Abstract
BACKGROUND: Whether differences in circulating long chain acylcarnitines (LCAC) are seen in heart failure (HF) patients with and without diabetes mellitus (DM), and whether these biomarkers report on exercise capacity and clinical outcomes, remains unknown. The objective of the current study was to use metabolomic profiling to identify biomarkers that report on exercise capacity, clinical outcomes, and differential response to exercise in HF patients with and without DM. METHODS: Targeted mass spectrometry was used to quantify metabolites in plasma from participants in the heart failure: a controlled trial investigating outcomes of exercise training (HF-ACTION) trial. Principal components analysis was used to identify 12 uncorrelated factors. The association between metabolite factors, diabetes status, exercise capacity, and time to the primary clinical outcome of all-cause mortality or all-cause hospitalization was assessed. RESULTS: A total of 664 participants were included: 359 (54%) with DM. LCAC factor levels were associated with baseline exercise capacity as measured by peak oxygen consumption (beta 0.86, p = 2 × 10(-7), and were differentially associated in participants with and without DM (beta 1.58, p = 8 × 10(-8) vs. 0.67, p = 9 × 10(-4), respectively; p value for interaction = 0.012). LCAC levels changed to a lesser extent in participants with DM after exercise (mean ∆ 0.09, p = 0.24) than in those without DM (mean ∆ 0.16, p = 0.08). In univariate and multivariate modeling, LCAC factor levels were associated with time to the primary outcome (multivariate HR 0.80, p = 2.74 × 10(-8)), and were more strongly linked to outcomes in diabetic participants (HR 0.64, p = 3.21 × 10(-9) v. HR 0.90, p = 0.104, p value for interaction = 0.001). When analysis was performed at the level of individual metabolites, C16, C16:1, C18, and C18:1 had the greatest associations with both exercise capacity and outcomes, with higher levels associated with worse outcomes. Similar associations with time to the primary clinical outcome were not found in a control group of patients without HF from the CATHeterization GENetics (CATHGEN) study. CONCLUSIONS: LCAC biomarkers are associated with exercise status and clinical outcomes differentially in HF patients with and without DM. Impaired fatty acid substrate utilization and mitochondrial dysfunction both at the level of the skeletal muscle and the myocardium may explain the decreased exercise capacity, attenuated response to exercise training, and poor clinical outcomes seen in patients with HF and DM. Trial Registration clinicaltrials.gov Identifier: NCT00047437.