How Has the Health of the United States Population Changed? Evidence From Biomarkers

Population health is a central input into social welfare, but health itself is complex and multidimensional. Mapping multiple dimensions of health into a single population health index may require strong assumptions about how to aggregate and weight across dimensions. This study adopts a novel nonparametric approach to rank individual health as a function of seven biomarkers associated with major domains of physiological health: body mass index (BMI), glycohemoglobin (HbA1c), total cholesterol, alanine aminotransferase (ALT), estimated glomerular filtration rate (eGFR), systolic blood pressure (SBP), and white blood cell (WBC). This approach avoids arbitrary weighting while requiring only monotonicity and quasi-concavity. Using NHANES data from 1988 to 2018, we document a significant deterioration in the U.S. health distribution, with declines concentrated in the lower tail. Counterfactual decompositions show that one-third of the negative shift can be attributed to an aging, more racially and ethnically diverse population, leaving a remaining portion unexplained by observed factors. The findings highlight the value of distribution-sensitive multidimensional health measures for monitoring health inequality and targeting preventive care.