Serum metabolomics improves risk stratification for incident heart failure

Area of Science:

• Biochemistry
• Cardiovascular Medicine
• Biomarker Discovery

Background:

• Heart failure (HF) poses a significant global health burden, impacting quality of life, survival rates, and healthcare costs.
• Early prediction and detection of HF are critical for timely intervention and improved patient outcomes.

Purpose of the Study:

• To investigate the predictive capability of serum metabolomics, specifically 168 metabolites measured by proton nuclear magnetic resonance (¹H-NMR) spectroscopy, for incident heart failure (HF).
• To compare the performance of metabolomics-based prediction models against a well-established clinical risk score.

Main Methods:

• Utilized data from 68,311 UK Biobank participants with over 0.8 million person-years of follow-up.
• Employed Cox proportional hazards models for individual metabolite associations and elastic net models for HF prediction using serum metabolomics.
• Benchmarked against the Pooled Cohort Equations to Prevent HF (PCP-HF) clinical risk score.

Main Results:

• Several metabolites demonstrated independent associations with incident HF, even after adjusting for clinical factors.
• Metabolomics-based risk models showed improved predictive performance when added to the PCP-HF score (e.g., increased Harrell's C-index and net reclassification improvement).
• Models incorporating age, sex, and metabolomics achieved predictive power comparable to existing clinical models.

Conclusions:

• Serum metabolomics significantly enhances the prediction of incident HF risk compared to the PCP-HF score alone.
• Metabolomics-based risk scores, using age and sex, offer a comparable predictive capacity to clinical scores, presenting a potentially scalable and cost-effective alternative for HF risk assessment.