Still no evidence brain training protects us against dementia


“Brain training computer game ‘reduced risk of dementia by almost a third’,” claims the Daily Mirror, arguably over-stating the importance of a US study’s results.

“Brain training computer game ‘reduced risk of dementia by almost a third’,” claims the Daily Mirror, arguably over-stating the importance of a US study’s findings.

“Brain training” refers to apps designed to stimulate cognitive (mental) functions, such as memory. Supporters of brain training have reported that the activity may play a role in dementia prevention, but hard evidence has been lacking.

Researchers looked at a group of people who had participated in a trial of three types of memory training game 10 years earlier. At the time, the study was designed to look at whether training improved aspects of memory and daily function, but the researchers wanted to see whether anyone in the study had gone on to develop dementia.

Overall the findings were not that convincing. There was a suggestion that people who had done a speed-based training game were less likely than the control group to have dementia, but this was right on the threshold of statistical significance, meaning we can't be sure of the finding.

Training games based on memory or reasoning had no effect on risk of dementia.

Additionally, these weren't valid clinical diagnoses of dementia, but rather self-reports of low cognitive scores.

From what we know about dementia, it would appear you should first focus on your heart health, then worry about your brain. Regular exercise, a healthy diet, quitting smoking if you smoke, maintaining a healthy weight, and drinking alcohol in moderation, may all help reduce your risk of dementia, as well as many other chronic diseases.

Where did the story come from?

The study was carried out by researchers in the US from the University of South Florida, Indiana University, Moderna Therapeutics, and Pennsylvania State University.

The original trial was funded by the National Institute of Nursing Research, the National Institute on Aging, the Indiana Alzheimer Disease Center, and the Cognitive and Aerobic Resilience for the Brain Trial. It was published in the peer-reviewed medical journal Alzheimer's & Dementia: Translational Research & Clinical Interventions.

Coverage was overly-optimistic in most cases. Out of the UK media, only The Guardian gave a good critical analysis of the research and carefully considered its limitations. Most of the other papers’ headlines suggested the results were more definitive than they actually were.

What kind of research was this?

This study carried out further analysis of an existing randomised controlled trial that looked at the effect of several computer-based memory games on the cognitive function of older adults (aged 65 or above). Here, the researchers followed up the remaining participants at 10 years after the initial study to see whether or not any of them had developed dementia.

Although a randomised controlled trial is the best way of testing an intervention like this, and it is a good idea to look at how people are at a later date, the initial study was not specifically designed to look at development of dementia. Also, it didn't look at validated clinical diagnoses of dementia. Some cases were assumed to be dementia based on below-average cognitive test scores, and others by the participants or their families telling the researchers that a diagnosis had been received.

What did the research involve?

The original study involved 2,785 people aged 65 or over who did not have dementia or significant cognitive impairment at the start of the study. People with visual impairment, stroke, certain cancers or communication problems were also excluded.

In the original trial, people were randomised to one of three computer-based memory training games, or to no training.

The three training programmes were:

  • memory training
  • reasoning training
  • speed training

Each one lasted for five to six weeks, during which time people had 10 sessions lasting 60-75 minutes. People who completed at least 80% of their initial training sessions were then offered booster sessions.

In the follow-up study, researchers looked to see who had gone on to develop dementia.

They defined this in one of three ways:

  • memory or day-to-day activity scores below average
  • a score of less than 22 points on the Mini-Mental State Examination (MMSE) test (a score of 25-30 is normal; 20 to 24 is usually interpreted as mild cognitive impairment)
  • the participant or their family informing the researchers that the participant had been diagnosed with dementia or Alzheimer’s disease during the follow-up

What were the basic results?

At 10 years the researchers found:

  • 1,220 people were still alive and able to participate in the follow-up study. This is 44% of the original sample; in the intervening time, 627 people had died and 938 were lost from the study for other reasons.
  • Across the whole study, 260 people were recorded as having “dementia” – about 1 in 5.
  • There was a suggestion that people who had done speed training were less likely to develop dementia than those in the control group (59 people in the speed training group, 75 in the control group, hazard ratio 0.71, 95% confidence interval 0.50 to 0.998). However, this is right on the threshold of statistical significance so this may just be a chance finding.
  • People who had done the memory or reasoning training were no more or less likely to have dementia than people in the control group.

How did the researchers interpret the results?

The researchers noted that the extent to which the training protected against dementia appeared to be influenced by how many sessions people had done. They did, however, acknowledge that the 5-year follow-up of this trial had found no benefit of any of the training types.


It has often been debated whether keeping mentally active or doing puzzles or activities that stimulate thinking and memory could help to prevent cognitive decline. It's an interesting idea, particularly as there is no established way to prevent Alzheimer's, the most common form of dementia.

However, this study brings us no further forward and fails to provide convincing evidence this kind of brain training is effective at preventing dementia.

There are several important limitations:

  • These were not valid clinical diagnoses of dementia. There's a high possibility that simply relying on self-reports of dementia or test scores below average could lead to errors and some people being incorrectly classified as having or not having it. For example, simply having a mini-mental score below 22/30 does not in any way relate to a confirmed diagnosis of dementia.
  • Although there was some suggestion that speed training could be beneficial, this was on the margin of statistical significance. In fact many statisticians would regard a confidence interval bordering on 0.998 (in other words 1.00) as no link at all. Therefore unless other studies can add further weight to this finding, this doesn't provide good evidence that speed training is helpful.
  • More than half the people included in the original study were not available at 10 years for follow up, either because of death or study drop-out. We do not know whether they had dementia or not and this would have affected the results.

It is good for all of us to keep our minds and bodies active as we age. However, we do not have enough evidence yet to say definitively whether brain training will protect us from dementia.

There are, however, established ways of protecting against cardiovascular disease, which in turn can protect against vascular dementia, even if the links with Alzheimer’s aren’t so clear. These include taking regular exercise, eating a healthy balanced diet, not smoking and limiting alcohol.

Article Metadata Date Published: Fri, 17 Nov 2017
Author: Zana Technologies GmbH
NHS Choices