Diet patterns are associated with progression of coronary artery calcification and differ across race/ethnicity: The Multi-Ethnic Study of Atherosclerosis.

Background: Inference regarding the longitudinal association between diet and coronary artery calcification (CAC) progression is limited, particularly when using empirically derived dietary patterns rather than hypothesis-driven diet indices. Prior studies have demonstrated discordance between calcification propensity and cardiovascular events across racial and ethnic groups, yet the effect of diet on CAC remains unknown. Methods: We analyzed 3182 participants from the Multi-Ethnic Study of Atherosclerosis (MESA) with CAC measurements and completed food frequency questionnaires (FFQ47) at Exams 1 (2000-2002) and 5 (2010-2012). Dietary patterns were identified using principal components analysis. CAC progression was defined by incidence, progression and slow/moderate/fast annual change rate as categorical variables. Associations between baseline dietary patterns, dietary changes, and CAC outcomes were assessed using ordinal regression models and logistic regression models, adjusted for baseline age, sex, race/ethnicity, smoke, diabetes, hypertension, baseline CAC score, total energy intake, education level, estimated glomerular filtration rate, and lipid-lowering medications. Multiplicative interaction terms between each dietary pattern PC and race/ethnicity were introduced in multivariable regression models to assess whether the associations differ by race/ethnicity. Results: Four major dietary patterns were identified: High Fat and Processed Meat; Vegetables and Fish; Whole Grains and Fruit; Beans, Tomatoes, and Refined grains. In multivariable models adjusting for demographic and cardiovascular risk factors, the Whole Grains and Fruit pattern was inversely associated with incident CAC (OR 0.89, 95% CI 0.80-0.99; P = 0.04), whereas the High Fat and Processed Meat pattern was not. For CAC progression, the High Fat and Processed Meat pattern was positively associated (OR 1.20, 95% CI 1.01-1.43; P = 0.03), while the Whole Grains and Fruit pattern was inversely associated (OR 0.87, 95% CI 0.76-0.98; P = 0.03). Similar associations were observed for annual CAC change rate. In ordinal logistic regression models that incorporated death within the CAC progression outcome, the High fat and Processed Meat dietary pattern was associated with higher odds of progression (OR = 1.12; 95% CI: 1.00-1.25; borderline significance, p = 0.05), whereas the Whole Grains and Fruit pattern was associated with lower odds (OR = 0.86; 95% CI: 0.79-0.93). No significant associations were found for the Vegetables and Fish or the Beans, Tomatoes, and Refined Grains patterns. Race/ethnicity stratified analyses showed significant modification effect for the High Fat and Processed Meat and Whole Grains and Fruit patterns. Associations between dietary patterns and CAC progression were evident among White participants but were attenuated among Black participants (P-interaction=0.009 and 0.006, respectively). Conclusions: Diets high in fat and processed meats are associated with increased CAC progression, whereas whole grain and fruit/greens-based diets are associated with lower odds of CAC incidence and progression. Compared with White participants, these associations were attenuated among Black participants.