Dynamic Measures of Balance during a 90° Turn in Self-Selected Gait in Individuals with Mild Parkinson’s Disease

Featured Application

The key finding of this project suggests biomechanical metrics that may be useful in screening individuals with Parkinson’s disease for loss of dynamic balance while walking.

Abstract

The risk of falls is common in the aging and Parkinson’s disease (PD) populations. There is limited research on dynamic gait stability, i.e., margin of stability (MOS), in the PD population while turning. The purpose of this exploratory study was to examine the dynamic balance control in individuals with mild to moderate PD and healthy individuals while walking and performing 90° turns utilizing computerized three-dimensional gait analysis. Specifically, we examined the anteroposterior and mediolateral margins of stability and their effect on the dynamic balance during turning in participants. A total of 11 individuals with mild to moderate idiopathic PD and 10 healthy controls (CON) participated in this study. Individuals with PD were tested during the “on phase” of PD medication. The Vicon Nexus camera system, force plates, and Visual3D software were utilized for three-dimensional motion analysis of three right and three left turning trials. A mixed-model ANOVA was used to analyze the primary dependent variables of dynamic balance (p < 0.0028) and the secondary dependent variables of spatiotemporal gait parameters (p < 0.0056). No significant differences in the spatiotemporal parameters or dynamic balance variables were observed between the groups. Gait velocity, center of mass-center of pressure (COM–COP) inclination angle at midstance, and MOS variables showed marginally significant group differences. We found no differences in dynamic balance during gait while performing turns betweenthe healthy elderly and individuals with PD. This finding may be related to the early stage of disease progression in the PD group.

1. Introduction

Fall incidence in adults aged 65 years and older is a critical health problem, as 28% report falling each year, resulting in 36 million falls per year (2020) in the United States [1]. Falls are significantly more prevalent in individuals with Parkinson’s disease (IPD), as they are five times more likely to fall compared to healthy older adults [2]. Prospective studies estimate that between 45% and 65% of IPD fall each year, with 39% experiencing recurrent falls and a high rate of injurious falls [3,4,5]. Parkinson’s disease (PD), the second most prevalent neurodegenerative disorder, is caused by a progressive loss of the dopaminergic neurons in the substantia nigra, which leads to dopamine deficiencies throughout the basal ganglia [6]. This progressive neurodegeneration leads to motor deficits in generating and executing voluntary movements, bradykinesia, rigidity, postural instability, and gait deficits [6,7,8]. PD–related gait impairments may include decreased gait velocity, hypokinesia (reduced step length and arm swing), shuffling gait with increased cadence, a narrower base of support (BOS), impaired gait rhythmicity and automaticity, freezing of gait (FOG), and impaired ability to perform dual tasks while walking [3,9,10,11]. These gait deficits contribute to increased fall risk during walking in the home and community, such as when navigating turns, obstacles, and varied surfaces. As the disease progresses, the decline in gait and postural control mechanisms results in high fall risk, reduced functional independence, and a lowered quality of life.

Falls are eight times more likely to occur during turns than straight–path walking in older adults [12]. Turns are an essential part of gait and often are the cause of falls, especially in the aging population with PD when considering the dynamic, unexpected, and fast turns of everyday life and the functional impairments experienced [4]. Previous research supports that IPD approach turns slower, turn wider with more steps, take longer to turn, turn at a slower peak turn speed, and are less accurate in their turns compared to healthy controls [13,14,15,16]. Compromised turning strategies were reported even in individuals with very mild PD [13,14,15,16]. Axial coordination and gait stability can be significantly disrupted in IPD due to increased muscle rigidity and anticipatory postural adjustment impairments [15]. Axial mobility, i.e., trunk rotation, intersegmental coordination, and dynamic changes in the BOS, are related to turning difficulty in IPD [13,16]. Individuals with mild PD display “en bloc turning”, an almost simultaneous rotation of their head and upper trunk instead of the cranial–to–caudal rotation pattern typically seen [16], which has been associated with lateral instability and reduced forward progression with extra steps during turning [15]. Interlimb temporal coordination, which is crucial for maintaining mediolateral (M/L) dynamic stability [17], is typically impaired along with gait rhythm in IPD and may be associated with increased instability during activities that require interlimb coordination [18]. Previous research on turning strategies during walking in IPD has focused primarily on temporal–distance gait parameters, such as step number, step length, and speed of turns; however, research that examines the dynamic control of balance during gait with turns is lacking. This research may provide insight into the mechanisms underlying postural instability during turning in IPD.

Using an inverted pendulum (IP) model, derived biomechanical variables have been applied to assist in understanding the mechanisms of standing postural control and dynamic balance during gait and turning [19,20] (Figure 1). In static situations, stability is maintained when the center of mass (COM) is constrained within the BOS [21]. Although, during gait, the COM vertical projection is outside of the BOS for the majority of the gait cycle, requiring complex balance control, healthy individuals can regulate it to maintain balance [9,17].

Recently, several biomechanical measures of balance during movement, e.g., walking or turning, have been described [15,19,21]. The COM–COP inclination angle represents the spatial relationship between the COM and COP and has been used to evaluate gait stability during 180° turning with IPD [15]. The position of the extrapolated COM (XCOM) is another variable that initially measured stability in static situations [21] but has also been used to determine dynamic stability in a recent study [19]. The XCOM accounts for the velocity of the COM [21]. The difference between the location of the XCOM and COM positioning is greater in less stable positions [21]. Balance can be predicted by knowing the XCOM and COP relative positions [19]. For example, generally, at the time of foot contact, walking stability is maintained when the COP is positioned at a specific distance behind and lateral to the XCOM [19].

The margin of stability (MOS) is the distance between the XCOM and the BOS. It is “proportional to the impulse needed to unbalance an individual” and has been validated for its application to examine and describe both static and dynamic balance/control [21,22,23]. When the BOS is decreased during walking, the MOS will decrease, and the XCOM may demonstrate a greater excursion [21]. A larger MOS may reflect increased dynamic stability, whereas a smaller MOS may reflect poorer balance control [17]. The MOS has been an effective variable in measuring dynamic stability [17,21] during gait in healthy individuals as well as those with neurological conditions [19,24], such as multiple sclerosis [24]. There is no previous research that has assessed the MOS during gait and turning in IPD, despite the known high fall risk of IPD during these activities. Therefore, the MOS may be useful to characterize dynamic balance control in IPD during turning activities.

Dynamic balance during walking with turns is impaired and often poses a high fall risk for IPD, which has been identified in previous research [13,14,15,16]. However, there are gaps in the literature regarding the underlying biomechanical deficits in the turning strategies for IPD. Few studies have examined biomechanical variables related to dynamic balance control in this population, such as COM–COP inclination angle and MOS, during walking and turning to identify possible underlying biomechanical deficits. Specifically, there is a lack of research that examines the relationship between MOS and dynamic stability in this population. Due to the limited studies on dynamic control of balance at any stage of PD, specifically regarding anteroposterior (A/P) and mediolateral (M/L) MOS, there may be features of gait and postural control during turns that are currently unknown. Therefore, the purpose of this exploratory study was to examine the dynamic balance control in individuals with mild to moderate PD and healthy individuals while walking while performing 90° turns utilizing computerized three-dimensional gait analysis. Specifically, we examined the A/P and M/L margins of stability and their effect on the dynamic balance during turning in IPD and healthy controls. These results may help inform researchers and clinicians about factors that contribute to fall risk during walking with turns in IPD.

2. Materials and Methods

2.1. Participants

The research was approved by the institutional Internal Review Board (17-267-H). Participants formed a sample of convenience from the local community and were recruited through informational flyers, newsletters, informational meetings, and conversations with medical professionals.

Eleven individuals with mild to moderate idiopathic PD and ten healthy controls (CON) participated in the study. Individuals were categorized as having mild to moderate idiopathic PD based on their clinical history, duration of symptoms, minimal to no FOG, and no fall history (Figure 2). One participant in the PD group demonstrated a score of 13/24 on the FOG Questionnaire and indicated that FOG was experienced rarely, about once a week. FOG was not experienced by this individual during data collection. Informed consent was obtained from all participants prior to data collection. Inclusion and exclusion criteria for PD and CON groups can be found in

Table 1. Inclusion and exclusion criteria for Parkinson’s disease and Control participants.

Inclusion	Exclusion
PD Group Able to safely ambulate ≥ 300 m with or without an assistive device (AD) Able to ambulate ≥14 m without AD Self-report of functional vision Diagnosis of idiopathic PD by a healthcare professional	PD Group Surgery or orthopedic injury within 6 months that may interfere with ambulation (LE joint replacements fractures, injuries) Significant medical conditions (hypertension, unstable heart disease, uncontrolled DM) Vestibular dysfunction within 6 months (BPPV, vertigo, Meniere’s disease, bilateral vestibular loss) Berg Balance Scale (BBS) ≤ 36 Montreal Cognitive Assessment (MoCA) <21 Self-report of ≥2 falls per month Diagnosis of another neurological disorder besides PD Deep brain stimulation
CON Group Same inclusion criteria as the first three items of the PD group Same decade of life and gender as a participant in the PD group	CON Group Same exclusion criteria as the first five items of the PD group Diagnosis of a neurological disease such as PD, multiple sclerosis, seizure disorder, brain injury, or history of stroke

CON (Control), PD (Parkinson’s Disease), AD (Assistive device), DM (Diabetes Mellitus), BPPV (Benign Paroxysmal Positional Vertigo).

. A score of less than 36 on the Berg Balance Scale (BBS) and/or a history of >2 falls per month indicate that the individuals are at a high risk of falling, or “fallers” [25], and were excluded from the study. Participants were also excluded if they scored less than 21 on the Montreal Cognitive Assessment (MoCA) because it indicated moderate cognitive impairment among IPD [26]. The healthy control group sample of convenience was matched by the ages (decade) and gender of participants with PD. A decade of life was defined as 50–59, 60–69, etc., and was used to categorize ages.

2.2. Instrumentation

Marker trajectories were captured at 120 Hz using the Vicon Nexus v2.6.1 motion capture software (Vicon Motion System Ltd., Oxford Metrics, UK), and 16 MX and T40 cameras. AMTI (Advanced Mechanical Technology Inc., Watertown, MA, USA) force plates were used to synchronously collect ground reaction force data (1200 Hz).

2.3. Procedures

Initial screening of potential participants was conducted over the phone to determine their eligibility for participation based on the inclusion and exclusion criteria (Table 1). Potential participants that passed the initial screening attended an in–person screening appointment in alignment with their “on” medication times. The MoCA and Berg Balance Scale were administered as screening measures for significant balance impairment, fall risk, and notable cognitive impairment. Once consent was obtained from qualified participants, they then completed the FOG Questionnaire, and data regarding demographic information (

Table 2. Demographic means ± SD for Parkinson’s (PD) and Control (CON) groups.

	PD (n = 11)	CON (n = 10)
Age (yrs)	67.5 ± 10.2	65.1 ± 7.4
Gender (M:F)	10:1	9:1
Height (cm)	1.8 ± 0.1	1.8 ± 0.1
Mass (kg)	90.5 ± 17.5	86.4 ± 12.6
BBS	53.4 ± 3.3	55.3 ± 1.6
MoCA	25.9 ± 2.8	27 ± 2.4
FOG-Q	2.7 ± 3.6	Not Tested

BBS = Berg Balance Scale, the highest score is 56 indicating no balance deficits. MoCA = Montreal Cognitive Assessment, the highest score is 30 indicating no cognitive deficits. FOG-Q = Freezing-of-Gait Questionnaire, the highest score is 24 indicating the most severe freezing of gait. FOG-Q was only used in the PD group.

), medical and PD history, fall history, and current medications were collected. A physical screen assessed postural alignment andgait, and anthropometric measurements were taken for knee and ankle widths, leg length, pelvic width, height, and depth for body modeling consistent with the Plug-In Gait (PIG) model (http://www.vicon.com (accessed on 17 April 2023)).

Data collection for three-dimensional (3D) motion walking trials with turns was performed in a single session in a quiet biomechanical laboratory setting. Participants wore their preferred athletic-type walking shoes. Men wore tight–fitting shorts and no shirt, and women donned tight-fitting shorts and a sports bra. A single trained (approximately 3 months) examiner placed 40 double–sided tape markers on the body following the full body PIG model with the use of extra tape to secure marker bases to the skin and shoes to prevent markers from falling or being kicked off when turning. Foot markers were placed directly on the participants’ shoes. The PIG model was modified by placing markers on the head of the fifth metatarsal. The PIG model marker location modifications also included thigh marker placement on the mid–lateral thigh and tibial markers on the mid–lateral shank. Wands were not used for thigh and tibial markers. See the Appendix A 

Table A1. Plug-in Gait Markers [39,40].

Marker Labels	Definition	Position on Patient
LFHD	Left front head	Left temple
RFHD	Right front head	Right temple
LBHD	Left back head	Left back of the head (defines the transverse plane of the head, together with the frontal markers)
RBHD	Right back head	Right back of the head (defines the transverse plane of the head, together with the frontal markers)
C7	7th cervical vertebra	On the spinous process of the 7th cervical vertebra
T10	10th thoracic vertebra	On the spinous process of the 10th thoracic vertebra
CLAV	Clavicle	On the jugular notch where the clavicles meet the sternum
STRN	Sternum	On the xiphoid process of the sternum
RBAK	Right back	Anywhere over the right scapula (no equivalent marker on the left side)
LSHO	Left shoulder	On the acromio-clavicular joint
LUPA *	Left upper arm	On the upper lateral 1/3 surface of the left arm (place asymmetrically with RUPA)
LELB	Left elbow	On the lateral epicondyle
LFRM	Left forearm	On the upper lateral 1/3 surface of the left arm (place asymmetrically with RUPA)
LWRA	Left wrist marker A	At the thumb side of a bar attached to a wristband on the posterior of the left wrist, as close to the wrist joint center as possible
LWRB	Left wrist marker B	At the little finger side of a bar attached to a wristband on the posterior of the left wrist, as close to the wrist joint center as possible
LFIN	Left finger	Just proximal to the middle knuckle on the left hand
RSHO	Right shoulder	On the acromio-clavicular joint
RUPA *	Right upper arm	On the upper lateral 1/3 surface of the right arm (place asymmetrically with LUPA)
RELB	Right elbow	On the lateral epicondyle approximating the elbow joint axis
RFRM *	Right forearm	On the lower lateral 1/3 surface of the right forearm (place asymmetrically with LFRM)
RWRA	Right wrist marker A	At the thumb side of a bar attached symmetrically to a wristband on the posterior of the right wrist, as close to the wrist joint center as possible
RWRB	Right wrist marker B	At the little finger side of a bar attached symmetrically to a wristband on the posterior of the right wrist, as close to the wrist joint center as possible
RFIN	Right finger	Just proximal to the middle knuckle on the right hand
SACR *	Sacral	On the skin mid-way between the posterior superior iliac spines (PSI) and positioned to lie in the plane formed by the ASIS and PSI points
LASI	Left ASIS	Left anterior superior iliac spine
RASI	Right ASIS	Right anterior superior iliac spine
LPSI	Left PSIS	Left posterior superior iliac spine (immediately below the sacroiliac joints, at the point where the spine joins the pelvis). This marker is used with the RPSI marker as an alternative to the single SACR marker.
RPSI	Right PSIS	Right posterior superior iliac spine (immediately below the sacroiliac joints, at the point where the spine joins the pelvis). This marker is used with the LPSI marker as an alternative to the single SACR marker.
LTHI	Left thigh	Over the lower lateral 1/3 surface of the left thigh in line with the hip and knee joint centers
LKNE	Left knee	Laterally, on the flexion-extension axis of the left knee
LTIB	Left tibia	Over the lower 1/3 surface of the left shank
LANK	Left ankle	On the lateral malleolus along an imaginary line that passes through the transmalleolar axis
LHEE	Left heel	On the calcaneus at the same height above the plantar surface of the foot as the toe marker
LTOE	Left toe	Over the second metatarsal head, on the mid-foot side of the equinus break between the forefoot and mid-foot
RTHI	Right thigh	Over the upper lateral 1/3 surface of the right thigh
RKNE	Right knee	Laterally, on the flexion-extension axis of the right knee
RTIB	Right tibia	Over the upper 1/3 surface of the right shank
RANK	Right ankle	On the lateral malleolus along an imaginary line that passes through the transmalleolar axis
RHEE	Right heel	On the calcaneus at the same height above the plantar surface of the foot as the toe marker
RTOE	Right toe	Over the second metatarsal head, on the mid-foot side of the equinus break between the forefoot and mid-foot

Note: Marker labels marked with an asterisk (*) are optional; however, using them improved marker tracking during dynamic trials. The model was modified in the present study with the addition of markers placed on the heads of the left and right fifth metatarsals. Additionally, for the subject calibration trial, markers were placed (later removed for walking trials) on the apex of the right and left medial femoral condyles and right and left medial malleoli.

for references regarding the validity and reliability of the PIG model, a list of markers, marker abbreviations, and placement descriptions.

Following marker placement, a static subject calibration trial was completed, which combined participant anthropometric and marker position data to create a unique skeletal model for each participant. Participants stood still in an anatomical position to capture marker positioning for 1–4 s. The ability of the cameras to see all markers was verified. The medial femoral condyle and medial malleolar markers were removed following the completion of the static calibration trial.

Following the calibration trial, each participant was given practice trials for walking the 10–m pathway at his or her self-selected pace to feel comfortable with the walkway and to assume his or her normal gait pattern. Participants were instructed to perform clean force plate strikes, defined as stepping in or near the middle of the force plate, during the walking tests. Data collection began after participants practiced self–paced forward walking until they were able to make clean force plate contacts without targeting the force plates, i.e., looking at them.

Walking data were collected and processed for one gait trial at a comfortable speed to check for accurate thigh, femoral epicondyle, and tibial marker placement. Frontal and transverse plane hip and knee joint angle data were determined, graphed, and compared to the normal walking database to ensure correct marker placement. If participants’ data fell within one standard deviation of the normal kinematic bands, the correct marker placement was verified, and testing continued. Following the kinematic check, forward walking data were collected until five clean force plate strikes were verified for both the right and left legs. To ensure the safety of the participants, a gait belt was available during the data collection process, but its use was up to the examiners’ discretion following the initial screening. Rest breaks were given as necessary.

Possible errors in joint angle and moment calculation following static calibration could occur if markers were placed incorrectly at anatomical locations or if anatomical markers moved during dynamic motions due to skin motion. In order to prevent these possible calculating errors, marker placement was checked frequently in between walking trials and during rest breaks.

Three walking task conditions, which had been approved by the institutional IRB were collected for the larger study: (1) termination of gait, (2) performing a 90° turn, and (3) stepping over an obstacle. All data for all conditions were collected in one session for each participant. However, only turning data were analyzed and reported for the present study.

For the walk–with–turn trials, participants followed a modified version of a procedure performed by Yang et al. [15]. Prior to data collection, participants performed multiple practice trials at a self-selected speed. For left turns, participants walked in the positive y–direction (laboratory coordinate system) at a self–selected speed until reaching the second force plate, where they placed their right foot squarely inside the plate, took a 90° turn to the left, and continued walking in the negative x–direction (Figure 3). A cone was set on the corner of the force plate to mark the desired location of the turn. For right turns, the cone was moved to the other side of the force plate in order for the participants to strike the force plate with their left foot and perform a 90° turn to the right. This procedure was repeated for right and left turns until five valid trials were obtained on each side. Participants had to step cleanly on the force plate with the appropriate foot, i.e., the outside foot, and make a 90° turn in the correct direction to count as a valid trial. Participants were allowed to look at the plate while walking to facilitate correct foot placement within the force plate. The rationale for permitting visual cues for turning trials was that people in this age group with PD likely routinely look at the ground while turning to ensure safety.

2.4. Data Processing

Three specific points in the representative gait cycle were chosen for analysis to assess dynamic stability: first double–support (FDS) at 6% of stance, midstance (MS) at 50% of stance, and second double–support (SDS) at 94% of stance, as defined by the gait events determined in Visual3D. These points were chosen because we hypothesized that they are the least stable points during gait while completing a turn. In FDS, one foot is in initial contact (IC) and the other foot is in preswing (PSw), with minimal contact on the ground with bilateral lower extremities as the weight shifts from the swing leg to the stance leg. In MS, there is only one foot in contact with the ground, making this phase more susceptible to M/L perturbations [24] and increasing the likelihood of falls. Second double–support is also a time of transition, with minimal foot contact on the ground as the weight shifts from the foot in stance to swing. Three right turns and three left turns were chosen for a total of six representative gait cycles. All study variables were assessed at FDS, MS, and SDS and included in the data analysis.

2.4.1. Data Reduction

Vicon Nexus 2.6.1 was used to reconstruct marker trajectories and fill gaps. Raw trajectory data were filtered using a Woltring filter (MSE 15) [27]. Ground reaction force data were smoothed with a 6 Hz zero-lag 4th order Butterworth filter [20,28]. The final data analysis only included trials with clean, i.e., entire foot within the confines of the plate, force plate strikes (five trials on each side). Nexus was used to trim full walking and turning trials, and three representative right and three representative left gait cycles were selected and exported to Visual3D (C–Motion, Inc., Germantown, MD, USA) for further dynamic data reduction and analysis.

The biomechanical model created in Visual3D included a virtual coordinate system attached to each foot (Figure 4). It was defined by markers placed on the medial and lateral malleoli and the toe (between the heads of the first and second metatarsal heads) that were projected onto the floor of the lab. In our model, the y-axis was anteroposterior and the x-axis was mediolateral. The virtual foot coordinate system was projected onto the floor since the COP is on the surface of the floor, and it estimates the position of the COP relative to the foot.

The primary variables of interest, COM–COP inclination angle (Figure 5), COP–COM A/P and M/L (Figure 5), COP–XCOM A/P and M/L (Figure 6), and U_MAX–XCOM A/P and M/L (Figure 6), were calculated to determine whether a difference in the reference limb existed between participants with PD and healthy CON participants at FDS, MS, and SDS of the gait cycle. Because the COM–COP inclination angle is a spatial angle, it was not determined in the foot coordinate system but relative to the laboratory coordinate system (Figure 3). Using a 15–link biomechanical model, total body COM was defined as the center of mass in three-dimensional space and was computed as the weighted average of each body segment’s COM in Visual3D. Segments were modeled as geometric shapes, e.g., cones, cylinders, spheres, and ellipsoids. Segment lengths between each joint were estimated per Dempster [29,30], and segment masses were expressed as a percentage of the total body mass [31]. The velocity of the COM was estimated using a first–order central difference formula.

Based on the rigid body and inverted pendulum model assumptions, the extrapolated center of mass ($XCOM$) was computed as follows:

$$XCOM=COM+\left(V_{COM}/{\omega }_0\right)$$

(1)

Where $COM$ and $V_{COM}$ are the instantaneous position and velocity of the whole-body $COM$, respectively, and ${\omega }_0$ is the angular eigenfrequency of the pendulum, which is calculated as follows:

$${\omega }_0=\sqrt{g/l}$$

(2)

where $g$ = 9.81 m/s² and $l$ is the leg length (defined as the distance from the $COM$ to the ankle joint center in meters) [21].

The boundaries of the BOS were defined by the foot markers, with the toe marker as the anterior limit (A/P U_MAX) and the fifth metatarsal marker as the lateral limit (M/L U_MAX). The $MOS$ in the A/P and M/L directions (in the virtual foot coordinate system) were determined relative to these boundaries. Dynamic stability during turning was calculated using the $MOS$ and $XCOM$, with the following equation:

$$MOS=U_{MAX}-XCOM$$

(3)

where $MOS$ is the shortest distance between the boundaries of the BOS and the $XCOM$, expressed in millimeters. The distances between both the COP and $XCOM$ and the COP and COM were also determined in the A/P and M/L directions.

The COM–COP inclination angle was defined by the 3D angle between the vertical projection on the ground of the COM and a line connecting the COM to the most lateral position of the COP.

Our secondary variables of interest included spatiotemporal parameters (ST): velocity, stride width, stride length, step length, cadence, stance time, and swing time. Following Visual3D data reduction, data were exported to Excel and SAS_JMP (JMP, Cary, NC, USA) for descriptive and statistical analyses.

2.4.2. Data Analysis

Primary and secondary variables of interest were analyzed using a mixed–model ANOVA. The fixed effects of the condition (PD and CON) along with the random effects of the subject and limb within the subject were included in the model. Left and right limb data were analyzed, as asymmetry in turning strategies in IPD has been previously reported [16]. Normality and homogeneity of variance were verified with residual analysis. Comparisons for the primary dependent variables described above were analyzed between foot strike and foot off in the gait cycle. The significance level was set at p < 0.05. A Bonferroni adjustment of the significance level was used for ST parameter variables (p < 0.0056) and dynamic balance variables (p < 0.0028) to decrease the risk of a type 1 error with the use of multiple statistical tests. Since this was an exploratory study, the impact of a type I error would be low. In fact, type II errors, i.e., missing significance, are probably more of a problem. Thus, values of p less than the Bonferroni cutoffs were declared significant. Values of p < 0.05 but greater than the Bonferroni cutoff are reported as marginally significant. One participant presented as an outlier in ST parameter analysis but was not an outlier in dynamic balance variables analysis and therefore was excluded from ST parameter analysis only. The effect size was calculated as the difference between the means of the two groups defined by the factor levels after correcting for the other factors in the model and is in the same units as the dependent variable. Since there is limited research regarding COP–COM A/P and M/L, COP–XCOM A/P and M/L, and U_MAX–XCOM A/P and M/L, and clinically meaningful effect sizes between these populations, mild, moderate, and large effect sizes were defined as mild effect sizes = 10–20 mm, moderate effect sizes = 20–30 mm, and large effect sizes ≥ 30 mm.

3. Results

3.1. ST Parameters

ST gait parameters for CON and PD groups were analyzed using right and left limbs combined since there were no differences between limbs. Gait velocity was the only parameter that showed a marginally significant difference (p = 0.011) between groups, with the PD group’s average velocity slower than the CON group; however, there were no significant differences (p < 0.0056) (

Table 3. Mean (standard deviation) spatiotemporal gait parameters for CON and PD groups while turning.

	CON	PD	p-Value	Effect Size *
Velocity (m/s)	1.02 (0.12)	0.87 (0.15)	0.011	0.15
Double limb support (s)	0.37 (0.06)	0.40 (0.06)	0.136	−0.04
Cycle time (s)	1.26 (0.12)	1.35 (0.15)	0.137	−0.09
Stance Time (s)	0.82 (0.08)	0.90 (0.12)	0.067	−0.03
Swing time (s)	0.44 (0.05)	0.45 (0.05)	0.805	0.00
Step Length (m)	0.66 (0.09)	0.58 (0.14)	0.138	0.07
Stride Length (m)	1.28 (0.16)	1.17 (0.20)	0.158	0.11
Cadence (steps/min)	96.6 (10.4)	90.2 (8.26)	0.075	6.23
Stride Width (m)	0.04 (0.03)	0.04 (0.03)	0.914	0.00

* Negative effect sizes indicate that the CON group values were less than the PD group values.

).

3.2. COM–COP Inclination Angle

The COM–COP inclination angle was not different between limbs, so right and left limb data were combined for both the CON and PD groups. At FDS, the PD group had a COM-COP mean inclination angle of 15.66° (±standard deviation (SD) = 1.57°), whereas the CON group had a mean of 16.39° (±1.73°). At MS, the PD group had a mean of 9.59° (±1.35°), whereas the CON group had a mean of 11.14° (±1.44°). At SDS, the PD group had a mean of 17.93° (±2.45°), whereas the CON group had a mean of 19.43° (±3.58°). There were no statistical differences between groups for the COM–COP inclination angle at FDS (p = 0.171) and SDS (p = 0.255). At MS, there was a marginally significant difference (p = 0.006) between groups; however, there were no significant differences (p < 0.0028). Both groups demonstrated a similarly shaped COM–COP inclination angle curve throughout the gait cycle (Figure 7). At FDS and SDS, the COM–COP inclination angle increased as both feet were in contact with the ground, whereas the angle decreased at MS.

3.3. COP–COM A/P and M/L

The COP–COM A/P and M/L variables can be used as a type of dynamic measure of balance that does not account for the velocity of the COM. These variables were not different between the limbs, so right and left limb data were combined for both CON and PD groups. There were no significant differences between groups for COP–COM M/L at FDS, MS, and SDS. There were also no significant differences between groups for COP–COM A/P at FDS and SDS (

Table 4. Mean (standard deviation), expressed in millimeters for A/P dynamic balance variables COP–COM A/P, COP–XCOM A/P, U_MAX–XCOM A/P for CON and PD group during turning.

	CON	PD	p–Value	Effect Size *
FDS
COP–COM	247.0 (59.6)	231.0 (61.5)	0.497	16.1
COP–XCOM	102.8 (26.9)	91.75 (17.96)	0.210	11.0
UMAX–XCOM	58.07 (35.5)	85.37 (35.2)	0.046	−27.3
MS
COP–COM	201.5 (32.36)	170.2 (30.49)	0.020	31.3
COP–XCOM	238.1 (84.6)	212.8 (47.7)	0.379	25.3
UMAX–XCOM	−137.6 (62.5)	−70.54 (78.6)	0.024	−67.1
SDS
COP–COM	308.3 (101.1)	299.5 (52.8)	0.800	8.89
COP–XCOM	504.2 (190.2)	487.7 (103.7)	0.800	16.5
UMAX–XCOM	−272.4 (142.7)	−199.9 (132.1)	0.197	−72.5

* Negative effect sizes indicate that the CON group values were less than the PD group values. Note FDS = first double–support; MS = midstance; SDS = second double–support.

and

Table 5. Mean (standard deviation), expressed in millimeters for M/L dynamic balance variables COP–COM M/L, COP–XCOM M/L, U_MAX–XCOM M/L for CON and PD group during turning.

	CON	PD	p–Value	Effect Size *
FDS
COP–COM	165.6 (57.3)	159.9 (51.3)	0.773	5.62
COP–XCOM	−97.5 (38.4)	−84.02 (42.9)	0.375	−13.5
UMAX–XCOM	173.5 (28.6)	155.5 (20.0)	0.076	18.0
MS
COP–COM	1.24 (63.4)	12.9 (52.2)	0.598	−11.7
COP–XCOM	−201.6 (73.0)	−146.7 (88.2)	0.100	−54.9
UMAX–XCOM	302.3 (90.5)	271.1 (48.5)	0.304	31.2
SDS
COP–COM	−100.4 (85.9)	−68.76 (73.0)	0.313	−31.6
COP–XCOM	−267.5 (138.6)	−202.3 (127.0)	0.222	−65.2
UMAX–XCOM	586.3 (193.0)	565.5 (104.3)	0.752	20.8

* Negative effect sizes indicate that the CON group values were less than the PD group values. Note FDS = first double–support; MS = midstance; SDS = second double–support.

). At MS, there was a marginally significant differences (p = 0.020) between groups; however, there were no significant differences (p < 0.0028).

Both groups demonstrated similarly shaped COP–COM A/P and M/L curves throughout the gait cycle (Figure 8). From FDS to SDS in the A/P direction, the COP moved from anterior to posterior in relation to the COM. At FDS and SDS, the distance between the COP and COM A/P increased as one limb was in initial contact and the opposite limb was in terminal stance. At MS, the distance between the COP and COM A/P decreased as the swing limb neared the reference (stance) limb prior to swinging past it.

From FDS to SDS, the COP remained lateral to the COM. At FDS and SDS, the distance between the COP and COM M/L was increased as both feet were in contact with the ground. This distance decreased at MS, where one limb was in contact with the ground, causing a medial shift of the COM.

3.4. COP–XCOM A/P and M/L

The COP–XCOM A/P and M/L variables were not different between limbs, so right and left limb data were combined for both CON and PD groups. There were no significant differences between groups for COP–XCOM A/P and M/L (Table 4 and Table 5).

Both groups demonstrated similarly shaped COP–XCOM A/P and M/L curves throughout the gait cycle (Figure 9). From FDS to SDS, the COP remained posterior and lateral to the XCOM. At FDS, the distance between the COP and the XCOM was the least in both A/P and M/L directions. The A/P and M/L distance increased from FDS to SDS as the XCOM moved in an anteromedial direction to maintain stability, causing the COP to move further posterolaterally in relation to the XCOM.

3.5. U_MAX–XCOM A/P and M/L

The U_MAX–XCOM A/P and M/L measures were not different between limbs, so right and left limb data were combined for both CON and PD groups. There were no significant differences between groups for U_MAX–XCOM M/L at FDS, MS, and SDS and U_MAX–XCOM A/P at SDS (Table 4 and Table 5). At FDS, there was a marginally significant differences (p = 0.046) between groups as well as at MS (p = 0.020); however, there were no significant differences (p < 0.0028). In the A/P direction, for U_MAX–XCOM, the effect size for FDS was moderate, and in MS and SDS, it was large, but this may be due to random variation when considering the larger p–values (Table 4).

Both groups demonstrated similarly shaped U_MAX–XCOM A/P and M/L curves throughout the gait cycle (Figure 10). At FDS, the U_MAX was anterior to the XCOM. From FDS to SDS, the XCOM moved anteriorly, causing the U_MAX to move posteriorly in relation to the XCOM. Also, from FDS to SDS, the distance between the U_MAX and the XCOM decreased following FDS and then increased as the groups progressed to MS and SDS. At SDS, this distance again decreased. The U_MAX remained lateral to the XCOM, and the distance between the U_MAX and the XCOM increased from FDS to SDS.

4. Discussion

Individuals with PD have potential postural instability when making 90° turns because they spend a greater percentage of time with the COM and XCOM outside of their BOS [14]. Currently, there is limited research using 3D motion analysis to examine the MOS in relation to dynamic stability in IPD. There are also no studies at this time that have analyzed the MOS using the marker placement on the foot that this study utilized. Therefore, MOS variables were examined in this sample of the PD population during gait and 90° turns to identify possible differences compared to healthy, age-matched controls.

All ST parameter variables were not significantly different between the CON and PD groups. Also, all dynamic balance variables were not statistically different between groups, although MS COM–COP inclination angle, MS COP–COM A/P, FDS U_MAX–XCOM A/P, and MS U_MAX–XCOM A/P demonstrated marginally significant differences. Three of these dynamic balance variables demonstrated marginally significant differences in the midstance phase of gait. This is important because MS is the single-limb support phase of the gait cycle, which is the most unstable phase of gait [32]. All of the dynamic balance variables that demonstrated marginally significant differences besides the COM–COP inclination angle were in the A/P direction. There is no prior research on PD gait during turns that examines these dynamic balance variables in the A/P direction, suggesting that these findings are novel and may be clinically relevant. Prior research supports the idea that a flexed posture, typically seen in IPD, is associated with a posterior COM during stance, which could lead to instability [9]. Increased cadence, decreased stride and step lengths, and decreased velocity in IPD may also affect the dynamic variables in the A/P direction [3,9,10,11], although the ST findings in this PD sample did not support group differences in these ST variables during gait with turns. These novel findings may help clinicians gain a better understanding of dynamic balance in the A/P direction in this population in order to decrease fall risk.

In examining the ST parameters, velocity was the only variable that demonstrated a marginally significant difference (p = 0.011) between groups. The mean velocity for the CON group was 1.02 m/s, whereas, for the PD group, it was 0.87 m/s. The lack of a significant difference between groups in gait velocity may be related to three things: (1) both groups were instructed to target the plate to strike as part of their turn; (2) the relatively mild manifestation of the disease in the PD group; and (3) relatively greater variability (i.e., SD) in gait velocity in both groups with a smaller sample size. It was expected that turn velocity may be slower in IPD as this turn task may reflect underlying motor deficits such as gait hypokinesia, bradykinesia, difficulty adapting gait such as executing direction changes, and impaired execution and sequencing of motor plans [3,9,10,11]. Slower velocity during turning was also expected, as slower peak turn speeds were reported in prior studies in individuals with mild and severe PD during 180° turning in similar gait trials and within the Timed Up and Go (TUG) test [13,16]. Perhaps the relatively mild disease stage in our sample may explain no significant differences in ST parameters between groups.

As mentioned previously, the COM–COP inclination angle may be a useful measure when examining dynamic balance [15]. Analysis of the COM–COP inclination angle at FDS, MS, and SDS revealed that MS was the only point in the gait cycle that demonstrated a marginally significant difference between groups. Of all the dynamic balance measures examined in this study, the COM–COP inclination angle at MS was the only variable with a p-value (p = 0.006) approaching the significance level (p < 0.0028). This finding may indicate that it is more likely that a true difference was observed between the PD and CON groups at MS as opposed to this difference being more likely due to random variation. The MS is the most unstable part of the gait cycle because only one limb is in contact with the ground as the other limb progresses in the swing phase. This dynamic balance demand in MS during 90° turning, coupled with a narrowed BOS typically seen in IPD, may contribute to postural instability. The CON group had a mid-stance COM–COP inclination angle that was 1.556° greater than that of the PD group. The closer the COM is to the COP, the smaller the COM-COP inclination angle and the reduced chance of a fall. Similar to walking on a balance beam, the closer the COM is to the COP during MS, the more likely it is that the individual will maintain balance. A reduced COM–COP inclination angle in the PD group may reflect a conservative strategy to reduce their fall risk. Additionally, it may suggest that IPD had increased difficulty controlling the COM position in relation to COP as compared to the CON group. As the variable of most interest, further research is warranted to investigate the COM–COP inclination angle at MS to characterize dynamic balance during walking and 90° turning in individuals with more advanced stage PD.

Our finding regarding differences in MS COM–COP inclination angle between IPD and healthy controls is consistent with previous research. Yang et al. [15] reported comparable group differences in peak COM–COP inclination angle when analyzing IPD and controls during 180° turning while completing the TUG. These researchers reported a difference in sagittal COM–COP inclination angles (0.96°), with IPD demonstrating a smaller inclination angle than controls. Yang et al. attributed this finding in IPD during turning to either a conservative shortening of steps as a compensatory strategy or unintentional small steps (gait hypokinesia) related to PD pathology triggered during turning [15]. There are notable differences in methodology between Yang’s study and this present study, as Yang’s PD sample included individuals with more advanced disease stages and FOG deficits. Additionally, Yang’s testing was performed during the off-medication times in IPD, and the task was a TUG test involving a 180° turn. In contrast, the present study examined individuals with mild-moderate PD without FOG symptoms who were tested during on-medication times during gait and 90° turns. Despite the milder disease stage of the individuals in this present study, the difference observed in the MS COM–COP inclination angle was greater than the findings in Yang’s study. Our data suggest that a difference of 1.556° observed may be clinically meaningful. Future research is warranted to examine 90° turning in individuals with more severe PD, FOG, and positive fall history to identify possible dynamic balance control problems.

Altered movement of COM in coordination with COP movement may demonstrate dynamic stability changes, and therefore, this relationship may allow identification and quantification of dynamic balance impairments in IPD of varying severity [33]. When examining COP–COM A/P and M/L at FDS, MS, and SD, a marginally significant difference between groups was observed only in the COP–COM A/P variable at MS (Table 4). The PD group demonstrated a distance between COP and COM in the A/P direction that was 31.3 mm less than the CON group. This finding may reflect strategies utilized by IPD to increase stability, similar to the strategy discussed regarding the COM–COP inclination angle, in which IPD bring their COP and COM as close as possible to reduce fall risk. Hass et al. [33] examined peak COP–COM distance at three points of gait initiation to measure dynamic balance control in IPD with varying severity with and without balance impairments. They observed a reduced peak distance (by 16%) between COP and COM at the phase of gait initiation where the COP moves anteriorly under the stance foot [33]. Hass et al. [33] suggested that a smaller COP–COM distance may be a conservative strategy adopted by IPD to reduce the mechanical and postural requirements and the magnitude of muscular strength required at this phase.

The only point in the gait cycle that demonstrated a marginally significant difference between groups regarding U_MAX–XCOM A/P and M/L was U_MAX–XCOM A/P at FDS and MS (Table 4). At MS, the CON group demonstrated a distance between U_MAX and XCOM in the A/P direction that was 67.07 mm less than the PD group. This finding may indicate that the PD demonstrated a wider BOS in the A/P direction during the turn, which is not typically seen in this population as it is more typical to see a narrowed BOS. One explanation for this finding may be that the visual cues (cones and force plates) provided during the turning task may have altered typical gait and turning patterns in the PD group. Additionally, an increased anterior XCOM owing to more forward-flexed postural alignment in IPD may have decreased the time spent in the less stable phases of gait during turns, FDS, and MS. Further research is needed to characterize this variable and the clinical relevance of U_MAX–XCOM A/P at MS and FDS as a parameter of dynamic balance control during gait and 90° turns in IPD.

With regard to U_MAX–XCOM M/L, one previous study observed that individuals with mild/moderate PD during 30–180° turning had a narrower BOS than healthy controls, which led to a reduced mean distance between the XCOM and the BOS, indicating some dynamic instability [14]. A reduced mean distance between U_MAX and XCOM in the M/L direction was expected in the present study based on this previous research finding. Based on our findings, the mean distances between U_MAX and XCOM in the M/L direction were larger in the CON group than in the PD group, although these differences were not statistically meaningful. Buurke et al. [17] suggested that a greater distance between XCOM and BOS may be an adaptation for unsteady walking and is associated with increased M/L stability during walking.

Although statistical analysis of primary and secondary variables showed minimally significant group differences across the data in this study, the application of 3D motion analysis of walking with turns may be clinically relevant. Completing turns during gait may be a more challenging task for IPD than straight–line walking as there are demands for coupling between gait and postural control as the individual modifies the locomotor pattern in this postural transition [13]. Three-dimensional motion analysis can be used to examine changes in dynamic balance over time in IPD to track disease progression and associated dynamic stability changes, which might provide insight into fall risk. This type of dynamic movement analysis during turns and other gait tasks is salient to daily ambulatory functional activities and balance demands in IPD. Since IPD are at high risk of falls, it is important to understand the changes in dynamic balance and MOS measures that are related to falls. On the other hand, the clinical application and utility of 3D motion analysis for fall prevention in the examination and treatment of IPD are limited due to the need for specialized equipment and time for complex data processing. While this research regarding the utility of the MOS in IPD is mechanistic and exploratory in nature, it can offer insight into dynamic balance variables to investigate potential differences between healthy elderly and IPD that may be associated with an increased risk of falls. The lack of statistically significant differences between the CON and PD groups in this study could be due to the small sample size and the higher level of gait and balance function in our sample of individuals with PD. Future research should examine a larger group of subjects with identified postural instability and fall history during later stages of PD in order to assess dynamic gait variables in a group that may have more variability.

There are several limitations to our study. Due to the small sample size in both groups, our analysis may be underpowered. Thus, we expressed some conservatism by highlighting selected variables that had a marginally significant difference between groups to reflect the uncertainty presented by this limitation. This study incorporated a sample of convenience, and participants were reportedly active community ambulators and did not have a fall history; therefore, the generalizability of our findings to less active individuals or IPD with more impaired gait and balance is limited. The traditional classification of disease severity in PD, i.e., Hoehn and Yahr staging, the Unified Parkinson’s Disease Rating Scale was not used in this study to categorize the participants due to the researchers’ limitations in training and formal certification in the administration and scoring of these standardized measures. However, given that the PD participants’ balance and gait function wasassessed using standard methods accepted by the clinical community, their description of having mild to moderate PD is accurate and reasonable. Participants were allowed to visually target the force plates when walking to achieve a clean force plate strike, which may or may not have been representative of their normal gait cycle or of walking and turning without visual cues.

Terry et al. [34] and Kazanski et al. [35] have suggested that Hof et al.’s. [21] original work on the margin of stability for static postures may need to be modified for use in gait, so additional work is needed in that area. In addition, future research should investigate turning with other task/environmental conditions, such as walking on an unstable surface, walking at a fast pace, or walking with cognitive dual task demands, since these factors represent salient “real world” demands and contexts that increase the risk of falls in IPD [36,37,38]. Falls in the home during gait navigation may occur on carpet or unlevel surfaces in the community and may occur when there is urgency in walking tasks, such as hurrying to the bathroom. Assistive devices were not used in this study, but the use of assistive devices, such as a cane or walker, while turning and their effect on MOS is another important area for investigation in IPD.

5. Conclusions

The purpose of this study was to examine the dynamic balance control in individuals with mild to moderate PD and healthy controls while walking while performing 90° turns. Three–dimensional motion analysis of dynamic balance variables using the margin of stability metrics and spatiotemporal gait parameters were analyzed to determine if there were significant differences between the groups. We found no statistically significant differences in the primary or secondary variables between the PD and Control groups; however, dynamic balance variables in the midstance phase of the gait cycle during a turn reflected trends in group differences. Interpretation of these preliminary findings may reflect a conservative postural control strategy in individuals with PD to reduce instability during turns, but further research is warranted. We believe that the MOS metrics may be useful to measure dynamic balance in individuals with more advanced PD who are likely to have gait and postural instability. It appears that the COM–COP inclination angle may be a useful measure of dynamic balance, as shown by our data, which corroborated earlier research. In addition, we believe that of the three other metrics, the COP–XCOM A/P and M/L may provide unique insight into possible postural control mechanisms underlying dynamic instability since they account for the response of the center of pressure.