(P14-012-23) Sociodemographic Predictors of Eating Patterns Associated With Higher Quality Diets

Objectives: Evidence supports a higher quality diet contributes to better health outcomes (e.g., reduced risk of cardiovascular disease, cancer). It remains unclear whether sociodemographic characteristics predict how likely one engages in day-level eating patterns (e.g., type of food, timing of eating episode) contributing to a higher quality diet. Thus, we examined associations between sociodemographic characteristics and eating patterns linked with higher quality diets.

Methods: Using diet data from 2015 and 2017 NHANES cycles (N recalls = 17088), we previously used decision tree modeling to determine a day-level eating pattern that contributed most to higher diet quality [i.e., HEI-2015 score ≥ 50.26]. A single eating pattern emerged: 1) any fruit and whole grain intake at breakfast and 2) intake of < 7.3 grams of solid fats at dinner. Logistic regression analysis was used to investigate associations between that single eating pattern with race/ethnicity, gender, age, income, and body mass index (BMI). We adjusted for energy intake and participant-level variability in the model.

Results: Approximately 11% of recalls (N recalls = 1964) met all three eating behaviors (i.e., eating pattern). Participants had higher odds of consuming this pattern if they were older [OR 1.02 (95% CI: 1.02, 1.03)], had higher income [OR 1.04 (95% CI: 1.02, 1.05)], and had a lower BMI [OR 0.98 (95% CI: 0.97, 0.99)]. Those who identified as males had 31% lower odds of consuming this pattern [OR 0.69 (95% CI: 0.60, 0.78)] than females. Compared to Caucasians, those of Hispanic ethnicity (outside of Mexican American) and non-Hispanic Asians had higher odds of meeting this eating pattern [OR 1.31 (95% CI: 1.07, 1.61); OR 1.45 (95% CI: 1.19, 1.77), respectively]. There were no significant differences in engagement between Caucasians with Mexican Americans (p-values ≥ 0.05).

Conclusions: Older age, higher income, and lower BMI status were associated with an eating pattern linked to a higher quality diet. Gender and race/ethnicity also predicted whether one engaged in health-promoting eating behaviors. We will apply these findings in future studies that design and test the efficacy of targeted diet quality messages for different populations.

Funding Sources: AWL is partly funded by NIDDK.