Health Measures and Health Inequality Over the Life Course
Final Report Abstract
This project investigated the role of self-reported and objective health measures in studying health inequality over the life course. Previous research has relied heavily on self-reported health measures, raising concerns about biases related to education, gender, age, cohort, and national context. This study systematically assessed whether such biases influence conclusions about health inequality and whether objective measures provide a more reliable alternative. The findings highlight three key insights. First, self-reported health measures are not necessarily more biased than objective ones. A comparison of self-rated health, the physical component scale from the SF-12, and grip strength demonstrated that self-reported measures produced more reliable findings on cumulative advantage and disadvantage than grip strength, particularly among men. A study on underreporting of chronic conditions such as diabetes and hypertension found that misreporting had only a minor impact on conclusions about health inequality. Further research comparing self-reported depressive symptoms and diagnosed depression raised doubts about the validity of diagnoses as a superior measure of mental health. Finally, an analysis of self-rated health across multiple countries found it to be interpreted consistently across age groups and genders, reinforcing its reliability for comparative health inequality research. Second, the accuracy of self-reported health measures varies by national context. A study on hypertension underreporting in the United Kingdom found minimal effects on inequality estimates, but a replication in China, where healthcare access is highly stratified, revealed substantial underreporting. Adjusting for misreporting significantly altered conclusions on gender, education, and rural-urban disparities in hypertension prevalence over the life course, underscoring the importance of considering healthcare access when interpreting self-reported health data. Third, given the robustness of self-reported health measures, the project explored efficient data collection methods for advancing research on health inequality. A benchmarking study of nonprobability surveys demonstrated that these surveys can reliably estimate certain health disparities—particularly those related to education and age—while offering a cost-effective alternative to traditional probability-based surveys. However, the study also showed that nonprobability surveys tend to overestimate gender disparities and introduce inconsistencies in racial and ethnic disparities.
Publications
-
Health Measurement and Health Inequality Over the Life Course: A Comparison of Self-rated Health, SF-12, and Grip Strength. Demography, 56(2), 763-784.
Leopold, Liliya
-
Education and age trajectories of chronic conditions: Are tests of the cumulative advantage and disadvantage hypothesis biased by underreporting?. Social Science & Medicine, 334, 116134.
Leopold, Liliya; van Valkengoed, Irene G.M. & Engelhardt, Henriette
-
Mental Health in China: Social Change in Life Course Trajectories. Population and Development Review, 51(2), 759-796.
Ge, Tingshuai; Van Leeuwen, Frans Johannes; Jiang, Quanbao & Leopold, Liliya
