Ancestry specific associations of a genetic risk score, dietary patterns and metabolic syndrome: a longitudinal ARIC study

Abstract

Background: Associations have been observed among genetic variants, dietary patterns, and metabolic syndrome (MetS). A gap in knowledge is whether a genetic risk score (GRS) and dietary patterns interact to increase MetS risk among African Americans. We investigated whether MetS risk was influenced by interaction between a GRS and dietary patterns among Whites and African Americans. A secondary aim examined if molecular genetic clusterings differed by racial ancestry.

Methods: We used longitudinal data over 4-visits (1987-1998) that included 10,681 participants aged 45-64y at baseline from the Atherosclerosis Risk in Communities study (8451 Whites and 2230 African Americans). We constructed a simple-count GRS as the linear weighted sum of high-risk alleles (0, 1, 2) from cardiovascular disease polymorphisms from the genome-wide association studies catalog associated with MetS risk. Three dietary patterns were determined by factor analysis of food frequency questionnaire data: Western, healthy, and high-fat dairy. MetS was defined according to the 2016 National Cholesterol Education Program Adult Treatment Panel III criteria but used 2017 American Heart Association/American College of Cardiology criteria for elevated blood pressure. Analyses included generalized linear model risk ratios (RR), 95% confidence intervals (CI), and Bonferroni correction for multiple testing.

Results: The Western dietary pattern was associated with higher risk for MetS across increasing GRS tertiles among Whites (p < 0.017). The high-fat dairy pattern was protective against MetS, but its impact was most effective in the lowest GRS tertile in Whites (RR = 0.62; CI: 0.52-0.74) and African Americans (RR = 0.67; CI: 0.49-0.91). Among each racial group within GRS tertiles, the Western dietary pattern was associated with development and cycling of MetS status between visits, and the high-fat dairy pattern with being free from MetS (p < 0.017). The healthy dietary pattern was associated with higher risk of MetS among African Americans which may be explained by higher sucrose intake (p < 0.0001). Fewer genes, but more metabolic pathways for obesity, body fat distribution, and lipid and carbohydrate metabolism were identified in African Americans than Whites. Some polymorphisms were linked to the Western and high-fat dairy patterns.

Conclusion: The influence of dietary patterns on MetS risk appears to differ by genetic predisposition and racial ancestry.

Keywords: Ancestry; Dietary patterns; Genetic risk score; Interaction; Metabolic syndrome; Polymorphism; Race.

Fig. 1

Entry of participants into the study and partitioned for analysis by racial ancestry. ARIC, Atherosclerosis Risk in Communities; MetS, metabolic syndrome; GRS, genetic risk score; PCA, principal components analysis

Fig. 2

Association between dietary patterns and developing MetS or being free from metabolic syndrome among Whites and African Americans. Participants’ MetS status change was in one direction only from visits 1 to 4. Key +  +  +  + versus −  −  −  −: Those with MetS at all 4 visits compared with those without MetS at all 4 visits; − # versus −  −  −  −: Those free of MetS at visit 1 but developed MetS at visit 4 compared with those without MetS at all 4 visits; +  = versus −  −  −  −: Those with MetS at visit 1 but free of MetS at visit 4 compared with those without MetS at all 4 visits. Bold indicates p values that were significant at p < 0.05. Bonferroni adjustment for multiple testing for dietary patterns (p = 0.05/4 = 0.017). Dietary patterns were calculated using factor analysis with the principal components factor option and the varimax rotation with correlation ≥ 0.3. MetS was regressed against the GRS adjusting for a covariate summary score composed of age, gender, sports physical activity (Baecke questionnaire), cigarette smoking status, drinker status, education level at visit 1, time in study, and 20 genetic principle components for admixture. Dietary patterns are from Additional file 1: Table S2. Dietary pattern contents for Whites: Western: fried foods, red meat, chips and fries, chicken with skin, processed meat, eggs, and condiments; Healthy: rice, pasta, vegetables, mashed potato, chicken without skin, lentils and beans; High-fat dairy: butter, whole milk, eggs. Dietary pattern contents for African Americans: Western: Eggs, processed meat, biscuit and cornbread, whole wheat bread, fried foods, white bread, and margarine-butter; Healthy: Chicken without skin, vegetables, lentils and beans, fruit, cooked breakfast cereal, fish, mashed potato, shellfish, cold breakfast cereal; High-fat dairy: Butter, margarine-butter, whole milk, cottage cheese

Fig. 3

Interactions between a GRS and dietary patterns for developing MetS or being free from metabolic syndrome among Whites and African Americans. Participants’ MetS status change was in one direction only from visits 1 to 4. Key: +  +  +  + versus −  −  −  −: Those with MetS at all 4 visits compared with those without MetS at all 4 visits; − # versus −  −  −  −: Those free of MetS at visit 1 but developed MetS at visit 4 compared with those without MetS at all 4 visits; +  = versus −  −  −  −: Those with MetS at visit 1 but free of MetS at visit 4 compared with those without MetS at all 4 visits. Bold indicates p values that were significant at p < 0.05. Bonferroni adjustment for multiple testing for dietary patterns (p = 0.05/3 = 0.017). Dietary patterns were calculated using factor analysis with the principal components factor option and the varimax rotation with correlation ≥ 0.3. MetS was regressed against the GRS adjusting for a covariate summary score composed of age, gender, sports physical activity (Baecke questionnaire), cigarette smoking status, drinker status, education level at visit 1, time in study, and 20 genetic principle components for admixture

Fig. 4

Molecular genetic clustering pathways for Whites. Molecular genetic clustering pathway analysis was performed using Literature Lab™ clustering software to find functional relationship differences among the genes by racial ancestry

Fig. 5

Molecular genetic clustering pathways for African Americans. Molecular clustering genetic pathway analysis was performed using Literature Lab™ clustering software to find functional relationship differences among the genes by racial ancestry