Neural correlates of recognition memory of social information in people with schizophrenia

Abstract

Background: Social dysfunction is a hallmark characteristic of schizophrenia. Part of it may stem from an inability to efficiently encode social information into memory and retrieve it later. This study focused on whether patients with schizophrenia show a memory boost for socially relevant information and engage the same neural network as controls when processing social stimuli that were previously encoded into memory.

Methods: Patients with schizophrenia and healthy controls performed a social and nonsocial picture recognition memory task while being scanned. We calculated memory performance using d'. Our main analysis focused on brain activity associated with recognition memory of social and nonsocial pictures.

Results: Our study included 28 patients with schizophrenia and 26 controls. Healthy controls demonstrated a memory boost for socially relevant information. In contrast, patients with schizophrenia failed to show enhanced recognition sensitivity for social pictures. At the neural level, patients did not engage the dorsomedial prefrontal cortex (DMPFC) as much as controls while recognizing social pictures.

Limitations: Our study did not include direct measures of self-referential processing. All but 3 patients were taking antipsychotic medications, which may have altered both the behavioural performance during the picture recognition memory task and brain activity.

Conclusion: Impaired social memory in patients with schizophrenia may be associated with altered DMPFC activity. A reduction of DMPFC activity may reflect less involvement of self-referential processes during memory retrieval. Our functional MRI results contribute to a better mapping of the neural disturbances associated with social memory impairment in patients with schizophrenia and may facilitate the development of innovative treatments, such as transcranial magnetic stimulation.

Fig. 1

Mean recognition memory performance for social and nonsocial pictures in healthy controls and patients with schizophrenia.

Fig. 2

Brain regions whose activity was significantly greater in patients with schizophrenia and controls during the processing of old versus new social pictures. Reported activations were thresholded at p < 0.005, with an extent threshold of 43 contiguous voxels, corresponding to a false-positive discovery rate of less than 5% across the whole brain. The shade bar represents t values.

Fig. 3

Brain regions showing greater activity in controls than patients with schizophrenia during the retrieval of old versus new social pictures. Reported activations were thresholded at p < 0.005, with an extent threshold of 43 contiguous voxels, corresponding to a false-positive discovery rate of less than 5% across the whole brain.

Fig. 4

Parameter estimates (β weights) of the blood oxygen level–dependent signal extracted from a cluster of activation in the dorsomedial prefrontal cortex. This cluster was significantly more active in controls than patients during successful retrieval of social pictures (i.e., social pictures correctly recognized v. social pictures correctly rejected).