Yuke Yang, Yufei Zhang, Zhihao Ma
Quality of life (QoL) in middle and later life is widely assessed using the CASP scale, yet its measurement structure and comparability across groups and longitudinal waves remain contested. Using data from five waves of the Survey of Health, Ageing and Retirement in Europe (SHARE), spanning nearly 10 years and 12 countries, this study examined the factor structure and measurement invariance of the SHARE CASP-12, applied a trait–state–occasion framework to distinguish stable trait-related from occasion-specific components of QoL, and assessed their social, economic, and health-related predictors. The results showed that the first-order three-factor model distinguishing Control/Autonomy, Pleasure, and Self-realization provided the best representation of the data. Scalar invariance was not supported across the grouping schemes examined, suggesting that unadjusted cross-group comparisons may be insufficiently justified. QoL comprised both stable and occasion-specific variation: Pleasure was predominantly occasion-specific, whereas Control/Autonomy and Self-realization were more trait-related. The relative contributions of these components varied across countries, sex, waves, and age. Education was more consistently associated with trait-related variation, while health status, chronic disease status, and household economic status showed more domain- and component-specific associations. Sensitivity analyses supported the main conclusions for Control/Autonomy and Self-realization, whereas findings for Pleasure were more sensitive to country composition. Overall, these results underscore the importance of measurement comparability and suggest that stable inequalities and shorter-term fluctuations in QoL may have distinct determinants.