Plasma Metabolite Associations with Incident Heart Failure with Reduced and Preserved Ejection Fraction

Background Prior studies on metabolite associations with incident heart failure (HF) used billing code-based definitions and lacked the data on left ventricular ejection fraction needed to determine associations with HF with reduced (HFrEF) and preserved (HFpEF) ejection fraction. Objectives Identify potentially causal plasma metabolite associations with HFrEF and HFpEF, ascertained using a validated machine learning- and natural language processing-based algorithm, in up to 38,000 individuals. Methods We included MGB Biobank participants who had available metabolomics data and no history of HF at baseline. The primary exposures were plasma levels of 42 metabolites measured using a H 1 nuclear magnetic resonance platform. The primary outcome was incident HF, ascertained by a validated machine learning- and natural language processing-based algorithm. Multivariable Cox proportional hazards regression to quantify the associations between a 1-SD difference in metabolite level and the time to incident HF. Mendelian randomization analysis was used test the potential causality of each metabolite-HF association. Results The final analytical cohort included 38,628 individuals with a mean age of 63 years (56% women). Higher plasma levels of glutamine associated with a higher risk of incident HF (HR [95% CI]: 1.21 [1.07-1.35]) while higher levels of docosahexaenoic acid (an omega-3 fatty acid) (0.85 [0.75-0.95]), phosphatidylcholines (0.85 [0.75-0.97]), phosphoglycerides (0.86 [0.76- 0.97]) and total cholines (0.85 [0.75-0.97]) associated with a lower HF risk. Docosahexaenoic acid and total omega-3 fatty acids associated with HFpEF. Associations with HFpEF tended to be stronger than with HFrEF for several fatty acids, including omega-3 fatty acids, in individuals with obesity, but not coronary artery disease or diabetes. Mendelian randomization analysis supported causal associations between higher levels of docosahexaenoic acid and omega-3 fatty acids with a higher risk of HF and greater left ventricular mass index. Conclusions Dysregulated omega-3 fatty acid metabolism may be causally associated with a higher risk of incident HFpEF. Condensed Abstract We investigated metabolic contributors to incident heart failure (HF) in the MGB Biobank, leveraging metabolomics and a validated machine learning/NLP algorithm for HF ascertainment. Among 38,628 participants without HF at baseline, plasma levels of 42 metabolites were evaluated using multivariable Cox models and Mendelian randomization. Higher glutamine levels were associated with increased HF risk, while docosahexaenoic acid (DHA), phosphatidylcholines, phosphoglycerides, and total cholines were inversely associated. DHA and total omega-3 fatty acids were linked to HFpEF. Mendelian randomization suggested causal associations between higher DHA and omega-3 fatty acid levels and elevated HF risk, implicating dysregulated omega-3 metabolism in HF development.