Journal of affective disorders · 2025
Key factors of depression in middle-aged and older adults based on social-ecological systems theory: an interpretable machine learning approach.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Chen Haihui, Fan Yunge, Zhou Licheng, Lin Junxu, Liu Dongming, Ji Ningna, Mo Xuan, Wang Dongfang, Ma Zijuan, Fan Fang
Аннотация
BACKGROUND: In the context of population aging, depression is becoming a problem that prevents middle-aged and older adults from achieving healthy aging. The impact of individual and environmental factors, on depressive symptoms has not been sufficiently examined in prior research. To identify the key impact factors of depression and explore their associations, this study developed an interpretable machine learning model based on the social ecosystems theory that encompassed multidimensional predictors.
METHODS: This study analyzed the data of 4634 adults aged above 40 years from the Psychology and Behavior Investigation of Chinese Residents. We included 32 factors at the individual, family, and social levels that may influence depression, and compared the performance of five models. An optimal model was identified and employed to detect the factors associated with depression and the direction of their associations.
RESULTS: The categorical boosting (CatBoost) model outperformed all other models. The top four features with the greatest impact were unhealthy eating behavior, family health, health literacy, and neuroticism trait, with SHAP values of 0.60, 0.53, 0.23, and 0.20, respectively. Unhealthy eating behaviors and higher neuroticism were positively associated with the likelihood of depression, while better family health was negatively associated. In addition, the association between health literacy and depression followed an inverted U-shaped curve.
CONCLUSIONS: It is recommended to apply the CatBoost machine learning model for screening depressive symptoms among middle-aged and older adults. Multifaceted factors at the individual and environmental levels are suggested to be included when exploring the likelihood of depression.
METHODS: This study analyzed the data of 4634 adults aged above 40 years from the Psychology and Behavior Investigation of Chinese Residents. We included 32 factors at the individual, family, and social levels that may influence depression, and compared the performance of five models. An optimal model was identified and employed to detect the factors associated with depression and the direction of their associations.
RESULTS: The categorical boosting (CatBoost) model outperformed all other models. The top four features with the greatest impact were unhealthy eating behavior, family health, health literacy, and neuroticism trait, with SHAP values of 0.60, 0.53, 0.23, and 0.20, respectively. Unhealthy eating behaviors and higher neuroticism were positively associated with the likelihood of depression, while better family health was negatively associated. In addition, the association between health literacy and depression followed an inverted U-shaped curve.
CONCLUSIONS: It is recommended to apply the CatBoost machine learning model for screening depressive symptoms among middle-aged and older adults. Multifaceted factors at the individual and environmental levels are suggested to be included when exploring the likelihood of depression.