Introduction: Proactive control, namely the ability to represent and manipulate goal information in working memory (WM), may be a common mechanism that drives cognitive deficits in schizophrenia (SZ). While WM studies in SZ have generally relied on visual paradigms, functional deficits in SZ, including hallucinations, are more prevalent and severe for the auditory than visual modality. We developed auditory analogs of visual Ignore/Suppress tasks conducted in conjunction with event-related potentials (ERPs) and oscillations (EROs) to examine the effects of selective attention and inhibitory control on auditory WM in SZ and healthy controls (HC).

Methods: ERPs (72 sites) were recorded from 31 HC and 10 SZ patients during encoding, maintenance, and item retrieval in 3 tasks (I: Ignore; S: Suppress; R: Remember) consisting of listening to a series of 4 letters alternately presented to each ear, followed by a 3-s maintenance interval and a probe (Fig. 1). Subjects selectively attended to letters presented to one ear and ignored those in the other ear (I), suppressed letters presented to one ear (S), or remembered all letters (R). The critical cue was provided either before (I) or after (S) the encoding series, or with the probe (R). After reference-free current source density (CSD) transformation, ERO measures were quantified for the entire 10-s trial interval using frequency principal components analysis (PCA). The CSD-fPCA solution identified a distinct factor with a spectral loadings peak at 6 Hz and maximum factor scores at mid-frontal sites (AFz, Fz, F1/2), which were statistically evaluated with nonparametric permutation tests.

Results: Distinct task-specific modulations of prominent mid-frontal theta (MFT) were seen for HC (Fig. 2), revealing significant event-related synchronization (ERS) of MFT during encoding (I only) and maintenance (I, S, R). Furthermore, MFT was selectively enhanced in HC for S compared to I and R during maintenance (i.e., not due to task difficulty or source memory requirements). In contrast, despite clear presence of MFT, SZ patients failed to show significant MFT ERS or task-specific MFT modulations.

Conclusions: These findings support previous evidence suggesting that MFT oscillations are closely related to WM processes in HC. Our new auditory WM paradigm allowed a functional dissociation of attentional “bottom-up” (I) and cognitive “top-down” (S) control processes, strongly suggesting that MFT is an electrophysiological correlate of inhibition of information stored in WM. Inasmuch as these measures reflect proactive control processes, SZ patients exhibited robust functional deficits of proactive control during these auditory WM tasks.