Exploring the perspectives of older adults who are pre-frail and frail to identify interventions to reduce sedentary behaviour and improve mobility: a thematic content analysis.

Older adults who are frail are one of the most sedentary and the least physically active age groups. Prolonged sedentary time is associated with increased risk of negative health outcomes. To help design effective and sustainable content and optimize the uptake of sedentary behaviour interventions, an in-depth understanding of older adults' perceptions of sedentary behaviour is needed; however, most qualitative studies have been conducted in healthy older adults. The aim of this study was to explore perspectives of sedentary behaviour within the context of older adults who are pre-frail and frail after the winter and spring. We included participants if they: (1) spoke English or attended with a translator or caregiver, (2) were ≥ 60 years, and (3) were frail on the Morley Frail Scale. We utilized a qualitative description methodology including a semi-structured in-depth interview and thematic content analysis. Concepts from the COM-B (Capability Opportunity Motivation-Behaviour) model were used to guide the semi-structured interviews and analysis. To ensure credibility of the data, we used an audit trail and analyst triangulation. We recruited 21 older adults (72 ± 7.3 years, 13 females, 13 frail) from southwestern Ontario, Canada. Two individuals were lost to follow-up due to medical mistrust and worse health. We transcribed 39 audio recordings. We identified three salient themes: (1) older adults rationalize their sedentary behaviours through cognitive dissonance (reflective motivation), (2) urban cities in southwestern Ontario may not be "age-friendly" (physical opportunity), and (3) exercise is something people "have to do", but hobbies are for enjoyment despite medical conditions (psychological capability). Perspectives of sedentary behaviour were different in the winter versus spring, with participants perceiving themselves to be less active in winter. Incorporating dissonance-based interventions as part of an educational program could be used to target the reflective motivation and psychological capability components. Future research should consider interdisciplinary collaborations with environmental gerontology to develop age-friendly communities that promote meaningful mobility to target physical opportunity.

Mapping sedentary behaviour (MAPS-B) in winter and spring using wearable sensors, indoor positioning systems, and diaries in older adults who are pre-frail and frail: A feasibility longitudinal study.

Older adults who are frail are likely to be sedentary. Prior interventions to reduce sedentary time in older adults have not been effective as there is little research about the context of sedentary behaviour (posture, location, purpose, social environment). Moreover, there is limited evidence on feasible measures to assess context of sedentary behaviour in older adults. The aim of our study was to determine the feasibility of measuring context of sedentary behaviour in older adults with pre-frailty or frailty using a combination of objective and self-report measures. We defined "feasibility process" using recruitment (20 participants within two-months), retention (85%), and refusal (20%) rates and "feasibility resource" if the measures capture context and can be linked (e.g., sitting-kitchen-eating-alone) and are all participants willing to use the measures. Context was assessed using a wearable sensor to assess posture, a smart home monitoring system for location, and an electronic or hard-copy diary for purpose and social context over three days in winter and spring. We approached 80 potential individuals, and 58 expressed interest; of the 58 individuals, 37 did not enroll due to lack of interest or medical mistrust (64% refusal). We recruited 21 older adults (72±7.3 years, 13 females, 13 frail) within two months and experienced two dropouts due to medical mistrust or worsening health (90% retention). The wearable sensor, indoor positioning system, and electronic diary accurately captured one domain of context, but the hard copy was often not completed with enough detail, so it was challenging to link it to the other devices. Although not all participants were willing to use the wearable sensor, indoor positioning system, or electronic diary, we were able to triage the measures of those who did. The use of wearable sensors and electronic diaries may be a feasible method to assess context of sedentary behaviour, but more research is needed with device-based measures in diverse groups.

Validation of physical activity levels from shank-placed Axivity AX6 accelerometers in older adults.

This cross-sectional study aimed to identify and validate cut-points for measuring physical activity using Axivity AX6 accelerometers positioned at the shank in older adults. Free-living physical activity was assessed in 35 adults aged 55 and older, where each participant wore a shank-mounted Axivity and a waist-mounted ActiGraph simultaneously for 72 hours. Optimized cut-points for each participant's Axivity data were determined using an optimization algorithm to align with ActiGraph results. To assess the validity between the physical activity assessments from the optimized Axivity cut-points, a leave-one-out cross-validation was conducted. Bland-Altman plots with 95% limits of agreement, intraclass correlation coefficients (ICC), and mean differences were used for comparing the systems. The results indicated good agreement between the two accelerometers when classifying sedentary behaviour (ICC = 0.85) and light physical activity (ICC = 0.80), and moderate agreement when classifying moderate physical activity (ICC = 0.67) and vigorous physical activity (ICC = 0.70). Upon removal of a significant outlier, the agreement was slightly improved for sedentary behaviour (ICC = 0.86) and light physical activity (ICC = 0.82), but substantially improved for moderate physical activity (ICC = 0.81) and vigorous physical activity (ICC = 0.96). Overall, the study successfully demonstrated the capability of the resultant cut-point model to accurately classify physical activity using Axivity AX6 sensors placed at the shank.

Longitudinal Monitoring of Biomechanical and Psychological State in Collegiate Female Basketball Athletes Using Principal Component Analysis.

Background: The growth in participation in collegiate athletics has been accompanied by increased sport-related injuries. The complex and multifactorial nature of sports injuries highlights the importance of monitoring athletes prospectively using a novel and integrated biopsychosocial approach, as opposed to contemporary practices that silo these facets of health. Methods: Data collected over two competitive basketball seasons were used in a principal component analysis (PCA) model with the following objectives: (i) investigate whether biomechanical PCs (i.e., on-court and countermovement jump (CMJ) metrics) were correlated with psychological state across a season and (ii) explore whether subject-specific significant fluctuations could be detected using minimum detectable change statistics. Weekly CMJ (force plates) and on-court data (inertial measurement units), as well as psychological state (questionnaire) data, were collected on the female collegiate basketball team for two seasons. Results: While some relationships (n = 2) were identified between biomechanical PCs and psychological state metrics, the magnitude of these associations was weak (r = |0.18-0.19|, p < 0.05), and no other overarching associations were identified at the group level. However, post-hoc case study analysis showed subject-specific relationships that highlight the potential utility of red-flagging meaningful fluctuations from normative biomechanical and psychological patterns. Conclusion: Overall, this work demonstrates the potential of advanced analytical modeling to characterize components of and detect statistically and clinically relevant fluctuations in student-athlete performance, health, and well-being and the need for more tailored and athlete-centered monitoring practices.

Accelerometer techniques for capturing human movement validated against direct observation: a scoping review.

INTRODUCTION: Accelerometers are devices commonly used to measure human physical activity and sedentary time. Accelerometer capabilities and analytical techniques have evolved rapidly, making it difficult for researchers to keep track of advances and best practices for data processing and analysis. OBJECTIVE: The objective of this scoping review is to determine the existing methods for analyzing accelerometer data for capturing human movement which have been validated against the criterion measure of direct observation. METHODS: This scoping review searched 14 academic and 5 grey databases. Two independent raters screened by title and abstract, then full text. Data were extracted using Microsoft Excel and checked by an independent reviewer. RESULTS: The search yielded 1039 papers and the final analysis included 115 papers. 71 unique accelerometer models were used across a total of 4217 participants. While all studies underwent validation from direct observation, most direct observation occurred live (55%) or using recordings (42%). Analysis techniques included machine learning approaches (22%), the use of existing cut-points (18%), ROC curves to determine cut-points (14%), and other strategies including regressions and non-machine learning algorithms (8%). DISCUSSION: Machine learning techniques are becoming more prevalent and are often used for activity identification. Cut-point methods are still frequently used. Activity intensity is the most assessed activity outcome; however, both the analyses and outcomes assessed vary by wear location. CONCLUSIONS: This scoping review provides a comprehensive overview of accelerometer analysis and validation techniques using direct observation and is a useful tool for researchers using accelerometers.

Exploring different models of pain phenotypes and their association with pain worsening in people with early knee osteoarthritis: The MOST cohort study.

OBJECTIVE: To determine i) pain phenotypes (PP) in people with early-stage knee osteoarthritis (EKOA); ii) the longitudinal association between the phenotypes and pain worsening at two years. DESIGN: We studied participants with EKOA from the Multicenter Osteoarthritis Study defined as pain intensity ≤3/10, Kellgren and Lawrence grade ≤2, intermittent pain none to sometimes, and no constant pain. Two models of PP were explored. Model A included pressure pain thresholds, temporal summation, conditioned pain modulation, pain catastrophizing, sleep quality, depression, and widespread pain (WSP). In Model B, gait characteristics, quadriceps strength, comorbidities, and magnetic resonance imaging features were added to Model A. Latent Class Analysis was used to create phenotypes, and logistic regression was used to determine their association with pain worsening. RESULTS: 750 individuals (60% females), mean age [standard deviation (SD)]: 60.3 (9.4) were included in Model A and 333 individuals (60% females), mean age (SD): 59.4 (8.1) in Model B. 3-class and 4-class solutions were chosen for Model A and Model B. In Model A, the most "severe" phenotype was dominated by psychosocial factors, WSP, and measures of nervous system sensitization. Similarly in Model B, the Model A phenotype plus gait variables, quadriceps strength, and comorbidities were dominant. Surprisingly, none of the phenotypes in either model had a significant relationship with pain worsening. CONCLUSION: Phenotypes based upon various factors thought to be important for the pain experience were identified in those with EKOA but were not significantly related to pain worsening. These phenotypes require validation with clinically relevant endpoints.

Reliability of waveforms and gait metrics from multiple outdoor wearable inertial sensors collections in adults with knee osteoarthritis.

Wearable sensors may allow research to move outside of controlled laboratory settings to be able to collect real-world data in clinical populations, such as older adults with osteoarthritis. However, the reliability of these sensors must be established across multiple out-of-lab data collections. Nine older adults with symptomatic knee arthritis wore wearable inertial sensors on their proximal tibias during an outdoor 6-minute walk test outside of a controlled laboratory setting as part of a pilot study. Reliability of the underlying waveforms, discrete peak outcomes, and spatiotemporal outcomes were assessed over four separate data collections, each approximately 1 week apart. Reliability at a different number of included strides was also assessed at 10, 20, 50, and 100 strides. The underlying waveforms and discrete peak outcome measures had good-to-excellent reliability for all axes, with lower reliability in frontal plane angular velocity axis. Spatiotemporal outcomes demonstrated excellent reliability. The inclusion of additional strides had little to no effect on reliability in most axes, but the confidence intervals generally became smaller across all axes. However, there was improvement in axes with lower (i.e., good) reliability. These data were collected in an out-of-lab setting, and the results are generally consistent with previous in-lab data collections, likely due to its semi-controlled nature. Additional out-of-laboratory research is required to investigate if these trends continue during truly free-living collections. This study provides support for increasing research conducted in out-of-lab data collections, as demonstrated by the good-to-excellent reliability of all axes.

Monitoring lower limb biomechanical asymmetry and psychological measures in athletic populations-A scoping review.

BACKGROUND: Lower limb biomechanics, including asymmetry, are frequently monitored to determine sport performance level and injury risk. However, contributing factors extend beyond biomechanical and asymmetry measures to include psychological, sociological, and environmental factors. Unfortunately, inadequate research has been conducted using holistic biopsychosocial models to characterize sport performance and injury risk. Therefore, this scoping review summarized the research landscape of studies concurrently assessing measures of lower limb biomechanics, asymmetry, and introspective psychological state (e.g., pain, fatigue, perceived exertion, stress, etc.) in healthy, competitive athletes. METHODS: A systematic search of MEDLINE, Embase, CINAHL, SPORTDiscus, and Web of Science Core Collections was designed and conducted in accordance with PRISMA guidelines. Fifty-one articles were included in this review. RESULTS: Significant relationships between biomechanics (k = 22 studies) or asymmetry (k = 20 studies) and introspective state were found. Increased self-reported pain was associated with decreased range of motion, strength, and increased lower limb asymmetry. Higher ratings of perceived exertion were related to increased lower limb asymmetry, self-reported muscle soreness, and worse jump performance. Few studies (k = 4) monitored athletes longitudinally throughout one or more competitive season(s). CONCLUSION: This review highlights the need for concurrent analysis of introspective, psychological state, and biomechanical asymmetry measures along with longitudinal research to understand the contributing factors to sport performance and injury risk from biopsychosocial modeling. In doing so, this framework of biopsychosocial preventive and prognostic patient-centered practices may provide an actionable means of optimizing health, well-being, and sport performance in competitive athletes.

Mechanisms hypothesized for pain-relieving effects of exercise in fibromyalgia: a scoping review.

Background: Exercise is one of the most recommended management strategies by treatment guidelines for fibromyalgia (FM); however, the mechanism through which exercise improves pain in FM is still unknown. Objective: We aimed to summarize the hypothesized theoretical mechanisms for the pain-relieving effects of exercise in people with FM. Eligibility Criteria: Randomized controlled trials (RCTs) in English reporting mechanisms for pain-relieving effects of exercise in the 'Introduction' and 'Discussion' sections and significant within- group or between-group effects of exercise interventions were included. Sources of Evidence: We searched the databases Ovid MEDLINE(R), EMBASE, CINAHL, COCHRANE, Sports Discuss, and AMED with the keywords: exercise and fibromyalgia until December 2021. Charting Methods: Two authors independently performed title/abstract, full-text review, and data abstraction using a data abstraction form. The hypothesized mechanisms from individual studies were grouped into three categories. Results: The literature search resulted in 2147 studies, out of which 220 studies were considered for full-text review. A total of 50 RCTs proposing 29 unique mechanisms for the pain-relieving effects of exercise were included. These mechanisms were divided into three categories: physical, neuro-physiological, and psychological. The neuro-physiological category was further subdivided into exercise-induced hypoalgesia (EIH), pain sensitization, the autonomic system, the immune system, the endocrine system, and miscellaneous categories. The most frequently hypothesized mechanisms were EIH (n = 15), autonomic modulation (n = 7), improved sleep (n = 6), muscle oxygenation (n = 6), self-efficacy (n = 5), mental health (n = 4), and benefits of the aquatic environment (n = 12). While all exercise interventions involved FM patients, most of the supporting evidence for these mechanisms was cited from previous studies conducted on healthy samples. No studies performed analyses to demonstrate causal associations between the mechanisms and outcomes. Conclusion: Multiple mechanisms were hypothesized for the positive influence of exercise in people with FM. Future studies using causal analyses, such as mediation analysis, are recommended to validate these mechanisms.

The Ecological Validity of Countermovement Jump to On-Court Asymmetry in Basketball.

Jump-based asymmetry is often used as an indicator of sport performance and may be used to discern injury susceptibility. Due to task specificity, however, countermovement jump asymmetry may not be representative of on-court asymmetry. As such, we assessed the association between countermovement jump asymmetry and on-court impact asymmetry metrics (n=3, and n=4, respectively) using linear regressions (α=0.05). Fifteen female basketball athletes completed countermovement jump and on-court sessions across a competitive season. A significant negative association was found between peak landing force asymmetry and both overall and medium acceleration on-court asymmetry (b=-0.1, R 2 =0.08, p<0.001; b=-0.1, R 2 =0.11, p<0.001, respectively), as well as between peak propulsive force asymmetry and on-court medium acceleration asymmetry (b=-0.24, R 2 =0.04, p=0.01). Alternatively, both peak landing and peak propulsive force asymmetry were significantly positively associated with on-court high acceleration asymmetry (b=0.17, R 2 =0.08, p<0.001; b=0.35, R 2 =0.02, p=0.04, respectively). While some overlap may exist, countermovement jump and on-court impact asymmetry appear to be independent. Thus, sport-specific monitoring may be necessary to adequately monitor injury susceptibility using asymmetry.

Original article: Validity and reliability of gait metrics derived from researcher-placed and self-placed wearable inertial sensors.

To compare the inter-session placement reliability for researcher-placed and self-placed sensors, and to evaluate the validity and reliability of waveforms and discrete variables from researcher-placed and self-placed sensors following a previously described alignment correction algorithm. Fourteen healthy, pain-free participants underwent gait analysis over two data collection sessions. Participants self-placed an inertial sensor on their left tibia and a researcher placed one on their right tibia, before completing 10 overground walking trials. Following an axis correction from a principal component analysis-based algorithm, validity and reliability were assessed within and between days for each sensor placement type through Euclidean distances, waveforms, and discrete outcomes. The placement location of researcher-placed sensors exhibited good inter-session reliability (ICC = 0.85) in comparison to self-placed sensors (ICC = 0.55). Similarly, waveforms from researcher-placed sensors exhibited excellent validity across all variables (CMC ≥ 0.90), while self-placed sensors saw high validity for most axes with reductions in validity for mediolateral acceleration and frontal plane angular velocity. Discrete outcomes saw good to excellent reliability across both sensor placement types. A simple alignment correction algorithm for inertial sensor gait data demonstrated good to excellent validity and reliability in self-placed sensors with no additional data or measures. This method can be used to align sensors easily and effectively despite sensor placement errors during straight, level walking to improve 3D gait data outcomes in data collected with self-placed sensors.

Sex differences in the regularity and symmetry of gait in older adults with and without knee osteoarthritis.

BACKGROUND: Three-dimensional (3D) motion measured at the lower back during walking can describe the regularity and symmetry of gait that may be related to osteoarthritis (OA) and functional status. However, gait speed and inherent sex differences, regardless of the presence of OA, may confound these measures. Therefore, there is a need to understand the effect of OA separately among males and females, without the confounding influence of gait speed. OBJECTIVE: To investigate the difference in 3D gait regularity and symmetry measures between gait speed-matched males and females with and without knee OA. METHOD: Gait regularity and symmetry were computed as autocorrelations of pelvic accelerations during treadmill walking in four groups of older adults: healthy asymptomatic females (AsymF; n = 44), healthy asymptomatic males (AsymM; n = 45), females diagnosed with knee OA (OAF; n = 44), and males diagnosed with knee OA (OAM; n = 45). Data were obtained from a larger research database, allowing for the matching of gait speed between groups. The main effect of OA, sex, and interaction effect between them was examined for the 3D gait regularity and symmetry measures at an alpha level of 0.05. RESULTS: There was no main effect of OA on any variable, but there was a significant main effect of sex on mediolateral and anteroposterior gait regularity measures. Specifically, females demonstrated significantly greater gait regularity, most notably in the mediolateral directions compared to males. CONCLUSION: Older adult females were found to display significantly greater mediolateral gait regularity as compared to males, regardless of the presence of OA. Further, this difference exists among matched gait speeds, suggesting it is not the result of gait speed. Overall, these results highlight the importance of sex-specific analyses and considering gait speed when examining gait acceleration patterns near the center of mass for both cross sectional and longitudinal gait assessments.

Wearable Inertial Sensors for Gait Analysis in Adults with Osteoarthritis-A Scoping Review.

Our objective was to conduct a scoping review which summarizes the growing body of literature using wearable inertial sensors for gait analysis in lower limb osteoarthritis. We searched six databases using predetermined search terms which highlighted the broad areas of inertial sensors, gait, and osteoarthritis. Two authors independently conducted title and abstract reviews, followed by two authors independently completing full-text screenings. Study quality was also assessed by two independent raters and data were extracted by one reviewer in areas such as study design, osteoarthritis sample, protocols, and inertial sensor outcomes. A total of 72 articles were included, which studied the gait of 2159 adults with osteoarthritis (OA) using inertial sensors. The most common location of OA studied was the knee (n = 46), followed by the hip (n = 22), and the ankle (n = 7). The back (n = 41) and the shank (n = 40) were the most common placements for inertial sensors. The three most prevalent biomechanical outcomes studied were: mean spatiotemporal parameters (n = 45), segment or joint angles (n = 33), and linear acceleration magnitudes (n = 22). Our findings demonstrate exceptional growth in this field in the last 5 years. Nevertheless, there remains a need for more longitudinal study designs, patient-specific models, free-living assessments, and a push for "Code Reuse" to maximize the unique capabilities of these devices and ultimately improve how we diagnose and treat this debilitating disease.

Validity and reliability of wearable inertial sensors in healthy adult walking: a systematic review and meta-analysis.

BACKGROUND: Inertial measurement units (IMUs) offer the ability to measure walking gait through a variety of biomechanical outcomes (e.g., spatiotemporal, kinematics, other). Although many studies have assessed their validity and reliability, there remains no quantitive summary of this vast body of literature. Therefore, we aimed to conduct a systematic review and meta-analysis to determine the i) concurrent validity and ii) test-retest reliability of IMUs for measuring biomechanical gait outcomes during level walking in healthy adults. METHODS: Five electronic databases were searched for journal articles assessing the validity or reliability of IMUs during healthy adult walking. Two reviewers screened titles, abstracts, and full texts for studies to be included, before two reviewers examined the methodological quality of all included studies. When sufficient data were present for a given biomechanical outcome, data were meta-analyzed on Pearson correlation coefficients (r) or intraclass correlation coefficients (ICC) for validity and reliability, respectively. Alternatively, qualitative summaries of outcomes were conducted on those that could not be meta-analyzed. RESULTS: A total of 82 articles, assessing the validity or reliability of over 100 outcomes, were included in this review. Seventeen biomechanical outcomes, primarily spatiotemporal parameters, were meta-analyzed. The validity and reliability of step and stride times were found to be excellent. Similarly, the validity and reliability of step and stride length, as well as swing and stance time, were found to be good to excellent. Alternatively, spatiotemporal parameter variability and symmetry displayed poor to moderate validity and reliability. IMUs were also found to display moderate reliability for the assessment of local dynamic stability during walking. The remaining biomechanical outcomes were qualitatively summarized to provide a variety of recommendations for future IMU research. CONCLUSIONS: The findings of this review demonstrate the excellent validity and reliability of IMUs for mean spatiotemporal parameters during walking, but caution the use of spatiotemporal variability and symmetry metrics without strict protocol. Further, this work tentatively supports the use of IMUs for joint angle measurement and other biomechanical outcomes such as stability, regularity, and segmental accelerations. Unfortunately, the strength of these recommendations are limited based on the lack of high-quality studies for each outcome, with underpowered and/or unjustified sample sizes (sample size median 12; range: 2-95) being the primary limitation.

Symptomatic knee osteoarthritis is associated with worse but stable quality of life and physical function regardless of the compartmental involvement: Data from the OAI.

Objective: To test whether combined patellofemoral and tibiofemoral osteoarthritis (OA), in addition to symptoms, is associated with greater changes in quality of life and objective physical function measures when compared with asymptomatic isolated tibiofemoral osteoarthritis. Design: Of the 4796 participants in the Osteoarthritis Initiative, 577 were categorized into four groups based on the presence of symptoms (asymptomatic and symptomatic) and the structural involvement within the knee, where tibiofemoral OA was graded with the Kellgren and Lawrence scale, while patellofemoral OA was based on the Magnetic Resonance Imaging Osteoarthritis Knee Scoring cartilage loss feature. Knee-related quality of life was examined using the Knee Injury and Osteoarthritis Outcome Scale quality of life subscale, and objective physical function was examined by the 20 m Walk Test, 30-s Chair Stand Test, and isometric knee strength. These outcomes were measured at Baseline, Year 2, and Year 4. Mixed effects models were fit to test whether the change in outcome, and the Baseline scores, differed based on group. Results: Quality of life worsened for the asymptomatic combined group but improved for the symptomatic combined group. However, these quality of life changes and changes in other outcomes were all within measurement error. Large between-group differences were found at Baseline, whereby individuals with symptoms had worse quality of life and physical function test scores. Conclusions: Quality of life and physical function are largely stable over four years. However, having symptoms is strongly associated with worse quality of life and physical function, regardless of structural disease distribution within the knee.

Individuals with knee osteoarthritis present increased gait pattern deviations as measured by a knee-specific gait deviation index.

BACKGROUND: A biomechanical analysis can provide valuable information on osteoarthritis (OA) gait, but important multidimensional interactions are often ignored. The Gait Deviation Index (GDI) was designed to address the issue of data complexity in gait analyses by providing a single, encompassing, value for one's deviation from a normative reference group. RESEARCH QUESTION: The primary aim of this study was to examine differences in a knee-specific GDI among young adults, and older individuals with and without knee OA. Secondarily, we aimed to examine these differences while controlling for gait speed. METHOD: Sagittal and frontal plane knee joint angles and moments were used in the computation of a GDI among young adults, and older individuals with and without knee OA. The GDI was calculated such that scores ≥100% were considered typical young-healthy gait and a 10% decrease below 100 equated to 1 standard deviation from typical gait. Scores were first examined using a one-way analysis of variance, and examined again after correcting for gait speed. RESULTS: The GDI was calculated for three groups: young-healthy adults (n = 52), older individuals without knee OA (n = 56), and individuals with knee OA (n = 191). Those with knee osteoarthritis exhibited a mean GDI of 87.2 (11.1), which was significantly lower than young adults (99.6 (10.6); p < 0.001) and older individuals without knee OA (94.3 (11.0); p < 0.001). Differences in GDI remained consistent after controlling for gait speed. Knee OA gait waveforms displayed significant variability across similar GDIs, specifically in frontal plane patterns. CONCLUSION: Those with knee osteoarthritis exhibited lower (worse) GDIs compared to those without knee osteoarthritis and young, healthy individuals. After correcting for gait speed, these findings did not change. The GDI highlighted the significant variability in gait waveforms within individuals with knee OA, but the clinical utility of the GDI score itself remains limited.

Subject-specific and group-based running pattern classification using a single wearable sensor.

The objective of this study was to determine whether subject-specific or group-based models provided better classification accuracy to identify changes in biomechanical running gait patterns across different inclination conditions. The classification process was based on measurements from a single wearable sensor using a total of 41,780 strides from eleven recreational runners while running in real-world and uncontrolled environment. Biomechanical variables included pelvic drop, ground contact time, braking, vertical oscillation of pelvis, pelvic rotation, and cadence were recorded during running on three inclination grades: downhill, -2° to -7°; level, -0.2° to +0.2°; and uphill, +2° to +7°. An ensemble and non-linear machine learning algorithm, random forest (RF), was used to classify inclination condition and determine the importance of each of the biomechanical variables. Classification accuracy was determined for subject-specific and group-based RF models. The mean classification accuracy of all subject-specific RF models was 86.29%, while group-based classification accuracy was 76.17%. Braking was identified as the most important variable for all the runners using the group-based model and for most of the runners based on a subject-specific models. In addition, individual runners used different strategies across different inclination conditions and the ranked order of variable importance was unique for each runner. These results demonstrate that subject-specific models can better characterize changes in gait biomechanical patterns compared to a more traditional group-based approach.

New considerations for collecting biomechanical data using wearable sensors: Number of level runs to define a stable running pattern with a single IMU.

Wearable technology can be used to quantify running biomechanical patterns in a runner's natural environment, however, changes in external factors during outdoor running may influence a runner's typical gait pattern. Therefore, the purpose of this study was to determine how many runs are needed to define a stable or typical running pattern. Six biomechanical variables were recorded using a single wearable sensor placed on the lower back during ten outdoor runs for twelve runners. Univariate and multivariate distributions were created and based on the probability density function, the percent of similar data points (within 95%) from each unique run for the same runner were determined. Stability was defined when the addition of data from a new run resulted in less than a 5% change in the probability density function. To cross-validate, the percent of similar data points at stability was compared between the same and different runners using a repeated-measures MANOVA (Bonferroni-corrected α = 0.007). The maximum number of runs needed to reach stability for univariate and multivariate analyses was four and five, respectively. There was a significant overall effect on similar data points between the same and different runners (p = 0.001), with a greater percent of similar data points for the same runner compared to other runners (p < 0.007). Based on biomechanical data collected using a single wearable sensor placed on the lower back, this is the first study to show that four (univariate) to five (multivariate) runs are needed to establish a stable running pattern in real-world settings.

Running patterns for male and female competitive and recreational runners based on accelerometer data.

The purpose of this study was to classify runners in sex-specific groups as either competitive or recreational based on center of mass (CoM) accelerations. Forty-one runners participated in the study (25 male and 16 female), and were labeled as competitive or recreational based on age, sex, and race performance. Three-dimensional acceleration data were collected during a 5-minute treadmill run, and 24 features were extracted. Support vector machine classification models were used to examine the utility of the features in discriminating between competitive and recreational runners within each sex-specific subgroup. Competitive and recreational runners could be classified with 82.63 % and 80.4 % in the male and female models, respectively. Dominant features in both models were related to regularity and variability, with competitive runners exhibiting more consistent running gait patterns, but the specific features were slightly different in each sex-specific model. Therefore, it is important to separate runners into sex-specific competitive and recreational subgroups for future running biomechanical studies. In conclusion, we have demonstrated the ability to analyze running biomechanics in competitive and recreational runners using only CoM acceleration patterns. A runner, clinician, or coach may use this information to monitor how running patterns change as a result of training.

Using wearable sensors to classify subject-specific running biomechanical gait patterns based on changes in environmental weather conditions.

Running-related overuse injuries can result from a combination of various intrinsic (e.g., gait biomechanics) and extrinsic (e.g., running surface) risk factors. However, it is unknown how changes in environmental weather conditions affect running gait biomechanical patterns since these data cannot be collected in a laboratory setting. Therefore, the purpose of this study was to develop a classification model based on subject-specific changes in biomechanical running patterns across two different environmental weather conditions using data obtained from wearable sensors in real-world environments. Running gait data were recorded during winter and spring sessions, with recorded average air temperatures of -10° C and +6° C, respectively. Classification was performed based on measurements of pelvic drop, ground contact time, braking, vertical oscillation of pelvis, pelvic rotation, and cadence obtained from 66,370 strides (~11,000/runner) from a group of recreational runners. A non-linear and ensemble machine learning algorithm, random forest (RF), was used to classify and compute a heuristic for determining the importance of each variable in the prediction model. To validate the developed subject-specific model, two cross-validation methods (one-against-another and partitioning datasets) were used to obtain experimental mean classification accuracies of 87.18% and 95.42%, respectively, indicating an excellent discriminatory ability of the RF-based model. Additionally, the ranked order of variable importance differed across the individual runners. The results from the RF-based machine-learning algorithm demonstrates that processing gait biomechanical signals from a single wearable sensor can successfully detect changes to an individual's running patterns based on data obtained in real-world environments.