A comparison of visual and acoustic mismatch negativity as potential biomarkers in schizophrenia

Abstract

Mismatch negativity (MMN) is an event-related potential (ERP) component generated when an unexpected deviant stimulus occurs in a pattern of standard stimuli. Several studies showed that the MMN response to both auditory and visual stimuli is attenuated in schizophrenia. While previous studies investigated auditory and visual MMN in different cohorts, here we examined the potential clinical utility of MMN responses to auditory and visual stimuli within the same group of patients. Altogether 39 patients with schizophrenia and 39 healthy controls matched in age, gender, and education were enrolled. We recorded EEG using 64 channels in eight experimental blocks where we presented auditory and visual stimulus sequences. Mismatch responses were obtained by subtracting responses to standard from the physically identical deviant stimuli. We found a significant MMN response to the acoustic stimuli in the control group, whereas no significant mismatch response was observed in the patient group. The group difference was significant for the acoustic stimuli. The 12 vane windmill pattern evoked a significant MMN response in the early time window in the control group but not in the patient group. The 6 vane windmill pattern evoked MMN only in the patient group. However, we found no significant difference between the groups. Furthermore, we found no correlation between the clinical variables and the MMN amplitudes. Our results suggest that predictive processes underlying mismatch generation in patients with schizophrenia may be more affected in the acoustic compared to the visual domain. Acoustic MMN tends to be a more promising biomarker in schizophrenia.

Figure 1

(a) Schematic illustration of the experimental paradigm. In the visual session, an oddball sequence of windmill patterns was presented for 200 ms followed by an inter-stimulus interval (ISI) of 600 ms. In the acoustic session, 100 and 200 ms long beeping sounds were presented, the stimuli were followed by an ISI of 600 ms. (b) The left scalp figure presents the acoustic regions of interest (green: left frontal (AF3, F5, FC5, FC3), red: frontocentral (F3, FZ, F4), blue: right frontal (AF4, F6, FC4, FC6)) the visual regions are shown in the right scalp figure (yellow:frontal(FP1, FPZ, FP2), red:left occipital (PO7, PO5, PO3), blue: midline occipital (PZ, POZ, OZ), green: right occipital (PO4, PO6, PO8)).

Figure 2

(a) Grand average ERPs and topoplots of the mismatch signal for long acoustic stimulus. The standard (red line) and the deviant stimuli (blue line), and the mismatch response (yellow line) from the three ROIs in the two study groups. The gray area shows the 120–200 ms time window. In the time windows marked with black * a significant mismatch signal was found. The blue star markes the significant different in the mismatch signal amplitude between the study groups. The topoplots show the mismatch responses in the 120–200 ms interval. (b) Grand average ERPs and topoplots of the mismatch signal for short acoustic stimulus.

Figure 3

(a) Grand average ERPs and topoplots of the 6 vane windmill pattern as visual stimulus. (b) Grand average ERPs and topoplots of the 12 vane windmill pattern as visual stimulus.