O-15 — Cohort-specific estimates before pooling; HPFS (men) crosses 1. See
Methodology
Prospective pooling of three Harvard cohorts free of T2D/CVD/cancer at baseline - Nurses’ Health Study (NHS, women, 1980-2012), NHS II (women, 1991-2017) and Health Professionals Follow-up Study (HPFS, men, 1986-2016), combined n=213,798, 5,529,959 person-years, 20,514 incident T2D cases. Egg intake from repeated FFQs entered as cumulative averages; Cox models (Model 3, fully adjusted) controlled for age, BMI, smoking, physical activity, alcohol, total energy and intake of foods co-consumed with eggs (bacon, red/processed meat, refined grains, potatoes, coffee, sugar-sweetened beverages) and overall diet quality (AHEI). Cohort HRs pooled by fixed-effects meta-analysis. The paper separately updates a dose-response meta-analysis to 16 prospective cohorts (589,559 participants, 41,248 cases), stratified by region.
Link to original