Several years ago, so-called “brain training” games from companies like Luminosity were all the rage. The various companies that brought these products to market claimed they could be used to offset the aging process by slowing or even reversing age-related cognitive declines. In 2014, two groups of scientists released two competing letters. One group argued there was no benefit to using this type of brain-training, while another group of scientists later claimed there was evidence that brain training could provide a benefit. Which group was right? That’s what a team of authors set out to determine, and the results of their meta-analysis don’t favor the pro-brain training camp.
The new study reviewed the results in over 130 papers that had previously been published and peer-reviewed in scientific journals. It’s long been known brains get better at performing specific tasks as you repeat them — if you play a lot of Counterstrike, you get better at Counterstrike. Play a lot of Tetris, and you’ll get better at Tetris. What makes brain training (hypothetically) different is that the various companies involved claim you can transfer skills you learn in their titles into generally improved work performance. With the over-65 population growing rapidly, medical costs rising, and an acute need for treatments that can stop or reverse the onset of Alzheimer’s disease, dementia, and other types of cognitive decline, the idea you could use specific types of puzzles and games to accomplish these ends was a potentially huge business.
To date, only a handful of studies have actually attempted to answer whether ‘brain training’ can have an impact on real-world performance. Most of the cited studies focus on neural plasticity, which is the brain’s ability to form new neural connections throughout life, and have assumed this means brain training would influence real-world outcomes. Companies that sell brain training products have commonly claimed their various software applications would improve memory, sharpen intellectual skills, and even prevent memory loss.
But statements of this sort aren’t just puffery — they can also fall under the rubric of health claims. The FTC fined Lumor Labs, makers of Luminosity, $2 million earlier this year for false advertising for this reason. Tide can theoretically claim its laundry detergent formula is “new and improved” without changing anything, but a company isn’t allowed to promise a product is effective as a medical treatment unless it provides data to prove it.
Scientists have been studying the nature of transference — our ability to take what we’ve learned in one situation and apply it to other situations — for more than a century, often with mixed results. One intrinsic problem in this field of study is the difference between various types of knowledge. Consider, for example, the difference between knowing how to thatch a grass hut and understanding elementary principles of logic. Both of these are types of knowledge, and both are transferable. But knowing how to build a roof with dry vegetation doesn’t tell you much about modern roofing methods, while the principles of logic can be applied to a dozen different scenarios a day.
Extensive existing research suggests that practice makes us better at specific tasks, but only within a relatively narrow field. If you practice memorizing long series of digits, you’ll become better at it. But you probably won’t see a benefit in your ability to memorize music as a result of memorizing numbers.
The full paper goes into a fair bit of detail, and I’d recommend reading it if you want an overview of the field. The authors write:
Nevertheless, we know of no evidence for broad-based improvement in cognition, academic achievement, professional performance, and/or social competencies that derives from decontextualized practice of cognitive skills devoid of domain-specific content. Rather, the development of such capacities appears to require sustained investment in relatively complex environments that afford opportunities for consistent practice and engagement with domain-related challenges…
Brain-training programs typically train performance on relatively simple skills in a limited range of contexts (typically on a home computer and with little involvement of substantive content or knowledge), but their marketing materials imply generalization to a wide range of skills in varied contexts with varied content.
The problem is, there’s no evidence in the existing literature or the best studies performed to date to support these claims. In one groundbreaking study from the late 1990s, there was evidence elderly individuals who received speed-of-processing training continued to receive a benefit from that training up to a decade later. But the other groups that received memory training or reasoning training showed no sustained benefit in followups five and ten years later. Once they stopped practicing what they’d learned, the benefits vanished. The speed training consisted of visual search and divided attention tasks, and came closest to the types of training that modern brain games offer.
But again, even when benefits were apparent, the benefits were limited to the specific tasks practiced.
The full paper weighs in at more than 50,000 words, but it’s surprisingly readable and fairly interesting. There are still questions about whether cognitive training can stave off some of the problems of old age, but so far, all available literature suggests that practicing the kinds of tasks brain training games present really only makes you good at those kinds of tasks. The search for a holy grail that would help us stave off age-related damage and dementia won’t be solved by cutesy apps or catchy titles.