A Surprising Brain Training Discovery: 25% Lower Dementia Risk After 20 Years

A groundbreaking study of over 2,800 older adults found that those who completed a specific type of brain training had a 25% lower risk of dementia diagnosis two decades later. The research, published in February 2025 in Alzheimer's & Dementia, represents the first large-scale randomized controlled trial—the gold standard in medical research—to show that cognitive training can meaningfully protect against dementia.

What Type of Brain Training Actually Works?

The key finding surprised even the researchers: it wasn't memory or reasoning exercises that made the difference, but "speed training." Participants in the study completed a computer-based task called Double Decision, which tests how quickly and accurately people can process visual information. In the game, a car and road sign briefly appear on screen before disappearing, and players must recall which car appeared and where the sign was located. The task adapts to become harder as performance improves.

"The thing that's so astonishing, it's really a very modest amount of training," said Marilyn Albert, a professor of neurology and director of the Johns Hopkins Alzheimer's Disease Research Center, who led the study. "Who would have thought?"

How Much Training Is Needed to See Results?

The study design was straightforward but rigorous. Researchers randomly assigned cognitively healthy people aged 65 and older to one of three types of brain training—speed, memory, or reasoning—or to a control group that received no training. Participants completed two 60- to 75-minute sessions per week for five to six weeks, totaling roughly 14 to 22 hours of training.

Here's where the results get interesting: about half of the participants in each group were randomly assigned to receive "booster" sessions. These boosters consisted of four additional one-hour sessions at the end of the first year and another four sessions at the three-year mark. Only the speed training group that received these booster sessions showed a significant reduction in dementia risk.

• Speed Training with Boosters: 25% lower risk of dementia diagnosis compared to control group
• Speed Training Without Boosters: No significant change in dementia risk
• Memory or Reasoning Training: No significant change in dementia risk, regardless of booster sessions

Why Might Speed Training Protect the Brain?

The exact mechanism remains unclear, but researchers have several theories. One possibility involves "implicit learning"—the kind of learning that happens without conscious awareness. "We know that changes that occur from this kind of learning are very long-lasting," Albert explained. Speed training demands intense focus and attention, and the difficulty increases as performance improves, creating a challenging cognitive workout.

Another theory suggests that speed training builds "brain reserve," a kind of cognitive buffer against damage. If cognitive decline occurs once brain damage reaches a certain threshold, then a brain with more neurons and connections would resist symptoms longer. Altered brain connectivity from speed training might also help people divide their attention more effectively, making it easier to navigate daily life as they age. This could reduce isolation, encourage greater activity, and increase social engagement—all factors known to support long-term brain health.

Should You Rush to Buy Brain Training Software?

Not necessarily. Brain training has a complicated history. Early enthusiasm for commercial brain-training programs waned after several companies were found to be overstating their benefits. In 2014, nearly 70 scientists signed an open letter stating there was no conclusive evidence that brain training produced real-world benefits. However, more than 100 scientists later countered those arguments.

Even the researchers behind this new study caution against rushing to purchase brain-training programs. "The take-home message is not that everyone should go into their windowless basement and start doing speed training games on their computer," said Andrew Budson, a neurologist at Boston University. Instead, experts suggest that activities using implicit learning—such as learning a new sport, vocation, or craft—may produce similar long-lasting beneficial changes in the brain.

What Do Experts Say About These Findings?

The study has generated cautious optimism among dementia researchers. "The analysis seems rigorous," said Torkel Klingberg at the Karolinska Institute in Stockholm, Sweden. "It is impressive to have a 20-year follow-up, and reducing the risk score for dementia is an impressive and important result".

However, some experts urge caution. The study measured many outcomes over the 20-year period, and researchers examined multiple statistical tests. As one expert noted, the more outcomes examined, the greater the likelihood that at least one result will appear meaningful by chance alone. Additionally, dementia diagnoses in the study were based on Medicare records rather than biological markers like amyloid plaques or tau proteins in the brain—the actual hallmarks of Alzheimer's disease.

"I'm both skeptical but also somewhat impressed," said Dr. Andrew Budson. "It's really quite rare for any computerized-game brain training study to show positive results—and to show effects on a diagnosis of Alzheimer's disease 20 years later".

The bottom line: while this study offers the most compelling evidence to date that cognitive training can help protect against dementia, researchers agree that more work is needed to confirm these findings and understand exactly why speed training appears beneficial. In the meantime, staying mentally engaged through learning new skills and remaining cognitively active remains one of the most evidence-based strategies for supporting brain health as you age.