Development of a cognitive composite for measuring change in progressive supranuclear palsy

Introduction: Individuals with progressive supranuclear palsy (PSP) experience cognitive changes that are challenging to follow without a validated neuropsychological test battery to measure progression. This study describes a composite measure to evaluate cognition in individuals with PSP. Methods: Baseline cognitive test data from 486 participants with PSP in the PASSPORT (NCT03068468) study included the Repeatable Battery for Assessment of Neuropsychological Status (RBANS), Color Trails Test (CTT) parts 1 and 2, letter-number sequencing, and letter fluency. Data were analyzed using summary statistics and a matrix of Pearson correlations. A hypothetical factor structure was constructed and validated. Results: Observed correlations were highest for scores between story memory and story recall (correlation coefficient = 0.78) and lowest for scores between letter fluency and picture naming (correlation coefficient = 0.11), and picture naming and figure copy (correlation coefficient = 0.12). After excluding picture naming and Color Trails Test (CTT) parts 1 and 2, a 3-factor structure was hypothesized for the remaining 13 tests. Confirmatory factor analysis demonstrated goodness of fit within acceptable limits (comparative fit index and Tucker-Lewis index = 0.98, standardized root-mean-square residual and root-mean-square error of approximation = 0.05-0.06). Mixed-model repeated-measures analysis of change from baseline to week 52 in RBANS and PSP cognitive composite score produced mean-to-standard-deviation ratios of 0.418 and 0.780, respectively. Conclusions: This novel composite endpoint, based on RBANS and designed to account for motor impairments in PSP, improves on current cognitive assessments PSP.