Multicenter Validation of Cerebellar Motor Learning Biomarker for Spinocerebellar Degenerations

Objectively quantifying cerebellar motor dysfunction remains a challenge in clinical trials. The Adaptability Index, derived from the Prism Adaptation Test, evaluates motor learning but lacks multicenter validation. We aimed to validate the Adaptability Index as a robust biomarker across multiple hospitals and investigate its relationship with the Scale for the Assessment and Rating of Ataxia. We enrolled 74 patients with cerebellar degeneration across seven hospitals using a standardized touchscreen system. We analyzed the correlation between Adaptability Index and the Scale for the Assessment and Rating of Ataxia and evaluated inter-hospital robustness using Linear Mixed-effects Modeling. Consistency of Adaptability Index variance between distinct genotypes (Spinocerebellar Ataxia type 6 and type 31) was assessed using Levene's test and normalized standard deviations. The Adaptability Index correlated negatively with the Scale for the Assessment and Rating of Ataxia (r = - 0.364, p < 0.01). Notably, Linear Mixed-effects Modeling revealed negligible inter-hospital variance, confirming robustness against site-specific biases. Levene's test indicated consistent Adaptability Index variance across these genotypes (p = 0.41). However, the Adaptability Index exhibited higher resolution than the Scale for the Assessment and Rating of Ataxia in detecting phenotypic variability (larger normalized standard deviations), particularly in the Spinocerebellar Ataxia type 31 cohort, likely reflecting cerebellar motor learning capacity distinct from motor execution deficits. The Adaptability Index is a valid, genotype-stable, and site-independent biomarker of cerebellar motor learning. Its high resolution and robustness make it a promising objective endpoint for multicenter interventional trials targeting cerebellar dysfunction.