In 2014, Luminosity — the leading brain training app at the time — settled with the Federal Trade Commission for $2 million over charges that it made false and unsubstantiated claims that its games could reduce cognitive impairment, prevent Alzheimer's disease, and improve real-world performance in work and school. The company, the FTC concluded, had failed to demonstrate that training on its games transferred to meaningful real-world cognitive improvements.
This does not mean cognitive training is ineffective. It means that not all cognitive activities are equally beneficial — and that the research distinguishes sharply between narrow training improvements and genuine cognitive enhancement.
The Core Problem: Near Transfer vs. Far Transfer
The central challenge in cognitive training research is the transfer problem. Near transfer refers to improvement on tasks very similar to the trained task. Far transfer refers to improvement on untrained tasks that require different cognitive processes. Commercial brain training apps reliably demonstrate near transfer — you get demonstrably better at the specific games in the app. What they fail to demonstrate, in most rigorous studies, is far transfer to real-world skills like driving, reading comprehension, navigation, or social functioning.
A landmark 2014 consensus statement signed by more than 70 leading cognitive neuroscientists and neuropsychologists concluded: "We object to the claim that brain games offer consumers a scientifically grounded avenue to reduce or reverse cognitive decline when there is no compelling scientific evidence to date that they do so." This position has been nuanced but not fundamentally overturned by subsequent research.
What the Research Actually Supports
1. Aerobic Exercise (Strongest Evidence)
Physical exercise consistently outperforms digital cognitive training in head-to-head comparisons for global cognitive benefit. A meta-analysis of 39 randomized controlled trials published in the British Journal of Sports Medicine found that aerobic exercise significantly improved memory, attention, processing speed, and executive function in adults. The mechanism involves BDNF upregulation, hippocampal neurogenesis, and improved cerebrovascular health — effects no computer game can replicate.
Recommended minimum: 150 minutes of moderate-intensity aerobic exercise per week, with resistance training 2 days per week for additional executive function benefits.
2. Learning a Musical Instrument
Musical training is among the most cognitively demanding activities a human can undertake. It simultaneously recruits fine motor control, auditory processing, working memory, executive attention, and emotional processing — engaging a wider neural network than virtually any other single activity. Research from Northwestern University shows that musicians have enhanced cognitive reserve, faster auditory processing, and better working memory than matched non-musicians. Even beginning musical training in adulthood produces measurable neural changes within months.
3. Learning a New Language
Bilingualism and active second-language learning are among the most extensively studied cognitive training interventions. A meta-analysis in Psychonomic Bulletin & Review found that bilingualism was associated with a 4.1-year delay in dementia symptom onset. The cognitive demands of maintaining and switching between two language systems appear to strengthen executive control networks — the same neural circuits that regulate attention, inhibition, and cognitive flexibility.
4. Meditation and Mindfulness Practice
Structured meditation practice — particularly mindfulness-based stress reduction (MBSR) — has demonstrated changes in gray matter density in areas associated with attention, learning, and memory, including the hippocampus and prefrontal cortex. A study from Harvard Medical School found that 8 weeks of MBSR produced measurable increases in hippocampal gray matter and corresponding improvements in working memory. The stress-reduction component also protects against cortisol-mediated hippocampal damage.
5. Complex Social Engagement
Social interaction is one of the most cognitively demanding activities humans perform — requiring simultaneous management of linguistic, emotional, social modeling, and memory systems. Longitudinal studies consistently find that socially isolated older adults show faster cognitive decline. The Rush Memory and Aging Project found that the most socially active participants had a 70% reduced rate of cognitive decline compared with the most isolated.
6. Reading (Active, Engaged Reading)
Regular engaged reading — particularly of complex fiction — has been associated with better theory of mind, sustained attention, vocabulary maintenance, and delayed cognitive decline. The key word is "engaged": passive reading with little comprehension has minimal cognitive benefit. Reading with discussion, annotation, or active recall (summarizing what you read) dramatically amplifies cognitive effects.
7. Strategy Games with Novel Challenge
While commercial brain training apps show weak evidence for far transfer, strategy games involving genuine novelty and increasing difficulty — chess, Go, bridge, complex tabletop games — show somewhat stronger evidence for executive function benefits. The critical factor is continued challenge: once you master a game, its cognitive benefits plateau. Seeking consistently novel challenges is essential.
8. Dancing
Dance combines aerobic exercise, coordination, rhythm processing, spatial navigation, social interaction, and often musical engagement — making it a uniquely multimodal cognitive activity. A 2017 study in Frontiers in Human Neuroscience found that dancing produced greater improvement in hippocampal subregion volume compared with conventional fitness training, with additional benefits in balance and processing speed.
The Novelty Principle
Across all forms of cognitive activity, the brain's neuroplastic response is driven by novelty and challenge. Comfortable, routine activities — regardless of how cognitively demanding they once were — produce limited ongoing plasticity. The brain adapts to mastered skills and stops responding neuroplastically to them. Sustained cognitive benefit requires sustained exposure to genuine challenge in unfamiliar territory.
The practical implication: the best cognitive exercise is whatever you have not yet mastered and continue to find challenging. A musician learning a new instrument gains less benefit than a non-musician learning their first. A lifelong chess player gets less neuroplastic stimulus from casual games than from competing at a higher level. Continual learning — in new domains — is the cognitive training strategy most supported by neuroscience.