Could this computer game delay Alzheimer’s symptoms? New study suggests it could.

You’re staring at a computer screen.

Two objects flash before your eyes, one directly in front of you, and the other off to the side, barely in view. In a split second, they’re both gone. Now the computer asks you: What were they? Where were they? Did you get a chance to see them both?

If you answer correctly, don’t relax yet. The next level will be harder.

But whatever you do, don’t give up.

A new analysis of previous research data announced at the Alzheimer’s Association International Conference this week tentatively suggests that this kind of game could decrease the risk of symptoms of dementia by almost half, compared to not having any brain training at all. The study presented is under peer review and hasn’t yet been published. (Studies can change dramatically from the conference setting to the pages of a journal, reminds PLOS blogger Hilda Bastian, so the findings should be considered preliminary for now.)

The game is called a speed-of-processing task. It’s one of three types of cognitive training that 2,800 people took part in during the Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) study, a randomized longitudinal study funded by NIH. The participants averaged 74 years old when the trial began. Scientists tracked them for 10 years, hoping to find out how cognitive training could impact the functioning of healthy older adults.

The participants were split into four groups. One played the speed-of-processing games, and two other groups took a memory or reasoning class. The last group did nothing, and served as a control. The memory classes taught tricks for memorization, like mnemonic devices or “methods of loci,” which is a tool to remember a series of objects by visualizing each one in a different physical location. The reasoning course taught logic and pattern recognition, and trained people how to choose the next letter in a series, based on the order of the ones that precede it.

The most recent ACTIVE paper was published in 2014 and concluded that the different cognitive training could in fact help a little with certain basic tasks, like driving or balancing a checkbook, as people got older.

Monday in Toronto, a research team led by Jerri Edwards at the University of South Florida announced that they had used the wealth of data from the ACTIVE study to ask a different and more provocative question: Could cognitive training delay the onset of dementia or cognitive decline related to Alzheimer’s?

Their findings showed that the group that completed 10 to 14 hours— that’s total, over 10 years — of the speed-of-processing games were 48 percent less likely to have developed Alzheimer’s or other forms of dementia, compared to those who received no brain training at all. These participants did 10 hours of game play in the first year of the study, and then were randomly selected to receive booster sessions up to four more hours throughout the rest of the trial.

“We believe this is the first time a cognitive training intervention has been shown to protect against cognitive impairment or dementia in a large, randomized, controlled trial,” Edwards said, in a news release.

The result caused a stir of cautious curiosity at its presentation, said one attendee, Penny Dacks, a neuroscientist and the director of aging and Alzheimer’s prevention at the Alzheimer’s Drug Discovery Foundation. It raises some provocative questions: Why did speed-of-processing show the strongest correlation and not, for example, the memory classes? And could it really be possible that only 10 to 14 hours had such a large effect, years later?

“I think it’s really exciting,” Dacks said in an interview with The Washington Post. “For one thing, it shows us that not all cognitive training is equal. This is not going get us all the way, but if it could help even a fraction of the population, I think that it should be applauded. And certainly pursued with more research.”

While the study shows a correlation rather than causation, the researchers claim it is the first evidence that could suggest a brain training game could alter the onset of dementia.“That’s a spectacular finding,” Susanne Jaeggi, the director of the Working Memory and Plasticity Laboratory at the University of California, Irvine told Dan Hurley in a New Yorker article. “We didn’t have any evidence that computerized training had any preventive effects on dementia. You could argue that this study provides evidence that it is possible.”

But even with this compelling ACTIVE data analysis, researchers are hesitant to accept the findings right away. There’s good reason to be a little skeptical. Beside the fact that it’s still under review, the 2014 ACTIVE study wasn’t designed to track dementia, Jonathan King told the Washington Post. He was the project director and co-author of the 2014 study and is a program director at the National Institute on Aging. Their goal was to see if they could help aging healthy adults achieve better functioning through the different kinds of classes and games.

And overall, they did. Those who took reasoning classes got better at reasoning. Those who played speed of processing games got better at driving, probably due to increasing their visual field. Edwards and a co-author on the new study, Lesley Ross, both worked on data collection for the ACTIVE study. When they saw that these games could positively affect the adults’ functioning they knew they wanted to take a closer look at the data.

“Given that one of the primary definitions of dementia has to do with both the person’s cognitive status but also the person’s functional status, that suggested to us that maybe we should go ahead and look at this again,” Ross told The Washington Post.

They reanalyzed the data, and checked how many people in the four groups had gotten dementia or experienced other disease-related cognitive decline. That’s when they discovered that those who had been in the speed-of-processing group were correlated with lower instances of dementia and Alzheimer’s.

“This finding is one of the first to say hey, there is something else going on here,” Ross said. “But we need more understanding. I don’t think anyone would make the argument that this is a cure.”

Researchers have been trying to find ways to prevent the cognitive decline and dementia that accompany diseases like Alzheimer’s. About 5.2 million peopleover the age of 65 have Alzheimer’s in the United States, and there’s no cure for the disease or its cognitive side effects.

In the past 10 years, there has been a boom in apps and online games that claimed to help, despite tenuous scientific backing. Lumosity, an app that claimed to improve cognition, memory and brain function, launched in 2007 and had 70 million users as of 2015. They charge $299.95 for lifetime access and state on their website that they work with “100+ independent researchers to investigate new areas in cognition and cognitive training.”

But in 2014, the Federal Trade Commission fined Lumosity $2 million because they “preyed on consumers’ fears about age-related cognitive decline, suggesting their games could stave off memory loss, dementia, and even Alzheimer’s disease,” said Jessica Rich, director of the FTC’s Bureau of Consumer Protection, in a news release.

At a meeting held by the Stanford Center on Longevity and the Berlin Max Planck Institute for Human Development in 2014, a group of over 70 scientists signed a statement to inform the public that there was no proof these kinds of games could benefit them in such a profound way.

“Many scientists cringe at exuberant advertisements claiming improvements in the speed and efficiency of cognitive processing and dramatic gains in ‘intelligence,’ in particular when these appear in otherwise trusted news sources,” the statement reads. “In the judgment of the signatories below, exaggerated and misleading claims exploit the anxiety of adults facing old age for commercial purposes. Perhaps the most pernicious claim, devoid of any scientifically credible evidence, is that brain games prevent or reverse Alzheimer’s disease.”

After it was released, another group of scientists published a rebuttal, saying that while they agreed with many points of the first statement, they didn’t believe that there had never been any sign that brain training could improve cognitive function.

Both stances have remained controversial since, Dacks said. What isn’t controversial, Dacks added, is the knowledge that people who have receive more education or who have intellectually engaging careers seem to exhibit a later onset of dementia and Alzheimer’s symptoms. But she’s not sold on the proposition that a game, specifically speed-of-processing, is helping the brain become “smarter.” Instead, she theorized that if the findings are verified and replicated, it might be because of an indirect result from the game: a prolonging of independence.

For example, the 2014 ACTIVE study showed that another benefit of speed of processing games was better driving and maneuvering skills.

“I wonder if this could be leading to changes in a person’s life with a reduced risk of accidents both in the car as well as walking around that allow them to continue to be independent for longer, and that in turn, improve their function,” Dacks said. “There’s a premise in neuroscience of use it or lose it. So perhaps by enabling better independence over time or protecting from decline, maybe this is where it’s coming from.”

Ross told The Washington Post that she suspects another factor, neuroplasticity, or the ability of the brain to form new pathways, may be playing a role.

“One of the things that makes the speed training a little different than the others is that it’s adaptive, meaning that if you get better at it, the game becomes harder,” she said. “And there’s been a growing area of research looking into that neural plasticity. Things that are speeded, timed, and adaptive, are tending to show more promise.”

There’s also the remarkable fact that participants only played the game for a total of 10 to 14 hours, all within the first half of the 10-year trial.

“It’s hard to understand how such a brief intervention could have a long-lasting impact,” Dr. Howard Fillit, executive director of the Alzheimer’s Drug Discovery Foundation, told Stat News.

It would be as if you exercised for 10 hours, 10 years ago, and remained in shape today. Dacks speculates that once the participants played the game, they continued to practice speed of processing techniques in their daily lives. Their use of speed-of-processing skills didn’t end with the game, it was applied to everyday rituals. But still, if challenging the brain in this way for 10 hours could really lead to drastic results, the next step is asking what could happen if they upped the “dose.”

“Only 10 hours did this,” King said. “Would 50 hours be five times as good?”

Ross said much more has to be done, but the findings might provide, finally, a clue as to what kind of brain training needs to be researched at a deeper level. After their study is published, Ross wants to next study a group of unhealthy adults who are at higher risk for dementia and Alzheimer’s, to see how speed-of-processing games affect them. Remember, speed-of-processing games will never be a fix-all but could be a resource for all those out there who want to take preventative action, whether they are at risk for Alzheimer’s or not.

“It’s just an extra tool in the battery we can use against cognitive decline,” Ross said. “My goal is to be able to go to my parents, who are in their 70s, and say here is this thing we have evidence for. I think you should start doing some of these.”