LAUSR

Objectives Recent research indicates that climate-friendly diets are healthier, but much of this evidence is based on aggregate consumption data or on individuals' dietary intake data from just one observation day. Analysis of individuals' self-selected diets is important for applying dietary guidance, but usual intake over time is needed to better assess relationships with human health and environmental outcomes. To address this gap, we developed a method to estimate the association of dietary carbon footprints with nutritional quality for usual intake based on multiple observation days. Methods Based on an exhaustive review of the sustainability science literature, we developed a database of Food Impacts on the Environment for Linking to Diets (dataFIELD), which has greenhouse gas emissions (GHGE) of 332 food commodities. We linked dataFIELD to the dietary recall data for each of two days in the 2005-2010 National Health and Nutrition Examination Survey (NHANES), and calculated GHGE per day for each individual. Finally, we adapted the multi-variate Markov chain Monte Carlo (MCMC) simulation procedure, originally developed by the National Cancer Institute to estimate the joint usual intake distributions of multiple dietary components. We included GHGE/1000 kcal along with the 12 dietary components in the Healthy Eating Index (e.g., total fruit intake/1000 kcal, sodium/1000 kcal, etc). Our MCMC model accounts for episodic consumption of foods (such as fish) and allows for inclusion of covariates, such as age, sex, or race-ethnicity to improve estimates. Results We demonstrate the method using two days of dietary intake for all adults with reliable data (n = 16,800). Previous results by our team using one-day data indicate a significant (P < 0.05) inverse relationship between Healthy Eating Index scores and GHGE. Analyses with this new method provide similar results. Conclusions This study demonstrates a method to estimate the relationships between healthfulness and carbon footprints for usual dietary intake. This is important because it allows for examination of diet quality and sustainability indicators of real, self-selected diets over time. These methods can be used to identify healthy, environmentally beneficial dietary changes for promotion in nutrition education. Funding Sources Wellcome Trust.