Abstract
BACKGROUND: Intake of foods high in nutrients of concern (moderation foods) contribute a large proportion of total energy intake in the U.S, but little is known about how moderation food intake varies across sociodemographic subgroups. OBJECTIVE: This observational study assessed sociodemographic differences in the contribution of moderation food to energy intake in U.S. adults. METHODS: Data from 24-hour dietary recalls of participants 20 - 80 years of age in the 2017-2020 National Health and Nutrition Examination Survey (n=7706) were used to calculate the percent of energy intake from foods meeting any of the following criteria: added sugar >20% energy, sodium >460mg per serving, refined grains >50% of total grains or >10:1 ratio of carbohydrate to fiber content, saturated fat >20% energy, total fat >9% by weight (applied only to vegetables, sweets, and snacks), and all alcoholic beverages. Moderation food intake was compared by age, sex, education, income poverty ratio (IPR), and race/ethnicity, and linear regressions with multiplicative interaction terms investigated whether race/ethnicity or education modified relationships of IPR with moderation food intake. RESULTS: Mean moderation food contributed 75% (95%CI:75%-77%) of energy intake in U.S. adults, and was higher in younger (mean=79%) vs. older (mean=72%-75%) adults, males (mean=77%) vs. females (mean=74%), and adults in the lowest (mean=79%) vs. highest (mean=74%) income quartile. Moderation food intake was lowest in Non-Hispanic Asian adults (mean=66%) versus other race/ethnicity groups (m=73% - 80%) and in the highest educated (mean=71%) vs. less-educated adults (m=75%-79%). The relationship of income with moderation food intake did not differ by race/ethnicity overall, but the relationship was significantly more negative for non-Hispanic Black adults than in Mexican American/Hispanic adults. Similarly, while the relationship of income with moderation food intake did not differ overall by education, marginal slopes estimates indicated an inverse association among high school graduates (second highest education group; β = -1.1 95%CI = -2.1 to -0.2) and null associations for all other adults. CONCLUSIONS: High moderation food intake across all U.S. adult subgroups suggests a need for population-wide changes in food selection to meet dietary guidelines.