Abstract
OBJECTIVE: This study evaluated differences in diet quality across multiple racial/ethnic groups with type 2 diabetes (T2D) experiencing food insecurity. DESIGN: Cross-sectional. Bilingual (English/Spanish, English/Marshallese) community health workers used a multiple-pass approach to collect 3 24-hour dietary recalls via phone. SETTING: Food pantries in Arkansas. PARTICIPANTS: Eighty-three adult food pantry clients with T2D (hemoglobin A1c ≥ 7.0%) experiencing food insecurity and self-identified race/ethnicity of Hispanic (39%), Marshallese Pacific Islander (i.e., Marshallese) (35%), or White (26%). MAIN OUTCOME MEASURE: Diet quality (Healthy Eating Index [HEI]-2015). ANALYSIS: Differences in diet quality scores by race/ethnicity were evaluated using one-way analysis of variance; Turkey-Kramer tests for pairwise comparisons assessed diet quality differences among Hispanic, Marshallese, and White participants. A food-level analysis identified top food categories contributing to diet quality. RESULTS: Diet quality was 71.7 ± 12.7 (mean ± standard deviation) among Hispanic, 56.9 ± 14.1 among Marshallese, and 45.9 ± 14.1 among White participants. Hispanic participants had a significantly higher HEI-2015 score than Marshallese (mean difference, 14.9 [95% confidence interval (CI), 6.6-23.2]; P < 0.001) and White (mean difference, 25.9 [95% CI, 16.9-34.8]; P < 0.001) participants. Marshallese participants had a significantly higher HEI-2015 score than White participants (mean difference, 11.0 [95% CI, 1.8-20.1]; P = 0.01). CONCLUSIONS AND IMPLICATIONS: Designers of nutrition-focused interventions may consider tailoring food and education around high and low diet quality component scores that vary across racial/ethnic groups (e.g., including culturally appropriate fruits/vegetables) to improve T2D management.