The Effect of Lee Silverman Voice Treatment (LSVT) on Parkinsonian Phonation: Nonlinear Dynamic, Perturbation, and Perceptual Analysis

Prof. Seong-Hee Choi, PhD, Department of Audiology and Speech-Language Pathology and Research Institute of Biomimetic Sensory Control, College of Medical Sciences, Catholic University of Daegu, 5 Gumgokri, Hayangup, Kyungsansi, Kyungsanbookdo, Korea e-mail: [email protected] tel.: +82 53 850 3841 Background & Objectives: Lee Silverman Voice Treatment (LSVT) has been widely accepted to improve Parkinsonian voice, speech, and swallowing. Acoustic analysis has been used to measure the treatment effectiveness for dysphonia. Unlike traditional perturbation analysis, a nonlinear dynamic approach has been applied to reliably quantify both periodic and aperiodic voice signals. The present study aims to investigate the effectiveness of LSVT on Parkinsonian voice with acoustic and perceptual analysis for evidence-based practice (EBP) and to evaluate whether nonlinear dynamic methods can be used to quantify the aperiodic phonation which is frequently exhibited in Parkinsonian voice. Methods: Fifteen Idiopathic Parkinson’s Disease (IPD) patients were enrolled and randomly assigned to a control (no treatment) (N=7) or treatment group (N=8). The standard LSVT program was administered by a speech language pathologist four times a week for four weeks, for a total of 16 sessions with the experimental PD group. Nonlinear dynamic methods with traditional perturbation measures were used to test treatment effects for a clinical quantification of aperiodic Parkinsonian phonation. All acoustic and perceptual parameters with /a/ sustained vowels were obtained before and four weeks after the voice therapy in the experimental group and before and four weeks after no treatment in the control group. Three listeners rated general vocal impairments with sustained /a/ vowel segments from a randomly ordered dataset for the control and experimental group before and four weeks after voice therapy. Results: Results demonstrated that both % jitter (p=0.039) and D2 (p=0.023) were significantly lower following voice therapy, whereas % shimmer values were not. For the PD control group (no treatment group), no significant differences were revealed before versus after four weeks in all acoustic parameters; % jitter (p=0.875), % shimmer (p=0.250), and D2 (p=0.078). In addition, perceived voice quality was improved following voice therapy for the treated group, while no change was found between before and four weeks after in the control group. Discussion & Conclusions: LSVT improves aperiodicity and supplements prior treatment outcomes. These findings suggest that the nonlinear measure of correlation dimension (D2) can be applicable for the characterization of aperiodic Parkinsonian voice, and the treatment effects of LSVT may provide useful information to clinicians for evidence-based decision making in Parkinsonian voices. (Korean Journal of Communication Disorders 2011;16:335-345)