Two decades after enrolling more than 2,800 older Americans in a randomized trial of brain exercises, researchers have finally answered a question the trial was never designed to answer on its own: does cognitive training change whether someone is later diagnosed with dementia in real-world medical records. Publishing in Alzheimer's & Dementia: Translational Research and Clinical Interventions, a team led by Norma B. Coe of the University of Pennsylvania linked participants from the original ACTIVE trial to Medicare claims running from 1999 to 2019, restricting the analysis to the 2,021 participants enrolled in traditional Medicare at baseline. Dementia diagnoses were identified using the federal Chronic Conditions Warehouse claims algorithm rather than clinical exams, meaning the outcome reflects what a doctor billed for, not a research diagnosis. The authors write that their own search of the literature found no prior study with a 20-year follow-up asking whether cognitive training changed the rate of a claims-based ADRD diagnosis, making this the longest such tracking to date. The ACTIVE trial originally randomized participants to one of three ten-session training arms, targeting verbal memory, inductive reasoning, or speed of visual processing, or to a no-contact control group; some participants later received booster sessions. Over the full 20-year follow-up, neither memory training nor reasoning training showed a statistically significant effect on the risk of a later ADRD diagnosis compared with the control group. The one significant finding was narrower: participants originally assigned to speed-of-processing training who went on to complete one or more booster sessions had a lower hazard of diagnosed ADRD, with a hazard ratio of 0.75 (95 percent confidence interval, 0.59 to 0.95). Speed-trained participants who did not receive a booster session showed no such reduction, with a hazard ratio of 1.01 (95 percent confidence interval, 0.81 to 1.27). Because booster attendance was not itself randomly assigned, this specific comparison is observational within an otherwise randomized trial, and the authors and outside researchers covering the study have noted that people who chose to return for boosters may have differed from those who did not in ways the data cannot fully account for.
Twenty-Year Medicare Tracking Finds No Broad Dementia Benefit From Cognitive Training, but a Signal in One Subgroup
A new analysis links the original ACTIVE trial, run in the late 1990s, to two decades of Medicare claims. Memory and reasoning training showed no effect on later dementia diagnosis; speed-of-processing training showed a reduced rate only among those who returned for booster sessions.
The National Institutes of Health, which funded the original trial, said in its own release on the findings that the strategies involved engaged mostly unconscious rather than deliberate thinking during the original training sessions. The authors describe the result as evidence that a specific, brief, computer-based training regimen, reinforced with booster sessions, is associated with a lower rate of later diagnosed dementia in this trial population; they do not describe it as evidence that cognitive training of any kind, or any commercial brain-training product, changes a person's risk of developing dementia.