Performance of the Movement Disorders Society criteria for prodromal Parkinson's disease: A population-based 10-year study.

OBJECTIVE: We aimed to identify prodromal Parkinson's disease (PD) and its predictive accuracy for incident PD in an unselected elderly population and to estimate the relevance of this approach for future neuroprotection trials. METHODS: We applied the recently published Movement Disorders Society (MDS) research criteria for prodromal PD to participants of the prospective population-based Bruneck Study of the 2005 assessment (n = 574, ages 55-94 years). Cases of incident PD were identified at 3-year, 5-year, and 10-year follow-up visits. We calculated predictive accuracies of baseline prodromal PD status for incident cases, and, based on them, estimated sample sizes for neuroprotection trials with conversion to PD as the primary outcome. RESULTS: Baseline status of probable prodromal PD (n = 12) had a specificity in predicting incident PD of 98.8% (95% confidence interval, 97.3%-99.5%), a sensitivity of 66.7% (29.6%-90.8%), and a positive predictive value of 40.0% (16.7%-68.8%) over 3 years. Specificity remained stable with increasing follow-up time, sensitivity decreased to 54.6% (28.0%-78.8%) over 5 years and to 35.0% (18.0%-56.8%) over 10 years, whereas positive predictive value rose to 60.0% (31.2%-83.3%) and 77.8% (44.3%-94.7%), respectively. Sample size estimates at 80% power in an intention-to-treat approach ranged from 108 to 540 patients with probable prodromal PD depending on trial duration (3-5 years) and effect size of the agent (30%-50%). CONCLUSIONS: Our findings show that the MDS criteria for prodromal PD yield moderate to high predictive power for incident PD in a community-based setting and may thus be helpful to define target populations of future neuroprotection trials. © 2018 International Parkinson and Movement Disorder Society.