Prediction of fall risk among community-dwelling older adults using a wearable system

Impact of Wearable Device-Based Walking Programs on Gait Speed in Older Adults: A Systematic Review and Meta-Analysis

Background

As walking abilities are widely affected among the aging population, investigating the effectiveness of wearable device-based walking programs is essential. The intentions of this meta-analysis were to investigate their effects on gait speed among older adults, as well as to include subgroup analysis to evaluate potential effects on individuals with aging-related conditions such as Parkinson's disease (PD) and stroke.

Methods

Systematic retrieval of Pubmed, The Cochrane Library, Embase and Web of Science databases were searched up to February 2024. Outcomes such as gait speed, balance, cadence, and stride length were extracted and analyzed. Study quality was evaluated using the Rob 2 tool and heterogeneity was tested using I2 statistics through STATA 16.

Results

Nine studies with 284 participants were analyzed. The intervention group showed a significant improvement in gait speed (weighted mean difference (WMD) 0.12; 95% CI 0.03 to 0.21). There is a subgroup analysis suggesting differential effects: significant improvements in PD and stroke subgroups, but not in the normal aging group. Balance (WMD: 1.93; 95% CI: 0.20 to 3.66) and stride length (WMD: 8.58; 95% CI: 3.04 to 14.12) were also shown to improve, but the heterogeneity across the studies was moderate (I2 = 63.91%). No significant changes were observed in the Timed Up and Go test, Gait Variability, and Step Width.

Conclusions

Wearable device-based walking programs improve gait speed in older adults, with top notch advantages in the ones tormented by PD or stroke. These findings advocate that such interventions can be a valuable part of individualized treatment strategies in geriatric care, aiming to enhance mobility and usual satisfactory of existence.

Fall risk prediction using temporal gait features and machine learning approaches

Introduction

Falls have been acknowledged as a major public health issue around the world. Early detection of fall risk is pivotal for preventive measures. Traditional clinical assessments, although reliable, are resource-intensive and may not always be feasible.

Methods

This study explores the efficacy of artificial intelligence (AI) in predicting fall risk, leveraging gait analysis through computer vision and machine learning techniques. Data was collected using the Timed Up and Go (TUG) test and JHFRAT assessment from MMU collaborators and augmented with a public dataset from Mendeley involving older adults. The study introduces a robust approach for extracting and analyzing gait features, such as stride time, step time, cadence, and stance time, to distinguish between fallers and non-fallers.

Results

Two experimental setups were investigated: one considering separate gait features for each foot and another analyzing averaged features for both feet. Ultimately, the proposed solutions produce promising outcomes, greatly enhancing the model's ability to achieve high levels of accuracy. In particular, the LightGBM demonstrates a superior accuracy of 96% in the prediction task.

Discussion

The findings demonstrate that simple machine learning models can successfully identify individuals at higher fall risk based on gait characteristics, with promising results that could potentially streamline fall risk assessment processes. However, several limitations were discovered throughout the experiment, including an insufficient dataset and data variation, limiting the model's generalizability. These issues are raised for future work consideration. Overall, this research contributes to the growing body of knowledge on fall risk prediction and underscores the potential of AI in enhancing public health strategies through the early identification of at-risk individuals.

Identifying determinants for falls among Iranian older adults: insights from the Bushehr Elderly Health Program

BACKGROUND

Falls are a common cause of fractures in older adults. This study aimed to investigate the factors associated with spontaneous falls among people aged ≥ 60 years in southern Iran.

METHODS

The baseline data of 2,426 samples from the second stage of the first phase of a prospective cohort, the Bushehr Elderly Health (BEH) program, were included in the analysis. A history of spontaneous falls in the year before recruitment was measured by self-report using a standardized questionnaire. Demographic characteristics, as well as a history of osteoarthritis, rheumatoid arthritis, low back pain, Alzheimer's disease, epilepsy, depression, and cancer, were measured using standardized questionnaires. A tandem gait (heel-to-toe) exam, as well as laboratory tests, were performed under standard conditions. A multiple logistic regression model was used in the analysis and fitted backwardly using the Hosmer and Lemeshow approach.

RESULTS

The mean (standard deviation) age of the participants was 69.34 (6.4) years, and 51.9% of the participants were women. A total of 260 (10.7%, 95% CI (9.5-12.0)%) participants reported a spontaneous fall in the year before recruitment. Adjusted for potential confounders, epilepsy (OR = 4.31), cancer (OR = 2.73), depression (OR = 1.81), low back pain (OR = 1.79), and osteoarthritis (OR = 1.49) increased the risk of falls in older adults, while the ability to stand ≥ 10 s in the tandem gait exam (OR = 0.49), being male (OR = 0.60), engaging in physical activity (OR = 0.69), and having high serum triglyceride levels (OR = 0.72) reduced the risk of falls.

CONCLUSION

The presence of underlying diseases, combined with other risk factors, is significantly associated with an increased risk of falls among older adults. Given the relatively high prevalence of falls in this population, it is crucial to pay special attention to identifying and addressing these risk factors.

Correlation enhanced distribution adaptation for prediction of fall risk

With technological advancements in diagnostic imaging, smart sensing, and wearables, a multitude of heterogeneous sources or modalities are available to proactively monitor the health of the elderly. Due to the increasing risks of falls among older adults, an early diagnosis tool is crucial to prevent future falls. However, during the early stage of diagnosis, there is often limited or no labeled data (expert-confirmed diagnostic information) available in the target domain (new cohort) to determine the proper treatment for older adults. Instead, there are multiple related but non-identical domain data with labels from the existing cohort or different institutions. Integrating different data sources with labeled and unlabeled samples to predict a patient's condition poses a significant challenge. Traditional machine learning models assume that data for new patients follow a similar distribution. If the data does not satisfy this assumption, the trained models do not achieve the expected accuracy, leading to potential misdiagnosing risks. To address this issue, we utilize domain adaptation (DA) techniques, which employ labeled data from one or more related source domains. These DA techniques promise to tackle discrepancies in multiple data sources and achieve a robust diagnosis for new patients. In our research, we have developed an unsupervised DA model to align two domains by creating a domain-invariant feature representation. Subsequently, we have built a robust fall-risk prediction model based on these new feature representations. The results from simulation studies and real-world applications demonstrate that our proposed approach outperforms existing models.

The Effect of Sensory Reweighting on Postural Control and Cortical Activity in Parkinson's Disease

Aims

Balance requires the cortical control of visual, somatosensory, and vestibular inputs. The aim of this cross-sectional study was to compare the contributions of each of these systems on postural control and cortical activity using a sensory reweighting approach between participants with Parkinson's disease (PD) and controls.

Methods

Ten participants with PD (age: 72 ± 9; 3 women; Hoehn & Yahr: 2 [1.5 - 2.50]) and 11 controls (age: 70 ± 3; 4 women) completed a sensory organization test in virtual reality (VR-SOT) while cortical activity was being recorded using electroencephalography (EEG). Conditions 1 to 3 were completed on a stable platform; conditions 4 to 6 on a foam. Conditions 1 and 4 were done with eyes open; conditions 2 and 5 in a darkened VR environment; and conditions 3 and 6 in a moving VR environment. Linear mixed models were used to evaluate changes in center of pressure (COP) displacement and EEG alpha and theta/beta ratio power between the two groups across the postural control conditions. Condition 1 was used as reference in all analyses.

Results

Participants with PD showed greater COP displacement than controls in the anteroposterior (AP) direction when relying on vestibular input (condition 5; p<0.0001). The mediolateral (ML) COP sway was greater in PD than in controls when relying on the somatosensory (condition 2; p = 0.03), visual (condition 4; p = 0.002), and vestibular (condition 5; p < 0.0001) systems. Participants with PD exhibited greater alpha power compared to controls when relying on visual input (condition 2; p = 0.003) and greater theta/beta ratio power when relying on somatosensory input (condition 4; p = 0.001).

Conclusions

PD affects reweighting of postural control, exemplified by greater COP displacement and increased cortical activity. Further research is needed to establish the temporal dynamics between cortical activity and COP displacement.

Reliability, Validity, and Identification Ability of a Commercialized Waist-Attached Inertial Measurement Unit (IMU) Sensor-Based System in Fall Risk Assessment of Older People

Falls are a prevalent cause of injury among older people. While some wearable inertial measurement unit (IMU) sensor-based systems have been widely investigated for fall risk assessment, their reliability, validity, and identification ability in community-dwelling older people remain unclear. Therefore, this study evaluated the performance of a commercially available IMU sensor-based fall risk assessment system among 20 community-dwelling older recurrent fallers (with a history of ≥2 falls in the past 12 months) and 20 community-dwelling older non-fallers (no history of falls in the past 12 months), together with applying the clinical scale of the Mini-Balance Evaluation Systems Test (Mini-BESTest). The results show that the IMU sensor-based system exhibited a significant moderate to excellent test-retest reliability (ICC = 0.838, p < 0.001), an acceptable level of internal consistency reliability (Spearman's rho = 0.471, p = 0.002), an acceptable convergent validity (Cronbach's α = 0.712), and an area under the curve (AUC) value of 0.590 for the IMU sensor-based receiver-operating characteristic (ROC) curve. The findings suggest that while the evaluated IMU sensor-based system exhibited good reliability and acceptable validity, it might not be able to fully identify the recurrent fallers and non-fallers in a community-dwelling older population. Further system optimization is still needed.

Assessment of fall-associated risk factors in the Muslim community-dwelling older adults of Peshawar, Khyber Pakhtunkhwa, Pakistan

BACKGROUND

Falls are the third-leading cause of disability among the elderly population worldwide. It is multifactorial, and the occurrence of falls depends on different factors, which can be different from context to context, and individual to individual. Therefore, regular assessment of fall risk factors is required to develop a strategy for fall prevention. The study aimed to identify fall-related risk factors in Pakistani healthy older adults at risk of developing physical disabilities. It also aimed to create a risk-predictive model for fall occurrence, offering evidence for preventive strategies.

METHODS

Data were collected from 140 Muslim older adults from two residential areas of Peshawar, Khyber Pakhtunkhwa, from July 2022 to August 25, 2022, after obtaining permission from the Zhengzhou University Ethical Review Board (ZZUIRB #202,254), and the District Health Department Office (DHO #14,207). Participants were informed, and consent was obtained before data collection. Data were collected using the Time Up and Go Test (TUGT) checklist, the Cognitive Screening Scores (CS-10) checklist, interviews regarding the prayer practice, fall history in the last six months, visual equity questions, and demographic variables.

RESULTS

Factors associated with falls were; age, gender, education, cognitive status, poor walking speed, lack of physical activity, poor vision, and history of falls in the last six months, with a significant P value of (P. < 0.05) in the Pearson correlation coefficient test. Poor cognition, low visual equity, poor walking speed, and lack of exercise increase the risk of falling in the future, with a prediction value of (P < 0.005) in Omnibus, Lemeshow score of (0.77).

CONCLUSION

Hence, our study provides a road map for future risk assessment of falls by adding the four mentioned risk factors in the proposed model to facilitate timely action to prevent fall-related infirmities in Pakistani healthy older adults.

Predictability of Fall Risk Assessments in Community-Dwelling Older Adults: A Scoping Review

Fall risk increases with age, and one-third of adults over 65 years old experience a fall annually. Due to the aging population, the number of falls and related medical costs will progressively increase. Correct prediction of who will fall in the future is necessary to timely intervene in order to prevent falls. Therefore, the aim of this scoping review is to determine the predictive value of fall risk assessments in community-dwelling older adults using prospective studies. A total of 37 studies were included that evaluated clinical assessments (questionnaires, physical assessments, or a combination), sensor-based clinical assessments, or sensor- based daily life assessments using prospective study designs. The posttest probability of falling or not falling was calculated. In general, fallers were better classified than non-fallers. Questionnaires had a lower predictive capability compared to the other assessment types. Contrary to conclusions drawn in reviews that include retrospective studies, the predictive value of physical tests evaluated in prospective studies varies largely, with only smaller-sampled studies showing good predictive capabilities. Sensor-based fall risk assessments are promising and improve with task complexity, although they have only been evaluated in relatively small samples. In conclusion, fall risk prediction using sensor data seems to outperform conventional tests, but the method's validity needs to be confirmed by large prospective studies.

Exploring Teslasuit's Potential in Detecting Sequential Slip-Induced Kinematic Changes among Healthy Young Adults

This study aimed to assess whether the Teslasuit, a wearable motion-sensing technology, could detect subtle changes in gait following slip perturbations comparable to an infrared motion capture system. A total of 12 participants wore Teslasuits equipped with inertial measurement units (IMUs) and reflective markers. The experiments were conducted using the Motek GRAIL system, which allowed for accurate timing of slip perturbations during heel strikes. The data from Teslasuit and camera systems were analyzed using statistical parameter mapping (SPM) to compare gait patterns from the two systems and before and after slip. We found significant changes in ankle angles and moments before and after slip perturbations. We also found that step width significantly increased after slip perturbations (p = 0.03) and total double support time significantly decreased after slip (p = 0.01). However, we found that initial double support time significantly increased after slip (p = 0.01). However, there were no significant differences observed between the Teslasuit and motion capture systems in terms of kinematic curves for ankle, knee, and hip movements. The Teslasuit showed promise as an alternative to camera-based motion capture systems for assessing ankle, knee, and hip kinematics during slips. However, some limitations were noted, including kinematics magnitude differences between the two systems. The findings of this study contribute to the understanding of gait adaptations due to sequential slips and potential use of Teslasuit for fall prevention strategies, such as perturbation training.

The comparative effect of exercise interventions on balance in perimenopausal and early postmenopausal women: A systematic review and network meta-analysis of randomised, controlled trials

In addition to a range of physiological and psychological symptoms, menopause causes a decrement to balance performance and risk of falls. This review aimed to determine the effects of exercise interventions on balance in perimenopausal and early postmenopausal women. Web of Science, PubMed, CINAHL, SPORTDiscus and Cochrane Central Register of Controlled Trials databases were searched. Randomised, controlled trials of exercise interventions in perimenopausal or early postmenopausal populations with an average age of 65 years or younger reporting balance measures were included. Risk of bias was assessed using Cochrane RoB 2. A random effects model network meta-analysis was performed to assess the effect of exercise on balance. Standardised mean differences with 95 % confidence intervals were used as the measure of effect. Twenty-six studies were included after screening. Network meta-analyses were conducted for 5 balance variables. Whole-body vibration (standardised mean difference: 2.25, confidence interval: 0.08; 4.43), balance (standardised mean difference: 1.84, confidence interval: 0.15; 3.53), balance + nutrition (standardised mean difference: 3.81, confidence interval: 1.57; 6.05) and resistance (standardised mean difference: 1.43, confidence interval: 0.41; 2.46) exercise improved Berg balance scale performance. Resistance + aerobic + balance exercise improved one-leg stance (standardised mean difference: 0.80, confidence interval: 0.39; 1.22) and whole-body vibration improved anterior-posterior (standardised mean difference: -0.89, confidence interval: -1.48; -0.31), medio-lateral (standardised mean difference: -0.58, confidence interval: -1.15; -0.01) postural sway and falls indices (standardised mean difference: -0.75, confidence interval: -1.45; -0.04). Exercise improved all balance measures and should be considered as an adjunct therapy in perimenopausal and postmenopausal women. Whole-body vibration was most frequently the highest ranked intervention; resistance and balance training also improved balance.

Responsiveness of Daily Life Gait Quality Characteristics over One Year in Older Adults Who Experienced a Fall or Engaged in Balance Exercise

Gait quality characteristics obtained from daily-life accelerometry are clinically relevant for fall risk in older adults but it is unknown whether these characteristics are responsive to changes in gait quality. We aimed to test whether accelerometry-based daily-life gait quality characteristics are reliable and responsive to changes over one year in older adults who experienced a fall or an exercise intervention. One-week trunk acceleration data were collected from 522 participants (65-97 years), at baseline and after one year. We calculated median values of walking speed, regularity (sample entropy), stability (logarithmic rate of divergence per stride), and a gait quality composite score, across all 10-s gait epochs derived from one-week gait episodes. Intraclass correlation coefficients (ICC) and limits of agreement (LOA) were determined for 198 participants who did not fall nor participated in an exercise intervention during follow-up. For responsiveness to change, we determined the number of participants who fell (n = 209) or participated in an exercise intervention (n = 115) that showed a change beyond the LOA. ICCs for agreement between baseline and follow-up exceeded 0.70 for all gait quality characteristics except for vertical gait stability (ICC = 0.69, 95% CI [0.62, 0.75]) and walking speed (ICC = 0.68, 95% CI [0.62, 0.74]). Only walking speed, vertical and mediolateral gait stability changed significantly in the exercisers over one year but effect sizes were below 0.2. The characteristic associated with most fallers beyond the LOA was mediolateral sample entropy (4.8% of fallers). For the exercisers, this was gait stability in three directions and the gait quality composite score (2.6% of exercisers). The gait quality characteristics obtained by median values over one week of trunk accelerometry were not responsive to presumed changes in gait quality after a fall or an exercise intervention in older people. This is likely due to large (within subjects) differences in gait behaviour that participants show in daily life.

Computer Vision and Machine Learning-Based Gait Pattern Recognition for Flat Fall Prediction

BACKGROUND

Gait recognition has been applied in the prediction of the probability of elderly flat ground fall, functional evaluation during rehabilitation, and the training of patients with lower extremity motor dysfunction. Gait distinguishing between seemingly similar kinematic patterns associated with different pathological entities is a challenge for the clinician. How to realize automatic identification and judgment of abnormal gait is a significant challenge in clinical practice. The long-term goal of our study is to develop a gait recognition computer vision system using artificial intelligence (AI) and machine learning (ML) computing. This study aims to find an optimal ML algorithm using computer vision techniques and measure variables from lower limbs to classify gait patterns in healthy people. The purpose of this study is to determine the feasibility of computer vision and machine learning (ML) computing in discriminating different gait patterns associated with flat-ground falls.

METHODS

We used the Kinect® Motion system to capture the spatiotemporal gait data from seven healthy subjects in three walking trials, including normal gait, pelvic-obliquity-gait, and knee-hyperextension-gait walking. Four different classification methods including convolutional neural network (CNN), support vector machine (SVM), K-nearest neighbors (KNN), and long short-term memory (LSTM) neural networks were used to automatically classify three gait patterns. Overall, 750 sets of data were collected, and the dataset was divided into 80% for algorithm training and 20% for evaluation.

RESULTS

The SVM and KNN had a higher accuracy than CNN and LSTM. The SVM (94.9 ± 3.36%) had the highest accuracy in the classification of gait patterns, followed by KNN (94.0 ± 4.22%). The accuracy of CNN was 87.6 ± 7.50% and that of LSTM 83.6 ± 5.35%.

CONCLUSIONS

This study revealed that the proposed AI machine learning (ML) techniques can be used to design gait biometric systems and machine vision for gait pattern recognition. Potentially, this method can be used to remotely evaluate elderly patients and help clinicians make decisions regarding disposition, follow-up, and treatment.

A Systematic Review of Wearable Sensor-Based Technologies for Fall Risk Assessment in Older Adults

Falls have been recognized as the major cause of accidental death and injury in people aged 65 and above. The timely prediction of fall risks can help identify older adults prone to falls and implement preventive interventions. Recent advancements in wearable sensor-based technologies and big data analysis have spurred the development of accurate, affordable, and easy-to-use approaches to fall risk assessment. The objective of this study was to systematically assess the current state of wearable sensor-based technologies for fall risk assessment among community-dwelling older adults. Twenty-five of 614 identified research articles were included in this review. A comprehensive comparison was conducted to evaluate these approaches from several perspectives. In general, these approaches provide an accurate and effective surrogate for fall risk assessment. The accuracy of fall risk prediction can be influenced by various factors such as sensor location, sensor type, features utilized, and data processing and modeling techniques. Features constructed from the raw signals are essential for predictive model development. However, more investigations are needed to identify distinct, clinically interpretable features and develop a general framework for fall risk assessment based on the integration of sensor technologies and data modeling.

Impact of mild COVID-19 on balance function in young adults, a prospective observational study

Balance is of essential importance in human life. The aim of the study is to examine the incidence of balance impairments in young adults who have recovered from mild COVID-19. The study involved 100 subjects, divided into two groups: the study group (50 individuals) comprised subjects who had recovered from mild COVID-19, and the control group (50 individuals) consisted of healthy subjects matched for gender and age. Balance was assessed using a force platform and clinical tests such as: timed up and go test, 15-s step test, sit-to-stand test and 6-min walk test. The assessment on the platform showed greater balance impairments in the trials with eyes closed; more specifically, compared to the controls, in trials with double-leg support the subjects from the study group acquired significantly higher scores in X average (lateral coordinates) (p < 0.05), Path length, V average (average Centre of Foot Pressure Velocity) (p < 0.05) and Area circular (p < 0.01), with even more significant results in trials with single-leg support in X average (p < 0.001), Y average (anterior–posterior coordinates) (p < 0.001) and Path length (p = 0.004). Higher scores in the timed up and go test were found in the study group (p = 0.013). The control group had higher scores in the remaining tests. The current findings show that mild COVID-19 may lead to balance impairments in young adults. Statistically significant differences in balance were found between the subjects in the study group and the healthy controls. Further studies in this area should take into account more age groups, and patients recovered from severe COVID-19, and should investigate long-term consequences of COVID-19 reflected by balance problems.

Quantitative Characterization of Motor Control during Gait in Dravet Syndrome Using Wearable Sensors: A Preliminary Study

Dravet syndrome (DS) is a rare and severe form of genetic epilepsy characterized by cognitive and behavioural impairments and progressive gait deterioration. The characterization of gait parameters in DS needs efficient, non-invasive quantification. The aim of the present study is to apply nonlinear indexes calculated from inertial measurements to describe the dynamics of DS gait. Twenty participants (7 M, age 9-33 years) diagnosed with DS were enrolled. Three wearable inertial measurement units (OPAL, Apdm, Portland, OR, USA; Miniwave, Cometa s.r.l., Italy) were attached to the lower back and ankles and 3D acceleration and angular velocity were acquired while participants walked back and forth along a straight path. Segmental kinematics were acquired by means of stereophotogrammetry (SMART, BTS). Community functioning data were collected using the functional independence measure (FIM). Mean velocity and step width were calculated from stereophotogrammetric data; fundamental frequency, harmonic ratio, recurrence quantification analysis, and multiscale entropy (τ = 1...6) indexes along anteroposterior (AP), mediolateral (ML), and vertical (V) axes were calculated from trunk acceleration. Results were compared to a reference age-matched control group (112 subjects, 6-25 years old). All nonlinear indexes show a disruption of the cyclic pattern of the centre of mass in the sagittal plane, quantitatively supporting the clinical observation of ataxic gait. Indexes in the ML direction were less altered, suggesting the efficacy of the compensatory strategy (widening the base of support). Nonlinear indexes correlated significantly with functional scores (i.e., FIM and speed), confirming their effectiveness in capturing clinically meaningful biomarkers of gait.