Alcohol-induced impairment of inhibitory control is linked to attenuated brain responses in right fronto-temporal cortex

Abstract

Background: A self-enhancing loop between impaired inhibitory control under alcohol and alcohol consumption has been proposed as a possible mechanism underlying dysfunctional drinking in susceptible people. However, the neural underpinnings of alcohol-induced impairment of inhibitory control are widely unknown.

Methods: We measured inhibitory control in 50 young adults with a stop-signal task during functional magnetic resonance imaging. In a single-blind placebo-controlled cross-over design, all participants performed the stop-signal task once under alcohol with a breath alcohol concentration of .6 g/kg and once under placebo. In addition, alcohol consumption was assessed with a free-access alcohol self-administration paradigm in the same participants.

Results: Inhibitory control was robustly decreased under alcohol compared with placebo, indicated by longer stop-signal reaction times. On the neural level, impaired inhibitory control under alcohol was associated with attenuated brain responses in the right fronto-temporal portion of the inhibition network that supports the attentional capture of infrequent stop-signals and subsequent updating of action plans from response execution to inhibition. Furthermore, the extent of alcohol-induced impairment of inhibitory control predicted free-access alcohol consumption.

Conclusions: We suggest that during inhibitory control alcohol affects cognitive processes preceding actual motor inhibition. Under alcohol, decreased brain responses in right fronto-temporal areas might slow down the attentional capture of infrequent stop-signals and subsequent updating of action plans, which leads to impaired inhibitory control. In turn, pronounced alcohol-induced impairment of inhibitory control might enhance alcohol consumption in young adults, which might promote future alcohol problems.

Trial registration: ClinicalTrials.gov NCT01097213.

Keywords: Acute alcohol intoxication; alcohol consumption; fMRI; inhibitory control; response inhibition; stop-signal task.

Copyright © 2014 Society of Biological Psychiatry. Published by Elsevier Inc. All rights reserved.

Figures

Figure 1

Sequence of experiments and fMRI time course. A: The experiment consisted of four sessions. First, we conducted free-access alcohol self-administration on two separate days (session 1+2). Second, the SST was performed during fMRI alcohol clamping, once under a constant alcohol exposure of 0.6 g/kg, and once under placebo (session 3+4). The order of the alcohol and placebo condition in the fMRI alcohol clamping part was randomized across participants. B: Timing of the fMRI alcohol clamping experiment with target and measured BrACs (mean, Error bars represent standard deviations). We also measured subjective perceptions of alcohol (SPA) and saccadic eye-movements (PS) at baseline and before the SST at T1 to track the “Level of alcohol intoxication” (see supplement). After the break, the experiment continued with other tasks (see supplement Figure S1 for complete time course). Abbreviations: ASL = arterial spin labeling, BrAC = breath alcohol concentration, PS = prosaccades, SPA = subjective perception of alcohol, SST = stop-signal task, T1 = time 1.

Figure 2

Timing of the SST. In go trials, the go stimulus was displayed until a response was recorded, but for a maximum of 1000 ms. In stop trials, the go stimulus was presented for the duration of the variable stop-signal delay (SSD; mean ±SD: alcohol = 184 ms ± 91; placebo = 196 ms ±85) followed by the stop-signal for 300 ms in successful stop trials (StopInhibit) or until a response was recorded in failed stop trials (StopFail). Go and stop trials were followed by the presentation of a central fixation cross for the duration of a jittered inter-trial interval (ITI, mean=900ms). In stop trials, the SSD (initial SSD=200 ms) was adapted dynamically according to the performance in the preceding stop trial (cf, 36): If participants successfully inhibited the response, the SSD was increased by 50 ms, if they failed to inhibit the response, the SSD was decreased by 50 ms. The SST lasted for 13 minutes. ms = milliseconds, occ/temp = occipito-temporal cortex, RT = reaction time, SD = standard deviation.

Figure 3

Activation of the fronto-subcortical motor inhibition network and bilateral occipito-temporo-parietal cortex during stop trials. Brain areas that were active across both stopping conditions (conjunction of StopInhibit and StopFail) are shown in yellow. Increased brain activation for StopInhibit is shown in green (StopInhibit > StopFail) and for StopFail in red (StopFail > StopInhibit). Some brain areas showed overlapping activation for both stopping conditions and increased activation for StopFail > StopInhibit, this is shown in orange. Voxel-wise significance threshold: p<.001 uncorrected with at least 50 connected voxels.

Figure 4. Whole-brain conjunction analysis

(A) Alcohol effects (alcoholp<.001 uncorrected, k>50). Mean brain responses for StopInhibit and StopFail (above the implicit go baseline) for alcohol and placebo are displayed for the two inhibition-related brain areas that exhibited decreased activation under alcohol in the whole-brain conjunction analysis: RIFG/insula (B), and the occipito-temporal cortex (C). Regional analyses: Correlation between alcohol-induced impairment of inhibitory control (SSRTalcohol-SSRTplacebo; >0: impaired; <0: improved) and alcohol effects on regional inhibition-related brain responses in the RIFG/insula (D), and the occipitotemporal cortex (E). Error bars represent the standard error of the mean. Locations are given in MNI-space. Abbreviations: mm3 = cubic millimeter, ms = milliseconds, occ/temp = occipito-temporal cortex, r = Pearson correlation coefficient, RIFG = right inferior frontal gyrus, SSRT = stop-signal reaction time, * = p <.025 (α-level corrected for multiple testing, p=.05/[number of tests=2]).

Figure 5

Correlation between alcohol-induced impairment of inhibitory control at 0.6 g/kg during fMRI alcohol clamping and number of alcohol requests during free-access ASA (n=38, valid free-access data). Abbreviations: ms = milliseconds, rs = Spearman correlation coefficient, SSRT = stop-signal reaction time, * = p <.05.