Edited by: Manuel Fernando Casanova, University of Louisville, USA
Reviewed by: Gabriele Miceli, University of Trento, Italy; Deniz Doruk, Harvard Medical School, USA
*Correspondence: Jessica W. Younger
This article was submitted to Neural Technology, a section of the journal Frontiers in Neuroscience
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Adults struggling with low reading skills are underserved by limited available treatments. While brain stimulation techniques such as transcranial direct current stimulation (tDCS) has the potential to improve a variety of cognitive functions, little work has been done examining its potential to treat reading disabilities. Research on the effects of tDCS on reading abilities has been somewhat inconsistent perhaps in part due to discrepancies between studies in the nature of the tasks. In the current study, we examined the effect of tDCS to the left inferior parietal lobe (L IPL) on two reading tasks in low-to-average readers. We compared performance on a sight word efficiency (SWE) task and a rhyme judgment task before and after either stimulation to the L IPL, right superior parietal lobe (R SPL), or sham stimulation. Readers who received stimulation to the L IPL showed greater improvements on the SWE task, but less improvement on the rhyme judgment task compared to the R SPL and sham groups. This study demonstrates for the first time both a positive and negative effect of stimulation under the same stimulation parameters within the same participants. The results highlight the need to consider multiple tasks when assessing the potential of using tDCS as a treatment.
Over the last decade, interest in using brain stimulation techniques as a therapeutic tool to treat cognitive impairment in adults has received increasing attention (Dubljević et al.,
More recently, research using brain stimulation to treat learning disorders, such as dyslexia and dyscalculia, has been called for (Cohen Kadosh et al.,
Neuroimaging research has identified several brain regions that show altered function in individuals with dyslexia that may serve as potential targets of brain stimulation. Three brain areas in particular have consistently shown altered functionality compared to typical readers—the inferior frontal gyrus (IFG), temporo-parietal areas, and occipito-temporal areas (Richlan et al.,
So far, neuromodulation studies examining the tool's potential to improve reading ability have stimulated regions shown to be underactivated in poor readers, and all have had some success (Costanzo et al.,
The Costanzo et al.'s TMS studies (Costanzo et al.,
This hypothesis is supported by second language learning studies with adults (Hashimoto and Sakai,
In order to test the hypothesis that tDCS to superior portions of the temporo-parietal cortex will lead to reading improvement for low ability readers, we stimulated the left IPL in low-to-average readers and measured their improvement on two reading tasks; single word reading efficiency and a rhyme judgment task. Both require the use of phonological and orthographic information, but in different ways. Single word reading efficiency requires articulating the phonological output from orthographic input. This skill has been shown to be related to both overall word reading ability and the ability to decode words based on grapheme-to-phoneme mappings (Adlof et al.,
This study was carried out in accordance with the recommendations of the University of Texas at Austin Institutional Review Board with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki.
In total, 100 right-handed 18–35 year-old native English speakers with normal or corrected-to-normal vision were screened for below average reading ability (< 100 standard score) as determined by the Sight Word Efficiency (SWE) subtest of the Test of Word Reading Efficiency (TOWRE; Torgesen et al.,
Of those who met all performance criteria, 11 (7 female) received real stimulation to the L IPL, 14 (6 female) received sham stimulation, and 11 (9 female) received real stimulation to the R SPL. Due to an imbalance in the run orders in the sham group, four participants were randomly eliminated for a final sample of 10 (4 female). One-way ANOVAs revealed no significant effects of group on all group characteristics and baseline measures as reported in Table
Age (years) | 26.8 (5.5) | 25.2 (3.5) | 26.2 (4.9) |
Gender (f) | 7 | 8 | 4 |
IQ | 112.1 (10.9) | 111.0 (11.70) | 109.3 (9.9) |
Pre-single word reading | 88.5 (8.4) | 88.1 (7.6) | 89.2 (7.9) |
Pre-rhyme judgment RT (ms) | 854 (124) | 972 (147) | 998 (193) |
Pre-rhyme judgment accuracy (%) | 91.6 (7.0) | 88.4 (9.5) | 87.9 (12.0) |
Participants took part in a single-blind, sham and stimulation controlled study comparing pre- and post-stimulation performance on two measures of reading ability: single word reading efficiency and rhyme judgment. Participants completed two sessions that took place 3–5 days apart. During the first session, participants completed standardized tests and baseline assessments of reading ability. During the second session, participants received either sham or real stimulation for 20 min, after which they completed an alternate form of the reading ability measures using different sets of stimuli. Alternate forms of the tasks were counterbalanced across participants.
Direct current was administered using a battery-driven DC stimulator device (NeuroConn) via two saline-soaked electrodes (5 × 5 cm; 25 cm2). The anode electrode was placed over either the L IPL (P3) or R SPL (CP4) according to the international 10–20 system for electroencephalography (EEG) electrode placement (Herwig et al.,
Word reading efficiency was measured via the SWE subtest of the TOWRE. This test is a measure of ability to read real words accurately and quickly. Participants were given 45 s to read aloud as many of 104 words as possible. A standard score is determined by the number of words read correctly within 45 s, and this score was used as the metric of single word reading efficiency.
The ability to map orthography to phonology and phonological working memory were assessed with a rhyme judgment task in which participants were presented with a series of visual word pairs and asked to indicate whether the words rhymed or not. Word pairs were designed to manipulate orthographic and phonological similarity to ensure participants could not rely on orthography alone and phonological representations had to be used to accurately complete the task. There were two congruent conditions in which word pairs had either similar orthography and phonology (e.g., CAGE-RAGE) or not (e.g., TRIAL-FALL), and two incongruent conditions in which word pairs had either similar orthography but dissimilar phonology (e.g., PINT-MINT) or dissimilar orthography but similar phonology (e.g., GRADE-PAID) pairs. Each condition had 12 trials for a total of 48 trials in each session.
All words were monosyllabic, having neither homophones nor homographs and were matched across condition for written word frequency in children (Zeno,
Participants were asked to respond as quickly and as accurately as possible. The first word was presented for 800 ms followed by a 200 ms inter-stimulus interval and the presentation of the second word. Participants could respond as soon as the second word was presented up to 2500 ms after the onset of the word. After the participant responded, a red fixation cross appeared signaling the inter-trial interval. The task was self-paced and participants were able to control when the next trial began. Average reaction times (RT) to correct trials trimmed to include only responses within 2.5 standard deviations from an individual's average reaction time were used as the metric of rhyme judgment ability due to ceiling effects on accuracy.
Performance on each experimental task was submitted to a 3 (Stimulation group; L IPL, Sham, R SPL) × 2 (Time; Session 1, Session 2) mixed-model ANOVA in order to determine whether a measure showed a Group × Time interaction. Planned follow-up tests were conducted using separate 2 (Stimulation group; L IPL, Sham or R SPL) × 2 (Time; Session 1, Session 2) mixed-model ANOVAs to examine potential Group × Time interactions for the L IPL group compared to the two control groups (Sham, R SPL) separately.
The 3 × 2 ANOVA revealed a significant main effect of Time [
Single word reading | 88.5 (8.4) | 98.8 (14.3) | 88.1 (7.6) | 92.2 (8.3) | 89.2 (7.9) | 91.7 (8.2) |
Rhyme judgment RT (ms) | 854 (124) | 811 (142) | 972 (147) | 885 (141) | 998 (193) | 820 (99) |
Rhyme judgment accuracy (%) | 91.7 (7.0) | 94.7 (7.5) | 88.4 (9.5) | 89.6 (10.6) | 87.9 (12.1) | 90.0 (9.0) |
All participants performed well on the rhyme judgment task as indicated by high accuracy at Time 1 and Time 2. A 3 × 3 ANOVA did not reveal a main effect of Time or any Group × Time interactions (
The 3 × 2 ANOVA again revealed a significant main effect of Time [
The goal of the current study was to assess whether stimulation of the L IPL improves multiple aspect of reading for low-to-average readers by measuring its impact on two tasks that tap into different subskills of reading. While L IPL stimulation did result in gains in single word reading efficiency, it resulted in relative impairment on the rhyme judgment task, demonstrating for the first time a significant positive and negative effect on two different tasks with the same stimulation parameters within the same group of participants. Although our results indicate stimulation to the L IPL may be a good site for improving reading fluency for low-to-average readers, the lack of improvement on the rhyme judgment task warrants caution in advocating the left IPL as a site to improve several aspects of reading.
The positive influence of L IPL stimulation on reading fluency measures for low-to-average readers was consistent with our hypothesis. Although there was a main effect of time, indicating there was a general practice effect for all groups, the L IPL stimulation resulted in greater improvement than the Sham or R SPL stimulation. Our finding that stimulation to the IPL led to greater improvements for low-to-average readers than previous reports of stimulation to the STG (Turkeltaub et al.,
Our results that left hemisphere stimulation leads to improvement in SWE is in contrast to Thomson et al. (
In contrast to the expected results of tDCS on the SWE, the negative effect of L IPL stimulation on improvement on the rhyming judgment task in low-to-average readers was unexpected. Previous neuroimaging work with both children and adults has shown that increases in activation in the left IPL are associated with better performance on the rhyming task (Hoeft et al.,
From a clinical perspective, perhaps the most important finding from the current study is that tDCS can positively impact one skill while negatively impacting another. Our results underscore the importance of including multiple tasks that potentially tap into different underlying cognitive processes in order to assess whether the potential costs of stimulation outweigh the potential gains. Assessing multiple tasks becomes especially important when considering whether tDCS should be recommended as a treatment. In the current study, the cost in practice effect on speed during a rhyme judgment task is probably worth the gains seen in single word reading for low-to-average readers; accuracy was not affected and speed did not decrease after stimulation. However, this population was low-to-average in skill, not impaired. The cost to benefit ratio may increase as reading skill decreases. Further studies examining how individual differences impact the effects of tDCS on multiple reading tasks are needed before being able to advocate for tDCS as a treatment for reading disabilities.
Further, our findings that tDCS had a differential effect on two aspects of reading in low-to-average readers has implications for the design of future tDCS studies. In the current study, both tasks were reading-related, but the differences between tasks in the working memory component are in line with the literature that anodal tDCS does not have the expected positive effect on working memory abilities. These results support the idea that tDCS can affect cognitive processes differently, depending on how the target or surrounding brain area is involved in a given cognitive process. Future research is still needed to determine the circumstances in which the conventional idea that anodal stimulation leads to enhancement of activation and behavior while cathodal stimulation leads to inhibition of activation and behavior holds true (De Berker et al.,
The current study used a between subjects design, meaning that the different stimulation groups were composed of different individuals. While the groups were equated on task behavioral abilities, it is virtually impossible to equate them on all factors that could potentially impact the effects of stimulation. For example, other research groups have shown that individual differences in physiological measures such as skull thickness, and levels of certain hormones and neurotransmitters such as GABA, can affect the way stimulation affects an individual (Krause and Cohen Kadosh,
Finally, as with all tDCS studies, without the use of neuroimaging techniques such as structural or functional magnetic resonance imaging (fMRI), we are unable to confirm that the stimulated area was in fact the targeted area. Individual differences in anatomy may have led to differences in how well the stimulation site aligned with the brain regions that are actually used to perform the tasks. Similarly, due to the distributed effects of tDCS we cannot make strong conclusions about whether the results are due to stimulation to the targeted region or surrounding and connected regions without neuroimaging measures. Future research with tDCS would benefit from using neuroimaging methods to have more precisely located targets and a better understanding of the locations that were actually affected in order to develop the most effective treatment methods.
Our study provides important cautionary evidence for the use of tDCS as a treatment for low reading ability. Although stimulation to the left IPL led to greater improvements in reading fluency than those previously demonstrated with a different stimulation site (2012), we also found a negative effect on another subcomponent of reading in low-to-average readers, i.e., rhyming two visually presented words. These positive and negative effects on two different subcomponents of reading were demonstrated using the same stimulation parameters within the same participants. These results stress the need for further research examining the effect of a set of stimulation parameters on complementary skills so that potential users of tDCS as a therapy can accurately weigh the costs and benefits of the treatment.
MR and JB conceived and designed the experiments. MR and JY performed the experiments. JY and JB analyzed and interpreted the data. JY drafted the manuscript. JY, MR, and JB performed a critical review of the manuscript. All the authors read and approved the final version of the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This research was supported by the University of Texas at Austin. The authors would also like to thank the numerous research assistants who assisted in data collection throughout the study.