Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 May 1.
Published in final edited form as: Neuropsychology. 2018 May;32(4):476–483. doi: 10.1037/neu0000368

Facial-affect recognition deficit as a predictor of different aspects of social communication impairment in traumatic brain injury

Arianna Rigon a,*, Lyn S Turkstra b, Bilge Mutlu c, Melissa C Duff a,d,e
PMCID: PMC5975647  NIHMSID: NIHMS965737  PMID: 29809034

Abstract

Objective

To examine the relationship between facial affect recognition and different aspects of self- and proxy-reported social communication impairment following moderate-severe traumatic brain injury (TBI).

Methods

Forty-six adults with chronic TBI (>6 months post injury) and 42 healthy comparison (HC) adults were administered the LaTrobe Communication Questionnaire (LCQ) Self and Other forms to assess different aspects of communication competence, and the Emotion Recognition Test (ERT) to measure their ability to recognize facial affects.

Results

Individuals with TBI underperformed HC adults in the ERT, and self-reported, as well as were reported by close others, as having more communication problems than HC adults. TBI group ERT scores were significantly and negatively correlated with LCQ-Other (but not LCQ-Self) scores (i.e., participants with lower emotion recognition scores were rated by close others as having more communication problems). Multivariate regression analysis revealed that adults with higher ERT scores self-reported more problems with disinhibition/impulsivity and partner sensitivity, and had fewer other-reported problems with disinhibition/impulsivity and conversational effectiveness.

Conclusions

Our findings support growing evidence that emotion recognition deficits play a role in specific aspects of social communication outcomes after TBI and should be considered in treatment planning.

INTRODUCTION

Many adults with moderate-severe traumatic brain injury (TBI) have difficulties in social functioning, including reduced social participation, increased socially inappropriate behaviors, problems producing and understanding verbal and non-verbal communication, and deficits in empathy (de Sousa et al., 2011; Dumont, Gervais, Fougeyrollas, & Bertrand, 2004; Milders, Fuchs, & Crawford, 2003; Rousseaux, Verigneaux, & Kozlowski, 2010). Social communication problems in particular (e.g., making inappropriate or irrelevant comments; monopolizing conversations; or being insensitive to the needs of others) have been extensively reported and have been associated with poor life outcomes such as maintaining employment, opportunities to experience positive social interactions, and reduced participation in social activities (Finset, Dyrnes, Krogstad, & Berstad, 1995; Struchen, Clark, et al., 2008; Wehman et al., 1993; Ylvisaker, 1989).

While social dysfunctions in adults with TBI have been well documented, their underlying causes remain poorly understood. One potential cause, particularly for problems in social communication, is impairment in recognizing facial affect. Facial-affect recognition impairments have been widely reported in both the acute and chronic stages after moderate-severe TBI (Green, Turner, & Thompson, 2004; Jackson & Moffat, 1987; Prigatano & Pribram, 1982; Rigon, Turkstra, Mutlu, & Duff, 2016; Rosenberg, McDonald, Dethier, Kessels, & Westbrook, 2014; Spikman, Milders, et al., 2013). Knox and Douglas (2009) observed a significant correlation between the ability to recognize facial affect and self-reported social integration in adults with severe TBI (Knox & Douglas, 2009); similarly, Petterson (1991) noted that emotion-recognition skills in children and adolescents with TBI were a predictor of frequency of parent-reported socially inappropriate behaviors (Pettersen, 1991). Spikman et al. (2013) found that the ability to recognize emotions was negatively correlated with both behavioral problems of adults with TBI (as indicated by proxy reports) and their degree of self-awareness (Spikman, Milders, et al., 2013). Taken together, these findings suggest a relationship between emotion recognition abilities and social functioning in TBI populations.

Facial-affect recognition impairments have also been linked specifically to social communication problems. Ryan and colleagues (2013) found that, among survivors of pediatric TBI, social-communication difficulty mediated the relationship between the ability to recognize emotions and occurrence of externalizing behaviors (Ryan et al., 2013). Watts and Douglas (2006) found a negative correlation between facial-affect recognition and proxy-reported communication problems (Watts & Douglas, 2006). The relationship between self-rated communication problems and emotion recognition was only marginal, and the authors suggested that a link between emotion recognition skills and communication competence might be visible with a sample size larger than n=12, although it is possible that the lack of a significant correlation is due to limited self-awareness in adults with TBI, leading to a failure to correctly appraise one’s communication difficulties.

The aim of the current study was to further examine the relationship between everyday social communication problems and affect recognition in a larger sample, and to investigate whether different types of communication problems are more likely to be associated with facial-affect recognition abilities. A better knowledge of the relationship between communication competence and facial-affect recognition can inform the development and administration of interventions for social communication problems after TBI, as well as interventions for a broader set of cognitive challenges, as impairments in social cognition can affect rehabilitation for other cognitive functions (Spikman, Boelen, et al., 2013). Indeed, there is mixed evidence supporting that treatments currently employed for social communication problems improve everyday social interactions (Dahlberg et al., 2007; McDonald et al., 2008; Ylvisaker, Turkstra, & Coelho, 2005). Evidence of a relationship between social cue perception deficit and specific types of communication impairment can guide the creation of new and more effective protocols that consider the underlying mechanism leading to problems in everyday life. In particular, it is important to examine the relationship between emotion recognition abilities and different kinds of social communication competence both as it is self-reported by adults with TBI and as it is perceived by their frequent communication partners: although not universally present, self-awareness problems have been reported in adults with chronic moderate-severe TBI (Prigatano & Altman, 1990; Vanderploeg, Belanger, Duchnick, & Curtiss, 2007), suggesting that in-depth characterization of post-injury deficits might require cooperation and reports from patients’ close others.

Here, we build on previous work investigating social functioning in TBI and examine the relationship between emotion recognition accuracy and frequency of self- and proxy-reported social communication problems in adults with TBI. We measured facial-affect recognition using the Emotion Recognition Test (ERT)(Kessels, Montagne, Hendriks, Perrett, & de Haan, 2014; Montagne, Kessels, De Haan, & Perrett, 2007) and social communication problems using the self-report and proxy-report forms of the LaTrobe Communication Questionnaire (LCQ) (Douglas, 2010b; Douglas, Bracy, & Snow, 2007; Douglas, O’Flaherty, & Snow, 2000). Based on previous findings (Watts & Douglas, 2006), we expected that participants with TBI would have both worse emotion recognition performance and also more self-reported and proxy-reported social communication problems than a demographically matched comparison group. In addition, we predicted a negative correlation between other-reported communication problems and affect recognition accuracy in adults with TBI. Although, as noted above, a previous study only found a marginal correlation between self-reported communication problems and emotion recognition, findings from other work finding associations between self-reported social integration and affect recognition skills (Knox & Douglas, 2009), led us to also expect a negative correlation between facial-affect recognition abilities and self-reported communication problems.

In order to examine the relationship between emotion recognition and different aspects of communication competence we used an exploratory multivariate regression approach, and we hypothesized that emotion recognition abilities would serve as the predictor (independent variable) of different aspects of social communication abilities (dependent variables), as measured by the four LCQ factors: initiation/conversational flow, disinhibition/impulsivity, conversational effectiveness, partner sensitivity, and reported emotion recognition ability (Struchen, Pappadis, et al., 2008).

METHODS

Participants

Forty-six adult individuals with a history of moderate-severe TBI (females=23) and forty-two healthy comparisons (females=21) participated in the study. Participants were recruited from two different sites. TBI severity was assessed according to the Mayo Classification System (Malec et al., 2007). Participants were considered moderate-severe if only one of the four following criteria was met: (1) Glasgow Coma Scale (GCS)<13, (2) positive acute CT findings or lesions visible on a chronic MRI, (3) loss of consciousness (LOC)>30 minutes or post-traumatic amnesia (PTA)>24 hours, and (4) retrograde amnesia>24 hours. All participants were at least 6 months post injury (range of time post injury in months was 6-506). Cause of injury included motor-vehicle accidents (18 individuals), falls (15), non-motored vehicle accidents (4), assaults (4), sport-related injuries (2), and other (3). Severity-related information was collected using available medical records and self-report.

The HC group was composed of individuals with no history of TBI, loss of consciousness, or neurological or psychiatric disorders affecting language or thinking. TBI and HC groups were matched for age (t(86)=.4, p=.69), education (t(81.25)=.96, p=.34), site of testing (χ2 (1)=.18, p=.83), and sex (χ2 (1)<.001, p=1) (See 1). Participants were excluded if they scored in the Aphasia range on the Western Aphasia Battery (WAB) screening test (Shewan & Kertesz, 1980). Approval by the appropriate Institutional Review Boards was obtained.

Communication Problems

Frequency of communication problems was assessed using the LCQ (Douglas et al., 2007). The LCQ was designed to measure communication problems in adults with TBI based on the literature on communication problems associated with TBI and Grice’s model of effective communication, and has both high validity (>.9) and reliability (>.8) (for further information, see Douglas et al, 2007). There are two versions of the LCQ: a self-report form completed by the participant (LCQ-SELF), and an other-report form completed by a close other nominated by the participant as a frequent communication partner (LCQ-OTHER)(Douglas et al., 2007). The LCQ-SELF and LCQ-OTHER are composed of the same 34 items, with items phrased in the second person for the LCQ-SELF (e.g., “When talking to others, do you leave out important details?”) and in the third person for the LCQ-OTHER (e.g., “When talking to others, does [name] leave out important details?”). Respondents answer each item on a 1–4 Likert scale, where 1 represents “Never or Rarely” and 4 represents “Usually or always.” Higher overall scores correspond to a higher frequency of communication problems. After participants completed the LCQ-SELF, they were asked to nominate a frequent communication partner to complete the LCQ-OTHER form, and study personnel mailed the form to that person.

Items from the LCQ have been found to load on four different factors: initiation/conversational flow, disinhibition/impulsivity, conversational effectiveness and partner sensitivity (Struchen, Pappadis, et al., 2008). Recently, the author of the survey added four new items aimed at investigating participants’ self- and other-reported ability to recognize emotions and moods (e.g., “When talking to others, do you/does [name] misinterpret the other person’s facial expression?”) (In a conversation with Professor J. Douglas, (March 3, 2016)). As these items were not included in the factor analysis performed by Struchen et al. (2008), in the current study they were grouped together and separately from the other subscales, and treated as items loading onto a separate factor (emotion recognition). Dependent variables for the LCQ were overall LCQ-OTHER and LCQ-SELF scores and average scores for each of the five factors.

To assess interpersonal closeness between the participant and the frequent communication partner, we administered the Inclusion of Other in the Self Scale (IOS) (Aron, Aron, & Smollan, 1992). The IOS allowed us to ensure that group differences in closeness within partner pairs did not influence group differences on the LCQ-OTHER, and that higher (or lower) instances of other-perceived communication problems in the TBI or HC groups were not due a systematic bias driven by within-pair closeness. On a visual rating scale that illustrates different levels of closeness between two entities, participants and communication partners were asked to indicate their interpersonal closeness on a scale from 1 (very low closeness) to 7 (very high closeness), yielding two separate IOS-SELF and IOS-OTHER scores. The IOS was not completed in one HC-communication partner pair, and it was not returned by two communication partners for the TBI group.

Emotion Recognition Abilities

Facial-affect recognition was measured using the Emotion Recognition Test (ERT) (Kessels et al., 2014). The ERT is a dynamic computer-based facial-affect recognition task, which has been used as an effective tool to examine subtle emotion recognition impairment in several neuropsychiatric patient groups, including populations with TBI (Law Smith, Montagne, Perrett, Gill, & Gallagher, 2010; Rosenberg, Dethier, Kessels, Westbrook, & McDonald, 2015). Participants view clips of faces that first appear neutral and then gradually morph to express one of six basic emotions (afraid, angry, disgusted, happy, neutral, sad, and surprised). For this study, we adopted the short version of the test, which consists of four morphs for each of the six basic emotions, transforming from neutral to four levels of emotion intensities (0 to 40%, 60%, 80%, and 100%) and makes up a total of 96 trials (4 morphs × 6 emotions × 4 intensities). The presentation order of morphs was fixed from the lowest to the highest emotion intensity, so that participants could not guess at lower-intensity morphs from having seen the stimulus as part of a higher-intensity image of the same face. For each item, participants were instructed to choose “the word that best describes what the person is feeling” from among six alternatives listed to the right of the stimulus. The dependent variable was the total number of correct responses across all intensities and emotions.

Statistical Analysis

Within group relationship between LCQ-SELF and LCQ-OTHER was calculated using bivariate correlation. An analysis of covariance (ANCOVA) was used to test the hypothesis that individuals with TBI would have lower accuracy than HC adults on the ERT, co-varying for sex, age, education and site. For the LCQ, we performed an ANCOVA, entering respondent (LCQ-SELF vs. LCQ-OTHER) and factor (initiation/conversational flow, disinhibition/impulsivity, conversational effectiveness, partner sensitivity, and reported emotion recognition) as within-subjects variables, group (TBI vs. HC) as a between-subjects variable, and sex, age, education, and site of testing as covariates. Similarly, for IOS data, we carried out an ANCOVA with respondent as a within-subjects factor and group as between-subjects factor, and co-varying for sex, age, education and site. We computed partial correlations between ERT and overall LCQ measures, correcting for sex, age, education, and site, in order to test our hypothesis that emotion recognition ability would negatively correlate with both other-perceived and self-perceived communication competence.

Finally, we conducted a multivariate linear regression analysis to explore whether different aspects of communication problems (i.e., separate LCQ-SELF and LCQ-OTHER factors) were linked to emotion recognition abilities, either separately or as a group. In each model, we set ERT performance as the independent variable and LCQ factors as dependent variables. Separate multivariate analyses were carried out for the TBI and HC groups.

RESULTS

Communication problems, interpersonal closeness, & emotion recognition: between-group comparison

For the ERT, a mixed-effect ANCOVA revealed significant effects of group (F1,87=10.86, p=.001, ηp2=.12; HC > TBI), sex (F1,87=8.55, p=.004, ηp2=.1; females > males) and age (F1,87=4.61, p=.04, ηp2=.05; younger > older), with no other significant effects (all F’s1,87<2.53, p’s>.11, ηp2′s<.30).

For LCQ data, we found a marginally significant correlation between LCQ-SELF and LCQ-OTHER within the TBI group (r=.25, p=.045) and a significant correlation within the HC group (r=.32 p=.02) group. An ANCOVA revealed that there was significant effect of group (F1,82=17.93, p<.001, ηp2=.18) but no significant effect of respondent (F1,82=2.96, p=.09, ηp2=.04) or factor (F2.87,235.09=1.95, p=.12, ηp2=.02) as well as no group-by-respondent interaction, no respondent-by-factor interaction, no group-by-factor interaction, and no group-by-factor-by-respondent interaction (all F<1.66, p=<.16, ηp2<.02) (See Table 2). There was no significant main effect of any of the covariates (sex, site, age, or education) (all F1,82<1.57, p>.21, ηp2<.02) and no significant interaction between respondent, factor and any of the covariates (all F<2.4, p>.06, ηp2<.03). Given the non-significant effect of factor, further examination of the main effect of group was carried out on the overall LCQ score. When all items were included, participants with TBI rated themselves (t(79.28)=5.28, p<.001, d=1.12) and were rated by their close others (t(86)=2.62, p=.01, d=.56) as having more communication problems than HCs. Taken together, these results indicate that individuals with TBI both perceived themselves and are perceived by others as having significantly more communication problems than healthy individuals. This impairment is present across the board, i.e., it is not restricted to specific areas of communication. Lastly, communication problems could not be explained by demographic factors such as age, sex, or education.

Table 2.

Mean and standard deviation of TBI and HC groups for LCQ, IOS and ERT data

HC
(Mean±SD) 		TBI
(Mean±SD) 		GROUP DIFFERENCE
(P-value, Cohen’s d)
LCQ-SELF 53.48±8.49 65.5±12.62 p<.001, d=1.12
LCQ-OTHER 52.57±16.23 61.89±17.01 p=.01, d=.56
IOS-SELF 5.56±1.25 5.15±1.49 p=.17, d=.3
IOS-OTHER 4.84±1.46 4.73±1.66 p=.73, d=.07
ERT 58.26±8.41 52.15±8.52 p=.001, d=.73
Open in a new tab

Note: HC=Healthy comparison participants, TBI=Traumatic brain injury, ERT=Emotion recognition task, LCQ=Latrobe Communication Questionnaire, IOS=Inclusion of the other in the self scale, SD=Standard deviation

Analysis of the IOS responses to examine interpersonal closeness between participants and respondents revealed a significant effect of respondent (F1,78=9.28, p=.02, ηp2=.06), but no main effect of group (F1,78=.092, p=.76, ηp2=.03), no group-by-respondent interaction (F1,78=2.1, p=.15, ηp2=.03), and no significant interaction between group or respondent and any of the covariates (all F’s1,78<2.36, p’s>.12, ηp’s2<.03). Further inspection of the respondent effect revealed that across groups closeness of communication partner was rated higher by the participants than by close others. Overall, this suggests that close others for the TBI and comparison groups did not differ in how close they rated themselves to participants and that group differences in other-perceived communication problems are unlikely to be due to differences in participant-respondent pairs’ interpersonal closeness.

Relationship between emotion recognition abilities and communication problems

Another goal of our analyses was to examine the association between frequency of communication problems and emotion recognition skills. Partial correlation analysis revealed that within the HC group there was no significant correlation between ERT score and total LCQ-SELF (r=−.07, p=.34) or LCQ–OTHER (r=−.2, p=.12). Within the TBI group, ERT scores were significantly and negatively correlated with total LCQ-OTHER scores (r=−.33, p=.02), indicating that participants with TBI who were rated as having more communication problems by their close others also had lower emotion recognition abilities. However, total LCQ-SELF and ERT scores did not significantly correlate (r=.22, p=.08).

Because ERT performance correlated significantly with LCQ-OTHER but not with LCQ-SELF, we used a Fisher r-to-z transform to test whether the difference between the two correlation coefficients was significantly different from 0. Our results showed a stronger relationship between emotion recognition abilities and other-perceived communication problems than between emotion recognition abilities and self-perceived communication problems (Z=2.44 p=.02) within the TBI group.

Our main aim was to investigate the role of emotion recognition performance as a predictor of different types of communication impairment (initiation/conversational flow, disinhibition/impulsivity, conversational effectiveness, partner sensitivity, and reported emotion recognition ability). Within the TBI group, multivariate analysis with ERT performance as predictor and the five LCQ-SELF factors as outcomes revealed that emotion recognition abilities were a significant predictor of composite self-reported social communication problems (Pillai’s Trace=.26, F5,40=2.87, p=.03) (See table 3). ERT score was a significant predictor of self-reported disinhibition/impulsivity (β=.02, CI 95% [.003, .04]) and partner sensitivity (β=.02, CI 95% [.0001, .04]). However, in the second multivariate model, when the five LCQ-OTHER factors were entered as outcomes, ERT performance was not a significant predictor of overall communication problems (Pillai’s Trace=.2, F5,40=1.97, p=.11). An inspection of the relationship between separate communication factors and ERT performance revealed that the latter was a significant predictor only of other-reported disinhibition/impulsivity (β=−.02, CI 95% [−.041, −.005]) and conversational effectiveness (β=−.03, CI 95% [−.051, −.008]). Interestingly, in our data, an increase in emotion recognition skills corresponded to both more self-reported problems in disinhibition/impulsivity and partner sensitivity and less other-reported problems in disinhibition/impulsivity and conversational effectiveness.

Table 3.

Multivariate linear regression analysis examining LCQ-SELF and LCQ-OTHER factors associated with ERT performance – participants with TBI

Value F B CI 95% P Value Adjusted R2
LCQ-SELF
 WHOLE MODEL
  Pillai’s Trace .27 2.87 .03
  Wilk’s Lambda .74 2.87 .03
 VARIABLES
  I/CF SELF .02 −.001, .03 .06 .06
  D/I SELF .02 .003, .04 .02 .1
  CE SELF .0003 −.02, .02 .98 −.02
  PS SELF .02 .0001, .04 .048 .07
  ER SELF −.02 −.04, .002 .08 .05
LCQ-OTHER
 WHOLE MODEL
  Pillai’s Trace .2 1.97 .11
  Wilk’s Lambda .8 1.97 .11
 VARIABLES
  I/CF OTHER −.02 −.03, .003 .11 .04
  D/I OTHER −.02 −.04, −.005 .01 .11
  CE OTHER −.03 −.05, −.008 .01 .13
  PS OTHER −.02 −.04, .008 .2 .02
  ER OTHER −.02 −.04, .001 .07 .05
Open in a new tab

Note: TBI=Traumatic brain injury, ERT=Emotion recognition task, LCQ=Latrobe Communication Questionnaire, CI=Confidence Interval, I/CF=Initiation/conversational flow, D/I=Disinhibition/impulsivity, CE=Conversational effectiveness, PS=Partner sensitivity, ER=Emotion recognition.

Within the HC group, ERT performance was not a significant predictor of overall self-reported (Pillai’s Trace=.04, F5, 38=.34, p=.89) or other-reported (Pillai’s Trace=.08, F5, 38=.62, p=.69) social communication problems; none of the individual communication outcomes were significantly predicted by emotion recognition abilities (Table S1).

DISCUSSION

In the current study, we examined the relationship between emotion recognition accuracy and self- and other-reported communication problems in adults with moderate to severe TBI. Individuals with TBI performed more poorly than demographically-matched healthy comparisons on a dynamic emotion recognition task. In addition, individuals with TBI rated themselves and were rated by their frequent communication partners as having more communication problems than individuals without TBI. We also observed that other-reported frequency of communication problems was significantly and negatively associated with the ability to recognize emotions from faces. Further examination using a multivariate model revealed that emotion recognition performance was a positive predictor of self-reported communication problems in the domains of partner sensitivity and impulsivity and a negative predictor of other-reported communication problems related to impulsivity and conversational effectiveness. We examine each of these findings in the following paragraphs.

Individuals with TBI have often been reported as having communication problems and facial-affect recognition impairments (Babbage et al., 2011; Douglas et al., 2007). Indeed, previous work using the LCQ and the ERT has observed that, at the group level, adults with TBI tend to show more problems than healthy comparison peers in both (Douglas, 2010a; Rigon et al., 2016; Rosenberg et al., 2014). Our current findings are consistent with prior work, confirming that TBI can have a profound effect on different aspects of social functioning even in the chronic stage. The main aim of the current study, however, was to investigate the specific association between two different aspects of social cognition: communication competence and the ability to recognize emotions. In particular, we hypothesized that an inability to correctly interpret social cues, such as a failure to successfully recognize others’ emotions, might contribute to a reduction in communication competence by depriving individuals with TBI of critical information that healthy individuals use to navigate social situations.

To assess the relationship between the ability to recognize facial affect and communication problems, we first computed the bivariate correlation between overall LCQ scores and ERT performance. Our data showed that, after correcting for demographic variables such as sex, age and education, individuals with TBI with more other-reported overall communication problems less accurately labeled facial affect. Previous studies have reported an association between social communication and emotion recognition (Knox & Douglas, 2009; Pettersen, 1991). Watts and Douglas (2006), who used the LCQ to measure social competence and the Emotion Evaluation Test (EET) section of the Awareness of Social Inference Test (McDonald, Flanagan, Rollins, & Kinch, 2003) to measure emotion recognition, found in a small sample of individuals with severe TBI (N=12) that LCQ-OTHER scores negatively correlated with EET performance. Our replication of these findings in a larger sample of individuals with moderate-severe TBI offers further support for the hypothesis that impairment in these two domains of social functioning may be closely related. Interestingly, the correlation between LCQ and emotion recognition abilities was significant within the TBI group but not within the HC group. This finding may be explained by a decline in emotion recognition skills influencing communication competence only when the ability to recognize emotions drops below a certain threshold. This explanation is supported by the fact that, as a group, individuals with TBI scored worse than HCs in both the ERT and the LCQ. These differences did not result from ceiling effects, as even within the HC group no ceiling effect was noted for the ERT, and HCs did report communication problems on the LCQ.

Our analysis also aimed to expand on previous analyses by examining how specific aspects of communication competence are influenced by the ability to recognize facial affect. We used a multivariate regression approach, which allowed us to test out the hypothesis that emotion recognition impairment is not only related, but it is a significant predictor of frequency of communication problems, and to examine how specific aspects of self- and other-perceived communication competence (i.e., initiation/conversational flow, disinhibition/impulsivity, conversational effectiveness, partner sensitivity, and reported emotion recognition ability) are affected by problems in emotion recognition. It should be noted that not all five subscales of the LCQ contribute equally to the overall score. For instance, ten items were found to load on initiation/conversational flow, while only four items loaded on partner sensitivity. For this reason, examining whether ERT performance can serve as a predictor for the average score on a specific subscale can yield different findings, as well as be more informative, than a simple bivariate correlation. Indeed, our multivariate analyses revealed that, within the TBI group, ERT score was a significant predictor of all LCQ-SELF subscales taken together (i.e., the multivariate model was significant), but not of all LCQ-OTHER subscales. Further exploration of the subscales of both measures revealed that ERT performance was a significant predictor of other-reported disinhibition/impulsivity and conversational effectiveness. This finding indicates that participants who were less accurate at labeling facial affect were more likely to be perceived by others as being more impulsive and disinhibited and less able overall to carry out effective social conversations.

Surprisingly, ERT scores were significant predictors of self-reported disinhibition/impulsivity and partner sensitivity but in a direction opposite to the one expected. Participants with TBI who were better at labeling facial affect also considered themselves worse at reading their communication partners’ needs and more impulsive and disinhibited, which partially contradicts the results presented by Watts and Douglas (2006). One potential explanation for this finding is that individuals with TBI who have better emotion recognition skills also have higher self-awareness of their deficits, which may render them more vigilant to their own social performance and lead them to overestimate—instead of underestimate—communication problems in specific domains. Partial support for this explanation comes from a study by Spikman and colleagues (2013), who reported that individuals with TBI who are better at recognizing emotions are also more self-aware of their deficits (Spikman, Milders, et al., 2013). In this sense, the discrepancy with previous studies may be due to differences in sample size and composition; we tested participants with moderate injuries as well as severe TBI and thus may have included individuals with heightened awareness. Also, half of our sample was female, compared to one out of 12 for the sample reported by Watts and Douglas (2006). The recent finding that males, but not females with TBI, underreport their communication deficits on the LCQ (Despins, Turkstra, Struchen, & Clark, 2015) suggests that data from female participants might account for our different findings.

Although we did not measure self-awareness directly, future work should examine its role as a mediator in the relationship between communication and emotion recognition deficits. In the current study, a correlational analysis between LCQ-SELF and LCQ-OTHER scores within both TBI group revealed a significant (albeit marginal) correlation between the two measures, indicating that, at least to a certain degree, individuals with TBI are aware of their own communication problems. This finding is consistent with previously reported results (Douglas et al., 2007), but it should be noted that as the correlation is relatively low, it likely reflects the fact that our sample includes both individuals with and without impaired self-awareness, thus mirroring the moderate-severe TBI population (Vanderploeg et al., 2007). This in turn may explain the different patterns of associations found between ERT performance and LCQ-SELF and LCQ-OTHER, and suggests a need to obtain both self- and proxy reports when assessing communication impairment in individuals with TBI. We also found no significant difference between LCQ-SELF and LCQ-OTHER ratings within the TBI group (See Results - Communication problems, interpersonal closeness, & emotion recognition: between-group comparison), suggesting that, on average and across all LCQ factors, participants did not rate their communication competence as significantly different than their close others did. Although there have been studies which have found that individuals with TBI underestimate their communication problems (Douglas et al., 2007), several other studies have found no differences between self- and other-reports (Snow, Douglas, & Ponsford, 1998; Struchen, Pappadis, et al., 2008): it has been hypothesized that the discrepancy of the SELF vs. OTHER comparison results is related to time post injury (with longer intervals corresponding to lower differences in SELF and OTHER ratings) (Struchen, Pappadis, et al., 2008), and this, together with sex composition of the sample, might have played a role in our study, as the mean time post injury is relatively high. It should also be considered that at least one previous study (Douglas et al, 2007) that has reported significant differences in SELF and OTHER scores following TBI, with individuals with TBI under-reporting their communication problems, also found a significant correlations between SELF and OTHER, indicating that the two situations are compatible and perhaps pointing to a systematic bias within TBI populations.

Interestingly, ERT performance was not a significant predictor of self- or other-reported emotion recognition (i.e., the fifth LCQ factor). One possible explanation for this seemingly contradictory finding is that, although the four items were created to measure a specific construct (affect recognition), they do not cluster together as intended. Indeed, as mentioned above, these items were added to the LCQ recently and were not included in Struchen’s factor analysis. For practical reasons, we chose to treat these items as an independent factor, but they may load onto one of the four subscales identified by Struchen and colleagues. Future studies with a larger sample size should explore whether the LCQ factor structure is changed by the addition of these four items. We speculate that errors in emotion recognition per se are not detectable to either people with TBI or their partners in an ongoing conversation but manifest themselves in higher-level problems such as “[carrying] on talking about things for too long in your conversations,” which could occur because the speaker fails to recognize the listener’s facial emotions accurately, and which might be mistakenly attributed to other causes (e.g., lack of empathy) (Mathersul et al., 2009).

In summary, the current study provided evidence that emotion recognition can predict specific elements of communication competence following TBI. Although longitudinal studies are needed to further examine the presence of a causal relationship between these two variables, our results support the idea that specific perceived communication problems, particularly disinhibition and conversation effectiveness, are linked to errors in facial-affect recognition. These findings highlight the importance of instructing communication partners about emotion recognition deficits and their potential impact on everyday interactions and may inform selection of treatment targets and intervention methods.

Supplementary Material

1

Table 1.

Demographic information of participants

N AGE
(Years)
(Mean±SD) 		SEX
(Females)		 EDUCATION
(Years)
(Mean±SD) 		CHRONICITY
(Months)
(Mean±SD) 		SITE Masked
HC 42 45.74±14.68 23 15.33±1.83 N/A 20
TBI 46 47.09±17.07 21 14.88±2.58 97.98±16.5 24
Group Differences (p) N/A .69 1 .34 N/A .83
Open in a new tab

Note: HC=Healthy comparison participants, TBI=Traumatic brain injury, N=Sample size, p=p-value, SD=Standard deviation

Masked Sentence 1.

Participants were recruited from the communities surrounding the University of Iowa (Iowa City, IA) and the University of Madison-Wisconsin (Madison, WI). The sample included in the current study partially overlaps with previous studies from our group (Rigon, Duff, McAuley, Kramer, & Voss, 2016; Rigon, Turkstra, et al., 2016; Rigon, Voss, Turkstra, Mutlu, & Duff, 2016).

Masked Sentence 2.

As reported in our previous work, individuals with TBI performed worse than demographically-matched healthy comparisons on a dynamic emotion recognition task (Rigon et al., 2016).

PUBLIC SIGNIFICANCE STATEMENT.

This study advances the idea that social cue perception, and in particular the ability to correctly interpret the emotion of others by looking at their face, plays a role in specifics aspects of social communication impairment following traumatic brain injury.

Acknowledgments

This work was supported by NICHD/NCMRR grant R01 HD071089.

References

  1. Aron A, Aron EN, Smollan D. Inclusion of Other in the Self Scale and the structure of interpersonal closeness. J Pers Soc Psychol. 1992;63(4):596–612. doi: 10.1037/0022-3514.63.4.596. [DOI] [Google Scholar]
  2. Babbage DR, Yim J, Zupan B, Neumann D, Tomita MR, Willer B. Meta-analysis of facial affect recognition difficulties after traumatic brain injury. Neuropsychology. 2011;25(3):277–285. doi: 10.1037/a0021908. [DOI] [PubMed] [Google Scholar]
  3. Dahlberg CA, Cusick CP, Hawley LA, Newman JK, Morey CE, Harrison-Felix CL, Whiteneck GG. Treatment efficacy of social communication skills training after traumatic brain injury: a randomized treatment and deferred treatment controlled trial. Arch Phys Med Rehabil. 2007;88(12):1561–1573. doi: 10.1016/j.apmr.2007.07.033. [DOI] [PubMed] [Google Scholar]
  4. de Sousa A, McDonald S, Rushby J, Li S, Dimoska A, James C. Understanding deficits in empathy after traumatic brain injury: The role of affective responsivity. Cortex. 2011;47(5):526–535. doi: 10.1016/j.cortex.2010.02.004. [DOI] [PubMed] [Google Scholar]
  5. Despins EH, Turkstra LS, Struchen MA, Clark AN. Sex-Based Differences in Perceived Pragmatic Communication Ability of Adults With Traumatic Brain Injury. Arch Phys Med Rehabil. 2015 doi: 10.1016/j.apmr.2014.06.023. [DOI] [PubMed] [Google Scholar]
  6. Douglas JM. Relation of executive functioning to pragmatic outcome following severe traumatic brain injury. J Speech Lang Hear Res. 2010a;53(2):365–382. doi: 10.1044/1092-4388(2009/08-0205). [DOI] [PubMed] [Google Scholar]
  7. Douglas JM. Using the La Trobe Communication Questionnaire to Measure Perceived Social Communication Ability in Adolescents With Traumatic Brain Injury. Brain Impairment. 2010b;11(02):171–182. [Google Scholar]
  8. Douglas JM, Bracy CA, Snow PC. Measuring perceived communicative ability after traumatic brain injury: reliability and validity of the La Trobe Communication Questionnaire. J Head Trauma Rehabil. 2007;22(1):31–38. doi: 10.1097/00001199-200701000-00004. [DOI] [PubMed] [Google Scholar]
  9. Douglas JM, O’Flaherty CA, Snow PC. Measuring perception of communicative ability: the development and evaluation of the La Trobe communication questionnaire. Aphasiology. 2000;14(3):251–268. doi: 10.1080/026870300401469. [DOI] [Google Scholar]
  10. Dumont C, Gervais M, Fougeyrollas P, Bertrand R. Toward an explanatory model of social participation for adults with traumatic brain injury. J Head Trauma Rehabil. 2004;19(6):431–444. doi: 10.1097/00001199-200411000-00002. [DOI] [PubMed] [Google Scholar]
  11. Finset A, Dyrnes S, Krogstad JM, Berstad J. Self-reported social networks and interpersonal support 2 years after severe traumatic brain injury. Brain Inj. 1995;9(2):141–150. doi: 10.3109/02699059509008187. [DOI] [PubMed] [Google Scholar]
  12. Green RE, Turner GR, Thompson WF. Deficits in facial emotion perception in adults with recent traumatic brain injury. Neuropsychologia. 2004;42(2):133–141. doi: 10.1016/j.neuropsychologia.2003.07.005. [DOI] [PubMed] [Google Scholar]
  13. Jackson HF, Moffat NJ. Impaired emotional recognition following severe head injury. Cortex. 1987;23(2):293–300. doi: 10.1016/s0010-9452(87)80039-4. [DOI] [PubMed] [Google Scholar]
  14. Kessels RP, Montagne B, Hendriks AW, Perrett DI, de Haan EH. Assessment of perception of morphed facial expressions using the Emotion Recognition Task: normative data from healthy participants aged 8-75. J Neuropsychol. 2014;8(1):75–93. doi: 10.1111/jnp.12009. [DOI] [PubMed] [Google Scholar]
  15. Knox L, Douglas J. Long-term ability to interpret facial expression after traumatic brain injury and its relation to social integration. Brain Cogn. 2009;69(2):442–449. doi: 10.1016/j.bandc.2008.09.009. [DOI] [PubMed] [Google Scholar]
  16. Law Smith MJ, Montagne B, Perrett DI, Gill M, Gallagher L. Detecting subtle facial emotion recognition deficits in high-functioning Autism using dynamic stimuli of varying intensities. Neuropsychologia. 2010;48(9):2777–2781. doi: 10.1016/j.neuropsychologia.2010.03.008. [DOI] [PubMed] [Google Scholar]
  17. Malec JF, Brown AW, Leibson CL, Flaada JT, Mandrekar JN, Diehl NN, Perkins PK. The mayo classification system for traumatic brain injury severity. J Neurotrauma. 2007;24(9):1417–1424. doi: 10.1089/neu.2006.0245. [DOI] [PubMed] [Google Scholar]
  18. Mathersul D, Palmer DM, Gur RC, Gur RE, Cooper N, Gordon E, Williams LM. Explicit identification and implicit recognition of facial emotions: II. Core domains and relationships with general cognition. J Clin Exp Neuropsychol. 2009;31(3):278–291. doi: 10.1080/13803390802043619. [DOI] [PubMed] [Google Scholar]
  19. McDonald S, Flanagan S, Rollins J, Kinch J. TASIT: A new clinical tool for assessing social perception after traumatic brain injury. J Head Trauma Rehabil. 2003;18(3):219–238. doi: 10.1097/00001199-200305000-00001. [DOI] [PubMed] [Google Scholar]
  20. McDonald S, Tate R, Togher L, Bornhofen C, Long E, Gertler P, Bowen R. Social skills treatment for people with severe, chronic acquired brain injuries: a multicenter trial. Arch Phys Med Rehabil. 2008;89(9):1648–1659. doi: 10.1016/j.apmr.2008.02.029. [DOI] [PubMed] [Google Scholar]
  21. Milders M, Fuchs S, Crawford JR. Neuropsychological impairments and changes in emotional and social behaviour following severe traumatic brain injury. J Clin Exp Neuropsychol. 2003;25(2):157–172. doi: 10.1076/jcen.25.2.157.13642. [DOI] [PubMed] [Google Scholar]
  22. Montagne B, Kessels RP, De Haan EH, Perrett DI. The Emotion Recognition Task: a paradigm to measure the perception of facial emotional expressions at different intensities. Percept Mot Skills. 2007;104(2):589–598. doi: 10.2466/pms.104.2.589-598. [DOI] [PubMed] [Google Scholar]
  23. Pettersen L. Sensitivity to emotional cues and social behavior in children and adolescents after head injury. Percept Mot Skills. 1991;73(3 Pt 2):1139–1150. doi: 10.2466/pms.1991.73.3f.1139. [DOI] [PubMed] [Google Scholar]
  24. Prigatano GP, Altman IM. Impaired awareness of behavioral limitations after traumatic brain injury. Arch Phys Med Rehabil. 1990;71(13):1058–1064. [PubMed] [Google Scholar]
  25. Prigatano GP, Pribram KH. Perception and memory of facial affect following brain injury. Percept Mot Skills. 1982;54(3):859–869. doi: 10.2466/pms.1982.54.3.859. [DOI] [PubMed] [Google Scholar]
  26. Rigon A, Turkstra L, Mutlu B, Duff M. The female advantage: sex as a possible protective factor against emotion recognition impairment following traumatic brain injury. Cogn Affect Behav Neurosci. 2016 doi: 10.3758/s13415-016-0437-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Rosenberg H, Dethier M, Kessels RP, Westbrook RF, McDonald S. Emotion perception after moderate-severe traumatic brain injury: The valence effect and the role of working memory, processing speed, and nonverbal reasoning. Neuropsychology. 2015;29(4):509–521. doi: 10.1037/neu0000171. [DOI] [PubMed] [Google Scholar]
  28. Rosenberg H, McDonald S, Dethier M, Kessels RP, Westbrook RF. Facial emotion recognition deficits following moderate-severe Traumatic Brain Injury (TBI): re-examining the valence effect and the role of emotion intensity. J Int Neuropsychol Soc. 2014;20(10):994–1003. doi: 10.1017/S1355617714000940. [DOI] [PubMed] [Google Scholar]
  29. Rousseaux M, Verigneaux C, Kozlowski O. An analysis of communication in conversation after severe traumatic brain injury. Eur J Neurol. 2010;17(7):922–929. doi: 10.1111/j.1468-1331.2009.02945.x. [DOI] [PubMed] [Google Scholar]
  30. Ryan NP, Anderson V, Godfrey C, Eren S, Rosema S, Taylor K, Catroppa C. Social communication mediates the relationship between emotion perception and externalizing behaviors in young adult survivors of pediatric traumatic brain injury (TBI) Int J Dev Neurosci. 2013;31(8):811–819. doi: 10.1016/j.ijdevneu.2013.10.002. [DOI] [PubMed] [Google Scholar]
  31. Shewan CM, Kertesz A. Reliability and validity characteristics of the Western Aphasia Battery (WAB) J Speech Hear Disord. 1980;45(3):308–324. doi: 10.1044/jshd.4503.308. [DOI] [PubMed] [Google Scholar]
  32. Snow P, Douglas J, Ponsford J. Conversational discourse abilities following severe traumatic brain injury: a follow-up study. Brain Inj. 1998;12(11):911–935. doi: 10.1080/026990598121981. [DOI] [PubMed] [Google Scholar]
  33. Spikman JM, Boelen DH, Pijnenborg GH, Timmerman ME, van der Naalt J, Fasotti L. Who benefits from treatment for executive dysfunction after brain injury? Negative effects of emotion recognition deficits. Neuropsychol Rehabil. 2013;23(6):824–845. doi: 10.1080/09602011.2013.826138. [DOI] [PubMed] [Google Scholar]
  34. Spikman JM, Milders MV, Visser-Keizer AC, Westerhof-Evers HJ, Herben-Dekker M, van der Naalt J. Deficits in facial emotion recognition indicate behavioral changes and impaired self-awareness after moderate to severe traumatic brain injury. PLoS One. 2013;8(6):e65581. doi: 10.1371/journal.pone.0065581. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Struchen MA, Clark AN, Sander AM, Mills MR, Evans G, Kurtz D. Relation of executive functioning and social communication measures to functional outcomes following traumatic brain injury. NeuroRehabilitation. 2008;23(2):185–198. [PubMed] [Google Scholar]
  36. Struchen MA, Pappadis MR, Mazzei DK, Clark AN, Davis LC, Sander AM. Perceptions of communication abilities for persons with traumatic brain injury: validity of the La Trobe Communication Questionnaire. Brain Inj. 2008;22(12):940–951. doi: 10.1080/02699050802425410. [DOI] [PubMed] [Google Scholar]
  37. Vanderploeg RD, Belanger HG, Duchnick JD, Curtiss G. Awareness problems following moderate to severe traumatic brain injury: Prevalence, assessment methods, and injury correlates. J Rehabil Res Dev. 2007;44(7):937–950. doi: 10.1682/jrrd.2006.12.0163. [DOI] [PubMed] [Google Scholar]
  38. Watts AJ, Douglas JM. Interpreting facial expression and communication competence following severe traumatic brain injury. Aphasiology. 2006;20(8):707–722. doi: 10.1080/02687030500489953. [DOI] [Google Scholar]
  39. Wehman P, Kregel J, Sherron P, Nguyen S, Kreutzer J, Fry R, Zasler N. Critical factors associated with the successful supported employment placement of patients with severe traumatic brain injury. Brain Inj. 1993;7(1):31–44. doi: 10.3109/02699059309008154. [DOI] [PubMed] [Google Scholar]
  40. Ylvisaker M. Cognitive and psychosocial outcome following head injury in children. In: Hoff JT, Anderson TE, Cole TM, editors. Mild to Moderate Head Injury. London: Blackwell Scientific Publications, Inc; 1989. pp. 203–216. [Google Scholar]
  41. Ylvisaker M, Turkstra LS, Coelho C. Behavioral and social interventions for individuals with traumatic brain injury: a summary of the research with clinical implications. Semin Speech Lang. 2005;26(4):256–267. doi: 10.1055/s-2005-922104. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS965737-supplement-1.docx (83.8KB, docx)

RESOURCES