Differentiating Gait Behaviors between Early-Stage Dementia with Lewy Bodies and Parkinson’s Disease

Abstract

Dementia with Lewy bodies (DLB) and Parkinson’s disease (PD) can be difficult to differentially diagnose, given the overlap in clinical and pathological features. Evidence suggests gait may be a sensitive and selective marker of neurodegeneration. The complexity of the behavior combined with the multilevel neural circuity required for effective gait provides an opportunity to assess how pathological overlap with these neural networks can result in distinct gait changes between DLB and PD. Given both PD and DLB patients have impaired basal ganglia function, research postulates that a compensatory shift to greater cortical control is employed to maintain safe walking. Therefore, it is theorized that a cognitively demanding task while walking (dual tasking) may unmask and amplify subtle differences in walking impairments between PD and DLB patients. Evidence shows that PDD patients have greater swing and stance time asymmetry compared to patients with DLB during self-paced walking. However, no studies to date have assessed gait profiles during single-task walking between individuals with early-stage PD and DLB or whether dual tasking may exacerbate subtle walking differences between these two Lewy body disorders. Thus, this thesis aimed to (i) comprehensively characterize gait in Early PD and Early DLB, and evaluate whether there are differences in discrete gait characteristics between Early PD and Early DLB patients during normal walking, (ii) assess the sensitivity and specificity of specific gait characteristics in accurately discriminating between individuals with Early PD and Early DLB and (iii) investigate whether increasing cognitive load by modifying task complexity (i.e., serial 1 vs. serial 7s dual-task walking) impacts gait behaviors between Early PD and Early DLB patients.

Forty-six Lewy Body disorder patients (26 PD, 20 DLB) that were within five years since diagnosis (‘early’ stage) as well as 16 healthy older adults walked across a 6-meter pressure sensor walkway under three conditions, (i) normal self-paced walking, (ii) walking while subtracting 1s from 100 and (iii) walking while subtracting 7s from 100. To determine whether walking differences existed between Early PD and Early DLB patients during normal walking, 16 spatiotemporal gait measures were evaluated during self-paced gait and compared between groups. Results showed during self-paced gait, Early DLB had significantly worse gait performance in some features of pace (velocity, p=0.008; step length, p=0.042) and rhythm (stance time, p=0.015) compared to Early PD patients. An assessment of outcome measure accuracy revealed step time (AUC=0.709), stance time (AUC=0.739) and step velocity variability (AUC = 0.719) were able to discriminate Early DLB patients from Early PD patients with moderate accuracy. To assess whether increasing cognitive load unmasked and/or exacerbated gait differences between the two Lewy body disorders, 16 spatiotemporal gait metrics were measured across the serial 1s and serial 7s dual-task conditions between the Early PD and Early DLB groups. The study found increasing cognitive load during dual-task walking (serial 1s vs. serial 7s) did not expose or intensify any gait differences between Early PD and Early DLB. However, a significant main effect of condition was seen for step velocity variability (F (1,38) =4.684, p=0.037) whereby step velocity variability was greater during the serial 7s dual task compared to the serial 1s task regardless of disease group. In closing, this study found a normal walking assessment revealed several differences in gait behaviors between Early DLB and Early PD, some of which had a moderate ability to discriminate between the groups. This study aids in understanding unique gait profiles at the early stages of the disease which is critical if gait is ever to be used as a tool for predicting disease trajectory in at-risk individuals.