Brain
Brain Games and Cognitive Aging
Brain games make you better at the games. Transfer to daily life is weak, but one speed-training program was linked to fewer dementia diagnoses.
A brain game is an unusually convenient health pitch: a few minutes on your phone, perhaps while the kettle gets on with its work. But what travels beyond the screen? The Humanic Blue question is how far that benefit travels—from a better game score to the much bigger claims about aging.
A review led by University of Illinois psychologist Daniel Simons found strong evidence for improvement on practiced tasks, but little for broad everyday benefits. One specific speed-training program has been linked to fewer dementia diagnoses in long-term follow-ups, with important limitations. That finding doesn’t establish that any brain-game app prevents dementia or helps people live longer.
Key Takeaways
- Practice improves performance on the trained task; broader benefit is harder to demonstrate.
- Cochrane reviewers found uncertain evidence for lasting, general improvements from computer training.
- One speed-training program has a possible long-term dementia signal, not a universal prevention result.
- These studies don’t establish lifespan extension or validate every commercial app.
What improves when you practice?
The publicly funded ACTIVE trial tested memory, reasoning and speed training in independent older adults. Participants improved the particular abilities they practiced, and those gains could last.
That part makes sense: practicing a task can make it easier. The harder question is how far the improvement travels. Faster at identifying shapes isn’t automatically better at remembering an appointment.
Does the benefit travel?
Simons and colleagues reviewed the published intervention studies cited by leading brain-training companies. Evidence weakened as they moved from the practiced task to related tasks, then to unrelated skills and everyday life.
Researchers call this transfer. We can use a simpler question: what got better outside the exercise itself? That is the question a broad “sharper brain” claim needs to answer.
When an app celebrates a rising score, what change beyond the screen would make that score meaningful to you?
What changed in everyday life?
The Simons review found little evidence of everyday cognitive improvement. Cochrane reviewers also found that small gains in older adults were uncertain and didn’t clearly persist.
In a separate review of people with mild cognitive impairment, a condition involving thinking difficulties beyond expected aging with everyday independence largely preserved, the trials didn’t report enough disease outcomes to establish dementia prevention. Thinking-test performance, daily independence and dementia diagnosis need separate evidence.
That doesn’t tell us a puzzle can’t be enjoyable. It tells us enjoyment and a medical promise should be judged separately. Time, subscription costs and the claim being sold all belong in the comparison.
What is the specific dementia finding?
ACTIVE’s speed training involved quickly spotting information on a screen while dividing attention. It was a particular training approach, not a test of brain games generally.
The latest long-term analysis linked speed training with additional practice sessions, called boosters, to fewer dementia diagnoses. Memory and reasoning training didn’t show that advantage, and the whole speed-training group wasn’t clearly different when people with and without boosters were combined.
These are follow-up analyses of one trial. Only people who completed most initial sessions were eligible for boosters, and the latest diagnoses came from billing records rather than new clinical assessments. Those limits leave uncertainty about prevention. Some researchers disclosed ties to the company marketing related training. Today’s commercial app wasn’t the version tested in ACTIVE. The follow-ups and disclosures explain the finding in full.
So there is a specific signal worth understanding. It doesn’t give every puzzle on the app store a prevention certificate.
How much evidence would you want before turning an enjoyable pastime into something you felt obliged to do for your health?
What happened when marketing went further?
The Federal Trade Commission brought deceptive-advertising charges against Lumosity’s maker over unsupported claims, including protection against dementia and age-related decline. The company settled.
The relevant lesson is about the gap between the study and the sales pitch. A real training effect doesn’t substantiate every broader claim attached to it.
What’s Interesting
The scientific question is more interesting than “games: good or bad.” It’s whether a learned skill travels to the rest of life. A score can rise while the bigger question stays open. That helps us read a result without either dismissing the whole field or giving the marketing all the credit.
What This Means in Practice
No result here establishes that games help people live longer. For living longer and better, enjoyment, daily functioning and medical outcomes are different things to weigh.
Simons and colleagues urge comparison with other uses of the same time and money. A useful question for a product is what its specific program has demonstrated, in whom, and outside the trained task. Protect Your Brain as You Age covers the broader health picture.
New memory concerns need clinical evaluation; a game score cannot explain them. At-Home Cognitive Tests describes screening’s separate role.
Bottom Line
Training improves practiced skills. Evidence for broad everyday benefit is weaker, and the dementia signal belongs to one particular speed-training program with important limits. Enjoyment and medical effectiveness are separate reasons for spending time on a game.
A NOTE FROM LONGEVITY365
We read research, consult expert perspectives and work to make complicated health information easier to understand. What you’re reading here reflects how we’ve interpreted and organized that information. It is educational content—not personal medical advice, diagnosis or treatment guidance.
We’re not perfect, and neither is the available research. Evidence changes, new studies emerge and what’s appropriate for one person may not be appropriate for another.
Use this article as a starting point for your own research and conversations with qualified professionals. Always talk with your doctor or another trusted healthcare professional before making decisions that could affect your health, particularly if you take medications, are receiving treatment, have existing health conditions, are pregnant or have other circumstances that could affect what’s appropriate for you.
Reader Questions
A few questions to carry into the comments or your next conversation. Please leave out identifying or personal medical details.
- 01What benefit outside a game would make a training claim meaningful to you?
- 02How would you weigh enjoyment separately from a promise about health?
Extra Insights & Sources
For readers who want the details behind each claim above.
What We Know
- Brain training improves performance on trained tasks, durably for years. Established.
- Transfer to unrelated tasks and daily life is weak. Supported.
- Computer-based training in healthy older adults showed small gains that didn’t clearly persist. Preliminary.
- The FTC has penalized a major brain-game company for unsupported dementia claims. Established.
What We Don’t Know
- Whether speed training prevents dementia, as opposed to delaying diagnosis. Uncertain; the evidence comes from one trial’s secondary analyses.
- Whether today’s commercial versions of speed training reproduce ACTIVE’s effect. Unknown.
- Whether brain games help people live longer. Unknown.
ACTIVE, the original trial
The ACTIVE trial (Advanced Cognitive Training for Independent and Vital Elderly) was a randomized, controlled, single-blind trial that recruited 2,832 volunteers aged 65 to 94 from senior housing, community centers and hospitals or clinics in six US metropolitan areas between March 1998 and October 1999. Participants were assigned to 10-session group training in memory (n = 711), reasoning (n = 705) or speed of processing (visual search and identification; n = 712), or to a no-contact control group (n = 704). Four-session booster training was offered to a random 60% of trained participants 11 months later. Thirty people were incorrectly randomized and excluded from analysis.
Each intervention improved its targeted ability, durable to two years (P < .001). Immediately after training, 87% of speed-trained, 74% of reasoning-trained and 26% of memory-trained participants showed reliable improvement. Boosters enhanced gains in speed (92% with booster vs 68% without) and reasoning (72% vs 49%). “No training effects on everyday functioning were detected at 2 years.” The authors wrote that the training effects were “equivalent to the amount of decline expected in elderly persons without dementia over 7- to 14-year intervals,” and that with so little decline in any group, “longer follow-up is likely required to observe training effects on everyday function.” Authorship was alphabetical by vote of the ACTIVE Steering Committee; Karlene Ball of the University of Alabama at Birmingham is listed first.
The brain-training review
Simons DJ, Boot WR, Charness N, Gathercole SE, Chabris CF, Hambrick DZ and Stine-Morrow EAL reviewed “all of the published peer-reviewed intervention studies cited on the websites of leading brain-training companies listed on Cognitive Training Data,” the site that hosted the proponents’ open letter. Their full conclusion: “we find extensive evidence that brain-training interventions improve performance on the trained tasks, less evidence that such interventions improve performance on closely related tasks, and little evidence that training enhances performance on distantly related tasks or that training improves everyday cognitive performance.” They also found that “many of the published intervention studies had major shortcomings in design or analysis that preclude definitive conclusions about the efficacy of training.” On cost, they wrote that “before starting a brain-training protocol, people should consider the comparative effectiveness of that protocol.”
The authors declared no conflicts of interest with respect to the article. They also disclosed research connections: Walter Boot and Neil Charness were evaluating an experimental product with Aptima software (paying Aptima from a National Institute on Aging grant); Boot had helped with an unpublished CogniFit study, with travel reimbursed by CogniFit; Susan Gathercole holds a Cogmed research license giving free product access; Christopher Chabris is a shareholder and scientific adviser for Knack.it, an assessment-games company; and Elizabeth Stine-Morrow signed the 2014 Stanford/Max Planck consensus letter.
The Cochrane reviews
In these results, global cognition means an overall thinking score. Episodic memory concerns events or experiences; working memory holds information in mind while using it. SMD (standardized mean difference) puts differences from varied tests on a common scale, rather than giving a percentage improvement. CI (confidence interval) shows uncertainty around an estimate.
Healthy older adults (2020 update). Gates NJ and colleagues searched to March 2018 for randomized trials of computerized training lasting at least 12 weeks in cognitively healthy people, at least 80% of them 65 or older. They found eight trials with 1,183 participants; training ran 12 to 26 weeks, and five trials lasted 12 or 13 weeks. Against active controls, two studies (232 participants) suggested slight improvement in global cognition at the end of training (SMD −0.31, 95% CI −0.57 to −0.05; low-quality evidence). One trial (77 participants) measured global cognition 12 months later and found no clear persistent effect (SMD −0.21, 95% CI −0.66 to 0.24). Training may make little or no difference to episodic or working memory. Against inactive controls, there was no data on global cognition and a possible slight improvement in episodic memory from one study of 150 people. The reviewers found no evidence on quality of life, activities of daily living or adverse effects. They said the small benefits were of uncertain clinical importance.
Mild cognitive impairment (2019). Eight trials with 660 participants, lasting 12 weeks to 18 months, met the criteria. None of the trials reported data on new cases of dementia. Most were at unclear or high risk of bias, and the evidence was of very low quality for most outcomes. The reviewers concluded that the evidence could not show whether computerized cognitive training prevents clinical dementia, or improves or maintains cognitive function in people who already show cognitive impairment. They added that direct adverse events are unlikely, but that the time, and sometimes the money, that such programs take may be significant burdens.
The dementia follow-ups
HR (hazard ratio) compares how quickly diagnoses accumulate during follow-up. A value of 1 means no difference; a confidence interval spanning 1 leaves the direction uncertain. Adjusted estimates account for measured differences between groups, but cannot remove every possible source of bias. The table’s percentages show recorded diagnoses; the hazard ratios also take timing into account.
10-year analysis (2017). Edwards JD and colleagues analyzed 2,802 ACTIVE participants. Training ran up to 10 sessions over six weeks, with up to four booster sessions at 11 months and again at 35 months. Dementia was defined with a combination of interview- and performance-based methods, and 260 cases were identified. Speed training was associated with lower dementia risk than control (HR 0.71, 95% CI 0.50–0.998, P = .049); memory (HR 0.79, P = .177) and reasoning (HR 0.79, P = .163) training were not. Each additional speed session was associated with a 10% lower hazard (unadjusted HR 0.90, 95% CI 0.85–0.95). The paper notes that an earlier ACTIVE analysis found no difference in dementia by training arm at five years.
Disclosures in the paper: Jerri Edwards worked from 1996 to 2005 as a consultant conducting related research studies for Visual Awareness, Inc., which owned the intellectual property for the speed-training software; worked for about three months in 2008 as a limited consultant to Posit Science to analyze data and prepare a publication; and serves on the data safety and monitoring board of NIH grants awarded to Posit Science employees. Coauthor Frederick Unverzagt received free Posit Science site licenses for investigator-initiated research (2009 to 2012). Coauthor Lin Guey was employed by Moderna Therapeutics.
20-year analysis (2026). Coe NB and colleagues linked ACTIVE records to Medicare claims from 1999 to 2019. Of the 2,802 original participants, 2,763 matched; the analysis excluded 725 people in Medicare Advantage at baseline (whose claims data are incomplete), eight who died in the enrollment year and nine already diagnosed, leaving 2,021 (72.1%). Dementia (Alzheimer’s disease and related dementias, ADRD) was identified from billing records using the Chronic Conditions Warehouse algorithm, a set of rules for finding diagnoses in those records. This measures recorded diagnoses, not a fresh clinical assessment of every participant. About 77% of the sample died during follow-up.
The last two rows below concern the 465 speed-trained people who completed at least eight of the first 10 sessions and were eligible for booster assignment. They are a subset of the full 512-person speed arm, not a division of everyone in it. As primary Table 2 and its footnotes specify, the hazard ratios in both rows compare those eligible subgroups with the 491-person no-contact control group; they do not directly compare boosters with no boosters. The direct booster comparison is reported after the table.
| Group (vs no-contact control) | Diagnosed ADRD over 20 years | Unadjusted HR (95% CI) | Adjusted HR (95% CI) |
|---|---|---|---|
| Control | 239 of 491 (48.7%) | reference | reference |
| Memory training, all | 231 of 516 (44.8%) | 0.86 (0.72–1.03) | 0.85 (0.70–1.02) |
| Reasoning training, all | 223 of 502 (44.4%) | 0.86 (0.72–1.04) | 0.88 (0.73–1.07) |
| Speed training, all | 223 of 512 (43.6%) | 0.85 (0.70–1.01) | 0.87 (0.72–1.05) |
| Booster-eligible speed: assigned boosters | 105 of 264 (39.8%) | 0.74 (0.59–0.93) | 0.75 (0.59–0.95) |
| Booster-eligible speed: assigned no boosters | 92 of 201 (45.8%) | 0.97 (0.78–1.21) | 1.01 (0.81–1.27) |
The authors wrote that participants in each training arm tended to have a lower risk of an ADRD diagnosis, but that these differences were not statistically significant. The results text labels the 0.74 figure for speed training with boosters as adjusted; the table and abstract give the adjusted figure as 0.75. Compared directly with speed-trained people randomized not to receive boosters, the booster group’s adjusted HR was 0.81 (95% CI 0.66–1.0). Booster eligibility required completing at least eight of the 10 initial sessions. The abstract describes the booster group as participants who completed at least one booster session, while the results describe them as those randomized to receive booster sessions. The authors wrote that selecting participants after the initial random assignment could still distort comparisons, that people with more education are more likely to receive dementia diagnoses, and that excluding the generally healthier Medicare Advantage enrollees could have made differences between groups look smaller. They acknowledged that a Type I error (a false-positive finding) was possible, given the many ACTIVE publications.
The paper’s conflict-of-interest statement discloses that Karlene Ball consults for and owns stock in Posit Science Inc. Posit Science acquired, and markets, the Useful Field of View Test and the speed-of-processing training software, which the Visual Awareness Research Group, Inc. originally developed and which was used in the ACTIVE trial. Ball still collaborates on designing and testing these assessment and training programs as a member of Posit Science’s Scientific Advisory Board. The other authors reported no disclosures.
Posit Science’s BrainHQ ACTIVE study page says: “The speed-of-processing training that ACTIVE study participants used was an earlier version of the BrainHQ exercise Double Decision,” and that the company’s science team worked to keep the exercise faithful to the original (as of September 26, 2026).
The FTC settlement
The FTC announced the settlement on January 5, 2016. According to the complaint, Lumosity consisted of 40 games, and the company advertised that training 10 to 15 minutes three or four times a week could help users achieve their “full potential in every aspect of life.” Subscriptions were then priced from $14.95 monthly to $299.95 for a lifetime membership. The FTC alleged that the company claimed the games would improve performance at school, work and sports; “delay age-related cognitive decline and protect against mild cognitive impairment, dementia, and Alzheimer’s disease”; and reduce cognitive impairment tied to conditions including stroke, traumatic brain injury, post-traumatic stress disorder (PTSD), attention-deficit/hyperactivity disorder (ADHD) and chemotherapy side effects. The proposed stipulated court order required “competent and reliable scientific evidence” for future claims and imposed a $50 million judgment against Lumos Labs, suspended after the company paid $2 million. The Commission vote was 4-0.
Sources
- Ball K, et al. ACTIVE trial, JAMA, 2002: PubMed 12425704
- Simons DJ, et al. Do “Brain-Training” Programs Work?, Psychological Science in the Public Interest, 2016: PubMed 27697851
- Gates NJ, et al. Computerised cognitive training in cognitively healthy people in late life, Cochrane Database of Systematic Reviews, 2020: PubMed 32104914
- Gates NJ, et al. Computerised cognitive training in mild cognitive impairment, Cochrane Database of Systematic Reviews, 2019: PubMed 30864747
- Edwards JD, et al. Speed of processing training and dementia, Alzheimer’s & Dementia: Translational Research & Clinical Interventions, 2017: PubMed 29201994
- Coe NB, et al. Cognitive training and claims-based diagnosed dementia over 20 years, Alzheimer’s & Dementia: Translational Research & Clinical Interventions, 2026: PubMed 41669119
- Federal Trade Commission: Lumosity to Pay $2 Million to Settle FTC Deceptive Advertising Charges
- Posit Science, BrainHQ: The ACTIVE Study
Notes
Note 1: The ACTIVE trial was funded by the National Institute on Aging and the National Institute of Nursing Research, according to the 2017 and 2026 follow-up papers. The 2026 analysis was funded by the National Institute on Aging (R01AG056486). The 2002 JAMA paper’s own funding statement isn’t in its public PubMed record.
Note 2: The Cochrane reviews’ funding sources aren’t listed in their public PubMed records. Their authors declared no interests related to cognitive training; two reported pharmaceutical advisory or consulting roles outside the reviews’ subject.
Leave a Reply