UNVEILING KEY DETERMINANTS OF SUBJECTIVE WELL-BEING AMONG EUROPEAN OLDER ADULTS: A MACHINE LEARNING APPROACH

Subjective well-being has become a central focus in aging research, reflecting the interplay between psychological, social, and economic factors in later life. This study investigates the key determinants of subjective well-being among older adults in Europe, utilizing data from the Survey of Health, Ageing and Retirement in Europe (SHARE) waves 7–9. By integrating penalized regression models (Lasso and Ridge) with machine learning techniques such as Gradient Boosting, the analysis offers a comprehensive perspective on the factors shaping life satisfaction. The results emphasize the essential role of mental health—particularly depression—as the most significant predictor of well-being, followed by social relationships and financial security. Notably, subjective financial stability exerts a stronger influence on well-being than actual income, highlighting the psychological dimensions of financial security. While aging itself is associated with a modest increase in well-being, physical health decline and functional limitations negatively impact life satisfaction, used as measure of subjective well-being. The use of machine learning enhances the ability to detect complex interactions between predictors, providing deeper insights beyond traditional statistical methods. These findings underscore the need for targeted policy interventions that prioritize mental health support, social inclusion, and financial resilience to promote well-being in aging populations.