GRAPHICAL ABSTRACT

INTRODUCTION

In Parkinson’s disease (PD), gait impairments are characterized by increased gait asymmetry, shorter steps, and slowed gait [1]. Gait speed is a critical variable that influences the range of activity, fall risk, quality of life (QOL), and social participation of people [2-4]. Although gait speed is affected primarily by cadence and step length, cadence is a principal contributor to gait speed in people with PD (PwPD) [5]. Previous studies have investigated participants who walk at their preferred or fast gait speed [6,7]. PwPD involves disease-specific narrowing of the range of gait speeds, from intentionally slow to fast, which is associated with step time and length [8]. Accurate evaluation of the adjustability of gait speed in diverse environments and contexts is clinically important for understanding the gait adaptability of PwPD.

Quantitative gait assessments have been conducted in specialized laboratories or clinical settings under controlled conditions; however, gait assessments in the clinic are likely to increase awareness of being observed, which reflects only the gait capacity of a person [9]. Conversely, gait in free-living environments is affected by divided attention, cluttered environments, diverse sensory experiences, and fatigue, which reflect the functional performance of a person [10,11]. The distribution of gait speed in free-living environments exhibits a bimodal Gaussian pattern, indicating that participants have two preferred gait speeds [12]. It is plausible that a lower speed aligns gait with high cognitive loads or in confined spaces, whereas a higher speed corresponds to reaching a certain target or gait in more open areas. By quantifying the gait speed distribution in this manner, a broader spectrum of information is obtained rather than being confined to a single mean and standard deviation. Functional impairments, as indicated by the bimodal distribution of gait speed, may limit the ability to adjust speed, which is evident in individuals with a history of falls or some with PD [13,14]. However, the relational factors affecting gait speed distribution in PwPD have not yet been identified. Characterizing gait speed distributions in relation to the environment and gait control will provide deeper insights into gait impairment in PD.

This study aimed to compare changes in gait parameters between clinic and free-living environments at the preferred lower and higher speeds of PwPD, categorized by gait speed distribution. The relationships among gait parameters, environmental factors, and PD-specific assessments were also investigated. We hypothesized that gait parameters change similarly according to gait speed in both clinical and free-living environments but deteriorate more in free-living environments than in the clinical setting. The gait distribution characteristics of PwPD are presumably related to environmental factors, PD severity, and spatiotemporal gait parameters, especially in free-living environments.

MATERIALS & METHODS

Participants and assessment

In the study by Atrsaei et al. [14], to determine the necessary sample size for repeated-measures analysis of variance (RMANOVA), 80% power with a type–I error of 2.5% was applied [14]. Prior power analysis and 15% dropout calculations indicated that a sample size of more than 40 people with PD diagnosed according to the United Kingdom Parkinson’s Disease Society Brain Bank was needed. The inclusion criteria were the absence of concurrent neurological, orthopedic, or internal diseases affecting motor function and no history of deep brain stimulation. The exclusion criteria were a diagnosis of atypical parkinsonism and stage ≥2 disease on the Functional Assessment Staging Test (FAST) of Alzheimer’s disease (AD). The FAST is a comprehensive tool used to assess the progression of AD by evaluating a patient’s ability to perform activities of daily living (ADL) through both direct observations of the patient and information provided by family members and caregivers. A stage ≥2 on the 7-stage FAST indicates the presence of AD.

This study protocol was conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent. The present study was approved by the Ethical Review Committee of our institution (approval number 054).

Study design and procedure

The experimental procedure involved clinical assessments and gait measurements both in a laboratory setting and in a free-living environment.

Data on demographics such as age and sex, body mass index, disease duration, number of falls within the last 3 months, and Walk Score were collected. The Walk Score is based on residential postal codes and analyzes factors such as distance to various categories of facilities, population density, and road connectivity (e.g., block length and intersection density), providing a measure of neighborhood walkability [15]. The Walk Score ranges from 0 to 100, with lower scores indicating greater reliance on cars for errands and higher scores suggesting that daily errands can be accomplished on foot. In this study, we adopted the Walk Score to elucidate the relationships between free-living gait parameters and outdoor environmental factors. Clinical assessments and gait measurements in the clinical setting were performed during the “ON” medication state. The Mini-Mental State Examination (MMSE), Hoehn–Yahr stage, Movement Disorder Society-Unified Parkinson’s Disease Rating Scale III (MDS-UPDRS-III), New Freezing of Gait Questionnaire (NFOGQ), and Parkinson’s Disease Questionnaire-39 (PDQ-39) scores were evaluated. The MDS-UPDRS III score was classified into four subgroups: axial scores (items 1, 10–13), bradykinesia (items 2, 4–9, 14), rigidity (item 3), and tremor (items 15–18) [16]. Moreover, we divided the PDQ-39 score into eight subdomains: mobility (items 1–10), ADL (items 11–16), emotional well-being (items 17–22), stigma (items 23–26), social support (items 27– 29), cognition (items 30–33), communication (items 34–36), and bodily discomfort (items 37–39) [17].

For gait measurements in the clinic, participants were outfitted with a single wearable triaxial accelerometer (AX3; Axivity, York, UK) placed on the fifth lumbar vertebra (L5). Extensive guidance was provided to the participants to ensure proper use of the device. The participants practiced until they could securely attach the device themselves. The accelerometer was configured to record data at a rate of 100 Hz with a sensitivity range of ±8 g [18]. In the clinical setting, participants were instructed to walk for 90 min, including rest periods, in an unrestricted indoor environment accompanied by a therapist. Following the clinical measurements, gait data were collected in a free-living environment. The participants were required to wear a triaxial accelerometer for 7 consecutive days and perform their typical daily routines in their regular living environment. The participants were instructed to wear the sensor at all times except when they were bathing, sleeping, or experiencing discomfort from wearing the device.

Data processing and analysis of spatiotemporal gait parameters

Acceleration-derived variables were analyzed using custom scripts in MATLAB R2021b (MathWorks Inc., Natick, MA, USA). To ensure analysis of the steady-state gait, walking bouts lasting <15 s were excluded. Previous studies have extensively detailed the algorithms and data segmentation techniques utilized for accelerometer data [18,19]. Wearing time was calculated using established methods from a previous study, ensuring that all participants consistently wore the device for a minimum of 8 hours each day throughout the entire measurement period [20]. Walking activities were detected on the basis of the sensor’s upright orientation and the predefined thresholds (mean >0.05 and standard deviation >0.77) of vertical acceleration [21]. Next, using continuous wavelet transformation, the vertical acceleration was analyzed to determine the initial and final contacts during the gait cycle for the spatiotemporal parameters [19]. The spatiotemporal gait parameters, including step velocity, step length, cadence, the walk ratio (step length divided by cadence), the double support ratio (the proportion of double support during the gait cycle), and the asymmetries of step length and step time, were determined for each gait cycle in both clinical and free-living environments. The step length was calculated as follows:

(1)

Step length = 2 (2lh-h2)

where l is the height of L5 (in meters), as measured using an accelerometer, and h is the change in the vertical position of the center of mass (in meters).

Cadence was calculated by dividing step time by 60 s, and the step velocity was calculated by dividing the step length by the step time. The walk ratio is not significantly different between individuals who walk at different preferred speeds, making it a reliable indicator of gait automaticity and central gait coordination [22]. Gait cycles with a speed <0.2 m/s were eliminated, as they potentially represent a pause or interruption. The distribution of daily life gait speed is known to follow a bimodal distribution [12-14]. We calculated the fit of the bimodal distribution for the distribution of gait speed via Ashman’s D (Equation 2). An Ashman’s D greater than 2 indicated a bimodal distribution.

(2)

Ashman’s D = |μh-μl|σh2+σl22

Here, μ_h is the average of each gait parameter in the higher gait speed distributions, μ_l is the average of each gait parameter in the lower gait speed distributions, and σh and σl are the standard deviations from each of the means.

The gait speed distribution of each person was divided into lower and higher gait speed distributions on the basis of the probability of attributing the gait speed to either speed. The higher speed ratio was calculated as the proportion of gait cycles at higher gait speeds relative to the total number of gait cycles. For each distribution, the gait parameters for each gait cycle were analyzed. To observe differences in gait parameters between the preferred gait speeds, we defined the ratio of gait parameters to represent the percentage difference between lower and higher gait speeds in both clinical and free-living environments.

(3)

Δμ = 2μh-μlμh+μl×100

Here, μ_h is the mean of gait parameters in higher gait speed distributions, and μ_l is the mean of gait parameters in lower gait speed distributions. We calculated the step length, cadence ratio, walk ratio, double support, step length asymmetry, and step time asymmetry.

Statistical analyses

R version 4.3.1 (R Foundation, Vienna, Austria) was used for the statistical analyses. In both clinical and free-living environments, the PwPD were divided on the basis of an Ashman’s D >2 and whether the gait speed distribution was bimodal, and the headcount ratios were analyzed using the chi-square test. To examine gait parameters, we performed a 2 × 2 (environment [clinic/free-living] × speed [lower/higher]) RM-ANOVA. Partial eta squared (ηp2) values were calculated to determine effect sizes, commonly categorized as small (ηp2 ≥ 0.01), medium (ηp2 ≥ 0.06), and large (ηp2 ≥ 0.14). We evaluated the associations between PD symptoms and the ratio of gait parameters via Bayesian Pearson’s correlation coefficients [23]. This analysis quantifies evidence supporting the Bayesian Pearson’s correlation between two variables via the Bayes factor (BF). The null hypothesis of no correlation between the variables was assessed using the BF in comparison to the alternative hypothesis of a correlation between the variables. The Pearson correlation coefficients, BF₁₀ values, and 95% confidence intervals were calculated. BF₁₀ is a measure that assesses the weighted average likelihood ratio of the data supporting the two opposing hypotheses, namely, the null hypothesis and alternative hypothesis. It gauges the strength of the evidence in favor of the null hypothesis, with interpretations as follows: anecdotal evidence (BF₁₀ ≥ 1), substantial evidence (BF₁₀ ≥ 3), strong evidence (BF₁₀ ≥ 10), very strong evidence (BF₁₀ ≥ 30), and decisive evidence (BF₁₀ ≥ 100) [23].

RESULTS

Demographic characteristics of PwPD

In total, 41 PwPD participated in this study, and Table 1 presents the demographic characteristics and symptoms of the PwPD. The average disease duration was 6 years, and Hoehn–Yahr stage III was the most common stage among the participants in this study.

Spatiotemporal gait parameters

Figure 1 depicts the histogram of gait speed for a typical patient, along with the Gaussian mixture models that were fitted during free-living environments. This figure reveals a bimodal distribution of gait speed in free-living environments, indicating that patients have two preferred gait speeds: a lower (μ_l) speed and a higher (μ_h) speed in free-living environments. Ashman’s D >2 was present in 20 and 30 PwPD in the clinical and free-living environments, respectively; Ashman’s D ≤2 was present in 21 and 11 PwPD, respectively. The chi-square test for headcount ratios yielded a value of 5.13 (p < 0.05).

Table 2 displays the results of the RM-ANOVA for the spatiotemporal gait parameters, which include the average and standard deviation of each gait parameter, F value, p value, and ηp2 value. The RM-ANOVAs of step velocity, step length, and double support all revealed significant main effects for environment and speed. The RM-ANOVA of cadence revealed significant main effects for speed but not for the environment. The RMANOVAs of the walk ratio, step-time asymmetry, and step-length asymmetry all revealed significant main effects for the environment but not speed.

Associations between PD symptoms and gait parameters

The associations between the number of falls, Walk Score, and change in gait parameters and the MDS-UPDRS-III, NFOGQ, and PDQ-39 scores are presented in Figure 2.

In the clinical setting, Ashman’s D had a positive correlation with step length (r = 0.49, BF₁₀ = 35.79 [very strong evidence]) and cadence (r = 0.53, BF₁₀ = 86.21 [very strong evidence]) and a negative correlation with step length asymmetry (r = -0.47, BF₁₀ = 25.27 [strong evidence]). The MDS-UPDRS-III total score was negatively correlated with step length (r = -0.49, BF₁₀ = 33.74 [very strong evidence]) and the walk ratio (r = -0.51, BF₁₀ = 58.82 [very strong evidence]). Notably, a negative correlation was observed between the bradykinesia score of the MDS-UPDRS-III subgroups and step length (r = -0.53, BF₁₀ = 94.44 [very strong evidence]). Cadence was negatively correlated with the walk ratio (r = -0.63, BF₁₀ > 1,000 [decisive evidence]) and step length asymmetry (r = -0.55, BF₁₀ = 168.16 [decisive evidence]).

In free-living environments, Ashman’s D was negatively correlated with the number of falls (r = -0.47, BF110 = 22.97 [strong evidence]) and the PDQ-39 total score (r = -0.37, BF₁₀ = 4.25 [substantial evidence]) and positively correlated with step length (r = 0.47, BF₁₀ = 23.37 [strong evidence]) and cadence (r = 0.37, BF₁₀ = 4.39 [substantial evidence]).

Positive correlations were observed between a higher velocity ratio and the Walk Score (r = 0.52, BF₁₀ = 71.56 [very strong evidence]) and between the step length and the walk ratio (r = 0.79, BF₁₀ > 1,000 [decisive evidence]). Negative correlations were observed between the MDS-UPDRS-III total score and step length (r = -0.54, BF₁₀ = 115.70 [decisive evidence]), between the MDS-UPDRS-III total score and the walk ratio (r = -0.47, BF₁₀ = 25.62 [strong evidence]), and between cadence and the walk ratio (r = -0.43, BF₁₀ = 10.34 [strong evidence]). Notably, a negative correlation was observed between the bradykinesia score of the MDS-UPDRS-III subgroups and step length (r = -0.56, BF₁₀ = 189.15 [decisive evidence]). In the PDQ-39 subdomain, the mobility score had a negative correlation with Ashman’s D (r = -0.44, BF₁₀ = 14.37 [strong evidence]). The ADL score was negatively correlated with step length (r = -0.39, BF₁₀ = 5.92 [substantial evidence]). The communication score had a negative correlation with Ashman’s D score (r = -0.49, BF₁₀ = 35.55 [very strong evidence]).

DISCUSSION

In this study, we compared spatiotemporal gait parameters between clinical and free-living environments, which were assessed for 7 consecutive days, at preferred lower and higher speeds divided by gait speed distribution in PwPD. The relationships among changes in gait parameters, environmental factors, and PD-specific assessments were investigated. We hypothesized that gait parameters change similarly according to gait speed in both clinical and free-living environments but deteriorate more in free-living environments compared with clinical settings. The gait distribution characteristics of PwPD are presumably related to step length, cadence, environmental factors, and PD severity, especially in free-living environments.

Impact of environment and gait speed on spatiotemporal gait characteristics in PwPD

Our results showed that the spatiotemporal gait parameters, excluding the walk ratio, were worse in the free-living environment than in the clinical setting and changed with gait speed. These findings, which indicate a tendency to walk faster, have a longer step, and increased asymmetry in the clinical setting than in the free-living environment, align with the findings of a previous study [24]. Previous studies also indicate that gait measurements obtained from free-living environments differ from those collected during laboratory or clinical tests [24,25]. A key factor contributing to these differences in PwPD could be reliance on gait mechanisms that require more attention and are less automatic [26]. Free-living environments require continuous adaptation of gait control in complex and unstructured environments, which makes daily gait more challenging.

Step velocity and length are affected by the environment and speed, whereas cadence is significantly affected only by speed. According to a previous study, the regulation of step length but not cadence is closely related to disease severity in PD patients [27], which is consistent with the present results. Step length is associated with dopaminergic system pathology and cognitive load, whereas cadence is associated with poor dopamine responsiveness and is less sensitive to cognitive load [28-30]. Hence, step length is considered more sensitive to environmental influences, whereas cadence, which remains intact in PD, is less influenced by the environment. Gait speed is determined by the product of step length and cadence, wherein the preferred combination of these two factors optimizes spatiotemporal control, as well as attentional demands for a given speed [31]. In both clinical and free-living environments, the relationship between bradykinesia and step length is very strongly or decisively evident. Our findings suggest that the difficulty in adjusting step length caused by bradykinesia is compensated for by an adjustment in cadence [32]. The present results also show that both step length and cadence increased with increasing gait speed in PwPD, but the optimization of these combinations requires further study. The walk ratio, which is calculated by dividing step length by cadence, is an indicator of the automaticity of gait and central gait coordination [22]. The present results showed that the walk ratio is invariant at preferred lower and higher gait speeds but is lower in free-living environments. In addition, the walk ratio was negatively associated with the MDS-UPDRS-III score in both clinical and free-living environments. Considering that the normal walk ratio ranges from approximately 0.65−0.65 cm/step/min, the walk ratio in the clinical setting is relatively close to the normal range, whereas the walk ratio in the free-living environment deviates from the normal range in PwPD [22]. A lower walk ratio typically indicates a strategy in which individuals maintain speed by increasing their cadence while reducing their step length, which may suggest cautious gait or impaired balance control.

Our findings also revealed that double support and asymmetries of step length and time were increased in free-living environments. Prolonged double support time may result from difficulties in weight transfer, reflecting neural strategies to mitigate fall risk and enhance postural control [33]. The asymmetrical loss of dopamine in the basal ganglia leads to gait asymmetry in PwPD, which is associated with an uneven onset of bradykinesia and rigidity [34]. In challenging environments, gait asymmetry, particularly step length asymmetry, becomes more pronounced and significantly affects gait and balance function [35]. Gait in the clinic represents what the individual can do, whereas gait in the free-living environment provides insights into actual performance, function, and behavior [10]. Therefore, evaluating gait in both clinical and free-living environments offers a complementary perspective. The findings of this study highlighted how gait control adaptations in response to speed vary across different environments, potentially offering new insights into gait disturbances in PwPD.

Characteristics of gait speed adjustability in PwPD

The headcount of the bimodal distribution in gait speed (Ashman’s D ≥ 2) was lower in the free-living environment than in the clinical setting, indicating that gait speed and its range in the free-living environment declined accordingly. Nonpreferred gait speeds are often associated with diminished dynamic balance [36]. Consequently, the gait speed range of PwPD may be reduced to maintain dynamic stability and minimize the risk of falls. Our results demonstrated that the Walk Score, which reflects the walkability of nearby destinations [15], was positively correlated with a higher gait speed ratio but not with Ashman’s D score. These results indicate that the home-surrounding environment does not affect the adjustability of gait speed but affects the breakdown between slow and fast speeds. Higher gait speeds are thought to reflect longer periods of outdoor walking, and these results support this hypothesis [14]. The poor neighborhood walkability of the home environment may limit the opportunity for outdoor walking or fast walking in PwPD. In addition, the present results showed that gait speed adjustability was positively correlated with step length and cadence in both clinical and free-living environments. These findings are consistent with those of a previous study in which the reduced range of gait speed of PwPD was associated with step length and time in a laboratory environment [8].

Interestingly, the adjustability of gait speed only in free-living environments was negatively related to the number of falls and the MDS-UPDRS-III and PDQ-39 total scores. Our finding that the relationship between the adjustability of gait speed and MDS-UPDRS-III score differs between gait capacity (clinic) and functional performance (free-living environment) cannot be explained by differences in the walking environment alone. PwPD in clinical settings may improve gait performance due to the Hawthorne effect, which occurs when participants potentially perform better when they are aware of being observed [37]. Alternatively, fluctuations between the “ON” and “OFF” states during daily activities might have affected gait measures throughout the week. The adjustability of gait speed to the complex context of the environment and reaching a target may be hampered by reduced automatic and attention-demanding gait mechanisms following the progression of the motor symptoms of PD. Difficulty in adjusting gait control hinders adaptation to the environment and contributes to falls [13]. Moreover, a previous study reported that gait speed was not associated with QOL in laboratory settings in PwPD [38]. In contrast, the adjustability of gait speed in free-living environments is associated with QOL and may be a clinically important factor in improving QOL in PwPD. PwPD patients who exhibit reduced adjustability of gait speed in free-living environments repeatedly experience difficulties in environmental adaptations, possibly resulting in increased subjective gait difficulties and limitations, as included in the mobility domain of the PDQ-39. The communication score of the PDQ-39 comprises speech and communication with other people. Communication impairment in PwPD may contribute to cognitive and motor impairments, including motor speech systems, and is associated with freezing of gait and gait speed [39,40].

Limitations

Our study had several limitations. First, the data were limited to one facility; hence, a larger sample size from various regions is needed to investigate the generalizability of the relationship between walkability at nearby destinations and gait. Second, falls were retrospectively determined and could therefore have been affected by recall bias. Third, the “ON” and “OFF” states of free-living environments were not evaluated, thereby precluding a comparison of the differences in the respective gait parameters. Future studies are warranted to investigate these disparities and enhance the understanding of gait in free-living environments. Fourth, a previous study demonstrated high validity for spatiotemporal gait parameters, except for step length asymmetry [18]. This issue may be addressed by incorporating foot-mounted sensors in addition to the low back-mounted sensor used in this study. Fifth, brain magnetic resonance imaging (MRI) data were not collected in this study. However, white matter hyperintensities from brain MRI, which are associated with PD executive dysfunction, including worse attention, working memory, and processing speed, provide objective information on the cognitive status of patients. Future studies should include brain MRI to examine the relationships between white matter hyperintensity and gait parameters. Sixth, we did not specifically assess the impact of environmental factors, such as obstacles, space constraints, or the layout of indoor environments. These factors can significantly influence gait patterns, particularly in free-living environments, where individuals frequently navigate through varied and often constrained indoor environments. Future studies should consider incorporating detailed assessments of indoor environmental characteristics, potentially using additional sensors, such as GPS devices, or integrating self-reported environmental data.

Conclusions

The findings of the present study suggest that gait control, which involves adjusting gait speed according to context, differs between clinical and free-living environments in PwPD. Although previously reported differences in gait quality between laboratory or clinical settings and free-living settings were based on mean and median comparisons, the present study has clinical implications, focusing on the bimodal distribution of gait speed and identifying differences in gait control at lower and higher gait speeds. Moreover, the adjustability of gait speed in free-living environments is related to the environment and severity of PD, which is related to clinically important factors, such as the number of falls and QOL. Assessing gait using clinical settings and daily life tests should be used to interpret gait impairment in PwPD in a complementary manner.

Notes

Conflicts of Interest

The authors have no financial conflicts of interest.

Funding Statement

This study was supported by the personal research funds of Yuki Nishi.

Author Contributions

Conceptualization: Yuki Nishi. Data curation: Shintaro Fujii, Koki Ikuno, Yuta Terasawa. Formal analysis: Yuki Nishi. Funding acquisition: Yuki Nishi. Investigation: Shintaro Fujii, Koki Ikuno, Yuta Terasawa. Methodology: Yuki Nishi, Shintaro Fujii, Koki Ikuno. Project administration: Shu Morioka. Software: Yuki Nishi. Supervision: Koki Ikuno, Shu Morioka. Visualization: Yuki Nishi. Writing—original draft: Yuki Nishi. Writing—review & editing: Shintaro Fujii, Koki Ikuno, Shu Morioka.

Acknowledgments

Last Author thanks Yuji Minamikawa for collaboration on the early stages of this work.

Figure 1.

An example of a gait speed probability density function for one of the patients. The distribution is shown on the basis of a free-living environment. The red fitted curves are the first and second terms of the bimodal Gaussian distribution introduced. The higher speed ratio was calculated as the number of gait cycles attributed to a higher gait speed distribution to the percentage of the total number of gait cycles. Ashman’s D was calculated using Equation 2. μ_l: 0.49 m/s. μ_h: 1.07 m/s. Higher speed ratio: 43.96%. Ashman’d D: 2.32.

Figure 2.

Heatmap showing correlation coefficients in the spatiotemporal gait parameters in clinical and free-living environments. Darker pixels reflect higher correlation values (red: positive, green: negative). The BF₁₀ and 95% confidence intervals are indicated only in the pixels when the correlation was significant at anecdotal evidence (BF₁₀ ≥ 1). *BF₁₀ ≥3 (substantial evidence); ^†BF₁₀ ≥30 (very strong evidence); and ^‡BF₁₀ ≥100 (decisive evidence). MDSUPDRS-III, Movement Disorder Society-Unified Parkinson’s Disease Rating Scale III; PDQ-39, Parkinson’s Disease Questionnaire-39; MMSE, Mini-Mental State Examination; NFOGQ, New Freezing of Gait Questionnaire; ADL, activities of daily living; supp, support; asy, asymmetry.

REFERENCES

• 1. Morris M, Iansek R, Matyas T, Summers J. Abnormalities in the stride length-cadence relation in parkinsonian gait. Mov Disord 1998;13:61–69.ArticlePubMed
• 2. Canning CG, Paul SS, Nieuwboer A. Prevention of falls in Parkinson’s disease: a review of fall risk factors and the role of physical interventions. Neurodegener Dis Manag 2014;4:203–221.ArticlePubMed
• 3. Sanders JB, Bremmer MA, Comijs HC, van de Ven PM, Deeg DJH, Beekman ATF. Gait speed and processing speed as clinical markers for geriatric health outcomes. Am J Geriatr Psychiatry 2017;25:374–385.ArticlePubMed
• 4. Santos García D, de Deus Fonticoba T, Suárez Castro E, Borrué C, Mata M, Solano Vila B, et al. Non-motor symptoms burden, mood, and gait problems are the most significant factors contributing to a poor quality of life in non-demented Parkinson’s disease patients: results from the COPPADIS study cohort. Parkinsonism Relat Disord 2019;66:151–157.PubMed
• 5. Peterson DS, Mancini M, Fino PC, Horak F, Smulders K. Speeding up gait in Parkinson’s disease. J Parkinsons Dis 2020;10:245–253.ArticlePubMedPMC
• 6. Costa P, De Jesus T, Torriani-Pasin C, Polese J. Functional capacity and walking speed reserve in individuals with chronic stroke: a cross-sectional study. Physiother Theory Pract 2022;38:2563–2567.ArticlePubMed
• 7. Knapstad MK, Steihaug OM, Aaslund MK, Nakling A, Naterstad IF, Fladby T, et al. Reduced walking speed in subjective and mild cognitive impairment: a cross-sectional study. J Geriatr Phys Ther 2019;42:E122–E128.ArticlePubMed
• 8. von der Recke F, Warmerdam E, Hansen C, Romijnders R, Maetzler W. Reduced range of gait speed: a Parkinson’s disease-specific symptom? J Parkinsons Dis 2023;13:197–202.ArticlePubMedPMC
• 9. Friesen KB, Zhang Z, Monaghan PG, Oliver GD, Roper JA. All eyes on you: how researcher presence changes the way you walk. Sci Rep 2020;10:17159.ArticlePubMedPMCPDF
• 10. Hillel I, Gazit E, Nieuwboer A, Avanzino L, Rochester L, Cereatti A, et al. Is every-day walking in older adults more analogous to dual-task walking or to usual walking? Elucidating the gaps between gait performance in the lab and during 24/7 monitoring. Eur Rev Aging Phys Act 2019;16:6.ArticlePubMedPMCPDF
• 11. Warmerdam E, Hausdorff JM, Atrsaei A, Zhou Y, Mirelman A, Aminian K, et al. Long-term unsupervised mobility assessment in movement disorders. Lancet Neurol 2020;19:462–470.ArticlePubMed
• 12. Van Ancum JM, van Schooten KS, Jonkman NH, Huijben B, van Lummel RC, Meskers CGM, et al. Gait speed assessed by a 4-m walk test is not representative of daily-life gait speed in community-dwelling adults. Maturitas 2019;121:28–34.ArticlePubMed
• 13. Brodie MA, Coppens MJ, Ejupi A, Gschwind YJ, Annegarn J, Schoene D, et al. Comparison between clinical gait and daily-life gait assessments of fall risk in older people. Geriatr Gerontol Int 2017;17:2274–2282.ArticlePubMedPDF
• 14. Atrsaei A, Corrà MF, Dadashi F, Vila-Chã N, Maia L, Mariani B, et al. Gait speed in clinical and daily living assessments in Parkinson’s disease patients: performance versus capacity. NPJ Parkinsons Dis 2021;7:24.ArticlePubMedPMCPDF
• 15. Koohsari MJ, Sugiyama T, Hanibuchi T, Shibata A, Ishii K, Liao Y, et al. Validity of Walk Score® as a measure of neighborhood walkability in Japan. Prev Med Rep 2018;9:114–117.ArticlePubMedPMC
• 16. Jenkinson C, Fitzpatrick R, Peto V, Greenhall R, Hyman N. The Parkinson’s disease questionnaire (PDQ-39): development and validation of a Parkinson’s disease summary index score. Age Ageing 1997;26:353–357.ArticlePubMed
• 17. Li X, Xing Y, Martin-Bastida A, Piccini P, Auer DP. Patterns of grey matter loss associated with motor subscores in early Parkinson’s disease. Neuroimage Clin 2017;17:498–504.ArticlePubMedPMC
• 18. Del Din S, Godfrey A, Rochester L. Validation of an accelerometer to quantify a comprehensive battery of gait characteristics in healthy older adults and Parkinson’s disease: toward clinical and at home use. IEEE J Biomed Health Inform 2016;20:838–847.ArticlePubMedPDF
• 19. McCamley J, Donati M, Grimpampi E, Mazzà C. An enhanced estimate of initial contact and final contact instants of time using lower trunk inertial sensor data. Gait Posture 2012;36:316–318.ArticlePubMed
• 20. van Hees VT, Renström F, Wright A, Gradmark A, Catt M, Chen KY, et al. Estimation of daily energy expenditure in pregnant and non-pregnant women using a wrist-worn tri-axial accelerometer. PLoS One 2011;6:e22922.ArticlePubMedPMC
• 21. Hickey A, Del Din S, Rochester L, Godfrey A. Detecting free-living steps and walking bouts: validating an algorithm for macro gait analysis. Physiol Meas 2017;38:N1–N15.ArticlePubMedPDF
• 22. Bogen B, Moe-Nilssen R, Ranhoff AH, Aaslund MK. The walk ratio: investigation of invariance across walking conditions and gender in community-dwelling older people. Gait Posture 2018;61:479–482.ArticlePubMed
• 23. Nuzzo RL. An introduction to Bayesian data analysis for correlations. PM R 2017;9:1278–1282.ArticlePubMedPDF
• 24. Del Din S, Godfrey A, Galna B, Lord S, Rochester L. Free-living gait characteristics in ageing and Parkinson’s disease: impact of environment and ambulatory bout length. J Neuroeng Rehabil 2016;13:46.PubMedPMC
• 25. Nishi Y, Shigetoh H, Fujii R, Osumi M, Morioka S. Changes in trunk variability and stability of gait in patients with chronic low back pain: impact of laboratory versus daily-living environments. J Pain Res 2021;14:1675–1686.ArticlePubMedPMCPDF
• 26. Yogev-Seligmann G, Hausdorff JM, Giladi N. The role of executive function and attention in gait. Mov Disord 2008;23:329–342.quiz 472. ArticlePubMedPMCPDF
• 27. Welzel J, Wendtland D, Warmerdam E, Romijnders R, Elshehabi M, Geritz J, et al. Step length is a promising progression marker in Parkinson’s disease. Sensors (Basel) 2021;21:2292.ArticlePubMedPMC
• 28. Peterson DS, Horak FB. Neural control of walking in people with parkinsonism. Physiology (Bethesda) 2016;31:95–107.ArticlePubMedPMC
• 29. Mondal B, Choudhury S, Banerjee R, Chatterjee K, Ghosal S, Anand SS, et al. Analysis of gait in Parkinson’s disease reflecting the effect of l-DOPA. Ann Mov Disord 2019;2:21–27.Article
• 30. Huang CY, Chen YA, Wu RM, Hwang IS. Dual-task walking improvement with enhanced kinesthetic awareness in Parkinson’s disease with mild gait impairment: EEG connectivity and clinical implication. Front Aging Neurosci 2022;14:1041378.ArticlePubMedPMC
• 31. Donelan JM, Kram R, Kuo AD. Mechanical and metabolic determinants of the preferred step width in human walking. Proc Biol Sci 2001;268:1985–1992.ArticlePubMedPMC
• 32. Morris ME, Iansek R, Matyas TA, Summers JJ. The pathogenesis of gait hypokinesia in Parkinson’s disease. Brain 1994;117(Pt 5):1169–1181.ArticlePubMed
• 33. Peppe A, Chiavalon C, Pasqualetti P, Crovato D, Caltagirone C. Does gait analysis quantify motor rehabilitation efficacy in Parkinson’s disease patients? Gait Posture 2007;26:452–462.ArticlePubMed
• 34. Djaldetti R, Ziv I, Melamed E. The mystery of motor asymmetry in Parkinson’s disease. Lancet Neurol 2006;5:796–802.ArticlePubMed
• 35. Orcioli-Silva D, Barbieri FA, Dos Santos PCR, Beretta VS, Simieli L, Vitorio R, et al. Double obstacles increase gait asymmetry during obstacle crossing in people with Parkinson’s disease and healthy older adults: a pilot study. Sci Rep 2020;10:2272.ArticlePubMedPMCPDF
• 36. Lee M, Youm C, Noh B, Park H, Cheon SM. Gait characteristics under imposed challenge speed conditions in patients with Parkinson’s disease during overground walking. Sensors (Basel) 2020;20:2132.ArticlePubMedPMC
• 37. Robles-García V, Corral-Bergantiños Y, Espinosa N, Jácome MA, García-Sancho C, Cudeiro J, et al. Spatiotemporal gait patterns during overt and covert evaluation in patients with Parkinson’s disease and healthy subjects: is there a Hawthorne effect? J Appl Biomech 2015;31:189–194.ArticlePubMed
• 38. Bettecken K, Bernhard F, Sartor J, Hobert MA, Hofmann M, Gladow T, et al. No relevant association of kinematic gait parameters with health-related quality of life in Parkinson’s disease. PLoS One 2017;12:e0176816.ArticlePubMedPMC
• 39. Moore O, Peretz C, Giladi N. Freezing of gait affects quality of life of peoples with Parkinson’s disease beyond its relationships with mobility and gait. Mov Disord 2007;22:2192–2195.ArticlePubMed
• 40. Ge Y, Zhao W, Zhang L, Zhao X, Shu X, Li J, et al. Correlation between motor function and health-related quality of life in early to mid-stage patients with Parkinson disease: a cross-sectional observational study. Front Aging Neurosci 2024;16:1399285.ArticlePubMedPMC

Figure & Data

References

Citations

Citations to this article as recorded by  

Comments on this article