Do Brain Games Prevent Dementia? What the Evidence Shows

Brain games are everywhere, and many promise to keep your mind young. If you care for someone with dementia, or worry about your own memory, it is natural to wonder whether playing them could help. Here is what the research actually shows, including the parts that are encouraging and the parts that are not.

This is general information, not medical advice. See our medical disclaimer.

Do brain games prevent dementia?

No brain game has been proven to prevent dementia. The National Institute on Aging (NIA) advises people to be wary of claims that computer and online games can improve memory, and says there is not enough evidence that any commercial brain-training app has the same effect as the training tested in clinical trials (NIA on preventing Alzheimer’s).

That does not make games useless. It means their value is in enjoyment, company and routine, not in protecting anyone from dementia.

What is “near transfer”, and why does it matter?

When you practice a task, you usually get better at that task. Researchers call this a training effect. The harder question is whether the improvement transfers to other things: to similar tasks (near transfer) or to everyday life, like remembering appointments or managing money (far transfer).

A major 2016 review examined the studies that brain-training companies cited as their best evidence. It found extensive evidence that people improve on the trained tasks, less evidence that they improve on closely related tasks, and little evidence that training improves everyday thinking. Many of the studies also had design problems that made firm conclusions impossible (Simons and colleagues, Psychological Science in the Public Interest).

What do reviews say about computer brain training?

Cochrane reviews, which pool the best available trials, have looked at computerized training lasting at least 12 weeks:

• Healthy adults 65 and older: eight trials with 1,183 people. The evidence was low or very low quality. There may be a slight benefit right after training, with no clear evidence that it lasts (Cochrane review, healthy older adults).
• People with mild cognitive impairment: eight trials with 660 people. No trial reported whether people went on to develop dementia, and the evidence was too weak to draw conclusions (Cochrane review, mild cognitive impairment).

What did the ACTIVE trial find?

ACTIVE is a large US trial that began in 1998 with 2,802 adults aged 65 and older who lived independently and did not have significant memory or thinking problems. People were randomly assigned to one of three types of small-group training (speed of processing, memory, or reasoning) or to a no-contact group. Training was up to ten sessions over five to six weeks. Some people who finished most sessions were also randomly offered booster sessions about one and three years later.

The speed training involved spotting and processing visual information in shorter and shorter glimpses, at a level that adjusted as people improved.

In 2026, researchers linked 2,021 participants to their Medicare records for 20 years (ACTIVE 20-year analysis). Here is exactly what they found:

• Speed training plus at least one booster session: a lower rate of diagnosed dementia than the no-contact group (a hazard ratio of 0.75, which roughly means a 25% lower rate).
• Speed training without boosters: no lower rate.
• Memory or reasoning training: no significant effect on dementia diagnoses.
• Each training group as a whole compared with the no-contact group: a tendency toward lower rates, but not statistically significant.

The limits of this finding

• It applies to one specific kind of supervised speed training with booster sessions, not to brain games in general.
• The significant result came from a subgroup (people who finished most of the first sessions and were offered boosters), not from the whole speed group.
• Dementia was measured using diagnosis codes in Medicare claims, not by examining each person. About a quarter of the original participants were left out because they were enrolled in Medicare Advantage, which lacks full claims records.
• Participants had no significant thinking problems when they started. The results say nothing about games for people already living with dementia.

It is an interesting result that deserves more research. It is not evidence that any app prevents dementia.

What did the FTC say about Lumosity?

In January 2016, Lumos Labs, the maker of Lumosity, agreed to pay $2 million to settle Federal Trade Commission charges of deceptive advertising. The FTC said the company claimed, without adequate science, that its games could improve performance at work and school and could delay age-related decline, mild cognitive impairment, dementia and Alzheimer’s. The order requires the company to have competent and reliable scientific evidence before making such claims in the future (FTC press release).

The lesson for families: be skeptical of any game, app or program that promises to protect memory or prevent dementia.

Can games help someone who already has dementia?

They can add to a good day, which matters a great deal. The best-studied approach is cognitive stimulation: enjoyable group activities such as discussion, reminiscence, word games, puzzles and music, usually in a small group led by a facilitator.

A 2023 Cochrane review of 37 trials with 2,766 people with mild to moderate dementia found (Cochrane review of cognitive stimulation):

• a small, short-term benefit on thinking tests (moderate-quality evidence);
• clinically relevant improvements in communication and social interaction (high-quality evidence, from a smaller number of studies);
• slight benefits for quality of life, mood and behavior.

Benefits were measured mostly right after programs of about ten weeks, and few studies checked whether they lasted. Cognitive stimulation is a structured group program, so these results do not automatically apply to a person playing a game alone on a tablet.

The 2024 Lancet Commission also says activity programs are important for people living with dementia, to maintain enjoyment and purpose (Lancet Commission).

What are games actually good for?

When chosen well, games can offer:

• Enjoyment. A pleasant half hour is worth having for its own sake.
• Connection. Playing side by side gives you something easy to share when conversation is hard. See how to play together.
• Routine. A familiar activity at the same time each day can bring comfort.
• Confidence. Activities matched to what a person can still do well let them feel capable, not tested.
• Conversation. A picture or a familiar saying can open a story. Try conversation cards or finish the saying.

What to avoid: timers, scores, streaks and anything that feels like a memory test. If the person gets frustrated, stop. Ending early, or simply sitting together, is a good outcome. Our guide to choosing apps and games has a checklist.

Where Golden Moments fits

Golden Moments games, such as picture pairs, picture puzzle and large-print word search, are made for enjoyment and shared time. They have no clock and no score. They are not a treatment, and they do not prevent or slow dementia. We would rather tell you that plainly than promise something no game can deliver.

Talk with a doctor

If you are worried about memory changes in yourself or someone you care for, see a doctor rather than relying on an app. A doctor can look for treatable causes and talk through the steps with the best evidence, such as managing blood pressure and checking hearing. See Can dementia be prevented? for more. For sudden confusion or other sudden changes, get urgent medical help.