Assessing cognitive change in Parkinson's disease: Development of practice effect-corrected reliable change indices

Abstract

Serial assessments of cognitive functioning in individuals with Parkinson's disease (PD) are commonly used for the detection of incident dementia and neurobehavioral changes associated with treatments using neuromodulation (e.g., deep brain stimulation) and pharmacological agents (e.g., cholinesterase inhibitors). This study provides test–retest stability, expected practice effects, and practice-corrected reliable change indices for several commonly used neuropsychological tests from 62 older adults with mostly mild PD who underwent repeat evaluations approximately 17 months apart. At the group level, results showed adequate test–retest reliability (Spearman's rho range = 0.24–0.86) and generally small practice effects (Cohen's d range = 0.00–0.50). Application of reliable change indices using 90% confidence intervals showed the expected number of individuals (generally 10% or fewer) with statistically meaningful improvements (range = 0–12%) or declines (range = 2–8%) in cognitive performance at retest. Limitations discussed include ceiling effects at test baseline, sample homogeneity, interpretative cautions, and generalizability of study results. These data may be useful to researchers and clinicians interested in determining the statistical significance of changes in cognitive test performance in persons with PD over a 1- to 2-year interval.