A sheen is starting to appear on Rocky Blumhagen’s forehead, just below his gray hair. He’s marching in place in a starkly lit room decked out with two large flatscreens. On both of the TVs, a volcano lets off steam through wide cracks glowing with lava, their roar muffling the Andean percussion and flutes on the soundtrack. Golden coins slide across the screen. Rocky reaches out his left hand, as if to grasp a coin from midair, and one of them disappears with a brrring.
“I don’t know if I can do it,” he says to a guy named Josh sitting nearby in a felt-covered lounge chair. He looks up from his iPad, watching Rocky, age 66, grab, jog, kick, and reach his way through the videogame. “Keep it up,” Josh says as the heart monitor in the corner of the screen reads 129.
Rocky and research assistant Josh Volponi are technically in a lab clinic at the University of California, San Francisco, but aside from the mannequin heads studded with electrodes, the room looks more like a man cave. But here, the videogames could halt the mental decay of aging. This is the premise that the university’s new research institute, named Neuroscape, was built to test.
This is Rocky’s 18th training session at Neuroscape, founded by neuroscientist Adam Gazzaley. Rocky is fit for his age—he works as a substitute yoga instructor, after retiring from careers producing radio and performing Cole Porter songs—but as he makes it to the end of the level, he looks exhausted. The game cuts to an animation of a jungle, birds chirping and light playing through the canopy as a list of his past scores pops up. This round wasn’t his best. “I haven’t been here for a week,” he says. Volponi asks him to rate his physical exertion level. Rocky gives it a 15 out of 20; Volponi marks it on the iPad. “I feel rusty,” he says, wiping his hands on his orange exercise shorts.
Over the next week, Rocky will return three more times, completing a two-month run of 21 gaming sessions. At the end, researchers will test his physical and cognitive abilities and scan his brain to compare to his baseline MRI before playing. Along with 19 other players and 20 controls, all age 55 to 85, Rocky’s results will suggest whether a brain-training exercise game could ever really have an impact on cognitive function.
But there’s a problem. It’s really difficult to design an experiment to convincingly show that brain training works. And there’s an even bigger problem: It’s also pretty hard to show that it doesn’t work.
Escape the Room
Brain training is a billion-dollar industry built on wishes. Around 2000, companies like Lumosity and Posit Science began combing through a vast literature of learning research. They took the experimental exercises used to test memory, attention, perception, and other cognitive abilities, and turned them into “delightful games,” as Lumosity calls them. They marketed the training to people for a whole host of ills, including aging—at least until 2016, in Lumosity’s case, when the Federal Trade Commission charged the company with deceiving consumers. The company settled for $2 million.
Before the settlement, Lumosity’s website touted specific benefits for different brain areas. Now it’s more circumspect. It showcases a single study from 2015 that showed playing Lumosity games led to better performance on a cognitive test than playing a crossword puzzle. “These results are promising,” the site says modestly, before adding a familiar caveat: “But we need to do more research to determine the connection between improved assessment scores and everyday tasks in participants’ lives.”
And there’s the rub. Playing a videogame will certainly make you better at that game. But the evidence that playing a game makes you better at life—remembering and focusing in everyday situations—is weak at best. Psychologists call these types of benefits “far transfer,” and they’re the ultimate goal of every brain game designer.
But scientists clash over whether that kind of skill transfer is even possible. In the past few years, they’ve engaged in a polite academic battle, lobbing opposing review papers and open letters back and forth. In 2014, 75 scientists declared that there was little evidence that brain training works. In response, 133 scientists—some affiliated with companies like Posit—returned fire, saying that “a substantial and growing body of evidence shows that certain cognitive training regimens can significantly improve cognitive function.” They cited 132 articles.
Then, a small group of psychologists started taking notice of the sweeping claims brain-training companies were making about their games. So these researchers—many of whom studied the effects of regular old commercial videogames on the brain—decided to read every paper cited by a major brain-training company, along with the 132 cited by the brain-training advocates. It took them a couple of years to review all 378 papers, but last December they finally published their results in the journal Psychological Science in the Public Interest. It wasn’t pretty—or brief, at 72 pages. To sum it up: “There’s no compelling evidence that practicing these games leads to real world improvements in daily tasks,” says Daniel Simons, a psychologist at the University of Illinois and the lead author of the review.
Then the big guns came out. On June 22 of this year, the National Academies of Sciences, Engineering, and Medicine issued a report on cognitive training’s efficacy, based on a review of high-quality clinical trials conducted by the Agency for Healthcare Research and Quality. It concluded that cognitive training is “supported by modest but inconclusive evidence at present.”
These reviewers aren’t saying that brain training will never work. They’re saying that it’s hard to know for sure. Like so much social science research, cognitive training experiments have been poorly designed. They have small sample sizes, inadequate controls, and methods that are bad at evaluating whether far transfer actually occurs. The National Academies report concluded that there’s only been one decent study of cognitive training that found any kind of improvements—and they were decidedly mild. Plus, that study included socializing, workshops and in-person booster sessions. “That’s very different from sitting down at a computer and doing Lumosity.com stuff,” says Story Landis, former director of the National Institute of Neurological Disorders and Stroke and a panelist on the National Academies’ report, because scientists suspect that old people who socialize have better cognition. With low-quality games to test, and even lower-quality research, the scientists are still at an impasse.
That’s why Gazzaley built Neuroscape. When he first landed at UCSF, the cognitive neuroscientist treated people with Alzheimer’s and dementia, but quickly grew frustrated. “I was still telling stories about what goes wrong as we get older,” Gazzaley says, “but not really doing anything to help people.” He started looking for a way to treat aging-related cognitive decline, testing Alzheimer’s drugs on people with normal age-related decline. But the results were mixed, and the drugs had side effects. So instead of using a drug to take the reins of the aging brain, he decided he would try to use the greatest mind-alterer of them all: experience.
In 2009, he reached out to his friend Matt Omernick, a former art director at LucasArts who worked on Star Wars videogames like The Force Unleashed. Together they developed Neuroracer, an iPad car racing game that forces a player to multitask by driving while picking out specific road signs flashing on the screen. They ran experiments with young and old people, measuring how their cognition and brain activity changed after playing the game. Gazzaley published in Nature in 2013, presenting evidence that cognition in elderly players improved—though scientists pointed out problems with the study’s sample size, controls, blinding, and outcome measures. All of those things could have undercut Gazzaley’s early, promising results.
Instead, Gazzaley’s lab slowly morphed from a traditional research group into a sprawling collection of scientists—other neuroscience faculty, post-docs, technicians, and a team of game developers. In 2016, they formalized into Neuroscape, with a glitzy website and 10 technologies at different stages of development. With Omernick, Gazzaley founded Akili Interactive, a company in Boston that’s running a Phase 3 clinical trial on an updated version of Neuroracer—called Project: EVO ADHD—to earn FDA approval as a treatment to reduce the severity of ADHD. If they manage to show that the game is effective, it could be in the clinic as a treatment, and potentially as an alternative to Adderall, for doctors to prescribe by 2018.
Rocky’s videogame, called Body Brain Trainer, is just one of nine Neuroscape projects, from iPad games to full-body motion capture virtual reality. VR could be a good gaming-slash-treatment platform once Gen Z hits the golden years, but Body Brain Trainer seems more suited for Rocky. The gameplay is intuitive, using Microsoft’s Kinect to track the position of the players so there’s no controller or hardware attached to the body. And it capitalizes on one of the only sure things in health: Exercise is good for you, and the National Academies report released in June suggests that it could help prevent cognitive decline too.
But Body Brain Trainer is designed to give the brain and the body a workout. For the brain, the game forces the player to quickly focus on one item while ignoring others—not unlike the basic premise in the original Neuroracer game. On one level, the word Vegetable or Color flashes on the screen. Then a mangled-looking vegetable—half pepper, half carrot—appears. The player has to judge whether it looks more like a pepper or a carrot, or whether its color is closer to red or orange, depending on which word flashed earlier. The game uses closed-loop feedback software so the better you do, the harder it gets, instructing the player to run in place to keep heart rate at a set level and increasing the difficulty of the cognitive task as the player gets better.
The hope is that the game will lube up overlapping modules of the brain’s prefrontal cortex—the stuff responsible for thinking, decisionmaking, and remembering. Maybe even enough to initiate that unicorn-like “far transfer” to daily skills. Gazzaley is especially hopeful that combining exercise with brain training could multiply the benefits of brain training—and by partnering with game designers and building multiple games at once, he is trying to build the best brain-training games out there. But there’s still one major roadblock to improving his studies: the placebo effect.
Rocky says he doesn’t know if he’s in the control or experimental group. (In his study, the control group stays home to play word games and a tai chi game on an iPad.) But he certainly thinks something is working. By the end of his 21 sessions, he claimed he was better at remembering phone numbers and recognizing familiar faces.
But that doesn’t mean he’s right. Just thinking that your brain is being trained causes a placebo effect. Last year, researchers at George Mason University conducted a study using two different recruiting flyers: one that promised a cognitive enhancement study to improve intelligence and one for a generic study. Both groups got the same exact brain-training exercise. Those recruited with the cognitive enhancement flyer increased their IQ by five to 10 points—they were smarter! But the group that came in through the vanilla flyer didn’t change. “That’s the problem when you actively recruit and you tip your hand,” says Pam Greenwood, the cognitive psychologist at George Mason University who led the study. “You may get what you expect because of the type of person that you actually recruited.”
In Gazzaley’s lab, it’s hard to imagine a subject coming in without some kind of exposure to the research. Rocky learned about Body Brain Trainer at an event for UCSF donors, where Gazzaley lectured about his technology. The scientist is a regular on the pop science circuit, with his hint of a New York accent and off-the-cuff style, and Rocky was impressed: He approached Gazzaley to sign up as a study participant after the talk. Many of Gazzaley’s other subjects come to him through advertisements on the Neuroscape website, which clearly describes its goals.
“Gazzaley does not concern himself with this placebo issue,” Greenwood says. “You can’t even tell by reading his papers whether he actually recruits people and tells them what he’s doing.” One potential issue with participants knowing about the experiment is that some will come in knowing that they’re supposed to be getting smarter—and therefore work really hard—and others may realize that they’ve been assigned to a control group if they’re playing on an iPad at home instead of coming to the lab, and not keep up with the training. Differences in motivation could affect the results; in the Body Brain Trainer study, the technicians admitted that the young participants were much less enthusiastic than the older participants.
Gazzaley is trying to minimize the differences between experimental groups that could lead to the appearance of cognitive enhancement due to factors other than playing the game. And he wants to make sure that it’s really the combination of exercise and cognitive training that causes the effect. He plans to test two additional control groups—one that plays an iPad game without doing the body motions and one that plays an exercise-only version. In addition, they’ve double-blinded the whole study, so the people evaluating Rocky don’t know what group he was in, either.
But building a perfect control is still far out of reach. Gazzaley thinks having a plushly produced, immersive experience is crucial to changing the brain, because that’s the way experiences are in the real world. “You need to have high level products and interactivity. That takes years to build,” Gazzaley says. And then, essentially, you have to build it again. In the case of a brain training study, Greenwood says researchers could tell every participant that they should expect to get smarter, and then give them games that look exactly the same except for the key element of the experience.
For now, Gazzaley is actually counting on some participants’ savvy. “I’m fairly public,” Gazzaley says. “There’s no hiding Neuroscape’s mission.” To counter this, Gazzaley says he tells all participants that they are likely to experience cognitive benefits from his games. By telling everyone they should get smarter, he’s giving them all the placebo effect, potentially making the difference between his control and experimental groups smaller than it would be otherwise—a risk Gazzaley says he’s willing to take to find something that actually works. Plus, he says, “If this eventually becomes a treatment, they’re going to know that they’re taking it.”
In the brain-training world, scientists often compare the brain to a muscle. And now that Rocky’s two-month training session has concluded, he’s adopted the metaphor. “We can keep the brain more active, longer,” Rocky says. “There’s no reason for it to atrophy.” He believes, as many baby boomers do, that the real problem is that old people don’t learn new things, and that simply doing more brain “workouts”—whatever those might be—can fight aging.
“When I play the game, I’m mentally and physically tired,” Rocky says at the end of a session. He thinks a lot of other exercise regimens are boring. And now that he’s getting older, Rocky worries about his memory, especially when he’s learning a new song to perform. He’s noticed that he’s pretty good with the old Cole Porter songs, but that learning totally new stuff is harder than it used to be. “What I’ve told my friends is, this is the exercise of the future,” he says. Now that his training is done, a post-doc will analyze his before-and-after cognitive results and brain scans to see if and when his abilities improved.
But even brain training proponents admit they have only vague ideas of how a game could alter the brain. Far transfer might theoretically be possible: People who learn a musical instrument at a young age have better hearing abilities when they’re old, for example—even if they just had piano lessons for a few years. But it’s possible that a young person who plays piano had different abilities from the start. In 2015, Greenwood wrote a review asking cognitive training scientists to stop and think for a minute: “If we want far transfer—which is what I think we want—then how do we do that? What’s the theoretical basis for that? I don’t think we really know.”
If Gazzaley’s game gets the stamp of the FDA, it’ll be a powerful validation of the cognitive training field.
In the meantime, the clinical trial of Akili’s game to treat ADHD is barreling along, skeptics be damned. And if it gets the stamp of the FDA, it’ll be a powerful validation of the cognitive training field.
And it will be a powerful way to get doctors to sell brain training to patients. “If you want physicians to think about treating patients with anything other than drugs or surgery, you have to go through the FDA,” Gazzaley says. “Doctors don’t have the time to break down how strong the evidence is.” Brain-training companies already use the authority of health practitioners to sell their games, Simons points out, by promoting their games to therapists. Then the therapists suggest the games to their patients. But “unlike some brain-training proponents, Adam has been appropriately careful not to market a product until he has more compelling evidence,” Simons says.
Still, selling a product will always be the goal. In fact, Neuroscape is working with consumer brain-training companies already, receiving sponsorship from Posit Science and the game company Zynga and publishing scientific papers with Posit Science’s chief scientific officer Michael Merzenich. Posit Science aggressively touts its scientific legitimacy in press releases, even claiming that the recent National Academies report shows that brain training is supported by “more evidence of efficacy than any other intervention examined”—which Landis says is simply not true. If the current state of brain-training commerce is any indication, whatever games Neuroscape comes up with—consumer-based products or clinically approved treatments—even the weakest evidence of benefits will be enough to make a lot of money.
“I think we will unlock the potential to optimize our cognitive abilities and our emotional regulation in a way that we have never seen before. I do believe that. Is that overselling it? Who knows,” Gazzaley says. “I’d rather be proven wrong in 15 years.” To keep investments for research, the company, and the tech development flowing, Gazzaley must remain cautiously optimistic about brain training’s abilities. “Are we there now?” he asks. “No. I would never say that. But I do think that we can get there.”
If brain training doesn’t really work, it could mean consumers spend more time and money on videogames rather than on developing real-life skills they need. But if a well-designed study does manage to show that a videogame is capable of far transfer, it will be a huge breakthrough. For Rocky, it could mean that, 20 years from now, when he stands onstage in his white jacket and red bowtie, tapping his toes along with the hi-hat, he’ll remember all those lyrics to the songs he loves.