Brain Games Are Being Sold on a Feeling. The Evidence Is Narrower Than That.

Most brain games are sold on a feeling. You finish a puzzle, you feel sharp, and you assume something good happened upstairs. The research is less cozy. Two pieces of it point at one kind of game, the kind that makes you react fast. They also show how easily a promising result gets blown up into a promise.

What the ACTIVE follow-up suggests

The most talked-about claim comes from ACTIVE, an NIH-funded trial of nearly 3,000 older adults. A 20-year follow-up, as described in a social media post, says people who did speed training saw a large drop in Alzheimer's and other dementias. Speed training means exercises for visual processing and decision-making. You see something briefly, and you have to respond quickly and correctly.

The same account says memory and reasoning exercises did not give lasting protection. If that holds up, it matters. It would mean a brain game is not just a brain game. Getting better at one narrow puzzle may not carry over to anything a study can measure in real life.

I am leaving out the specific risk-reduction and training-hours figures the post quotes. The only source for them is that social post. The primary trial paper is not in my sources, and the CNN and AARP pages I had contained only headlines and navigation. Until someone checks the trial report itself, treat the whole ACTIVE story as a lead, not a result. A social post is a thin base for a claim this big.

What the McGill study measured

The second piece is a McGill-led trial published October 29, 2025. It enrolled 92 healthy adults aged 65 and older. According to the ScienceDaily summary, 10 weeks of BrainHQ, a speed-based game app, brought cholinergic function back to levels typical of someone about 10 years younger. The cholinergic system is a set of brain cells that use the chemical acetylcholine to pass signals. It is tied to learning and memory. When it works well, the brain is better at picking up and holding on to new information.

Two details make this study better than average. The comparison group played recreational computer games, so the study separated structured training from simply sitting at a screen and staying busy. And the measurement was real: a specialized PET scan and tracer that only a few centres worldwide can run.

Still, the trial is small, short, and tracks one biological marker. It shows a plausible mechanism. It does not show anyone avoided dementia. The researcher also said crosswords and reading may not be enough, since not all activities promote neuroplasticity. That is a reasonable guess, and it fits the ACTIVE pattern, but the trial did not test crosswords or reading. Read it as interpretation.

Where the claims go too far

Some coverage runs well ahead of the data. The social post calls the results astonishing and the benefits permanent and durable. The ScienceDaily headline says brain training reverses 10 years of aging, and its summary says it proved it can make aging brains 10 years younger. The trial measured one chemical system in 92 people over 10 weeks. Nothing in it establishes permanence.

The McGill study says its findings help explain dementia-risk results. That is a statement about a possible mechanism. A changed biomarker and a lower rate of disease are different things, and only the second is what most of us care about.

The post also credits implicit learning and cognitive reserve for the lasting effect. Those are the poster's explanations. Neither study established them. They may be worth chasing, but they are not results.

Questions to ask before trying a brain-training program

You do not need a lab to be a sharp shopper. Four questions cover most of it.

• Does it include speed-based tasks? That is the type linked to lasting benefit in the ACTIVE account. Memory drills and general quizzes are not the same thing.
• Was the benefit measured in healthy adults, and for how long? McGill covered 10 weeks. A promise of decades goes far beyond that.
• What was the control group? McGill used an active group playing recreational games. A study against no activity cannot separate a specific program from general engagement.
• Is it being sold as a replacement for care? The researchers floated brain training as a safer alternative to medication or a partner to it. That is a suggestion, not a tested conclusion. Talk to a clinician before changing anything medical.

The researchers did note that BrainHQ is commercially available and that clinicians can discuss it with patients who want to maintain or improve brain health. That is a fair, modest framing.

What remains unknown

A lot. My sources give no pricing, subscription, or device details for BrainHQ or any other app, so I cannot tell you what this costs. There is no independent replication of the McGill result and no follow-up past 10 weeks. The ACTIVE claims also still need to be checked against the trial's own published paper.

The FTC has a case file on Lumos Labs, the maker of Lumosity, but the page I had held only site navigation. I cannot say what it found. It is a reminder that marketing for these products has drawn regulators before.

Nothing here tells us whether short daily puzzles or trivia, the kind found on everyday brain-teaser sites, do anything like what speed training did. Safety, privacy, and data handling for these apps are also missing from the record. That deserves a look before you hand over your performance data.

So here is my plain read. Speed-based training looks like the most promising format so far, but the evidence is early and one of its two pillars is unverified. A daily puzzle may still be a pleasant way to spend twenty minutes. The studies here do not show it does the same job.