Robin Hansen was hoping to help her son pass a community-college baking course when she heard about Cogmed software.
Hansen paid a psychologist $1,500 for the five-week training program, sold as a learning and memory aid. Her son, diagnosed with attention deficit hyperactivity disorder and dyslexia, worked 50 minutes, five days a week, on Cogmed’s memory tasks.
The psychologist detected gains from Cogmed, but they weren’t obvious “and didn’t translate into any real-world skills,” said Hansen, a design engineer in San Francisco. Her son, Warren Leon, failed the baking class, denying him the certificate he wanted to become a pastry apprentice.
“I felt devastated,” said Leon, now 23. Two years after the Cogmed course, he is attending a school for adults and children with learning difficulties.
As U.S. sales of Cogmed are increasing, so are questions about the effectiveness of the software, sold by U.K.-based publisher Pearson Plc through psychologists and to schools.
At least three studies or multi-study overviews published in peer-reviewed journals since 2012 have called Cogmed ineffective for ADHD. A fourth review included Cogmed with other products, and found the group ineffective.
Some 6.4 million U.S. children have been diagnosed with ADHD, defined as persistent inattention or hyperactivity that impairs learning or socialization.
The recent research suggests that some parents and schools using an emerging array of “brain training” programs to battle the condition are finding disappointment, as the Hansens did.
On its website, Pearson says Cogmed is “proven to increase working memory, which underlies attention, behavior, and the capacity to learn.” The company lists dozens of published studies on Cogmed on the site -- both pro and con -- and cites its own 2012 analysis of research to back its claims.
Working with psychologists, “80 percent of Cogmed end-users report they experience improvements in inattention by 30 percent on average,” the company said in a statement. “Cogmed has therapeutic utility for persons with ADHD,” the statement said. Lynne Baldwin, the psychologist who worked with Warren Leon, declined to comment.
Cogmed is part of a $1 billion brain-training market, according to SharpBrains, a San Francisco research house that forecasts growth to $6 billion by 2020.
Most of the purveyors, including the widely advertised Lumosity website, are oriented toward general consumers for self-help such as memory improvement. Part of the market, occupied by Cogmed and others, caters to psychologists, families and schools dealing with learning disorders and attention-deficit problems.
One of the latest studies in the field said that Cogmed should not be used to treat ADHD children. The study, led by Anil Chacko, a psychology professor at Queens College in New York, appeared in the March issue of the Journal of Child Psychology and Psychiatry.
Chacko’s look at 85 children had “statistical power,” was 50 percent larger than most previous studies and “marks a significant point” in research on the product, according to an editorial by Susan Gathercole, a cognitive psychologist at Cambridge University in England.
Eleven percent of U.S. kids between 4 and 17 had an ADHD diagnosis in 2012, up from 7.8 percent in 2003, according to the latest data from the U.S. Centers for Disease Control.
At least half of the children with ADHD have other learning or developmental problems, making treatment more difficult, according to Susanna Visser, who studies the attention condition for the CDC.
As ADHD cases multiply, so does interest in potential software-based treatments, according to SharpBrains. It estimates that sales of programs aimed at learning disabilities, including Cogmed, will triple to about $600 million in 2020 from $210 million in 2013.
Pearson declined to disclose sales of Cogmed, other than to say they are growing in the U.S. The company, which publishes the Financial Times, competes with Bloomberg LP, the parent of Bloomberg News, in providing business news and information.
The standard treatments for children with ADHD include therapy to modify behavior and stimulant medications such as Ritalin. Not all children benefit from the drugs and there has been debate over their side effects, potential abuse and over-prescription. Psychologists say Cogmed should be used as a supplement to other treatments.
“At no time should Cogmed replace medication,” said Sheldon Kaplan, a psychologist in Jacksonville, Florida, who has used Cogmed with patients. Chacko said his study took into account whether children were on medication or not -– about 30 percent were -- and found no difference in results.
Cogmed runs like a series of video games, without the flashy design and rapid-fire action of Grand Theft Auto and its ilk. Accompanied by a cartoon-robot guide, users have to blast moving asteroids in a certain order, whack green aliens or remember the sequence of blinking lights on a cube rotating in outer space. A narrator praises users when they do well, and the game gets harder. After mistakes, it gets easier.
Parents are urged to offer stickers or treats to get children to complete the training. Psychologists can monitor the patient’s progress in real time online.
“It’s like a gym for the brain,” said Heather Goldman, a psychologist in New York City who charges $2,250 for the training, including weekly meetings with patients.
Cogmed’s sales will grow to more than $15 million this year, which will make it the largest brain-training product aimed at learning difficulties, overtaking Scientific Learning Corp.’s Fast ForWord software, according to Alvaro Fernandez, the chief executive officer of SharpBrains. Scientific Learning declined to comment.
Sellers of these products say they promote neuroplasticity, the ability of the brain to change -- for better or worse -- over a lifespan. Neuroplasticity is real, but not all of the claims made in the marketplace are, some research scientists said.
“It’s such a buzzword right now,” said Tom Redick, a psychology professor at Purdue University who studies working memory. He said he isn’t involved with any commercial products.
Pearson bought Cogmed in 2010 from Sweden’s Karolinska Development AB, a commercial arm of the Stockholm research institute of the same name. The price was less than $100 million, according to J. Travis Millman, vice president for new business and innovation at Pearson’s clinical assessment group.
Though that’s relatively small for Pearson, whose sales last year were 5.18 billion pounds ($8.72 billion), Cogmed fits with the company’s strategy of moving into digital products and services to reduce reliance on printed media.
The shift is especially important in the U.S., where declining college enrollments have squeezed textbook sales. North American education accounted for about 55 percent of Pearson’s revenue and 61 percent of operating profit in 2013.
The company gets 13 percent of its profit from testing for K-12 students, including assessments used to measure ADHD symptoms, according to Tim Nollen, an analyst at Macquarie Capital USA in New York. Cogmed is sold both by Pearson salespeople that market clinical assessments to psychologists and by those that offer educational products to schools, according to Mark Yaphe, general manager for the product.
The U.S. is Cogmed’s fastest-growing market, according to SharpBrains. Pearson sells licenses for the software to between 500 and 1,000 psychologists in the U.S., for prices of about $100-$200 per patient, practitioners said. They said they charge patients between $650 and $2,250 for Cogmed training, including testing and supervision services.
Pearson’s Millman said the company has stepped up marketing to schools in the past year. Schools pay a price per student similar to the licensing fee for psychological patients, Millman said.
Cogmed has drawn research interest partly because it targets working memory, which is like a mental jotting pad needed to remember and manipulate information for performing complex tasks such as math problems or analytical reasoning.
Working memory is regarded by psychologists as essential for academic success. Pearson’s meta-analysis in 2012 covering research until then found average gains of 26 percent in tests of visuo-spatial working memory and 23 percent in verbal working memory.
Chacko said these measurements don’t reflect true working memory. Children in his study showed no improvement in reading, math or spelling. “Our data suggest that Cogmed really only improves the more basic, and arguably less important, short-term memory on tasks that closely resemble Cogmed training tasks,” Chacko said.
Scientists call this divide the difference between near transfer and far transfer.
To explain the terms, Zach Shipstead, a professor at Arizona State University who published a critical review of Cogmed in 2012, used an analogy to music. Someone who trains on the piano might get better on the organ -- an example of near transfer, he said. To see whether they gained more broadly and achieved far transfer, “you should be testing them on a drum set or a tuba,” he said.
When it comes to showing whether Cogmed users attain far transfer, “the research is less mature in that area,” Pearson’s Millman said.
Psychologists and school officials who have adopted Cogmed said in interviews that surveys of parents and their own impressions of patients after training showed the program works.
Some scientists say these reports are questionable because there is a well-known bias at work when evaluators know their subjects received the training being studied. For that reason, rigorous studies often keep evaluators in the dark about which study subjects received a treatment, and which didn’t.
Mark Rapport at the University of Central Florida found that “unblinded raters reported significantly larger benefits” of working memory programs, including Cogmed, in a review published last year of 25 research papers.
“People have a tendency to look for behaviors that confirm their expectations,” Shipstead said, especially if they are working with the child or have paid for the training.
Cogmed’s website features testimonials from school special-education staff who have tried the program. In at least two of the cases cited, the districts where Cogmed was used have discontinued it.
Rhonda Cunningham is quoted on the site saying Cogmed “was very, very successful” in an experiment with five learning-disabled children in the Natalia Independent School District in Texas. Cunningham, formerly a psychologist in the district, wrote her doctoral dissertation at Walden University, an online college, on a dozen Natalia special-education students who later used Cogmed.
The dissertation reported results similar to Chacko’s -- improvement on some kinds of working memory but no change in academic performance. In an interview, Cunningham said three-quarters of the dozen students made academic gains that came several months later, too late for the dissertation.
Natalia has no further plans to use Cogmed, said Melissa Cortez, the district’s special-education facilitator. She said she didn’t know whether the program worked.
Others say ADHD students using Cogmed in combination with other strategies have improved. Ninth graders in the Digital Literacy class in Bay Village City Schools, near Cleveland, typically average a sixth-grade reading level when they start the school year and improve by three grade levels by the end, according to Barbara Marsh, who teaches the course.
About seven or eight of the 30 kids in the class, which uses Cogmed and other interventions, have an ADHD diagnosis. Each year, about half of the ADHD students improve to normal level, according to Richard Bogielski, the school psychologist.
“I can’t say we could draw a causative link” to Cogmed, he said. “It’s one of the ingredients.”
Goldman, the New York psychologist, said she trains about 20 children a year on Cogmed and finds it difficult to attribute their improvements solely to the program. Between the therapy, medication and online training used on her patients, “it’s hard to tell what’s what,” she said.
In San Francisco, meanwhile, Warren Leon said he’s looking forward to finishing the school year. Once it ends in June, he said he’ll spend summer vacation with his family before figuring out what to do next. He has no plans to try to take the baking course again. “I’m afraid of having a repeat performance,” he said.