Identification of Gene-Diet Interactions Impacting Glycemic Biomarkers in the Multi-Ethnic TOPMed Cohorts. (7th June 2021)

• Record Type: Journal Article
• Authors: Westerman, Kenneth; Walker, Maura; Merino, Jordi; Manning, Alisa
• Is Part Of: Current developments in nutrition. Volume 5(2021)Supplement 2
• Page Start: 952
• Page End: 952
• Publication Date: 2021-06-07
• Subjects: Nutrition -- Periodicals; Nutritional Physiological Phenomena; Nutrition; Periodicals; Periodicals; Fulltext; Internet Resources; Periodicals; 612.3
• DOI: 10.1093/cdn/nzab050_019
• Languages: English
• ISSNs: 2475-2991

Abstract

Abstract: Objectives: Identification of robust gene-diet interactions impacting cardiometabolic traits has been limited due to low statistical power and poor replication across populations. Emerging statistical methods increase power by simultaneously testing genetic interactions with multiple exposures, an especially appealing strategy for complex and highly-correlated dietary traits. Furthermore, meta-analysis across ancestrally and behaviorally diverse populations may allow for more robust discoveries. Here, our objective was to leverage multi-exposure interaction tests in a diverse set of cohorts to identify interactions involving macronutrient ratios and impacting multiple biomarkers of glycemia. Methods: Seven cohorts from the TOPMed consortium (total N ∼ 20, 000) contributed whole-genome sequencing data, self-reported dietary data, and glycemic trait measurements (fasting glucose and insulin (FG and FI) and hemoglobin A1c). Three macronutrient ratios were defined to model realistic dietary exchanges and minimize collinearity: carbohydrate: fat, polyunsaturated: saturated fat, and fiber: carbohydrate. For each glycemic trait outcome and common genetic variant, we fit a model including all three ratios and their genetic interactions, with joint significance testing in each cohort followed by cross-cohort meta-analysis. Results: Four variants showed promising sub-threshold signals (p < 1 × 10 −7 ), though none reached genome-wide significance. For example, diet ratios collectively interacted with rs2276620 in the oxysterol-binding OSBPL6 gene to influence FG (p = 6.5 × 10 −8 ) and rs114448070 in the beta cell function-associated KAT2B gene (p = 5.4 × 10 −8 ) to influence FI. These associations were contributed to by multiple cohorts and multiple dietary factors, emphasizing the value of this multi-cohort and multi-exposure approach to GDI variant detection. Conclusions: Our approach takes advantage of population diversity and multi-faceted dietary exposures to understand genetic effects on the diet-glycemia relationship and support the development of genome-based precision nutrition. Our results will be updated with additional cohorts and gene-based tests using rare genetic variants, which are less common but may have stronger effects. Funding Sources: KEW is supported by an NIDDK T32 award.