Skip Navigation

Researchers Use Brain Scans to Predict Early Reading Difficulties

Skip sharing on social media links
Share this:
This news is archived NICHD Archive
Note: Information on this page was accurate at the time of publication.
This page is no longer being updated.

NIH-funded study could help reduce reading problems for some children

Wednesday, October 29, 2014

Listen to this podcast (MP3 - 5.5 MB).

Barrett Whitener: Researchers have used brain scans to track how young children learn to read, raising the possibility that the method could be used to diagnose young children with dyslexia and other reading disorders before they experience problems in school. Once identified, the children could be fast-tracked to interventions designed to help them overcome their reading difficulties.

From the National Institutes of Health, I’m Barrett Whitener. This is “Research Developments,” a podcast from the NIH’s Eunice Kennedy Shriver National Institute of Child Health and Human Development—the NICHD.

With me today is the study’s senior author, Fumiko Hoeft, an associate professor of child and adolescent psychiatry at UC San Francisco and a member of the UCSF Dyslexia Center.

Dr. Hoeft and her colleagues analyzed brain scans of 38 kindergarteners as they learned to read, and tracked the development of their brains’ so-called “white matter” until third grade.

White matter results from myelination—a process that reinforces brain-cell networks with an insulating material called myelin. Myelin covers neurons much like a covering insulates an electrical wire, and speeds the transmission of nerve impulses. Myelination coordinates communication among different regions of the brain and is essential for perceiving, thinking, and learning.

The researchers found that the developmental course of the children’s white matter predicted their ability to read. The research was funded, in part by the National Institutes of Health.

Thank you for joining us today, Dr. Hoeft.

Fumiko Hoeft: Thank you for having me.

Mr. Whitener: Dr. Hoeft, what methods do doctors typically use to measure reading ability? And why would analyzing brain scans make it easier to identify poor readers?

Dr. Hoeft: Typically, psychologists used what’s known as paper-and-pencil measures, or neuropsychological tests, or psycho-educational battery, as it’s known sometimes. They ask whether children can read often real words that we use in everyday life or read sentences or fake words accurately and fluently. They might also ask if children can understand the written text and spell words—a process that we call “recoding,” instead of “decoding,” for reading.

Additionally, practitioners often ask if a children’s home, school or peer environment—to see if their literacy environment is adequate. With an impoverished environment, we know that children have difficulty thriving. This is not just for reading, but for many other abilities. Practitioners often ask about family history because we know that if you had a parent who struggled to read, you will have a 50% or so chance to have similar problems.

Also, in terms of your second question about brain scans and how it could make it easier to identify poor readers—we previously showed two things in 2007 and 2011. Number 1, that brain measures can have an additive explanatory power to these behavior measures in predicting how well children will be reading in the future—even though, until then, people believed that behavior and brain may just be two sides of the same coin. Number 2, we showed that brain measures can be predictive, even when we don’t have good behavioral markers that predict outcomes. So in this study, we looked at children already diagnosed with dyslexia and tried to predict how well some kids will learn to compensate. We were able to use brain to relatively accurately predict outcome 3 years later, or 2½ years later, even though behavioral measures—the typical standard measures—were not able to predict outcomes.

Mr. Whitener: So you mentioned measuring the brain scans over time and that’s something I believe is also different about your study, compared to previous ones. Is that right?

Dr. Hoeft: Yes, actually that’s a very good point. In previous studies, we measured it at baseline at the very beginning of our study and then followed them up for a year or 2½ years in these different studies—and asked if we could predict reading outcomes from these baseline brain measures and behavioral measures.

In the current study, we looked at development over time from kindergarten to the beginning of third grade. There’s a difference there, also.

Now in this new study, we can make a third point, which is that brain measures can explain and predict outcome above and beyond not just behavioral measures that we showed previously, but also above and beyond other important clues that practitioners used to predict outcome, which we talked about already, such as environment or literacy environment, socioeconomic status, family history, and other paper-and-pencil tests.

So if all things being equal and possible, there is room for brain measures, I think, to make important contributions to clinical practice.

Mr. Whitener: I gave a brief description of your study results a minute ago. Could you tell us more about what you found?

Dr. Hoeft: Yes. So I guess the results—the bottom line of the results—are pretty simple: that we found this region called the left temporal parietal white matter, which is just above and behind your left ear, and the white matter clusters or white matter region—brain region within this region—was most likely associated with a fiber known as arcuate fasciculus. This arcuate fasciculus is known to be important for speech and language processing. We know that from many previous studies.

This white matter fiber or white matter region was predictive of reading acquisition from kindergarten to third grade, above and beyond the typical measures that we use in practice to predict outcome—which was the pre-literacy measures, the paper-and-pencil tests, general cognitive abilities, such as IQ—expressive and receptive language measures—and environment, as well as family history.

Mr. Whitener: Based on what you found in the study, do you have any particular advice or suggestions for parents and caregivers of young children?

Dr. Hoeft: Yes. Sometimes when we publish work like this, we get calls—this is more of a word of caution, I suppose—but sometimes when we publish work like this, we get calls all the way from India, wanting their children to be tested in our laboratory.

While we believe that the current findings help move the field forward, it is currently far from being reliable and useful on an each-child basis, and hence, to be useful in clinical practice. In fact, some people think that there are concerns that this may never be useful in clinical practice.

For example, even if this were a really accurate tool, currently the cost of a scan can be hundreds of dollars to thousands of dollars. Many towns don’t have access to MRI scanners like the ones we do, or some research institutions do. Also, most MRI research reports on a very small number of participants. As you just said, we looked at this in 38 young children from our Bay Area region. We reported on group findings—not findings on each child.

Also, white matter development from kindergarten to grade 3 was used to predict grade 3 reading outcomes in the study, as you said. So some might argue that we can hardly call this prediction, because we’re using information from kindergarten all the way to grade 3 to predict grade 3 outcomes. In practice, we want to be able to predict much earlier than this: for example, take preschool-children’s brain scans and predict reading outcome, or even in infancy or even prenatally, before they’re born.

So that will be the ideal scenario. We need to use approaches that can create models a priori, so, beforehand, and predict the outcome of each child individually. Currently the technology doesn’t allow us to predict each child one at a time, accurately as we want to. We don’t have enough data to be able to say that it’s going to work for children on the east coast, children who speak a different language. So there’s a lot of variability that we have to kind of overcome before we can use this for practice.

Mr. Whitener: What is the next step for your research in this area? What are you planning next?

Dr. Hoeft: We’re interested in two main topics. The first one is continuing to work on early identification and early intervention. The second piece is on social and emotional health.

For the first part, we’re interested in continuing our work on early identification. Now we’re starting to look at parents, and parents’ brain and behavior, not just children’s brain and behavior. We want to see if parents’ brain scans can predict the child’s reading outcome; we’re having promising results.

The second part is on social and emotional health. We’re very invested and we think that it’s very important to focus on social and emotional health to enhance motivation and self-esteem in children with learning disabilities.

So we’re developing assessments, we’re developing apps to train social and emotional health, or enhance social and emotional health in children with learning disabilities. We’re also creating teacher training material. Those are our next steps.

Mr. Whitener: I’ve been speaking with Dr. Fumiko Hoeft, senior author of the study, “White Matter Morphometric Changes Uniquely Predict Children’s Reading Acquisition,” which was published online in the journal, Psychological Sciences. Thank you for joining me today, Dr. Hoeft.

Dr. Hoeft: Thank you very much, Barrett.

###

About the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD): The NICHD sponsors research on development, before and after birth; maternal, child, and family health; reproductive biology and population issues; and medical rehabilitation. For more information, visit the Institute's website at http://www.nichd.nih.gov/.​​​​​

Contact Information

NICHD Research Developments
NICHD Press Office
nichdpress@mail.nih.gov
301-496-5133
BOND National Institues of Health Home Home Division of Intramural Population Health Research