Brain games and dementia: most don’t work, but one very specific kind does

Brain games and dementia: most don't work the way people hope, but one very specific kind does

A landmark trial found one narrow, specific type of cognitive training cut dementia risk by roughly 29% over 10 years, while most commercial brain-training apps show no such effect.

A landmark trial found one narrow, specific type of cognitive training cut dementia risk by roughly 29% over 10 years, while most commercial brain-training apps show no such effect.

Time to effect

Years (long-term risk reduction)

Years (long-term risk reduction)

Core practice

Complete a structured speed-of-processing training program (commercially available as BrainHQ’s Double Decision exercise, the specific task used in the ACTIVE trial), including booster sessions, rather than generic multi-domain brain-training apps

Complete a structured speed-of-processing training program (commercially available as BrainHQ’s Double Decision exercise, the specific task used in the ACTIVE trial), including booster sessions, rather than generic multi-domain brain-training apps

▪ The challenge at hand

Commercial brain-training apps promise broad cognitive benefits and dementia prevention, and the honest research picture for that general category is disappointing: meta-analyses consistently find these programs improve the specific skill practiced in the game, without meaningfully transferring to general cognition, everyday function, or dementia risk itself. A major review of cognitive training trials rated the evidence for actual dementia prevention as insufficient.

One narrow exception genuinely stands apart from this pattern. A specific type of training called speed-of-processing training, a repetitive visual task that trains how quickly you can identify and locate objects in your field of view, showed a real, replicated effect in a large randomized trial: participants who completed this training had a 29% lower rate of dementia diagnosis 10 years later, an effect that held up at 20-year follow-up. This is not the same as generic 'brain games,' and the distinction matters enormously for anyone deciding where to spend their time.

▪ What it is

This is speed-of-processing training, a specific, narrow visual-processing exercise (commercially available as BrainHQ's Double Decision), distinct from generic multi-domain brain-training or memory apps, with genuine dementia-risk-reduction evidence behind it.

Why this is surprising

Commercial brain-training apps promise broad dementia prevention, but meta-analyses consistently find they improve only the specific skill practiced, without transferring to general cognition or dementia risk, a major review rated the dementia-prevention evidence for cognitive training overall as insufficient. The genuine, non-obvious exception: one narrow, specific type, speed-of-processing training, showed a real, replicated 29% lower dementia diagnosis rate at 10-year follow-up in a large randomized trial, holding up even at 20 years. Most brain games aren't this; this one specific kind is.

▪ How it works

One narrow skill, a surprisingly durable effect.

Speed-of-processing training repeatedly exercises the ability to quickly identify and localize visual information across an increasingly wide and cluttered field of view, under time pressure that adapts to your performance. Unlike memory or puzzle-style brain games, this specific training targets a foundational cognitive process, visual processing speed, tied to broader everyday function like safe driving, and appears to produce more durable, generalizable benefit than narrower recall or pattern-matching exercises.

▪ The research

What the evidence says

The ACTIVE trial, a large NIH-funded randomized controlled trial of nearly 3,000 older adults, found that participants who completed speed-of-processing training, particularly those who also received booster sessions, had a 29% lower rate of dementia diagnosis at 10-year follow-up compared with a no-contact control group, with the effect persisting at 20-year follow-up. In contrast, a separate meta-analysis of commercial computerized cognitive games found improvements limited to trained skills (processing speed, working memory, executive function) with no established link to dementia prevention specifically.

Edwards JD et al. Alzheimers Dement (N Y). 2017;3(4):603-611. (ACTIVE trial 10-year follow-up.) Also: Ball K et al., JAMA. 2002;288(18):2271-81, original ACTIVE trial.

WE'VE COACHED THOUSANDS OF USERS WITH THEIR BRAIN FOG

WE'VE COACHED THOUSANDS OF USERS WITH THEIR BRAIN FOG

Brain games and dementia, in practice

Brain games and dementia, in practice

Brain games and dementia, in practice

Cognitive improvements are genuinely hard to notice in yourself without a reference point. Here's how it played out for people actually tracking it.

Cognitive improvements are genuinely hard to notice in yourself without a reference point. Here's how it played out for people actually tracking it.

Cognitive improvements are genuinely hard to notice in yourself without a reference point. Here's how it played out for people actually tracking it.

349

349

started

63%

63%

completed

43%

43%

noticed a change

14%

14%

made it routine

Self-reported by Coco users. Not a clinical outcome.

Self-reported by Coco users. Not a clinical outcome.

Data across the Coco Health user base, not a clinical outcome.

Coco is the AI health coach that runs experiments like this one with you

Know exactly what to do: Coco sets the protocol and checks in by call or message

See what's actually changing: Coco tracks your symptoms and synthesizes the trend

Get a real answer: Coco tells you whether the data supports continuing or stopping

▪ What to expect over time

The trial's dementia-risk benefit was measured at 10 and 20 years, this is a long-horizon investment, not something with a short-term cognitive readout, though processing speed itself improves within the training period.

Side effects

None. Requires sustained engagement, benefit in the trial was strongest with completed booster sessions.

Who should be cautious

None. This is a complement to, not a substitute for, the other well-established dementia-risk-reduction levers (exercise, blood pressure control, hearing correction, social engagement).

FAQ

Can I use any brain-training app and get this benefit?

Do I need to do booster sessions, or is the initial training enough?

Is Coco a replacement for my doctor?

Coco helps you turn health ideas like this into small, trackable experiments you can actually stick with.

The hard part isn't starting — it's knowing if it's working

Stay consistent: Coco checks in so you don't have to rely on motivation

See clearly: Coco reads your symptom data so you can trust what you're seeing

Get a real answer: Coco tells you whether it's working, even if it isn't

Educational only. This is not medical advice. Always talk with a qualified clinician before changing medications, supplements, or care plans.