Stress-sensitive inference of task controllability.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Psychiatry

Background:

• Accurate estimation of environmental controllability is crucial for adaptive behavior and decision-making.
• Understanding the neural mechanisms of controllability estimation and its modulation by stress is vital for mental health research.

Purpose of the Study:

• To elucidate the computational and neural mechanisms underlying human controllability estimation.
• To investigate how stress exposure impacts controllability inference and decision-making processes.

Main Methods:

• Utilized an explore-and-predict task with trial-by-trial analysis of choices, decision times, and neural activity (blood oxygen-level dependent responses).
• Employed computational models, including an 'actor' and a 'spectator' model, to analyze statistical learning and prediction errors.
• Examined the effects of inescapable stress exposure on controllability estimates and model reliance.

Main Results:

• Human controllability estimation involves comparing predictions from actor and spectator models of the environment.
• Neural activity in striatal and medial prefrontal areas correlated with prediction error differences between the two models.
• Activity in posterior cingulate, temporoparietal, and prefrontal cortices reflected changes in estimated controllability.
• Inescapable stressors reduced controllability estimates and increased spectator model reliance in an anxiety-dependent manner.

Conclusions:

• The findings provide a mechanistic account of how the brain infers environmental controllability.
• Stress exposure distorts controllability estimates through altered statistical learning and increased reliance on a less adaptive spectator model.
• This research offers insights into the neural basis of anxiety and stress-related disorders.