Do gambling and cocaine addiction impair learning from feedback differently?
Heavier gamblers were slower to adapt when the rewarded option switched and showed abnormal brain responses to feedback, whereas cocaine users' problems were tied to their lifetime amount of use and showed up as poorer steady-state choices.
Source
Cocaine Dependent Individuals and Gamblers Present Different Associative Learning Anomalies in Feedback-Driven Decision Making: A Behavioral and ERP Study
Study at a glance
- Design
- Case-control — Three-group comparison (cocaine dependent, pathological gamblers, healthy controls) on a probabilistic reversal learning task with EEG recorded to feedback; exposure measures used as covariates within clinical groups
- N
- N=64 · 20 cocaine-dependent individuals, 21 pathological gamblers and 23 healthy controls; two gamblers and two controls were dropped from EEG analyses due to equipment problems
- Population
- Abstinent adults in treatment for cocaine dependence or pathological gambling in Granada and Málaga, Spain, and sociodemographically matched controls
- Outcome
- Correct choices after each reversal (reversal cost) and late in each phase (asymptotic learning), decision latencies, feedback-related negativity (FRN) and its estimated cortical sources
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The researchers compared 20 abstinent cocaine-dependent individuals, 21 pathological gamblers and 23 matched healthy controls on a probabilistic reversal learning task. On each trial participants picked one of two coloured squares; one was usually rewarded and the other usually punished, and the rules reversed three times during the task. EEG was recorded to measure the feedback-related negativity, a brain response to outcomes, and interviews estimated each patient's monthly (intensity) and lifetime (severity) amount of gambling or cocaine use.
What they found
Group differences in learning were subtle overall and did not survive corrected post hoc tests. Within the gamblers, those who gambled more money per month made fewer correct choices right after reversals (a larger reversal cost) and were slower to decide, while lifetime gambling severity did not predict reversal cost. Within the cocaine group, greater lifetime cocaine use predicted fewer correct choices once learning had levelled off, not slower reversal. The feedback-related negativity was smaller in gamblers than in controls, and source analysis suggested gamblers used frontal, insular and cingulate regions for feedback processing in a pattern that differed from both controls and cocaine users.
The limits
What it doesn't show
Because this is a group-comparison study, it cannot show whether the learning differences caused the addictions or resulted from them; the authors say prospective studies are needed. Samples were small, the main group differences were weak, and several findings come from within-group covariate analyses. Cocaine users also used more alcohol and cannabis than the other groups, and although these did not directly affect the results, they could have shaped the chronic effects. The task had too few trials to track feedback responses separately after reversals, and gamblers could not be split by type of gambling, most being male slot-machine players.
Key terms
- Probabilistic reversal learning
- A task where people learn which of two options is usually rewarded, then must adjust when the reward rules are secretly swapped.
- Reversal cost
- The drop in correct choices just after the rewarded option switches, reflecting how slowly someone lets go of the old rule.
- Asymptotic learning
- The stable level of performance reached late in a phase, once learning has levelled off.
- Feedback-related negativity (FRN)
- An event-related potential that appears a few hundred milliseconds after feedback, typically larger for unexpected losses or negative outcomes.
- Intensity vs severity of use
- Intensity is the average amount used or gambled per month (acute exposure); severity is the estimated lifetime total (cumulative exposure).
Flashcards
0 of 12 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
In gamblers, what predicted a larger reversal cost?
Common questions
Why compare gamblers with cocaine users at all?
The two addictions share behavioural and neurobiological features, but gambling involves no drug toxicity, so contrasting them can hint at which learning problems reflect addiction in general and which reflect the drug's effects.
Why does it matter that reversal cost linked to intensity rather than severity?
If reversal cost tracked lifetime gambling, it might reflect accumulated practice or brain adaptation; linking it to current monthly intensity suggests it is not a cumulative effect of gambling.
What does a smaller FRN in gamblers suggest?
Since the FRN reflects evaluating outcomes as good or bad, a reduced FRN suggests gamblers may weigh negative feedback less, which could make it harder to learn from losses.
More on Addiction