Mind Games

A Workout for the Brain Offers Hope for Kids With Autism

By Anita Harrison

Six-year-old Alex Harvey can’t wait for the game to start. Sitting on his mother’s lap, with two electrode sensors on his head and a sensor clip on each ear, he fidgets and fusses. “Mom, I want to say, ‘Go!’” he says. Finally, Guy McCormack, chair of MU’s Department of Occupational Therapy and Occupational Science, finishes with setup. Alex has requested Chomper, a game similar to Pac-Man.

“Are you ready?” McCormack asks.

“Yeah,” Alex says. “Go.”

The game begins and Alex’s fidgets stop. Relaxed against his mother, he moves Chomper through a simple maze. The on-screen graphics may be a flashback to the 80s, but the game is still incredible to watch. There’s no handheld controller, no joystick, no mouse. Alex makes Chomper move with his brainpower alone.

“It really runs on players’ attention,” McCormack says. “The better they pay attention, the faster it goes.”

McCormack is using the game to provide Alex with neurofeedback training as a therapy for autism, a condition Alex was diagnosed with when he was 3. Neurofeedback training is also called EEG biofeedback because it is based on the electroencephalogram. McCormack describes it as something like “a workout for the brain,” a mental exercise regimen that offers hope of significant improvements in the lives of thousands of autistic kids.

“The therapist acts as a personal trainer,” he says. “The neurofeedback equipment works like a ‘mind mirror’ that shows the child instantly how his or her brain is functioning.”

The electrode sensors on Alex’s head measure and amplify electrical impulses emitted from his brain’s neurons and send those signals to the computer. If Alex is paying attention, his brainwaves register in the targeted “alpha” range, and Chomper gobbles up the dots. If Alex is not well-focused, the sensors direct Chomper to become less voracious. 

If Alex is, for example, experiencing anxiety during the game, the sensors read spikes in fast “beta” waves. If he is bored, they show slow “theta” waves. In both instances Chomper becomes sluggish or even stops. “The basic principle behind neurofeedback is operant conditioning,” McCormack says. “The aim of neurofeedback is to enable the child to consciously control his or her brainwave activity by being rewarded for his ability to attend.”

Neurofeedback was developed by NASA to monitor and train pilots in flight simulations; in the last 30 years it has been explored as a therapy for numerous disorders, ranging from PMS to epilepsy, and even as a way to train athletes to “get in the zone.”

Remaining focused is particularly difficult for Alex and others with autism because of how their brains are “wired.” Although the causes of autism remain unknown, imaging studies have shown that children with autism often have what scientists call an over-proliferation of white matter in some areas of their brains and an under-proliferation in other areas. White matter carries messages between different parts of the brain. Too much results in extra connections; too little produces too few.

“Many times the children report that sounds, touch, and visual [stimulation] — and even eye contact — are difficult for them to process,” McCormack says. “In general, children with autism have a lot of anxiety and difficulty adapting to new situations.”

Alex’s parents, Greg and Kathy Harvey, signed him up for McCormack’s experiment with the hope that neurofeedback will “retrain” their son’s brain to perform better, thus improving his ability to pay attention and more effectively deal with everyday stress. Alex is on the high-functioning end of the “autism spectrum disorder,” often abbreviated to ASD. The phrase indicates how autism has varied manifestations, from hardly noticeable to extremely debilitating. The diagnosis of autism in the United States has increased tenfold in the last decade to 1 in 150 children, and the Centers for Disease Control and Prevention have identified autism as a national public health crisis.

In response to that crisis, MU in 2005 launched the Thompson Center for Autism and Neurodevelopmental Disorders, a facility providing diagnostic, assessment and treatment services for children, youth and young adults. In 2008 it served 1,736 individuals.

Researchers from across the MU campus are engaged in more than 30 autism-related research projects, and the Thompson Center is a partner in three major national research networks: the Autism Treatment Network, Simons Simplex Collection and the Interactive Autism Network.

McCormack’s project ranks as one of the more high-profile of these investigations. The Sinquefield Charitable Foundation, based in Westphalia, Mo., contacted the Thompson Center two years ago about funding a study on neurofeedback as a treatment for autism. McCormack, who came to MU in 2003 with almost 30 years of experience in academia and occupational therapy, agreed to take on the project.

“I like to investigate on the edge of science to demonstrate that interventions with clinical effectiveness have research significance,” he says. “This is where much breakthrough research is accomplished.”

He began his pilot study in summer 2008 with eight boys ranging in age from 6 to 16, each of whom was on the high-functioning end of the autism spectrum. All the subjects were male not by design but because autism is much more prevalent in the male population. McCormack says the study did not include children with more severe autism because neurofeedback training tends to require more cooperation and focus than these children can give.

The study followed a single-subject, multiple-measure design. To get a “before” picture of the children, McCormack used a variety of assessments. He weighed the children on a standardized social responsiveness scale developed by John Constantino from Washington University in St. Louis. The 65-item rating scale, according to a description by Constantino published on the website of the Western Psychological Services company, “measures the severity of autism spectrum symptoms as they occur in natural social settings.  Completed by a parent or teacher in just 15 to 20 minutes, the SRS provides a clear picture of a child's social impairments, assessing social awareness, social information processing, capacity for reciprocal social communication, social anxiety/avoidance, and autistic preoccupations and traits.”

McCormack next gave the children a “test for variables of attention.” This involved asking the kids to watch a box displayed on a computer screen. A small black square constantly popped up in different areas. When the black square appeared at the box’s top, the children were instructed to press a handheld clicker. The computer scored the children’s reaction times in microseconds, while also recording the times they clicked too soon, clicked in error or failed to click at all.

Finally, McCormack had parents fill out questionnaires that measured the children’s overall behavior and sleep patterns, along with the frequency of their autistic behaviors at home.

McCormack received neurofeedback software training through EEG Spectrum, a company that sells its own neurofeedback system. He also studied the literature on neurofeedback and autism to determine the best protocol, such as where on the children’s heads to place the electrode sensors. He found the experts were not in agreement. In fact, they weren’t even close.

“It was a little frustrating to me because I got on a conference call with people who were identified as the gurus in neurofeedback, and they couldn’t agree with each other,” he says. “They were all getting some results, but they didn’t agree with each other.”

After experimenting with different placements, McCormack settled on locating the sensors over the frontal lobe, a part of the brain involved with reasoning, emotions and judgment. Placement of the sensors matters, McCormack says, because one of the goals is to get neurons in specific areas of the children’s brains to fire in synchrony with each other. Placement is also important, he adds, because it is possible to be training one area of the brain to attend to various stimuli while other areas continue to function at lower and higher frequencies.

Upon arrival at the lab, a nondescript room in the basement of MU’s Clark Hall, the children are wired up to receive a 15-minute training session. There are two computers: One displays the game, and one displays the children’s electroencephalograms. McCormack uses the electroencephalograms to monitor the children’s theta, alpha and beta wave levels; if the game slows, he can see whether the problem is anxiety or boredom. He can also use the EEG computer to change the game parameters, allowing the children to succeed with less focus if the game is too hard or requiring them to focus more if the game is too easy. 

The children can choose from several games, including the aforementioned Chomper and a space race that challenges them to keep a middle rocket (representing their alpha brainwaves) ahead of two competing rockets (one representing theta waves, the other beta waves). As the games progress, sensors provide a continuous flow of information to the computer, with updates occurring multiple times a second and the results showing up in the game’s performance. That flow of feedback is what makes neurofeedback unique, McCormack says.

“It’s kind of like having someone say, ‘Nice job, Johnny,’ thousands of times in a second,” McCormack says. “A teacher or therapist could not offer feedback that continuous.”

Though the game “tells” children how their brains are performing, neither it nor McCormack ever offers instructions on how to get their brains into the targeted alpha zone. Why this happens is something of a mystery to the nation’s autism researchers.

“Somehow the brain is able to figure out what it needs to do and can make it happen,” says neuropsychologist Robert Coben, a licensed psychologist in New York and one of the most well-known practitioners of neurofeedback for people with autism. “How the brain is doing it is a mystery, but almost anyone can do it. It is rather amazing that a person can learn to control a specific part of their brain in one or two minutes, but it happens.”

McCormack goes back to operant conditioning: the training changes behaviors through a system of rewards and penalties.

“The children feel differently when their brains are operating in different ways,” he says. “If they see it on the screen and they link that up with how they feel, they can almost consciously bring that back at another time.

“Ultimately, the control falls back on them, and that’s the beauty of it,” he says. “They gain ways to recognize when their brain is working well, and otherwise, they wouldn’t know that.”

However, operant conditioning is not all that McCormack thinks is going on during neurofeedback training. He, along with Coben and others, also theorize that neurofeedback creates physical changes in the brain.

“It’s based on neuroplasticity,” McCormack says. “Where you put the sensors, it increases the metabolism of the brain — increases the blood supply and oxygen — so you get more sprouting of ‘dendrites,’ or neuron branches, in those areas.

“It gets pretty dense in the neuroscience at the molecular level,” McCormack says, but the basic idea is that neurofeedback not only retrains but also rewires the brain, laying down new neural pathways that result in improved focus, attention, social interactions, sleep and appetite.

Whether neurofeedback can truly deliver those results for any neurological disorder is still up for debate in the scientific community. The effects of neurofeedback on Attention-Deficit Hyperactivity Disorder (ADHD) have been studied far more than those on autism, and though practitioners swear by the treatment, some scientists still question its value. Dr. Russell Barkley, a clinical professor of psychiatry at the Medical University of South Carolina, is one of the most prominent skeptics of using neurofeedback for ADHD. He says he remains unconvinced because there have been no rigorous studies showing the treatment is effective.

 “We have a lot of papers claiming an effect,” he said, “but they didn’t include either random assignment or placebo. Without those controls, we don’t know what the results are due to.” He adds that there have been a couple of “scientifically rigorous” studies, and they “haven’t shown much.”

“I would describe biofeedback as experimental at best and disproven at worst,” he says. “It should not be offered to the public as though it worked.”

McCormack does not claim his study will end debate about biofeedback’s effectiveness. But as a pilot study, he says, it is “rather promising.” After 20 sessions, five of the eight children showed measurable improvements and two showed no change (one, who got worse early on, did not complete the 20 sessions). On the “test for variables of attention,” the children averaged a 21 percent improvement on reaction time, an 18 percent improvement in response sensitivity, an 11 percent decrease in hitting the button too soon and a 28 percent decrease in omission errors.

“That test has really been remarkable because it’s very objective,” McCormack says. “It’s not like the self-reporting tests that are filled out by the parents, and what we saw were some great gains.”

On the standardized social responsiveness scale, five of the eight children improved by 20 percent; the others showed no significant change. Meanwhile, qualitative data pulled from the parent questionnaires were highly positive; parents reported their children were less anxious, more flexible and better able to control their emotions and behaviors.

“This one little boy, he would run from wall to wall, literally bouncing off the walls; that stopped,” McCormack says. “One boy would take hot showers to calm himself down, and he stopped needing to do that. That was interesting. Two of the boys got their driving permits. One finished a scuba diving certificate. Some of the outbursts subsided.

"One of the older boys said he started dating a cheerleader.” McCormack laughs. “Things are looking up for him, I guess.”

As for Alex, his mom says the effects have been “huge.”

“He’s more relaxed,” she says. “The school reported within the first two weeks that he finally started talking to them and was actually doing the work. He’s much more flexible, too. It has improved his ability to sit still and attend to task enormously. “I’d gotten the impression before that the school was hanging on by their fingernails to keep him for three hours, but this year, he’s been doing wonderful.”

McCormack beams at this report but later tries to put it into perspective. “I don’t want to claim too much,” he says. “The preliminary data look promising, but more research needs to be done. So far, though, it seems to have made a difference in the lives of some of the children.”