Abstract 14801: Protective Associations of Dairy Foods and Cardiometabolic Markers Differ by Food Type and Fat Content: The Boston Puerto Rican Health Study

Abstract

Introduction: The relation of dairy intake with cardiometabolic (CM) markers remains unclear due to a complex nutrient profile across dairy types. This study aimed to evaluate associations between dairy intake and CM markers among Puerto Rican adults with excessive CM burden. Methods: Baseline and 2-y follow-up data on diet and CM markers were available for 1110 participants aged 46-78y. Diet was assessed with a validated food frequency questionnaire. Dairy groups (total, milk, yogurt, cheese, cream, total with any fat, total without fat) were created from disaggregated foods and presented as servings/d/2000 kcal (s/d). Mean dietary intakes across two visits were used. CM biomarkers were assessed from fasted blood at 2-y and included waist circumference (WC), HDL and LDL cholesterol, TG, insulin, HbA1c, HOMA-IR, homocysteine (Hcy), C-reactive protein (CRP), blood pressure (BP) and Framingham risk score (FRS). Generalized linear and logistic regression were used to determine associations between dairy variables and each CM marker, adjusted for age, sex, BMI, kcal, alcohol and fruit /vegetable intakes, CVD/diabetes medications, language acculturation score, smoking, perceived stress, and physical activity. Multiple comparisons were adjusted by FDR correction. Results: Mean age was 59.3±7.7y, 70% female, 80% with metabolic syndrome. Total dairy intake was 2.3±1.1 s/d; 2.1±1.0 from dairy with fat and 0.2±0.5 without fat. Total dairy, dairy with fat, and milk intakes were associated with lower Hcy (β -0.37, SE 0.12, P =0.0012; β -0.39, SE 0.13, P =0.003; β -0.42, SE 0.13, P =0.002; respectively); non-fat dairy with lower WC (β -1.41, SE 0.44, P =0.001); milk with lower odds of high BP (OR 0.84, P =0.04); cheese with lower TG (β -18.5, SE 9.2, P= 0.04); yogurt with lower HbA1c (β -0.46, SE 0.17, P =0.008); cream with lower HbA1c (β -0.20, SE 0.10, P =0.04) and HOMA-IR (β -1.33, SE 0.56, P =0.02). No associations were seen with cholesterol, CRP, insulin, or TG. After FDR correction, findings remained significant for total dairy, dairy with fat, and milk with Hcy. Conclusions: Protective associations of dairy with specific CM markers vary by type of dairy and fat content. Associations were independent of overall dietary pattern and other risk factors for CM disease.