The first controlled trial worldwide demonstrating a preventive effect on dementia over 20 years was published in February 2026 in the journal Alzheimer's & Dementia: Translational Research & Clinical Interventions. Result: older adults who completed computerized speed training and attended booster sessions were 25 percent less likely to receive a dementia diagnosis over two decades compared to a control group without training. Total training duration was ten sessions of one hour each, spread over six weeks.
What is the ACTIVE program?
ACTIVE stands for "Advanced Cognitive Training for Independent and Vital Elderly." The U.S. study, funded by the National Institutes of Health, began in 1998 and enrolled 2,802 independently living adults with mean age 74 at six sites. All participants were dementia-free at baseline. Three training types were tested: memory, reasoning, and processing speed.
Speed training uses adaptive visual tasks. Participants identify objects on a screen while simultaneously locating peripheral stimuli. Difficulty and pace increase automatically. The program is now commercially available as BrainHQ, distributed by Posit Science. Co-author Karlene Ball, who originally developed the speed training, has connections to this company, mentioned in study discussions.
Why only speed training worked
After 20 years, researchers led by lead author Norma B. Coe of the University of Pennsylvania found: only speed training statistically reduced dementia diagnosis. Memory training and reasoning training showed no significant long-term effect. The explanatory hypothesis: speed training uses implicit learning and builds skills, comparable to learning an instrument. Memory and reasoning training activate more explicit, conscious learning. Both engage different brain regions. Why this long-term difference exists remains unclear, according to Coe.
Crucially, the 25 percent reduction applies only to participants who, beyond the basic course, attended booster sessions in the eleventh and thirty-fifth months after initial training. Those completing only baseline training showed no measurable advantage over control after 20 years.
What the 25 percent means and does not mean
The 25 percent represents relative risk reduction. In absolute terms: in the control group, 48.7 percent received a dementia diagnosis within 20 years. In the trained group with boosters, approximately 40 percent. The absolute difference is roughly nine percentage points. For a disease currently affecting 57 million people worldwide and projected to reach 139 million by 2050, even this would be substantial impact.
Experts view the findings with caution. Dr. Baptiste Leurent from University College London emphasized that the statistically significant finding comes from a subgroup analysis, not the pre-specified primary comparison of all training groups. The confidence interval (Hazard Ratio 0.59 to 0.95) allows a range between 41 and 5 percent risk reduction. Dr. Susan Kohlhaas of Alzheimer's Research UK pointed out that dementia diagnoses come from Medicare billing data, not clinical evaluations by specialists.
In comparison: what other studies show
The ACTIVE study is the first randomized controlled trial demonstrating preventive effect on dementia diagnosis over 20 years. This is methodologically novel. Earlier longitudinal studies primarily measured cognitive decline, not diagnosis emergence.
The Finnish FINGER study, published in Lancet 2015, was the first randomized intervention trial showing measurable cognitive benefit through a combined program: nutrition adaptation, physical training, cognitive training, and blood pressure management together improved cognitive performance after two years versus control. Individual components alone did not deliver this effect. The 2024 Lancet Dementia Prevention Commission identified 14 total modifiable risk factors, including hypertension, hearing loss, and physical inactivity, potentially responsible for up to 45 percent of all dementia cases.
In Germany, 1.8 million people currently live with dementia. Without structural prevention, numbers could reach 2.8 million by 2050. Even modest preventive effect, as the ACTIVE study suggests, would shift or prevent hundreds of thousands of cases.
Three conditions for broader recommendation
For cognitive speed training to become a general prevention recommendation, three things must be established. First, replication: ACTIVE is currently the sole source for the 20-year effect. An independent study without commercial ties of individual authors, ideally conducted in Europe or Asia, would substantially strengthen evidence. Second, mechanism: without understanding how speed training influences cognitive reserve, it remains unclear whether the specific BrainHQ program is necessary or whether comparable activities like certain video games or concentrated driving produce the same effect. Third, accessibility: BrainHQ is a paid service. To embed prevention in public health systems requires either inexpensive alternatives or insurance reimbursement logic.
The NIH-funded ACTIVE consortium is not currently planning immediate Phase 3 replication. Whether health insurers in the United States or Europe classify the program as reimbursable depends on regulatory decisions falling in coming years.