Making decisions immediately post-stress: Evidence for dorsolateral prefrontal cortex involvement with an fMRI study

Abstract

Identifying neural markers might help to investigate the specific effects of stress on risky decision-making and to understand the behavioral interventions for risk adjustment. It is proposed that stress makes decision-making less deliberate. We employ fMRI to investigate this, with stress and post-stress decision-making occurring during the same fMRI session with a minimal stress-to-decision-making latency and concurrent electrodermal activity (EDA) measurement. In this within-subject study (40 participants, 18 males, M_age = 24.6 ± 3.9 years), participants performed the "decision-making under risk" task twice: immediately following stress induction, and after a control condition. The order of the stress and control conditions was counterbalanced. Stress was induced by asking participants to solve mental arithmetic tasks under time pressure while receiving negative feedback. During the decision-making task, participants chose between a safe and a risky option with monetary incentives. Behavioral and EDA (n = 34; due to technical issues) data confirmed that stress induction was successfully implemented. Participants took less risky decisions post-stress than post-control. An fMRI contrast analysis revealed that the right fronto-opercular and the left anterior part of the dorsolateral prefrontal cortex (dlPFC) exhibited significantly lower activation during decisions post-stress than decisions post-control. The results indicate that decisions taken immediately after exposure to the acute stressor are less risky. Furthermore, decisions post-stress are associated with reduced activation in the regions of the dlPFC, possibly leading to lower cognitive control and less deliberate decision-making post-stress. Interventions to increase dlPFC activation might be suitable to improve the decision-making post-stress.

Keywords: Decision-making; FMRI; Functional neuroimaging; Risk; Stress.

Fig. 1

Experimental design of the study for the control (a), the stress condition (b), and the decision-making under risk task (c). In stress and control conditions, the duration of the fixation cross phase was jittered between 1 to 2 s. The maximal allowed response time (RT) was dynamically adapted to the mean RT of the previous trials, with RTs being lower than the mean of previous RTs in the stress condition (b) and being higher in the control condition (a). In both conditions, the response was followed by a feedback screen displayed for a fixed 2-s interval (English translation of original German text is displayed). In contrast to the control condition, feedback screens in the stress condition consisted of an additional horizontal bar indicating the individuals’ performance. In the decision-making under risk task (c), participants had to choose between two vertical bars indicating the probability of wins with a safe option (100%) on the left and risky options (25%, 10%, and 50%, from top to bottom order) on the right. The max. RT for the respective decision was 4 s followed by a feedback screen, risky negative feedback (0€), safe positive feedback (3€), and risky positive feedback (4€)

Fig. 2

a) Participants’ self-reported stress rating on a scale of 0 to 10 (p < .0001, Z = −5.37). b) Tonic electrodermal activity level (EDA, in microSiemens, µS) during control and stress conditions (p = .02, Z = 2.49). c) Math performance (percent correct) during control and stress (p < .0001, Z = −5.51). d) Decision-making behavior demonstrated by percent of risky decisions during post-control and post-stress (p = .028, Z = −2.2)

Fig. 3

a) fMRI contrast of decision-making post-control > decision-making post-stress (cluster size 110 voxels, p _{FWE corrected (cluster level)} < .05, df = [1, 39] as per nonparametric permutation testing. Right frontal opercular and left middle frontal clusters show higher activation during decision-making post-control compared with decision-making post-stress. Overlayed on MNI152_2009 using AFNI_22.1.06 Antoninus Pius. Color scale indicates t values. b) Beta estimates from the 8-mm sphere centered at the peak activation of the right frontal clusters with MNI co-ordinates (20 54 14). For both safe (p < .001, t_[37] = 4.8) and risky conditions (p = .017, t_[36] = 2.5), we found significant differences between decision-making post-control and post-stress