0
0
A new study from Caltech’s John O’Doherty reveals that problem gamblers may be more prone to escalating losses due to their brains relying on a slower form of learning when responding to negative feedback. This slow learning, while beneficial in stable environments, seems to hinder the ability to adjust quickly to the changing dynamics of gambling, where immediate responses to losses are crucial.
O’Doherty, Fletcher Jones Professor of Decision Neuroscience at Caltech, has long been interested in understanding how the brain processes feedback, both positive and negative, and how individuals use this information to guide future decisions. “We think that people learn via a thing called prediction error: the difference between what you’re expecting to get and what you actually get,” says O’Doherty. “If there’s a big discrepancy, you’ll have a big prediction error, prompting an update in your learning to improve future predictions.”
In a recent study published in The Journal of Neuroscience, O’Doherty’s team examined the decision-making behavior of 40 gamblers, half of whom were classified as problem gamblers. These participants were recruited based on questionnaires indicating symptoms of gambling disorder. The other half consisted of recreational gamblers who did not exhibit these symptoms. By excluding individuals currently in treatment for gambling disorders, the researchers aimed to avoid any influence from treatment-based learning changes.
During the study, participants underwent functional magnetic resonance imaging (fMRI) while performing two decision-making tasks. One task involved learning from rewards, where participants developed strategies to maximize earnings. The second task required participants to adjust their behavior to minimize losses, simulating the decision-making process in gambling. The goal was to examine how different brain regions responded to the prediction errors caused by gains and losses.
The study found that both groups of gamblers used similar brain regions. However, there was a key difference in how they processed losses. Problem gamblers were found to rely more on slow-learning processes, which meant they did not update their decision-making quickly enough to avoid further losses. This sluggish response to negative feedback resulted in their inability to effectively reduce losses compared to recreational gamblers.
The research highlighted significant brain activity in areas such as the anterior cingulate cortex and insular cortex in problem gamblers. These regions are involved in the integration of information over time and in tracking slow-learning prediction errors, respectively. Interestingly, the insular cortex is also implicated in substance use disorders, suggesting a possible connection between slow learning and a broader spectrum of addiction-related behaviors.
While the activation of these brain regions provides valuable insight, O’Doherty acknowledges that problem gambling is a complex disorder with many contributing factors. “The activation of these brain regions is not the whole story,” he explains. “There are probably individual differences in how these brain systems work that we will need to explore further with larger participant pools.”
Looking ahead, O’Doherty and his team hope to extend their research into other mental health conditions, such as depression and obsessive-compulsive disorder, to identify similar patterns in decision-making processes. By understanding the precise computational mechanisms behind these disorders, researchers aim to map them to specific neural circuits. This could open the door to targeted interventions, including brain stimulation and pharmacological treatments, to address the underlying causes of these conditions.
This study is a step forward in computational psychiatry, offering new insights into how the brain’s learning processes influence behavior and mental health, particularly in conditions like problem gambling.
For more details, see: Kiyohito Iigaya et al, “Computational and neural evidence for altered fast and slow learning from losses in problem gambling,” The Journal of Neuroscience (2024).
