A 20-year federally funded study of 2,802 adults found that one specific type of online brain game reduces dementia risk by 25%.
It’s not crosswords, Sudoku or memory puzzles
Most brain games are a waste of time if your goal is to prevent dementia. Crossword apps, word searches, memory matching games might keep you entertained, but a major 20-year study just confirmed that none of these activities reduce your risk of Alzheimer’s disease or related dementias.
One type of game does.
It’s called speed-of-processing training, and it works by forcing your brain to identify objects on a screen in fractions of a second while tracking other targets in your peripheral vision. A federally funded study found that people who did eight to ten sessions of this specific training (plus booster sessions) were 25% less likely to be diagnosed with dementia over the following two decades.
“We now have a gold-standard study that tells us that there is something we can do to reduce our risk for dementia.”
Let that sink in. Not a drug. Not a surgical procedure. A computer game.
What the study found
The research comes from the ACTIVE study, the largest randomized controlled trial of cognitive training in older adults ever conducted in the United States. It was funded by the National Institutes of Health and began in 1998.
Researchers split 2,802 adults (average age around 74) into four groups. Three groups received different types of brain training: memory strategies, reasoning strategies or visual speed-of-processing exercises. The fourth group received no training at all.
Each intervention group got just 10 sessions of 60 to 75 minutes each, spread over five to six weeks. Some participants also received short “booster” blocks of four sessions at 11 months and 35 months afterward.
The 2026 follow-up used Medicare records to track dementia diagnoses across all groups from 1999 to 2019. The results were stark:
Only one type of training worked. And it only worked when people did the booster sessions.
“It’s super-exciting to see that these effects are still holding 20 years out.”
Why speed training works
The answer comes down to the difference between two types of learning. Memory games and reasoning puzzles rely on explicit learning: conscious strategies like mnemonics or pattern recognition. You’re learning rules and trying to apply them. This type of learning depends heavily on the brain areas that are among the first damaged by Alzheimer’s disease.
Speed-of-processing training targets something different: implicit learning. This is the same unconscious, automatic learning that lets you ride a bike, catch a ball or tie your shoelaces.
“We know that implicit learning operates differently in the brain and has more long-lasting effects. You can learn to ride a bike in about 10 hours of training. And even if you don’t practice for the next 20 years, you still have a bike-riding brain.”
A concurrent study from McGill University adds a biochemical explanation. Using PET brain imaging, researchers measured the density of a key chemical transporter linked to the attention and learning network that deteriorates in Alzheimer’s disease. After 10 weeks of speed training using the same exercises from the ACTIVE study, participants showed a 2.3% increase in cholinergic activity. That doesn’t sound like much until you learn that healthy aging reduces this marker by about 2.5% per decade. The training effectively reversed roughly a decade of age-related brain change in this system.
Speed training doesn’t just teach you a skill. It changes your brain’s chemistry.
What the game looks like
The specific exercise is based on the Useful Field of View (UFOV) paradigm, developed by researchers Karlene Ball and Daniel Roenker. The commercial version is called Double Decision, and is available on the BrainHQ website.
Here’s how it works:
A car or truck flashes in the center of your screen for a split second. Simultaneously, a road sign appears somewhere in your peripheral vision. You have to identify whether the central image was a car or a truck AND click where the road sign appeared.
“If you had all day to look at that, anyone could figure it out. But it doesn’t give you all day. It shows the image on the screen to you very quickly and then it goes away.”
Three features make this training effective where other brain games fall short:
George Kovach, 74, from Vienna, Virginia, has been doing BrainHQ speed exercises for a decade. He’s completed over 1,300 sessions.
“These things are hard, but you do get better at it,” Kovach says. “I look at it like doing sit-ups.”
How much training do you need?
The ACTIVE study gives us a concrete answer, and it’s surprisingly modest.
Earlier analyses from the ACTIVE data showed a dose-response pattern. People who completed 11 to 14 sessions had about a 48% lower dementia risk at 10 years, compared to the 29% average for the full speed-training group. More training correlated with greater protection.
An ongoing NIH-funded study called PACT (Preventing Alzheimer’s with Cognitive Training) is testing whether 45 sessions over several years offers even stronger protection. It has enrolled about 7,500 people aged 65 and older. Results are expected around 2028.
“The results with ACTIVE suggest that just 10 hours of training, with some booster sessions, can make a difference.”
What about other games?
BrainHQ’s Double Decision is the only game with direct, randomized trial evidence linking it to lower dementia rates. But other games share its core mechanics (speed, divided attention, adaptive difficulty) and may offer similar benefits, even though they haven’t been tested in long-term dementia trials.
Lumosity’s Eagle Eye game requires you to identify a bird in the center of the screen while locating a number in the periphery. It’s the closest mainstream approximation to the ACTIVE protocol for people who already have a Lumosity subscription.
CogniFit’s UFOV module explicitly targets Useful Field of View training. CogniFit has shown benefits for attention and processing speed in clinical populations, though not specifically for dementia incidence.
Fast-paced action video games (first-person shooters) force players to track peripheral threats while maintaining central focus, which maps onto the same cognitive demands. Research has shown that action game players have superior visual processing speed and reduced attentional blink.
These alternatives are reasonable based on their mechanistic similarity to the ACTIVE paradigm, but the dementia-prevention claim is an inference, not something proven by their own trials.
Off-screen alternatives
You don’t need a computer to train implicit learning and processing speed. Several physical activities engage the same basal ganglia and cerebellar pathways:
Tai Chi looks slow, but neurobiologically, it’s demanding. Practitioners memorize and execute complex motor sequences while maintaining broad spatial awareness, similar to the peripheral monitoring in UFOV tasks. A 2024 review linked regular Tai Chi practice to increased cortical thickness and improved working memory.
Dual-task walking combines physical movement with a cognitive challenge. Walk briskly while counting backward from 100 by 7s, or scan your environment for specific objects without stopping. This forces your brain to split resources between gait stability (implicit) and working memory (explicit) simultaneously.
Juggling requires visual tracking of multiple objects in motion and rapid motor responses. Even learning a simple three-ball cascade produces measurable increases in gray matter density in visual processing areas.
Konklusjon
The ACTIVE study gives us a specific, accessible and relatively low-effort intervention with 20 years of data behind it. You don’t need a prescription, a clinic visit or expensive equipment.
Note that speed training works best as part of a broader strategy. Do high-intensity aerobic workouts and follow a heart-healthy diet. Cognitive training should complement physical activity, cardiovascular health management and social engagement, not replace them.
The numbers from ACTIVE are encouraging but sober. The speed-training-plus-boosters game isn’t a cure. It’s a meaningful reduction in risk. But 25% is 25%. For an intervention that costs a few dollars a month and takes about the same time as a Netflix episode, that’s a strong return.
Gi tilbakemelding om dette