**1. Introduction**

Freezing of gait (FOG) is a dramatic phenomenon frequently affecting patients with Parkinson's disease (PD) [1], causing falls, mobility restrictions, and poor quality of life [2–4]. FOG is defined as a brief, episodic absence or marked reduction of forward progression of the feet despite the intention to walk [5], which typically occurs when initiating or modulating gait (e.g., turning, obstacle crossing, and so on).

Gait initiation (GI) is a highly challenging task for the balance control system and is of particular interest in the study of neural control of upright posture maintenance during whole-body movement [6]. Specifically, this task allows the precise assessment of anticipatory postural adjustments (APAs; i.e., muscular synergies that precede GI), aiming to destabilize the antigravity postural set by shifting the center of pressure (CoP) to generate a gravitational moment favoring the center of mass (CoM) forward acceleration [7]. APAs are considered a motor program controlled by feedforward mechanisms regulated by the supraspinal locomotor network [8–12]. The selection and scaling of appropriate APAs rely on the ability to use sensory information to determine the body positioning relative to the environment prior to step execution [13,14] and on the intended forthcoming movement (natural, slow, fast, obstacle, and so on) [7,15–17]. Striatal dopamine loss, a pathophysiological hallmark of PD, greatly impacts the production of APAs at GI and particularly

**Citation:** Palmisano, C.; Beccaria, L.; Haufe, S.; Volkmann, J.; Pezzoli, G.; Isaias, I.U. Gait Initiation Impairment in Patients with Parkinson's Disease and Freezing of Gait. *Bioengineering* **2022**, *9*, 639. https://doi.org/ 10.3390/bioengineering9110639

Academic Editors: Christina Zong-Hao Ma, Zhengrong Li and Chen He

Received: 27 September 2022 Accepted: 15 October 2022 Published: 2 November 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

the CoP displacement and velocity [12]. Only a few studies have specifically investigated the GI task in Parkinsonian patients with a history of FOG (PDF), non-implanted for deep brain stimulation (DBS), and after withdrawal of dopaminergic medication (meds-off state). The stimulation and medication condition should be carefully considered, as both DBS and dopaminergic drugs can variably influence posture and gait in PD [12,18–26]. Overall, these studies showed conflicting results, with APAs being reported as normal [27–29] or multiple and hypometric [10,30]. Several methodological discrepancies may account for such different findings, including a non-standardized meds-off state [27], imposed (predefined) feet positioning [29], cueing [10,29,30], and specific instructions on the execution of the GI task (e.g., to start walking as quickly as possible [30,31] or while performing a cognitive task [28]). All of these factors can significantly impact and alter APA expression at GI. Specifically, a cued start signal influences motor programming towards normalization, especially in PDF [9,18], similar to the improvements seen with the administration of levodopa for self-generated step initiation [18]. Moreover, the initial feet position [12,22,32] and posture [33–35] can significantly impact the biomechanical features of APAs at GI.

Postural changes in particular would have a detrimental impact on APA production. An altered representation of the body position (egocentric representation) may determine a functional re-organization of the supplementary motor area (SMA)-proper, hampering selection and re-scaling of APAs to adapt to the altered postural framework and bradykinetic stepping [33,34,36–38].

Our study aims to describe GI alteration in patients with PD and FOG, accounting for the influence of anthropometric measurements (AMs) and the base of support (BoS) and investigating their relationship with the initial posture. We have also addressed the relative timing and movement sequence of each body segment subserving GI.

#### **2. Materials and Methods**

#### *2.1. Subjects*

We recruited 23 patients with idiopathic PD (according to the U.K. Brain Bank criteria) and an unambiguous, previous history of FOG (PDF; i.e., patients reporting episodes of FOG on a daily basis prior to the experiment). On the day of the experiment, the presence of FOG was confirmed with a clinical evaluation by an experienced neurologist (I.U.I.). In addition, 20 patients with PD and no previous history of FOG (PDNF) and 23 healthy controls (HCs) were also included. HCs and PDNF patients were chosen to match in terms of demographic and clinical data with the PDF group. Subjects with neurological diseases other than PD, including cognitive decline (i.e., Mini-Mental State Examination score < 27), vestibular disorders, and orthopedic impairments that could interfere with gait were excluded. Disease severity was evaluated with the Unified Parkinson's Disease Rating Scale motor part (UPDRS-III).

#### *2.2. Experimental Protocol*

Patients were investigated in practical meds-off state, i.e., in the morning after overnight withdrawal (>12 h) of all dopaminergic drugs.

Kinematic data were recorded using an optoelectronic system with six cameras (sampling rate 60 Hz, SMART 1.10, BTS, Garbagnate Milanese, Italy) and a set of 29 markers placed on anatomical landmarks (temples, acromions, lateral humeral condyles, ulnar styloids, anterior superior iliac spines, middle thighs, lateral femoral condyles, fibula heads, tibial anterior side, lateral malleoli, Achilles tendon insertion, fifth metatarsal heads, halluxes, the seventh cervical vertebra [C7], point of maximum kyphosis, and middle point between the posterior superior iliac spines) [39,40]. Eight additional technical markers were placed on the trochanters, the medial condyles, the medial malleoli, and the first metatarsi for a short calibration trial, which allowed the computation of the AMs and BoS measurements [11,12,41]. Markers traces were filtered with a fifth-order lowpass Butterworth filter (cut-off frequency: 10 Hz [41]). Dynamic measurements were recorded with a force plate working at a sampling rate of 960 Hz (KISTLER 9286A, Winterthur, Switzerland). The

resulting signal was low-pass filtered (fifth-order lowpass Butterworth filter) with a cut-off frequency of 30 Hz [11,42].

At the beginning of each trial, subjects stood upright on the force platform at a comfortable stance position for about 30 s. The initial stance position was not standardized to prevent modification of the subject's usual motor strategy to initiate gait [12].

Participants were instructed to start walking after a self-selected period from a verbal signal, in order to avoid any effect of cueing on GI. The instruction given was as follows: "Start walking at the moment of your choice". Subjects were not instructed on the stepping leg to use and they moved at their own pace until the end of the walkway. After a training session, at least three consecutive trials were recorded. The principal investigator supervised all participants during the experiment.
