(P14-034-23) The Association Between Dietary Pattern Indices and Glycemic Control in African American Adults with Type 2 Diabetes in TX STRIDE

Objectives: We evaluated the relationship between common measures of diet quality and HbA1c in a cohort of African American (AA) adults with type 2 diabetes (T2D).

Methods: This cross-sectional analysis used data from participants in TX STRIDE, a clinical trial of a resilience-based diabetes education program delivered to AA adults with T2D. At baseline, we collected: demographic data via survey, HbA1c via blood draws, and two diet recalls, from which we calculated healthy eating index, alternative healthy eating index, and alternative Mediterranean diet scores. We used the NOVA classification system to calculate the percentage that ultra-processed foods (UPF) and unprocessed or minimally processed foods (UMPF) contributed to the calories and grams of each participant’s diet. HbA1c data were divided into statistical quintiles for analysis. We analyzed the association between each measure of diet quality using linear regression, adjusting for covariates using stepwise regression.

Results: The analytic sample included 175 AA participants, mean T2D duration of 10.9 years, average age was 62 years, 72% were female, and 78% were obese. The percentage of grams from UPF in the diet was positively associated with HbA1c (β = 0.015, p=0.024) while UMPF was inversely associated with HbA1c (β = -0.014, p=0.042). Following adjustment for insulin use and employment status, UPF and UMPF remained significantly associated with HbA1c (β = 0.014, p=0.029 and β = -0.013, p=0.045, respectively). Exploratory analyses showed that the results were driven, in part, by higher intake of UPF beverages among participants with higher HbA1c, particularly artificially sweetened beverages. No other dietary pattern was associated with HbA1c.

Conclusions: In this, the first study to find a significant relationship between any dietary quality pattern and glycemic control in AA adults with T2D, the degree of processing was the only measure of diet quality that significantly correlated with HbA1c. The positive relationship between UPF intake and T2D risk is well established, but this is one of the first studies to find an association between UPF intake and glycemic control. We cannot infer directionality in this analysis but plan to examine changes in UPF and HbA1c throughout the TX STRIDE trial.

Funding Sources: National Institutes of Health R01DK123146