Predicting depression among rural and urban disabled elderly in China using a random forest classifier

Machine learning judged neutral facial expressions as key factors for a "good therapist" within the first five minutes: An experiment to simulate online video counselling

Objective

Machine learning models were employed to discern patients' impressions from the therapists' facial expressions during a virtual online video counselling session.

Methods

Eight therapists simulated an online video counselling session for the same patient. The facial emotions of the therapists were extracted from the session videos; we then utilized a random forest model to determine the therapist's impression as perceived by the patients.

Results

The therapists' neutral facial expressions were important controlling factors for patients' impressions. A predictive model with three neutral facial features achieved an accuracy of 83% in identifying patients' impressions.

Conclusions

Neutral facial expressions may contribute to patient impressions in an online video counselling environment with spatiotemporal disconnection.

Innovation

Expression recognition techniques were applied innovatively to an online counselling setting where therapists' expressions are limited. Our findings have the potential to enhance psychiatric clinical practice using Information and Communication Technology.

Using random forest to identify correlates of depression symptoms among adolescents

PURPOSE

Adolescent depression is a significant public health concern, and studying its multifaceted factors using traditional methods possess challenges. This study employs random forest (RF) algorithms to determine factors predicting adolescent depression scores.

METHODS

This study utilized self-reported survey data from 56,008 Canadian students (grades 7-12) attending 182 schools during the 2021/22 academic year. RF algorithms were applied to identify the correlates of (i) depression scores (CESD-R-10) and (ii) presence of clinically relevant depression (CESD-R-10 ≥ 10).

RESULTS

RF achieved a 71% explained variance, accurately predicting depression scores within a 3.40 unit margin. The top 10 correlates identified by RF included other measures of mental health (anxiety symptoms, flourishing, emotional dysregulation), home life (excessive parental expectations, happy home life, ability to talk to family), school connectedness, sleep duration, and gender. In predicting clinically relevant depression, the algorithm showed 84% accuracy, 0.89 sensitivity, and 0.79 AUROC, aligning closely with the correlates identified for depression score.

CONCLUSION

This study highlights RF's utility in identifying important correlates of adolescent depressive symptoms. RF's natural hierarchy offers an advantage over traditional methods. The findings underscore the importance and additional potential of sleep health promotion and school belonging initiatives in preventing adolescent depression.

Ensemble of hybrid model based technique for early detecting of depression based on SVM and neural networks

The prevalence of depression has increased dramatically over the last several decades: it is frequently overlooked and can have a significant impact on both physical and mental health. Therefore, it is crucial to develop an automated detection system that can instantly identify whether a person is depressed. Currently, machine learning (ML) and artificial neural networks (ANNs) are among the most promising approaches for developing automated computer-based systems to predict several mental health issues, such as depression. This study propose an ensemble of hybrid model-based techniques that aims to build a strong detection model that considers many psychological and sociodemographic characteristics of an individual to detect whether a person is depressed. Support vector machines (SVM) and multilayer perceptrons (MLP) are the two fundamental methods used to construct the suggested ensemble approach. The hybrid DeprMVM served as a meta-learner. In this study, the hybrid DeprMVM is a level-1 learner, whereas the SVM and MLP networks are level-0 learners. After the classifiers are trained and tested at level 0, their outputs are based on both the independent and dependent variables in the new data set that was used to train the meta-classifier. The training data class imbalance was reduced by applying the synthetic minority oversampling technique (SMOTE) and cluster sampling together, which improved the accuracy for detecting depression. Additionally, it can effectively reduce the risk of over-fitting from simply duplicating data points. To further confirm the effectiveness of the proposed method, various performance evaluation metrics were calculated and compared with previous studies conducted on this specific dataset. In conclusion, among all the techniques for identifying depression, the suggested ensemble approach had the best accuracy, at 99.39%, and an F1-score of 99.51%.

Age, period, cohort effects in trends of depressive symptoms among middle-aged and older Chinese adults

Objectives

To investigate the effects of age, period, and cohort on the trends of depression; and to examine the influence of these three temporal effects on residential disparities in depression.

Methods

Using data from the China Health and Retirement Longitudinal Study (CHARLS) during 2011 to 2020, involving 77,703 respondents aged 45 years old and above. The measurement of depressive symptoms was the score of 10-question version of the Center for Epidemiologic Studies Depression Scale (CES-D 10). The hierarchical age-period-cohort cross-classified random effects models were conducted to examine trends in depressive symptoms related to age, period and cohort.

Results

CES-D scores increased with age and slightly decreased at older age. The cohort trends mostly increased except for a downward trend among those born in 1950s. As for the period effect, CES-D scores decreased gradually from 2011 to 2013 followed by a upward trend. Rural residents were associated with higher level of depression than those live in urban area. These residence gaps in depression enlarged before the age of 80, and then narrowed. The urban-rural disparities in CES-D scores gradually diminished across cohorts, while the corresponding period-based change in urban-rural gaps was not significant.

Conclusion

When age, period, cohort factors are considered, the age effects on depression dominated, and the period and cohort variations were relatively small. The residence disparities in depression reduced with successive cohorts, more attention should be paid to the worsening depression condition of younger cohorts in urban areas.

A machine-learning approach to model risk and protective factors of vulnerability to depression

There are multiple risk and protective factors for depression. The association between these factors with vulnerability to depression is unclear. Such knowledge is an important insight into assessing risk for developing depression for precision interventions. Based on the behavioral data of 496 participants (all unmarried and not cohabiting, with a college education level or above), we applied machine-learning approaches to model risk and protective factors in estimating depression and its symptoms. Then, we employed Random Forest to identify important factors which were then used to differentiate participants who had high risk of depression from those who had low risk. Results revealed that risk and protective factors could significantly estimate depression and depressive symptoms. Feature selection revealed four key factors including three risk factors (brooding, perceived loneliness, and perceived stress) and one protective factor (resilience). The classification model built by the four factors achieved an ROC-AUC score of 75.50% to classify the high- and low-risk groups, which was comparable to the classification performance based on all risk and protective factors (ROC-AUC = 77.83%). Based on the selected four factors, we generated a mood vulnerability index useful for identifying people's risk for depression. Our findings provide potential clinical insights for developing quick screening tools for mood disorders and potential targets for intervention programs designed to improve depressive symptoms.

Mitochondrial genome plasticity of mammalian species

There is an ongoing process in which mitochondrial sequences are being integrated into the nuclear genome. The importance of these sequences has already been revealed in cancer biology, forensic, phylogenetic studies and in the evolution of the eukaryotic genetic information. Human and numerous model organisms' genomes were described from those sequences point of view. Furthermore, recent studies were published on the patterns of these nuclear localised mitochondrial sequences in different taxa.However, the results of the previously released studies are difficult to compare due to the lack of standardised methods and/or using few numbers of genomes. Therefore, in this paper our primary goal is to establish a uniform mining pipeline to explore these nuclear localised mitochondrial sequences.Our results show that the frequency of several repetitive elements is higher in the flanking regions of these sequences than expected. A machine learning model reveals that the flanking regions' repetitive elements and different structural characteristics are highly influential during the integration process.In this paper, we introduce a general mining pipeline for all mammalian genomes. The workflow is publicly available and is believed to serve as a validated baseline for future research in this field. We confirm the widespread opinion, on - as to our current knowledge - the largest dataset, that structural circumstances and events corresponding to repetitive elements are highly significant. An accurate model has also been trained to predict these sequences and their corresponding flanking regions.

Evaluating Machine Learning Stability in Predicting Depression and Anxiety Amidst Subjective Response Errors

Major Depressive Disorder (MDD) and Generalized Anxiety Disorder (GAD) pose significant burdens on individuals and society, necessitating accurate prediction methods. Machine learning (ML) algorithms utilizing electronic health records and survey data offer promising tools for forecasting these conditions. However, potential bias and inaccuracies inherent in subjective survey responses can undermine the precision of such predictions. This research investigates the reliability of five prominent ML algorithms-a Convolutional Neural Network (CNN), Random Forest, XGBoost, Logistic Regression, and Naive Bayes-in predicting MDD and GAD. A dataset rich in biomedical, demographic, and self-reported survey information is used to assess the algorithms' performance under different levels of subjective response inaccuracies. These inaccuracies simulate scenarios with potential memory recall bias and subjective interpretations. While all algorithms demonstrate commendable accuracy with high-quality survey data, their performance diverges significantly when encountering erroneous or biased responses. Notably, the CNN exhibits superior resilience in this context, maintaining performance and even achieving enhanced accuracy, Cohen's kappa score, and positive precision for both MDD and GAD. This highlights the CNN's superior ability to handle data unreliability, making it a potentially advantageous choice for predicting mental health conditions based on self-reported data. These findings underscore the critical importance of algorithmic resilience in mental health prediction, particularly when relying on subjective data. They emphasize the need for careful algorithm selection in such contexts, with the CNN emerging as a promising candidate due to its robustness and improved performance under data uncertainties.

The impact of long-term care insurance pilot on the mental health of older adults: Quasi-experimental evidence from China

The Chinese government launched pilot programs for a long-term care insurance system in response to the ongoing increase in the aging population. This study uses the difference-in-differences (DID) model to analyze the impact of long-term care insurance on older adults' mental health based on China Health and Retirement Longitudinal Study (CHARLS) four-period panel data from 2011 to 2018. This study found that long-term care insurance reduced Center for Epidemiological Studies Depression Scale (CES-D) scores among older adults by 1.059 points. Moreover, there was an improvement of 0.181 and 0.870 points in mental status and scenario memory scores, respectively. The impact of the long-term care insurance pilot program on improving the mental health of older adults was more pronounced, especially for those with chronic diseases or disabilities as well as those living in rural and western regions. This study also revealed that long-term care insurance enhances mental health by reducing medical expenses and increasing daily companionship and social interaction. Therefore, a pilot study of long-term care insurance showed a significant improvement in the mental health of older adults. To provide a comprehensive care service system for older adults, the government should expand the scope of the pilot program and increase the accessibility of mental health services for older adults.

Study on the trajectory of depression among middle-aged and elderly disabled people in China: Based on group-based trajectory model

Background

Previous studies have shown that middle-aged and elderly adults with disabilities are at higher risk for depressive symptoms. However, there are few studies on the long-term trajectories of depressive symptoms in the Chinese middle-aged and elderly disabled population.

Objective

This study aimed to identify the different development trajectories of depressive symptoms and their influencing factors in middle-aged and elderly people with disabilities in China.

Methods

Using data from the China Health and Retirement Longitudinal Study (CHARLS) in 2011, 2013, 2015, and 2018, a longitudinal cohort was formed for the study. A total of 2053 participants underwent at least two measures of depressive symptoms, assessed using the Center for Epidemiological Studies Depression Scale (CES-D10), a depression symptom assessment scale. We constructed a Group-Based Trajectory Model (GBTM) to identify the development trajectory of depressive symptoms in 2053 middle-aged and elderly disabled individuals, screened the potential predictors using lasso regression, and analyzed the factors affecting the development trajectory of depression through multivariate logistic regression.

Results

We identified four depression symptom trajectories throughout the follow-up process: "low depressive symptom group", "worsening depressive symptom group", "relieved depressive symptom group", and "high depressive symptom group". We found that there were differences in basic characteristics among different subgroups of depression trajectory. However, middle-aged and elderly disabled women living in rural areas, with limited ADL or IADL, physical pain, poor self-reported health and self-reported memory, short sleep time, and no relatives and friends to take care of them were the key groups for the prevention and treatment of depressive symptoms.

Conclusion

There is heterogeneity in the trajectories of depressive symptoms in the Chinese middle-aged and elderly disabled population, it is necessary to focus on the characteristics of the trajectories of different subgroups.

Social determinants of health and gender differences in depression among adults: A cohort study

The role of social determinants of health (SDoH) in gender differences in depression remains unclear among Chinese adults. We aimed to explore the association between SDoH and depression and investigate their role in explaining gender differences in depression. This prospective longitudinal cohort study used four wave surveys (2012, 2016, 2018, and 2020) of the China Family Panel Study (CFPS). Fourteen SDoH variables were assessed, and depression was measured using the 8-item short version of the Center for Epidemiologic Studies Depression Scale. The Cox proportional hazards regression and multiple mediation analysis were performed to estimate the effect sizes. The longitudinal sample included 18,874 participants aged 18-92 years (51.4 % males and 48.6 % females). Women had higher risk of depression than men. Unfavorable SDoH were associated with higher risk of depression. After including multiple SDoH in mediation analysis, multiple SDoH mediated 15.7 % of the total effect of gender on depression. In sum, SDoH significantly influenced depression, and specific factors explained gender differences in depression. Supporting women in education, employment, and community involvement could help reduce gender differences in depression.

Urban-rural disparities in the prevalence and trends of depressive symptoms among Chinese elderly and their associated factors

BACKGROUND

This study aimed to examine urban-rural disparities in the prevalence and trends of depressive symptoms (DS) among Chinese elderly and associated factors.

METHODS

A total of 8025, 7808, and 4887 respondents aged 60 years and above were selected from the China Family Panel Studies (CFPS) in 2016, 2018, and 2020, respectively. DS was assessed using a short version of Center for Epidemiologic Studies Depression Scale (CES-D). Twenty-two associated factors from six categories were included in random forest (RF) models. All urban-rural comparisons were conducted based on good model performance.

RESULTS

The DS prevalence among all rural elderly was significantly higher than corresponding urban elderly. This disparity continued to widen among younger elderly, while it continued to narrow among older elderly. The top 10 common leading factors were sleep quality, self-rated health, life satisfaction, memory ability, child relationship, IADL disability, marital status, educational level, and gender. Urban-rural disparities in sleep quality, interpersonal trust, and child relationship continued to widen, while disparities in multimorbidity, hospitalization status, and frequency of family dinner continued to narrow.

LIMITATION

This study may exist recall bias and lacks causal explanation.

CONCLUSIONS

Significant and continuing disparities in the DS prevalence were observed between urban and rural elderly in China, showing opposite trends in younger and older elderly. The top 10 leading associated factors for DS were nearly consistent across urban and rural elderly, with sleep quality having strongest contribution. Urban-rural disparities in associated factors also showed different trends. This study provides a reference for mental health promotion among Chinese elderly.

Predicting unmet activities of daily living needs among the oldest old with disabilities in China: a machine learning approach

Background

The ageing population in China has led to a significant increase in the number of older persons with disabilities. These individuals face substantial challenges in accessing adequate activities of daily living (ADL) assistance. Unmet ADL needs among this population can result in severe health consequences and strain an already burdened care system. This study aims to identify the factors influencing unmet ADL needs of the oldest old (those aged 80 and above) with disabilities using six machine learning methods.

Methods

Drawing from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) 2017-2018 data, we employed six machine learning methods to predict unmet ADL needs among the oldest old with disabilities. The predictive effects of various factors on unmet ADL needs were explored using Shapley Additive exPlanations (SHAP).

Results

The Random Forest model showed the highest prediction accuracy among the six machine learning methods tested. SHAP analysis based on the Random Forest model revealed that factors such as household registration, disability class, economic rank, self-rated health, caregiver willingness, perceived control, economic satisfaction, pension, educational attainment, financial support given to children, living arrangement, number of children, and primary caregiver played significant roles in the unmet ADL needs of the oldest old with disabilities.

Conclusion

Our study highlights the importance of socioeconomic factors (e.g., household registration and economic rank), health status (e.g., disability class and self-rated health), and caregiving relationship factors (e.g., caregiver willingness and perceived control) in reducing unmet ADL needs among the oldest old with disabilities in China. Government interventions aimed at bridging the urban-rural divide, targeting groups with deteriorating health status, and enhancing caregiver skills are essential for ensuring the well-being of this vulnerable population. These findings can inform policy decisions and interventions to better address the unmet ADL needs among the oldest old with disabilities.

A Fast and Minimal System to Identify Depression Using Smartphones: Explainable Machine Learning-Based Approach

BACKGROUND

Existing robust, pervasive device-based systems developed in recent years to detect depression require data collected over a long period and may not be effective in cases where early detection is crucial. Additionally, due to the requirement of running systems in the background for prolonged periods, existing systems can be resource inefficient. As a result, these systems can be infeasible in low-resource settings.

OBJECTIVE

Our main objective was to develop a minimalistic system to identify depression using data retrieved in the fastest possible time. Another objective was to explain the machine learning (ML) models that were best for identifying depression.

METHODS

We developed a fast tool that retrieves the past 7 days' app usage data in 1 second (mean 0.31, SD 1.10 seconds). A total of 100 students from Bangladesh participated in our study, and our tool collected their app usage data and responses to the Patient Health Questionnaire-9. To identify depressed and nondepressed students, we developed a diverse set of ML models: linear, tree-based, and neural network-based models. We selected important features using the stable approach, along with 3 main types of feature selection (FS) approaches: filter, wrapper, and embedded methods. We developed and validated the models using the nested cross-validation method. Additionally, we explained the best ML models through the Shapley additive explanations (SHAP) method.

RESULTS

Leveraging only the app usage data retrieved in 1 second, our light gradient boosting machine model used the important features selected by the stable FS approach and correctly identified 82.4% (n=42) of depressed students (precision=75%, F1-score=78.5%). Moreover, after comprehensive exploration, we presented a parsimonious stacking model where around 5 features selected by the all-relevant FS approach Boruta were used in each iteration of validation and showed a maximum precision of 77.4% (balanced accuracy=77.9%). Feature importance analysis suggested app usage behavioral markers containing diurnal usage patterns as being more important than aggregated data-based markers. In addition, a SHAP analysis of our best models presented behavioral markers that were related to depression. For instance, students who were not depressed spent more time on education apps on weekdays, whereas those who were depressed used a higher number of photo and video apps and also had a higher deviation in using photo and video apps over the morning, afternoon, evening, and night time periods of the weekend.

CONCLUSIONS

Due to our system's fast and minimalistic nature, it may make a worthwhile contribution to identifying depression in underdeveloped and developing regions. In addition, our detailed discussion about the implication of our findings can facilitate the development of less resource-intensive systems to better understand students who are depressed and take steps for intervention.

A qualitative study of the interaction experiences between family caregivers and community nurses for disabled elderly people at home

BACKGROUND

Family members are currently the main caregivers of the disabled elderly people at home. With declining health and increasing frailty, caregiving of disabled elderly people becomes a task of family caregivers in conjunction with community nurses. Interaction between family caregivers and community nurses can effectively improve the quality of home care for the disabled elderly people. This study aimed to investigate the interaction experiences between family caregivers and community nurses for disabled elderly people at home.

METHODS

This research was a study of qualitative descriptions based on semi-structured face-to-face interviews. This study was to purposefully select family caregivers of the disabled elderly and community nurses in Zhengzhou city, Henan Province and explore the interaction patterns between them. Directed content analysis method was used to generate qualitative codes and identify themes.

RESULTS

A total of 12 interviews were completed, including 7 family caregivers and 5 community nurses. Four themes were identified: 1) Information interaction; 2) Emotional interaction; 3) Practical interaction; 4) Factors that promote and hinder the interaction.

CONCLUSIONS

It was found that the interaction between family caregivers and community nurses was not optimistic. Lack of communication and collaboration between community nurses and caregivers. Providing a new perspective that we can develop and implement intervention to facilitate positive interactions, which will reduce the burden of family caregivers, bring the highest quality of care to older adults with disabilities and improve the quality of care for disabled elderly people.

TRIAL REGISTRATION

Registered in the Chinese Clinical Trial Registry on April 19, 2021, number ChiCTR2100045584.

Network analysis of anxiety and depression in the functionally impaired elderly

Background

Evidence from previous studies has confirmed that functionally impaired elderly individuals are susceptible to comorbid anxiety and depression. Network theory holds that the comorbidity emerges from interactions between anxiety and depression symptoms. This study aimed to investigate the fine-grained relationships among anxiety and depression symptoms in the functionally impaired elderly and identify central and bridge symptoms to provide potential targets for intervention of these two comorbid disorders.

Methods

A total of 325 functionally impaired elderly individuals from five communities in Xi'an, China, were recruited for our investigation. The GAD-7 and PHQ-9 were used to measure anxiety and depression, respectively. SPSS 22.0 software was used for descriptive statistics, and R 4.1.1 software was used for network model construction, expected influence (EI) evaluation and bridge expected influence (BEI) evaluation.

Results

In the network, there were 35 edges (indicating partial correlations between symptoms) across the communities of anxiety and depression, among which the strongest edge was A1 "Nervousness or anxiety"-D2 "Depressed or sad mood." A2 "Uncontrollable worry" and D2 "Depressed or sad mood" had the highest EI values in the network, while A6 "Irritable" and D7 "Concentration difficulties" had the highest BEI values of their respective community. In the flow network, the strongest direct edge of D9 "Thoughts of death" was with D6 "Feeling of worthlessness."

Conclusion

Complex fine-grained relationships exist between anxiety and depression in functionally impaired elderly individuals. "Uncontrollable worry," "depressed or sad mood," "irritable" and "concentration difficulties" are identified as the potential targets for intervention of anxiety and depression. Our study emphasizes the necessity of suicide prevention for functionally impaired elderly individuals, and the symptom "feeling of worthlessness" can be used as an effective target.

Homeownership, Depression, and Life Satisfaction in China: The Gender and Urban-Rural Disparities

This study examines how depression and life satisfaction are associated with assets in the form of homeownership in China and whether their relationships differ between men and women, and between urban and rural areas. While the psychological benefits of homeownership are well-documented, how gender makes a difference in this relationship remains unclear. Given the dynamic housing market conditions characterized by the urban-rural divide and the notable gender gap in psychological well-being, China can provide a relevant context to address this knowledge gap. A series of linear regression analyses based on the China Family Panel Studies (CFPS) data show that homeownership is positively associated with life satisfaction and negatively related to depression, and this relationship is driven by men. While the homeownership-life satisfaction relation does not differ between urban and rural areas, the negative association between homeownership and depression is seen only among rural residents. The gender difference could be explained by the salient role of the financial security obtained from homeownership, whereas the regional difference seems to be supported by the social comparison theory. This study contributes to the knowledge of how a biological determinant, i.e., gender, interacts with a social determinant, i.e., homeownership, to affect psychological well-being.