The potential of the EAT-Lancet reference diet, which promotes a healthy diet within planetary limits, to reduce greenhouse gas emissions (GHGe) remains understudied. This study examines the role of nutritional and acceptability constraints in reducing GHGe through diet optimization, and tests the alignment between GHGe reduction and the EAT-Lancet score. The study used data from 29,413 NutriNet-Santé participants to model French diets and evaluate their environmental, nutritional, economic, and health impact. The Organic Food Frequency Questionnaire was used to assess organic and conventional food consumed, and the Dialecte database was used to estimate the diet environmental impacts. Quality of diets were also evaluated based using the PNNS-GS2 (Programme National Nutrition-Santé 2 guidelines score). When testing minimizing GHGe under strict nutritional and acceptability constraints, it was possible to reduce GHGe up to 67 % (from 4.34 in the observed diet to GHGe = 1.45 kgeqCO2/d) while improving the EAT score by 103 % with 91 % of the food as organic. Greater reductions required relaxation of some constraints. When testing maximizing EAT score under gradual reduction in GHGe, the adherence to the EAT-Lancet diet was not significantly affected by the gradual reduction in GHGe. To maximize EAT score with 75 % reduction in GHGe (1.09 kgeqCO2/d), less strict constraints on the bioavailability of iron and zinc are necessary. The EAT score improved by 141 %, while land occupation decreased by 57 %, compared to the observed value. The diet contained 94 % of organic foods. There was some alignment between the degree of adherence to the EAT-Lancet diet and the reduction in GHGe, but other diets may also lead to a strong reduction in GHGe. Thus, GHGe can be greatly reduced by dietary choices, but require profound reshaping of diets which must be coupled with changes in other areas of the food chain.
