Abstract P097: Sociodemographic Differences in the Dietary Quality of Food Purchases Among Participants in the Nationally-Representative Food Acquisition and Purchase Study (FoodAPS)

Abstract

Background: Self-reported dietary data may not accurately represent usual dietary intake. Grocery purchase records provide an objective measure of diet, but little research has explored whether the diet quality of grocery purchases reflects individual-level diet quality and whether it differentiates between groups with known differences in dietary quality. This study evaluated whether there are sociodemographic differences in the dietary quality of grocery purchases akin to those observed in existing individual-level national data. Hypothesis: The diet quality of grocery purchases will reflect trends observed with individual-level diet quality. Methods: Grocery purchasing data from 3,961 households in the nationally-representative Food Acquisition and Purchase Study (FoodAPS) were used. Dietary data was collected using hand scanners over 7-days, and demographic data was self-reported. The Healthy Eating Index (HEI) 2015 was used to determine grocery purchase quality, and multivariable-adjusted regression with Tukey adjustment or planned contrasts were used to determine whether HEI-15 scores differed according to food insecurity, Supplemental Nutrition Assistance Program (SNAP) participation, obesity within a household, race/ethnicity, and interactions between these variables. FoodAPS sample weights and Taylor series linearization method for variance estimation accounted for the complex sample design. Results: Primary respondents were, on average, 50.6 years, non-Hispanic White (70.3%), female (70.2%) and had attended some college (57.8%). The mean HEI-15 score was 54.7; 14.0% of households were food insecure and 12.7% received SNAP benefits. HEI-15 scores differed across all sociodemographic predictors (p<0.05). Additionally, interactions (p<0.1) were detected between SNAP participation and (1) food insecurity and (2) obesity within the household as well as race/ethnicity and obesity within the household. Food secure, non-SNAP households had higher HEI-15 scores than food insecure, SNAP participating households (53.9±0.5 vs. 50.3±0.7, p=0.007) in models adjusted for age, income, smoking, and education. Similarly, non-SNAP households without obesity present had significantly higher HEI-15 scores than all other households. Obesity presence within a household was associated with lower HEI-15 scores in non-Hispanic White (50.8±0.5 vs. 52.5±0.7, p=0.046), and Black households (48.8±1.5 vs. 53.1±1.4, p=0.018), but not Hispanic households (54.4±1.0 vs. 52.2±1.2, p=0.21). Conclusions: Disparities in the diet quality of grocery purchases reflect individual-level trends and vary based on sociodemographic factors. This analysis highlights the interrelationships between SNAP, food insecurity, race/ethnicity, and obesity within a household, demonstrating how the co-occurrence of risk factors may modify associations with diet quality.