Cross-sectional evaluation of the metabolic vulnerability index in heart failure populations

TL;DR

This study compares the Metabolic Vulnerability Index (MVX) in heart failure patients from a clinical trial and a community cohort, finding differences in its distribution and clinical correlations.

Contribution

The study introduces new insights into the MVX score's variability and clinical relevance across different heart failure populations.

Findings

The median MVX score was lower in the clinical trial compared to the community cohort.

Male sex and hyperlipidemia were associated with higher MVX scores in the clinical trial.

NT-proBNP levels were significantly linked to MVX scores in the community cohort.

Abstract

The Metabolic Vulnerability Index (MVX) is a novel multi-marker risk score derived from nuclear magnetic resonance (NMR) measures and has shown predictive value for mortality in heart failure. Hence, we aimed to evaluate the distribution of MVX and its clinical correlates within a clinical trial population and a comparable subpopulation of patients with heart failure with reduced ejection fraction and ischemic heart disease within a community cohort.

We studied a clinical trial (2016–2018) and a community cohort (2003–2012), matched based on ejection fraction category and presence of ischemic heart failure. NMR LipoProfile analyses of plasma from both populations provided measures of valine, leucine, isoleucine, citrate, GlycA, and small high-density lipoprotein particles used to compute sex-specific MVX scores. Univariable and multivariable regression models assessed the relationship between MVX (modeled continuously), and selected demographic and clinical covariates.

Clinical trial patients (N = 101, median age: 63, 93% male, median EF: 28%) were younger and predominantly male compared to the cohort (N = 288, median age: 75, 70% male, median EF: 30%). The median MVX score was lower in the clinical trial (50, 42–61) compared to the cohort (66, 58–73). Male sex and hyperlipidemia were linked to higher MVX scores in the clinical trial, while obesity and NT-proBNP were linked to lower and higher MVX scores, respectively, in the cohort (p <.05). After adjusting for significant covariates from univariable analyses and age in multivariable analyses, only the associations between male sex and MVX scores in the clinical trial, and NT-proBNP levels with MVX in the cohort remained significant.

This study highlights significant differences in MVX distribution and its clinical correlates between a clinical trial and a community cohort despite matched heart failure subtypes. These findings have important implications for interpreting and applying the score in diverse study settings.

Figures