Pools 9 US prospective cohorts (n=103,811) and re-estimates egg-intake associations with incident type-2 diabetes and CHD after adding an overall dietary-pattern/diet-quality adjustment on top of standard covariates (age, ethnicity, BMI, exercise, smoking, alcohol) — an explicit empirical test of whether confounding by a healthier or unhealthier overall diet explains the raw egg association, rather than assuming it away or asserting it. Even after this adjustment, ≥2 eggs/week remained associated with elevated T2D risk (27% higher at 7+/week vs none, 95% CI 16-37%), and a sensitivity analysis in older adults found a CHD signal at 5-6 eggs/week (+30%, CI 3-56%); no overall CHD association was found at lower/typical intakes.
relevance_note: Egg-specific empirical test of the healthy/unhealthy-user-bias hypothesis via diet-quality adjustment — here the confounding-adjustment does not fully explain away the signal.