Fig 1.
Auditory-evoked response at the level of scalp electrodes, tACS-induced electric field reconstruction, and group difference at baseline at 30 Hz.
(A) The average response over time (dyslexia group, all AM frequencies considered) showed the typical auditory N100 and P200 components and related topography. The strongest peak-to-peak response amplitude was recorded at electrode FCz for both the N100 and P200 components. (B) Simulation of the electric field induced by tACS using a high-definition 4 × 1 electrode configuration, displayed on a head model from an individual subject. The model has been obtained with the freeware software SimNIBs [37] and Gmsh (www.gmsh.info). The electric field elicited by the tACS can be observed selectively in the left hemisphere, more prominently on auditory brain regions, whereas in the contralateral hemisphere, the resultant electric field can be estimated to zero (see S8 Fig). (C) Source reconstruction of the 30-Hz EEG response to 30-Hz AM before tACS. In the left auditory cortex, the dyslexia group showed a reduced response as compared to the no-dyslexia group (left); no difference between the groups was found in the right hemisphere (right). Numerical data used to generate this figure can be found at https://osf.io/6j49q/. Significance is denoted with * for p < 0.05. AM, amplitude-modulated; EEG, electroencephalography; tACS, transcranial alternating current stimulation.
Fig 2.
tACS-induced power modulation in ASSR at the scalp level.
ASSR in power (dB) to pure sounds modulated in amplitude (AM) with specific frequencies (from 28 Hz to 62 Hz, x-axis) recorded at electrode FCz. This electrode was chosen because it displayed the strongest evoked response in the time domain (Fig 1A). We considered the EEG power at the frequency corresponding to that of the AM pure tones (e.g., 30-Hz EEG power in response to 30-Hz AM tones). Both the dyslexia (A, left) and no-dyslexia groups (A, right) showed the strongest response for 40-Hz AM tones and the weakest for 30-Hz AM tones. In the dyslexia group, the 30-Hz tACS elicited a selective 30-Hz power increase for the 30-Hz AM sounds (A). This effect was absent in the sham and 60-Hz conditions in the dyslexia (B) as well as in the no-dyslexia (A, right) group. Numerical data used to generate this figure can be found at https://osf.io/6j49q/. Significance is denoted with ** for p < 0.01. AM, amplitude-modulated; ASSR, auditory steady-state response; EEG, electroencephalography; tACS, transcranial alternating current stimulation.
Fig 3.
tACS stimulation effectiveness: Behavioral results.
Changes in performance occurring after tACS on phonemic awareness (A) and reading accuracy (C) for each tACS condition (sham, 30 Hz, 60 Hz) in the no-dyslexia (shades of blue) and dyslexia (shades of red) groups. Each variable is obtained as the difference between the before versus after tACS measurements. For each of the 2 metrics, we performed a repeated-measures ANOVA with group (dyslexia, no dyslexia) as a between-subjects factor and stimulation condition (sham, 30 Hz, 60 Hz) as a within-subject factor. For both variables, the improvement after 30-Hz tACS was stronger in the dyslexia than in the control group, as well as after the sham in the same group and 60 Hz for the phonemic awareness index only. The improvement in phonemic awareness immediately after 30-Hz tACS in the dyslexic group was not accompanied by changes in syllable short-term memory (B). In the dyslexia group, performance increased after 30-Hz tACS for both phonemic awareness and reading accuracy and decreased 1 hour after for the phonemic awareness index only (D). Numerical data used to generate this figure can be found at https://osf.io/6j49q/. Significance is denoted with * for p < 0.05, ** for p < 0.01, *** for p < 0.001. tACS, transcranial alternating current stimulation.
Fig 4.
A 30-Hz EEG power increase after 30-Hz tACS in auditory cortex and STG.
Average power (dB) of ASSRs to 30-Hz AM pure tones over 2 ROIs in each hemisphere, before (white whisker plots) and after (colored whisker plots) 30-Hz tACS. Differences were tested by considering as fixed effects time (before/after tACS) and hemisphere (left/right) separately in the auditory cortex (A) and the STG (B). In auditory cortex, 30-Hz tACS increased responses bilaterally (A, right: average activity of both hemispheres for each time, main effect of time), an effect driven by a significant power increase in left auditory cortex (A, left). A significant interaction between time and hemisphere in the STG (B) revealed that EEG power after tACS was significantly higher than before on the left hemisphere. In addition, after the stimulation, power in the left hemisphere was stronger than in the right one. The relation between 30-Hz power gain in left auditory cortex and behavioral variables was considered across groups (solid gray lines). Changes in power correlated negatively with dyslexia severity (C, left: r = −0.46, p = 0.01) and positively (trend) with phonemic improvement (C, right: r = 0.36, p = 0.06). Within-group trend lines are displayed with dashed lines (red: dyslexia group, r = −0.15, p = 0.59 for ECLA, r = 0.05, p = 0.85 for phonemic awareness; blue: control group, r = 0.35, p = 0.22 for ECLA, r = −0.036, p = 0.9 for phonemic awareness). Numerical data used to generate this figure can be found at https://osf.io/6j49q/. Significance is denoted with * for p < 0.05, ** for p < 0.01. AM, amplitude-modulated; ASSR, auditory steady-state response; ECLA16+, Évaluation de Compétences de Lecture chez l'Adulte de plus de 16 ans; EEG, electroencephalography; ROI, region of interest; STG, superior temporal gyrus; tACS, transcranial alternating current stimulation.