February 10, 2026
The Times
Rhys Blakely

Exercises to sharpen the brain’s processing speed could reduce the likelihood of a diagnosis by as much as 25 per cent, according to an American study.

Scientists claim to have found the first evidence that a targeted form of brain training can significantly reduce the risk of Alzheimer’s disease.

The findings, from a large American trial that tracked older adults for two decades, suggest that exercises designed to sharpen the brain’s processing speed may cut the likelihood of a dementia diagnosis by as much as 25 per cent.

The results raise the prospect that mental workouts, even when started in later life, could delay the disease by years.

Dementia is one of the most pressing health challenges. About 900,000 people in the UK are living with the condition, with Alzheimer’s the most common type. There is no cure. Recently developed drugs can, at best, slow decline modestly in some patients.

There is evidence, however, that a healthy lifestyle can help. Regular exercise, controlling blood pressure and diabetes, avoiding smoking and staying socially active are all associated with a lower risk.

The new results, published in the journal Alzheimer’s & Dementia: Translational Research & Clinical Interventions, come from the so-called Active study. In the late 1990s, 2,802 healthy adults aged 65 and over were randomly assigned to one of four groups where they got either speed training for the brain, memory training, reasoning training or no training.

Participants attended ten sessions, lasting about an hour, over five to six weeks. Some of those who completed most of the sessions were offered additional “booster” sessions 11 months later and again after 35 months. In total, it amounted to between ten and 22½ hours of training spread over three years.

Memory training taught mnemonic techniques. Reasoning training focused on spotting patterns and solving problems. Speed training, delivered via a computer screen, required participants to identify and locate visual information.

In one game, Double Decision, participants were shown a car and a road sign in a landscape. They appeared for a very short time and then the player had to say where on the screen they were. As their performance improved, the tasks became progressively harder.

The researchers analysed healthcare records to track who went on to be diagnosed with Alzheimer’s disease or related dementias. Nearly half of the control group received a dementia diagnosis over the 20-year follow-up period. The rate was essentially the same for those who received memory or reasoning training.

The story was different for those who completed the speed training and also received the top-up sessions in the years that followed. Compared with the control group, their risk of a dementia diagnosis was 25 per cent lower.

The findings must be interpreted with caution, the researchers have said. The participants who appeared to benefit were those who completed the initial training and went on to attend additional “boosters”. It is possible that there was already something distinctive about this group — possibly more motivation, or better underlying health — that made them more likely both to persevere with the exercises and to be less vulnerable to dementia in the first place.

The researchers attempted to address this potential “healthy user bias” by randomly allocating booster sessions among those who had completed the initial course, and by comparing those assigned to receive boosters with those who were eligible but not selected. They also adjusted their analysis for age, education, baseline cognitive performance and a range of health risk factors. Even so, no statistical method can entirely eliminate the possibility that the most engaged or healthiest participants were also the most likely to benefit.

Henry Mahncke, the chief executive of Posit Science, the company behind the training software used in the study, said: “This is an astonishing result, and it comes at a critical time. These results change what’s possible — now — for better brain health and Alzheimer’s prevention.”

If the findings are borne out, the implications would be considerable. Delaying dementia by even a year or two across the population could ease pressure on families, the NHS and social care.

The academic authors of the study, which included researchers from Johns Hopkins School of Medicine, the University of Pennsylvania and the University of Washington, were more cautious, concluding that “cognitive training involving speed of cognitive processing has the potential to delay the diagnosis” of Alzheimer’s disease and related dementias.