Rethinking Body Ownership in Schizophrenia: Experimental and Meta-analytical Approaches Show no Evidence for Deficits

Abstract

Schizophrenia is a severe psychiatric disorder, in which patients experience an abnormal sense of self. While deficits in sensorimotor self-representation (agency) are well documented in schizophrenia, less is known about other aspects of bodily self-representation (body ownership). Here, we tested a large cohort (N = 59) of chronic schizophrenia patients and matched controls (N = 30) on a well-established body illusion paradigm, the Full Body Illusion (FBI). In this paradigm, changes in body ownership are induced through prolonged multisensory stimulation, in which participants are stroked on their back while seeing the stroking on the back of a virtual body. When the felt and seen stroking are synchronous, participants typically feel higher identification with the seen body as well as a drift in self-location towards it. However, when the stroking is asynchronous, no such changes occur. Our results show no evidence for abnormal body ownership in schizophrenia patients. A meta-analysis of previous work corroborates this result. Thus, while schizophrenia patients may be impaired in the sense of agency, their multisensory bodily self-representation, as tested here, seems to be unaffected by the illness.

Figures

Fig. 1.

(A) Setup for the Full Body Illusion. (B) The 8 questions of the body ownership questionnaire.

Fig. 2.

Schizophrenia patients and controls answers for the subjective experience. We found a main effect of the questions type (experimental vs control questions), for the synchronous compared to asynchronous conditions (higher responses for the synchronous condition), as well as a group effect, with higher scores for schizophrenia patients compared to controls.

Fig. 3.

Ranking analysis for schizophrenia patients and controls. We calculated the mean rank across all participants for the experimental questions (EXPQ) and the drift. We then ran 10 000 simulations for a random set of participants, where the scores for the questions would be independent of the score for another question or from the drift. The simulation matched the data. Error bars are the SDs of the ranks.

Fig. 4.

Forest plot of 3 studies considering the interaction between the condition (synch/asynch) and the group (schizophrenic patients/healthy controls [SCZ/HTY]) (N = 134, k = 3, g = 0.005, P = .992).