Excessive state switching underlies reversal learning deficits in cocaine users

Area of Science:

• Neuroscience
• Psychology
• Addiction Research

Background:

• Chronic cocaine exposure impacts neural mechanisms.
• The probabilistic reversal-learning task assesses cognitive dysfunction in addiction.
• Performance deficits in cocaine users require further clarification.

Purpose of the Study:

• To investigate decision-making deficits in cocaine users using a probabilistic reversal-learning task.
• To disambiguate the nature of performance impairments in individuals with cocaine addiction.

Main Methods:

• Utilized a probabilistic reversal-learning task with 45 cocaine users and 41 controls.
• Applied a Bayesian hidden Markov model (HMM) to analyze task performance data.

Main Results:

• Cocaine users showed a significant deficit in reaching the reversal criterion compared to controls.
• This deficit was attributed to excessive switching behavior, including heightened sensitivity to false feedback and spontaneous switching.
• A single parameter in the HMM effectively captured this decision-making behavior without needing a perseverative error parameter.

Conclusions:

• Cocaine users display a distinct pattern of excessive switching on reversal-learning tasks.
• Impulsive decision-making, rather than solely compulsive behavior, appears more relevant to the observed impairments.
• Findings support the development of diagnostic tools to differentiate and quantify decision-making dysfunctions in addiction treatment.