Edited by:
Reviewed by:
*Correspondence:
†
This article was submitted to Psychopathology, a section of the journal Frontiers in Psychology.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
There has been extensive research on impaired emotion recognition in schizophrenia in the facial and vocal modalities. The literature points to biases toward non-relevant emotions for emotional faces but few studies have examined biases in emotional recognition across different modalities (facial and vocal). In order to test emotion recognition biases, we exposed 23 patients with stabilized chronic schizophrenia and 23 healthy controls (HCs) to emotional facial and vocal tasks asking them to rate emotional intensity on visual analog scales. We showed that patients with schizophrenia provided higher intensity ratings on the non-target scales (e.g., surprise scale for fear stimuli) than HCs for the both tasks. Furthermore, with the exception of neutral vocal stimuli, they provided the same intensity ratings on the target scales as the HCs. These findings suggest that patients with chronic schizophrenia have emotional biases when judging emotional stimuli in the visual and vocal modalities. These biases may stem from a basic sensorial deficit, a high-order cognitive dysfunction, or both. The respective roles of prefrontal-subcortical circuitry and the basal ganglia are discussed.
Chronic schizophrenia is a disabling disease that encompasses both cognitive and emotional disorders (
The processing of emotional stimuli in schizophrenia has been investigated in several modalities. Some studies, for instance, have explored the visual modality with faces (see
In addition, few studies have adopted an approach across sensorial modalities, exposing the same patients to both vocal and facial stimuli in two independent tasks. In their review,
The present study was designed to test emotional biases in chronic schizophrenia in two modalities (facial and vocal) of emotion recognition. To avoid categorization biases, we asked participants to provide intensity ratings on a set of emotional visual analog scales in two facial and vocal emotion tasks taken from
Twenty-three (15 men and 8 women) patients with chronic schizophrenia and 23 (12 men and 11 women) HCs participated in this study. All participants were native French speakers.
Patients who tested as clinically stable were recruited from outpatient units at Guillaume Régnier Hospital (Rennes, France). The diagnosis of schizophrenia was established by a clinically trained psychiatrist according to the Mini International Neuropsychiatric Inventory (
We also recruited 23 HCs. Inclusion criteria for HCs were no current or past mental illness or any psychotropic treatment.
Exclusion criteria for all participants were neurological and systemic illness, head injury with loss of consciousness longer than 15 min, significantly impaired vision or auditory acuity, and a score below 130 on the Mattis Dementia Rating Scale (MDRS;
Patients and HCs were matched on sex, age, education level and handedness (Edinburgh Handedness Inventory;
Clinical and demographic characteristics of patients with schizophrenia and healthy control participants.
Schizophrenia patients |
Healthy controls |
Stat. val. | ||
---|---|---|---|---|
Age (years) | 33.92 (7.27) | 36.51 (7.23) | 0.46 | |
Sex (male/female) | 15/8 | 12/11 | χ2 = 0.81 | 0.37 |
Edinburgh Handedness Quotient | 85.87 (33.05) | 85 (32.79) | 0.93 | |
Education (years) | 12.70 (2.01) | 13.13 (1.79) | 0.44 | |
Duration of illness (years) | 12.43 (6.50) | |||
PANSS (overall score) | 66.30 (18.58) | |||
PANSS (positive subscale) | 10.91 (3.54) | |||
PANSS (negative subscale) | 23.52 (9.11) | |||
PANSS (general psychopathology) | 31.86 (8.64) | |||
SANS score | 46.09 (25.37) | |||
Neuroleptic dosage (CPZ equivalent) | 680.39 (461.56) | |||
Patients on antidepressants (%) | 26.09% |
Written informed consent was obtained from each participant and the study was conducted in accordance with the Declaration of Helsinki. The study was approved by the local ethics committee (CPP Ouest II- Angers; no. 2012/16).
The current severity of the patients’ psychiatric symptoms was assessed using the Positive and Negative Syndrome Scale (
In order to assess the relationship between cognitive dysfunction and emotion processing, participants underwent a neuropsychological assessment by a trained neuropsychologist (
Neuropsychological background of patients with schizophrenia and healthy control participants.
Schizophrenia patients |
Healthy controls |
Stat. val. |
|||
---|---|---|---|---|---|
MDRS (max. 144) | 137.74 (6.64) | 141.43 (1.75) | 2.58 | 0.01 | |
Digit span | Forward | 5.91 (1.28) | 5.74 (0.92) | -0.53 | 0.60 |
Backward | 4.87 (1.14) | 4.61 (1.47) | -0.67 | 0.50 | |
Verbal fluency | Categorical (2 min) | 24.91 (8.02) | 31.13 (6.97) | 0.89 | 0.38 |
Phonemic (2 min) | 19.52 (7.54) | 21.39 (6.65) | 2.81 | 0.007 | |
Stroop | Interference | -0.55 (8.56) | 3.85 (8.14) | 1.79 | 0.08 |
TMT | A (seconds) | 46.78 (17.24) | 31.96 (10.31) | -3.54 | <0.001 |
B (seconds) | 100.65 (53.63) | 64.35 (19.30) | -3.05 | 0.004 | |
B–A (seconds) | 53.87 (43.92) | 32.39 (15.36) | -2.21 | 0.03 | |
MCST | Categories (max. = 6) | 5.87 (0.43) | 6 (0) | 1.45 | 0.15 |
Errors | 2.87 (2.75) | 2.70 (7.40) | -0.11 | 0.91 | |
Perseverative errors | 0.87 (1.49) | 0.04 (0.21) | -2.64 | 0.01 | |
Digit Symbol-Coding (WAIS III) | 53.62 (14.31) | 74.13 (19.52) | 4.06 | <0.001 | |
PEGA (max. 30) | 29.04 (1.15) | 29.34 (0.93) | -0.69 | 0.49 | |
Benton Facial Recognition Test | 46.57 (3.26) | 45.61 (5.79) | 0.99 | 0.33 |
This task featured two different sets of 56 emotional facial expressions produced by eight actors (four male and four female) per set. Six emotions (fear, disgust, anger, sadness, surprise, and happiness) were depicted, alongside neutral faces. For each emotion, there were four male faces and four female ones, making a total of eight stimuli per emotional category. These photographs were taken from Ekman and Friesen’s Pictures of Facial Affect (
The vocal stimuli were taken from the Montreal Affective Voices database developed and validated by
We administered two emotional tasks: a facial emotion recognition task featuring facial stimuli and a vocal emotion recognition task featuring vocal stimuli. The facial emotion recognition task was always performed before the vocal emotion recognition task. The procedure was the same for both. Participants were seated in a quiet room, in front of a computer. Each condition (faces or voices) was displayed by an Authorware program.
At the beginning of each trial, a progress bar appeared on the computer screen. This was followed by the stimulus and participants were asked to rate its emotional content on scales that were simultaneously displayed on the screen. More specifically, participants were instructed to indicate the extent to which the different emotional categories were expressed on visual analog scales ranging from
The entire protocol was completed in 90 min with a pause between clinical assessments and evaluation of emotion recognition.
For the two emotion recognition tasks (facial and vocal), we performed two levels of analysis.
First, we compared the percentages of correct responses for the two tasks. For each trial, we compared the rating of intensity for each scale. A response was deemed to be correct when a participant rated the target scale higher than all the non-target scales. If the intensity were higher for the target scale (e.g., the anger scale for an angry face), we scored “1”; if not we scored “0.” We summed the score for each task and for each emotional category and then, we calculated the percentage of correct responses. We performed a repeated measures ANOVA with group (two levels: patients and HCs) as the between-subjects variable and task (two levels: visual and. vocal) and emotion (seven levels: anger, fear, sadness, disgust surprise, happiness, and neutral) as the within-subjects variable.
Then, we compared the ratings given by the two groups for each task (facial and vocal) on the scales for each type of emotion and for each individual scale, distinguishing between (1) the target scales, that is, the mean ratings on the scales (e.g., Anger scale) corresponding to the relevant stimuli (e.g., anger stimulus), and (2) the non-target scales, that is, the mean ratings on the scales that did not correspond to the stimuli (e.g., Fear scale for anger stimulus). This second analyses enabled us to take into account target ratings on incorrect responses (e.g., when recognizing an “anger” stimuli, rating 80% on the “anger” scale and “90%” on the surprise scale). In order to pinpoint impaired emotion biases in schizophrenia, we performed contrasts for the non-target ratings for each condition. To this end, we ran repeated-measures ANOVA with group (patients and HCs; two levels) as a between-participants factor, and emotion (seven levels) and scale (seven levels) as within-participants factors. We compared the HC and patient groups on each emotional category and each rating scale.
Sociodemographic, psychiatric and neuropsychological data were compared using at-test for two independent groups for two independent groups (patients and HCs).
Versions A and B of the emotional recognition tasks were compared using a
Correlations between (1) clinical assessments and daily neuroleptic dose and vocal and facial emotion recognition (2) executive functions performances and the results of the emotion recognition tasks were assessed using Spearman’s rank correlation coefficient for the patients group.
Statistical analyses were performed using Statistica 8.0. The significance threshold was set at
The patients with schizophrenia scored significantly lower than HCs on the MDRS (
Number of correct responses, expressed as a percentage of total responses (standard errors, SE) for categorical judgments in the facial and vocal emotion recognition tasks of patients with schizophrenia and healthy controls.
Schizophrenia patients | Healthy controls | ||
---|---|---|---|
Mean (±SE) | Mean (±SE) | ||
Facial emotion | Anger | 68.24 (19.28) | 79.34 (20.00) |
Recognition task | Disgust | 58.93 (27.64) | 82.61 (17.57) |
Fear | 54.89 (24.35) | 52.17 (26.56) | |
Happiness | 90.22 (14.58) | 97.83 (4.84) | |
Neutral | 70.65 (24.89) | 84.24 (17.76) | |
Sadness | 67.93 (22.87) | 72.28 (19.93) | |
Surprise | 88.04 (20.81) | 97.83 (6.14) | |
Overall | 71.28 (10.93) | 80.90 (8.28) | |
Vocal emotion | Anger | 41.74 (19.92) | 60.87 (17.56) |
Recognition task | Disgust | 62.61 (30.33) | 83.48 (19.68) |
Fear | 53.04 (24.58) | 73.04 (26.01) | |
Happiness | 86.09 (24.45) | 99.13 (4.17) | |
Neutral | 74.78 (33.15) | 86.96 (21.41) | |
Sadness | 82.61 (22.81) | 92.17 (11.66) | |
Surprise | 51.30 (30.05) | 48.70 (20.74) | |
Overall | 64.60 (13.91) | 77.76 (7.21) |
Analysis revealed an Emotion × Task interaction [
Analysis revealed a main effects of task [
There was no Group × Emotion interaction [
No significant difference was found between the accuracy (percentage of correct responses) in the HC group for version A and B for facial recognition task (
We found a significant correlation between the PANSS scores and the overall scores on the facial (
Spearman’s correlation coefficients showed a significant correlation between digit span forward and the overall score on the vocal emotion recognition task (
In this study, we sought to pinpoint the presence of emotional biases in both facial and vocal emotion recognition in chronic schizophrenia, controlling for confounding factors. In both the visual and vocal modalities, our main findings pointed to emotional biases in the patients’ ratings of emotional intensity of all the target emotions.
Using an original emotion recognition procedure that had already been validated (
At a second level of analysis, the patients with schizophrenia were found to be less discriminative than HCs in their intensity ratings for both modalities (facial and vocal). Although they provided the same intensity ratings on the target scales (i.e., categorical judgments), they responded differently on the non-target scales thus exhibiting biases in their responses. These results are consistent with previous studies reporting emotional biases (
Delusions and hallucinations in schizophrenia may contribute to the introduction of biases. Patients with greater positive (including delusions and hallucinations) and general symptoms (including anxiety, attentional deficit, and depression) achieve lower overall recognition scores in the two emotional paradigms (vocal and facial).
Moreover, patients with poorer digit span forward scores had poorer overall scores on the facial emotion recognition task, while a higher number of perseverations on the MCST was related to a lower overall score on the vocal recognition task.
The biases in emotion processing observed here can be explained in three ways. The first explanation refers to the sensory deficit in schizophrenia. Several studies have reported a basic sensory disturbance in the processing of emotion in the facial (
The second explanation concerns the lack of cognitive control in schizophrenia (
Third and last, the mutual influence of sensory disorder and cognitive control impairment in schizophrenia may explain the emotional disorders described in the literature. In a recent study using event-related potentials,
Schizophrenia is related to brain dysfunction, especially in the prefrontal striatonigral circuit. Some studies have shown that a dysfunction of the prefrontal cortex is related to the identification of emotional context in schizophrenia (
This study had several limitations that need to be addressed. First, only 23 patients were assessed. However, we took care to recruit patients with the same clinical characteristics, in order to construct a homogenous group. Second, as described by
Further research is needed to investigate the involvement of cognitive control in the management of top-down and bottom-up processing, using both implicit and explicit emotion processing tasks.
In summary, this study showed that chronic schizophrenia induces emotional biases for all emotions in two sensory modalities (visual and vocal), and appears to cause interference in emotion recognition. There are now at least two mechanisms that need to be, considered if we are to explain impairments in emotion recognition: a deficit in sensory functions and a lack of cognitive control. These results could help to enhance current cognitive and emotional remediation in schizophrenia.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We would like to thank Elizabeth Wiles-Portier for revising the English style as well as the reviewers for their precious comments on the earlier versions of this manuscript. We thank Dr. Florian Naudet (CIC INSERM 0203; Rennes) for his help in the statistical analysis of this study. The study was carried out at the Psychiatry Unit (Prof. Bruno Millet) of Guillaume Régnier Hospital in Rennes, France. This work was supported by a grant from the Fondation de la Recherche and the Fondation Pierre Deniker. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.