Transcranial random noise stimulation and cognitive training to improve learning and cognition of the atypically developing brain: A pilot study

Abstract

Learning disabilities that affect about 10% of human population are linked to atypical neurodevelopment, but predominantly treated by behavioural interventions. Behavioural interventions alone have shown little efficacy, indicating limited success in modulating neuroplasticity, especially in brains with neural atypicalities. Even in healthy adults, weeks of cognitive training alone led to inconsistent generalisable training gains, or "transfer effects" to non-trained materials. Meanwhile, transcranial random noise stimulation (tRNS), a painless and more direct neuromodulation method was shown to further promote cognitive training and transfer effects in healthy adults without harmful effects. It is unknown whether tRNS on the atypically developing brain might promote greater learning and transfer outcomes than training alone. Here, we show that tRNS over the bilateral dorsolateral prefrontal cortices (dlPFCs) improved learning and performance of children with mathematical learning disabilities (MLD) during arithmetic training compared to those who received sham (placebo) tRNS. Training gains correlated positively with improvement on a standardized mathematical diagnostic test, and this effect was strengthened by tRNS. These findings mirror those in healthy adults, and encourage replications using larger cohorts. Overall, this study offers insights into the concept of combining tRNS and cognitive training for improving learning and cognition of children with learning disabilities.

Conflict of interest statement

R.C.K. serves on the scientific advisory boards of Neuroelectrics Inc., InnoSphere Inc. and The Cognitive Enhancement Foundation (non-profit organisation). The other authors declare no conflicts of interest.

Figures

Figure 1

Transcranial random noise stimulation coupled with cognitive training to improve learning of children with mathematical learning disabilities at school (a) Illustration of a child moving from side-to-side to map a number on a number line while receiving transcranial random noise stimulation from a wireless brain stimulator. Response was registered for each trial when both hands were raised. Body movements were detected by a time-of-flight camera, KinectTM. (b) Examples of feedback on correct and incorrect responses. The game was adaptive to children’s performance; every 3 consecutive correct answers promoted the following trial to a more difficult level and vice versa.

Figure 2

Improved accuracy with transcranial random noise stimulation (tRNS). Only children in the real tRNS group showed improved accuracy with training. Those in the sham group showed little improvement, characteristic of those with learning disabilities. (a) Bar graph indicates the mean accuracy of each group on each training day. Error bars indicate one standard error of the mean (SEM). (b) Scatter plot displays the mean accuracy of each individual on each training day, labelled according to group.

Figure 3

Steeper learning with tRNS. Children in the real tRNS group showed a steeper learning slope compared to those in the sham group. Data points indicate the averaged group level at the end of each training day. Error bars indicate one standard error of the mean (SEM).

Figure 4

Sham group showed reduced response times with day compared with the real tRNS group, but was not associated with speed-accuracy trade-off (a) Bar graph displays the mean response time of each group on each training day. Error bars represent one standard error of the mean (SEM). (b) Scatter plot shows the mean response times of each individual on each training day, labelled by group.

Figure 5

Transfer of training gains to real-life achievement (a) Improvement in training transferred to performance on a standardised diagnostic mathematics test (MALT), a measure of achievement at school. Note that while we presented the actual values, the correlation was calculated using rank data (Spearman correlation) to account for the sample size and to avoid spurious results due to outliers. These points are not labelled by groups, as we found no effect of group. (b) Model 1 shows the non-significant effect of tRNS on training across Days 1 to 5 and gain in maths performance as indicated by the MALT (mathematical age-equivalence). Model 2 presents the significant relationship between gains in accuracy across Days 6 to 9 and gain in maths performance as indicated by the MALT (mathematical age-equivalence), which was moderated by group (tRNS vs. sham). Models 1 and 2 demonstrate the influence of tRNS on transfer, which depend on its effect on cognitive training. For the sake of transparency, we presented the unstandardised regression coefficients in the figure, and reported the corresponding beta weights in the text. ** = p < 0.02, *** = p < 0.001.