Abstract P482: Associations Between Determinates of Food Choice and Healthy Eating Index-2015 in Adults: An NHANES 2017 - 2018 Analysis

Introduction: Previous work in nationally representative US samples has focused on associations between single determinants of food choice and diet quality assessed by the Health Eating Index-2015 (HEI-2015). Limited research has examined the relationship between multiple domains of food choice and diet quality. A greater understanding of the association between individual-level determinants of food choice and diet quality may inform the development of interventions to improve diet quality. Aim: The aim was to examine the association between potentially modifiable factors known to affect individual-level food choice and HEI-2015. Methods: This cross-sectional analysis included adult participants in NHANES 2017 - 2018. Determinates of individual-level food choice were based on a previously proposed framework. Variables assessed in NHANES were mapped to the following four domains: 1) food-external factors (use of nutrition labels, hours worked weekly, typical work schedule); 2) personal-state factors (health status, consumption of food away from home); 3) cognitive factors (perceived diet quality, knowledge of MyPlate); 4) society-related features (monthly grocery expenditures, food security status, income to poverty ratio). Univariate regression assessed the relationship between food choice determinates and HEI-2015 score. Statistically significant variables were entered into multivariate regression models predicting HEI-2015 score. Significant variables from the final model were used in multivariate regression models predicting adequacy and moderation component sub scores of HEI-2015. All models were adjusted for non-modifiable determinates of food choice, including age, sex, race, BMI, education, and relationship status. All analyses were survey weighted. Results: The weighted mean HEI-2015 score for the sample (N=1467) was 51.69 points (95% CI 49.55, 53.84; weighted n=95.6 million). In the final model, perceived diet quality, food away from home, and use of nutrition labels were significant predictors of HEI-2015 (R 2 adj =0.25, p<0.001). Those with excellent/very good perceived diet quality had a 4.88-pt higher HEI-2015 score than those with fair/poor perceived diet quality (p<0.001). Consuming food away from home ≥5 times per week resulted in a 5.92-pt lower HEI-2015 score than 0 times per week (p=0.02). Always/most of the time using nutrition labels was associated with a 7.85-pt higher HEI-2015 score compared to rarely/never using labels (p<0.001). Variables accounted for a higher proportion of variance in adequacy score (R 2 adj =0.10, p<.001) than moderation score (R 2 adj =0.24, p<.001). Conclusion: Food-external, personal-state, cognitive, and socio-cultural factors are associated with diet quality. Future research should consider multiple individual-level food-choice determinates when developing interventions to improve diet quality.