Longitudinal validity of self-rated health: the presence and impact of response shift

Objective: This paper aimed to examine the longitudinal validity of self-rated health (SRH) and whether it would be affected by possible changes in evaluation standards (i.e., response shift) over time.Design: Data are from a longitudinal survey of a nationally representative sample in China. Analytical sample was restricted to respondents aged 45 and above (n = 15,893). Individual fixed effects models were used to analyze changes in ratings on health anchoring vignettes and self-rated health over time.Main outcome measures: SRH at two time points with a -two-year span.Results: Both SRH and anchoring vignettes ratings displayed changes over a two-year span for all the studied age groups. Compared with the self-assessed change in health ("How would you rate your health as compared to that of last year?"), changes in SRH reported over time displayed a more stable and optimistic pattern. SRH responded to doctor diagnosed chronic disease and changes in functional limitation, before and after adjusting for evaluation standards.Conclusion: SRH is responsive to the newly diagnosed chronic disease and functional limitation, regardless of whether we consider response shift within the same respondents over time.