{"title":"Using machine learning to identify predictors of self-perceptions of aging among older adults in England","authors":"Mohsen Joshanloo","doi":"10.1016/j.archger.2025.105922","DOIUrl":null,"url":null,"abstract":"<div><h3>Objectives</h3><div>This study aimed to identify the strongest predictors of self-perceptions of aging (SPA) from an extensive set of 55 variables, including demographic, psychological, social, and economic factors.</div></div><div><h3>Methods</h3><div>Data were drawn from the English Longitudinal Study of Ageing, comprising over 7,000 adults aged 50 and older. Two advanced machine learning models, Random Forest and extreme gradient boosting (XGBoost), were used for data analysis. This approach allowed for a comprehensive evaluation of the relative importance of each predictor.</div></div><div><h3>Results</h3><div>Psychological factors emerged as the strongest predictors of positive SPA, outweighing health-related and demographic variables. Sense of control was identified as the strongest predictor, followed by pleasure derived from daily life experiences and the perception of life as worthwhile. Other significant predictors included components of emotional well-being (e.g., anxiety and happiness), autonomy, self-realization, and the quality of interpersonal relationships.</div></div><div><h3>Discussion</h3><div>The results indicate that a sense of competence, autonomy, and relatedness forms a vital foundation for positive self-perceptions of aging. These factors are enhanced by both hedonic and eudaimonic experiences, which contribute to the emotional and existential richness of the aging process. The findings highlight opportunities for targeted interventions and the refinement of existing theoretical models.</div></div>","PeriodicalId":8306,"journal":{"name":"Archives of gerontology and geriatrics","volume":"137 ","pages":"Article 105922"},"PeriodicalIF":3.5000,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Archives of gerontology and geriatrics","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0167494325001797","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"GERIATRICS & GERONTOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Objectives
This study aimed to identify the strongest predictors of self-perceptions of aging (SPA) from an extensive set of 55 variables, including demographic, psychological, social, and economic factors.
Methods
Data were drawn from the English Longitudinal Study of Ageing, comprising over 7,000 adults aged 50 and older. Two advanced machine learning models, Random Forest and extreme gradient boosting (XGBoost), were used for data analysis. This approach allowed for a comprehensive evaluation of the relative importance of each predictor.
Results
Psychological factors emerged as the strongest predictors of positive SPA, outweighing health-related and demographic variables. Sense of control was identified as the strongest predictor, followed by pleasure derived from daily life experiences and the perception of life as worthwhile. Other significant predictors included components of emotional well-being (e.g., anxiety and happiness), autonomy, self-realization, and the quality of interpersonal relationships.
Discussion
The results indicate that a sense of competence, autonomy, and relatedness forms a vital foundation for positive self-perceptions of aging. These factors are enhanced by both hedonic and eudaimonic experiences, which contribute to the emotional and existential richness of the aging process. The findings highlight opportunities for targeted interventions and the refinement of existing theoretical models.
期刊介绍:
Archives of Gerontology and Geriatrics provides a medium for the publication of papers from the fields of experimental gerontology and clinical and social geriatrics. The principal aim of the journal is to facilitate the exchange of information between specialists in these three fields of gerontological research. Experimental papers dealing with the basic mechanisms of aging at molecular, cellular, tissue or organ levels will be published.
Clinical papers will be accepted if they provide sufficiently new information or are of fundamental importance for the knowledge of human aging. Purely descriptive clinical papers will be accepted only if the results permit further interpretation. Papers dealing with anti-aging pharmacological preparations in humans are welcome. Papers on the social aspects of geriatrics will be accepted if they are of general interest regarding the epidemiology of aging and the efficiency and working methods of the social organizations for the health care of the elderly.