Swedish scientists from the Karolinska Institute studied how people with psychotic-spectrum disorders recognize emotions from faces, voices and body movements. It turned out that those who also have a history of interpersonal aggression have the lowest recognition accuracy, and they more often confuse negative emotions with positive ones. The work was published in the journal Schizophrenia.
The ability to understand other people's emotions is important for communication: facial expression, intonation and posture signal what the interlocutor feels. If these signals are misinterpreted, inadequate reactions are possible. In people with psychotic disorders (in the study these were schizophrenia, schizoaffective disorder, delusional disorder and bipolar disorder with psychotic features), such difficulties occur often, especially with negative emotions.
The study involved three groups: 79 people with a psychotic disorder and a history of aggression (from a forensic psychiatric clinic in Stockholm), 72 people with the same disorders but without aggression (from outpatient clinics), and 86 healthy volunteers. The groups were matched by age and sex; men made up 69% to 75% in each group, and the mean age was 35 to 39 years. Participants with disorders had been clinically stable for at least three months.
The subjects took tests of intelligence and reaction speed, selected synonyms for 14 words denoting emotions, and then performed a task to recognize 12 emotions that actors expressed with face, voice and body while uttering meaningless phrases. The videos were presented only as image, only as sound, or together. Participants chose an emotion from a list.
As expected, the group with disorder and aggression performed worst: accuracy ranged from 25% (despair) to 54% (anger). Participants with disorder but without aggression showed results from 29% to 66%, and healthy participants from 39% to 68%. All groups exceeded chance level, and the gaps between them were similar for different emotions and presentation methods. Emotions were recognized best when video and sound were combined.
The most noticeable difference was the pattern of errors. In the aggression group, negative emotions were taken for positive ones in 14% of cases, about 1.2 times more often than without aggression, and twice as often as in healthy participants. The authors suggest that such errors may prevent a person from noticing that their behavior is upsetting those around them and may remove a signal capable of stopping a growing conflict.
Among participants with disorders, the best predictors (factors predicting the result) of emotion recognition were fluid intelligence and understanding of words denoting emotions, followed by sex (women performed better) and education. Unexpectedly, a history of past substance abuse was associated with a better result after accounting for other factors; the authors warn that this may be a statistical artifact (a chance finding). A history of aggression remained associated with worse recognition even after adjustment for these factors, although it explained only about 2% of additional variance.
"People with psychotic spectrum and a history of interpersonal aggression have more pronounced emotion recognition deficits than people with the same disorder without aggression. This highlights the potential benefit of including assessment and training of emotion recognition in clinical interventions to reduce the risk of aggression and improve social functioning," the authors conclude.
The study does not allow conclusions about cause-and-effect relationships. The groups differed in education, intelligence and substance use, and participants were tested in a stable state, so the results do not reflect acute psychosis. Emotions were portrayed by actors rather than expressed spontaneously; clips created in Switzerland with white European actors may have been perceived differently by participants of other origins.