Acute stress selectively impairs learning to act

Abstract

Stress interferes with instrumental learning. However, choice is also influenced by non-instrumental factors, most strikingly by biases arising from Pavlovian associations that facilitate action in pursuit of rewards and inaction in the face of punishment. Whether stress impacts on instrumental learning via these Pavlovian associations is unknown. Here, in a task where valence (reward or punishment) and action (go or no-go) were orthogonalised, we asked whether the impact of stress on learning was action or valence specific. We exposed 60 human participants either to stress (socially-evaluated cold pressor test) or a control condition (room temperature water). We contrasted two hypotheses: that stress would lead to a non-selective increase in the expression of Pavlovian biases; or that stress, as an aversive state, might specifically impact action production due to the Pavlovian linkage between inaction and aversive states. We found support for the second of these hypotheses. Stress specifically impaired learning to produce an action, irrespective of the valence of the outcome, an effect consistent with a Pavlovian linkage between punishment and inaction. This deficit in action-learning was also reflected in pupillary responses; stressed individuals showed attenuated pupillary responses to action, hinting at a noradrenergic contribution to impaired action-learning under stress.

Action boosts episodic memory encoding in humans via engagement of a noradrenergic system

“…Note that the behavioral task used in Exp 4, 5 and 6 was identical, with Exp 4 and 6 showing significantly better remember accuracy for Go vs. NoGo stimuli. To test for an interaction between encoding and button press in pupil diameter, the raw pupil responses (Figure 3g = 17.78; P < 0.001) as is clear from the raw pupil traces, where an initial pupil constriction, due to the light-reflex to stimulus presentation, is followed by a later differential dilation (Figure 3g), in line with previous observations in human 41,42 and nonhuman primates 18 . Critically, a significant interaction between response type (Go vs.…”

Early life stress, FK506 binding protein 5 gene (FKBP5) methylation, and inhibition-related prefrontal function: A prospective longitudinal study

“…For example, slower reaction times on go trials would allow more time to recruit inhibition-related circuitry on no-go trials (Bogacz, Wagenmakers, Forstmann, & Niuenhuis, 2009). Slowed responses have also been tied to stress exposure suggesting that stress may impair action production leading not only to slower responses but also to inaction such as freezing during threat (de Berker et al, 2016).…”

Threat of shock and aversive inhibition: Induced anxiety modulates Pavlovian-instrumental interactions.

“…Thus, in addition to the accumulating evidence of anxiety-potentiated “neutral” response inhibition ( Aylward & Robinson, 2017 ; Grillon et al, 2016 ; Robinson, Krimsky, et al, 2013 ; Torrisi et al, 2016 ), we show that anxiety specifically promotes response inhibition in a situation where the prepotent response is inhibition (i.e., when the context is specifically aversive). This suggests that ToS, by virtue of being a global aversive context, may promote a “generic” bias toward inhibition on the SART and other tasks because inhibition is the prepotent response to aversive contexts ( Aylward & Robinson, 2017 ; de Berker et al, 2016 ; Grillon et al, 2016 ; Robinson, Krimsky, et al, 2013 ; Torrisi et al, 2016 ).…”