Identifying subjective life expectancy risk factors in physically active and inactive middle-aged and older adults using machine learning models

Hepatitis B virus (HBV) infection remains a public health concern in China, especially among middle-aged and older adults. Accurate prediction models can assist in early detection and targeted interventions. However, predicting HBV infection using population-level data is challenging due to its relatively low prevalence, which creates highly imbalanced datasets that undermine the performance of traditional predictive models. This study aimed to develop a machine learning model to predict HBV infection using an ensemble approach designed to handle class imbalance. Our study conducted a cross-sectional analysis of participants aged 45 and above from the China Health and Retirement Longitudinal Study 2011. To address the class imbalance in HBV infection, the synthetic minority oversampling technique (SMOTE) was employed. We developed machine learning models, including logistic regression (LR), support vector machine (SVM), random forest (RF), gradient boosting machine (GBM), adaptive boosting classifier (AdaBoost), extreme gradient boosting (XGBoost), and stacking ensemble model (SEM), to predict HBV infection and identify predictors. Model performance was evaluated using area under the receiver operating characteristic curve (AUC), area under the precision-recall curve (AP), accuracy, sensitivity, specificity, precision, F1 score, and Brier score. Shapley additive explanation (SHAP) analysis was applied to interpret each predictor’s individual contribution and direction to the model’s output. In the final analysis, 8,954 participants were included, among whom 0.64% self-reported HBV infection, confirming highly imbalanced data. The SEM model achieved balanced performance (AUC: 87.93%, AP: 18.85%, F1: 15.66%, Brier score: 0.03). Machine learning models also identified several new important predictors, including birth season, depression, and knee height. An ensemble machine learning model incorporating techniques to manage data imbalance can accurately predict HBV infection in Chinese middle-aged and older adults. The novel predictors we identified offer new insights into HBV risk in ageing populations and support the development of more targeted prevention strategies.

BackgroundSarcopenia is a common muscle disorder in older adults, and its early identification and management in middle-aged populations are essential for ensuring a healthier later life. Detecting sarcopenia at an earlier stage may reduce the future burden on health care systems and enhance the quality of life in older adults. Machine learning (ML) models can evaluate large datasets, identify essential variables, and find complicated correlations between input variables. However, using ML models to detect sarcopenia remains an unsatisfied need.ObjectiveThis study aimed to develop and externally validate an ML model to predict sarcopenia risk among middle-aged adults using a nationally representative dataset.MethodsWe analyzed data from 1926 participants aged 40 to 64 years and enrolled in the 2022 Korea National Health and Nutrition Examination Survey (KNHANES). Sarcopenia was diagnosed and defined based on the 2019 Asian Working Group for Sarcopenia criteria, which incorporate both low muscle mass and reduced muscle strength. Muscle mass was assessed using bioelectrical impedance analysis with cutoffs of <7.0 kg/m² for men and <5.7 kg/m² for women. Muscle strength was measured via handgrip strength using a digital dynamometer with thresholds of <28 kg for men and <18 kg for women. Participants meeting both criteria were classified as those with sarcopenia. Four ML algorithms, random forest, support vector machine, extreme gradient boosting, and logistic regression, were used to identify risk factors of sarcopenia and predict its likelihood. The top-performing model was subsequently validated in an external cohort of 2247 middle-aged adults from the 2023 KNHANES. Model performance was assessed using the F2-score, area under the curve of a receiver operating characteristic curve, and sensitivity. All analyses were conducted using Python 3.13.2 (Python Software Foundation).ResultsAmong the 4 models, the logistic regression model demonstrated the strongest performance, yielding an area under the curve of 0.85, a sensitivity of 0.92, and an F2-score of 0.66. External validation using the 2023 KNHANES dataset confirmed the model’s robust performance, indicating its potential for widespread applications.ConclusionsThis study developed and externally validated an ML model that accurately identified sarcopenia in middle-aged adults. Leveraging data from a comprehensive national survey, our findings underscore the significance of early detection and customized interventions in midlife to mitigate sarcopenia risk and optimize long-term health outcomes.

Subjective life expectancy subjectively refers to an individual's expectation of the possibility of survival in the future, and is not a simple subjective evaluation, but an important health indicator that allows people to make decisions, plan, and control their own behavior in relation to health promotion. This study analyzed data from the Korean Longitudinal Study of Aging (KLoSA) to determine whether frailty and oral health affect the subjective life expectancy of patients with arthritis aged 55 or older. In order to identify factors influencing the participants' subjective life expectancy, multiple linear regression was used to analyze the results. In the final model the subjective life expectancy decreased as age increase. Subjective life expectancy was higher in the case of having a spouse, high income, working, the better the subjective health status. Compared to healty participants, the subjective life expectancy was lower in the pre-frailty and the frailty. The higher the oral health, the higher the subjective life expectancy. Based on the results of this study, if the oral health is maintained positively and subjective life expectancy is increased by providing interventions to prevent frailty, arthritis patients will live healthier lives with high expectations regarding subjective life expectancy in a positive direction.

Comparative study of different machine learning models in landslide susceptibility assessment: A case study of Conghua District, Guangzhou, China

Cardiometabolic multimorbidity (CMM) is a prevalent syndrome among middle-aged and older adults, significantly impairing quality of life and imposing substantial health and economic burdens on China's aging healthcare system. The development of timely predictive models is crucial for enabling early intervention. This study aimed to integrate multidimensional data from the China Health and Retirement Longitudinal Study (CHARLS) to develop an effective predictive model for assessing the five-year risk of CMM onset, thereby facilitating early intervention and management for individuals at risk in China. We analyzed data from the 2015 to 2020 CHARLS surveys, involving 5,388 middle-aged and older adults initially free of CMM. The dataset was randomly split into a training set (70%) and a validation set (30%). Key predictors were identified from 31 potential variables using LASSO regression with 10-fold cross-validation. Selected predictors underwent correlation analysis and were used to construct both Extreme Gradient Boosting (XGBoost) and Logistic Regression (LR) models. Variable contributions in the XGBoost model were interpreted using SHapley Additive exPlanations (SHAP) values, while the LR model was visualized via a nomogram. The performance of the superior model was evaluated using receiver operating characteristic curves, calibration curves, and decision curve analysis. LASSO regression screened significant variables from the initial 31 candidates. Correlation analysis ultimately identified nine key predictors: systolic blood pressure, BMI, fasting blood glucose, total cholesterol, triglycerides, uric acid, age, comorbidities, and pain. The logistic regression model demonstrated superior and more stable predictive performance on the validation set, with an area under the curve (AUC) of 0.732 (95% CI: 0.703-0.761). Calibration curves indicated reliable predictive accuracy, and decision curve analysis established the clinical net benefit across a range of risk thresholds. This study developed and validated a reliable, clinically applicable logistic regression model, supplemented by a nomogram, for assessing the five-year risk of CMM in Chinese middle-aged and older adults. The model effectively identifies high-risk individuals, supporting targeted early intervention and management strategies to alleviate the health and economic burdens associated with CMM in this population.

Uric acid to high-density lipoprotein cholesterol ratio (UHR) is considered a novel marker of inflammation. The aim of this study was to investigate the association between the UHR and chronic kidney disease (CKD) in middle-aged and older adults. This study used data from the National Health and Nutrition Examination Survey (NHANES) and the China Health and Retirement Longitudinal Study (CHARLS). The association between UHR and CKD prevalence was analyzed using weighted multivariable logistic regression models. Restricted cubic spline (RCS) models were used to analyze nonlinear relationships. Weighted subgroup analyses were performed to validate the robustness of the findings. The diagnostic performance of UHR for CKD was evaluated using receiver operating characteristic (ROC) curves. This study included 10,968 participants from NHANES and 9,012 participants from CHARLS. The mean age of NHANES participants was 59.60 years (standard deviation [SD] = 10.97), with 5,575 men (50.83%). CHARLS participants had a mean age of 59.58 years (SD = 9.40), and 4,203 (46.64%) were men. Weighted logistic regression analysis revealed a positive association between UHR and CKD prevalence (CHARLS: OR = 1.83, 95% CI = 1.44–2.32, p < 0.001; NHANES: OR = 1.81, 95% CI = 1.58–2.07, p < 0.001). The RCS regression models demonstrated a significant nonlinear relationship between UHR and CKD. ROC analysis indicated that UHR demonstrated moderate CKD discrimination (CHARLS: AUC = 0.707; NHANES: AUC = 0.645). UHR is positively associated with an increased prevalence of CKD in middle-aged and older adults, and this association exhibits cross-ethnic consistency.

To analyze the impact of the social environment of a community on the self-rated health of middle-aged and older adults living in the community, and to provide a basis for building favorable community social environment that promotes the health of middle-aged and older adults. Using the community and individual survey data from China Health and Retirement Longitudinal Study (CHARLS) done in 2011, 7822 middle-aged and older adults living in 397 communities were selected as the subjects of the study. Community social environment under study included community socioeconomic status, community social security and community services. Self-rated health was measured with a five-level score. A three-level hierarchical generalized linear model (HGLM) was used to analyze the impact of community social environment on the self-rated health of middle-aged and older adults. The proportion of subjects who reported self-rated poor health was 28.8%. According to the results of the three-level HGLM, the self-rated health of middle-aged and older adults showed differences at both the city and the community levels, and the variance reduction ratio reached 35.71% at the community level. The self-rated health status of residents in the communities with medium and high net per capita income was better than that in the communities with low net per capita income ( OR=0.84, P<0.05; OR=0.70, P<0.05). The self-rated health status of residents in the communities with secondary and higher education levels was better than that in the communities with low education level ( OR=0.78, P<0.05; OR=0.73, P<0.05). The self-rated health of middle-aged and older adults in the communities with unemployment subsidies was much better ( OR=0.77, P<0.05). The self-rated health status of middle-aged and older adults with chronic diseases was worse ( OR=4.46, P<0.05). Middle-aged and older adults living in cities had better self-rated health status ( OR=0.67, P<0.05). Community socioeconomic status and community unemployment subsidy are positively correlated with the self-rated health of middle-aged and older adults living in the community. It is necessary to pay special attention to the health status of middle-aged and older adults in communities of low socioeconomic status, improve unemployment insurance policies, and promote the health of middle-aged and older adults.

Sarcopenia has emerged as a significant public health issue that threatens the health of older adults. The association between sleep duration and sarcopenia is receiving increasing attention. Based on data from the China Health and Retirement Longitudinal Study (CHARLS), this study investigates the dynamic characteristics of sleep duration trajectories and their relationship with the onset of sarcopenia. A retrospective cohort study. The study used data from the CHARLS across various regions of China. Data were drawn from the 2011, 2013 and 2015 waves of CHARLS, including 2521 participants aged 45 years or older at baseline who had complete 3-year records of sarcopenia diagnosis. Sarcopenia was defined according to the 2019 criteria established by the Asian Working Group for Sarcopenia and was assessed across three dimensions: grip strength, estimated appendicular skeletal muscle mass and physical function. Participants were categorised by changes in sleep duration, and linear mixed-effects models were used to identify three trajectories of sleep duration change (decrease-increase, decrease-no recovery and continuous increase). Multivariate logistic regression was conducted to examine the association between these trajectories and the risk of sarcopenia, adjusting for confounders, such as education level, body mass index (BMI) and lifestyle factors. The median age of the cohort was 57.00 years, 61.40% of participants were female and the median BMI was 23.44. During the 5-year follow-up, the prevalence of sarcopenia was 14.09%. Compared with the non-sarcopenia group, individuals with sarcopenia had a higher median age (60.00 vs 56.00 years, p<0.001), lower BMI (23.06 vs 23.44, p=0.020), a higher proportion of unmarried/divorced individuals (16.34% vs 11.5%, p=0.010) and lower education levels (p<0.001). After adjusting for gender, age, BMI, residential area, education level, marital status, health insurance status, smoking and alcohol consumption, both increased and decreased sleep durations were associated with higher odds of sarcopenia relative to unchanged sleep duration, though the associations were not statistically significant (increased sleep duration: OR=1.081, 95% CI 0.807 to 1.449, p=0.601 and decreased sleep duration: OR=1.225, 95% CI 0.928 to 1.618, p=0.153). These results remained non-significant after further multivariate adjustment. Analysis of the three sleep duration trajectories identified via the linear mixed-effects model showed that, after multivariate adjustment and in comparison with trajectory 1 (sleep duration decreased and then increased significantly, exceeding baseline levels), both trajectory 2 (sleep duration decreased and then slightly increased but remained below baseline) and trajectory 3 (a consistent increase in sleep duration) were associated with a potential reduction in sarcopenia risk. However, these associations were not statistically significant (trajectory 2: OR=0.833, 95% CI 0.620 to 1.120, p=0.226 and trajectory 3: OR=0.844, 95% CI 0.647 to 1.101, p=0.210). This study developed a novel sleep trajectory model using longitudinal data, revealing the non-linear, multidimensional nature of the relationship between sleep duration and sarcopenia risk. Although no independent association was identified between 5-year sleep duration trajectories and sarcopenia in Chinese middle-aged and older adults, the study provides a valuable methodological framework for exploring multifactorial interactions in the ageing process and establishes a foundation for future research.

The primary goal was to assess the connection between sarcopenic obesity (SO) and progression to advanced cardiovascular-kidney-metabolic (CKM) syndrome in middle-aged and older Chinese adults. Our analysis utilized data collected during the 2011 and 2015 of the China Health and Retirement Longitudinal Study (CHARLS). CKM stages were identified based on the AHA criteria adapted for CHARLS. SO was defined with the concurrence of possible sarcopenia and obesity. The relationship between SO and advanced CKM stages was assessed using logistic regression models. In cross-sectional analysis of 8,448 participants, after accounting for potential confounders, individuals with SO showed significantly higher odds of advanced CKM syndrome (OR 1.51, 95% CI 1.21-1.88), compared to the normal group. Over an average of 4-year follow-up, 829 of 4,454 individuals with early CKM stages progressed to advanced CKM stages. The fully adjusted model revealed that participants with SO showed a substantially higher risk of CKM syndrome progression compared to normal group (OR: 1.81; 95% CI: 1.34-2.45). In subgroup analyses, SO maintained a positive correlation with advanced CKM stages, especially among men, older adults, smokers, alcohol consumers, and diabetic patients. Sex and age may act as modifiers in the relationship between SO and advanced CKM syndrome (P for interaction, 0.03 and < 0.001, respectively). Our cross-sectional and longitudinal analyses of CHARLS data underscore the significant association between SO and the risk of progression to advanced CKM syndrome. Incorporating SO assessments may facilitate earlier identification and prevention of CKM syndrome progression. The online version contains supplementary material available at 10.1007/s40200-026-01886-8.

Background: China’s rapidly aging population presents challenges for cognitive health and mental well-being among the older adults. This study examines how the number of children affects cognitive function in middle-aged and older adults and whether depressive symptoms mediate this relationship. Methods: This study analyzed data from waves 1 to 5 (2011–2020) of the China Health and Retirement Longitudinal Study (CHARLS), involving 5932 participants aged 45 and older. Participants were grouped by the number of children: childless, only child and multiple children. We used Logarithmic Generalized Linear Models (LGLMs) to explore the relationships among the number of children, depressive symptoms, and cognitive function. Indirect effect coefficients and 95% bias-corrected and accelerated confidence intervals (BCaCI) were estimated using Simultaneous Equation Models (SEM) with three-stage least squares (3SLS) and the bootstrap method to assess the mediating effect of depressive symptoms. Results: In middle-aged and older adults, a negative association was observed between the number of children and overall cognitive functioning (all p < 0.01). This association remained significant even after adjusting for covariates in groups with three (β = −0.023, p < 0.05) and four or more children (β = −0.043, p < 0.001). Conversely, the positive association between the number of children and depression also persisted after adjusting for covariates, although it weakened as the number of children increased (all p < 0.01). Depressive symptoms consistently correlated negatively with overall cognitive function (p < 0.001) and partially mediated the relationship between the number of children and cognitive function (pMe = 20.36%, p < 0.05). The proportion of the mediating effect attributed to depression was more pronounced in middle-aged and older adults who had experienced the loss of children (pMe = 24.31%) or had two children (pMe = 25.39%), with stronger mediating effects observed in males (pMe = 48.84%) and urban residents (pMe = 64.58%). Conclusions: The findings indicate that depressive symptoms partially mediate the relationship between the number of children and cognitive function in middle-aged and older adults in China. These results highlight the significance of considering mental health factors when studying cognitive function in this demographic. Notably, in families without children and those with two children, depressive symptoms play a crucial role in explaining the decline in cognitive function.

To analyze the current status of social support for middle-aged and older adults with multimorbidity and to explore the correlation between different dimensions of social support and multimorbidity and the related outcomes on the basis of China Health and Retirement Longitudinal Study (CHARLS) 2015 survey data so as to reveal the complex social background of multimorbidity and the impact of social support on multimorbidity. A total of 9168 valid samples, with an average age of 59.60 years, were included in the study. Using the social support-related variables of the respondents, we conducted factor analysis and constructed regression models of common factors of social support and multimorbidity-related outcomes, intending to analyze the impact of common factors of social support on multimorbidity in the middle-aged and older adults. The multimorbidity of middle-aged and older adults in China was related to multiple factors of social support, and the differences were statistically significant. Logistic regression showed that social support in the form of activity/recreational facilities and medical resources was a protective factor of multimorbidity, that family emotional support and economic support had a positive effect on life satisfaction of comorbid patients, and that social support in the form of education, social life and housing conditions was negatively correlated with catastrophic medical expenditure of the comorbid population ( P<0.05). Social support for middle-aged and older adults in China is unevenly distributed. Social support in the form of activity/recreational facilities and medical resources may reduce the risks of multimorbidity among middle-aged and older adults. Good family economic and emotional support can improve the life satisfaction of middle-aged and older adults with multimorbidity. Social support in the form of education, social life and housing conditions may reduce the risk of catastrophic medical expenditure in middle-aged and older adults with multimorbidity.

BackgroundCognitive impairment (CI), an early sign of Alzheimer's disease, may be identified using the metabolic score for insulin resistance (METS-IR), a novel insulin resistance marker.ObjectiveThis nationwide cohort study in China examined the association between METS-IR and CI in middle-aged and older adults, providing preliminary evidence for early prevention.MethodsData were obtained from the China Health and Retirement Longitudinal Study (CHARLS) 2011-2015. METS-IR was calculated using fasting plasma glucose, triglycerides, body mass index, and high-density lipoprotein cholesterol. CI was evaluated through episodic memory and mental status. Descriptive statistics, multivariable logistic regression, restricted cubic spline (RCS), subgroup, and interaction analyses were used to assess the association of METS-IR with CI, with sensitivity analysis confirming robustness.ResultsThe study included 5842 middle-aged and older adults in China. The incidence of CI throughout the follow-up period was 14.36%. The multivariable logistic regression models indicated that participants in Q3 and Q4 of METS-IR had 1.375-fold and 1.564-fold higher risks of CI, respectively, compared with those in Q1 (Q3: OR = 1.375, 95% CI = 1.075-1.758; Q4: OR = 1.564, 95% CI = 1.220-2.005). The trend analysis of METS-IR was also statistically significant. A substantial linear relation between METS-IR and CI was confirmed by the RCS analysis (p for nonlinearity = 0.462). Moreover, consistent associations were observed across extensive subgroup analyses.ConclusionsElevated METS-IR levels were significantly associated with increased CI risk in middle-aged and older Chinese adults. These findings provide support for early identification and intervention of CI.

The world population is rapidly aging, and depression mainly affects middle-aged and older adults with chronic diseases and cognitive impairments. The sample for this study was obtained from the China Health and Retirement Longitudinal Study (CHARLS) public database. The sample size for inclusion was 12,767. There were 6,773 females and 5,994 males, with an overall low level of education. This study aims to provide a theoretical and practical reference basis for the clinical non-pharmacological treatment of depression in middle-aged and older adults (age ≥ 50 years) with chronic diseases. Additionally, the study seeks to promote the development of mental health interventions for middle-aged and older adults (age ≥ 50 years) with chronic diseases, ultimately enhancing the sense of well-being and quality of life for this demographic. Cognitive functioning and depressive symptoms of the study participants were assessed using the Mini-Mental State Examination Scale (MMSE) and the short version of the Center for Epidemiological Studies Depression Scale (CESD-10). Binary logistic regression results showed that among middle-aged and older adults (age ≥ 50 years) with chronic diseases, participation in physical activity [OR = 1.397; 95% CI (1.181-1.651); p < 0.05] was more effective than participation in social activities [OR = 0.997; 95% CI (0.924-1.076); p < 0.05] for preventing depression. Those with cognitive impairment [OR = 1.206; 95% CI (1.089-1.335); p < 0.05] were more likely to experience depression than those without cognitive impairment. Activity participation (physical activity and social activity) had a more significant effect on mild and moderate depression compared to no depression, and cognitive level had a more pronounced effect on moderate depression [OR = 1.491; 95% CI (1.278-1.740); p < 0.05] and major depression [OR = 2.231; 95% CI (1.282-3.884); p < 0.05]. Within the specific cohort of middle-aged and older adults (age ≥ 50 years) with chronic diseases, both activity participation and cognitive level exert a significant influence on the prevention and intervention of depression. Engagement in physical activity, participation in social activities, and enhanced cognitive functioning emerged as protective factors against depression. Therefore, the policy-maker should strengthen the prevention and treatment of depression in a comprehensive manner through the promotion of physical and social activities and the enhancement of cognitive level, so as to safeguard the mental health of middle-aged and older adults with chronic diseases.

BackgroundPopulation aging is a major global trend with significant social, economic, and health implications. In China, the increasing aging population presents challenges, including increased chronic diseases and disabilities. Social capital has emerged as vital in determining health outcomes for middle-aged and older adults. This study seeks to examine the impact of social capital and educational attainment on physical functioning in middle-aged and older adults, with particular emphasis on the moderating effect of education within this relationship.MethodsThis study utilized data from the China Health and Retirement Longitudinal Study (CHARLS) from 2018 to 2020, involving 9,497 participants aged 45 and older. Physical function was assessed using the Activities of Daily Living (ADL) and Instrumental Activities of Daily Living (IADL) scales. Social capital was measured in four dimensions: social trust, social support, social participation, and reciprocity. Educational attainment was categorized into four levels: below primary school, primary school, middle school, and high school or above. Cox proportional hazards regression and moderating effect models were used for data analysis, adjusting for demographic and health-related variables.ResultsOur findings highlight the significant roles of social participation (aHR = 0.856, 95%CI: 0.675–0.809) and reciprocity (aHR = 0.700, 95%CI: 0.626–0.784) in improving physical function. Education enhanced the positive effects of social participation (aHR = 0.923, 95%CI: 0.840–0.980). Subgroup analyses showed that social support was a protective factor for females (aHR = 0.857, 95% CI: 0.737–0.998), while social trust negatively affected urban residents (aHR = 1.330, 95%CI: 1.135–1.560).ConclusionThe findings underscore the importance of social participation, reciprocity, and education in enhancing physical function among middle-aged and older adults. Tailored interventions addressing gender and residential differences are essential to meet the unique needs of various subgroups. Understanding the relationship between social capital, education, and health can inform strategies to improve this population’s well-being.

BackgroundAdverse childhood experiences (ACEs) are associated with an increased risk of depressive symptoms (DS) in older adults. This study investigated the role of social activity in mediating the relationship between ACEs and DS among middle-aged and older Chinese adults.MethodsData were derived from the China Health and Retirement Longitudinal Study (CHARLS). The discovery dataset included 10,164 participants from 2018, matched with life history data from 2014, while the replication dataset comprised 8,899 participants from 2020. DS was measured using the Center for Epidemiologic Studies Depression Scale (CES-D). Multiple linear regression and mediation analysis were conducted.ResultsIn the discovery dataset, ACEs were positively correlated with DS (r = 0.17, p < 0.001), while social activity was negatively correlated with both DS (r = –0.11, p < 0.001) and ACEs (r = –0.03, p = 0.01). Mediation analysis indicated that ACEs significantly predicted DS (estimate = 0.51, 95% CI 0.43 to 0.60), and social activity partially mediated this relationship (estimate = –0.01, bootstrap 95% CI –0.01 to –0.001), particularly among middle-aged adults (indirect effect estimate = –0.01, bootstrap 95% CI –0.01 to –0.001). Additionally, social activity notably mediated the relationship between childhood violence exposure and DS (estimate = –0.02, bootstrap 95% CI –0.04 to –0.003). These results were robustly validated through replication analysis, reinforcing the reliability of our conclusions.ConclusionsSocial activity mediates the relationship between ACEs and DS, highlighting the importance of social engagement to reduce depression risk in this population.