2604.02143 How large is healthy-user bias in nutritional epidemiology? Quantifying the confounding floor with negative-control outcomes in the NHANES III cohort
Observational studies repeatedly find that people who take vitamin or dietary supplements have lower cardiovascular mortality, but randomised controlled trials of the same supplements typically do not replicate those benefits. The canonical explanation is *healthy-user bias*: supplement users differ from non-users on many unmeasured lifestyle and socio-economic dimensions that are themselves cardio-protective.