Abstract P207: Blood Metabolomic Signatures of 24-hr Urinary Sodium and Potassium Excretion Were Associated With Cardiometabolic Diseases in the Hispanic Community Health Study/Study of Latinos

Abstract

Background: 24hr urinary sodium (Na) potassium (K) excretions are considered as accurate measures of sodium and potassium intake, but data are limited in large population-based studies due to the difficulties of 24hr urine sample collection. Hypothesis: 24hr urinary Na and K excretion can be predicated by blood metabolomic signatures, which are associated with the risk of cardiometabolic diseases. Methods: We assessed 24hr urinary Na, K excretion and sodium/potassium ratio (Na/K) in 447 apparently healthy individuals from the SOL-Nutrition & Physical Activity Assessment Study (SOLNAS) of the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). We calculated urinary Na, K excretion, and Na/K metabolite scores by using the Least Absolute Shrinkage and Selection Operator regression on 199 serum metabolites in SOLNAS, and evaluated their associations with incident CVD (heart failure, stroke, myocardial infarction), type 2 diabetes (T2D), and hypertension (HTN) in up to 5738 individuals in HCHS/SOL. Results: There were 41, 22, and 71 metabolites selected to be predictive of urinary Na, K excretion, and Na/K, respectively. Metabolite scores were significantly correlated with measured urinary Na (r=0.35, p < 5e -14 ), K (r=0.58, p< 5e -50 ) excretion and Na/K (r=0.6, p< 5e -50 ). After multivariate adjustment, metabolite scores of urinary Na excretion and Na/K were significantly associated with incident CVD risk (hazard ratio=2.6 for Na and 1.91 for Na/K, per 1 SD increase in score), incident T2D (relative risk [RR]=1.33 for Na and 1.16 for Na/K), and incident HTN (RR=1.53 for Na and 1.35 for Na/K) over a median 7.6 years of follow-up. The metabolite score of urinary K excretion was not associated with CMD risk. Conclusion: In US Hispanics/Latinos, higher levels of 24hr urinary Na excretion and Na/K indicated by blood metabolomics signatures were associated with increased risk of CVD, HTN, and T2D.