SOCIAL NEUROSCIENCE, 2015 Vol. 10, No. 1, 27–34, http://dx.doi.org/10.1080/17470919.2014.959618
Facial affect recognition linked to damage in specific white matter tracts in traumatic brain injury Helen M. Genova1,2, Venkateswaran Rajagopalan1,2, Nancy Chiaravalloti1,2, Allison Binder3,4, John Deluca1,2, and Jeannie Lengenfelder1,2 1
Neuropsychology and Neuroscience Laboratory, Kessler Foundation Research Center, West Orange, NJ, USA 2 Department of Physical Medicine and Rehabilitation, Rutgers – New Jersey Medical School, Newark, NJ, USA 3 Department of Psychology, University of Massachusetts, Amherst, MA, USA 4 Department of Clinical Psychology, Suffolk University, Boston, MA, USA
Emotional processing deficits have recently been identified in individuals with traumatic brain injury (TBI), specifically in the domain of facial affect recognition. However, the neural networks underlying these impairments have yet to be identified. In the current study, 42 individuals with moderate to severe TBI and 23 healthy controls performed a task of facial affect recognition (Facial Emotion Identification Test (FEIT)) in order to assess their ability to identify and discriminate six emotions: happiness, sadness, anger, surprise, shame, and fear. These individuals also underwent structural neuroimaging including diffusion tensor imaging to examine white matter (WM) integrity. Correlational analyses were performed to determine where in the brain WM damage was associated with performance on the facial affect recognition task. Reduced performance on the FEIT was associated with reduced WM integrity (fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity) in the inferior longitudinal fasciculus and inferior-fronto-occipital fasciculus in individuals with TBI. Poor performance on the task was additionally associated with reduced gray matter (GM) volume in lingual gyrus and parahippocampal gyrus. The results implicate a pattern of WM and GM damage in TBI that may play a role in emotional processing impairments.
Keywords: Traumatic brain injury; Facial affect recognition; Diffusion tensor imaging; Emotional processing; Inferior longitudinal fasciculus; Inferior-fronto-occipital fasciculus.
Individuals with traumatic brain injury (TBI) can experience a host of symptoms including physical, cognitive, and those associated with mood. There is emerging evidence that individuals with TBI demonstrate deficits in emotional processing, specifically facial affect recognition (i.e., ability to identify emotions based on facial features) (Babbage et al., 2011). However, the neuropathology underlying this impairment is unclear. A network of brain
regions critical for facial affect recognition have been identified) and include the prefrontal cortex (i.e., ventromedial prefrontal cortex and the orbitofrontal cortex), limbic structures (i.e., amygdala, temporal lobes and fusiform gyrus, and regions of the parietal lobe) (Adolphs, 2010; Miyahara, Harada, Ruffman, Sadato, & Iidaka, 2013). One study to date by Philippi et al. (2009) has examined the relationship between white matter
Correspondence should be addressed to: Helen M. Genova, Neuropsychology and Neuroscience Laboratory, Kessler Foundation Research Center, West Orange, NJ, USA. E-mail: [email protected] This work was supported by the New Jersey Commission on Brain Injury Research (NJCBIR), Research Grants [10-3218-BIR-E-0] to JL and [CBIR13IRG026] to JL and HG.
© 2014 Taylor & Francis
28
GENOVA ET AL.
(WM) integrity and facial affect recognition in a “brain-damaged” sample and found that damage to association fiber tracts (inferior-fronto-occipital fasciculus (IFOF) and inferior longitudinal fasciculus (ILF)) (assessed by lesion mapping) connecting the occipital lobes to regions involved in facial affect recognition, including limbic structures, was associated with poorer performance on a facial affect recognition task. However, because this study included a mixed neurological sample, it precludes any definite conclusion regarding the impact of TBI on emotional processing, as TBI results in dual pathology of focal damage with additional traumatic axonal injury (Andriessen, Jacobs, & Vos, 2010; Johnson, Stewart, & Smith, 2013). It is thus important that individuals with TBI be examined to better understand the specific neuropathology underlying facial affect recognition impairments. The IFOF and ILF are both WM tracts that are of particular interest, therefore, in the current study. The ILF is a WM tract connecting the occipital and anterior temporal lobes (Catani, Jones, Donato, & Ffytche, 2003; Polyack, 1957). Its branches in the occipital lobe pass through extrastriate cortical regions, posterior lingual gyrus, fusiform gyri, and cuneus, and its branches in the anterior temporal lobe connect to the superior, middle and inferior temporal lobes as well as the uncus/parahippocampal gyrus near the amygdala (Catani et al., 2003). Several functions have been reported to involve the ILF, including facial recognition as well as processing of emotional stimuli. The branches of the IFOF connect the occipital cortex (superior, middle, and inferior) to the orbitofrontal cortex and pass through the medial temporal cortex and parietal lobe (Martino, Brogna, Robles, Vergani & Duffau, 2010). The IFOF’s roles include the sending of processed visual information from the occipital lobes to the frontal lobes as well as processing semantic information (Martino et al., 2010). It is likely that because both of these tracts have roles in connecting visual areas to limbic regions, they play a role in the processing of emotional stimuli. We hypothesized that (1) individuals with TBI would demonstrate impaired performance on a facial affect recognition task; (2) these impairments would be associated with reductions in WM integrity [indicated by reduced fractional anisotropy (FA) and elevated mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD)] specifically in the IFOF and ILF due to their connections between regions involved in the effective emotional processing of faces; and (3) reductions in gray matter (GM;
indicating atrophy) would be associated with facial affect recognition deficits.
METHOD Participants Participants consisted of 42 individuals with moderateto-severe TBI and 23 healthy controls (HC). All participants were between the ages of 19 and 60. Individuals with TBI were at least one year postinjury at the time of study participation. Injury severity was confirmed via medical record review and significant other interview. Moderate-to-severe TBI was defined by loss of consciousness greater than 30 minutes at the time of injury. TBI severity was confirmed via Glascow Coma Scale scores (where available), medical records, or corroboration with family members. Exclusionary criteria included a history of any neurologic disease/injury (other than the brain injury among TBI participants), a history of alcohol or drug abuse, or major psychiatric disturbance (defined by psychiatric inpatient stay or diagnosis of psychiatric disorder prior to brain injury). The individuals with TBI ranged in age from 19 to 55, (M = 38.83, SD = 11.66) with a mean of 13.76 years (SD = 1.76) of education. Healthy individuals ranged in age from 19 to 60 (M = 35.22, SD = 12.90) and had a mean of 14.83 years of education (SD = 1.78). There was no significant difference between the two groups on age (t(63) = 1.15, p = .254), but the HCs did have significantly more years of education than the TBI group (t(63) = −2.32, p = .024). Education was thus utilized as a covariate in the main analyses. There was no significant difference between groups in gender composition (X2 (1) = 2.042. p = .153). The individuals with TBI were 9.9 years since injury (range 13–311 months). The sample characteristics are described in Table 1. TABLE 1 Sample characteristics
Demographics Age Education Premorbid IQ (WTAR) Gender Depression CMDI Emotional processing FEIT Identification FEIT Discrimination
HC Mean (SD)
TBI Mean (SD)
p
35.2 (12.9) 13.8 (1.76) 111.4 (9.3) 10 M, 13 F
38.8 (11.7) 14.8 (1.8) 101.2 (13.7) 26 M, 16 F
.245 .024 .019 .153
44.6 (7.7)
56.6 (14.6)
.001
14.0 (2.3) 26.4 (2.1)
11.7 (2.7) 24.9 (2.8)
.001 .037
Note: WTAR, Weschler Test of Adult Reading.
FACIAL AFFECT RECOGNTION LINKED
Procedures All procedures were approved by the Institutional Review Boards of Kessler Foundation Research Center and the University of Medicine and Dentistry of New Jersey (now Rutgers, New Jersey Medical School). Informed consent was obtained prior to enrollment in the study. All participants were paid for their participation. All participants were administered a battery of neuropsychological measures which assessed premorbid intellectual functioning, mood, and facial affect recognition. Cognitive measures were also administered, but will be reported elsewhere. Diffusion tensor imaging (DTI) procedures were completed by all participants on a different day, within 2 weeks of the neuropsychological testing.
Neuroimaging procedures Neuroimaging was performed on a Siemens 3T Allegra scanner. A high-resolution 3D anatomical T1-weighted image was obtained using magnetization prepared acquisition of a gradient echo sequence. The imaging parameters include TR = 2500 ms, TE = 3.5 ms, flip angle = 8°, effective TI = 1200 ms, 256 × 256 matrix, FOV = 256 mm, NEX = 1, 192 slices, in-plane resolution = 1 × 1 mm, 1 mm slice thickness). GM voxel-based morphometry (VBM) analysis was performed using the T1-weighted image.
29
suggesting that it is a reliable and valid measure of premorbid IQ (Green et al., 2008). The Facial Emotion Identification Test (FEIT) (Kerr & Neale, 1993) consists of two tasks. (1) Identification: 19 black-and-white photographs of models with differing facial emotions are presented on a computer screen. After each photograph, the subjects were asked to identify which one of six emotions (i.e., happy, angry, afraid, sad, surprised, and ashamed) best described the emotion in the photograph. The DV in the current study was the number of correct responses, with normal performance demonstrated to fall at a total score of 14 (SD = 2) in normative studies (Kerr & Neale, 1993). (2) Discrimination: a set of 30 pairs of photographs are presented to the subject, side by side, and the subject is asked to determine if the pair of photographs are showing the same or different emotion. The DV in the current study is the total number of correct responses. The Chicago Multiscale Depression Inventory (CMDI) (Nyenhuis et al., 1998) was developed as a measure of depression specifically for use with neurological samples. The CMDI consists of three subscales (mood, evaluative, and vegetative) which measure different components of depression. The participants respond on a 5-point Likert scale and rate the extent to which a word or brief phrase describes them during the past week with 1 indicating “not at all” and 5 indicating “extremely”. The DV in the current study was the total score.
Data analysis Diffusion tensor imaging Statistical analysis DTI data were acquired using a 12-directional echoplanar sequence (TR = 7300 ms, TE = 88 ms, FOV = 210 mm, matrix = 128 × 128, slice thickness 4 mm, 26 slices, no gap, b = 1000 s/mm2, NEX = 7). After acquisition, data were transferred off-line to a Linux-based workstation and processed using FSL software (FMRIB software library, Functional Magnetic Resonance Imaging of the Brain Software Library, http://www.fmrib.ox.au.uk/fsl). Neuropsychological performance Weschler Test of Adult Reading (WTAR) (Wechsler, 2001) is a list of 50 words used to assess premorbid (preinjury) level of intellectual functioning. Subjects are asked to read the words aloud, and the dependent variable (DV) is the number of correctly pronounced words. Research indicates that WTAR performance remains stable following TBI recovery,
Statistical analyses on performance variables were performed using SPSS. Analyses of covariance were performed to examine behavioral differences between TBI and HC groups on mood (CMDI), premorbid IQ (WTAR), and emotional processing measures (discrimination and identification subscores on the FEIT). Education level served as a covariate in all analyses given the differences in education level between the groups. Pearson correlations were performed to examine associations between emotional processing performance and depression, demographic variables (age and education), and premorbid intelligence. Correlational analyses were run on the entire sample (collapsed across both groups) and then each group (HC, TBI) individually. An additional correlation was run to examine the relationship between months since TBI and performance on the emotional processing measures.
30
GENOVA ET AL.
Tract-based spatial statistics FSL’s Brain Extraction Tool (BET) was used to remove all nonbrain tissue from each image by creating a brain mask from the b = 0 (nondiffusion weighted) images. Each subject’s data were then corrected for distortions due to eddy currents and head motion using affine transformations. Finally, using DTIFIT (http://fsl. fmrib.ox.ac.uk/fsl/fsl4.0/fdt/fdt_dtifit.html) from FSL, a diffusion tensor model was fitted at each voxel and maps of FA, MD, AD, and RD were obtained. After the preprocessing steps were completed, Tract-Based Spatial Statistics (TBSS) was performed on FA, MD, AD, and RD in FSL 5.0. (Smith et al., 2006) Briefly, TBSS include the following steps: all subjects’ FA maps were aligned to MNI152 space using FNIRT (nonlinear registration). FA skeleton was created from the mean FA skeleton image. Each subject’s FA image was then projected onto this skeleton image, and the resulting images were subjected to group comparison by regressing out age and education. Significance was set at p < 0.05, corrected for multiple comparisons using family-wise error rate (FWER) to compare controls with TBI participants. Correlation between performances on the FEIT score with DTI metrics (in the tracts from TBSS skeleton) was performed to identify WM tracts that show changes with FEIT scores in TBI group. Voxel-based morphometry GM VBM analysis was carried out using openware FSL version 5.0 (http://www.fmrib.ox.ac.uk/fsl/) (Smith et al., 2004) adopting the standard VBM pipeline. An optimized VBM approach, developed by Good and colleagues, was adopted (Ashburner & Friston, 2000; Good et al., 2001). Data processing was divided into four major steps: (1) brain extraction was performed using BET (Smith, 2002) adopting the methods of Popescu et al. (2012). (2) Brain extracted images were segmented into WM, GM, and cerebrospinal fluid volume probability maps using FAST (http://fsl.fmrib. ox.ac.uk/fsl/fslwiki/FAST; Zhang, Brady, & Smith, 2001). (3) A study-specific GM template was created. This was done by registering the 23 HCs and 23 of the 42 TBI subjects (randomly chosen to avoid bias during the registration process) into the MNI152 space using the affine registration tool FLIRT (Jenkinson & Smith, 2001; Zhang et al., 2001) and then by nonlinear registration using FNIRT (www.fmrib.ox.ac.uk/analysis/techrep). The resulting images were averaged to create the template. (4) All the native GM images were nonlinearly re-registered to the template and modulated. (5) These images were then smoothed using a full-width half-
maximum of 7 mm. (6) General linear model was used to compare voxel-wise differences in GM volume between the TBI and HC groups after regressing out age and education. Nonparametric statistics was performed using “randomize” with 5000 permutations and using threshold-free cluster enhancement option. Also, GM volume in TBI patients was correlated with FEIT discrimination and identification scores. A p < 0.05 corrected for multiple comparisons using FWER was used as the level of significance.
RESULTS Behavioral performance The behavioral performance is described in Table 1. After controlling for variance associated with education, the TBI group demonstrated significantly increased levels of depression on the CMDI total score compared to HCs (F(1, 55) = 12.39, p = .001). Similarly, on the WTAR, individuals with TBI demonstrated significantly lower estimated premorbid IQ (F(1, 62) = 5.77, p = .019) after controlling for variance associated with education. The TBI group also performed significantly worse than HCs on both the discrimination trial (F(1, 62) = 4.53, p = .037) and the identification trial (F(1, 62) = 11.22, p = .001) of the FEIT, after controlling for differences between the groups in education. Correlational analyses indicated that in either group individually, or the sample as a whole, no significant correlations were found between demographic variables (age, education) and either emotional processing variables. Further, no correlations were found between depression and either emotional processing measure in the whole sample, or in either group. However, in the TBI group only, there was a significant association between premorbid intelligence and FEIT identification (R = .319, p = .039), indicating that individuals with higher premorbid intelligence performed better on the identification trial of the FEIT. No relationship was observed between time since injury and accuracy on either FEIT identification or discrimination.
Relationship between WM integrity and facial affect recognition A significant relationship was noted between reduced performance on the discrimination trial of the FEIT and reduced WM integrity in two specific WM tracts: right IFOF (x = 33, y = 6, z = 1, Figure 1) and right
FACIAL AFFECT RECOGNTION LINKED
ILF (x = 48, y = 0, z = −17, Figure 2). In the IFOF, there was a positive correlation between FA and FEIT discrimination (p < .05). In the ILF, there was a
31
significant negative correlation between AD, MD, and RD with FEIT discrimination (p < .05). No significant relationships were observed between DTI metrics and the identification trial of the FEIT.
Differences in DTI metrics in ROIs between groups The TBSS analysis demonstrated that WM integrity in the ILF and IFOF was significantly lower in persons with TBI as compared with HCs. In the IFOF, AD, MD, and RD values in the TBI group were significantly higher than the HCs (p < .05), but no significant difference in FA values was observed between groups in that tract. In the ILF, significantly higher MD and RD levels were noted in the TBI group compared to HCs (p < .05), but no significant difference was noted between the groups in FA or AD.
Relationship between GM volume and facial affect recognition Figure 1. A significant relationship was noted between reduced performance on the discrimination trial of the FEIT and reduced WM integrity in right IFOF. A, B, and C refer to coronal, sagittal and axial images, respectively.
Voxel-wise correlations were run to examine the relationship between GM volume and performance on FEIT discrimination and FEIT identification. One cluster encompassing two regions was observed to be positively correlated with FEIT identification: right lingual gyrus and right hippocampus/parahippocampus (p < .05) (see Figure 3). Reduced GM volume in these regions was associated with poorer performance on the FEIT. No significant correlations were noted between FEIT discrimination and GM volume.
DISCUSSION
Figure 2. A significant relationship was noted between reduced performance on the discrimination trial of the FEIT and reduced WM integrity in right ILF. A, B, and C refer to coronal, sagittal and axial images, respectively.
The current study examined the relationship between TBI-related pathology and performance on a facial affect recognition task. Individuals with TBI were significantly more impaired on a task of facial affect recognition, both in terms of identifying emotions and in terms of discriminating between emotions. Reduced WM integrity in the ILF and IFOF was associated with poor performance on the facial affect recognition task. Additionally, atrophy of the hippocampus/parahippocampus and lingual gyrus was associated with impaired facial affect recognition. These findings demonstrate that pathology of both WM and GM contributes to impairment in the
32
GENOVA ET AL.
Figure 3. Reduced GM volume in right lingual gyrus and right hippocampus/parahippocampus was associated with poorer performance on the FEIT. A, B, and C refer to coronal, sagittal and axial images, respectively.
ability to identify emotions of faces in persons with TBI. The finding that reduced WM integrity in the ILF and IFOF was related to worse facial affect recognition abilities in TBI is consistent with what has been reported in other clinical populations. For example, in individuals with Parkinson’s disease, FA levels of the IFOF and ILF have been shown to be associated with impaired ability to identify specific emotions (Baggio et al., 2012). Similarly, WM lesions of the ILF and IFOF were associated with an impaired ability to identify facial expression in individuals with multiple sclerosis (Mike et al., 2013). In autism, a disorder in which facial affect recognition impairments are prevalent, multiple studies have found abnormal diffusion compared to HCs in the ILF and IFOF (Jou et al., 2011; Shukla & Keehn, 2011). Interestingly, these findings suggest a “disconnection,” where compromised connections between several critical regions (such as occipital lobes, limbic structures), rather than damage only to the regions themselves, can result in impairment in emotional processing. Additionally, in the current study, compromised GM (atrophy) in the TBI group was noted to be related to impaired facial affect recognition, specifically in the lingual gyrus and the hippocampal/parahippocampal regions. While the lingual gyrus has been shown to be involved in the perception of faces (Puce, Allison,
Asgari, Gore, & McCarthy, 1996), there is a growing body of evidence for a specific role in the processing of emotional facial expressions (e.g., Fusar-Poli et al., 2009). The hippocampus and parahippocampal regions have also been shown to be involved in identification of emotional facial expressions (Gur et al., 2002; Trautmann, Fehr, & Herrmann, 2009). In multiple studies involving functional magnetic resonance imaging (fMRI) examining facial emotion processing in individuals with major depressive disorder, the hippocampus and parahippocampal gyrus show abnormal brain activation patterns compared to HCs (Fu et al., 2008; Surguladze et al., 2005; Suslow et al., 2010). Thus, using different neuroimaging methodologies, the converging evidence suggests that damage to the hippocampus/parahippocampal region and lingual gyrus via either atrophy or insult results in impaired emotional processing in TBI. Interestingly, in our study, we found a relationship between premorbid intelligence and performance on the emotional processing measure, suggesting that certain preinjury variables may influence one’s emotional processing abilities. To date, there is little to no research focused on the relationship between premorbid intelligence and facial affect recognition, but recent research has shown that there is a significant relationship between facial affect recognition and levels of cognitive functioning following a brain injury. For example, it has been shown that functioning in other cognitive domains such as verbal and nonverbal memory, processing speed, and working memory are associated with facial affect recognition (Yim, Babbage, Zupan, Neumann, & Willer, 2013). Examination of the relationship between emotional processing and both premorbid intelligence and cognitive outcome following a TBI may be an interesting avenue for future research. Our findings represent an important first step in understanding impaired emotional processing in individuals with TBI. Although impairments in emotional processing in TBI have been well documented (specifically in regards facial affect recognition), there is a lack of understanding as to why this deficit exists, or when these deficits occur. For example, it is possible that damage to brain structure leads to impaired facial affect recognition (as our findings in the WM tracts suggest). With the passage of time, impaired emotional processing in individuals with TBI may lead to reduced social functioning and a shrinking social network, reinforcing the neural networks which are impaired, leading to even greater emotional processing impairments (Kennedy & Adolphs, 2012). This cyclical relationship (impaired emotional processing —> decreased social functioning —> greater impaired emotional processing) may continue throughout the life span of the individual with TBI.
FACIAL AFFECT RECOGNTION LINKED
Therefore, an important future avenue of research would be to determine when in the recovery process these impairments arise, so that the deficits could be potentially treated. There were several limitations to the current study. First, the task that we used, while examining multiple emotions, does not lend itself to easy examination of individual emotions. The number of trials for each emotion is not equally distributed, and therefore, it is difficult to examine which emotions were more difficult for the individuals with TBI to identify. Examining the emotions separately will enable researchers to further delineate and identify distinct emotional processing networks of the brain. Additionally, because the DTI data were acquired as part of a larger fMRI study, the number of directions was limited which may result in reduced signal. Finally, the sample of individuals with TBI showed higher levels of depression than HCs. Although this is expected, it is hard to determine whether the impairments observed in the current study were due to the TBI or the depression, as individuals with depression have been shown to have both WM and GM abnormalities and impairments in facial affect recognition abilities (Aldinger et al., 2013; Bessette, Nave, Caprihan, & Stevens, 2013; Grieve, Korgaonkar, Koslow, Gordon, & Williams, 2013) In conclusion, damage to WM connections and GM atrophy in individuals with TBI is linked to impairments in a key social behavior: the ability to recognize facial affect. The association between this damage and impairments in distinguishing emotions occurred in specific WM tracts and GM regions involved in pathways connecting the occipital lobes to the orbital frontal cortex and involves key limbic structures. Functional neuroimaging may provide additional avenues of research in which aspects of functional connectivity can be examined in an effort to better understand the relationship between emotional processing abilities and brain networks involved in social functioning. Original manuscript received 7 April 2014 Revised manuscript accepted 25 August 2014 First published online 16 September 2014
REFERENCES Adolphs, R. (2010). What does the amygdala contribute to social cognition?. Annals of the New York Academy of Sciences, 1191, 42–61. doi:10.1111/j.1749-6632.2010. 05445.x Aldinger, M., Stopsack, M., Barnow, S., Rambau, S., Spitzer, C., Schnell, K., & Ulrich, I. (2013). The
33
association between depressive symptoms and emotion recognition is moderated by emotion regulation. Psychiatry Research, 205(1–2), 59–66. doi:10.1016/j. psychres.2012.08.032 Andriessen, T. M., Jacobs, B., & Vos, P. E. (2010). Clinical characteristics and pathophysiological mechanisms of focal and diffuse traumatic brain injury. Journal of Cellular and Molecular Medicine, 14(10), 2381–2392. doi:10.1111/j.1582-4934.2010.01164.x Ashburner, J., & Friston, K. J. (2000). Voxel-based morphometry: The methods. Neuroimage, 11(6), 805–821. doi:10.1006/nimg.2000.0582 Babbage, D. R., Yim, J., Zupan, B., Neumann, D., Tomita, M. R., & Willer, B. (2011). Meta-analysis of facial affect recognition difficulties after traumatic brain injury. Neuropsychology, 25(3), 277–285. doi:10.1037/a0021908 Baggio, H. C., Segura, B., Ibarretxe-Bilbao, N., Valldeoriola, F., Marti, M. J., & Compta, Y. (2012). Structural correlates of facial emotion recognition deficits in Parkinson’s disease patients. Neuropsychologia, 50(8), 2121–2128. doi:10.1016/j.neuropsychologia.2012.05.020 Bessette, K. L., Nave, A. M., Caprihan, A., & Stevens, M. C. (2013). White matter abnormalities in adolescents with major depressive disorder. Brain Imaging and Behavior, 1–11. Advance online publication. doi:10.1007/s11682013-9274-8 Catani, M., Jones, D. K., Donato, R., & Ffytche, D. H. (2003). Occipito-temporal connections in the human brain. Brain, 126(9), 2093–2107. doi:10.1093/brain/awg203 Fu, C. H., Williams, S. C., Cleare, A. J., Scott, J., Mitterschiffthaler, M. T., Walsh, N. D., & Murray, R. M. (2008). Neural responses to sad facial expressions in major depression following cognitive behavioral therapy. Biological Psychiatry, 64(6), 505–512. doi:10.1016/j.biopsych.2008.04.033 Fusar-Poli, P., Placentino, A., Carletti, F., Allen, P., Landi, P., Abbamonte, M., & Politi, P. L. (2009). Laterality effect on emotional faces processing: ALE meta-analysis of evidence. Neuroscience Letters, 452(3), 262–267. doi:10.1016/j.neulet.2009.01.065 Good, C. D., Johnsrude, I. S., Ashburner, J., Henson, R. N., Friston, K. J., & Frackowiak, R. S. (2001). A voxelbased morphometric study of ageing in 465 normal adult human brains. Neuroimage, 14(1), 21–36. doi:10.1006/nimg.2001.0786 Green, R. E. A., Melo, B., Christensen, B., Ngo, L., Monette, G., & Bradbury, C. (2008). Measuring premorbid IQ in traumatic brain injury: An examination of the validity of the Wechsler Test of Adult Reading (WTAR). Journal of Clinical Experimental Neuropsychology, 30(2), 163–172. doi:10.1080/ 13803390701300524 Grieve, S. M., Korgaonkar, M. S., Koslow, S. H., Gordon, E., & Williams, L. M. (2013). Widespread reductions in gray matter volume in depression. NeuroImage: Clinical, 3, 332–339. doi:10.1016/j.nicl.2013.08.016 Gur, R. C., Schroeder, L., Turner, T., McGrath, C., Chan, R. M., Turetsky, B. I., & Gur, R. E. (2002). Brain activation during facial emotion processing. Neuroimage, 16(3), 651– 662. doi:10.1006/nimg.2002.1097 Jenkinson, M., & Smith, S. (2001). A global optimisation method for robust affine registration of brain images. Medical Image Analysis, 5(2), 143–156. doi:10.1016/ S1361-8415(01)00036-6
34
GENOVA ET AL.
Johnson, V. E., Stewart, W., & Smith, D. H. (2013). Axonal pathology in traumatic brain injury. Experimental Neurology, 246, 35–43. doi:10.1016/j.expneurol. 2012.01.013 Jou, R. J., Jackowski, A. P., Papademetris, X., Rajeevan, N., Staib, L. H., & Volkmar, F. R. (2011). Diffusion tensor imaging in autism spectrum disorders: Preliminary evidence of abnormal neural connectivity. Australian and New Zealand Journal of Psychiatry, 45(2), 153–162. doi:10.3109/00048674.2010.534069 Kennedy, D. P., & Adolphs, R. (2012). The social brain in psychiatric and neurological disorders. Trends in Cognitive Sciences, 16(11), 559–572. doi:10.1016/j. tics.2012.09.006 Kerr, S. L., & Neale, J. M. (1993). Emotion perception in schizophrenia: Specific deficit or further evidence of generalized poor performance?. Journal of Abnormal Psychology, 102(2), 312–318. doi:10.1037/0021843X.102.2.312 MAGNIMS Study Group, Popescu, V., Battaglini, M., Hoogstrate, W. S., Verfaillie, S. C., Sluimer, I. C., & Van Schijndel, R. A. (2012). Optimizing parameter choice for FSL-Brain Extraction Tool (BET) on 3D T1 images in multiple sclerosis. Neuroimage, 61(4), 1484– 1494. Martino, J., Brogna, C., Robles, S. G., Vergani, F., & Duffau, H. (2010). Anatomic dissection of the inferior fronto-occipital fasciculus revisited in the lights of brain stimulation data. Cortex, 46(5), 691–699. doi:10.1016/j. cortex.2009.07.015 Mike, A., Strammer, E., Aradi, M., Orsi, G., Perlaki, G., Hajnal, A., & Illes, Z. (2013). Disconnection mechanism and regional cortical atrophy contribute to impaired processing of facial expressions and theory of mind in multiple sclerosis: A structural MRI study. PLoS One, 8(12), e82422. doi:10.1371/journal.pone.0082422 Miyahara, M., Harada, T., Ruffman, T., Sadato, N., & Iidaka, T. (2013). Functional connectivity between amygdala and facial regions involved in recognition of facial threat. Social Cognitive & Affective Neuroscience, 8(2), 181–189. doi:10.1093/scan/nsr085 Nyenhuis, D. L., Luchetta, T., Yamamoto, C., Terrien, A., Bernardin, L., Rao, S. M., & Garron, D. C. (1998). The development, standardization, and initial validation of the Chicago Multiscale Depression Inventory. Journal of Personality Assessment, 70(2), 386–401. doi:10.1207/ s15327752jpa7002_14 Philippi, C. L., Mehta, S., Grabowski, T., Adolphs, R., & Rudrauf, D. (2009). Damage to association fiber tracts impairs recognition of the facial expression of emotion. The Journal of Neuroscience, 29(48), 15089–15099. doi:10.1523/JNEUROSCI.0796-09.2009 Polyack, S. (1957). The vertebrate visual system. University of Chicago Press.
Puce, A., Allison, T., Asgari, M., Gore, J. C., & McCarthy, G. (1996). Differential sensitivity of human visual cortex to faces, letterstrings, and textures: A functional magnetic resonance imaging study. The Journal of Neuroscience, 16(16), 5205–5215. Shukla, D. K., Keehn, B., & Müller, R.-A. (2011). Tractspecific analyses of diffusion tensor imaging show widespread white matter compromise in autism spectrum disorder. Journal of Child Psychology and Psychiatry, 52(3), 286–295. doi:10.1111/j.1469-7610.2010.02342.x Smith, S. M. (2002). Fast robust automated brain extraction. Human Brain Mapping, 17(3), 143–155. doi:10.1002/ hbm.10062 Smith, S. M., Jenkinson, M., Johansen-Berg, H., Rueckert, D., Nichols, T. E., Mackay, C. E., & Behrens, T. E. (2006). Tract-based spatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage, 31(4), 1487–1505. doi:10.1016/j.neuroimage.2006.02.024 Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E., Johansen-Berg, H., & Matthews, P. M. (2004). Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage, 23(Suppl 1), SS208–219. doi:10.1016/j.neuroimage. 2004.07.051 Surguladze, S., Brammer, M. J., Keedwell, P., Giampietro, V., Young, A. W., Travis, M. J., & Williams, S. C. (2005). A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder. Biological Psychiatry, 57(3), 201–209. doi:10.1016/j.biopsych.2004.10.028 Suslow, T., Konrad, C., Kugel, H., Rumstadt, D., Zwitserlood, P., Schöning, S., . . . Dannlowski, U. (2010). Automatic mood-congruent amygdala responses to masked facial expressions in major depression. Biological Psychiatry, 67(2), 155–160. doi:10.1016/j. biopsych.2009.07.023 Trautmann, S. A., Fehr, T., & Herrmann, M. (2009). Emotions in motion: Dynamic compared to static facial expressions of disgust and happiness reveal more widespread emotion-specific activations. Brain Research, 1284, 100–115. doi:10.1016/j.brainres. 2009.05.075 Wechsler, D. (2001). The Wechsler Test of Adult Reading. San Antonio, TX: The Psychological Corporation. Yim, J., Babbage, D. R., Zupan, B., Neumann, D., & Willer, B. (2013). The relationship between facial affect recognition and cognitive functioning after traumatic brain injury. Brain Injury, 27(10), 1155–1161. doi:10.3109/ 02699052.2013.804203 Zhang, Y., Brady, M., & Smith, S. (2001). Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Transactions on Medical Imaging, 20(1), 45–57. doi:10.1109/42.906424
Copyright of Social Neuroscience is the property of Psychology Press (UK) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Facial affect recognition linked to damage in specific white matter tracts in traumatic brain injury Helen M. Genova1,2, Venkateswaran Rajagopalan1,2, Nancy Chiaravalloti1,2, Allison Binder3,4, John Deluca1,2, and Jeannie Lengenfelder1,2 1
Neuropsychology and Neuroscience Laboratory, Kessler Foundation Research Center, West Orange, NJ, USA 2 Department of Physical Medicine and Rehabilitation, Rutgers – New Jersey Medical School, Newark, NJ, USA 3 Department of Psychology, University of Massachusetts, Amherst, MA, USA 4 Department of Clinical Psychology, Suffolk University, Boston, MA, USA
Emotional processing deficits have recently been identified in individuals with traumatic brain injury (TBI), specifically in the domain of facial affect recognition. However, the neural networks underlying these impairments have yet to be identified. In the current study, 42 individuals with moderate to severe TBI and 23 healthy controls performed a task of facial affect recognition (Facial Emotion Identification Test (FEIT)) in order to assess their ability to identify and discriminate six emotions: happiness, sadness, anger, surprise, shame, and fear. These individuals also underwent structural neuroimaging including diffusion tensor imaging to examine white matter (WM) integrity. Correlational analyses were performed to determine where in the brain WM damage was associated with performance on the facial affect recognition task. Reduced performance on the FEIT was associated with reduced WM integrity (fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity) in the inferior longitudinal fasciculus and inferior-fronto-occipital fasciculus in individuals with TBI. Poor performance on the task was additionally associated with reduced gray matter (GM) volume in lingual gyrus and parahippocampal gyrus. The results implicate a pattern of WM and GM damage in TBI that may play a role in emotional processing impairments.
Keywords: Traumatic brain injury; Facial affect recognition; Diffusion tensor imaging; Emotional processing; Inferior longitudinal fasciculus; Inferior-fronto-occipital fasciculus.
Individuals with traumatic brain injury (TBI) can experience a host of symptoms including physical, cognitive, and those associated with mood. There is emerging evidence that individuals with TBI demonstrate deficits in emotional processing, specifically facial affect recognition (i.e., ability to identify emotions based on facial features) (Babbage et al., 2011). However, the neuropathology underlying this impairment is unclear. A network of brain
regions critical for facial affect recognition have been identified) and include the prefrontal cortex (i.e., ventromedial prefrontal cortex and the orbitofrontal cortex), limbic structures (i.e., amygdala, temporal lobes and fusiform gyrus, and regions of the parietal lobe) (Adolphs, 2010; Miyahara, Harada, Ruffman, Sadato, & Iidaka, 2013). One study to date by Philippi et al. (2009) has examined the relationship between white matter
Correspondence should be addressed to: Helen M. Genova, Neuropsychology and Neuroscience Laboratory, Kessler Foundation Research Center, West Orange, NJ, USA. E-mail: [email protected] This work was supported by the New Jersey Commission on Brain Injury Research (NJCBIR), Research Grants [10-3218-BIR-E-0] to JL and [CBIR13IRG026] to JL and HG.
© 2014 Taylor & Francis
28
GENOVA ET AL.
(WM) integrity and facial affect recognition in a “brain-damaged” sample and found that damage to association fiber tracts (inferior-fronto-occipital fasciculus (IFOF) and inferior longitudinal fasciculus (ILF)) (assessed by lesion mapping) connecting the occipital lobes to regions involved in facial affect recognition, including limbic structures, was associated with poorer performance on a facial affect recognition task. However, because this study included a mixed neurological sample, it precludes any definite conclusion regarding the impact of TBI on emotional processing, as TBI results in dual pathology of focal damage with additional traumatic axonal injury (Andriessen, Jacobs, & Vos, 2010; Johnson, Stewart, & Smith, 2013). It is thus important that individuals with TBI be examined to better understand the specific neuropathology underlying facial affect recognition impairments. The IFOF and ILF are both WM tracts that are of particular interest, therefore, in the current study. The ILF is a WM tract connecting the occipital and anterior temporal lobes (Catani, Jones, Donato, & Ffytche, 2003; Polyack, 1957). Its branches in the occipital lobe pass through extrastriate cortical regions, posterior lingual gyrus, fusiform gyri, and cuneus, and its branches in the anterior temporal lobe connect to the superior, middle and inferior temporal lobes as well as the uncus/parahippocampal gyrus near the amygdala (Catani et al., 2003). Several functions have been reported to involve the ILF, including facial recognition as well as processing of emotional stimuli. The branches of the IFOF connect the occipital cortex (superior, middle, and inferior) to the orbitofrontal cortex and pass through the medial temporal cortex and parietal lobe (Martino, Brogna, Robles, Vergani & Duffau, 2010). The IFOF’s roles include the sending of processed visual information from the occipital lobes to the frontal lobes as well as processing semantic information (Martino et al., 2010). It is likely that because both of these tracts have roles in connecting visual areas to limbic regions, they play a role in the processing of emotional stimuli. We hypothesized that (1) individuals with TBI would demonstrate impaired performance on a facial affect recognition task; (2) these impairments would be associated with reductions in WM integrity [indicated by reduced fractional anisotropy (FA) and elevated mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD)] specifically in the IFOF and ILF due to their connections between regions involved in the effective emotional processing of faces; and (3) reductions in gray matter (GM;
indicating atrophy) would be associated with facial affect recognition deficits.
METHOD Participants Participants consisted of 42 individuals with moderateto-severe TBI and 23 healthy controls (HC). All participants were between the ages of 19 and 60. Individuals with TBI were at least one year postinjury at the time of study participation. Injury severity was confirmed via medical record review and significant other interview. Moderate-to-severe TBI was defined by loss of consciousness greater than 30 minutes at the time of injury. TBI severity was confirmed via Glascow Coma Scale scores (where available), medical records, or corroboration with family members. Exclusionary criteria included a history of any neurologic disease/injury (other than the brain injury among TBI participants), a history of alcohol or drug abuse, or major psychiatric disturbance (defined by psychiatric inpatient stay or diagnosis of psychiatric disorder prior to brain injury). The individuals with TBI ranged in age from 19 to 55, (M = 38.83, SD = 11.66) with a mean of 13.76 years (SD = 1.76) of education. Healthy individuals ranged in age from 19 to 60 (M = 35.22, SD = 12.90) and had a mean of 14.83 years of education (SD = 1.78). There was no significant difference between the two groups on age (t(63) = 1.15, p = .254), but the HCs did have significantly more years of education than the TBI group (t(63) = −2.32, p = .024). Education was thus utilized as a covariate in the main analyses. There was no significant difference between groups in gender composition (X2 (1) = 2.042. p = .153). The individuals with TBI were 9.9 years since injury (range 13–311 months). The sample characteristics are described in Table 1. TABLE 1 Sample characteristics
Demographics Age Education Premorbid IQ (WTAR) Gender Depression CMDI Emotional processing FEIT Identification FEIT Discrimination
HC Mean (SD)
TBI Mean (SD)
p
35.2 (12.9) 13.8 (1.76) 111.4 (9.3) 10 M, 13 F
38.8 (11.7) 14.8 (1.8) 101.2 (13.7) 26 M, 16 F
.245 .024 .019 .153
44.6 (7.7)
56.6 (14.6)
.001
14.0 (2.3) 26.4 (2.1)
11.7 (2.7) 24.9 (2.8)
.001 .037
Note: WTAR, Weschler Test of Adult Reading.
FACIAL AFFECT RECOGNTION LINKED
Procedures All procedures were approved by the Institutional Review Boards of Kessler Foundation Research Center and the University of Medicine and Dentistry of New Jersey (now Rutgers, New Jersey Medical School). Informed consent was obtained prior to enrollment in the study. All participants were paid for their participation. All participants were administered a battery of neuropsychological measures which assessed premorbid intellectual functioning, mood, and facial affect recognition. Cognitive measures were also administered, but will be reported elsewhere. Diffusion tensor imaging (DTI) procedures were completed by all participants on a different day, within 2 weeks of the neuropsychological testing.
Neuroimaging procedures Neuroimaging was performed on a Siemens 3T Allegra scanner. A high-resolution 3D anatomical T1-weighted image was obtained using magnetization prepared acquisition of a gradient echo sequence. The imaging parameters include TR = 2500 ms, TE = 3.5 ms, flip angle = 8°, effective TI = 1200 ms, 256 × 256 matrix, FOV = 256 mm, NEX = 1, 192 slices, in-plane resolution = 1 × 1 mm, 1 mm slice thickness). GM voxel-based morphometry (VBM) analysis was performed using the T1-weighted image.
29
suggesting that it is a reliable and valid measure of premorbid IQ (Green et al., 2008). The Facial Emotion Identification Test (FEIT) (Kerr & Neale, 1993) consists of two tasks. (1) Identification: 19 black-and-white photographs of models with differing facial emotions are presented on a computer screen. After each photograph, the subjects were asked to identify which one of six emotions (i.e., happy, angry, afraid, sad, surprised, and ashamed) best described the emotion in the photograph. The DV in the current study was the number of correct responses, with normal performance demonstrated to fall at a total score of 14 (SD = 2) in normative studies (Kerr & Neale, 1993). (2) Discrimination: a set of 30 pairs of photographs are presented to the subject, side by side, and the subject is asked to determine if the pair of photographs are showing the same or different emotion. The DV in the current study is the total number of correct responses. The Chicago Multiscale Depression Inventory (CMDI) (Nyenhuis et al., 1998) was developed as a measure of depression specifically for use with neurological samples. The CMDI consists of three subscales (mood, evaluative, and vegetative) which measure different components of depression. The participants respond on a 5-point Likert scale and rate the extent to which a word or brief phrase describes them during the past week with 1 indicating “not at all” and 5 indicating “extremely”. The DV in the current study was the total score.
Data analysis Diffusion tensor imaging Statistical analysis DTI data were acquired using a 12-directional echoplanar sequence (TR = 7300 ms, TE = 88 ms, FOV = 210 mm, matrix = 128 × 128, slice thickness 4 mm, 26 slices, no gap, b = 1000 s/mm2, NEX = 7). After acquisition, data were transferred off-line to a Linux-based workstation and processed using FSL software (FMRIB software library, Functional Magnetic Resonance Imaging of the Brain Software Library, http://www.fmrib.ox.au.uk/fsl). Neuropsychological performance Weschler Test of Adult Reading (WTAR) (Wechsler, 2001) is a list of 50 words used to assess premorbid (preinjury) level of intellectual functioning. Subjects are asked to read the words aloud, and the dependent variable (DV) is the number of correctly pronounced words. Research indicates that WTAR performance remains stable following TBI recovery,
Statistical analyses on performance variables were performed using SPSS. Analyses of covariance were performed to examine behavioral differences between TBI and HC groups on mood (CMDI), premorbid IQ (WTAR), and emotional processing measures (discrimination and identification subscores on the FEIT). Education level served as a covariate in all analyses given the differences in education level between the groups. Pearson correlations were performed to examine associations between emotional processing performance and depression, demographic variables (age and education), and premorbid intelligence. Correlational analyses were run on the entire sample (collapsed across both groups) and then each group (HC, TBI) individually. An additional correlation was run to examine the relationship between months since TBI and performance on the emotional processing measures.
30
GENOVA ET AL.
Tract-based spatial statistics FSL’s Brain Extraction Tool (BET) was used to remove all nonbrain tissue from each image by creating a brain mask from the b = 0 (nondiffusion weighted) images. Each subject’s data were then corrected for distortions due to eddy currents and head motion using affine transformations. Finally, using DTIFIT (http://fsl. fmrib.ox.ac.uk/fsl/fsl4.0/fdt/fdt_dtifit.html) from FSL, a diffusion tensor model was fitted at each voxel and maps of FA, MD, AD, and RD were obtained. After the preprocessing steps were completed, Tract-Based Spatial Statistics (TBSS) was performed on FA, MD, AD, and RD in FSL 5.0. (Smith et al., 2006) Briefly, TBSS include the following steps: all subjects’ FA maps were aligned to MNI152 space using FNIRT (nonlinear registration). FA skeleton was created from the mean FA skeleton image. Each subject’s FA image was then projected onto this skeleton image, and the resulting images were subjected to group comparison by regressing out age and education. Significance was set at p < 0.05, corrected for multiple comparisons using family-wise error rate (FWER) to compare controls with TBI participants. Correlation between performances on the FEIT score with DTI metrics (in the tracts from TBSS skeleton) was performed to identify WM tracts that show changes with FEIT scores in TBI group. Voxel-based morphometry GM VBM analysis was carried out using openware FSL version 5.0 (http://www.fmrib.ox.ac.uk/fsl/) (Smith et al., 2004) adopting the standard VBM pipeline. An optimized VBM approach, developed by Good and colleagues, was adopted (Ashburner & Friston, 2000; Good et al., 2001). Data processing was divided into four major steps: (1) brain extraction was performed using BET (Smith, 2002) adopting the methods of Popescu et al. (2012). (2) Brain extracted images were segmented into WM, GM, and cerebrospinal fluid volume probability maps using FAST (http://fsl.fmrib. ox.ac.uk/fsl/fslwiki/FAST; Zhang, Brady, & Smith, 2001). (3) A study-specific GM template was created. This was done by registering the 23 HCs and 23 of the 42 TBI subjects (randomly chosen to avoid bias during the registration process) into the MNI152 space using the affine registration tool FLIRT (Jenkinson & Smith, 2001; Zhang et al., 2001) and then by nonlinear registration using FNIRT (www.fmrib.ox.ac.uk/analysis/techrep). The resulting images were averaged to create the template. (4) All the native GM images were nonlinearly re-registered to the template and modulated. (5) These images were then smoothed using a full-width half-
maximum of 7 mm. (6) General linear model was used to compare voxel-wise differences in GM volume between the TBI and HC groups after regressing out age and education. Nonparametric statistics was performed using “randomize” with 5000 permutations and using threshold-free cluster enhancement option. Also, GM volume in TBI patients was correlated with FEIT discrimination and identification scores. A p < 0.05 corrected for multiple comparisons using FWER was used as the level of significance.
RESULTS Behavioral performance The behavioral performance is described in Table 1. After controlling for variance associated with education, the TBI group demonstrated significantly increased levels of depression on the CMDI total score compared to HCs (F(1, 55) = 12.39, p = .001). Similarly, on the WTAR, individuals with TBI demonstrated significantly lower estimated premorbid IQ (F(1, 62) = 5.77, p = .019) after controlling for variance associated with education. The TBI group also performed significantly worse than HCs on both the discrimination trial (F(1, 62) = 4.53, p = .037) and the identification trial (F(1, 62) = 11.22, p = .001) of the FEIT, after controlling for differences between the groups in education. Correlational analyses indicated that in either group individually, or the sample as a whole, no significant correlations were found between demographic variables (age, education) and either emotional processing variables. Further, no correlations were found between depression and either emotional processing measure in the whole sample, or in either group. However, in the TBI group only, there was a significant association between premorbid intelligence and FEIT identification (R = .319, p = .039), indicating that individuals with higher premorbid intelligence performed better on the identification trial of the FEIT. No relationship was observed between time since injury and accuracy on either FEIT identification or discrimination.
Relationship between WM integrity and facial affect recognition A significant relationship was noted between reduced performance on the discrimination trial of the FEIT and reduced WM integrity in two specific WM tracts: right IFOF (x = 33, y = 6, z = 1, Figure 1) and right
FACIAL AFFECT RECOGNTION LINKED
ILF (x = 48, y = 0, z = −17, Figure 2). In the IFOF, there was a positive correlation between FA and FEIT discrimination (p < .05). In the ILF, there was a
31
significant negative correlation between AD, MD, and RD with FEIT discrimination (p < .05). No significant relationships were observed between DTI metrics and the identification trial of the FEIT.
Differences in DTI metrics in ROIs between groups The TBSS analysis demonstrated that WM integrity in the ILF and IFOF was significantly lower in persons with TBI as compared with HCs. In the IFOF, AD, MD, and RD values in the TBI group were significantly higher than the HCs (p < .05), but no significant difference in FA values was observed between groups in that tract. In the ILF, significantly higher MD and RD levels were noted in the TBI group compared to HCs (p < .05), but no significant difference was noted between the groups in FA or AD.
Relationship between GM volume and facial affect recognition Figure 1. A significant relationship was noted between reduced performance on the discrimination trial of the FEIT and reduced WM integrity in right IFOF. A, B, and C refer to coronal, sagittal and axial images, respectively.
Voxel-wise correlations were run to examine the relationship between GM volume and performance on FEIT discrimination and FEIT identification. One cluster encompassing two regions was observed to be positively correlated with FEIT identification: right lingual gyrus and right hippocampus/parahippocampus (p < .05) (see Figure 3). Reduced GM volume in these regions was associated with poorer performance on the FEIT. No significant correlations were noted between FEIT discrimination and GM volume.
DISCUSSION
Figure 2. A significant relationship was noted between reduced performance on the discrimination trial of the FEIT and reduced WM integrity in right ILF. A, B, and C refer to coronal, sagittal and axial images, respectively.
The current study examined the relationship between TBI-related pathology and performance on a facial affect recognition task. Individuals with TBI were significantly more impaired on a task of facial affect recognition, both in terms of identifying emotions and in terms of discriminating between emotions. Reduced WM integrity in the ILF and IFOF was associated with poor performance on the facial affect recognition task. Additionally, atrophy of the hippocampus/parahippocampus and lingual gyrus was associated with impaired facial affect recognition. These findings demonstrate that pathology of both WM and GM contributes to impairment in the
32
GENOVA ET AL.
Figure 3. Reduced GM volume in right lingual gyrus and right hippocampus/parahippocampus was associated with poorer performance on the FEIT. A, B, and C refer to coronal, sagittal and axial images, respectively.
ability to identify emotions of faces in persons with TBI. The finding that reduced WM integrity in the ILF and IFOF was related to worse facial affect recognition abilities in TBI is consistent with what has been reported in other clinical populations. For example, in individuals with Parkinson’s disease, FA levels of the IFOF and ILF have been shown to be associated with impaired ability to identify specific emotions (Baggio et al., 2012). Similarly, WM lesions of the ILF and IFOF were associated with an impaired ability to identify facial expression in individuals with multiple sclerosis (Mike et al., 2013). In autism, a disorder in which facial affect recognition impairments are prevalent, multiple studies have found abnormal diffusion compared to HCs in the ILF and IFOF (Jou et al., 2011; Shukla & Keehn, 2011). Interestingly, these findings suggest a “disconnection,” where compromised connections between several critical regions (such as occipital lobes, limbic structures), rather than damage only to the regions themselves, can result in impairment in emotional processing. Additionally, in the current study, compromised GM (atrophy) in the TBI group was noted to be related to impaired facial affect recognition, specifically in the lingual gyrus and the hippocampal/parahippocampal regions. While the lingual gyrus has been shown to be involved in the perception of faces (Puce, Allison,
Asgari, Gore, & McCarthy, 1996), there is a growing body of evidence for a specific role in the processing of emotional facial expressions (e.g., Fusar-Poli et al., 2009). The hippocampus and parahippocampal regions have also been shown to be involved in identification of emotional facial expressions (Gur et al., 2002; Trautmann, Fehr, & Herrmann, 2009). In multiple studies involving functional magnetic resonance imaging (fMRI) examining facial emotion processing in individuals with major depressive disorder, the hippocampus and parahippocampal gyrus show abnormal brain activation patterns compared to HCs (Fu et al., 2008; Surguladze et al., 2005; Suslow et al., 2010). Thus, using different neuroimaging methodologies, the converging evidence suggests that damage to the hippocampus/parahippocampal region and lingual gyrus via either atrophy or insult results in impaired emotional processing in TBI. Interestingly, in our study, we found a relationship between premorbid intelligence and performance on the emotional processing measure, suggesting that certain preinjury variables may influence one’s emotional processing abilities. To date, there is little to no research focused on the relationship between premorbid intelligence and facial affect recognition, but recent research has shown that there is a significant relationship between facial affect recognition and levels of cognitive functioning following a brain injury. For example, it has been shown that functioning in other cognitive domains such as verbal and nonverbal memory, processing speed, and working memory are associated with facial affect recognition (Yim, Babbage, Zupan, Neumann, & Willer, 2013). Examination of the relationship between emotional processing and both premorbid intelligence and cognitive outcome following a TBI may be an interesting avenue for future research. Our findings represent an important first step in understanding impaired emotional processing in individuals with TBI. Although impairments in emotional processing in TBI have been well documented (specifically in regards facial affect recognition), there is a lack of understanding as to why this deficit exists, or when these deficits occur. For example, it is possible that damage to brain structure leads to impaired facial affect recognition (as our findings in the WM tracts suggest). With the passage of time, impaired emotional processing in individuals with TBI may lead to reduced social functioning and a shrinking social network, reinforcing the neural networks which are impaired, leading to even greater emotional processing impairments (Kennedy & Adolphs, 2012). This cyclical relationship (impaired emotional processing —> decreased social functioning —> greater impaired emotional processing) may continue throughout the life span of the individual with TBI.
FACIAL AFFECT RECOGNTION LINKED
Therefore, an important future avenue of research would be to determine when in the recovery process these impairments arise, so that the deficits could be potentially treated. There were several limitations to the current study. First, the task that we used, while examining multiple emotions, does not lend itself to easy examination of individual emotions. The number of trials for each emotion is not equally distributed, and therefore, it is difficult to examine which emotions were more difficult for the individuals with TBI to identify. Examining the emotions separately will enable researchers to further delineate and identify distinct emotional processing networks of the brain. Additionally, because the DTI data were acquired as part of a larger fMRI study, the number of directions was limited which may result in reduced signal. Finally, the sample of individuals with TBI showed higher levels of depression than HCs. Although this is expected, it is hard to determine whether the impairments observed in the current study were due to the TBI or the depression, as individuals with depression have been shown to have both WM and GM abnormalities and impairments in facial affect recognition abilities (Aldinger et al., 2013; Bessette, Nave, Caprihan, & Stevens, 2013; Grieve, Korgaonkar, Koslow, Gordon, & Williams, 2013) In conclusion, damage to WM connections and GM atrophy in individuals with TBI is linked to impairments in a key social behavior: the ability to recognize facial affect. The association between this damage and impairments in distinguishing emotions occurred in specific WM tracts and GM regions involved in pathways connecting the occipital lobes to the orbital frontal cortex and involves key limbic structures. Functional neuroimaging may provide additional avenues of research in which aspects of functional connectivity can be examined in an effort to better understand the relationship between emotional processing abilities and brain networks involved in social functioning. Original manuscript received 7 April 2014 Revised manuscript accepted 25 August 2014 First published online 16 September 2014
REFERENCES Adolphs, R. (2010). What does the amygdala contribute to social cognition?. Annals of the New York Academy of Sciences, 1191, 42–61. doi:10.1111/j.1749-6632.2010. 05445.x Aldinger, M., Stopsack, M., Barnow, S., Rambau, S., Spitzer, C., Schnell, K., & Ulrich, I. (2013). The
33
association between depressive symptoms and emotion recognition is moderated by emotion regulation. Psychiatry Research, 205(1–2), 59–66. doi:10.1016/j. psychres.2012.08.032 Andriessen, T. M., Jacobs, B., & Vos, P. E. (2010). Clinical characteristics and pathophysiological mechanisms of focal and diffuse traumatic brain injury. Journal of Cellular and Molecular Medicine, 14(10), 2381–2392. doi:10.1111/j.1582-4934.2010.01164.x Ashburner, J., & Friston, K. J. (2000). Voxel-based morphometry: The methods. Neuroimage, 11(6), 805–821. doi:10.1006/nimg.2000.0582 Babbage, D. R., Yim, J., Zupan, B., Neumann, D., Tomita, M. R., & Willer, B. (2011). Meta-analysis of facial affect recognition difficulties after traumatic brain injury. Neuropsychology, 25(3), 277–285. doi:10.1037/a0021908 Baggio, H. C., Segura, B., Ibarretxe-Bilbao, N., Valldeoriola, F., Marti, M. J., & Compta, Y. (2012). Structural correlates of facial emotion recognition deficits in Parkinson’s disease patients. Neuropsychologia, 50(8), 2121–2128. doi:10.1016/j.neuropsychologia.2012.05.020 Bessette, K. L., Nave, A. M., Caprihan, A., & Stevens, M. C. (2013). White matter abnormalities in adolescents with major depressive disorder. Brain Imaging and Behavior, 1–11. Advance online publication. doi:10.1007/s11682013-9274-8 Catani, M., Jones, D. K., Donato, R., & Ffytche, D. H. (2003). Occipito-temporal connections in the human brain. Brain, 126(9), 2093–2107. doi:10.1093/brain/awg203 Fu, C. H., Williams, S. C., Cleare, A. J., Scott, J., Mitterschiffthaler, M. T., Walsh, N. D., & Murray, R. M. (2008). Neural responses to sad facial expressions in major depression following cognitive behavioral therapy. Biological Psychiatry, 64(6), 505–512. doi:10.1016/j.biopsych.2008.04.033 Fusar-Poli, P., Placentino, A., Carletti, F., Allen, P., Landi, P., Abbamonte, M., & Politi, P. L. (2009). Laterality effect on emotional faces processing: ALE meta-analysis of evidence. Neuroscience Letters, 452(3), 262–267. doi:10.1016/j.neulet.2009.01.065 Good, C. D., Johnsrude, I. S., Ashburner, J., Henson, R. N., Friston, K. J., & Frackowiak, R. S. (2001). A voxelbased morphometric study of ageing in 465 normal adult human brains. Neuroimage, 14(1), 21–36. doi:10.1006/nimg.2001.0786 Green, R. E. A., Melo, B., Christensen, B., Ngo, L., Monette, G., & Bradbury, C. (2008). Measuring premorbid IQ in traumatic brain injury: An examination of the validity of the Wechsler Test of Adult Reading (WTAR). Journal of Clinical Experimental Neuropsychology, 30(2), 163–172. doi:10.1080/ 13803390701300524 Grieve, S. M., Korgaonkar, M. S., Koslow, S. H., Gordon, E., & Williams, L. M. (2013). Widespread reductions in gray matter volume in depression. NeuroImage: Clinical, 3, 332–339. doi:10.1016/j.nicl.2013.08.016 Gur, R. C., Schroeder, L., Turner, T., McGrath, C., Chan, R. M., Turetsky, B. I., & Gur, R. E. (2002). Brain activation during facial emotion processing. Neuroimage, 16(3), 651– 662. doi:10.1006/nimg.2002.1097 Jenkinson, M., & Smith, S. (2001). A global optimisation method for robust affine registration of brain images. Medical Image Analysis, 5(2), 143–156. doi:10.1016/ S1361-8415(01)00036-6
34
GENOVA ET AL.
Johnson, V. E., Stewart, W., & Smith, D. H. (2013). Axonal pathology in traumatic brain injury. Experimental Neurology, 246, 35–43. doi:10.1016/j.expneurol. 2012.01.013 Jou, R. J., Jackowski, A. P., Papademetris, X., Rajeevan, N., Staib, L. H., & Volkmar, F. R. (2011). Diffusion tensor imaging in autism spectrum disorders: Preliminary evidence of abnormal neural connectivity. Australian and New Zealand Journal of Psychiatry, 45(2), 153–162. doi:10.3109/00048674.2010.534069 Kennedy, D. P., & Adolphs, R. (2012). The social brain in psychiatric and neurological disorders. Trends in Cognitive Sciences, 16(11), 559–572. doi:10.1016/j. tics.2012.09.006 Kerr, S. L., & Neale, J. M. (1993). Emotion perception in schizophrenia: Specific deficit or further evidence of generalized poor performance?. Journal of Abnormal Psychology, 102(2), 312–318. doi:10.1037/0021843X.102.2.312 MAGNIMS Study Group, Popescu, V., Battaglini, M., Hoogstrate, W. S., Verfaillie, S. C., Sluimer, I. C., & Van Schijndel, R. A. (2012). Optimizing parameter choice for FSL-Brain Extraction Tool (BET) on 3D T1 images in multiple sclerosis. Neuroimage, 61(4), 1484– 1494. Martino, J., Brogna, C., Robles, S. G., Vergani, F., & Duffau, H. (2010). Anatomic dissection of the inferior fronto-occipital fasciculus revisited in the lights of brain stimulation data. Cortex, 46(5), 691–699. doi:10.1016/j. cortex.2009.07.015 Mike, A., Strammer, E., Aradi, M., Orsi, G., Perlaki, G., Hajnal, A., & Illes, Z. (2013). Disconnection mechanism and regional cortical atrophy contribute to impaired processing of facial expressions and theory of mind in multiple sclerosis: A structural MRI study. PLoS One, 8(12), e82422. doi:10.1371/journal.pone.0082422 Miyahara, M., Harada, T., Ruffman, T., Sadato, N., & Iidaka, T. (2013). Functional connectivity between amygdala and facial regions involved in recognition of facial threat. Social Cognitive & Affective Neuroscience, 8(2), 181–189. doi:10.1093/scan/nsr085 Nyenhuis, D. L., Luchetta, T., Yamamoto, C., Terrien, A., Bernardin, L., Rao, S. M., & Garron, D. C. (1998). The development, standardization, and initial validation of the Chicago Multiscale Depression Inventory. Journal of Personality Assessment, 70(2), 386–401. doi:10.1207/ s15327752jpa7002_14 Philippi, C. L., Mehta, S., Grabowski, T., Adolphs, R., & Rudrauf, D. (2009). Damage to association fiber tracts impairs recognition of the facial expression of emotion. The Journal of Neuroscience, 29(48), 15089–15099. doi:10.1523/JNEUROSCI.0796-09.2009 Polyack, S. (1957). The vertebrate visual system. University of Chicago Press.
Puce, A., Allison, T., Asgari, M., Gore, J. C., & McCarthy, G. (1996). Differential sensitivity of human visual cortex to faces, letterstrings, and textures: A functional magnetic resonance imaging study. The Journal of Neuroscience, 16(16), 5205–5215. Shukla, D. K., Keehn, B., & Müller, R.-A. (2011). Tractspecific analyses of diffusion tensor imaging show widespread white matter compromise in autism spectrum disorder. Journal of Child Psychology and Psychiatry, 52(3), 286–295. doi:10.1111/j.1469-7610.2010.02342.x Smith, S. M. (2002). Fast robust automated brain extraction. Human Brain Mapping, 17(3), 143–155. doi:10.1002/ hbm.10062 Smith, S. M., Jenkinson, M., Johansen-Berg, H., Rueckert, D., Nichols, T. E., Mackay, C. E., & Behrens, T. E. (2006). Tract-based spatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage, 31(4), 1487–1505. doi:10.1016/j.neuroimage.2006.02.024 Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E., Johansen-Berg, H., & Matthews, P. M. (2004). Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage, 23(Suppl 1), SS208–219. doi:10.1016/j.neuroimage. 2004.07.051 Surguladze, S., Brammer, M. J., Keedwell, P., Giampietro, V., Young, A. W., Travis, M. J., & Williams, S. C. (2005). A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder. Biological Psychiatry, 57(3), 201–209. doi:10.1016/j.biopsych.2004.10.028 Suslow, T., Konrad, C., Kugel, H., Rumstadt, D., Zwitserlood, P., Schöning, S., . . . Dannlowski, U. (2010). Automatic mood-congruent amygdala responses to masked facial expressions in major depression. Biological Psychiatry, 67(2), 155–160. doi:10.1016/j. biopsych.2009.07.023 Trautmann, S. A., Fehr, T., & Herrmann, M. (2009). Emotions in motion: Dynamic compared to static facial expressions of disgust and happiness reveal more widespread emotion-specific activations. Brain Research, 1284, 100–115. doi:10.1016/j.brainres. 2009.05.075 Wechsler, D. (2001). The Wechsler Test of Adult Reading. San Antonio, TX: The Psychological Corporation. Yim, J., Babbage, D. R., Zupan, B., Neumann, D., & Willer, B. (2013). The relationship between facial affect recognition and cognitive functioning after traumatic brain injury. Brain Injury, 27(10), 1155–1161. doi:10.3109/ 02699052.2013.804203 Zhang, Y., Brady, M., & Smith, S. (2001). Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Transactions on Medical Imaging, 20(1), 45–57. doi:10.1109/42.906424
Copyright of Social Neuroscience is the property of Psychology Press (UK) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
