Global self-rated health data from a longitudinal study predicted mortality better than comparative self-rated health in old age

Abstract

Background and objective: Self-rated health (SRH) has proved to be a predictor of subsequent mortality in old age. This study examines if the different question wording in SRH questions influences the association of SRH with mortality. Two SRH measures are examined, an age group comparative question and a global question with no explicit point of reference.

Methods: The data are from the Tampere Longitudinal Study on Ageing, consisting 944 respondents aged 60-89 years. The association between mortality and self-rated health was studied at 5, 10, and 20 years follow-up using Cox proportional hazard models.

Results: As crude measures, global SRH was significantly associated with mortality after 5, 10, and 20 years follow-up, but the comparative SRH was not. After adjustment for age and several social and health indicators both SRH measures were associated with increased mortality risk even after 20 years of follow-up.

Conclusions: Because the age-sensitivity of the comparative SRH the global SRH may be a more appropriate measure in studies where the study population has a large age range and also as a health measure in clinical settings.