Carbon footprint and nutrition of self-selected US diets

To obtain the dietary data, participants were asked to recall what they had eaten in the last 24 hours, aided by items such as glasses and measuring spoons to help participants estimate portion sizes.

Greenhouse gas emissions data for production and (in some cases) primary processing were gathered via a literature survey. The emissions data for different food types will become available soon here. The diets of each participant were ranked according to greenhouse gas emissions per 1000 kcal. The figure below shows the distribution of emissions, with the average emissions intensity being 2.21 kg CO₂ eq. per 1000 kcal and the average overall dietary emissions per day being 4.72 kg CO₂ eq. per person per day.

Image: Figure 2, Rose et al. Distribution of dietary GHGE from 1-d diets, NHANES 2005–2010, kg CO₂-eq/1000 kcal. CO₂-eq, carbon dioxide equivalent; GHGE, greenhouse gas emissions.​

Health was assessed using both levels of specific nutrients and the USDA’s Healthy Eating Index (HEI), which scores diets on a 100-point scale (with a higher score being healthier). The lowest quintile of diets (defined in terms of emissions per calorie) had significantly higher (P < 0.001) HEI scores by 2.3 ± 0.7 points on a 100-point scale (an average score of 50.3, compared to an HEI of 48.0 for the highest quintile of diets by emissions). The lowest-emitting quintile of diets had healthier HEI scores than the highest-emitting quintile of diets on some components of the scale, including whole fruit, whole grains, seafood and plant proteins, and ratio of unsaturated to saturated fatty acids. The highest-emitting quintile of diets had healthier HEI scores for total protein foods, dairy, and total vegetables. The high-emitting diets also had higher levels of some nutrients, including vitamins A and D, choline, calcium, iron, and potassium.

The study also looked at demographics and behaviour in relation to diets. It found that women, African-Americans and people ages 18 to 19 were more likely to have diets in the lowest quintile of emissions per calorie. Those who considered themselves to be vegetarian had, on average, diets with lower emissions per calorie (by 0.8 kg CO₂ eq. per 1000 kcal) than nonvegetarians. However, the paper doesn’t specify how daily caloric intake (and hence total dietary emissions, as opposed to emissions per calorie) compare for vegetarians and nonvegetarians.

Abstract

Background

A substantial portion of greenhouse gas emissions (GHGE) has been attributed to the food sector, but little is known about the association between the carbon footprint of individual self-selected diets in the United States and nutritional quality.

Objectives

The aims of this study were to assess the GHGE from individual self-selected diets in the United States and examine their association with nutritional quality of the diets, demographic patterns, and food-related behaviours.

Methods

The dietary GHGE from US adults (>18 y, N = 16,800) in the 2005–2010 National Health and Nutrition Examination Survey (NHANES) were calculated by linking all foods consumed in their 24-h recall diets to our new database of food environmental impacts. Diets were ranked by GHGE/1000 kcal. Those in the top and bottom quintiles were compared on the US Healthy Eating Index (HEI) and on the amounts of specific nutrients known to be under- or overconsumed in the US population. Demographic and behavioural variables from the NHANES were also correlated to these dietary carbon footprints.

Results

Diets in the bottom quintile accounted for one-fifth the total emissions (GHGE/1000 kcal) of those in the top quintile, yet had significantly higher (P < 0.001) HEI scores by 2.3 ± 0.7 points on a 100-point scale. These low-GHGE diets contained higher amounts of fiber and vitamin E and lower amounts of sodium and saturated fats, whereas high-GHGE diets contained higher amounts of vitamins A and D, choline, calcium, iron, and potassium. Low-GHGE diets had less meat, dairy, and solid fats, and more poultry, plant protein foods, oils, whole and refined grains, and added sugars.

Conclusions

Food patterns responsible for lower GHGE had a better overall diet quality and were more nutritious on several key dimensions, although not all. These results can inform dietary guidance and other policies that seek to address the goals of improved dietary intakes and reduced food-related emissions.

Reference

Rose, D., Heller, M. C., Willits-Smith, A. M., and Meyer, R. J., 2019. Carbon footprint of self-selected US diets: nutritional, demographic, and behavioral correlates, The American Journal of Clinical Nutrition, nqy327.

Read the full paper here. See also the Foodsource chapter What is a healthy sustainable eating pattern? and the paper Greenhouse gas emissions and energy use associated with production of individual self-selected US diets.