Is Subjective Health Reliable as a Proxy Variable for True Health? A Comparison of Self-rated Health and Self-assessed Change in Health among Middle-aged and Older South Koreans

Abstract

In numerous studies, subjective health is widely used as a proxy for actual health status. However, subjective health questions are exposed to reference group bias, adaptability bias, ceiling effect and floor effect, retention bias, and inter-individual heterogeneity, and may have significant measurement errors. Therefore, it is important to verify that subjective health is an effective proxy for actual health status. The purpose of this study was to examine whether self-rated health and self-assessed change in health predict the actual health status (diagnosis of disease and experience of hospitalization and accident) of middle-aged and older South Koreans, using the Korea Longitudinal Study of Aging. In order to control individual heterogeneity, we analyzed whether the subjective health of health status change between two time points effectively predicts the occurrence of actual health problems between two time points. To ensure the validity of the study, various health variables were analyzed. The analysis showed that the negative changes of SRH and SACH in all models without age group significantly predicted the occurrence of health problems. It was also found that SACH predicts health problems much better than SRH. However, by analyzing by age group, the predictive power of SRH decreased significantly, but the predictive power of SACH remained high. Therefore, it is effective to use subjective health as a surrogate variable of actual health status in Korea. However, if possible, using SACH as a proxy for actual health status would be preferable to using SRH.