Development and evaluation of a nutrient-based method of classifying moderation foods
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
While dietary guidelines recommend limiting foods high in added sugars, saturated fat, refined grains, and sodium, there is no criteria for identifying foods high in these components.
Objective
To evaluate a novel nutrient-based method to classify moderation foods and compare with two alternative methods.
Design
Face validity was assessed by examining the proportion of recommended and non-recommended foods classified as moderation using the 2017-2018 Food and Nutrient Database for Dietary Studies (FNDDS, n=6909). Convergent validity was evaluated by comparing nutrient density in moderation vs. non-moderation foods in FNDDS, and investigating correlations of moderation food intake with diet quality from 2-day dietary recalls. Results were compared with the hyperpalatable food and NOVA ultra-processed food classifications.
Participants
National Health and Nutrition Examination Survey (2017-2018) non-pregnant participants aged 2+ years (n=6136) were included.
Main outcome measures
The Nutrient Rich Foods 9.3 index (NRF) measured nutrient density; Healthy Eating Index-2020 scores (HEI-2020) measured diet quality.
Statistical Analyses
T-tests and ANOVA evaluated differences in NRF by moderation, hyperpalatable, and ultra-processed classifications. Fisher’s z-transformation compared associations of HEI-2020 with intake (%kcal) from moderation, hyperpalatable, and ultra-processed foods.
Results
More non-recommended (e.g., 97% of snacks) than recommended (e.g., 18% of vegetables) food groups were classified as moderation. NRF was significantly lower in moderation foods than non-moderation foods (mean diff: -74.8, 95%CI:-70.6 - - 78.9). The difference in NRF between moderation vs. non-moderation foods was larger than that between ultra-processed vs. non-ultra-processed (mean diff=-29.7, 95%CI:-25.1- - 34.2) and hyperpalatable vs. non-hyperpalatable foods (mean diff=-53.2, 95%CI:-49.3- - 57.1)(p-diff<0.001). Moderation food intake (% kcal) was correlated with HEI-2020 (r=-0.72), and associations were stronger than those with hyperpalatable (r=-0.40) or ultra-processed food intake (r=-0.49).
Conclusions
The moderation food classification method demonstrated strong face and convergent validity, and may improve diet quality assessment and public health interventions.
