Table 1.

Summary of population-based studies assessing association of Mediterranean diet and cardiometabolic outcomes in minority racial/ethnic populations in the USA.

Study and Reference	Mediterranean Diet Construct	Race/Ethnicities	Sample Size and Age	Key Findings
Multi-Ethnic Study of Atherosclerosis (MESA) [32]	Sum of population sex-specific median of 9 groups: vegetables; whole grains; nuts; legumes; fruits; MUFA:SFA; red and processed meat; dairy; fish; alcohol	Whites, African Americans, Hispanics, and Chinese	45–84 years n = 5390	Higher MedDiet score was associated with lower baseline mean insulin levels and lower glucose levels but was not significantly associated with diabetes risk. No significant interaction by race.
Multi-Ethnic Study of Atherosclerosis (MESA) [33]	Sum of population sex-specific median of 9 groups: vegetables; whole grains; nuts; legumes; fruits; MUFA:SFA; red and processed meat; dairy; fish; alcohol	Whites, African Americans, Hispanics, and Chinese	45–84 years n = 4497	MedDiet associated with modestly better left ventricular (LV) structure and function (For each +1-U difference in score: LV volume was 0.4 (95% CI: 0.0, 0.8 mL) higher, the stroke volume was 0.5 (95% CI: 0.2, 0.8 mL) higher, and the ejection fraction was 0.2 percentage points (95% CI: 0.1, 0.3) higher. The study did not stratify by race/ethnicity nor reported testing for interaction.
Coronary Artery Risk Development in Young Adults (CARDIA) [34]	Sum of population median of: whole grains, fruit, vegetables, fruit and vegetable juice, legumes, nuts, poultry, fish, eggs, coffee and tea, MUFA+PUFA:SFA, red and processed meat, dairy products, fried vegetables, refined grain, sauces, snack foods, sugar-sweetened beverages, diet beverages, alcohol	African Americans and whites	18–30 years n = 4713	For the overall cohort, hazard ratio for metabolic syndrome 0.67 (0.49, 0.90) vs. 0.82 (0.67, 1.01), as well as incidence of its components of abdominal obesity (41.9 vs. 59.4%), elevated triglycerides (21.6 vs. 37.3%), and low HDL-C (59.3 vs. 68.4%), was better in those with higher MedDiet scores (top quintile) compared to lower scores (lowest quintile). No significant interaction by race.
Multiethnic Study (MEC) [35]	Sum of population median of 10 groups: vegetables without potatoes, fruits, whole grains, nuts, legumes, fish, red and processed meat, alcohol consumption, MUFA:SFA	Whites, Native Hawaiians, and Japanese Americans living in Hawaii and California	45–75 years n = 89,185	Higher adherence to MedDiet was related to a 13–28% lower risk of T2D in white participants but not in other ethnic groups (HR (95% CI): 0.90 (0.84, 0.95) white men; 0.95 (0.88, 1.03) Native Hawaiian men; 0.98 (0.93, 1.02) Japanese American men; 0.93 (0.86, 1.00) white women; 0.97 (0.90, 1.05) Native Hawaiian women; 1.00 (0.95, 1.05) Japanese American women).
Multiethnic Study (MEC) [36]	Sum of population median of 10 groups: vegetables without potatoes, fruits, whole grains, nuts, legumes, fish, red and processed meat, alcohol consumption, MUFA:SFA	White, African Americans, Native Hawaiians, Japanese Americans and Latinos	45–75 years n = 215,782	The MedDiet was associated with lower risk of CVD mortality only in whites participants (0.70 (0.59, 0.84) in men, 0.77 (0.62, 0.95) in women); African American men [HR (95% CI) 0.75 (0.62, 0.90)], and women [0.82 (0.70, 0.97)] and Japanese American men [0.80 (0.68, 0.94)] and women [0.72 (0.59, 0.87)] but not for in Latino or Native Hawaiian men and women.
Boston Puerto Rican Health Study [20]	Sum of sex-specific energy-adjusted population median for 9 components: vegetables, fruits, whole grains, nuts and legumes, meat, fish, dairy products, MUFA: SFA, and alcohol	Puerto Ricans living in Boston	45–75 years n = 1194	A higher MedDiet score was associated with 2-years lower waist circumference (β coefficient ± SE: −0.52 ± 0.26); BMI (−0.23 ± 0.08); log-insulin (−0.06 ± 0.02); log-homeostasis model assessment of insulin resistance (−0.05 ± 0.02), and log-C-reactive protein (−0.13 ± 0.03). Traditional foods consumed at high MedDiet included vegetables (e.g., root crops, green bananas) and meats in homemade soups, orange juice, oatmeal, beans, legumes, fish (e.g., cod, canned tuna), whole milk, corn oil, beer.
Racial Differences in Stroke (REGARDS) [37]	Sum of population-based quintiles of 11 components: vegetables, fruits, lean meats, fish, nuts, MUFA:SFA, red and processed meats, sodium, dairy foods, grains and starches, and alcohol	Black and white men and women	≥4 years n = 21,423	Compared with those in the lowest MedDiet score quintile, participants with the highest MedDiet adherence had 32% (95% CI: 47%, 12%) lower risk of CVD mortality after 6.25 years of follow-up. The associations were similar across race.
Northern Manhattan Study (NOMAS) [25]	Sum of sex-specific energy-adjusted population median for 9 components: fruits and nuts, vegetables, legumes, cereals and grains, fish, meat, dairy products, MUFA:SFA alcohol	Hispanics, non-Hispanic Blacks, and non-Hispanic whites from New York city	>40 years n = 2568	Compared with those in the first MedDiet score quintile, participants in the top quintile had 28% (95% CI: 4%, 46%) lower risk of the combined vascular events (ischemic stroke, myocardial infarction, and vascular death). Diet only was inversely associated with vascular death only. No significant interaction by race-ethnicity.
Northern Manhattan Study (NOMAS) [38]	Sum of sex-specific energy-adjusted population median for 9 components: fruits and nuts, vegetables, legumes, cereals and grains, fish, meat, dairy products, MUFA:SFA, alcohol	Hispanics, non-Hispanic Blacks, and non-Hispanic whites from New York city	>40 years n = 1937	Greater adherence to MedDiet was associated with lower left ventricular mass (1.98 g lower per 1-point of the diet score). Non-significant interactions by race/ethnicity.
Northern Manhattan Study (NOMAS) [39]	Sum of sex-specific energy-adjusted population median for 9 components: fruits and nuts, vegetables, legumes, cereals and grains, fish, meat, dairy products, MUFA:SFA, alcohol	Hispanics, non-Hispanic Blacks, and non-Hispanic whites from New York city	>40 years n = 1374	MedDiet was not associated with carotid intima media thicknesses in the whole multi-ethnic cohort. No association between MedDiet and plaque thickness nor area in African Americans or whites. For Hispanics, an inverse association was found between MedDiet adherence and the 75th percentile of plaque thickness (beta-coefficient, (95% CI): −0.0906 (−0.1541, −0.0271) change in mm).
Washington Heights-Inwood Community Aging Project (WHICAP) [40]	Sum of sex-specific energy-adjusted population median for 9 components: fruits and nuts, vegetables, legumes, cereals, fish, meat, dairy products, MUFA:SFA, alcohol	White, Hispanics, and African Americans living in Manhattan	≥ 65 years n = 4308	MedDiet score was associated with leukocyte telomere length only in whites (β = 48.3) but not among Hispanics or African Americans.

Note: All studies used a validated self-administered food frequency questionnaire (FFQ) for dietary assessment except CARDIA (Coronary Artery Risk Development in Young Adults) that used a diet history questionnaire. The following FFQ were used: MESA (Multi-Ethnic Study of Atherosclerosis): 127-item FFQ; MEC (Multiethnic Study): a quantitative FFQ; BPRHS (Boston Puerto Rican Health Study): a semi-quantitative FFQ; REGARDS: a Block FFQ; NOMAS (Northern Manhattan Study): a modified Block National Cancer Institute FFQ; WHICAP (Washington Heights-Inwood Community Aging Project): Willett’s semi-quantitative FFQ. Abbreviations: MedDiet: Mediterranean Diet; MUFA: Monounsaturated fats; PUFA: Polyunsaturated fats; SFA: Saturated fats; LV: Left ventricular; HR: Hazard Ratio; CVD: Cardiovascular disease.