Abstract P1108: Plasma metabolome of a cholesterol-lowering plant-based diet and cardiometabolic disease risk

Introduction: The portfolio diet, a cholesterol-lowering plant-based diet, is recommended for lowering low-density lipoprotein cholesterol (LDL-C), however, less is known about its long-term impact on cardiometabolic disease. Objective metabolic response to diet using metabolomics can provide additional insight into how diet affects disease risk. Objective: To investigate whether a metabolomic signature can characterize adherence to the portfolio diet and whether this signature is associated with cardiometabolic disease. Methods and Results: The metabolomic signature discovery cohort included 633 plasma samples from four previously conducted randomized controlled trials (RCTs) on the portfolio diet. The validation cohorts included 11,985 participants from the Nurses’ Health Studies (NHS) I&II and Health Professionals Follow-up Study (HPFS). Adherence to the portfolio diet was assessed using a portfolio diet score, and the plasma metabolome was profiled by liquid chromatography-tandem mass spectrometry. Using elastic net regression, a metabolomic signature of 56 metabolites was found, most of which were amino acids and amines (34%), glycerophosphocholines (14%), glycerophosphoethanolamines (7%), and indoles (7%). The signature was highly correlated with portfolio diet adherence in the RCTs (r=0.87, p<0.001). In the NHS/HPFS cohorts, a simplified signature showed an attenuated relationship with the portfolio diet score (r=0.26, p<0.001). In the RCTs, using linear mixed models, the signature was associated with lower total cholesterol, LDL-C, non-high-density lipoprotein cholesterol (non-HDL-C), apolipoprotein B, triglycerides, and systolic blood pressure. In the NHS/HPFS cohorts, using multivariable linear regression, the signature was associated with lower total cholesterol, LDL-C, non-HDL-C, and HDL-C. In multivariable Cox regressions in the NHS/HPFS cohorts, the signature was inversely associated with total mortality (HR: 0.94 [95% CI: 0.89, 0.99]) and type 2 diabetes (0.92 [0.84, 0.99]), but not cardiovascular disease (0.96 [0.86, 1.08]). Conclusions: A metabolomic signature that reflects adherence and metabolic response to the cholesterol-lowering plant-based portfolio diet was identified, which was then associated with several cardiovascular disease risk factors in both the RCTs and cohorts, and lower risk of mortality and type 2 diabetes in the cohorts. Further research on this dietary pattern and replication in other cohort studies is warranted.