Low-frequency oscillations reflect aberrant tone restoration during the auditory continuity illusion in schizophrenia

Abstract

Patients with schizophrenia (ScZ) often show impairments in auditory information processing. These impairments have been related to clinical symptoms, such as auditory hallucinations. Some researchers have hypothesized that aberrant low-frequency oscillations contribute to auditory information processing deficits in ScZ. A paradigm for which modulations in low-frequency oscillations are consistently found in healthy individuals is the auditory continuity illusion (ACI), in which restoration processes lead to a perceptual grouping of tone fragments and a mask, so that a physically interrupted sound is perceived as continuous. We used the ACI paradigm to test the hypothesis that low-frequency oscillations play a role in aberrant auditory information processing in patients with ScZ (N = 23). Compared with healthy control participants we found that patients with ScZ show elevated continuity illusions of interrupted, partially-masked tones. Electroencephalography data demonstrate that this elevated continuity perception is reflected by diminished 3 Hz power. This suggests that reduced low-frequency oscillations relate to elevated restoration processes in ScZ. Our findings support the hypothesis that aberrant low-frequency oscillations contribute to altered perception-related auditory information processing in ScZ.

Figure 1

Continuity rating across conditions. Control data (H) is arranged on the left side of the figure (light grey boxes); patient data (P) on the right (dark grey boxes). Grey points indicate mean values. Brackets describe all significant results from planned post-hoc comparisons (see "Analysis of behavioural data" section for details of planned comparisons included in the analysis).

Figure 2

Induced 3 Hz power across conditions. Control data (H) is arranged on the left of the figure (light grey boxes); patient data (P) on the right (dark grey boxes). Grey points indicate mean values. Significance bracket describes the result of an independent samples t-test for the IN healthy versus IN patients contrast.

Figure 3

EEG spectrograms and scalp topographies for all conditions. The spectrograms show the power change (relative to baseline), over the electrodes indicated by dots in the scalp maps. Mask onsets and offsets are indicated by dashed grey lines. The topographies show the 3 Hz power change, averaged over the analysed time window (1.2–1.6 s). Mediocentral 3 Hz power in the IN condition is larger in healthy participants compared to patients.

Figure 4

Spectrograms and topographic distributions of contrasts of interest. (A) Top row: details a time–frequency plot for the IN healthy controls versus IN patients contrast, indicating increased induced 3 Hz power during the gap/mask in the control group. Noise mask on- and offset are indicated by dashed lines. Bottom row: topographic distribution of induced 3 Hz power change within the time window of interest. Analysed channels are indicated by dots. (B) Same as in A, for the contrast between controls and patients pooled across experimental conditions.

Figure 5

Schematic representation of different stimulus types employed in the experimental paradigm.