ARTICLE IN BRIEF
A new study used fMRI images to compare brain activity in adults with dyslexia versus normal readers, finding that dyslexics had a faulty connection between the areas of the brain that process acoustic/phonetic information and the area that uses that information for language.
Dyslexia can cause lifelong difficulties in a world that is built on the written word. Now, a team of scientists from the University of Leuven in Belgium, Oxford University, and University College London present data that challenges the most prominent theory of why so many people have problems reading and spelling.
Bart Boets, PhD, and his colleagues combined functional MRI with multi-voxel pattern analysis in adults with dyslexia. Their experiments were designed to tease out whether dyslexics have acoustic deficits that lead to fuzzy representations or whether they have normal phonetic representations but difficulty accessing these representations further upstream for language processing.
Many studies done with dyslexic children over the past several decades have supported the first idea — that dyslexics have trouble with acoustic and phonetic representation. However, this study uncovered evidence through neural imaging that the dyslexics had a faulty connection between the areas of the brain that process acoustic/phonetic information and the area that uses that information for language.
Speech sound units (phonemes) are represented by visual letters (graphemes) when reading. Learning the relationship between phonemes and graphemes requires an understanding that words can be broken down into smaller units of sound (phonemes) and it is these sounds that the letters represent. It has long been thought that people with dyslexia have problems with analyzing the acoustics needed to learn the sounds of language.
“The leading view on dyslexia hypothesizes that these phonetic representations are somehow degraded in individuals with dyslexia; they might be less robust and less distinct. Yet, this has mainly been inferred indirectly on the basis of performance on behavioral phonological tasks,” said Dr. Boets. “We aimed to objectively assess the quality of these representations by measuring their robustness and distinctness directly at the cortical brain level.”
In the study, published in the Dec. 6 issue of Science, the researchers collected whole-brain functional MRI scans from 23 adults with dyslexia and 22 age-matched normal readers. They were asked to listen to four speech sounds and carry out a phoneme discrimination task. Once they listened to the sounds of the snippets of words, the researchers gave them another sequence to listen to.
They wanted to determine whether dyslexics' phonological deficits are caused by poor quality of the phonetic representations or by difficulties in accessing intact phonetic representations.
They looked for differences in the ability of dyslexics and normal readers to discriminate vowels and stop-consonants. [The letters b, d, g, p, t, and k are stop consonants, otherwise referred to as plosives. When producing stop consonants, the air is stopped briefly by closure somewhere in the mouth. These are also the speech sounds that require the most rapid acoustic processing.]
Other studies have suggested that people with dyslexia have problems discriminating between stop-consonants, and therefore they expected to see the biggest difference in brain activation to consonant syllables between normal readers and those adults diagnosed with dyslexia.
But they did not. They analyzed information from a dozen regions in the left and right hemisphere that are involved in speech processing in both dyslexics and normal readers and found no difference in the quality of the phonetic representation. The responses were as robust and distinct in adult dyslexics as they were in adults with no history of reading or language problems. However, the study subjects with dyslexia took 50 percent longer to answer the questions, suggesting that they were having difficulty with processing the information as efficiently as the normal readers did.
NEURAL IMAGING RESULTS
So what is going wrong? Were other brain regions having problems accessing the phonetic representation? The investigators focused on the front of the brain to Broca's area, and in particular the left inferior frontal gyrus (IFG) pars opercularis, because it is known to be involved with sensory-motor integration and phonological processing. According to Dr. Boets and his colleagues, this region has to “access the representations in primary and secondary auditory cortices to compute the required phonological manipulations.”
The investigators reported that both groups showed strong connectivity among bilateral temporal regions — primary and secondary auditory cortices — and these areas were functionally connected to the left IFG but the connection was smaller in the dyslexic group. The difference was most prominent in the left superior temporal gyrus (STG) and the right posterior anterior cingulate.
Through diffusion tensor imaging studies, they identified significant reductions in the white matter integrity in the left arcuate fasciculus in the dyslexics (p=0.019), which scientists say is “neuroanatomical evidence that corroborates the deficiency in functional connectivity between the left inferior frontal gyrus and left STG.”
These functional and structural measures accounted for 35 percent of the variance in reading and writing abilities among dyslexics. “Individual differences in the quality of this functional connection were highly significantly correlated with individual differences in reading, spelling, and phonological skills,” said Dr. Boets.
Still, they said in the paper's conclusion that it is not “time to abandon the influential phonological deficit hypothesis.”
The researchers acknowledge that the brain differences captured in this adult sample may have little to do with what is going on during early childhood when dyslexics are learning to read. “It may be that dyslexic readers achieve normal neural representations through greater than normal effort,” they wrote.
During the task, their speed of response was slower but accuracy was intact. “Although we cannot rule out that dyslexics' neural representations may have been less specified at a younger age or would follow a different temporal trajectory detectable through techniques such as electroencephalography, our results indicate that the phonetic representations can be intact in adult dyslexics despite persisting reading difficulties,” they noted.
STILL MORE QUESTIONS
The study left the investigators with a lot more questions. Are the intact phoneme representations present during the entire development of reading? When does this kind of ‘connectivity’ problem develop during the process of learning to read? How and when do the connections involved in the reading network develop during the reading instruction phase?
To begin to answer some of these questions they have identified a group of five-year old children who are at genetic risk for dyslexia. They scanned them before they learned to read, and will do so after they know how to read.
If similar structural and functional changes are seen in dyslexic children, it might be helpful to develop interventions that strengthen the communication pathway from the auditory to higher-level phonological regions, said Dr. Boets. The findings “inform us about a very specific dysfunctional connection, and this knowledge should be taken into account while designing the most appropriate intervention techniques.” He added that traditional interventions to strength phonological skills and phoneme-grapheme coupling would also help improve access to this representation.
EXPERTS WEIGH IN
“It is never one simple story,” said John Gabrieli, PhD, director of the Athinoula A. Martinos Imaging Center at the McGovern Institute at Massachusetts Institute of Technology (MIT) and co-director of the MIT Clinical Research Center. “The scientists have used a clever and thoughtful design to show that adults with dyslexia are processing language sounds correctly but that there is a communication problem between posterior and frontal areas that manipulate and use this information to read.”
“A small weakness in some areas becomes important when you are pressured to learn to read,” he said.
There may be ways to strengthen the communication between these brain regions, he added. It may be possible to use devices such as transcranial direct current stimulation or transcranial magnetic stimulation that uses stimulating electrodes to enhance communication between brain regions. “There is reason to start thinking about these types of interventions,” said Dr. Gabrieli. “Electrical stimulation would make people cringe for kindergarten children but the evidence is overwhelming that interventions are much more effective from kindergarten through second grade than after that period.”
Dr. Gabrieli said he and colleagues have recently completed a study in kindergarten children showing that the brain pathway is different in those with language problems. “Our results align with the current finding, even before kids start learning how to read.” Dr. Gabrieli and his Boston colleagues are tracking behavior with brain measurements over time to see if they can identify factors that are present in those who will go on to become poor readers.
Paula Tallal, PhD, a professor of neuroscience and co-director of the Center for Molecular and Behavioral Neuroscience at Rutgers University, has spent her career studying individual differences in language and reading development and how to use the brain's own neuroplasticity to help remediate struggling learners.
Dr. Tallal said that she agrees with both Dr. Boets and Dr. Gabrieli that brain differences assessed in adult samples likely reflect compensatory outcomes rather than the cause of dyslexia. “It is important to understand that the neural mechanisms that are involved in developing a complex function such as reading are not the same as those needed to maintain that same function once developed. If we want to understand the causes of developmental disabilities we need to study development as it unfolds,” she said.
Dr. Tallal said that she found the analysis complicated. Looking at the composite brain scans, she said that it seems that people with dyslexia performed equally well on acoustic processing of speech compared with normal readers whose left hemisphere is primed to process this information more quickly and efficiently. “We need specialization for rapid processing,” said Dr. Tallal. “I am questioning whether the investigators looked at unilateral versus bilateral processing. Even though the dyslexics show more right hemisphere activity than normal readers they conclude that there is no difference in the interpretation of sound in the two groups — dyslexics and normal readers.
“Dyslexics may be getting there, but not the same way as normal readers,” she added. “The brain is processing the acoustic information but through an alternate pathway that is not normal.”
She said that she agrees “that there is a disconnection in the pathway between the temporal parietal region and the inferior frontal gyrus (Broca's area) in the left hemisphere, but the disruption may occur because the acoustic information itself is represented more bilaterally by dyslexics and in different regions of the brain.”
Dr. Tallal added that studying adult dyslexics doesn't offer much insight to the broader population of children with dyslexia. Their brains have compensated for their reading difficulty, she said.
She worries about the implications of the current finding in Science. “If this study says that there is no difference in the processing of acoustic information, then some people will believe that it is not important to intervene. There is an abundance of data showing it is important to improve the processing of acoustic speech if you are going to help children.”
“Their study needs to be carried out in children,” she concluded.