Kinetic–rest tremor dominance for differentiating dystonic tremor from Parkinson's disease

Abstract

Background Differentiating dystonic tremor (DT) from Parkinson's disease (PD) remains challenging, particularly when dystonic signs are subtle, and DT is a major contributor to scans without evidence of dopaminergic deficit. Although dopamine transporter imaging can support the distinction, it is costly and not universally accessible. Simple diagnostic indicators less reliant on subjective clinical interpretation are needed. Methods In a dopamine transporter imaging–supported cohort (n = 102; DT, 51; PD, 51), we used accelerometry to characterize ordinal patterns of tremor amplitude across rest, postural, and kinetic (spiral drawing) conditions. Candidate ordinal rules were evaluated, and the rule with the highest classification accuracy was selected for subsequent blinded visual assessment by three raters within the same cohort. Results Accelerometry showed greater kinetic tremor amplitude in DT and greater rest tremor amplitude in PD, whereas postural tremor amplitude did not differ significantly between groups. Among candidate ordinal rules, the pairwise comparison of kinetic versus rest tremor amplitude achieved the highest overall diagnostic performance for identifying DT (90% sensitivity, 85% specificity, 88% accuracy). Blinded visual classification of this rule showed excellent inter-rater agreement (Fleiss’ κ = 0.92), high concordance with accelerometer-based classification (Cohen's κ = 0.86), and 90% sensitivity, 88% specificity, and 89% accuracy. Conclusions A simple visual comparison of tremor amplitude during spiral drawing versus at rest required no specialized equipment and showed high reproducibility in blinded ratings. Although probabilistic and requiring validation in independent cohorts, this approach may serve as a practical bedside adjunct for distinguishing DT from PD.