A four-week cognitive-training app trial reported improvements in selected memory and decision-task outcomes without a significant advantage in drug-related attention bias. The intervention did not demonstrate improvement in every domain it was designed to address.
Zhu and colleagues assigned 40 men with methamphetamine use disorder to computerized cognitive addiction therapy plus usual rehabilitation or usual rehabilitation alone. The app combined cognitive exercises with methamphetamine-related pictures. Participants received 20 sessions over four weeks and completed laboratory assessments before and after the intervention. Cognitive training, rather than methamphetamine administration, was the experimental intervention.
The study reported improvements in memory-related measures and significant group-by-time effects on delay discounting, Iowa Gambling and Balloon Analog Risk tasks. These outcomes address different processes, including remembering information, valuing delayed rewards and choosing under uncertainty. A change on one task does not establish an equal change across all forms of executive control.
The group-by-time interaction for attention bias was nonsignificant (p = 0.34), although there was an overall time effect. Thus, the evidence did not show that adding the app reduced attention bias more than usual treatment. Social-emotional cognition also did not significantly improve within the app group. These findings qualify a broad claim that the app corrected cognition or eliminated cue bias.
The comparison group did not receive an equivalent active computerized training program. The study therefore could not isolate every contribution of practice, engagement or additional contact. The small, all-male rehabilitation sample limits generalization, and the authors noted that they could not disentangle the role of the drug-related pictures added to training. The reported cognitive changes did not establish durable improvement in everyday function or reduced methamphetamine relapse. No drug-concentration threshold or impairment rule follows from these task results. A defensible treatment summary names both the domains that changed and the intended target that did not.
See also why cognitive flexibility is not a single outcome.