Dystonia in Parkinson's disease: Clinical characteristics and predictors of treatment response.
BACKGROUND Dystonia causes pain and disability in Parkinson's disease (PD) but remains inadequately characterized. OBJECTIVE To characterize temporal features, somatotopic distribution, clinical correlates, and treatment response of dystonia in PD. METHODS We performed a retrospective, observational case-control study with longitudinal treatment-response assessment of patients seen between 2014 and 2025 at University of Cincinnati. Data from medical charts of individuals with PD and dystonia were compared to those without dystonia from the same source population. Clinical variables included age at PD onset, disease duration, dystonia phenotype and timing, levodopa equivalent daily dose (LEDD), Hoehn and Yahr (H&Y) stage, and Movement Disorder Society-Unified Parkinson's Disease Rating Scale Part III (MDS-UPDRS-III) scores. Multivariable logistic regression analyses were performed to summarize adjusted associations. RESULTS Data on 246 PD patients (164 with and 82 without dystonia) were analyzed. Focal dystonia was the predominant phenotype (75.0%), most commonly involving lower limbs, followed by cervical dystonia. Age (odds ratio [OR], 0.93; P < 0.001), disease duration (OR, 0.90; P = 0.002), and MDS-UPDRS III scores (OR, 0.96; P = 0.002) were inversely associated with dystonia, whereas advanced H&Y stage ≥3 (OR, 3.61; P = 0.002) was directly associated with it. Levodopa dose increases improved dystonia in 51.6% of patients, and botulinum neurotoxin (BoNT) improved dystonia (OR, 3.01; P = 0.05). CONCLUSION PD-associated dystonia most commonly affects lower limbs and is associated with younger age at PD onset, shorter disease duration, and milder overall motor severity of PD.