There’s personalized education. And then there’s precision education.
The use of genetic information in health has opened vast new areas for medical research and treatments in the past decade, along with questions about how personal genetic information will be used and who will benefit. And debates over those potential benefits and concerns are starting to enter the education field.
The New Haven, Conn., school district is working with a team of education, genetics, and neuroscience researchers from Yale University in what may be the first attempt to design so-called “precision” gene-based education help for an academic disorder, dyslexia.
The controversial $20 million project is supported by the nonprofit Manton Foundation. As part of the project, more than 450 New Haven students who entered school with literacy scores in the bottom 20 percent were given four years of two intensive, widely used reading programs, Reading Recovery and Empower, to provide at least an hour of supplemental support five days a week each school year. But the researchers are not evaluating Reading Recovery or Empower. Instead, near the end of the study, the students spit into a test tube, and researchers sequence the students’ full genome to look for differences between the students who responded to the intervention over the years, and those who continued to struggle in reading.
In the longer term, studying phenotypes—the interactions between genetic variations and environmental influences—may allow scientists to find the underlying mechanisms that cause dyslexia and develop more tailored approaches to correct it. In the nearer term, identifying specific variations of genes associated with a disorder may be used to create a tool to screen children for potential risk of developing dyslexia long before they start school, and allow for earlier reading interventions.
“Twenty-five percent of children don’t respond to high-quality interventions now,” said Yale pediatric geneticist Jeffrey Gruen, who leads the New Haven study. “The idea is that if you could come up with a panel that could identify a child at risk [of dyslexia] long before they got to school, you could track them and intervene at a very early age.”
An Earlier Red Flag
That timing is important. Many schools already use behavioral screening tools for dyslexia—including a well-known one developed by fellow Yale dyslexia researcher Sally Shaywitz—yet only 18 states require universal screening for the disorder. An estimated 1 in 5 school-age children have the reading disability, but federal education statistics note that little more than 3 percent of such children receive special education services for any specific learning disability (including dyslexia).
A significant percentage of students with dyslexia go undiagnosed until the late elementary or secondary grades—a challenge, Gruen said, because earlier intervention seems to be key. Prior studies have found about 25 percent of students below 4th grade still don’t respond to otherwise highly effective reading interventions, and that ratio flips to 75 percent for students who are not identified until later grades.
Rebecca Tabak Roberts, a 2nd grade teacher at the Strong 21st Century Communications Magnet School in New Haven, was one of those students as a young child. She was bright and found teachers reluctant to refer her for testing even after her mother repeatedly raised concerns; she was finally diagnosed in late 4th grade.
Tabak Roberts’ school is participating in the study, and in the past two years she has seen students in her class improve as a result of the interventions, but she said she hoped the genetic research would also bear fruit.
“I think that this can end up helping students in the long run. I really do,” she said. “If this comes out to have a valid test to link our DNA and dyslexia, you better believe my children are going to end up taking that test before they go to preschool so they can get the attention they deserve.”
Risk and Reward
Dyslexia, the most common learning disability, affects how students process and link spoken and written language. Educators and researchers have known for nearly a century that the reading disorder can run in families, but in the last 15 years, scientists have uncovered much more of the physical evidence for such a link.
“We are several steps behind where precision medicine is,” Gruen said. “We’ve found maybe a dozen or so genes associated with reading, but ... we’re just now starting to identify the variants [of those genes] that confer risk and starting to see enough fluid readers and non-fluid readers to estimate those risks” of developing dyslexia.
In part, genomics has been applied to education more slowly because it’s more difficult to get access to genetic material for education studies than for cancer research, and there can be even more complex interactions between genetic and environmental influences in education than in medicine. But Gruen said, “probably the bigger barrier is we’re all sensitized to the idea of using genetics [in education], because people have misused these terms in ways that are totally inappropriate.”
In particular, such research raises the specter of eugenics-based racial discrimination in schools and stories like those of Henrietta Lacks, a black woman in Baltimore whose cancer cells were used without her consent to create HeLa, one of the foundational stem-cell lines in research.
While the New Haven study includes students of different racial groups, it draws students from the overwhelmingly black and Latino New Haven public schools. Gruen said parents explicitly opted to participate in the study after detailed explanations in English or Spanish. But one of the study’s most powerful benefits for parents or the district itself—the opportunity for free, intensive, one-on-one reading support for the district’s most struggling young students—also may make it difficult for potential participants to refuse.
Moreover, applying genetics to problems in health or education requires a very long view. Understanding the ways genes and environment enhance or inhibit each other remains incredibly tricky and complex, requiring massive (and often expensive) studies that may or may not improve on existing interventions. In a 2017 article in the Harvard Medical School’s Bioethics Journal, researchers in precision medicine cautioned that it’s easy to raise hopes prematurely, and argued that practitioners should not use “precision” in the form of genetic indicators to replace a more holistic view of people’s background, contexts, and behavior.
Fear of Stigmatization
Susanne Haga, a bioethicist and associate professor of medicine at Duke University’s Center for Applied Genomics and Precision Medicine said any potential genetic screening for dyslexia will have to balance the benefit of early identification with the potential harms; if early interventions don’t keep pace with identification, for example, toddlers could be labeled as at-risk readers without much earlier recourse for parents.
Earlier this summer, some of those concerns bubbled up in New Haven, as incoming school board members expressed concern over the project during an annual renewal of the district’s participation in the project. “If we have a strong research base [for existing reading interventions] ... are we setting up, with the DNA knowledge, a self-fulfilling prophesy that this person is going to have difficulty learning to read because he’s genetically limited?”, adding, “I’m at a loss to understand why our district isn’t utilizing those practices without the possibility of stigmatizing kids with DNA analysis.”
The discussion also led to a blizzard of questions from parents and the public: Does testing change students’ DNA? Can it be used to create independent stem-cell lines? Can the researchers sell the data after the study? Do researchers explain exactly what will happen to parents and get their explicit consent? (No, no, no, and yes.)
The board ultimately voted to continue its partnership with the project—which has provided about $1.5 million to the district in salary and benefits for six teachers trained in the interventions—but it added a new committee to brief all district parents, in addition to the researchers’ briefing, before they decide to participate. It also updated data-privacy protections and asked Gruen’s team to try to increase compensation for participating students, which is now $10 per reading assessment.
New Haven’s mayor, superintendent, and school board did not respond to requests for comment.
At the end of the current seven-year study, the researchers will destroy the key that links the genetic data to individual students, but the anonymous data will continue to be used in follow-up studies on dyslexia—and potentially other research, if parents opt to allow that.
Olena Lennon’s son Evan participated in both the intervention study and a smaller related brain-imaging study from 1st through 4th grades.
Researchers explained that her child’s DNA would be collected as part of the study, but Lennon said, “It’s never been an issue.”
“I had no problem with the DNA [collection],” Lennon said. “I know that’s sort of controversial, not just in this context but in general. But I still feel, you know, those risks must be taken, and we have to look at radical ways of doing research to come up with effective interventions.”
Her son enjoyed the interventions—and even the often awkward brain-scanning sessions—and she said the study team sent regular updates on test results. Ultimately, Evan was never diagnosed with a formal reading disability, but Lennon said she saw progress while her son was participating in the interventions. “There are certain areas where he’s had weaknesses ... but by the time he entered 3rd grade he was reading at grade level,” she said.
The debate is likely to grow. While school board members were debating whether to keep New Haven schools in the research study,on how genes affect how long people will stay in school. International research teams from the Social Science Genetic Association Consortium and the commercial genetics site 23andMe analyzed the full genomes of more than 1.1 million people, including about 300,000 from the commercial site. While all people share more than 99 percent of their DNA, researchers found what they described as a “treasure trove” of more than 1,200 differences in that last 1 percent that are associated with educational attainment.
The genetic differences affected an array of functions, many involving how the brain communicates, but the study could not predict schooling trajectories for any individual child. However, it opened hundreds of new avenues to explore not just genes, but how home and school environments might exaggerate or reduce genetic differences. All that suggests the New Haven study could be just the tip of the iceberg to come in research.
Other, similar studies are also exploring genetic links to autism, attention deficit disorders, and Down syndrome, among other learning-related conditions.
A version of this article appeared in the September 12, 2018 edition of Education Week as DNA Testing Seeks Answers for Dyslexia