We’ve become good at keeping people alive into old age. We remain bad at keeping their minds intact. The newest drugs slow decline modestly, and lifestyle programs have yet to show they prevent actual dementia diagnosis. Against this backdrop, a study published in early 2026 stands out: a specific form of cognitive training, totalling about 22 hours and delivered two decades ago, reduced the risk of dementia diagnosis by 25% (ref 1).

The Study

The story starts in the late 1990s with the ACTIVE trial — a large randomized controlled trial that enrolled 2,802 adults aged 65 and older (ref 2).Participants were assigned to one of four groups: training in memory strategies, reasoning strategies, speed of processing, or no training at all. The initial program ran for about 10–12 hours over six weeks. Some participants then received booster sessions at 11 and 35 months, bringing the total to roughly 22 hours.

Twenty years later, Coe and colleagues linked these participants to US Medicare claims data — documenting every dementia diagnosis recorded by any physician, independently of the study team.. The result: speed-trained participants who also completed the booster sessions had a 25% lower risk of being diagnosed with Alzheimer’s disease or related dementias. Speed-trained participants without boosters showed no significant benefit. Neither memory nor reasoning training reached statistical significance.

Why this training worked

A key question is why the speed-training, but not the reasoning or episodic memory training worked. The memory and reasoning programs taught explicit strategies — tricks for remembering, rules for solving patterns. In the speed training, participants sat at a computer, identified objects flashing onscreen, and tracked peripheral targets simultaneously with no instructions. It is largely an implicit training of attention, not teaching explicit strategies.

The speed-training also differed in that it was adaptive: the computer raised the difficulty as participants improved. A meta-analysis has shown that adaptivity is a key ingredient in training (ref 3). The focus on cognitive function (attention), adaptivity of difficulty and length of training might all be important ingredients for the outcome.

The brain-training debate

In 2014, a group of scientists organized by the Stanford Center on Longevity warned that “brain training” – in general – was ineffective, thus discouraging further research in the field (ref 4). Dismissing an entire field in general is questionable scientific logic, which was pointed out by another group of researchers, including me (ref 5). Luckily, Coe and her colleagues were not discouraged. The field of cognitive training is still emerging and we need more of the kind of large-scale studies like the ACTIVE study.

Twenty-two hours of focused mental effort, spread over three years, to cut one’s dementia risk by a quarter. That is a remarkably good deal — and are minder that the brain, like the body, responds to training.

‍

1. Coe NB, Miller KEM, Sun C, et al. Impact of cognitive training on claims-based diagnosed dementia over 20 years: evidence from the ACTIVE study. Alzheimer’s & Dementia:TRC&I. 2026;12:e70197.

2. Ball K, Berch DB, Helmers KF, et al. Effects of cognitive training interventions with older adults: a randomized controlled trial. JAMA. 2002;288:2271–2281.

3. Edwards JD, Fausto BA, Tetlow AM, et al. Systematic review and meta-analysis of useful field of view cognitive training. Neurosci Biobehav Rev. 2018;84:72–91.

4. Simons DJ, Boot WR, Charness N, et al. Do “brain-training” programs work? Psychol SciPublic Interest. 2016;17:103–186.

5. Cognitive Training Data Response Letter