Current Status of Treatments for Dyslexia: Critical Review

Ann W. Alexander, MD; Anne-Marie Slinger-Constant, MD


J Child Neurol. 2004;19(10):744-758. 

In This Article

Neuroimaging Studies and Response to Treatment

Neuroimaging studies have demonstrated the impact of intensive and explicit phonologically based remedial interventions on cortical activation. Simos and colleagues used magnetic source imaging to evaluate changes in spatiotemporal brain activation profiles while children performed a pseudoword reading task before and after treatment.[25] Treatment consisted of 8 weeks of intensive (one on one, 1 to 2 hours/day), phonologically based interventions in eight children aged 7 to 17 years (mean age 11.4 years) with dyslexia. Six of the eight children scored below the 4th percentile, and two scored at or below the 18th percentile on reading measures (the Basic Reading Skills cluster of the Woodcock-Johnson Test of Achievement, Third Edition.)[13] Intelligence was in the average range (mean IQ 102 ± 4.5 on the Wechsler Intelligence Scale for Children, Third Edition [WISC-III]).[14] Six of the eight were diagnosed with attention-deficit disorder and were treated with psychostimulant medication throughout the study period. The comparison group consisted of eight children aged 8 to 14.2 years old (mean age 10.3 years) with reading scores above the 50th percentile and a mean IQ score of 107 (+ 10.5). One of the eight was diagnosed with attention-deficit disorder and was treated with medication. All 16 subjects were right-handed, native English speakers. Following intervention, reading accuracy in all subjects was above the 37th percentile. On the imaging studies prior to intervention, the subjects with dyslexia displayed the characteristic cortical activation profile of the individual with dyslexia.little to no activation in the left temporoparietal areas and strong activation in the homotopic right hemisphere. In contrast, the controls had little activation on the right and strongly activated the temporoparietal areas on the left, believed to be involved in phonologic processing. After treatment, the cortical activation patterns of the subjects with dyslexia resembled much more closely those of the normal controls. A dramatic increase was noted on the left, most pronounced in the left superior temporal gyrus, and increased activation in the inferior parietal areas approached significance. There was a moderate decrease in homotopic right hemisphere activation. The cortical activation patterns of the normal controls did not change. Of note, the imaging study also revealed that the subjects with dyslexia engaged the left superior temporal gyrus more slowly than did the controls, despite remediation, suggesting that the new circuitry might not be as efficient, and fluency would remain a problem. The authors stated that the small sample size calls for caution in interpreting the results. However, with the significant improvement noted both behaviorally and physiologically, they suggested that these findings indicate a "normalization" of functional brain organization following intensive intervention.

These findings have been replicated in a methodologically rigorous longitudinal study by Shaywitz and colleagues.[26] A large cohort (77 6- to 9-year-old children) with reading disabilities received 86 to 115 hours of daily, individual, and evidence-based phonologically mediated reading intervention at school. There were two control groups.a group of reading-disabled children who received the typical interventions of the school and private tutoring and a group of normal readers. The children were assessed before, following, and 1 year after treatment ended. Behavioral gains on the Gray Oral Reading Test, Third Edition, were found to be significant in comparison with the control reading-disabled group and the normal readers immediately post-treatment. The scale score improved from 5.4 to 7.0 for the treatment group; the scale score decreased from 5.4 to 4.9 for the reading-disabled control group on the passage score, a combination of accuracy and rate (mean 7, SD ±3). There could be no comparison at 1 year because only two of the reading-disabled control group and two of the normal reading group returned. The treatment group evidenced a decline in performance from a standard score of 7.0 to 6.4. However, an effect size calculation of the group's gains during treatment and of gains maintained from pretreatment to 1-year follow-up revealed an effect for both periods (effect size 0.52 and 0.43, respectively). Effect size in the 0.5 to 0.79 range is considered moderate.[27] Of note, the mean reading comprehension as measured by the Gray Oral Reading Test, Third Edition, improved during the treatment and at the 1-year follow-up, moving into and staying in the average range (pretest mean standard score 5.7; post-test mean standard score 8.0; standard score at 1-year follow-up 8.5). Other studies have demonstrated a greater ability with text comprehension than with word-level reading before treatment[11] and that, following treatment, comprehension exceeds the 30th percentile, although word reading ability still lags behind in the 13th to 22nd percentile. This suggests that enhanced phonologic decoding allows the person with dyslexia to better use other top-down processing strategies, which can be strengths.

Physiologic measures also revealed significant changes. Using functional magnetic resonance imaging (MRI) during a letter identification task, Shaywitz et al. also demonstrated the shift from right hemisphere to increased left hemisphere activation reported by Simos et al.[25,26] Interestingly, although the treatment subjects remained in the impaired range on reading measures (Gray Oral Reading Test, Third Edition, passage score.5.4 preintervention, 7.0 immediately postintervention, and 6.4 1 year later) in contrast to the normal readers, both groups demonstrated similar brain activation patterns, with increased activation in the left inferior gyrus and posterior middle temporal gyrus immediately after treatment. At the 1-year follow-up, only the treatment group underwent a repeated imaging study, which revealed that these children activated the fast-paced occipitotemporal word form area serving skilled reading, the bilateral inferior frontal gyri, and left superior temporal regions.

Richards et al. used magnetic resonance spectroscopy (specifically, proton echo-planar spectroscopy) to evaluate the response to a less intensive but phonologically based intervention on brain lactate metabolism during "reading-related tasks" in eight boys with dyslexia and seven control boys between the ages of 10 and 13 years.[28] After the 3-week intervention period, which consisted of 15 2-hour group sessions, performance on behavioral measures of phonologic processing improved. Prior to intervention, the dyslexic participants manifested significantly greater lactate metabolism in the anterior quadrant of the left hemisphere during a reading task compared with controls, suggesting that they had greater difficulty with the task that necessitated increased use of the frontal cortex. One year after intervention, repeat magnetic resonance spectroscopy revealed a metabolic pattern similar to that of controls during the phonologic task.

To evaluate the effects of treatment using two different linguistic interventions, phonologic and morphologic, Berninger et al. conducted a study with children in fourth to sixth grade who were part of a family genetics study and were found to be dyslexic.[29] The children were of normal intelligence (mean Wechsler Verbal IQ was 110.6 ± 11), and although they did not have oral language problems or attention deficit, rapid naming deficits were present. The subjects were randomly assigned to a phonologic awareness or a morphologic awareness treatment group. They received daily treatment amounting to 28 hours over a period of 3 weeks. The behavioral results revealed significant gains in phonemic decoding skill and the rate from pretest to post-test for both groups. Phonologic decoding (the Woodcock-Johnson Reading Mastery Test Word Attack subtest) improved significantly for both groups. Interestingly, the morphology treatment group showed a significantly greater level of improvement than the phonology treatment group in phonologic decoding efficiency, as measured by the Pseudoword Reading Efficiency subtest of the Test of Word Reading Efficiency.[30] Berninger et al. suggested that the efficiency of phonologic decoding depends on an interaction between phonologic, morphologic, and orthographic representations, as described by Harm and Seidenberg's computer simulation studies below.[31] Moreover, following morphology treatment, magnetic resonance spectroscopy revealed decreased lactate activation in the left frontal region during a rhyming task in all but one child with dyslexia, and the activation pattern resembled that of normal readers. However, in the phonologic treatment group, the pattern of activation remained the same or increased in all of the subjects. A reduction in lactate activation might reflect increased efficiency of mental processing. Good readers have less activation in the left frontal region during this phonologic judgment task, indicating less need to enlist prefrontal executive resources. A second imaging study using functional MRI revealed that normally reading controls showed different patterns of activation when performing phonologic and morphologic judgment tasks that were stable over time. In contrast, the pretreatment activation patterns of the subjects with dyslexia differed from those of the controls but resembled them after treatment.

The study groups were small, and the intervention was brief; therefore, the interpretation of these results merits caution. Berninger et al. suggested that morphologic awareness training can help children develop the coordination for all of the language codes needed for word learning.[29] If this is so, they suggest that morphology provides a built-in executive function to do so. They recommend that future treatment design not only be aimed at orthographic and phonologic awareness but should also include training in morphologic awareness and the interrelationship of the three word forms.phonologic, morphologic, and orthographic.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: