The present study aims to define a protocol of electrical stimulation of the cerebellum via transcranial direct current stimulation (tDCS) combined with a virtual reality protocol to assist the rehabilitation of social skills in adolescents and young adults with childhood ataxia.

Taking into account the high neuronal density of the cerebellum, its strong connection with the cerebral cortex, and its involvement in motor, cognitive and affective processes, as well its involvement in social prediction abilities, the investigators hypothesized that excitatory stimulation of the cerebellum might improve social prediction abilities in adolescents and young adults with childhood ataxia. Moreover, as suggested by previous studies investigating the effect of tDCS in paediatric population, the investigators expected that tDCS will be safe and well tolerated. Such a result would encourage the use of non-invasive brain stimulation in the rehabilitation of social skills in childhood ataxia.

The investigators planned a single centre, randomized stratified, double-blind, sham- controlled design.

Adolescents and young adults with childhood ataxia will be recruited and randomly assigned to two different groups: the active-tDCS group and the sham-tDCS group. Each group will undergo a multi-sessions (8 sessions) intervention during which tDCS will be delivered over the cerebellum. The stimulation will be paired with a virtual reality VR training. The Virtual Reality (VR) training will exploit a design based on probabilistic learning of social events in child-friendly environments. During the training, participants will be asked to conquer some goal/objects by predicting the behaviour of some competing virtual avatars whose actions should be probabilistically learned. Based on the same structures, two different child-friendly scenarios will be created and they will be respectively used in the pre- and in the post-training evaluation sessions (scenario A) or in association with the tDCS protocol throughout the 8 sessions of intervention (scenario B).

Participants' abilities of social prediction (primary outcome) will be tested through a validated computer based Action prediction task assessing participants' abilities in predicting others' actions based on previous experience. This experimental paradigm comprises a probabilistic learning (familiarization) phase and a testing phase. In the familiarization phase participants are asked to observe an actor performing two different types of grasping actions (such as grasping movement of an apple for eating the apple or for offering the apple to another partner) in different colour-cued contexts. They are asked to recognise actor's intention.

Crucially, the probability of co-occurrence between one action and the colour-cued context is implicitly biased with pre-established probability of association. In the testing phase, the same videos are presented but their length is dramatically shortened via temporal occlusion before the action is completed. In this way, since the movement kinematics is ambiguous, an observer would use the previously learned association with context to predict the fate of the action, and responses should be biased toward the contextual priors. A control non-social prediction task with a similar structure will be also used. A standard neuropsychological assessment (NEPSY-II) before and after the training will allow assessing the generalizability of the effects to general social perception abilities, in particular Theory of Mind and affect recognition (Secondary Outcomes). In the post-training and in the follow-up evaluation session (one month after the end of the intervention) the training acceptability and the quality of life assessments will be performed.

The protocol will allow testing the efficiency of the combined tDCS+ VR training in:

• enhancing social prediction abilities in childhood ataxia;
• enhancing implicit learning abilities, even in non social contexts;
• improving theory of mind abilities;
• improving patients' quality of life;
• further investigating the safety and tolerability of tDCS.

Inclusion criteria:

• Adolescents and young adults who aged 11-25 years
• Childhood ataxia
• Intelligence Quotient > 80

Exclusion Criteria:

• Presence of contraindication to tDCS (history of epilepsy, migraine, severe brain trauma; presence of metal in the brain/skull or implanted neurostimulator, cardiac pacemaker; state of pregnancy)
• Intake of or withdrawal from some drugs potentially changing the seizure threshold
• Presence of comorbidity with an important medical conditions
• Severe sensorial, motor and/or behavioural problems that could interfere with the use of GRAIL/VR technology