In addition, dyslexics exhibited enhanced response entrainment in the right PT at 30 Hz, contralateral to the left location where there was an entrainment deficit (Figure 3F). Our next aim was to relate the ASSR asymmetry (left minus right) in the PT within the 25–35 Hz window to behavioral measures. We first checked whether reading fluency (as assessed by reading speed) correlated with ASSRs in the low-gamma band. We found a significant correlation in controls on both sides Cell Cycle inhibitor (Figure 4A, black frames) but no correlation in dyslexics on either side. To explore this global effect in greater depth, we conducted correlation analyses with scores from tests of phonological skills that are presumed to underlie the

reading deficit (Table 1; Table S1). A principal component analysis performed on behavioral data revealed two well-known factors, one loading on rapid naming tasks, and the other on phonological awareness (nonword repetition, spoonerisms, and digit span). By hypothesis, each task contributing to the PHONO factor relies on early auditory cortical sampling processes but investigating RAN was also of interest to us because it requires coordination of left temporal and prefrontal cortices (Holland et al., 2011). Subsequent analyses

were therefore conducted on the average Z-score of rapid naming tasks (RAN, Table 1), and the average Z-score of spoonerisms, nonword repetition PF-02341066 concentration and digit span tasks (PHONO, Table 1). We tested for correlations between the ASSR power in the 25–35 Hz window and each of these two composite phonological variables. In controls, we found no significant correlation with RAN on either side (a positive trend in Figure 4B), and a positive correlation with PHONO in the left PT only (Figure 4C). In dyslexics, there was no correlation with RAN and PHONO in the left PT (Figures 4B and 4C, upper panels). Conversely, in the right PT there was a negative correlation with RAN and a positive correlation with PHONO (note that there was also a positive correlation with nonword repetition when tested on its own). With respect Olopatadine to asymmetry (left-right, Figures 4B and 4C, lower panels), the correlation appeared positive

for RAN in dyslexics due to the strong negative correlation in the right PT. The correlation was positive for PHONO in controls and negative in dyslexics. To understand how individual subjects contributed to these effects, we first plotted the two behavioral variables against one another (Figure 5A). Usually, there is a positive correlation between the phonological scores, i.e., RAN and PHONO (Wolf et al., 2002). Our data overall confirmed this relationship in controls (C, r = 0.532, p = 0.013), but not in dyslexics (r = −0.413, p = 0.070). Instead, and consistent with Wolf et al. (2002), most dyslexic individuals show both deficits (Figure 5A) but frequently either a PHONO or a RAN deficit subtype (circles). We then computed the correlations between ASSR magnitude asymmetry in the PT at 30 Hz, i.e.