A gambling-task study found weaker links between reward information and momentary happiness ratings in men with methamphetamine use disorder. Its computational results did not establish why any individual used drugs or prove that mood dysregulation caused later use.
The researchers compared 27 men with methamphetamine use disorder with 49 male controls. Participants chose between certain and uncertain rewards and rated their current happiness every few trials. Models estimated how expected value and reward prediction errors contributed to those ratings. This was an observational human group comparison, not methamphetamine administration or a mood-treatment experiment.
The clinical group made more choices classified as irrational under the task’s criteria. Model estimates also indicated weaker influence of reward information on reported mood, with selected associations related to substance-use severity. Expected value and prediction error were calculated task variables; they were not direct measurements of dopamine or proof of a person’s conscious reasoning.
The sample was small and all male, and model conclusions depend on the candidate structures and assumptions tested. A momentary happiness rating during monetary decisions is different from diagnosing depression, measuring all emotional regulation or forecasting relapse. Cross-sectional severity associations cannot establish that the modeled coupling produced more substance use. The study supplies a useful hypothesis about how reward and feeling may be linked differently, while clinical mechanisms and treatment implications require prospective testing. Its strongest contribution is connecting decisions with repeated subjective ratings rather than treating every reward-task outcome as the same phenomenon.
See reward learning and effort as separable laboratory outcomes.