Predicting Future Mobility Limitation in Older Adults: A Machine Learning Analysis of Health ABC Study Data

Identifying and predicting physical limitation and cognitive decline trajectory group of older adults in China: A data-driven machine learning analysis

OBJECTIVE

This study aimed to utilize data-driven machine learning methods to identify and predict potential physical and cognitive function trajectory groups of older adults and determine their crucial factors for promoting active ageing in China.

METHODS

Longitudinal data on 3026 older adults from the Chinese Longitudinal Healthy Longevity and Happy Family Survey was used to identify potential physical and cognitive function trajectory groups using a group-based multi-trajectory model (GBMTM). Predictors were selected from sociodemographic characteristics, lifestyle factors, and physical and mental conditions. The trajectory groups were predicted using data-driven machine learning models and dynamic nomogram. Model performance was evaluated by area under the receiver operating characteristics curve (AUROC), area under the precision-recall curve (PRAUC), and confusion matrix.

RESULTS

Two physical and cognitive function trajectory groups were determined, including a trajectory group with physical limitation and cognitive decline (14.18 %) and a normal trajectory group (85.82 %). Logistic regression performed well in predicting trajectory groups (AUROC = 0.881, PRAUC = 0.649). Older adults with lower baseline score of activities of daily living, older age, less frequent housework, and fewer actual teeth were more likely to experience physical limitation and cognitive decline trajectory group.

LIMITATION

This study didn't carry out external validation.

CONCLUSIONS

This study shows that GBMTM and machine learning models effectively identify and predict physical limitation and cognitive decline trajectory group. The identified predictors might be essential for developing targeted interventions to promote healthy ageing.

Verification of a comprehensive framework for mobility using data from the Canadian Longitudinal Study on Aging: a structural equation modeling analysis

BACKGROUND

Mobility within and between life spaces is fundamental for health and well-being. Our objective was to verify a comprehensive framework for mobility.

METHODS

This was a cross-sectional study. We used structural equation modeling to estimate associations between latent factors with data from the Canadian Longitudinal Study on Aging for participants 65-85 years of age (65+, n = 11,667) and for adults with osteoarthritis (OA) aged 45-85 (n = 5,560). Latent factors included life space mobility, and physical, psychosocial, environmental, financial, and cognitive elements. Personal variables (age, sex, education) were covariates.

RESULTS

The models demonstrated good fit (65+: CFI = 0.90, RMSEA (90% CI) = 0.025 (0.024, 0.026); OA: CFI = 0.90, RMSEA (90% CI) = 0.032 (0.031, 0.033)). In both models, better psychosocial and physical health, and being less afraid to walk after dark (observed environmental variable) were associated with greater life space mobility. Greater financial status was associated with better psychosocial and physical health. Higher education was related to better cognition and finances. Older age was associated with lower financial status, cognition, and physical health. Cognitive health was positively associated with greater mobility only in the 65 + model. Models generated were equivalent for males and females.

CONCLUSIONS

Associations between determinants described in the mobility framework were verified with adults 65-85 years of age and in an OA group when all factors were considered together using SEM. These results have implications for clinicians and researchers in terms of important outcomes when assessing life space mobility; findings support interdisciplinary analyses that include evaluation of cognition, depression, anxiety, environmental factors, and community engagement, as well as physical and financial health. Public policies that influence older adults and their abilities to access communities beyond their homes need to reflect the complexity of factors that influence life space mobility at both individual and societal levels.

Disability risk prediction model based on machine learning among Chinese healthy older adults: results from the China Health and Retirement Longitudinal Study

Background

Predicting disability risk in healthy older adults in China is essential for timely preventive interventions, improving their quality of life, and providing scientific evidence for disability prevention. Therefore, developing a machine learning model capable of evaluating disability risk based on longitudinal research data is crucial.

Methods

We conducted a prospective cohort study of 2,175 older adults enrolled in the China Health and Retirement Longitudinal Study (CHARLS) between 2015 and 2018 to develop and validate this prediction model. Several machine learning algorithms (logistic regression, k-nearest neighbors, naive Bayes, multilayer perceptron, random forest, and XGBoost) were used to assess the 3-year risk of developing disability. The optimal cutoff points and adjustment parameters are explored in the training set, the prediction accuracy of the models is compared in the testing set, and the best-performing models are further interpreted.

Results

During a 3-year follow-up period, a total of 505 (23.22%) healthy older adult individuals developed disabilities. Among the 43 features examined, the LASSO regression identified 11 features as significant for model establishment. When comparing six different machine learning models on the testing set, the XGBoost model demonstrated the best performance across various evaluation metrics, including the highest area under the ROC curve (0.803), accuracy (0.757), sensitivity (0.790), and F1 score (0.789), while its specificity was 0.712. The decision curve analysis (DCA) indicated showed that XGBoost had the highest net benefit in most of the threshold ranges. Based on the importance of features determined by SHAP (model interpretation method), the top five important features were identified as right-hand grip strength, depressive symptoms, marital status, respiratory function, and age. Moreover, the SHAP summary plot was used to illustrate the positive or negative effects attributed to the features influenced by XGBoost. The SHAP dependence plot explained how individual features affected the output of the predictive model.

Conclusion

Machine learning-based prediction models can accurately evaluate the likelihood of disability in healthy older adults over a period of 3 years. A combination of XGBoost and SHAP can provide clear explanations for personalized risk prediction and offer a more intuitive understanding of the effect of key features in the model.

Weight-cycling over 6 years is associated with pain, physical function and depression in the Osteoarthritis Initiative cohort

Body weight significantly impacts health and quality of life, and is a leading risk factor for the development of knee osteoarthritis (OA). Weight cycling may have more negative health consequences compared to steady high or low weight. Using the Osteoarthritis Initiative dataset, we investigated the effects of weight cycling on physical function, quality of life, and depression over 72-months compared to stable or unidirectional body weight trajectories. Participants (n = 731) had knee OA and were classified as: (1) stable-low (BMI < 25), (2) stable-overweight (BMI = 25-29.9), and (3) stable-obese (BMI ≥ 30); (4) steady-weight-loss; (5) steady-weight-gain (weight loss/gain ≥ 2.2 kg every 2-years); (6) gain-loss-gain weight cycling, and (7) loss-gain-loss weight cycling (weight loss/gain with return to baseline), based on bi-annual assessments. We compared Knee Injury and Osteoarthritis Outcome Knee-Related Quality of Life, Function in Sports and Recreation, Physical Activity in the Elderly, Short Form SF-12, repeated chair rise, 20-m gait speed, and Center for Epidemiological Studies Depression using repeated-measures ANOVA. The steady weight loss group demonstrated the worst pain, physical function, and depressive symptoms over time (p's < 0.05). More research is needed to confirm these findings, and elucidate the mechanisms by which steady weight loss is associated with functional decline in knee OA.

Feasibility of a Smartwatch Platform to Assess Ecological Mobility: Real-Time Online Assessment and Mobility Monitor

BACKGROUND

Early detection of mobility decline is critical to prevent subsequent reductions in quality of life, disability, and mortality. However, traditional approaches to mobility assessment are limited in their ability to capture daily fluctuations that align with sporadic health events. We aim to describe findings from a pilot study of our Real-time Online Assessment and Mobility Monitor (ROAMM) smartwatch application, which uniquely captures multiple streams of data in real time in ecological settings.

METHODS

Data come from a sample of 31 participants (Mage = 74.7, 51.6% female) who used ROAMM for approximately 2 weeks. We describe the usability and feasibility of ROAMM, summarize prompt data using descriptive metrics, and compare prompt data with traditional survey-based questionnaires or other established measures.

RESULTS

Participants were satisfied with ROAMM's function (87.1%) and ranked the usability as "above average." Most were highly engaged (average adjusted compliance = 70.7%) and the majority reported being "likely" to enroll in a 2-year study (77.4%). Some smartwatch features were correlated with their respective traditional measurements (eg, certain GPS-derived life-space mobility features (r = 0.50-0.51, p < .05) and ecologically measured pain (r = 0.72, p = .01), but others were not (eg, ecologically measured fatigue).

CONCLUSIONS

ROAMM was usable, acceptable, and effective at measuring mobility and risk factors for mobility decline in our pilot sample. Additional work with a larger and more diverse sample is necessary to confirm associations between smartwatch-measured features and traditional measures. By monitoring multiple data streams simultaneously in ecological settings, this technology could uniquely contribute to the evolution of mobility measurement and risk factors for mobility loss.

Effects of visual restoration on gait performance and kinematics of lower extremities in patients with age-related cataract

BACKGROUND

Visual inputs are critical for locomotor navigation and sensorimotor integration in the elderly; however, the mechanism needs to be explored intensively. The present study assessed the gait pattern after cataract surgery to investigate the effects of visual restoration on locomotion.

METHODS

The prospective study recruited 32 patients (70.1 ± 5.2 years) with bilateral age-related cataracts in the Department of Ophthalmology at Peking University Third Hospital from October 2016 to December 2019. The temporal-spatial gait parameters and kinematic parameters were measured by the Footscan system and inertial measurement units. Paired t -test was employed to compare data normally distributed and Wilcoxon rank-sum test for non-normally distributed.

RESULTS

After visual restoration, the walking speed increased by 9.3% (1.19 ± 0.40 m/s vs. 1.09 ± 0.34 m/s, P =0.008) and exhibited an efficient gait pattern with significant decrease in gait cycle (1.02 ± 0.08 s vs. 1.04 ± 0.07 s, P =0.012), stance time (0.66 ± 0.06 s vs. 0.68 ± 0.06 s, P =0.045), and single support time (0.36 ± 0.03 s vs. 0.37 ± 0.02 s, P =0.011). High amplitude of joint motion was detected in the sagittal plane in the left hip (37.6° ± 5.3° vs. 35.5° ± 6.2°, P =0.014), left thigh (38.0° ± 5.2° vs. 36.4° ± 5.8°, P =0.026), left shank (71.9° ± 5.7° vs. 70.1° ± 5.6°, P =0.031), and right knee (59.1° ± 4.8° vs. 56.4° ± 4.8°, P =0.001). The motor symmetry of thigh improved from 8.35 ± 5.30% to 6.30 ± 4.73% ( P =0.042).

CONCLUSIONS

The accelerated gait in response to visual restoration is characterized by decreased stance time and increased range of joint motion. Training programs for improving muscle strength of lower extremities might be helpful to facilitate the adaptation to these changes in gait.

CCR5-Δ32 polymorphism-a possible protective factor from gait impairment amongst post-stroke patients

BACKGROUND AND PURPOSE

Stroke and small vessel disease cause gait disturbances and falls. The naturally occurring loss-of-function mutation in the C-C chemokine receptor 5 gene (CCR5-Δ32) has recently been reported as a protective factor in post-stroke motor and cognitive recovery. We sought to examine whether it also influences gait and balance measures up to 2 years after stroke.

METHOD

Participants were 575 survivors of first-ever, mild-moderate ischaemic stroke or transient ischaemic attack from the TABASCO prospective study, who underwent a 3 T magnetic resonance imaging at baseline and were examined by a multi-professional team 6, 12 and 24 months after the event, using neurological, neuropsychological and mobility examinations. Gait rhythm and the timing of the gait cycle were measured by force-sensitive insoles. CCR5-Δ32 status and gait measures were available for 335 patients.

RESULTS

CCR5-Δ32 carriers (16.4%) had higher gait speed and decreased (better) stride and swing time variability 6 and 12 months after the index event compared to non-carriers (p < 0.01 for all). The association remained significant after adjustment for age, gender, education, ethnicity and stroke severity.

CONCLUSIONS

Significant associations were found between gait measurements and CCR5-Δ32 loss-of-function mutation amongst stroke survivors. This is the first study showing that genetic predisposition may predict long-term gait function after ischaemic stroke.

Predicting restriction of life-space mobility: a machine learning analysis of the IMIAS study

BACKGROUND

Some studies have employed machine learning (ML) methods for mobility prediction modeling in older adults. ML methods could be a helpful tool for life-space mobility (LSM) data analysis.

AIM

This study aimed to evaluate the predictive value of ML algorithms for the restriction of life-space mobility (LSM) among elderly people and to identify the most important risk factors for that prediction model.

METHODS

A 2-year LSM reduction prediction model was developed using the ML-based algorithms decision tree, random forest, and eXtreme gradient boosting (XGBoost), and tested on an independent validation cohort. The data were collected from the International Mobility in Aging Study (IMIAS) from 2012 to 2014, comprising 372 older patients (≥ 65 years of age). LSM was measured by the Life-Space Assessment questionnaire (LSA) with five levels of living space during the month before assessment.

RESULTS

According to the XGBoost algorithm, the best model reached a mean absolute error (MAE) of 10.28 and root-mean-square error (RMSE) of 12.91 in the testing portion. The variables frailty (39.4%), mobility disability (25.4%), depression (21.9%), and female sex (13.3%) had the highest importance.

CONCLUSION

The model identified risk factors through ML algorithms that could be used to predict LSM restriction; these risk factors could be used by practitioners to identify older adults with an increased risk of LSM reduction in the future. The XGBoost model offers benefits as a complementary method of traditional statistical approaches to understand the complexity of mobility.