| Title: | Plasma metabolomics uncovers key metabolites and enhances prediction of aortic stenosis: insights from prospective cohort and Mendelian randomisation studies |
| Journal: | Heart |
| Published: | 4 Aug 2025 |
| Pubmed: | https://pubmed.ncbi.nlm.nih.gov/40759471/ |
| DOI: | https://doi.org/10.1136/heartjnl-2025-326161 |
Publication 16033
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
OBJECTIVE: To examine the association between nuclear magnetic resonance (NMR)-based metabolomics and aortic stenosis (AS) risk, and determine whether metabolomic profiling can enhance AS prediction beyond conventional clinical risk factors.</p>
METHODS: We included 168 metabolites in our study. The primary outcome of interest was incident AS. The secondary outcome of interest was incident AS-related interventions or deaths. Prospective and two-sample Mendelian randomisation (MR) analyses were used to investigate the relationship between metabolites and AS risk. Harrell's C index, net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were employed to assess whether adding metabolites could improve risk discrimination and reclassification for AS prediction beyond clinical risk factors.</p>
RESULTS: During a median follow-up of 13.7 years, 2562 AS events, along with 1039 AS-related interventions or deaths, were recorded. In observational analyses, 77 of 168 metabolites were significantly associated with AS risk after adjusting for covariates and correcting for multiple testing. In MR analyses, 39 metabolites, including apolipoprotein B, apolipoprotein B-containing lipoprotein particles and lipid constituents within lipoprotein subclasses, remained causally associated with AS risk. Besides, all of the 39 metabolites were associated with incident AS-related interventions or deaths. Using LASSO (least absolute shrinkage and selection operator) regression, we identified 12 metabolites in the training set (n=146 252). Adding the 12 metabolites to clinical risk factors-based Cox regression model significantly improved AS risk prediction. The C index improved from 0.780 to 0.791 in the testing set (n=62 678) and from 0.764 to 0.786 in the external validation cohort (n=21 546). The NRI and IDI were also significantly improved.</p>
CONCLUSIONS: Our study identified 39 metabolites with causal associations with AS risk, particularly apolipoprotein B-containing lipoproteins. The derived 12-metabolite panel significantly improved AS risk prediction, demonstrating the potential of NMR-based metabolomics to enhance AS risk stratification and targeted prevention.</p>
10 Authors
- Hao Xiang
- Yanjun Zhang
- Yuanyuan Zhang
- Ziliang Ye
- Sisi Yang
- Xiaoqin Gan
- Yu Huang
- Yiting Wu
- Yiwei Zhang
- Xianhui Qin
Enabling scientific discoveries that improve human health