Development and Application of the Dietary Environmental Index for the Assessment of Foods and Dietary Patterns in the United States
El-Abbadi, Naglaa.
2018
-
Abstract:
Introduction: Identifying and promoting food intake patterns that enhance overall diet
quality and health outcomes within ecologically viable parameters is crucial for dietary
sustainability and food security. Diet composition strongly influences dietary
environmental impact because of the wide variation in life cycle impacts among foods,
even among those with similar nutrition and ... read morehealth effects. Often processed foods high
in sugar, fat, or refined carbohydrates are low in greenhouse gas emissions; therefore,
focusing solely on minimizing environmental impact risks leaning too much on foods that
are less healthy. At the same time, diets that are nutrient-rich and diverse may be
associated with a heavy environmental burden. Future dietary recommendations must place
nutritional health in the context of environmental effects; to date, we lack methods for
aligning health and sustainability concerns in our dietary choices. Objectives: The
central objective of this dissertation research was the construction of a Dietary
Environmental Index as a comprehensive scoring system that includes both nutritional and
environmental impact assessment, to be utilized in scoring a wide range of foods
consumed in the United States. This tool is further applied to evaluate observed food
intake patterns, and ultimately used to project how they may be modified to optimize
nutritional health while mitigating environmental harm. Methods: For Aim 1, the DEX was
developed as the ratio of a Nutrient Density Score (NDS), derived from the Nutrient Rich
Foods Index 9.3, to an Environmental Impact Score (EIS) composed of four life cycle
environmental impacts representing climate change potential, land use, water resource
depletion, and marine eutrophication. DEX scores were calculated for 8,076 food products
from USDA databases. For Aim 2, the DEX scores were applied to evaluate 24-hr dietary
intake reported by adults in the 2007-08 National Health and Nutrition Examination
Survey (NHANES), standardized to 2000 kcal. Each participant's total food intake for the
day was classified into one of five patterns: lacto-ovo vegetarian (plant-based foods
plus dairy and egg); pesce-pollotarian (vegetarian plus any seafood and/or poultry); and
omnivore at 3 levels (all foods, with the proportion of red meat (in grams) out of total
food consumed at <5%, 5-15%, and >15%). For Aim 3, DEX-modeled food patterns were
constructed based on the NHANES 2007-08 reported food intake. A single high-DEX food was
selected to represent each of the 150 What We Eat in America food categories, and used
to substitute any reported food within that category which ranked below the 75th
percentile for DEX, at equal mass as originally reported per participant. The
DEX-modeled food patterns were standardized to 2000 kcal, and resultant changes in DEX,
NDS, and EIS scores were evaluated. Results from Aims 2 and 3 were compared against
nutritional intake recommendations in the 2015 Dietary Guidelines for Americans (DGA),
and the direct environmental impacts were calculated. Results: Aim 1: DEX scores were
calculated for 8,076 foods from the 2007-2008 National Health and Nutrition Examination
Survey, median [25th, 75th percentiles]: 39.1 [34.0, 43.7], higher scores favored.
Median DEX scores for food groups, highest to lowest, were legumes/nuts/seeds (48.2),
grains (44.7), vegetables (43.0), fruits (41.8), dairy (40.4), poultry (36.8), seafood
(32.7), pork (32.0), eggs (31.3), beef (28.1). The coefficient of variation of the DEX
score within food group categories ranged from 13% to 32%. Aim 2: Average DEX score of
dietary intake among NHANES participants was 39.6 ± 3.0. Among the food intake
patterns, more favorable DEX scores were associated with lower consumption of meat,
particularly red meat. The mean DEX of the lacto-ovo vegetarian category was highest
(41.8), with similar mean scores between pesce-pollotarian (40.8) and light-meat
omnivore (40.9), slightly lower for medium-omnivore (40.0), and with the lowest for
heavy-meat omnivorous intake (37.6). Overall, participants in the highest DEX tertile
category had superior HEI-2010 scores (p<0.001), and were better able to meet most
DGA recommendations (p<0.01). Aim 3: Eliminating foods below the 75th percentile of
DEX per WWEIA food category still allowed over 2000 food options, for which DEX, NDS,
and EIS scores showed improvement by 15%, 9%, and 7%, respectively. After substituting
all lower-scoring foods consumed with the high-DEX category representative, average DEX
score of the modeled dietary intake was 44.0 ± 3.0. DEX-modeled diets showed
overall higher intake of fruits, vegetables, and whole grains, and lower intake of meat.
As a result, the distribution of DEX-modeled food intake patterns shifted significantly
from their original counterparts observed in NHANES data. DEX-modeled diets were also
associated with greater levels of beneficial nutrients, including fiber, iron, calcium,
and vitamins A, C, and E, ranging from 32% to 82%. Levels of nutrients to discourage
were modestly reduced, including saturated fat, added sugar, and sodium, from -12% to
-16%. Individual direct environmental impacts improved by 11-17%. Conclusions: Our
results demonstrate the importance of considering the relative
nutrition-to-environmental impact of food choices, as only a modest correlation was
found between nutrient density of foods and their environmental impact, and substantial
variation was seen in DEX scores even within food groups. Our application of the DEX
towards the assessment of observed versus modeled food intake patterns provides insight
on how dietary intake reported by a nationally representative sample of the American
population fares in aligning diet quality and sustainability, and how moving towards
food decisions which incorporate environmental consideration may improve these results.
Future evaluations may explore the potential association between the DEX and
diet-related health outcomes. Additional considerations to enhance this research would
be the inclusion of food prices as an economic indicator of sustainability. This work
can further the development of a usable tool to promote sustainable public health
nutrition, and aid individuals trying to identify ways to be environmentally
conscientious in their dietary choices.
Thesis (Ph.D.)--Tufts University, 2018.
Submitted to the Dept. of Nutritional Epidemiology.
Advisor: Paul Jacques.
Committee: Timothy Griffin, Miriam Nelson, and Christian Peters.
Keywords: Nutrition, Epidemiology, and Environmental science.read less - ID:
- gq67k3886
- Component ID:
- tufts:26045
- To Cite:
- TARC Citation Guide EndNote
