JUDY WOODRUFF: This weekend, the crossword puzzle marks its hundredth birthday. First published in a NewYork newspaper, it's become a daily ritual for many and even been rumored to help stave off dementia. But there isn't much evidence to back up that claim.

On the other hand, special correspondent Jake Schoneker reports on new research using video games to sharpen an aging mind.

JAKE SCHONEKER: Fifty-seven-year-old Ashley Wolff has been a self-employed children's author and illustrator for 25 years, working out of her small home studio in San Francisco. She says she loves being her ownboss, but that working from home can be a challenge.

ASHLEY WOLFF, author and illustrator: Working from home allows me really to just let my attention deficit problem fly.

JAKE SCHONEKER: Like many people her age, she's recently found herself forgetting things more often, andgetting easily distracted from work. She was worried about these problems getting worse as she got older, especially because her mother, at age 85, was beginning to exhibit signs of Alzheimer's disease.

ASHLEY WOLFF: My sister and I are watching our mother kind of lose her memory. And we thought, gee, wow,she always did the New York times crossword puzzle, and always seemed so sharp, and here she is, none of that helped her. And we thought, we should try something.

JAKE SCHONEKER: So she decided so try cognitive training, a new breed of video game designed to exercisethe brain. She now gets a daily reminder to log on to her laptop for a 15-minute mental workout.

ASHLEY WOLFF: She wants a BLT and coffee. He wants a garden salad.

JAKE SCHONEKER: In this game, called "Familiar Faces," Wolff is a waitress at a cafe. She has to remember the names of people who come in, as well as what they order.

ASHLEY WOLFF: Cheeseburger.

JAKE SCHONEKER: "Familiar Faces" is one of many games designed by Lumosity, a San Francisco based startup that launched in 2007. The company markets itself as a kind of gym for the brain, complete with monthly membership fees.

The service creates a personalized training program for users based on their needs that includes exercisesfor attention, speed, and memory. On most days, Wolff will play five different games selected for her by the program. Even though her memory isn't what it used to be, she says she's noticed a modest improvement since she started training. Increasing evidence from the field of neuroscience suggests it's never too latefor the brain to change.

Companies like Lumosity have built a billion-dollar business out of a very simple premise: that no matter your age, you can improve your brain's performance through cognitive training.

Joe Hardy, the head of the science team at Lumosity, says the idea of being able to improve and train the brain as we get older is relatively new.

JOE HARDY, Lumosity: Previous to maybe 30 years ago, neuroscientists believed that the brain was effectively fixed in its ability to process information, pay attention, plan, remember. We now know that the brain is constantly changing the way that it operates in response to the challenges and activities that it's engaged in.

JAKE SCHONEKER: The concept is called neuroplasticity, meaning the brain continues to adapt, change, andperhaps be trained even as we age. It's the underlying foundation for the many startup companies that are developing brain fitness programs and bringing them to market. But are the claims of these companies supported by the science?

LAURA CARSTENSEN, Stanford University Center on Longevity: The claims being made by most of these companies that are selling products to improve your brain are exaggerated.

JAKE SCHONEKER: Laura Carstensen is director of the Stanford University Center on Longevity, an expert and author on the aging brain.

LAURA CARSTENSEN: When marketers tell you that you can increase your I.Q. by 20 points or that you can, with 15 minutes a day of training, improve your cognitive control and executive functioning, and that thiswill make you smarter in everyday life, those kinds of claims are just clearly unsupportable by scientificevidence.

JAKE SCHONEKER: She says brain games have potential, but more research is needed to understand if, and how, the brain benefits from training.

At the University of California, San Francisco, Adam Gazzaley is one of many neuroscientists working to answer those questions.

ADAM GAZZALEY, University of California-San Francisco: Well, I think that, very frequently, there's a mismatch between being based on science and claims being validated by the scientific method. And I think that's what the real goal, is that they're more than based on science; they're validated by scientific methodology.

JAKE SCHONEKER: So Gazzaley and his research team set out to design a game that started with the science. They chose a brain function known to decline with age, multitasking, and tried to find a way to slow that decline through training. The game they came up with, called "NeuroRacer," challenges subjects to drive a virtual car down a winding road while simultaneously recognizing and responding to road signs.

MAN: I'm dangerous here. This is -- I'm a menace to society here.

JAKE SCHONEKER: Like commercial games, this game gets progressively harder as the player improves, creating a challenging virtual environment for the brain to adapt to.

As subjects trained in "NeuroRacer," they improved dramatically at multitasking. Study participants in their 70s and 80s who trained for one month performed better than 20-year-olds playing for the first time. But perhaps the most interesting part of the study was that those older players improved in other areas as well, like working memory and sustained attention.

That's a big deal because there's evidence that training in one task can lead to benefits throughout the brain.

ADAM GAZZALEY: This is a measure of what we call functional connectivity, which is a reflection of how your brain functions as a network, so that different parts of the brain are not acting in isolation, but acting as a network. And that's what we see, that that improves as well.

JAKE SCHONEKER: Gazzaley's study made a splash on the scientific community, making the cover of the science journal "Nature."

Gazzaley is now working to build a better, more interactive version of "NeuroRacer" that the FDA could approve as a therapy for ADHD. But that pathway could take years or decades to complete. And, until then, Gazzaley says he can't make any strong recommendations for the use of cognitive training.

ADAM GAZZALEY: We do need better, more carefully controlled studies in order to make really strong prescriptive advice.

That being said, in general, I think if you find these games fun, at least there's no clear evidence that they have detrimental effects, so I usually don't disrecommend them.

ASHLEY WOLFF: I think the proof will be in the pudding. It's not going to happen to me now, when I'm in my 50s. But if I'm still able to do stuff like this in my 80s, I will be thrilled.

JAKE SCHONEKER: By then, in 30 years, who knows what science will tell us about how middle-aged people like Wolff can keep their minds sharp.

But with five million Americans suffering from Alzheimer's disease today, and with that number due to risesharply in the coming decades, those solutions can't come soon enough.

ASHLEY WOLFF: Good brain training for the day.