Machine Learning IDs Metabolites in Walnuts Linked with CVD, Diabetes Risk Reduction

Machine learning models were able to tease out some of the metabolite components in walnuts that are potentially linked to reduced risk for type 2 diabetes and cardiovascular disease.

“Walnut consumption is associated with lower risk of type 2 diabetes and cardiovascular disease,” the researchers wrote. “However, it is unknown whether plasma metabolites related to walnut consumption are also associated with lower risk of cardiometabolic diseases.”

The researchers sought to highlight the metabolites associated with the consumption of walnuts and their associations with the risk for type 2 diabetes and cardiovascular disease. The analysis included 1,833 participants from the PREDIMED study who were at high cardiovascular risk at study baseline (57% women in the study population). Of these patients, 1,522 had metabolic data available, comprising an internal validation cohort. The research team performed plasma metabolomics analysis and, using elastic net continuous regression analysis, cross-sectional analyses of the links between 385 known metabolites and the consumption of walnuts. The also performed a 10-cross validation procedure and assessed Pearson correlation coefficients to identify the associations between weighted metabolite models and self-reported walnut consumption. They also used Cox regression models to estimate prospective associations between the study subject metabolite profile and incident type 2 diabetes.

According to the study results, 19 metabolites were associated with walnut consumption, and the association was statistically significant. The metabolites included lipids, purines, acylcarnitines, and amino acids. The 10-cross-validation Pearson correlation coefficients between self-reported walnut consumption and metabolite profile were 0.16 for the discovery population and 0.15 for the validation population. Metabolite profiles were inversely associated with the incidence of type 2 diabetes. The hazard ratio per one-standard deviation of cardiovascular disease incidence was 0.71 (95% CI, 0.60 to 0.85; P<0.001).

“With data-driven technologies, we are able to enhance our understanding of the relationship between diet and disease and take a personalized approach to nutrition which will lead to better prevention and management of various health conditions,” said lead author Dr. Marta Guasch-Ferré, of the Department of Nutrition at Harvard T.H. Chan School of Public Health, and  Instructor in Medicine at Harvard Medical School and Brigham and Women’s Hospital, said in a press release. “In this study, we revealed the unique metabolomic signature of walnuts, which brings us one step closer to understanding ‘how’ walnuts are good for our health. These cutting-edge technologies are shaping the future of nutrition recommendations.”

The study was published in The Journal of Nutrition.