CCChun-Yuan Chen

Chun-Yuan Chen

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Chun-Yuan Chen has a total of 28 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Area Between Trajectories: Insights into Optimal Group Selection and Trajectory Heterogeneity in Group-Based Trajectory Modeling

    by Yi-Chen Hsiao, Chun-Yuan Chen, Mei-Fen Tang

    Group-based trajectory modeling (GBTM) is commonly used to identify longitudinal patterns in health outcomes among older adults, with determining the optimal number of groups being a crucial step. While statistically grounded criteria are primarily relied upon, clinical relevance is gradually emphasized in medicine to ensure that the identified trajectory heterogeneity appropriately reflects changes in a disease or symptom over time. However, such considerations are often judged through visual comparisons, without concrete approaches for their application. To address this, the area between trajectories (ABT) was introduced as insights for quantifying trajectory group differences. Using a simulated sleep quality dataset, GBTM was applied to build and compare models. Subsequently, ABT was demonstrated to show how it works, while also highlighting its limitations and potential applications.

  • Open Access

    Development of a prediction model for infant hospitalization and death using clinical features assessed by community health workers during routine postnatal home visits in Dhaka, Bangladesh

    by Alastair Fung, Marimuthu Sappani, Cole Heasley, Chun-Yuan Chen, Shaun K. Morris, Peter J. Gill, Diego G. Bassani, Davidson H. Hamer, Prakesh S. Shah, S. M. Abdul Gaffar, et al.

    Introduction: To improve upon the World Health Organization (WHO) 8 danger signs used to identify young infants (<2 months) requiring referral during community health worker (CHW) home visits, aggregative features (e.g., cumulative visits with fever) rather than visit-specific features (e.g., fever at a single visit), and a machine learning random forest model, may enhance predictive performance. Applying these approaches, we aimed to develop a prediction model for infant hospitalization and/or death using CHW-assessed clinical features during home visits in Dhaka, Bangladesh. Methods: We analyzed data from generally healthy infants prospectively enrolled at birth and assessed at 11 scheduled CHW visits from 3-60 days of age. To predict first hospitalization or death, we developed two models – time-varying Cox regression and random forest – using the same set of candidate predictors (45 clinical features of which 8 were WHO danger signs, and 12 additional covariates) with aggregative features incorporated. We evaluated discrimination (C-statistic) and calibration (calibration plots). Performance was compared to a time-varying Cox model using only WHO danger signs. Results: Among 1906 infants, 176 (9.2%) had an event (173 hospitalizations, 3 deaths). The best-performing Cox model (C-statistic=0.71; 95% CI 0.68-0.75) consisting of three baseline covariates (any perinatal/delivery complication, umbilical cord care, gestational age) and four visit-specific clinical features (nasal congestion, cough, jaundice, skin rash), and a Cox model with these four features plus WHO danger signs (C-statistic=0.70; 95% CI 0.67-0.74), demonstrated higher discrimination than WHO danger signs alone (C-statistic=0.56; 95% CI 0.54-0.60), with similar calibration. A random forest model (42 predictors) was well-calibrated with comparable discrimination (C-statistic=0.69; 95% CI 0.64-0.73). Conclusion: Aggregative features and random forest did not outperform a time-varying Cox model using baseline covariates and visit-specific features. Adding four features to WHO danger signs may improve predictive performance by capturing a broader spectrum of infant illnesses requiring hospitalization. What is already known on this topic: During community health worker (CHW) home visit assessments of young infants (<2 months), use of World Health Organization (WHO)-recommended danger signs to predict hospitalization and/or death may have limited sensitivity and may miss cases of severe illness requiring referral. Summarizing repeated assessments of clinical features during sequential home visits as aggregative predictors (e.g., cumulative visits with fever) rather than visit-specific predictors (e.g., fever at a single visit), and machine learning models such as random forest, have not been previously evaluated for prediction of infant hospitalization and/or death and may improve predictive performance compared to WHO danger signs. What this study adds: Random forest, and use of aggregative predictors in both a random forest model and a time-varying Cox model, did not improve prediction of infant hospitalization and/or death during CHW routine home visits compared to a time-varying Cox model consisting of baseline covariates and visit-specific clinical features. Adding four visit-specific clinical features to the WHO danger signs improved prediction of hospitalization and/or death during CHW routine home visit assessments of young infants born generally healthy in an urban setting. How this study might affect research, practice or policy: The findings support future research evaluating whether adding visit-specific clinical features to the WHO danger signs algorithm can improve identification of infants needing referral across diverse settings and with varying baseline risks.

  • Open Access

    Area between trajectories: Insights into optimal group selection and trajectory heterogeneity in group-based trajectory modeling

    by Yi-Chen Hsiao, Chun-Yuan Chen, Mei-Fen Tang

    Group-based trajectory modeling (GBTM) is commonly used to identify longitudinal patterns in health outcomes among older adults, with determining the optimal number of groups being a crucial step. While statistically grounded criteria are primarily relied upon, clinical relevance is gradually emphasized in medicine to ensure that the identified trajectory heterogeneity appropriately reflects changes in a disease or symptom over time. However, such considerations are often judged through visual comparisons, without concrete approaches for their application. To address this, the Area Between Trajectories (ABTs) was introduced as insights for quantifying trajectory group differences. Using a simulated sleep quality dataset, GBTM was applied to build and compare models. Subsequently, ABTs was demonstrated to show how it works, while also highlighting its limitations and potential applications.

  • Dual trajectories of loneliness and depression and their baseline correlates over a 14‐year follow‐up period in older adults: Results from a nationally representative sample in Taiwan

    by I Liu, Yu‐Jen Huang, Liang‐Kai Wang, Yi‐Hsuan Tsai, Sheng‐Lun Hsu, Chun‐Jui Chang, Ying‐Hsien Li, Yi‐Chen Hsiao, Chun‐Yuan Chen, Shue‐Ren Wann

    Aims: To explore the codevelopment between loneliness and depression in older adults, and to identify its potential baseline individual, family and extrafamilial correlates.Background: The number of older adults around the world has steadily increased over the last decades. Later life is a particularly vulnerable life stage due to multiple unfavourable conditions, and mental health in this stage appears to become an inescapable issue. Previous research has found the cross‐sectional association between loneliness and depression, but their codevelopment has been understudied. Therefore, exploring the codevelopment and its correlates has significant implications for prevention and healthcare professionals.Design: A longitudinal follow‐up study.Methods: The study used nationally representative data over a 14‐year follow‐up period from the Taiwan Longitudinal Study on Ageing focused on Taiwanese aged 60 years and above (n = 4049). Group‐based trajectory modelling, group‐based dual‐trajectory modelling and multinomial logistic regression were the primary analytical methods.Results: We identified three distinct dual trajectories of loneliness and depression: longitudinal low‐frequency lonely depressed (29.3%), longitudinal moderate‐frequency lonely depressed (59.4%) and longitudinal high‐frequency lonely depressed (11.3%). After considering several demographic and background characteristics, difficulty in physical functioning, number of physical symptoms and diseases, sleep quality and number of child deaths were found to be significantly associated.Conclusion: Across the three identified dual‐trajectory groups, they all showed a stable loneliness frequency pattern over time; however, the moderate‐frequency group and high‐frequency group both had a trajectory of increasing depression. It seems that depression tends to change over time in a worsening direction, especially for those with a certain frequency of loneliness. Furthermore, differences in individual and family correlates were found across the groups.Implications for practice: Interventions focusing on the specific factors may help hinder coexisting loneliness and depression, and have implications for developing health promotion strategies and chronic disease care plans.


  • A longitudinal, cross-lagged panel analysis of loneliness and depression among community-based older adults

    by Yu-Chun Hsueh, Chun-Yuan Chen, Yi-Chen Hsiao, Cheng-Ching Lin

    In the present study, we intended to examine the temporal association between loneliness and depression among community-based older adults. We analyzed data on 3,920 persons with a mean age of 67.6 years at baseline from the Taiwan Longitudinal Study on Aging. Structural equation modeling was used to investigate the temporal association between loneliness and depression over a period of 14 years. The results showed that the temporal association between loneliness and depression was bidirectional but stronger and more robust for depression as the initial symptom. Loneliness at Time 1 and 3, respectively, significantly predicted depression at Time 2 and 4. Depression at each time point had a significant effect on subsequent loneliness, even considering covariates. Future prevention programs for older adults with depression should also target at dealing with their loneliness, and vice versa. This study also highlights the need to assess and to manage the two symptoms simultaneously.


  • Exploring Individual, Family, and Extrafamilial Factors Associated With Depression Among Elderly Residents of Care Settings

    by Yi-Chen Hsiao, Chun-Yuan Chen

    This study investigated simultaneously the associations among individual, family, and extrafamilial factors and depression among elderly residents of care settings in Taiwan. The data for this study were obtained from the Vulnerability and Social Exclusion among Different Groups of Disadvantaged Elderly in an Aging Society: Phenomena and Strategies (1/2) project, which was conducted in Taipei City and Taipei County in 2007. We applied multiple regression procedures to a sample of 327 residents of care settings (48.3% women) who were mild physical impairment or physically independent and cognitively clear for determining factors associated with depression and their cumulative effects. We discovered that 24.5% of the residents were depressed. Physical mobility, number of chronic diseases, and self-assessed health status were significantly associated factors in all models. In the final model, we determined that one third of extrafamilial factors were significantly associated with depression, after controlling for all other variables. In addition, we observed an overall trend of increased mean scores of depression with an increase in the number of risk factors. The results imply that interventions aimed at preventing depression in elderly persons living in care settings, particularly persons with multiple risk conditions such as impaired physical mobility and medical comorbidities, should be designed to include extrafamilial elements. Establishing strong connections between elderly care setting residents and the institutional and neighborhood environments can help prevent depression.