Pub Date : 2026-01-01Epub Date: 2025-09-01DOI: 10.1016/j.gaitpost.2025.08.075
Hadis Imani, Ben Godde
Background
Age-related declines in dynamic balance and cognitive control increase fall risk in older adults (OA). Non-invasive brain stimulation, such as anodal transcranial direct current stimulation (a-tDCS), may enhance training outcomes. However, it remains unclear whether stimulation over motor or prefrontal regions is more effective for improving dynamic balance training (DBT) in OA.
Objective
To compare the effects of a-tDCS over the primary motor cortex (M1) vs. dorsolateral prefrontal cortex (DLPFC) on DBT performance and transfer to static balance in OA, and to explore whether baseline resting-state EEG predicts training outcomes.
Methods
In a randomized crossover design, 18 healthy OA completed DBT training during three stimulation conditions (M1, DLPFC, and sham). Static balance performance was assessed before and after training. Resting-state EEG was recorded to assess predictors of training success.
Results
DBT performance improved significantly more with DLPFC stimulation than with M1 or sham. Negative transfer effects were observed on untrained static balance tasks. Lower baseline alpha power predicted stronger training gains with DLPFC stimulation but weaker gains with M1 stimulation.
Conclusion
DLPFC-targeted a-tDCS enhances DBT in OA more effectively than M1 stimulation. Baseline oscillatory brain activity may inform individualized stimulation protocols to optimize balance training outcomes in OA.
{"title":"In older adults resting-state alpha power is associated with stronger effects of anodal tDCS over prefrontal cortex on dynamic balance","authors":"Hadis Imani, Ben Godde","doi":"10.1016/j.gaitpost.2025.08.075","DOIUrl":"10.1016/j.gaitpost.2025.08.075","url":null,"abstract":"<div><h3>Background</h3><div>Age-related declines in dynamic balance and cognitive control increase fall risk in older adults (OA). Non-invasive brain stimulation, such as anodal transcranial direct current stimulation (a-tDCS), may enhance training outcomes. However, it remains unclear whether stimulation over motor or prefrontal regions is more effective for improving dynamic balance training (DBT) in OA.</div></div><div><h3>Objective</h3><div>To compare the effects of a-tDCS over the primary motor cortex (M1) vs. dorsolateral prefrontal cortex (DLPFC) on DBT performance and transfer to static balance in OA, and to explore whether baseline resting-state EEG predicts training outcomes.</div></div><div><h3>Methods</h3><div>In a randomized crossover design, 18 healthy OA completed DBT training during three stimulation conditions (M1, DLPFC, and sham). Static balance performance was assessed before and after training. Resting-state EEG was recorded to assess predictors of training success.</div></div><div><h3>Results</h3><div>DBT performance improved significantly more with DLPFC stimulation than with M1 or sham. Negative transfer effects were observed on untrained static balance tasks. Lower baseline alpha power predicted stronger training gains with DLPFC stimulation but weaker gains with M1 stimulation.</div></div><div><h3>Conclusion</h3><div>DLPFC-targeted a-tDCS enhances DBT in OA more effectively than M1 stimulation. Baseline oscillatory brain activity may inform individualized stimulation protocols to optimize balance training outcomes in OA.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"123 ","pages":"Article 109957"},"PeriodicalIF":2.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145008613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-01Epub Date: 2025-09-12DOI: 10.1016/j.gaitpost.2025.09.004
Juliane Mueller , Julia Simmer , Stefan Schmid , Christoph Zinnen , Steffen Mueller
Background
Backpacks are essential in the daily lives of children. Carrying a heavy backpack affects trunk posture during standing. It remains unclear, whether this effect is also observed during gait.
Research question
How do different backpack weights affect trunk kinematics during walking in children?
Methods
Sixteen children stood and walked on a 5 m walkway with a custom load-carrying-system simulating unloaded and loaded backpacks (10 %;20 %;30 % of body mass (BM). A marker-based 3D motion analysis system captured whole-body kinematics (Rizzoli model). During walking, the primary outcomes were the maximum ranges of motion (RoM;[°]) of thoracic and lumbar trunk segmental angles in three planes. During standing, the average angles over 5 s were measured in three planes. Secondary measures included stride length, stride time, and velocity during walking. The children's own backpacks' weights were measured and expressed as a percentage of body mass. Statistical analysis was performed using repeated-measures ANOVA (α=0.05) and Tukey-Kramer post hoc test.
Results
The average weight of the children’s own backpack was 15.4 ± 7.4 %BM. For the experimental conditions, the average weights added to the load-carrying system were 3.3 ± 0.8 kg (10 %BM), 6.5 ± 1.7 kg (20 %BM), and 9.8 ± 2.5 kg (30 %BM). During standing, the average trunk flexion angles (sagittal plane) of the lumbar trunk segment significantly increased with increased backpack weight (p = 0.002). During walking, no changes in sagittal plane RoM but significant decreases in lumbar and thoracic transversal and frontal plane RoM (p < 0.001), stride length (p = 0.047) and velocity (p = 0.041) were observed with additional weight. No significant differences were observed for stride time between the conditions.
Significance
Added backpack weight led to a more flexed trunk posture during standing and reduced transversal and frontal plane trunk movement, stride length, and gait velocity during walking. These adjustments likely compensate for the dorsally displaced center of mass and minimize energy expenditure by reducing trunk-backpack-angular momentum during walking.
{"title":"Increased backpack weight might lead to increased trunk stiffness during walking in primary school aged children: A pilot study","authors":"Juliane Mueller , Julia Simmer , Stefan Schmid , Christoph Zinnen , Steffen Mueller","doi":"10.1016/j.gaitpost.2025.09.004","DOIUrl":"10.1016/j.gaitpost.2025.09.004","url":null,"abstract":"<div><h3>Background</h3><div>Backpacks are essential in the daily lives of children. Carrying a heavy backpack affects trunk posture during standing. It remains unclear, whether this effect is also observed during gait.</div></div><div><h3>Research question</h3><div>How do different backpack weights affect trunk kinematics during walking in children?</div></div><div><h3>Methods</h3><div>Sixteen children stood and walked on a 5 m walkway with a custom load-carrying-system simulating unloaded and loaded backpacks (10 %;20 %;30 % of body mass (BM). A marker-based 3D motion analysis system captured whole-body kinematics (Rizzoli model). During walking, the primary outcomes were the maximum ranges of motion (RoM;[°]) of thoracic and lumbar trunk segmental angles in three planes. During standing, the average angles over 5 s were measured in three planes. Secondary measures included stride length, stride time, and velocity during walking. The children's own backpacks' weights were measured and expressed as a percentage of body mass. Statistical analysis was performed using repeated-measures ANOVA (α=0.05) and Tukey-Kramer post hoc test.</div></div><div><h3>Results</h3><div>The average weight of the children’s own backpack was 15.4 ± 7.4 %BM. For the experimental conditions, the average weights added to the load-carrying system were 3.3 ± 0.8 kg (10 %BM), 6.5 ± 1.7 kg (20 %BM), and 9.8 ± 2.5 kg (30 %BM). During standing, the average trunk flexion angles (sagittal plane) of the lumbar trunk segment significantly increased with increased backpack weight (p = 0.002). During walking, no changes in sagittal plane RoM but significant decreases in lumbar and thoracic transversal and frontal plane RoM (p < 0.001), stride length (p = 0.047) and velocity (p = 0.041) were observed with additional weight. No significant differences were observed for stride time between the conditions.</div></div><div><h3>Significance</h3><div>Added backpack weight led to a more flexed trunk posture during standing and reduced transversal and frontal plane trunk movement, stride length, and gait velocity during walking. These adjustments likely compensate for the dorsally displaced center of mass and minimize energy expenditure by reducing trunk-backpack-angular momentum during walking.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"123 ","pages":"Article 109970"},"PeriodicalIF":2.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145088699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-01Epub Date: 2025-10-08DOI: 10.1016/j.gaitpost.2025.110002
Sally A. Kenworthy , Stacey L. Gorniak
Background
Trailing limb angle (TLA) has emerged as A potentially clinically feasible measure of propulsion during gait rehabilitation.
Research Question
The aim of this state-of-the-art review was to: (1) determine how propulsion is typically defined in studies using TLA, (2) map the varying definitions of TLA in the literature, and (3) identify what timing indicators are provided in TLA definitions.
Methods
A six-stage approach to the state-of-the-art review is conducted to finalize the research questions, determine the timeline of the review, develop the search strategy, and analyze the results. The literature search was performed using PubMed, EMBASE, CINAHL, Medline, and Web of Science databases.
Results
Seven articles published between 2006 and 2011 were identified as studies key to the development of TLA as a key contributor to propulsion function. Thirty-two articles published between 2011 and 2024 were included in the primary analysis.
Significance
Studies conducted prior to the emergence of TLA in the literature utilized the impulse of the anterior ground reaction force (AGRF) to quantify a propulsion during walking. However, recent studies investigating TLA more frequently use the peak GRF to characterize propulsion. Both propulsion and TLA definitions have evolved in small yet critical ways that may limit TLA use in clinical settings. There is a need to report clinically accessible timing indicators of TLA measurement to ensure reliable and accurate use of this measurement both within and beyond the research setting.
背景：在步态康复过程中，后肢角度（TLA）已成为一种潜在的临床可行的推进力测量方法。研究问题：这篇最新综述的目的是：(1)确定在使用TLA的研究中通常如何定义推进，(2)绘制文献中TLA的不同定义，(3)确定TLA定义中提供了哪些定时指标。方法：采用六个阶段的方法对最新的综述进行最终的研究问题，确定综述的时间表，制定检索策略和分析结果。文献检索使用PubMed、EMBASE、CINAHL、Medline和Web of Science数据库。结果：2006年至2011年间发表的7篇论文被确定为TLA作为推进功能关键贡献者发展的关键研究。2011年至2024年间发表的32篇文章被纳入初步分析。意义：文献中在TLA出现之前进行的研究利用了前地面反作用力（AGRF）的脉冲来量化行走过程中的推进力。然而，最近关于TLA的研究更多地使用峰值GRF来表征推进。推进力和TLA的定义都在小而关键的方面发生了变化，这可能会限制TLA在临床环境中的应用。有必要报告临床可获得的TLA测量的定时指标，以确保在研究环境内外可靠和准确地使用该测量。
{"title":"Trailing limb angle as a clinically feasible measure of propulsion: A state-of-the-art review","authors":"Sally A. Kenworthy , Stacey L. Gorniak","doi":"10.1016/j.gaitpost.2025.110002","DOIUrl":"10.1016/j.gaitpost.2025.110002","url":null,"abstract":"<div><h3>Background</h3><div>Trailing limb angle (TLA) has emerged as A potentially clinically feasible measure of propulsion during gait rehabilitation.</div></div><div><h3>Research Question</h3><div>The aim of this state-of-the-art review was to: (1) determine how propulsion is typically defined in studies using TLA, (2) map the varying definitions of TLA in the literature, and (3) identify what timing indicators are provided in TLA definitions.</div></div><div><h3>Methods</h3><div>A six-stage approach to the state-of-the-art review is conducted to finalize the research questions, determine the timeline of the review, develop the search strategy, and analyze the results. The literature search was performed using PubMed, EMBASE, CINAHL, Medline, and Web of Science databases.</div></div><div><h3>Results</h3><div>Seven articles published between 2006 and 2011 were identified as studies key to the development of TLA as a key contributor to propulsion function. Thirty-two articles published between 2011 and 2024 were included in the primary analysis.</div></div><div><h3>Significance</h3><div>Studies conducted prior to the emergence of TLA in the literature utilized the impulse of the anterior ground reaction force (AGRF) to quantify a propulsion during walking. However, recent studies investigating TLA more frequently use the peak GRF to characterize propulsion. Both propulsion and TLA definitions have evolved in small yet critical ways that may limit TLA use in clinical settings. There is a need to report clinically accessible timing indicators of TLA measurement to ensure reliable and accurate use of this measurement both within and beyond the research setting.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"123 ","pages":"Article 110002"},"PeriodicalIF":2.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145260299","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-01Epub Date: 2025-10-02DOI: 10.1016/j.gaitpost.2025.110000
Yun-Ru Lai , Chih-Cheng Huang , Chia-Yi Lien , Yi-Fang Chiang , Chi-Ping Ting , Chien-Feng Kung , Cheng-Hsien Lu
Background
Gait impairments are common for Parkinson’s disease (PD). With the development of artificial intelligence (AI) technology and three-dimensional Kinect V2 Detectors, it is possible to enable more accurate characterization of gait impairment. We develop a comprehensive prediction model by combining skeleton gait energy image with relative distance and angle for PD motor phenotypes.
Research question
Does the hybrid convolutional neural network-long short-term memory (CNN-LSTM) deep learning model improve diagnostic accuracy and outperform CNN or LSTM models in diagnosing different motor phenotypes of PD?
Method
We implemented and compared three deep learning architectures—CNN, LSTM, and a hybrid CNN-LSTM model. To mitigate class imbalance and enhance classification accuracy, the Synthetic Minority Oversampling Technique was applied. Feature relevance was determined using Random Forest (RF) and SHapley Additive exPlanations (SHAP), facilitating the identification of key predictors. Participants were stratified into three groups—healthy controls, non-postural instability, and gait disturbance (non-PIGD), and PIGD—based on mean scores from selected items of the Unified Parkinson’s Disease Rating Scale.
Results
The CNN–LSTM model demonstrated the highest predictive performance for PIGD classification during straight and turning walking in the off-medication state (AUC = 0.94 for both), followed by the CNN (AUC = 0.85 and 0.88) and LSTM models (AUC = 0.81 and 0.72). Moreover, the CNN–LSTM model achieved the highest classification accuracy across both on- and off-medication conditions. Using the DeLong test, we compared ROC curves of the CNN, LSTM, and hybrid CNN–LSTM models for PIGD classification across straight and turning walking tasks under both on- and off-medication conditions. The hybrid CNN–LSTM model consistently achieved significantly higher AUCs than the CNN and LSTM models in all settings.
Conclusion
Our study demonstrated that using a hybrid CNN-LSTM deep learning model in combination with RF and/or SHAP-based feature analysis, can achieve high classification performance.
{"title":"A comprehensive deep learning model for motor phenotypes of Parkinson's disease using three-dimensional kinect V2 detectors","authors":"Yun-Ru Lai , Chih-Cheng Huang , Chia-Yi Lien , Yi-Fang Chiang , Chi-Ping Ting , Chien-Feng Kung , Cheng-Hsien Lu","doi":"10.1016/j.gaitpost.2025.110000","DOIUrl":"10.1016/j.gaitpost.2025.110000","url":null,"abstract":"<div><h3>Background</h3><div>Gait impairments are common for Parkinson’s disease (PD). With the development of artificial intelligence (AI) technology and three-dimensional Kinect V2 Detectors, it is possible to enable more accurate characterization of gait impairment. We develop a comprehensive prediction model by combining skeleton gait energy image with relative distance and angle for PD motor phenotypes.</div></div><div><h3>Research question</h3><div>Does the hybrid convolutional neural network-long short-term memory (CNN-LSTM) deep learning model improve diagnostic accuracy and outperform CNN or LSTM models in diagnosing different motor phenotypes of PD?</div></div><div><h3>Method</h3><div>We implemented and compared three deep learning architectures—CNN, LSTM, and a hybrid CNN-LSTM model. To mitigate class imbalance and enhance classification accuracy, the Synthetic Minority Oversampling Technique was applied. Feature relevance was determined using Random Forest (RF) and SHapley Additive exPlanations (SHAP), facilitating the identification of key predictors. Participants were stratified into three groups—healthy controls, non-postural instability, and gait disturbance (non-PIGD), and PIGD—based on mean scores from selected items of the Unified Parkinson’s Disease Rating Scale.</div></div><div><h3>Results</h3><div>The CNN–LSTM model demonstrated the highest predictive performance for PIGD classification during straight and turning walking in the off-medication state (AUC = 0.94 for both), followed by the CNN (AUC = 0.85 and 0.88) and LSTM models (AUC = 0.81 and 0.72). Moreover, the CNN–LSTM model achieved the highest classification accuracy across both on- and off-medication conditions. Using the DeLong test, we compared ROC curves of the CNN, LSTM, and hybrid CNN–LSTM models for PIGD classification across straight and turning walking tasks under both on- and off-medication conditions. The hybrid CNN–LSTM model consistently achieved significantly higher AUCs than the CNN and LSTM models in all settings.</div></div><div><h3>Conclusion</h3><div>Our study demonstrated that using a hybrid CNN-LSTM deep learning model in combination with RF and/or SHAP-based feature analysis, can achieve high classification performance.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"123 ","pages":"Article 110000"},"PeriodicalIF":2.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145236614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-01Epub Date: 2025-09-30DOI: 10.1016/j.gaitpost.2025.109994
Shang-Hsi Lin , Ting-Ming Wang , Tung-Wu Lu
Background
Botulinum toxin type A (BoNT-A) injection into the gastrocnemius muscle is widely used to reduce calf hypertrophy, particularly for aesthetic purposes. However, given the essential role of the gastrocnemius muscle in propulsion and postural control during walking, weakening this muscle may subtly affect dynamic balance.
Research question
Does gastrocnemius BoNT-A injection alter whole-body balance control during level walking over time?
Methods
Fifteen healthy female adults received BoNT-A injections to bilateral gastrocnemius. Participants were assessed at baseline, 1 month, and 3 months post-injection. Measurements included shank circumferences and volume, temporospatial gait parameters, COM-COP inclination angles (IA), and their rate of change (RCIA) in both sagittal and frontal planes.
Results
Shank volume showed a statistically significant reduction at one month and three months post-injection. Temporospatial parameters, including step length, cadence, and speed, remained statistically unchanged. While COM-COP IA in both planes was preserved, RCIA increased significantly in the sagittal plane during the first and second half of double limb support (p = 0.033 and 0.050) and in the frontal plane during early single limb support (p = 0.038). These findings suggest subtle changes in dynamic balance control, even without observable gait temporospatial deviations.
Significance
This study highlights that gastrocnemius BoNT-A injections, although preserving basic gait performance, can compromise fine-tuned balance regulation during walking. Clinicians should consider the implications for postural stability, particularly in populations at risk of falls. Future studies should assess the combined effects of footwear (e.g., high-heels) and uneven terrains on post-injection gait balance control.
{"title":"Alterations in dynamic balance control over time following botulinum toxin injection for calf hypertrophy","authors":"Shang-Hsi Lin , Ting-Ming Wang , Tung-Wu Lu","doi":"10.1016/j.gaitpost.2025.109994","DOIUrl":"10.1016/j.gaitpost.2025.109994","url":null,"abstract":"<div><h3>Background</h3><div>Botulinum toxin type A (BoNT-A) injection into the gastrocnemius muscle is widely used to reduce calf hypertrophy, particularly for aesthetic purposes. However, given the essential role of the gastrocnemius muscle in propulsion and postural control during walking, weakening this muscle may subtly affect dynamic balance.</div></div><div><h3>Research question</h3><div>Does gastrocnemius BoNT-A injection alter whole-body balance control during level walking over time?</div></div><div><h3>Methods</h3><div>Fifteen healthy female adults received BoNT-A injections to bilateral gastrocnemius. Participants were assessed at baseline, 1 month, and 3 months post-injection. Measurements included shank circumferences and volume, temporospatial gait parameters, COM-COP inclination angles (IA), and their rate of change (RCIA) in both sagittal and frontal planes.</div></div><div><h3>Results</h3><div>Shank volume showed a statistically significant reduction at one month and three months post-injection. Temporospatial parameters, including step length, cadence, and speed, remained statistically unchanged. While COM-COP IA in both planes was preserved, RCIA increased significantly in the sagittal plane during the first and second half of double limb support (p = 0.033 and 0.050) and in the frontal plane during early single limb support (p = 0.038). These findings suggest subtle changes in dynamic balance control, even without observable gait temporospatial deviations.</div></div><div><h3>Significance</h3><div>This study highlights that gastrocnemius BoNT-A injections, although preserving basic gait performance, can compromise fine-tuned balance regulation during walking. Clinicians should consider the implications for postural stability, particularly in populations at risk of falls. Future studies should assess the combined effects of footwear (e.g., high-heels) and uneven terrains on post-injection gait balance control.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"123 ","pages":"Article 109994"},"PeriodicalIF":2.4,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145260321","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-19DOI: 10.1016/j.gaitpost.2025.110088
Jill Streamer , Vaibhavi Rathod , Robin M. Queen
Background
Symmetry is often defined as a limb difference of less than 10 %, a threshold originally established using sport-specific hop tests. However, its applicability to daily functional tasks in healthy populations remains unclear. This study aimed to evaluate differences among common symmetry indices and assess whether 10 % or 15 % thresholds are appropriate for healthy young adults during everyday activities.
Methods
72 young healthy adults were enrolled. A load sensing insole (200 Hz) was used to collect plantar loading data during level walking (LW), stair ascent (SA), stair descent (SD), and sit-to-stand (STS). Bilateral peak impact force (PIF) and average loading rate (ALR) were used to calculate limb symmetry with three indices (Absolute Symmetry Index (ASI), relative Limb Symmetry Index (rLSI), absolute Normalized Symmetry Index (aNSI)). A chi-squared analysis determined the appropriateness of the 10 % or 15 % threshold. Absolute agreement ICC values assessed agreement between symmetry indices.
Findings
10 % symmetry threshold was only appropriate for PIF symmetry in SA and SD (χ2 = 3.84, α = 0.05). In contrast, a 15 % threshold captured a substantially higher percentage of trials across all tasks, particularly for PIF in gait-related activities. ICC values demonstrated excellent agreement between the ALR and PIF ASI and rLSI. Poor agreement was observed between ALR ASI and aNSI and ALR rLSI and aNSI during walking and SD task.
Interpretation
15 % threshold may better account for natural variability in symmetry, especially for gait-related tasks. Increasing disagreement between symmetry indices shows the need for careful consideration of the symmetry index choice.
{"title":"Assessing functional load symmetry: A case for a 15 % threshold in healthy young adults","authors":"Jill Streamer , Vaibhavi Rathod , Robin M. Queen","doi":"10.1016/j.gaitpost.2025.110088","DOIUrl":"10.1016/j.gaitpost.2025.110088","url":null,"abstract":"<div><h3>Background</h3><div>Symmetry is often defined as a limb difference of less than 10 %, a threshold originally established using sport-specific hop tests. However, its applicability to daily functional tasks in healthy populations remains unclear. This study aimed to evaluate differences among common symmetry indices and assess whether 10 % or 15 % thresholds are appropriate for healthy young adults during everyday activities.</div></div><div><h3>Methods</h3><div>72 young healthy adults were enrolled. A load sensing insole (200 Hz) was used to collect plantar loading data during level walking (LW), stair ascent (SA), stair descent (SD), and sit-to-stand (STS). Bilateral peak impact force (PIF) and average loading rate (ALR) were used to calculate limb symmetry with three indices (Absolute Symmetry Index (ASI), relative Limb Symmetry Index (rLSI), absolute Normalized Symmetry Index (aNSI)). A chi-squared analysis determined the appropriateness of the 10 % or 15 % threshold. Absolute agreement ICC values assessed agreement between symmetry indices.</div></div><div><h3>Findings</h3><div>10 % symmetry threshold was only appropriate for PIF symmetry in SA and SD (χ2 = 3.84, α = 0.05). In contrast, a 15 % threshold captured a substantially higher percentage of trials across all tasks, particularly for PIF in gait-related activities. ICC values demonstrated excellent agreement between the ALR and PIF ASI and rLSI. Poor agreement was observed between ALR ASI and aNSI and ALR rLSI and aNSI during walking and SD task.</div></div><div><h3>Interpretation</h3><div>15 % threshold may better account for natural variability in symmetry, especially for gait-related tasks. Increasing disagreement between symmetry indices shows the need for careful consideration of the symmetry index choice.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"125 ","pages":"Article 110088"},"PeriodicalIF":2.4,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788222","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-19DOI: 10.1016/j.gaitpost.2025.110087
Elvira Molinero-Martín, Eduardo Villamil-Cabello, Antonio Luque-Casado, Miguel Fernandez-del-Olmo
Background
Postural instability, gait disturbances, and impaired functional mobility are key contributors to falls in Parkinson’s disease (PD). Although their interactions have been explored, the influence of cognitive load and visual information on these relationships remains poorly understood.
Objectives
To examine how static postural control, gait, and functional mobility interact in PD, and to determine how these associations are modulated by cognitive demands and the availability of visual feedback.
Methods
Thirty individuals with PD (Hoehn & Yahr I–III) were assessed in the ON-medication state. Participants completed gait tasks under self-selected, maximal, and dual-task conditions; the Timed Up and Go (TUG) test; and static posturography under four conditions (single/dual task × eyes open/closed). Correlation analyses, complementary linear mixed-effects models (LMM) and exploratory factor analysis (EFA) were conducted.
Results
Significant interactions emerged between dual-task step length and mediolateral sway metrics (path length, velocity and range) and the direction of these associations was context-dependent. TUG performance was associated with postural sway only in low-demand conditions. EFA identified three partially independent factors: anterior–posterior postural control, medial-lateral postural control, and gait.
Conclusions
These preliminary findings suggests that gait and posture rely on partially overlapping but flexibly interacting control networks, which become more interdependent as automaticity declines, underscoring the need for rehabilitation strategies that explicitly consider sensory and cognitive influences on motor control.
{"title":"Context-dependent coupling of posture, gait, and functional mobility under cognitive and sensory demands in Parkinson’s disease","authors":"Elvira Molinero-Martín, Eduardo Villamil-Cabello, Antonio Luque-Casado, Miguel Fernandez-del-Olmo","doi":"10.1016/j.gaitpost.2025.110087","DOIUrl":"10.1016/j.gaitpost.2025.110087","url":null,"abstract":"<div><h3>Background</h3><div>Postural instability, gait disturbances, and impaired functional mobility are key contributors to falls in Parkinson’s disease (PD). Although their interactions have been explored, the influence of cognitive load and visual information on these relationships remains poorly understood.</div></div><div><h3>Objectives</h3><div>To examine how static postural control, gait, and functional mobility interact in PD, and to determine how these associations are modulated by cognitive demands and the availability of visual feedback.</div></div><div><h3>Methods</h3><div>Thirty individuals with PD (Hoehn & Yahr I–III) were assessed in the ON-medication state. Participants completed gait tasks under self-selected, maximal, and dual-task conditions; the Timed Up and Go (TUG) test; and static posturography under four conditions (single/dual task × eyes open/closed). Correlation analyses, complementary linear mixed-effects models (LMM) and exploratory factor analysis (EFA) were conducted.</div></div><div><h3>Results</h3><div>Significant interactions emerged between dual-task step length and mediolateral sway metrics (path length, velocity and range) and the direction of these associations was context-dependent. TUG performance was associated with postural sway only in low-demand conditions. EFA identified three partially independent factors: anterior–posterior postural control, medial-lateral postural control, and gait.</div></div><div><h3>Conclusions</h3><div>These preliminary findings suggests that gait and posture rely on partially overlapping but flexibly interacting control networks, which become more interdependent as automaticity declines, underscoring the need for rehabilitation strategies that explicitly consider sensory and cognitive influences on motor control.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"125 ","pages":"Article 110087"},"PeriodicalIF":2.4,"publicationDate":"2025-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145788223","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-14DOI: 10.1016/j.gaitpost.2025.110084
Mark S. Redfern
Background
There are numerous measures of standing balance using force plates presented in the literature. Two of the most common measures are the root-mean-square (RMS) and mean velocity (MV) of the Center of Pressure (CoP). The purpose of this short communication is to promote a greater understanding of the implications of these two common metrics.
Research question
What aspects of postural control do the RMS and MV of the CoP measure?
Methods
CoP time series measured with a force plate and CoM calculated from motion capture during quiet standing in the AP and ML directions were analyzed. The RMS and MV of the CoP, Center of Mass (CoM), and the difference between the CoP and the CoM (CoP-CoM) were calculated. The relationships among these measures are presented.
Results
The CoPRMS was highly correlated with the CoMRMS (r > .96), indicating that CoPRMS measures the amount of sway. The CoPMV was highly correlated with (CoP-CoM)RMS (r > .90). The (CoP-CoM)RMS is related to the torque generation used to maintain stability; therefore CoPMV is related to the control effort used. The AP and ML measures do have some different characteristics due to the mechanism by which stability is maintained.
Significance
The RMS and MV of CoP are effective at capturing two different fundamental aspects of stability: the amount of sway and control effort. This highlights the importance of reporting both CoPRMS and CoPMV to allow for a better understanding of standing balance.
{"title":"Interpreting common standing postural sway measures","authors":"Mark S. Redfern","doi":"10.1016/j.gaitpost.2025.110084","DOIUrl":"10.1016/j.gaitpost.2025.110084","url":null,"abstract":"<div><h3>Background</h3><div>There are numerous measures of standing balance using force plates presented in the literature. Two of the most common measures are the root-mean-square (RMS) and mean velocity (MV) of the Center of Pressure (CoP). The purpose of this short communication is to promote a greater understanding of the implications of these two common metrics.</div></div><div><h3>Research question</h3><div>What aspects of postural control do the RMS and MV of the CoP measure?</div></div><div><h3>Methods</h3><div>CoP time series measured with a force plate and CoM calculated from motion capture during quiet standing in the AP and ML directions were analyzed. The RMS and MV of the CoP, Center of Mass (CoM), and the difference between the CoP and the CoM (CoP-CoM) were calculated. The relationships among these measures are presented.</div></div><div><h3>Results</h3><div>The <em>CoP</em><sub><em>RMS</em></sub> was highly correlated with the <em>CoM</em><sub><em>RMS</em></sub> (r > .96), indicating that <em>CoP</em><sub><em>RMS</em></sub> measures the amount of sway. The <em>CoP</em><sub><em>MV</em></sub> was highly correlated with (<em>CoP-CoM)</em><sub><em>RMS</em></sub> (r > .90). The (<em>CoP-CoM)</em><sub><em>RMS</em></sub> is related to the torque generation used to maintain stability; therefore <em>CoP</em><sub><em>MV</em></sub> is related to the control effort used. The AP and ML measures do have some different characteristics due to the mechanism by which stability is maintained.</div></div><div><h3>Significance</h3><div>The RMS and MV of CoP are effective at capturing two different fundamental aspects of stability: the amount of sway and control effort. This highlights the importance of reporting both <em>CoP</em><sub><em>RMS</em></sub> and <em>CoP</em><sub><em>MV</em></sub> to allow for a better understanding of standing balance.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"125 ","pages":"Article 110084"},"PeriodicalIF":2.4,"publicationDate":"2025-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145784092","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-13DOI: 10.1016/j.gaitpost.2025.110083
Jianqi Pan , Zixiang Gao , Zhanyi Zhou , Diwei Chen , Fengping Li , Julien S. Baker , Yaodong Gu
Objective
Artificial intelligence (AI) methods have been widely applied in gait analysis, yet quantitative comparisons across models and their input–output specifications remain limited. This study aims to systematically review and synthesize the existing literature to evaluate the effectiveness of AI methods in predicting lower limb joint moments during typically developed (TD) gait.
Methods
Relevant studies published before July 1, 2025, were retrieved from five databases (PubMed, Scopus, IEEE Xplore, ScienceDirect, and Web of Science) using Boolean logic operations and were screened according to predefined criteria. Risk of bias and applicability were assessed with PROBAST. Meta-analyses were performed in R using a multilevel random-effects model to examine differences in predictive performance across AI model group, signal input type, and output joints.
Results
Eleven studies involving 371 TD participants met the inclusion criteria. Deep neural networks (DNN) showed the best performance for R2 (0.88, 95 %CI 0.52–1.24), while traditional machine learning (ML) models demonstrated relative superiority for nRMSE (0.11, 95 %CI 0.06–0.29). Among input types, surface EMG (sEMG) achieved the highest R2 (0.96, 95 %CI 0.04–1.89), whereas all inputs except “kinematic and speed and anthropometrics” performed well in the nRMSE analysis. For output joints, the ankle was significantly superior to both the knee (p < 0.001) and the hip (p < 0.001) in terms of R2 and nRMSE.
Conclusion
AI methods can effectively predict lower limb joint moments during TD gait, but differences exist across model group, input type, and output joints. DNN show advantages in fitting complex data, while traditional ML demonstrates greater robustness in small-sample settings. The sEMG, as a process-related input, exhibits high potential, and predictions for the ankle joint are generally superior. Future studies should expand sample size, explore multimodal inputs and advanced modeling strategies, and further validate the applicability of AI methods in pathological gait.
{"title":"Artificial intelligence in lower limb joint moment prediction during typically developed gait: A systematic review and multilevel random-effects meta-analysis","authors":"Jianqi Pan , Zixiang Gao , Zhanyi Zhou , Diwei Chen , Fengping Li , Julien S. Baker , Yaodong Gu","doi":"10.1016/j.gaitpost.2025.110083","DOIUrl":"10.1016/j.gaitpost.2025.110083","url":null,"abstract":"<div><h3>Objective</h3><div>Artificial intelligence (AI) methods have been widely applied in gait analysis, yet quantitative comparisons across models and their input–output specifications remain limited. This study aims to systematically review and synthesize the existing literature to evaluate the effectiveness of AI methods in predicting lower limb joint moments during typically developed (TD) gait.</div></div><div><h3>Methods</h3><div>Relevant studies published before July 1, 2025, were retrieved from five databases (PubMed, Scopus, IEEE Xplore, ScienceDirect, and Web of Science) using Boolean logic operations and were screened according to predefined criteria. Risk of bias and applicability were assessed with PROBAST. Meta-analyses were performed in R using a multilevel random-effects model to examine differences in predictive performance across AI model group, signal input type, and output joints.</div></div><div><h3>Results</h3><div>Eleven studies involving 371 TD participants met the inclusion criteria. Deep neural networks (DNN) showed the best performance for R<sup>2</sup> (0.88, 95 %CI 0.52–1.24), while traditional machine learning (ML) models demonstrated relative superiority for nRMSE (0.11, 95 %CI 0.06–0.29). Among input types, surface EMG (sEMG) achieved the highest R<sup>2</sup> (0.96, 95 %CI 0.04–1.89), whereas all inputs except “kinematic and speed and anthropometrics” performed well in the nRMSE analysis. For output joints, the ankle was significantly superior to both the knee (<em>p</em> < 0.001) and the hip (<em>p</em> < 0.001) in terms of R<sup>2</sup> and nRMSE.</div></div><div><h3>Conclusion</h3><div>AI methods can effectively predict lower limb joint moments during TD gait, but differences exist across model group, input type, and output joints. DNN show advantages in fitting complex data, while traditional ML demonstrates greater robustness in small-sample settings. The sEMG, as a process-related input, exhibits high potential, and predictions for the ankle joint are generally superior. Future studies should expand sample size, explore multimodal inputs and advanced modeling strategies, and further validate the applicability of AI methods in pathological gait.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"125 ","pages":"Article 110083"},"PeriodicalIF":2.4,"publicationDate":"2025-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145784038","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-11DOI: 10.1016/j.gaitpost.2025.110082
Maxwell D. Smith, Rebecca L. Wong, Derek N. Pamukoff
Background
Bone responds to loading by accruing areal bone mineral density (BMD). Distance runners experience a ground reaction force (GRF) during exercise which contributes to bone loading. Sex differences in BMD reflect that males and females respond differently to running-imposed GRF.
Research question
What is the relationship between GRF and BMD in male and female runners?
Methods
Forty participants (20 male; age=25.1 ± 4.5 years; height=1.7 ± 0.1 m; mass=67.2 ± 11.5 kg) who routinely participated in distance running (44.0 ± 26.1 km/week over 4.5 ± 1.5 weekly sessions for the past 6.9 ± 5.2 years) underwent dual x-ray absorptiometry to calculate BMD and ran on a force-instrumented treadmill to measure vertical GRF characteristics at a self-selected (SS) and at a standardized pace (SP; 3.33 m/s). Independent samples t-tests compared outcomes between males and females. Pearson correlation examined associations between GRF and BMD separately based on sex.
Results
Absolute GRF and BMD outcomes were consistently lower in females compared with males (all p < 0.05). At SS, greater BMD in some sites was associated with greater vertical GRF (r = 0.582–0.793, p < 0.001–0.007), vertical loading rate (r = 0.459–0.626, p = 0.003–0.042), and vertical impulse (r = 0.518–0.759, p < 0.001–0.019) in males. Greater BMD in some sites was also associated with greater vertical GRF (r = 0.550–0.736, p < 0.001–0.012), vertical loading rate (r = 0.495–0.718, p < 0.001–0.026), and vertical impulse (r = 0.478–0.755, p < 0.001–0.033) in males at SP. There were no associations between BMD and GRF in females at either pace (r = -0.095–0.360, p = 0.130–0.983).
Significance
The associations between GRF and BMD in runners differ between males and females. Supplemental training methods may be necessary for female runners to influence BMD.
{"title":"Association between bone mineral density and ground reaction force in male and female runners","authors":"Maxwell D. Smith, Rebecca L. Wong, Derek N. Pamukoff","doi":"10.1016/j.gaitpost.2025.110082","DOIUrl":"10.1016/j.gaitpost.2025.110082","url":null,"abstract":"<div><h3>Background</h3><div>Bone responds to loading by accruing areal bone mineral density (BMD). Distance runners experience a ground reaction force (GRF) during exercise which contributes to bone loading. Sex differences in BMD reflect that males and females respond differently to running-imposed GRF.</div></div><div><h3>Research question</h3><div>What is the relationship between GRF and BMD in male and female runners?</div></div><div><h3>Methods</h3><div>Forty participants (20 male; age=25.1 ± 4.5 years; height=1.7 ± 0.1 m; mass=67.2 ± 11.5 kg) who routinely participated in distance running (44.0 ± 26.1 km/week over 4.5 ± 1.5 weekly sessions for the past 6.9 ± 5.2 years) underwent dual x-ray absorptiometry to calculate BMD and ran on a force-instrumented treadmill to measure vertical GRF characteristics at a self-selected (SS) and at a standardized pace (SP; 3.33 m/s). Independent samples t-tests compared outcomes between males and females. Pearson correlation examined associations between GRF and BMD separately based on sex.</div></div><div><h3>Results</h3><div>Absolute GRF and BMD outcomes were consistently lower in females compared with males (all p < 0.05). At SS, greater BMD in some sites was associated with greater vertical GRF (r = 0.582–0.793, <em>p</em> < 0.001–0.007), vertical loading rate (r = 0.459–0.626, <em>p</em> = 0.003–0.042), and vertical impulse (r = 0.518–0.759, <em>p</em> < 0.001–0.019) in males. Greater BMD in some sites was also associated with greater vertical GRF (r = 0.550–0.736, <em>p</em> < 0.001–0.012), vertical loading rate (r = 0.495–0.718, <em>p</em> < 0.001–0.026), and vertical impulse (r = 0.478–0.755, <em>p</em> < 0.001–0.033) in males at SP. There were no associations between BMD and GRF in females at either pace (r = -0.095–0.360, <em>p</em> = 0.130–0.983).</div></div><div><h3>Significance</h3><div>The associations between GRF and BMD in runners differ between males and females. Supplemental training methods may be necessary for female runners to influence BMD.</div></div>","PeriodicalId":12496,"journal":{"name":"Gait & posture","volume":"125 ","pages":"Article 110082"},"PeriodicalIF":2.4,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145776376","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号