Reporting heterogeneity and health measurement bias in older adults: nationwide cross-sectional study.

Abstract

Purpose: To examine reporting heterogeneity in self-rated health (SRH) among middle-aged and older adults using anchoring vignettes, and evaluate the impact of reporting heterogeneity on further measures of health inequalities.

Methods: Data on a sample of 13,323 respondents aged 45 years and over from the China Family Panel Studies (CFPS) were used. An ordered probit model and the hierarchical ordered probit (HOPIT) model were used to compare the estimated differences in pre- and post-adjusted SRH, and the concentration index was applied to measure health inequalities.

Results: After correcting for reporting heterogeneity, the health advantage of these middle-aged and older adults was further expanded to varying degrees, and those living in urban areas had better health compared to those living in rural areas (β = 0.057, P = 0.046). In addition, respondents with a high school education or higher and in urban areas use a lower cut-off point when rating a given health condition as poor and a higher cut-off point when rating a given health condition as good. Significant increases in the inequalities in SRH between urban and rural areas and across regions after adjustment for reporting heterogeneity. In comparison to urban areas, the trend that higher SRH was more concentrated among the better educated is more pronounced in rural areas.

Conclusion: These findings highlight the importance of addressing reporting heterogeneity in SRH to ensure equitable healthcare interventions and resource allocation, particularly to underserved regions and populations.