The long term objective of this research is to identify quantitative biomechanical parameters of postural instability in patients with Parkinson's disease (PD) that can in turn be used to assess fall risk. Currently, clinical assessments in PD are not sufficiently sensitive to predict fall risk, making a history of falls to be the best predictor of a future fall. Identifying biomechanical measures to predict risk of falls in PD would provide a quantitative justification to implement fall-reducing therapies prior to a first fall and help prevent the associated debilitating fractures or even morbidity. While past biomechanical studies have shown the presence of balance deficits in PD patients, which often include a broad spectrum of disease stages, compared to healthy controls (HC), no studies have assessed whether such parameters can distinguish the onset of postural instability prior to clinical presentation, and if such parameters persist following clinical presentation of postural instability. Toward this end this study had three goals:​

1. Determine if biomechanical assessment of a quasi-static task, postural sway, could provide preclinical indication of postural instability in PD.
2. Define a mathematical model (based on principal component analysis, PCA) with biomechanical and clinical measures as inputs to quantitatively score earlier postural instability presence and progression in PD.
3. Investigate if biomechanical assessment of a dynamic task, gait initiation, could provide preclinical indication of postural instability in PD.

Specific Aim 1 determined that some biomechanical postural sway variables showed evidence of preclinical postural instability and increased with PD progression. This aim distinguished mild PD (Hoehn and Yahr stage (H&Y) 2, without postural deficits) compared to HC suggesting preclinical indication of postural instability, and confirmed these parameters persisted in moderate PD (H&Y 3, with postural deficits). Specifically, trajectory, variation, and peak measures of the center of pressure (COP) during postural sway showed significant differences (p < .05) in mild PD compared to healthy controls, and these differences persisted in moderate PD. Schwab and England clinical score best correlated with the COP biomechanical measures. These results suggest that postural sway COP measures may provide preclinical indication of balance deficits in PD and increase with clinical PD progression.

Specific Aim 2 defined a PCA model based on biomechanical measures of postural sway and clinical measures in mild PD, moderate PD, and HC. PCA modeling based on a correlation matrix structure identified both biomechanical and clinical measures as the primary drivers of variation in the data set. Further, a PCA model based on these selected parameters was able to significantly differentiate (p < .05) all 3 groups, suggesting PCA scores may help with preclinical indication of postural instability (mild PD versus HC) and could be sensitive to clinical disease progression (mild PD versus moderate PD and moderate PD versus HC). AP sway path length and a velocity parameter were the 2 primary measures that explained the variability in the data set, suggesting further investigation of these parameters and mathematical models for scoring postural instability progression is warranted.

Specific Aim 3 determined that a velocity measure from biomechanical assessment of gait initiation (peak COP velocity towards the swing foot during locomotion) showed evidence of preclinical postural instability in PD. Because balance is a complex task, having a better understanding of both quasi-static (postural sway) and dynamic (gait initiation) tasks can provide further insight about balance deficits resulting from PD. Several temporal and kinematic parameters changed with increasing disease progression, with significant difference in moderate PD versus HC, but missed significance in mild PD compared to HC. Parkinson's Disease Rating Scale (UPDRS) and Pull Test clinical scores best correlated with the biomechanical measures of the gait initiation response. These results suggest dynamic biomechanical assessment may provide additional information in quantifying preclinical postural instability and progression in PD.

In summary, reducing fall risk in PD is a high priority effort to maintain quality of life by allowing continued independence and safe mobility. Since no effective screening method exists to measure fall risk, our team is developing a multi-factorial method to detect postural instability through clinical balance assessment, and in doing so, provide the justification for implementing fall reducing therapies before potentially debilitating falls begin