Non-invasive vagus nerve stimulation in epilepsy patients enhances cooperative behavior in the prisoner’s dilemma task

Figure 1. Experimental design and electrode placement. (A) The stimulation probe was attached by an independent clinician to the left ear and covered by a headband to conceal stimulation condition to the experimenter. (Ai) Following the guidelines of the manufacturer, for the tVNS condition the electrode was applied to the left cymba conchae to stimulate the auricular branch of the vagus nerve. (Aii) For the sham condition, the probe was placed in the center of the left lobule (image from50). (B) Graphical representation of the Prisoner’s dilemma task. (C) Points awarded in the prisoner’s dilemma task depending on the players’ and the opponent’s choices. (D) Likeability rating of each opponent using the SAM-scale ranging from one (least likeable) to nine (most likeable).

Figure 2. Effect of stimulation on cooperation and reaction times. (A–C) Average cooperation scores (%) (A) for each stimulation condition (tVNS/sham) and (B) each opponent (human/computer). Repeated measures ANOVA revealed a main effect of “stimulation” (F1,18 = 5.17, p = 0.035) and “opponent” (main effect opponent: F1,18 = 24.21, p < 0.001). (C) Individual cooperation score (%) for each patient during both stimulation conditions. (D, E) Average reaction time for each (D) stimulation condition and (E) opponent. Repeated measures ANOVA demonstrated a main effect “opponent” on reaction times (F1,18 = 4.76, p = 0.043). Bar graphs illustrate mean ± SEM parameters across participants.

Figure 3. Effect of stimulation on DDM parameters. A chi-square difference test comparing the goodness of fit of the two competing models used in DDM analysis demonstrated stimulation effects on (A) starting bias (p = 0.01) and (B) drift-rate (p < 0.001). Stimulation increased the starting bias towards cooperation and shifted drift-rate towards zero. Bar graphs illustrate mean ± SEM parameters across participants.