The Frontal Cortex

Brain Fitness Programs

Posted by Jonah Lehrer on May 5, 2008
(12)
More »

The Times recently had an article on the booming business of brain fitness:

Decaying brains, or the fear thereof, have inspired a mini-industry of brain health products — not just supplements like coenzyme Q10, ginseng and bacopa, but computer-based fitter-brain products as well.

Nintendo’s $19.99 Brain Age 2, a popular video game of simple math and memory exercises, is one. Posit Science’s $395 computer-based “cognitive behavioral training” exercises are another. MindFit, a $149 software-based program, combines cognitive assessment of more than a dozen different skills with a personalized training regimen based on that assessment. And for about $10 a month, worried boomers can subscribe to Web sites like Lumosity.com and Happy-Neuron.com, which offer a variety of cognitive training exercises.

Alvaro Fernandez, whose brain fitness and consulting company, SharpBrains, has a Web site focused on brain fitness research. He estimates that in 2007 the market in the United States for so-called neurosoftware was $225 million.

It’s a fine article, but I still wish there was mention of the fact that so many of these brain fitness products are utter nonsense. Last year, I wrote a short article for Seed on these over-hyped products:

It’s a trend that began in Japan. In the summer of 2005, Nintendo introduced Japanese consumers to its Brain Age software for its handheld gaming system. Since then, it has sold more than four million copies worldwide. The conceit of the game is simple, if woefully unscientific: After assessing your initial “brain age” with a brief series of mental tasks, the program runs you through a gauntlet of basic exercises like reading comprehension and sudoku before measuring you again. I was happy to see that my improvement was nothing short of miraculous: After less than an hour with Nintendo’s Brain Age, my own brain age had decreased by 20 years, from 65 to 45. Of course, when the science is pretend, great results are easy to come by.

Although Nintendo employs Ryuta Kawashima, a famous Japanese neuroscientist, to sell Brain Age, the company carefully avoids making any specific scientific claims. “We’re in the entertainment business,” says Perrin Kaplan, head of marketing for Nintendo’s U.S. operations. But the company coats Brain Age in a veneer of neuroscience. Potential customers find colorful fMRI images and admonishments to “get the most out of your prefrontal cortex” on the game’s promotional Web site. This entertainment is masquerading as neurology.

That said, there are some programs that come with real empirical proof. (Update: see some important qualifications in the comments.) PositScience, developed by UCSF neuroscientist Michael Merzenich, is damn expensive ($395!) and rather boring to play, but it has undergone some real testing:

In a recent PNAS paper, Merzenich’s lab announced that PositScience was able to reverse “age-related cognitive decline” in a randomized and controlled study of 182 subjects. Of those trained with PositScience, 93 percent showed significant cognitive improvement. “We’ve demonstrated that you can take the brain of a 75-year-old,” Merzenich says, “and make it function like the brain of a 35- or 40-year-old. It takes training, and some hard work, but it’s possible.” Preliminary results of a second trial study suggest that PositScience can even help stave off memory loss in the early stages of Alzheimer’s patients. After just four weeks, senile patients showed significant cognitive improvement. The control group, on the other hand, continued to decline.

Merzenich has big plans for PositScience. He wants to launch programs that target the visual cortex, working memory, and executive control. With those goals in mind, he has partnered with the Mayo Clinic to conduct an expansive trial study. Moreover, Merzenich has begun studying the positive effects of PositScience on schizophrenic patients. The results so far have been “extremely encouraging”: “If we got these same improvements with a pill,” Merzenich says, “we’d be counting the money already. We’d have billions in sales. But this isn’t a pill–it’s much better than that.”

(12)
More »

Comments

  1. #1 Hank Roberts
    May 5, 2008

    I wish they’d sell a lifetime membership or something. Pricey!

    Tempting, though. Very …..

  2. #2 OftenWrongTed
    May 5, 2008

    As a personal alternative I have begun to read the books contained in the bibliography of Proust Was a Neuroscientist, by Jonah Lehrer. The results after six months are promising since my professional education was science devoid of art and literature. .

  3. #3 Maja
    May 5, 2008

    What happened with good old fashion dancing for brain training?

  4. #4 Lee Pirozzi
    May 5, 2008

    Research your ancestors and eat the herbs from their homeland. Unless of course you want to wait for mock dinosaur. What else are they doing in the world today?
    My 9 year son just aimed his low-impact PLASTIC dart gun at his 12 year old brother – who then yelled, “Ow – My scapula!” Fortunately they are laughing,
    and they missed a painting.

  5. #5 jay french
    May 5, 2008

    Approaching 70 years of age, the claims of these brain exercisers has tempted me. I have not until now read Johan’s article debunking these programs and I am delighted I did not succumb. Rather I have relied on timing my weekly effort to complete the New York Times Sunday Magazine crossword. In the last thirty years, the loss has been less than fifteen minutes, and defensively I attribute that to the fact that I cannot identify the TV sitcom characters that seem to increasingly inhabit the clues during the last decade. (Admittedly, doing that puzzle for decades gets one into the mind of the editor and makes solving an acquired skill.) But if the timing starts to slip, there will likely be a check in the mail to Professor Merzenich. To paraphrase, “A mind is a terrible thing to waste, particularly if it’s yours!”

  6. #6 Peter Reiner
    May 5, 2008

    I concur with most of Jonah’s comments, but I do think it is worth taking a peek at a paper that has been posted online ahead of publication in PNAS. Using a novel training paradigm, Jaeggi et al. were able to show improvement in modalities other than those being trained, in particular modalities that include fluid intelligence, and that the improvement was dose dependent: the more one trained, the more one improved. [Before anyone gets too carried away, it is also worth noting that the duration of the improvement is not known - it could be transient or long-lasting - only time will tell.] Nonetheless, this result is unprecedented, and may open the way for further ‘ games’ that truly enhance cognitive abilities.

  7. #7 WetBlanket
    May 6, 2008

    I assume that the paper Prof. Merzenich is referring to in the quotation above is the paper that his company, Posit Science, published in 2006:
    http://www.pnas.org/cgi/content/full/103/33/12523
    I remember our lab reading that paper quite closely as we are interested in rehabilitative training using computers, not to mention the sterling reputation Dr. Merzenich has in the area of brain plasticity (e.g. his experimental work on auditory plasticity in rats is first rate).

    To make a long story short, we were underwhelmed with the evidence. I’ll just make a few points about the paper:

    a) The paper was “Contributed” by Dr. Merzenich (this implies that he’s a member of NAS so he can publish in PNAS without review, correct?). So the paper being published in PNAS might be due to his reputation, not the paper’s merits.

    b) The training effects in that paper are pretty small: the basic, generalizable effect is (being optimistic) a z=0.25 (z = mean/standard deviation) improvement. Really, though, you should subtract off z=0.1 from that because that’s the improvement the controls showed in being retested in memory performance. So you are left with a z=0.15 improvement for shelling out $400 and doing brain exercises for 8-10 weeks, 5 days/ week, 1 hour/ day. A z=0.15 improvement means that you have a 56% chance of doing better on the memory test after all that training than if you did not (=50% chance).

    c) Some of the training benefit might be due to the relatively high number of subjects (>10%) who dropped out over the course of the training. Also, they got a little unlucky in randomizing subjects because the trained subjects were 3 years younger on average than the key control subjects – that might have had an effect in outcomes.

    d) The bigger boost of z=0.6 [z=0.3 after subtracting off controls improvement] reported for subjects who started out less proficient in doing the brain exercises (so they had a higher “ceiling”) was discovered post-hoc – meaning that Posit didn’t plan on looking for this. Post-hoc results always have to be taken with a grain of salt [though this one isn't too unreasonable, really - it's not like they dug up a 100% improvement in left-handed latino ladies, or something completely arbitrary like that].

    But go read the paper for yourself and see what you think.

    We have found that in this area of research the devil really is in the details – effects tend to be strong for improvements in the specific tasks being trained on, but pretty weak when tested in more general abilities. Too bad, really, because this kind of training could help a lot of people (computers are cheap). Hopefully the Jaeggi paper points to a way forward, but we’ll see.

  8. #8 Peter Reiner
    May 6, 2008

    Actually, the Jaeggi et al. paper is not from Mezernich’s lab…

  9. #9 Steve Higgins
    May 6, 2008

    “this implies that he’s a member of NAS so he can publish in PNAS without review”

    I think they still get reviewed.

  10. #10 Lizzie
    May 22, 2008

    Yes, they still get “reviewed,” but I’ve never heard of one getting rejected. I agree that Merzenich’s results are less than stellar. As much as I respect his work with auditory plasticity and cochlear implants, he’s been somewhat obnoxious with Posit Science.

    On the website, he has an extraordinarily long list of publications which he claims “investigate the effectiveness of Posit Science programs”–but they actually have nothing to do with Posit Science, at least not directly. He put them there because he recently started collaborating with those researchers so he just pasted up their publication records, deceptively insinuating that they’re actually directly relevant to Posit Science’s games.

    That said, Posit Science tries harder to be scientific than most programs, like Brain Age and Happy Neuron. I think the best brain games are at Lumosity. To be honest, I know the people who founded Lumosity, so I’m biased, but I’m also aware of all the active scientific research they’re doing on their games in collaboration with researchers at very respectable scientific institutions. And the games are cheap and VERY FUN.

  11. #11 michael s. logan
    May 30, 2008

    This is what I have been looking for, some discussion of the research. Thankyou, thankyou. Mike Logan

  12. #12 geciktirici
    February 15, 2009

    It’s a trend that began in Japan. In the summer of 2005, Nintendo introduced Japanese consumers to its Brain Age software for its handheld gaming system. Since then, it has sold more than four million copies worldwide. The conceit of the game is simple, if woefully unscientific: After assessing your initial “brain age” with a brief series of mental tasks, the program runs you through a gauntlet of basic exercises like reading comprehension