By Laila Qureshi
A Billion-Dollar Promise Built on Thin Evidence
Open any app store and search “brain training” and you will find dozens of apps promising sharper memory, faster processing speed, and protection against age-related decline. The category has grown into a business worth well over a billion dollars a year, built almost entirely on a simple pitch: play daily puzzle-style games and your brain gets measurably better at everything else too. That pitch is intuitive. It is also, according to the bulk of controlled research, not what the science supports.
This is an evidence review, not a takedown or an endorsement. The goal is to walk through what researchers have actually measured, distinguish between the kind of improvement these apps reliably produce and the kind they advertise, and lay out what the data says works better for long-term cognitive health. Along the way, one specific regulatory case matters a lot: the Federal Trade Commission’s 2016 action against Lumosity, one of the largest and most visible companies in the space, for making health claims its own science did not back up.
Near Transfer Versus Far Transfer: The Distinction That Matters Most
Almost every claim in this space hinges on one methodological distinction that rarely makes it into marketing copy: the difference between near transfer and far transfer.
Near transfer means improvement on tasks that are structurally similar to the trained task. If someone practices a working-memory game for six weeks, they typically get better at that game and at very similar working-memory tests. This effect is well documented and largely uncontroversial: practice improves performance on practiced or closely related tasks, the same way practicing scales makes you better at scales.
Far transfer means improvement on tasks that draw on the same broad cognitive ability but look and feel different, for example whether working-memory training makes someone better at everyday planning, reading comprehension, or fluid reasoning on an unrelated test. Far transfer is the claim brain-training companies actually sell: not “you’ll get better at our game” but “you’ll get a better brain.” It is also the claim with the weakest support.
What the Large Controlled Studies Found
The single most cited piece of evidence here is a study led by neuroscientist Adrian Owen and colleagues, published in Nature in 2010 under the title “Putting brain training to the test.” The study recruited participants through a BBC science program, which allowed for an unusually large sample: more than 11,000 people completed the trial. Participants were randomly assigned to train on a battery of tasks designed to target reasoning, memory, planning, and visuospatial skills, or to a control condition, for six weeks. Every participant was also tested before and after on cognitive tests that were not part of the training itself.
The result: trained groups improved substantially on the specific tasks they practiced. But when tested on the separate benchmarking tasks, the trained groups showed no significant advantage over controls, including on tests that measured the exact broad abilities, like reasoning and working memory, the training programs were designed to target. Practice produced task-specific gains. It did not produce a detectable improvement in general cognitive function.
That single study does not stand alone. In 2016, a team of sixteen researchers led by Daniel Simons published a comprehensive review in Psychological Science in the Public Interest titled “Do ‘Brain-Training’ Programs Work?” The review examined the entire published literature on commercial brain-training claims against a set of methodological criteria: randomized controlled design, active control groups, blinded assessment, and transfer to real-world or standardized far-transfer measures. Their conclusion was blunt: the field’s evidence base was weak, plagued by small samples, absent or inadequate control groups, and publication bias, and it did not support the claim that commercial brain-training programs produce generalized cognitive improvement in healthy adults.
The debate has not been one-sided within the scientific community, which is part of what makes it a genuine evidence review rather than a settled verdict. In 2014, the Stanford Center on Longevity and the Max Planck Institute for Human Development coordinated an open letter, eventually signed by more than 70 cognitive scientists and neuroscientists, stating that the marketing claims of the brain-training industry “exceed what the current scientific evidence indicates.” A separate group of researchers with more favorable findings, organized as the Cognitive Training Data consortium, issued a competing statement defending the plausibility of transfer effects in specific, well-designed studies. Both letters agreed on one point: the average commercially marketed product had not demonstrated the broad benefits it advertised.
The Lumosity Case: When Marketing Outran the Science
The clearest real-world consequence of this evidence gap came from a regulator rather than a journal. In January 2016, the Federal Trade Commission announced a settlement with Lumos Labs, maker of the Lumosity brain-training app, resolving charges that the company had deceived consumers through false advertising.
According to the FTC’s complaint, Lumosity’s marketing told users its games could help them perform better at work and in school, delay age-related cognitive decline, and reduce cognitive impairment associated with conditions including Alzheimer’s disease, traumatic brain injury, PTSD, and ADHD. The FTC found that the company lacked adequate scientific evidence to back these claims and had exploited consumers’ fears about age-related memory loss to sell subscriptions.
The settlement required Lumos Labs to pay $2 million, notify subscribers who signed up for auto-renewal between 2009 and 2015 about the settlement, and stop making unsupported claims about real-world performance benefits, brain disease prevention, or protection against cognitive impairment tied to a specific health condition, unless the claims were backed by competent and reliable scientific evidence. The company was not barred from operating or claiming that its games improve performance on the games themselves, the near-transfer claim, only from asserting the far-transfer, medical-adjacent benefits it had not proven.
The Lumosity settlement matters for this review because it is not a case of a fringe app making outlandish claims. Lumosity was, at the time, one of the most scientifically engaged companies in the category, publishing its own research and partnering with academic labs. If the leading commercial product in the space could not substantiate its central marketing claims to a federal regulator’s standard, that is a meaningful data point about the gap between industry marketing and the underlying evidence.
Why Study Design Changes What You See
Part of why this topic generates so much confusion is that not all brain-training studies are built the same way, and design choices strongly shape the results a company can advertise. A study with a no-contact control group, where the comparison participants simply take the same tests twice with no activity in between, tends to produce larger apparent training effects than a study with an active control group, where the comparison participants do an equally engaging but different activity. The gap between these two designs is not a minor statistical detail. It largely accounts for why some smaller, industry-funded studies report dramatic improvements while larger, independently funded trials with active controls report little to no far-transfer benefit.
Sample size matters just as much. Many of the studies cited in brain-training marketing materials involve a few dozen participants, tracked for a few weeks, with outcome measures chosen after the fact. The Owen et al. trial in Nature is frequently cited precisely because it avoided these pitfalls: a sample size in the tens of thousands, a pre-registered set of benchmark tests unrelated to the training tasks, and random assignment to training or control conditions. When a claim rests on a small, uncontrolled pilot study rather than a large randomized trial, that is a reasonable signal to treat the claim with more caution.
Publication bias adds another layer. Studies that find a positive, marketable result are more likely to get published, funded, and cited in advertising than studies that find no effect. The Simons et al. review specifically flagged this pattern, noting that the file drawer of unpublished null results in cognitive training research is likely substantial, which means the published literature probably overstates the average effect even before marketing departments get involved.
Key Terms in Brain-Training Research
A short glossary helps make sense of the vocabulary that shows up throughout this field, in both scientific papers and marketing copy.
- Near transfer: Improvement on a task that is structurally similar to the one that was practiced, such as one working-memory game improving performance on a closely related working-memory test.
- Far transfer: Improvement on a task that draws on the same general cognitive ability but differs in structure or content from the trained task, such as everyday planning or unrelated reasoning tests.
- Active control group: A comparison group in a study that completes an alternative, similarly engaging activity instead of the trained intervention, used to separate a true training effect from the effect of simply doing something new and stimulating.
- Cognitive reserve: The brain’s capacity to sustain function despite age-related change or damage, thought to be built through a lifetime of varied mental, physical, and social engagement rather than any single activity.
- Executive function: A set of higher-order mental processes, including planning, working memory, and cognitive flexibility, frequently used as an outcome measure in both brain-training and exercise research.
Who Might Still Get Value from These Apps
None of the caution above means these apps have zero place in a person’s routine. Three groups plausibly get real value from them, distinct from the disease-prevention claims regulators have pushed back on. People recovering specific skills after an injury, under clinical supervision and using tools chosen by a therapist, may benefit from the near-transfer, task-specific practice these apps are good at providing. People who simply enjoy the games as a form of mentally engaging entertainment get the same benefit anyone gets from an enjoyable hobby, independent of any cognitive claim. Older adults specifically doing speed-of-processing training modeled on the ACTIVE trial protocol, rather than a generic app bundle, have the most direct evidence behind a possible dementia-risk association, though this remains one narrow finding rather than a blanket endorsement.
What the evidence does not support is treating any commercial app as a substitute for medical care, or expecting a subscription to meaningfully lower personal risk of Alzheimer’s disease or other diagnosed cognitive conditions. That specific gap between hope and evidence is exactly what the FTC’s Lumosity settlement addressed.
What the Evidence Actually Supports
Sorting through the research honestly produces a narrower, more specific set of conclusions than either brain-training marketing or its harshest critics tend to offer:
- Task-specific improvement is real and reliable. People who practice a cognitive task get better at that task. This is not controversial and is consistent across essentially every study in the field.
- Near-transfer to structurally similar tasks is often measurable. Practicing one working-memory task can improve performance on a closely related working-memory task, particularly when the tasks share underlying demands.
- Far transfer to broad, everyday cognitive function is weakly supported at best. The largest and most methodologically rigorous studies, including Owen et al.’s 11,000-person trial, have not found convincing evidence that commercial brain-training routines improve general intelligence, everyday memory, or unrelated reasoning tasks.
- Some structured, intensive training in older adults shows narrower, domain-specific durability. The ACTIVE trial (Advanced Cognitive Training for Independent and Vital Elderly), a large NIH-funded study following thousands of older adults for a decade, found that a specific speed-of-processing training module produced effects on that trained skill detectable years later, and secondary analyses linked it to a reduced relative risk of dementia diagnosis over follow-up. This finding is notable, narrower than typical brain-training marketing, and specific to one training type rather than a general endorsement of the app category.
- Expectation and motivation effects are a real confound. Several trials that lacked active control groups (meaning the control group did nothing rather than doing a comparably engaging alternative task) showed larger apparent training effects, suggesting some reported benefits are partly attributable to participants expecting to improve, a placebo-adjacent effect that better-controlled studies reduce or eliminate.
- Regulatory action has followed the strongest marketing claims, not the training itself. The FTC’s Lumosity case specifically targeted claims about disease prevention and real-world performance, not the basic mechanic of the games.
Evidence Summary at a Glance
| Claim | Evidence strength | Key source |
|---|---|---|
| Improves performance on the trained game itself | Strong | Owen et al., 2010, Nature |
| Improves closely related (near-transfer) tasks | Moderate, task-dependent | Simons et al., 2016, Psych. Science in the Public Interest |
| Improves general/unrelated (far-transfer) cognition | Weak | Owen et al., 2010; Simons et al., 2016 |
| Reduces dementia risk (specific speed-of-processing module) | Moderate, narrow scope | ACTIVE trial, NIA/NIH-funded |
| Prevents or treats Alzheimer’s, PTSD, ADHD-related decline | Not substantiated | FTC v. Lumos Labs, 2016 settlement |
What Shows Better-Supported Cognitive Benefits
If the goal is genuinely protecting long-term cognitive health rather than just getting better at a puzzle app, the research points toward a different, less app-centric set of habits with broader and more consistently replicated support.
Aerobic exercise has one of the more consistent evidence bases in cognitive aging research. Multiple randomized trials and meta-analyses associate regular aerobic activity with improvements in executive function and, in some studies, increases in hippocampal volume in older adults, a structural change linked to memory. Unlike most brain-training results, exercise effects have shown up across varied cognitive measures rather than only on trained tasks.
Learning a genuinely new, complex skill, such as a language, a musical instrument, or a demanding craft, engages multiple cognitive systems simultaneously (motor coordination, memory encoding, sustained attention) in a way that single-mechanic games do not. A widely cited study from the University of Texas at Dallas, the Synapse Project, found that older adults who spent roughly 15 hours a week learning new skills such as quilting or digital photography showed broader memory improvements than a comparison group doing familiar, low-challenge activities.
Sleep quality and duration affect memory consolidation directly; sleep-deprivation research consistently shows measurable next-day deficits in attention and recall, and this relationship is far more experimentally robust than most brain-training transfer claims.
Cardiovascular and metabolic health (blood pressure, blood sugar, and cholesterol management) correlates with dementia risk in large longitudinal cohort studies, including the Lancet Commission’s ongoing reviews of modifiable dementia risk factors, which list physical inactivity, hearing loss, social isolation, and untreated hypertension among the larger addressable contributors to later-life cognitive decline.
Social engagement also shows up repeatedly in longitudinal aging cohorts as a predictor of slower cognitive decline, independent of formal cognitive training.
Better-Supported Alternatives to a Training App Subscription
- 150 minutes a week of moderate aerobic activity (brisk walking counts)
- Learning a new, effortful skill, such as language, instrument, or craft, for several hours weekly
- Consistent sleep schedule with 7-9 hours nightly
- Managing blood pressure and blood sugar through regular checkups
- Regular in-person social contact and group activities
How to Read a Brain-Training App’s Claims Critically
None of this means brain-training apps are worthless as entertainment, or that no one should use them. Puzzle games can be enjoyable, and enjoyment has its own value. The distinction worth holding onto is between what a company advertises and what a controlled study actually measured. A few practical checks:
- Look for the words “near transfer” versus “far transfer” in any research the company cites, or ask what the untrained comparison task was.
- Be skeptical of claims tied to specific diseases (Alzheimer’s, dementia, ADHD); this is exactly the category the FTC targeted in the Lumosity case.
- Check whether cited studies used an active control group (a comparably engaging alternative activity) rather than a no-contact control, since no-contact controls inflate apparent effects.
- Treat “clinically proven” as a marketing phrase, not a technical one, unless a specific peer-reviewed citation is provided.
Frequently Asked Questions
Do brain-training apps actually work?
They reliably improve performance on the specific tasks practiced within the app. The evidence that this translates into broader, everyday cognitive improvement, meaning better memory, focus, or reasoning outside the app, is weak, based on the largest controlled trials to date, including the 2010 Owen et al. study in Nature.
What happened with Lumosity and the FTC?
In 2016, the Federal Trade Commission settled charges against Lumos Labs, the maker of Lumosity, for advertising that its app could delay cognitive decline and reduce impairment linked to conditions like Alzheimer’s disease and ADHD without adequate scientific support. The company paid $2 million and agreed to stop making those specific unsupported claims.
Is any brain-training research positive?
Yes, with an important caveat. The ACTIVE trial found that a specific speed-of-processing training module showed durable, trained-domain benefits in older adults over a long follow-up period, with some association with reduced dementia risk. This is a narrower and more specific finding than the general claims made by most commercial apps.
What actually helps protect cognitive health as people age?
Regular aerobic exercise, learning genuinely new and complex skills, adequate sleep, cardiovascular and metabolic health management, and sustained social engagement all have broader and more consistently replicated support in the research literature than commercial brain-training apps.
Should I stop using brain-training apps entirely?
That is a personal choice. If a person enjoys the games, there is little evidence they cause harm. The main caution is not to substitute app use for the higher-evidence activities above, or to expect the app to deliver medical-grade cognitive protection it has not been shown to provide.
What is the difference between near transfer and far transfer?
Near transfer is improvement on tasks very similar to the one practiced. Far transfer is improvement on different, broader cognitive tasks or real-world functioning. Brain-training apps consistently show near transfer; the marketed benefit is usually far transfer, which has much weaker support.
References
- Owen, A. M., et al. (2010). “Putting brain training to the test.” Nature, 465, 775-778.
- Simons, D. J., et al. (2016). “Do ‘Brain-Training’ Programs Work?” Psychological Science in the Public Interest, 17(3), 103-186.
- Federal Trade Commission. (2016). “Lumosity to Pay $2 Million to Settle FTC Deceptive Advertising Charges for Its ‘Brain Training’ Program.” FTC press release.
- Stanford Center on Longevity & Max Planck Institute for Human Development. (2014). “A Consensus on the Brain Training Industry from the Scientific Community” (open letter).
- Park, D. C., et al. (2014). “The Impact of Sustained Engagement on Cognitive Function in Older Adults: The Synapse Project.” Psychological Science, 25(1), 103-112.
- Ball, K., et al. (2002); Rebok, G. W., et al. (2014). ACTIVE study findings, published in JAMA and Journal of the American Geriatrics Society.
- Livingston, G., et al. (2020). “Dementia prevention, intervention, and care: 2020 report of the Lancet Commission.” The Lancet, 396(10248), 413-446.





























