Electrophysiological correlates for response inhibition in a Go/NoGo task

Abstract

Objective: Event-related brain potentials (ERPs) during a Go/NoGo task were investigated to elucidate the electrophysiological basis for executive and inhibitory control of responses.

Methods: We studied Go/NoGo ERPs in 13 healthy subjects during a modified continuous performance test using high-density electroencephalogram (EEG) recording. We measured peak latency, amplitude, and topographic distribution of the components, and analyzed the neural sources using low-resolution electromagnetic tomography.

Results: There were no differences between the Go and NoGo conditions in the latency, amplitude, scalp topography, or the electrical source localization of the P1 and N1 components. The N2 component was seen only in the NoGo ERP, and its source was located in the right lateral orbitofrontal and cingulate cortex. The NoGo-P3 component had larger amplitude and longer latency, and was more anteriorly localized than Go-P3; Go-P3 was located mainly in the medial part of the parietal cortex, whereas the NoGo-P3 activity was observed in the left lateral orbitofrontal cortex.

Conclusions: These results suggest that the lateral orbitofrontal and anterior cingulate areas play critical roles in the inhibitory control of behavior and that both hemispheres are involved in inhibitory cognitive function.

Introduction

The inhibition of action and the prevention of detrimental events are important executive mechanisms in human behavior. Normann and Shallice (1986) argued that executive control was engaged in situations requiring decision-making, conflict resolution, error correction, and response inhibition. Several studies have investigated the neuronal networks involved in executive function. Studies involving positron emission tomography (PET) (Buchsbaum et al., 1990, Kawashima et al., 1996), near-infrared spectroscopy (Fallgatter and Strik, 1997), and functional magnetic resonance imaging (fMRI) (Casey et al., 1997, Konishi et al., 1998), have indicated that activation of the inferior frontal areas is involved in the inhibition of motor responses. In addition to the inferior frontal area, the anterior cingulate cortex has often been implicated as an important locus in the network that exerts inhibitory control over human behaviors, e.g. Stroop interference condition (Bench et al., 1993, Taylor et al., 1994, Carter et al., 1997, Cabeza and Nyberg, 2000).

Since hemodynamic responses are slow, PET and early fMRI techniques have limited use in studies of cognitive processes unfolding in the sub-second range. Event-related, evoked brain potentials (ERPs) provide enhanced temporal resolution of brain activities. Two major ERP components have been investigated in relation to response inhibition. The first component (the N2 component) is generated in a NoGo condition and comprises a negative shift between 200 and 300 ms (Jodo and Kayama, 1992, Eimer, 1993, Fallgatter et al., 1999, Fallgatter and Strik, 1999). Falkenstein et al. (1999) reported that NoGo-N2 was attenuated and delayed in subjects with high false-alarm rates compared to those with low false-alarm rates. These authors suggested that the N2 component was a real-time correlate of a modality-specific, non-motor inhibition process, since frontal lobe N2 varies with stimulus modality and performance. Geczy et al. (1999) suggested that increased N2 amplitude in response to NoGo stimuli after Go cues might be related to increased efforts to activate the response inhibition system and to interrupt preparations for response execution. Furthermore, Pliszka et al. (2000) reported that normal children produced a large negative wave (N2) over the right inferior frontal cortex when response inhibition was required, whereas N2 was markedly reduced in attention-deficit/hyperactivity-disorder children. These findings suggest that NoGo-N2 represents an inhibitory neuronal process.

The second major ERP component (the P3 component) is a positive wave that peaks between 300 and 600 ms, and is also modulated in Go/NoGo conditions. The frontal P3 amplitude is larger in the NoGo than in the Go condition (Eimer, 1993, Kopp et al., 1996). Although the overlapping of movement-related activities may influence the difference between Go- and NoGo- ERPs within this time range (Falkenstein et al., 1999), P3 modulation is generally considered to be an inhibitory mechanism. Topographically, P3 in the Go condition was maximal at centroparietal sites, whereas in the NoGo condition P3 was maximal at frontocentral sites (Fallgatter et al., 1999, Fallgatter and Strik, 1999). Three-dimensional source localization analysis with low-resolution electromagnetic tomography (LORETA) showed that significantly stronger neural activity occurred in the right frontal lobe during NoGo trials than in Go trials (Strik et al., 1998). In contrast, Weisbrod et al. (2000) reported that NoGo-P3 was lateralized to the left side, over the frontal area, thus emphasizing the importance of the left frontal area in executing behavioral control.

Due to relatively poor spatial resolution, previous studies have not precisely mapped the locations of brain activity during Go/NoGo tasks (Keilp et al., 1997, Reinvang, 1998, Fallgatter and Strik, 1999, Volz et al., 1999). Better resolution might be achieved with higher density electrode arrays combined with a source localization method. The aim of this study was to accurately identify areas undergoing activation during Go/NoGo tasking. We hypothesized that the frontal lobe, in particular the inferior frontal area and the anterior cingulate cortex, might be involved in generating ERPs under NoGo conditions.