Plasma Metabolite Profiles for Vascular Aging Biomarker Development

2005–2013 data from two European cohorts comprises plasma metabolomic profiles for 719 Flemish and 811 Spanish participants. Researchers used nuclear magnetic resonance to identify 38 plasma metabolites and trained an elastic net regression model (pMTB-age) using 18 metabolites to predict biological age. The dataset was published by figshare admin karger in 2024.

Use Cases

• Predict biological age (pMTB-age) from a panel of 18 plasma metabolites, including six amino-acids, using elastic net regression.
• Analyze the association between specific metabolite levels (e.g., glycine, serine, threonine) and cardiovascular complications or mortality over a 12.3-18.8 year follow-up.
• Assess feature importance of the 18 retained metabolites for age prediction using SHAP (SHapley Additive exPlanation) values.
• Validate a metabolomic signature of vascular aging (pMTB-age-R) by correlating it with cardiovascular risk factors across distinct European populations.

Strengths

• Data from two independent cohorts provides internal (N=580/719) and external (N=811) replication.
• Longitudinal follow-up for clinical outcomes spans a median of 12.3 years in Flemish and 18.8 years in Spanish cohorts.
• The derived pMTB-age model explained 22.9% to 28.6% of chronological age variance in the data.

Limitations

• The dataset is small (3.0 MB), suggesting it contains summary or model results rather than raw individual-level data.
• Specific metabolite concentration values and individual participant data are not confirmed to be included.
• Geographic coverage is limited to two European cohorts, potentially limiting generalizability.

Provenance

Source

figshare admin karger

Collection Method

Nuclear magnetic resonance (NMR) profiling of 38 plasma metabolites from blood samples.

Time Range

2005–2013 (Flemish cohort: 2005–2010 discovery, 2009–2013 replication; Spanish cohort: 2001).

Freshness

Last updated in April 2026.

Geography

Flemish (Belgium) and Spanish (HORTEGA cohort) populations.