Do self-reported data accurately measure health inequalities in risk factors for cardiovascular disease?

Abstract

This study aimed to compare the magnitude of educational inequalities in self-reported and examination-based hypertension and hypercholesterolemia and to assess the impact of self-reported measurement error on health inequality indicators. We used the Portuguese National Health Examination Survey data (n = 4911). The slope index of inequality (SII) and the relative index of inequality (RII) were used to determine the magnitude of absolute and relative education-related inequalities. Among the 25-49-year-old (yo) men, absolute and relative inequalities were smaller for self-reported than for examination-based hypertension (SIIeb = 0.18 vs. SIIsr = - 0.001, p < 0.001; RIIeb = 1.99 vs. RIIsr = 0.86, p = 0.031). For women, the relative inequalities were similar despite differences in self-reported and examination-based hypertension prevalence. For hypercholesterolemia, self-reported relative inequalities were larger than examination-based inequalities among the 50-74-yo men (RIIsr = 2.28 vs. RIIeb = 1.21, p = 0.004) and women (RIIsr = 1.22 vs. RIIeb= 0.87, p = 0.045), while no differences were observed among 25-49-yo. Self-reported data underestimated educational inequalities among 25-49-yo men and overestimated them in older individuals. Inequality indicators derived from self-report should be interpreted with caution, and examination-based values should be preferred, when available.