Non-linear Analysis of Heart rate Variability Improves Differential Diagnosis Between Parkinson Diseases and Multiple System Atrophy

Aims: Parkinson's disease (PD) and multiple system atrophy (MSA) are neurodegenerative disorders characterized by motor "parkinsonian" symptoms and non-motor symptoms related to autonomic nervous system (ANS) dysfunction. The latter can be quantified with the analysis of Heart Rate Variability (HRVa) and of its complexity. In this study nonlinear (NL) HRV complexity parameters were calculated to assess their predictive accuracy as markers of “disease” useful for early differentiation between PD and MSA in parkinsonian syndromes of uncertain diagnosis. Study Design: Observational study. Place and Duration of Study: Clinical Physiology-Biomagnetism Center, Policlinico A. Gemelli, Rome Italy. Patients enrolled from January 2010 to October 2013. Methodology: 51 patients [25 with “certain” diagnosis of PD, 9 with a “highly probable” diagnosis of MSA and 17 with parkinsonian syndromes of uncertain neurological definition (6 with “undefined parkinsonism” and 11 with “suspected MSA”)] and 40 age-matched healthy control subjects were studied. Short-term NL HRVa was performed during daily activity and during REM/NREM sleep from 24 h ECG recordings. Discriminant analysis (DA) was used to identify which NL HRV parameters (or their combination) were efficient to differentiate between PD and MSA in cases of uncertain diagnosis. Results: Compared with healthy controls, most NL HRV parameters were significantly altered in patients (p<0.05), during both active and passive awakeness and during sleep. Most evident HRV abnormalities were found during active awakeness in MSA. DA of recurrence plot parameters provided the best predictive accuracy (76.5%) for the classification of parkinsonian patients with uncertain diagnosis. Conclusion: NL HRVa is efficient in differentiating MSA from PD and may improve earlier diagnosis in patients with parkinsonian symptoms of uncertain nature, useful to address second level diagnostic steps and to guide more individualized drug treatment.