Development of a comprehensive food glycomic database and its application: Associations between dietary carbohydrates and insulin resistance

Authors: SUAREZ, CHRISTOPHER - University Of California, Davis; CHEANG, SHAWN - University Of California, Davis; Larke, Jules; JIANG, JIANI - University Of California, Davis; WENG, CHENG-YU - University Of California, Davis; STACY, AARON - University Of California, Davis; COUTURE, GARRET - University Of California, Davis; CHEN, YE - University Of California, Davis; BACALZO, NIKITA - University Of California, Davis; SMILOWITZ, JENNIFER - University Of California, Davis; GERMAN, J - University Of California, Davis; MILLS, DAVID - University Of California, Davis; Lemay, Danielle; LEBRILLA, CARLITO - University Of California, Davis

Submitted to: Food Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/17/2025
Publication Date: 1/20/2025
Citation: Suarez, C., Cheang, S.E., Larke, J.A., Jiang, J., Weng, C.C., Stacy, A., Couture, G., Chen, Y., Bacalzo, Jr., N.P., Smilowitz, J.T., German, J.B., Mills, D.A., Lemay, D.G., Lebrilla, C.B. 2025. Development of a comprehensive food glycomic database and its application: Associations between dietary carbohydrates and insulin resistance. Food Chemistry. 473. Article 142977. https://doi.org/10.1016/j.foodchem.2025.142977.
DOI: https://doi.org/10.1016/j.foodchem.2025.142977

Interpretive Summary: Detailed information about carbohydrates in food has not previously existed beyond estimates of total fiber, soluble fiber, total carbohydrates, and simple sugars. Using advanced analytical chemistry, we produced a state-of-the-art glycan encyclopedia of food, Davis Food Glycopedia 2.0 (DFG2). This new food composition database includes monosaccharides, oligosaccharides, glycosidic linkages, and polysaccharides for 250 of the most commonly consumed foods in America. To demonstrate the utility of this new resource, dietary records from NHANES were mapped to foods to estimate quantities of glycans consumed by individuals. We then developed models to predict insulin resistance and found that glycan features were better predictors than traditional markers of carbohydrate intake, such as dietary fiber, simple sugars, and total carbohydrates.

Technical Abstract: The critical role of carbohydrates in the human diet has been extensively reported. While it is generally accepted that carbohydrates are an integral part of a healthy diet, the types and amounts consumed are a source of considerable disagreement amongst nutritional and medical practitioners. These disputes likely arise from an ignorance of their chemical structures. Monosaccharides, disaccharides, oligosaccharides, and polysaccharides each possess inherent biological activities based on their unique structural properties such as glycosidic linkage connectivity, degree of polymerization, and monomeric composition. However, upon close inspection of the commonly encountered nutrition label, this precision and nuance is nowhere to be found. The vast biological and structural complexities of dietary carbohydrates are distilled into vague terms such as total carbohydrates, sugars, starches, and soluble/insoluble fiber. To address this critical gap, we present a state-of-the-art glycan encyclopedia of food, Davis Food Glycopedia 2.0 (DFG2), that utilizes multi-glycomic analysis. Dietary carbohydrates, including monosaccharides, oligosaccharides, glycosidic linkages, and polysaccharides were determined for 250 of the most commonly consumed foods in America. To demonstrate the utility of the encyclopedia, the library was mapped to dietary records from the NHANES to predict insulin resistance (IR) from glycan features alongside traditional markers of carbohydrate intake (dietary fiber, simple sugars, and total carbohydrates). Several features from the DFG2 were more powerful predictors compared to traditional measures of carbohydrates indicating the need for molecular information in guiding precision nutrition