Game theoretical approach to theory of mind deficits in schizophrenic patients with delusion(s) of reference

Abstract

BACKGROUND: There is already ample evidence that theory of mind (ToM) is impaired in people with schizophrenia. However, a specific relationship between impaired ToM and paranoid delusions, while intuitively reasonable, has only been demonstrated in few studies. Psychometric properties of ToMtasksemployed in these studiesmay be a complicating factor in drawing conclusions about the relationships. Because most tests of ToM focus on the third-person, the stimuli presented are static which does not involve the dynamic nature in daily social interactions. The tasks fail to simulate the cognitive demands faced by individuals in real social situations,which give rise to our ability to ascribe mental states to our social partners. The current study proposed how the games traditionally used in game theoretic research can engagemechanisms specific to reasoning about themental states of a social partner and addressed how the mentalising deficits in schizophrenia may impair performance in the paradigms. We also aimed to clarify whether particular paranoid symptoms, such as delusions of reference (DOR), are implicated in mentalising impairment. METHODS: Three groups of subjects were recruited: (a) patients with DOR as main positive symptom (n=17); (b) patients with no clinically significant DOR symptoms (n=22); and (c) normal controls (n=20). An iterated prisoner's dilemma (PD) game was administered to all 3 groups of subjects. After completion of the PD task, a semi-structured interview was conducted to draw forth the subjects' strategic reasoning when playing the game, as well as their conception of the thought process going through their opponent's mind. Mentalising performance was rated on three criteria: whether the participant reported taking the opponent into account when making his or her own choices, whether the participant recognized the correct strategy of the opponent, and whether the participant's strategic behavioral choices are influenced by the opponent. All subjects were also administered the Character Intention Task (CIT; Sarfati et al., 1997), which was devised to test understanding of the intentions of nonverbal comic strip characters, and a neuropsychological test battery including measures of intelligence, memory, attention and executive functioning. Current symptom profiles of the schizophrenic subjects were assessed using the Scale for the Assessment of Positive and Negative Symptoms. RESULTS: The schizophrenic groups exhibited significant ToM impairments in the CIT compared to healthy control group (F=9.407, p<0.001). However, the mentalising deficit in the PD task was only observed for those schizophrenic patients who had DOR (F=10.960, p<0.001). In the patient groups, severity of DOR was negatively correlated with mentalising measures in the PD task (r=-0.584, p<0.001), but was not correlated with CIT scores. Of all other psychotic symptoms, only positive formal thought disorder was correlated with CIT performance (r=-0.518, p<0.001). DISCUSSION: The findings evidence ToM deficits in schizophrenia and the study, using a game theoretic approach, underlies the existence of a specific link between DOR and mentalising impairments in schizophrenia. Implications of these findings are discussed with reference to theoretical and methodological issues in current schizophrenia research. Further investigation is needed to establish the links between symptomatology and mentalising, particularly comparing symptom-specific groups and investigating the psychometric properties of ToM tasks. Reference: Sarfati, Y, Harde-Bayle, MC, Besche, C, et al. (1997). Attribution of intentions to others in people with schizophrenia. Schizophr. Res. 25, 199–209.