Transcranial Direct Current Stimulation (TDCS) Investigation of the Role of the Cerebellum in Motor, Language, and Social Prediction

Neuroimaging, cerebellar lesion studies, and animal studies have all provided evidence that the cerebellum is involved in cognitive processing. In particular, it has been hypothesized that the cerebellum uses an internal model to predict and prepare for future events. Our aim was to determine if neuromodulation of the cerebellum influences predictive processing in motor, language, and social domains. We used anodal and cathodal transcranial direct current stimulation (tDCS) to apply 2 mA of current to right lobule VII of the cerebellum (4 cm lateral to the inion and 1 cm down); in a third condition, no significant current was delivered (sham tDCS). TDCS is thought to modulate neural activity through either depolarizing (anodal, or excitatory, tDCS) or hyperpolarizing (cathodal, or inhibitory, tDCS) the resting membrane potential. Participants (14 males; mean age 19.93 ± 1.86 years old) completed a motor learning task (the serial response time task [SRT]), a sentence completion task, and the Cyberball social ball-game task after receiving 20 minutes of anodal, cathodal, and sham tDCS in three sessions spaced one week apart. Each task had both predictive and non-predictive conditions. For the SRT, there were repeated motor sequences which enabled prediction of the next button-press, as well as randomly-ordered trials. In the sentence completion task, some sentences had contexts that very strongly predicted the final word of the sentences, whereas others provided no strong indication of how the sentence should be completed. In the Cyberball task, there was a condition in which the other players tossed the ball in a predictable way, and blocks in which there was no predictable pattern to their behavior. We expected that cerebellar tDCS would specifically alter performance during predictive conditions, with less effect on non-predictive trials. Consistent with this hypothesis, trends showed that response times were slower during predictive blocks of the SRT following anodal and cathodal stimulation, with no effect of stimulation during the randomly-ordered blocks. In the Cyberball task, participants learned to throw the ball to more favorable players earlier following active tDCS than following sham tDCS. In both of these tasks, participants performed better when they did not receive active tDCS, suggesting cerebellar neuromodulation disrupted optimal performance. However, in the sentence completion task, trends indicated that anodal stimulation led to faster response times when words were presented in a random, scrambled order; there was no effect of cerebellar tDCS during predictive and non-predictive sentences. Although statistically insignificant and therefore not definitive, the findings from this study do support the idea that the cerebellum is involved in predictive processing, and suggest that tDCS can modulate the cerebellum’s ability to implicitly learn patterns in motor, cognitive, and social tasks.