participants in the posterior cingulate and precuneus, whereas the schizophrenia sample displayed greater activation to negative stimuli in these regions as well as ACC and orbitofrontal regions. Their task involved explicit judgments of the emotional valence of sentences describing social situations. Our results extend this general pattern to implicit modulation by perceived emotional valence and suggest that patients show an orientation bias towards negatively valenced stimuli. The involvement of frontal regions is also consistent with valence-related modulation in healthy samples typically involving more frontal and left-biased brain activation in association with more positive and self-referent processing (Dolcos & Denkova, 2008; Mickley & Kensinger, 2008; Moran et al., 2006; Northoff et al., 2009). A closer comparison of the current findings with those of Menon et al. (2011) supports areas of overlap as well as distinct regions. Both highlight bilateral medial PFC, particularly the right SMA (supplementary motor area), and the left insula and left precentral gyrus. Whereas Menon et al. reported differential striatal activation in the nucleus accumbens and putamen between groups, we observed differential correlations in the caudate. Whereas they found differential activation in the right posterior cingulate/precuneus, we observed differential correlations on the left. The brain areas reported by Menon et al. reflect those where healthy participants demonstrated greater activation for items endorsed as selfreferent compared to those not endorsed, whereas activation in the patient group failed to differentiate endorsement. Interestingly, they also observed main effects of self-referent endorsement in the left caudate, anterior cingulate, and precuneus that failed to interact with group in terms of magnitude of activation, whereas we report here differential modulation of these left-hemisphere regions in relation to valence between groups. Thus, the current findings complement and extend our understanding in showing that activation elicited in the context of forming self-referent judgments is differentially correlated with perceived valence in schizophrenia and healthy participants. The differential involvement of the precuneus and insula are also interesting in light of their associations with self-related processing and positive symptoms of psychosis. Siemerkus, Irle,
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
Schmidt-Samoa, Dechent, and Weniger (2012) reported a positive correlation between positive symptoms of psychosis and rightprecuneus activation during navigation of a virtual maze task using a body-centred frame of reference. As noted above, Menon et al. (2011) reported bilateral posterior cingulate/precuneus related to self-referential processing, with a decreased modulation by endorsement on the right in patients, but neither correlated with symptoms of psychosis. Patients with referential delusions in the current study failed to reveal a relation between left-precuneus activation and valence ratings, which was observed among comparison participants. Whereas the current results and task-related findings of Menon et al. also implicated left frontal-insular regions, Menon et al. found an insula cluster in the right hemisphere correlated with severity of referential delusions. These findings are not inconsistent, yet suggest that lateralized involvement of regions involved in self-referential cognitive processes among those with psychosis deserves further attention. In addition to symptoms, it will be useful to explore the role of individual difference factors more generally that might underlie these results in the future. The observation that the delusion group yielded more variance in their valence ratings than the healthy group, yet demonstrated less robust correlations with self-referent related BOLD activation, argues against a range restriction artefact for the latter. The wider range of ratings also argues against a blunting of emotion perception. Moreover, in some regions (frontal, striatal) the brain-behaviour relation is not just muted, but substantially opposite in direction. As suggested above, this pattern is consistent with a bias toward perception of negative emotional information in schizophrenia. In this regard, we ran post-hoc analyses that revealed higher doses of medication (CPZe) and greater symptoms of formal thought disorder were related to more negative valence ratings among patients (r ¼ 0.39 and 0.34, respectively). It will be important to better understand the mechanisms underlying these patterns and their relations to patient characteristics, including symptoms, neurocognitive abilities, and medication effects. When our data were collapsed across valence for more general analysis of affective salience, the results failed to reach statistical significance. Speechley, Woodward, and Ngan (2013) recently failed to find corroborating imaging evidence for differences in reflexive emotional modulation in schizophrenia. However, their measure of emotional salience was a summed score across valence and arousal ratings. Given the current valence-related findings, it is possible that different responses to specific emotional stimuli may have cancelled each other out. Williams et al. (2007) have also observed a disconnection of increased autonomic arousal and decreased brain activation (medial PFC, ACC, insula, amygdala) in relation to specific negative emotions among patients with paranoid delusions. Together these findings suggest that investigation of differential modulation by emotional domains may yield important insight into mechanisms related to aberrant thought processes associated with delusions. In sum, this study supports the importance of considering specific emotional attributes (i.e., valence) in discrete domains of delusional thought (i.e., self-referential communication). The current findings warrant future investigation regarding the anatomical and functional mechanisms underlying aberrant processing of emotional salience and self-referential thought. This work would benefit from longitudinal studies with large samples capable of disentangling potential relations with patient characteristics. For instance, Fitzsimmons et al. (2014) recently reported differential relations of integrity of the cingulum with delusion severity between samples with first-episode psychosis versus more chronic schizophrenia. In addition, potential moderation by sex, medication, and relations to functional outcome will prove informative
5
(Fitzsimmons et al., 2014; Williams et al., 2007). Conflict of interest The authors confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome. Acknowledgement We thank S. Kapur, G. Remington, J. Addington, T. Schmitz, A. Anderson, A. Graff, M. Korostil, D.Mamo, P.Gerretsen, and F. Carravaggio, as well as W. Mar, A. Naber, P. Barsoum, H. Marcon, and Sofia Raitsin for their roles in the original study by Menon et al. (2011) at the Centre for Addiction and Mental Health (Toronto, ON). This project contributed in part to a thesis at Ryerson University by L. L., who is now at Laurentian University. References Aleman, A., & David, A. S. (2006). How to fill a half-full glass: Emotion and schizophrenia. American Psychologist, 61, 75e76. http://dx.doi.org/10.1037/ 0003-066X.61.1.75. Andreasen, N. C. (1984). The scale for the assessment of positive symptoms (SAPS). Iowa City, IA: University of Iowa. Appelbaum, P. S., & Grisso, T. (1995). The Macarthur treatment competence study. I. Mental illness and competence to consent to treatment. Law and Human Behavior, 19, 105e126. Blackwood, N. J., Bentall, R. P., Ffytche, D. H., Simmons, A., Murray, R. M., & Howard, R. J. (2004). Persecutory delusions and the determination of selfrelevance: An fMRI investigation. Psychological Medicine, 34, 591e596. Breier, A., & Berg, P. H. (1999). The psychosis of schizophrenia: Prevalence, response to atypical antipsychotics, and prediction of outcome. Biological Psychiatry, 46, 361e364. http://dx.doi.org/10.1016/S0006-3223(99)00040-2. Christensen, B. K., & Bilder, R. M. (2000). Dual cytoarchitectonic trends: An evolutionary model of frontal lobe functioning and its application to psychopathology. Canadian Journal of Psychiatry, 45, 247e256. Dolcos, F., & Denkova, E. (2008). Neural correlates of encoding emotional memories: A review of functional neuroimaging evidence. Cell Science Reviews, 5, 78e122. Fitzsimmons, J., Schneiderman, J. S., Whitford, T. J., Niznikiewicz, M. A., Pelavin, P. E., Terry, D. P., et al. (2014). Cingulum bundle diffusivity and delusions of reference in first episode and chronic schizophrenia. Psychiatry Research: Neuroimaging, 224, 124e132. http://dx.doi.org/10.1016/j.pscychresns.2014.08.002. Genovese, C. R., Lazar, N. A., & Nichols, T. (2002). Thresholding statistical maps in functional neuroimaginge using the false discovery rate. NeuroImage, 15, 870e878. http://dx.doi.org/10.1006/nimg.2001.1037. Gerber, A. J., Posner, J., Gorman, D., Colibazzi, T., Yu, S., Wang, Z., Kangarlu, A., et al. (2008). An affective circumplex model of neural systems subserving valence, arousal, and cognitive overlay during the appraisal of emotional faces. Neuropsychologia, 46, 2129e2139. Gibbs, A. A., & David, A. S. (2003). Delusion formation and insight in the context of affective disturbance. Epidemiologica e Psichiatria Sociale, 12, 167e174. http:// dx.doi.org/10.1017/S1121189X00002943. Holt, D. J., Lakshmanan, B., Freudenreich, O., Goff, D. C., Rauch, S. L., & Kuperberg, G. R. (2009). Dysfunction of a cortical midline network during emotional appraisals in schizophrenia. Schizophrenia Bulletin, 37, 164e176. Kapur, S. (2003). Psychosis as a state of aberrant salience: A framework linking biology, phenomenology, and pharmacology in schizophrenia. American Journal of Psychiatry, 160, 13e23. van der Meer, L., Costafreda, S., Aleman, A., & David, A. S. (2010). Self-reflection and the brain: A theoretical review and meta-analysis of neuroimaging studies with implications for schizophrenia. Neuroscience & Biobehavioral Reviews, 34, 935e946. Menon, V. (2011). Large-scale brain networks and psychopathology: A unifying triple network model. Trends in Cognitive Sciences, 15, 483e506. http:// dx.doi.org/10.1016/j.tics.2011.08.003. Menon, M., Schmitz, T. W., Anderson, A. K., Graff, A., Korostil, M., Mamo, D., Gerretsen, P., et al. (2011). Exploring the neural correlates of delusions of reference. Biological Psychiatry, 70, 1127e1133. http://dx.doi.org/10.1016/ j.biopsych.2011.05.037. Mickley, K. R., & Kensinger, E. A. (2008). Neural processes supporting subsequent recollection and familiarity of emotional items. Cognitive, Affective, & Behavioral Neuroscience, 8, 153e164. http://dx.doi.org/10.3758/CABN.8.2.143. Moran, J. M., Macrae, C. N., Heatherton, T. F., Wyland, C. L., & Kelley, W. M. (2006). Neuroanatomical evidence for distinct cognitive and affective components of self. Journal of Cognitive Neuroscience, 18, 1e9. Northoff, G., & Bermpohl, F. (2004). Cortical midline structures and the self. Trends
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
6
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
in Cognitive Sciences, 8, 102e107. Northoff, G., Heinzel, A., Greek, M. D., Bermpohl, F., Dobrowolny, H., & Panksepp, J. (2006). Self-referential processing in our brainda meta-analysis of imaging studies on the self. Neuroimage, 31, 440e457. http://dx.doi.org/10.1016/ j.neuroimage.2005.12.002. Northoff, G., Schneider, F., Rotte, M., Matthiae, C., Tempelmann, C., Wiebking, C., et al. (2009). Differential parametric modulation of self-relatedness and emotions in different brain regions. Human Brain Mapping, 30, 369e382. http:// dx.doi.org/10.1002/hbm.20510. Phillips, L. K., & Seidman, L. J. (2008). Emotion processing in persons at risk for Schizophrenia. Schizophrenia Bulletin, 34, 888e903. Sass, L. A., & Parnas, J. (2003). Schizophrenia, consciousness, and the self. Schizophrenia Bulletin, 29, 427e444. Schmitz, T. W., & Johnson, S. C. (2007). Relevance to self: A brief review and framework of neural systems underlying appraisal. Neuroscience & Biobehavioral Reviews, 31, 585e596. Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., et al. (1998). The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for
DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59(suppl 20), 22e57. Siemerkus, J., Irle, E., Schmidt-Samoa, C., Dechent, P., & Weniger, G. (2012). Egocentric spatial learning in schizophrenia investigated with functional magnetic resonance imaging. NeuroImage: Clinical, 1, 153e163. Speechley, W. J., Woodward, T. S., & Ngan, E. T. (2013). Failure of conflict to modulate central executive network activity associated with delusions in schizophrenia. Frontiers in Psychiatry, 4, 113. http://dx.doi.org/10.3389/fpsyt.2013.00113. Startup, M., Bucci, S., & Langdon, R. (2009). Delusions of reference: A new theoretical model. Cognitive Neuropsychiatry, 14, 110e126. http://dx.doi.org/10.1080/ 13546800902864229. Wilkinson, G. S., & Jastak, J. (1993). Wide range achievement testdthird edition (WRAT-3). Wilmington, DE: Wide Range. Williams, L. M., Das, P., Liddell, B. J., Olivieri, G., Peduto, A. S., David, A. S., et al. (2007). Fronto-limbic and autonomic disjunctions to negative emotion distinguish schizophrenia subtypes. Psychiatry Research: Neuroimaging, 155, 29e44. http://dx.doi.org/10.1016/j.pscychresns.2006.12.018. Woods, S. W. (2003). Chlorpromazine equivalent doses for the newer atypical antipsychotics. Journal of Clinical Psychiatry, 64, 663e667.
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
Contents lists available at ScienceDirect
Journal of Behavior Therapy and Experimental Psychiatry journal homepage: www.elsevier.com/locate/jbtep
Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference Todd A. Girard a, *, Louis Lakatos b, Mahesh Menon c a
Ryerson University, Toronto, Ontario, Canada Laurentian University, Sudbury, Ontario, Canada c University of British Columbia, Vancouver, British Columbia, Canada b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 April 2016 Received in revised form 15 November 2016 Accepted 16 November 2016 Available online xxx
Background and objectives: Delusions of reference are thought to reflect abnormally heightened attributions of salience to mundane events or stimuli that lead to convictions that they are personally significant or directed at the observer. Recent findings highlight abnormal recruitment of brain regions associated with self-referential processes among patients with referential delusions. Given the inherent overlap of emotion, incentive salience, and self-relevance, as well as with aberrant thought processes in psychosis, this study investigated the implicit relations between participants’ perception of the emotional valence of stimuli on neural correlates of self-referent judgments among schizophreniaspectrum patients with referential delusions. Methods: During fMRI scanning, participants indicated whether sentences describing personal characteristics seemed to refer specifically to them. Subsequently, participants rated their perceived emotional valence of each statement. Results: Regression analyses revealed differential relations between groups across regions associated with self-referential processing, including prefrontal regions, anterior cingulate, insula, precuneus, and dorsal striatum. Within these regions, greater activation related to sentences rated as more positive among healthy comparison participants and more negative among patients. Limitations: The current results warrant replication and extension with larger and longitudinal samples to assess potential moderating relations of clinical and demographic individual differences. Conclusions: These findings support aberrant brain activation associated with emotional and salience brain networks in schizophrenia and highlight the importance of considering specific emotional attributes (valence) in discrete domains of delusional thought (self-referential communication). © 2016 Elsevier Ltd. All rights reserved.
Keywords: Referential delusions Emotional valence Self-reference fMRI Psychosis Schizophrenia-spectrum disorders
1. Introduction Schizophrenia has been characterized as a disorder of the self (e.g., Sass & Parnas, 2003) that is often accompanied by delusions (Breier & Berg, 1999). Although different types of delusions have long been defined, recent work highlights the potential importance of attending to their distinctions. Delusions of reference are among the most common type of delusions, and refer to instances where mundane events, stimuli, or other persons hold strong personal significance to, or directed towards, the observer (Startup, Bucci, & Langdon, 2009) and fit well with the view that delusions generally
* Corresponding author. Department of Psychology, Ryerson University, 350 Victoria St., Toronto, Ontario, M5B 2K3, Canada. E-mail address: [email protected] (T.A. Girard).
reflect exaggerated or abnormal attribution of salience (Kapur, 2003). David and colleagues (Aleman & David, 2006; Gibbs & David, 2003) have also promoted need to investigate neurocognitive mechanisms underlying such symptoms of psychosis in relation to emotional processes. These features have sparked interest in better understanding the appraisal of self-relevance in psychosis and potential relations to emotional content (e.g., Blackwood et al., 2004; Menon et al., 2011). Here we investigate the emotional modulation of brain activation during self-referential processing. In this context, selfreferential processing refers to cognitive mechanisms through which individuals deem the relevance of stimuli to their ‘self’ in a personal and environmental context (Northoff et al., 2006; van der Meer, Costafreda, Aleman, & David, 2010). Cortical midline structures in the brain play key roles in self-referential processing
http://dx.doi.org/10.1016/j.jbtep.2016.11.007 0005-7916/© 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
2
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
(Menon et al., 2011; Northoff & Bermpohl, 2004; Northoff et al., 2006; van der Meer et al., 2010). Core regions of this network include medial and inferior prefrontal cortices (PFC), anterior (ACC) and posterior cingulate cortices, medial-parietal cortex, insula, as well as connected sub-cortical regions. Despite overall convergence in supporting self-referential processing, these regions may form different subsystems with unique roles. For instance, a recent meta-analysis of self-referential processing involving conscious self-reflective decisions generally corroborates differential involvement of cortical-midline structures, and suggests that the dorsal-medial PFC may play a more general role in reflection and decision processes (van der Meer et al., 2010). Menon et al. (2011) observed expected widespread cortical-midline and limbic activation that was greater in response to items endorsed than those not deemed self-referent by a healthy sample. However, a schizophrenia sample with prominent delusions failed to show this decision-related distinction in activation within the dorsalmedial PFC, insula, ventral and dorsal striatum; i.e., unlike healthy individuals, these regions revealed comparable activation to items endorsed and non-endorsed as self-referent. Interestingly, the right ventral striatum and insula were among regions specifically revealing a positive correlation between endorsed items and the intensity of referential delusions in the patient sample. Overall, these findings are consistent with aberrant attribution of salience to stimuli accompanying schizophrenia, in combination with reduced subsequent reflection (Menon et al., 2011). Similar to the dorsal and ventral systems reviewed above in relation to self-referential processing, Phillips and Seidman (2008) differentiate dual brain systems involved in emotional processes. These systems complement each other and enable efficient perception, expression and experience of emotion. A ventral stream includes the amygdala, anterior insula, ventral ACC, orbital and ventral-lateral PFC. These structures are thought to be engaged in processes supporting both attribution of salience and emotional feeling. The deliberate mediation of these emotional processes is associated with more dorsal and archicortical structures including the dorsal ACC, dorsolateral PFC, and the hippocampus (Christensen & Bilder, 2000; Gerber et al., 2008; Phillips & Seidman, 2008). Importantly, however, emotional valence modulates the engagement of these systems. For instance, positive emotional stimuli may engage brain regions involved in more holistic and self-referential processing to a greater degree; whereas negative stimuli evoke regions linked to more detail-oriented sensory processing (e.g., Mickley & Kensinger, 2008). Moreover, despite similar response profiles among sub-cortical regions, Northoff et al. (2009) observed differential modulation of activation in the dorsal-medial PFC and ventral striatum by reported selfrelatedness, positive emotional valence, and emotional intensity. In contrast, Moran, Macrae, Heatherton, Wyland, and Kelley (2006) observed valence-related modulation of self-referential processing in the ventral ACC, with greater decreases in activation to negative stimuli. Despite some inconsistencies that may relate to specific task demands, it appears that emotional processing plays inherent but also unique roles in the phenomenological experience and neural systems involved in forms of self-referential processing. Here we assess the relation between emotional valence and brain activation during self-referent judgments among patients with referential delusions. The current study builds upon that by Menon et al. (2011) in which participants reflected on the extent to which descriptive sentences (e.g., “he likes to drink coffee”) referred specifically to them. More specifically, we investigate correlations between participants’ own valence ratings of these stimuli and their patterns of brain activation, as measured by fMRI (functional magnetic resonance imaging) during this self-reflective task. We hypothesized that we would observe aberrant modulation
of brain activation by affective valence across cortical-midline structures during self-referential processing among patients with prominent delusions of reference, compared to a healthy sample. 2. Methods 2.1. Participants The patient sample comprised 14 participants with a DSM-IV diagnosis of schizophrenia or schizoaffective disorder and prominent delusions of reference (score 4 on this item of the Schedule for Assessment of Positive Symptoms, SAPS; Andreasen, 1984). Additional inclusion criteria included fluency in English and the ability to provide voluntary consent (confirmed using the MacArthur Test of Competence; Appelbaum & Grisso, 1995). Exclusion criteria included serious unstable medical illness, concomitant major medical or neurological illness, history of head trauma that resulted in >30 min of unconsciousness, acute suicidal or homicidal ideation, formal thought disorder rating >2 on the SAPS, DSM-IV substance dependence (except caffeine and nicotine) within one month prior study entry, and MRI-related exclusions (e.g., metal implants, claustrophobia, worked with metal, pregnancy). Inclusion/exclusion criteria for healthy comparison participants (n ¼ 15) were consistent with the above with the addition of no history of any Axis I conditions as determined using the Mini International Neuropsychiatric Interview (MINI; Sheehan et al., 1998), and no reported use of illegal psychotropic drugs within the past two years. The patient group was recruited through the Centre for Addiction and Mental Health (Toronto, ON) and the non-psychiatric comparison group was recruited through advertisements in the surrounding community. The study was approved by the institutional REB and all participants provided written consent for voluntary participation. The patient and comparison samples were matched on age (Patient, M ± SD ¼ 40.6 ± 12.8; Comparison, 35.9 ± 6.9) and gender (10 males/group). Consistent with the nature of schizophrenia, the patient group had fewer years of education (Patient, M ± SD ¼ 12.9 ± 2.5; Comparison, 16.9 ± 2.1), t(27) ¼ 4.75, p < 0.001, d ¼ 1.76, and lower estimated premorbid IQ (Patient, M ± SD ¼ 97.7 ± 15.3; Comparison, 107.8 ± 8.9), t(27) ¼ 2.19, p < 0.05, d ¼ 0.81, as estimated by the WRAT-Reading test (Wide Range Achievement Test; Wilkinson & Jastak, 1993). Nonetheless, these means reflect that the patient sample had at least high-school education and were in the Average range of intelligence. All patients were on atypical antipsychotic medication (Mean chlorpromazine equivalent ¼ 413 mg; Woods, 2003). The mean SAPS score for delusions of reference was 4.3 (SD ¼ 0.9). See Menon et al. (2011) for further details. 2.2. Self-reference paradigm The experimental paradigm is described in detail by Menon et al. (2011). Sixty sentence stimuli were presented using E-Prime software (Psychology Software Tools, Pittsburgh, Pennsylvania). Each statement was presented for 5 s, with a variable inter-stimulus interval (ISI) of 1.5e3 s during which participants saw a fixation cross. The sentences described personal characteristics in the third person with the pronoun matched to participant sex, yet ambiguous in reference (e.g., “She is lazy”). Participants completed a detailed instruction phase and three practice trials to confirm that they were to make a binary decision regarding whether they judged each sentence gave rise to the feeling that it was written specifically about them, not just whether it was self-descriptive or true of them. The sentences were presented twice each during scanning (across 2 runs) in order to increase power and reliability of
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
the imaging data. Subsequently (“off-line”), participants separately rated the emotional valence of each statement in the set using an 11-point visual analogue scale ranging from very negative (0) to very positive (10). This paradigm design affords the ability to assess the relation between participants' independent ratings of emotional valence and their brain activation during self-referent processing. In order to assess the potential impact of rating strength independent of valence per se, we also derived a measure of “affective salience” by taking the absolute deviation of the ratings relative to neutral (i.e., the absolute difference score from ‘5’which was neutral. Thus, ‘very negative’ and ‘very positive’ were both coded as ‘5’, and neutral scored as 0). 2.3. Image acquisition MR images were acquired using a GE Signa 1.5 T scanner (General Electric, Waukesha, WI) equipped with a standard head coil. For localization of fMRI activation, we acquired IR-prepped 3D FSPGR recalled T1-weighted anatomic images: 120 contiguous axial 1.1-mm thick slices; 12-ms TR; 5.4-ms TE; 20 flip angle; 256 256 matrix; 200 200 mm2 FOV. The self-referential ideation task was administered across two scan sessions, each with 60 trials (30 statements, each presented twice/scan). Participants viewed backprojected images on a screen placed at the foot of the scanner bed via an adjustable mirror located above their eyes. In each fMRI session, 152 whole-brain volumes were acquired using a T2*sensitive spiral sequence: 28 contiguous axial 4.4-mm thick slices; 2300-ms TR; 40-ms TE; 85 flip; 64 64 matrix; 200 200 mm2 FOV. The first three volumes per session were discarded to allow for T1 equilibrium effects. 2.4. Image processing and analyses Consistent with the original pre-processing of the data (Menon et al., 2011), analyses were conducted using SPM5 (The Wellcome Department of Cognitive Neurology, London, UK). In order to analyze data across time within and across participants, the functional images needed to be spatially aligned. First, all functional images for a given participant were realigned to their first volume using a six-parameter rigid body transformation. Following qualitychecking of the data, those from one patient were distorted following normalization and were excluded from further processing (corresponding data were also omitted from behavioural analyses). A mean functional image was created for each participant as a representative image of their brain in order to determine parameters required to spatially normalize their brain into standard Montreal Neurological Institute (MNI) space using the EPI brain template. These computed transformation parameters were then applied to all functional images, which were also interpolated to 2mm3 isotropic voxels. These normalized images were smoothed using an 8-mm FWHM isotropic Gaussian kernel to further address anatomical variability and meet statistical assumptions. Primary whole-brain analyses assessed the correlations between BOLD (blood-oxygen level dependent) responses to items during the self-referent processing task and participants' subjective valence ratings of the stimuli. At the first-level (single person), two regression analyses provided statistical parametric correlation maps revealing brain regions where activation during the selfreferential task correlated 1) positively or 2) negatively with each participants’ independent valence ratings (on the above-described 0e10 negative-positive valence scale). That is, clusters showing negative correlations reflect greater regional brain activation in response to stimuli deemed emotionally more negative/less positive in valence by a given participant; whereas, positively correlated activation indicates regions with greater BOLD activation during
3
the decision task when presented items that were later rated as emotionally more positive/less negative. Subsequent second-level (group) t-tests were used to assess group differences in these brain correlation maps. Two contrasts were defined to assess brain regions where the healthy group correlations were greater than for patients, as these were redundant with the converse (i.e., regions where the healthy group shows a greater positive/less negative correlation equates to where patients show a greater negative/less positive correlation). Results were evaluated using a whole-brain false discovery rate (FDR) of q < 0.05 (Genovese, Lazar, & Nichols, 2002) and an extent threshold of 190 voxels corresponding to a cluster-wise p < 0.05 for this dataset. Secondary analyses investigated the extent to which the overall affective salience ratings correlated with regional brain activation. Thus, whereas our original paper (Menon et al., 2011) had examined the effect of selfreference per se, in the current analysis we examine the effects of affective valence on regional brain activation in the context of selfreferential processing. 3. Results Groups did not differ significantly in their mean valence ratings, t (adjusted df ¼ 14.33) ¼ 1.58, p ¼ 0.132. Overall, the mean ratings from both groups were near neutral (Patient M ¼ 4.35; Comparison M ¼ 4.98), but the patient sample's ratings were more variable (SD ¼ 1.37, range from 3.00 to 7.89) than the healthy comparison sample (SD ¼ 0.46, range from 4.13 to 5.58), Levene's test, F ¼ 7.02, p ¼ 0.014. Whole-brain analysis revealed greater positive correlations between BOLD signal and valence ratings among the healthy comparison group relative to the patient sample across a wide range of brain regions with most peaks in the left hemisphere, but extending bilaterally, including inferior, middle, medial, and precentral frontal regions, anterior cingulate, insula, precuneus, and caudate nucleus clusters (See Fig. 1 and Table 1). These significant clusters reflect brain regions where greater activation during self-
Fig. 1. Brain regions more positively correlated with valence ratings among healthy comparison than patient participants (voxel-wise FDR q ¼ 0.05; cluster threshold ¼ 1520 mm3, p < 0.05). Regions shown include supplementary motor area, anterior cingulate, caudate nucleus, insula, and precuneus (see Table 1 for further details). The colour bar reflects SPM T-scores. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
4
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
Table 1 Brain regions more positively correlated with valence ratings within the comparison than patient sample. Region
Cluster size (mm3)
Medial-lateral frontal Supplementary Motor Area Middle frontal g. Anterior cingulate
21,944
L Frontal-insular cortex Pars orbitalis Insula Pars triangularis
5000
L Precentral g.
MNI Coordinate of peak voxel (x, y, z)
z value
8 22 8
18 50 38
46 28 10
4.20 4.14 3.92
48 30 40
16 26 24
8 2 2
4.11 4.00 3.95
3576
38 40 26
4 6 12
54 34 56
3.98 3.55 3.49
L Precuneus
1560
6
70
38
4.38
L Caudate
1688
6
6
6
3.89 3
Voxel-wise and cluster-extent thresholds (190 voxels ¼ 1520 mm ) set at FDR ¼ 0.05.
referent judgements was correlated with more positive/less negative valence ratings of the corresponding sentences by healthy participants. Conversely, activation in these regions was related to less positive/more negative valence ratings among patients. Consistent with the group-level data at the peak voxels within these clusters illustrated in Fig. 2, further interrogation of the data indicated that the comparison group revealed robust positively correlated activation with valence across these regions; whereas within-group analysis of brain activation in the patient group demonstrated negative correlations that failed to reach significance at the FDR < 0.05 threshold. Analyses of activation correlated with affective salience failed to survive FDR correction. 4. Discussion Results highlight differential relations between participants'ratings of emotional valence and brain activity during appraisal of
Fig. 2. Plot of contrasts for activation correlated with participants' valence ratings by group corresponding to the regions of peak activation in Table 1. Activation in the healthy comparison (HC) group (rightward bars) was positively correlated with participants' valence ratings (i.e., greater activation during self-referential processing is associated with stimuli they rated more as more positive). In contrast, patient (SCZ) activation (leftward bars) tended to be minimally or negatively correlated with their valence ratings (i.e., correlations were less robust and in the direction of greater activation for more negatively valenced stimuli). The plotted data are the mean (þSE) contrast estimates extracted from the peak voxels in each cluster as reported in Table 1. SMA ¼ Supplementary motor area.
self-reference among patients with prominent delusions compared to a healthy comparison sample. More specifically, the healthy comparison group revealed greater positive correlations between their emotional valence ratings and fMRI BOLD signal than the patient sample across several core regions associated with selfreferential processing (Moran et al., 2006; Northoff & Bermpohl, 2004; Northoff et al., 2006; van der Meer et al., 2010), including clusters in medial and inferior PFC (extending to ACC and insula), precuneus, and dorsal striatum (caudate) (See Fig. 1 and Table 1). The current analysis examines the implicit effects of affective valence during self-referential judgments. Thus, these regions were associated with greater activation among the healthy comparison group than patients for sentences they later rated more positively. In contrast, activation in these regions was less, and if anything negatively correlated with valence ratings in the patient group (See Fig. 2). These different brain-behaviour correlations are not simply reflections of basic behavioural differences (both groups had comparable mean valence ratings) or an artefact of range restrictions. The frontal, limbic, and striatal regions differentially modulated by valence between groups are also consistent with aberrant brain activation associated with emotional and salience networks in schizophrenia (Menon, 2011; Phillips & Seidman, 2008; Schmitz & Johnson, 2007). The results are also partially consistent with observations of Holt et al. (2009) of differential modulation of activation among cortical-midline structures in schizophrenia. They observed greater activation to positive stimuli among healthy participants in the posterior cingulate and precuneus, whereas the schizophrenia sample displayed greater activation to negative stimuli in these regions as well as ACC and orbitofrontal regions. Their task involved explicit judgments of the emotional valence of sentences describing social situations. Our results extend this general pattern to implicit modulation by perceived emotional valence and suggest that patients show an orientation bias towards negatively valenced stimuli. The involvement of frontal regions is also consistent with valence-related modulation in healthy samples typically involving more frontal and left-biased brain activation in association with more positive and self-referent processing (Dolcos & Denkova, 2008; Mickley & Kensinger, 2008; Moran et al., 2006; Northoff et al., 2009). A closer comparison of the current findings with those of Menon et al. (2011) supports areas of overlap as well as distinct regions. Both highlight bilateral medial PFC, particularly the right SMA (supplementary motor area), and the left insula and left precentral gyrus. Whereas Menon et al. reported differential striatal activation in the nucleus accumbens and putamen between groups, we observed differential correlations in the caudate. Whereas they found differential activation in the right posterior cingulate/precuneus, we observed differential correlations on the left. The brain areas reported by Menon et al. reflect those where healthy participants demonstrated greater activation for items endorsed as selfreferent compared to those not endorsed, whereas activation in the patient group failed to differentiate endorsement. Interestingly, they also observed main effects of self-referent endorsement in the left caudate, anterior cingulate, and precuneus that failed to interact with group in terms of magnitude of activation, whereas we report here differential modulation of these left-hemisphere regions in relation to valence between groups. Thus, the current findings complement and extend our understanding in showing that activation elicited in the context of forming self-referent judgments is differentially correlated with perceived valence in schizophrenia and healthy participants. The differential involvement of the precuneus and insula are also interesting in light of their associations with self-related processing and positive symptoms of psychosis. Siemerkus, Irle,
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
Schmidt-Samoa, Dechent, and Weniger (2012) reported a positive correlation between positive symptoms of psychosis and rightprecuneus activation during navigation of a virtual maze task using a body-centred frame of reference. As noted above, Menon et al. (2011) reported bilateral posterior cingulate/precuneus related to self-referential processing, with a decreased modulation by endorsement on the right in patients, but neither correlated with symptoms of psychosis. Patients with referential delusions in the current study failed to reveal a relation between left-precuneus activation and valence ratings, which was observed among comparison participants. Whereas the current results and task-related findings of Menon et al. also implicated left frontal-insular regions, Menon et al. found an insula cluster in the right hemisphere correlated with severity of referential delusions. These findings are not inconsistent, yet suggest that lateralized involvement of regions involved in self-referential cognitive processes among those with psychosis deserves further attention. In addition to symptoms, it will be useful to explore the role of individual difference factors more generally that might underlie these results in the future. The observation that the delusion group yielded more variance in their valence ratings than the healthy group, yet demonstrated less robust correlations with self-referent related BOLD activation, argues against a range restriction artefact for the latter. The wider range of ratings also argues against a blunting of emotion perception. Moreover, in some regions (frontal, striatal) the brain-behaviour relation is not just muted, but substantially opposite in direction. As suggested above, this pattern is consistent with a bias toward perception of negative emotional information in schizophrenia. In this regard, we ran post-hoc analyses that revealed higher doses of medication (CPZe) and greater symptoms of formal thought disorder were related to more negative valence ratings among patients (r ¼ 0.39 and 0.34, respectively). It will be important to better understand the mechanisms underlying these patterns and their relations to patient characteristics, including symptoms, neurocognitive abilities, and medication effects. When our data were collapsed across valence for more general analysis of affective salience, the results failed to reach statistical significance. Speechley, Woodward, and Ngan (2013) recently failed to find corroborating imaging evidence for differences in reflexive emotional modulation in schizophrenia. However, their measure of emotional salience was a summed score across valence and arousal ratings. Given the current valence-related findings, it is possible that different responses to specific emotional stimuli may have cancelled each other out. Williams et al. (2007) have also observed a disconnection of increased autonomic arousal and decreased brain activation (medial PFC, ACC, insula, amygdala) in relation to specific negative emotions among patients with paranoid delusions. Together these findings suggest that investigation of differential modulation by emotional domains may yield important insight into mechanisms related to aberrant thought processes associated with delusions. In sum, this study supports the importance of considering specific emotional attributes (i.e., valence) in discrete domains of delusional thought (i.e., self-referential communication). The current findings warrant future investigation regarding the anatomical and functional mechanisms underlying aberrant processing of emotional salience and self-referential thought. This work would benefit from longitudinal studies with large samples capable of disentangling potential relations with patient characteristics. For instance, Fitzsimmons et al. (2014) recently reported differential relations of integrity of the cingulum with delusion severity between samples with first-episode psychosis versus more chronic schizophrenia. In addition, potential moderation by sex, medication, and relations to functional outcome will prove informative
5
(Fitzsimmons et al., 2014; Williams et al., 2007). Conflict of interest The authors confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome. Acknowledgement We thank S. Kapur, G. Remington, J. Addington, T. Schmitz, A. Anderson, A. Graff, M. Korostil, D.Mamo, P.Gerretsen, and F. Carravaggio, as well as W. Mar, A. Naber, P. Barsoum, H. Marcon, and Sofia Raitsin for their roles in the original study by Menon et al. (2011) at the Centre for Addiction and Mental Health (Toronto, ON). This project contributed in part to a thesis at Ryerson University by L. L., who is now at Laurentian University. References Aleman, A., & David, A. S. (2006). How to fill a half-full glass: Emotion and schizophrenia. American Psychologist, 61, 75e76. http://dx.doi.org/10.1037/ 0003-066X.61.1.75. Andreasen, N. C. (1984). The scale for the assessment of positive symptoms (SAPS). Iowa City, IA: University of Iowa. Appelbaum, P. S., & Grisso, T. (1995). The Macarthur treatment competence study. I. Mental illness and competence to consent to treatment. Law and Human Behavior, 19, 105e126. Blackwood, N. J., Bentall, R. P., Ffytche, D. H., Simmons, A., Murray, R. M., & Howard, R. J. (2004). Persecutory delusions and the determination of selfrelevance: An fMRI investigation. Psychological Medicine, 34, 591e596. Breier, A., & Berg, P. H. (1999). The psychosis of schizophrenia: Prevalence, response to atypical antipsychotics, and prediction of outcome. Biological Psychiatry, 46, 361e364. http://dx.doi.org/10.1016/S0006-3223(99)00040-2. Christensen, B. K., & Bilder, R. M. (2000). Dual cytoarchitectonic trends: An evolutionary model of frontal lobe functioning and its application to psychopathology. Canadian Journal of Psychiatry, 45, 247e256. Dolcos, F., & Denkova, E. (2008). Neural correlates of encoding emotional memories: A review of functional neuroimaging evidence. Cell Science Reviews, 5, 78e122. Fitzsimmons, J., Schneiderman, J. S., Whitford, T. J., Niznikiewicz, M. A., Pelavin, P. E., Terry, D. P., et al. (2014). Cingulum bundle diffusivity and delusions of reference in first episode and chronic schizophrenia. Psychiatry Research: Neuroimaging, 224, 124e132. http://dx.doi.org/10.1016/j.pscychresns.2014.08.002. Genovese, C. R., Lazar, N. A., & Nichols, T. (2002). Thresholding statistical maps in functional neuroimaginge using the false discovery rate. NeuroImage, 15, 870e878. http://dx.doi.org/10.1006/nimg.2001.1037. Gerber, A. J., Posner, J., Gorman, D., Colibazzi, T., Yu, S., Wang, Z., Kangarlu, A., et al. (2008). An affective circumplex model of neural systems subserving valence, arousal, and cognitive overlay during the appraisal of emotional faces. Neuropsychologia, 46, 2129e2139. Gibbs, A. A., & David, A. S. (2003). Delusion formation and insight in the context of affective disturbance. Epidemiologica e Psichiatria Sociale, 12, 167e174. http:// dx.doi.org/10.1017/S1121189X00002943. Holt, D. J., Lakshmanan, B., Freudenreich, O., Goff, D. C., Rauch, S. L., & Kuperberg, G. R. (2009). Dysfunction of a cortical midline network during emotional appraisals in schizophrenia. Schizophrenia Bulletin, 37, 164e176. Kapur, S. (2003). Psychosis as a state of aberrant salience: A framework linking biology, phenomenology, and pharmacology in schizophrenia. American Journal of Psychiatry, 160, 13e23. van der Meer, L., Costafreda, S., Aleman, A., & David, A. S. (2010). Self-reflection and the brain: A theoretical review and meta-analysis of neuroimaging studies with implications for schizophrenia. Neuroscience & Biobehavioral Reviews, 34, 935e946. Menon, V. (2011). Large-scale brain networks and psychopathology: A unifying triple network model. Trends in Cognitive Sciences, 15, 483e506. http:// dx.doi.org/10.1016/j.tics.2011.08.003. Menon, M., Schmitz, T. W., Anderson, A. K., Graff, A., Korostil, M., Mamo, D., Gerretsen, P., et al. (2011). Exploring the neural correlates of delusions of reference. Biological Psychiatry, 70, 1127e1133. http://dx.doi.org/10.1016/ j.biopsych.2011.05.037. Mickley, K. R., & Kensinger, E. A. (2008). Neural processes supporting subsequent recollection and familiarity of emotional items. Cognitive, Affective, & Behavioral Neuroscience, 8, 153e164. http://dx.doi.org/10.3758/CABN.8.2.143. Moran, J. M., Macrae, C. N., Heatherton, T. F., Wyland, C. L., & Kelley, W. M. (2006). Neuroanatomical evidence for distinct cognitive and affective components of self. Journal of Cognitive Neuroscience, 18, 1e9. Northoff, G., & Bermpohl, F. (2004). Cortical midline structures and the self. Trends
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007
6
T.A. Girard et al. / J. Behav. Ther. & Exp. Psychiat. xxx (2016) 1e6
in Cognitive Sciences, 8, 102e107. Northoff, G., Heinzel, A., Greek, M. D., Bermpohl, F., Dobrowolny, H., & Panksepp, J. (2006). Self-referential processing in our brainda meta-analysis of imaging studies on the self. Neuroimage, 31, 440e457. http://dx.doi.org/10.1016/ j.neuroimage.2005.12.002. Northoff, G., Schneider, F., Rotte, M., Matthiae, C., Tempelmann, C., Wiebking, C., et al. (2009). Differential parametric modulation of self-relatedness and emotions in different brain regions. Human Brain Mapping, 30, 369e382. http:// dx.doi.org/10.1002/hbm.20510. Phillips, L. K., & Seidman, L. J. (2008). Emotion processing in persons at risk for Schizophrenia. Schizophrenia Bulletin, 34, 888e903. Sass, L. A., & Parnas, J. (2003). Schizophrenia, consciousness, and the self. Schizophrenia Bulletin, 29, 427e444. Schmitz, T. W., & Johnson, S. C. (2007). Relevance to self: A brief review and framework of neural systems underlying appraisal. Neuroscience & Biobehavioral Reviews, 31, 585e596. Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., et al. (1998). The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for
DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59(suppl 20), 22e57. Siemerkus, J., Irle, E., Schmidt-Samoa, C., Dechent, P., & Weniger, G. (2012). Egocentric spatial learning in schizophrenia investigated with functional magnetic resonance imaging. NeuroImage: Clinical, 1, 153e163. Speechley, W. J., Woodward, T. S., & Ngan, E. T. (2013). Failure of conflict to modulate central executive network activity associated with delusions in schizophrenia. Frontiers in Psychiatry, 4, 113. http://dx.doi.org/10.3389/fpsyt.2013.00113. Startup, M., Bucci, S., & Langdon, R. (2009). Delusions of reference: A new theoretical model. Cognitive Neuropsychiatry, 14, 110e126. http://dx.doi.org/10.1080/ 13546800902864229. Wilkinson, G. S., & Jastak, J. (1993). Wide range achievement testdthird edition (WRAT-3). Wilmington, DE: Wide Range. Williams, L. M., Das, P., Liddell, B. J., Olivieri, G., Peduto, A. S., David, A. S., et al. (2007). Fronto-limbic and autonomic disjunctions to negative emotion distinguish schizophrenia subtypes. Psychiatry Research: Neuroimaging, 155, 29e44. http://dx.doi.org/10.1016/j.pscychresns.2006.12.018. Woods, S. W. (2003). Chlorpromazine equivalent doses for the newer atypical antipsychotics. Journal of Clinical Psychiatry, 64, 663e667.
Please cite this article in press as: Girard, T. A., et al., Aberrant modulation of brain activation by emotional valence during self-referential processing among patients with delusions of reference, Journal of Behavior Therapy and Experimental Psychiatry (2016), http://dx.doi.org/ 10.1016/j.jbtep.2016.11.007