Introduction:
There has been considerable work performed on nutrient profiling to assess the nutritional contribution of a food to a healthy dietary pattern. Most profiling approaches have focused on nutrients to limit and nutrients to encourage. A few profiling approaches have also included certain food groups in the profiling algorithm.
Objectives:
The objective of this study was to develop a nutrient density score, based on the Nutrient Rich Food Index (NRF) 6.3, that includes food groups and validate the score against a gold-standard marker of diet quality, the Healthy Eating Index (HEI) 2015.
Methods:
Stepwise regression was used to develop a nutrient density score based on the day 1 total dietary intake of the U.S. population 2 years and older (excluding pregnant and lactating women) from the National Health and Nutrition Examination Survey (NHANES) 2011-2016 (n=23,743). Intake of food groups was taken from the Food Patterns Equivalent Database (FPED) 2011-2016. Sixteen nutrients (as a percent of the Daily Value) as well as five food groups (as a percentage of recommended intake in 2015-2020 Dietary Guidelines) were considered in the score.
Results:
When only the 16 nutrients were included in the score, 66% of the variability in the HEI 2015 could be accounted for (R
2
= 0.66). When only the five food groups were considered, the maximum R
2
with the HEI 2015 was 0.50. However, when both nutrients and foods groups were considered, the model explained 74% of the variability in the HEI 2015 (Table). The increase in the R
2
begins to plateau after the inclusion of 10 elements: 3 nutrients to encourage (fiber, potassium and unsaturated fat), 4 food groups (dairy, fruit, whole grains, and nuts and seeds) and 3 nutrients to limit (added sugar, saturated fat, sodium).
Conclusion:
A nutrient density score that includes both nutrients and foods groups best predicts diet quality as measured by the HEI 2015.