Metabolomic profiling refines cardiovascular risk stratification beyond SCORE2-diabetes in patients with concurrent type 2 diabetes and pulmonary dysfunction
Aug 2026· Frontiers in Endocrinology· Vol 17· 0 citations· 63 references
Medicine
TL;DR
In participants with T2D and LFI, adding a metabolomic signature to SCORE2-Diabetes produced a statistically significant but moderate improvement in risk discrimination, which supports further evaluation in independent cohorts and prospective impact studies.
Abstract
Background Patients presenting with both type 2 diabetes (T2D) and lung function impairment (LFI) are exposed to an amplified risk of major adverse cardiovascular events (MACE). Although the SCORE2-Diabetes algorithm is widely endorsed for clinical risk assessment, its prognostic accuracy within this specific multimorbid phenotype remains inadequately explored. This study sought to evaluate the baseline utility of SCORE2-Diabetes and to investigate whether integrating a novel, machine learning-derived metabolomic signature could optimize 10-year MACE prediction in this highly vulnerable population. Methods We conducted a prospective analysis of UK Biobank participants with type 2 diabetes (T2D) and no cardiovascular disease at baseline. A reference cohort of 12,130 participants with preserved lung function was used to assess the base clinical model, and the primary cohort comprised 3,706 participants with lung function impairment (LFI). We profiled 249 plasma metabolites using nuclear magnetic resonance (NMR) spectroscopy. A machine-learning framework combining least absolute shrinkage and selection operator (LASSO) Cox regression, random forest survival analysis, and extreme gradient boosting (XGBoost) was used to derive a consensus signature. Internal validation used 1,000 bootstrap resamples; the complete modeling pipeline was repeated within each resample. Incremental performance beyond SCORE2-Diabetes was assessed using Harrell’s concordance index (C-index), net reclassification improvement (NRI), calibration, and decision curve analysis (DCA). Results During a median follow-up of 12 years, SCORE2-Diabetes showed lower discrimination in the LFI cohort than in the reference cohort (C-index, 0.670 vs. 0.707). Among the 3,706 participants with LFI, 845 developed major adverse cardiovascular events (MACE). Adding the 12-metabolite signature increased the C-index from 0.670 to 0.722 (absolute difference, 0.052; P < 0.001), representing a statistically significant but moderate improvement in discrimination. The five-metabolite model had a C-index of 0.708, while the two-metabolite sensitivity model yielded a C-index of 0.699 (95% CI, 0.683-0.716). Using the 2023 European Society of Cardiology SCORE2-Diabetes categories, the categorical NRI was 14.8% (95% CI, 10.8%-18.8%). Internal calibration and DCA suggested improved agreement and net benefit, but these results require external validation. Conclusions In participants with T2D and LFI, adding a metabolomic signature to SCORE2-Diabetes produced a statistically significant but moderate improvement in risk discrimination. These internally validated findings support further evaluation in independent cohorts and prospective impact studies; they do not establish immediate clinical usefulness.
BACKGROUND
Chronic kidney disease (CKD) is a major global health concern associated with increased mortality and cardiovascular events. Traditional clinical models fail to capture the complex metabolic disturbances in CKD progression. This study aimed to assess the value of circulating Nuclear Magnetic Resonance (NMR) metabolic biomarkers in predicting all-cause and cardiovascular mortality in CKD patients.
METHODS
Using NMR-based metabolomic data from the UK Biobank, we analyzed baseline plasma samples from 16,306 participants. The multivariable-adjusted Cox proportional hazards models were applied to evaluate associations between NMR metabolic biomarkers and mortality in CKD. The full cohort was randomly assigned to a training set and test set to develop and validate the CKD prognostic risk prediction model using the least absolute shrinkage and selection operator (LASSO) regression and Cox proportional hazards regression analyses. Predictive performance was assessed using Harrell's C-index, with incremental value measured by continuous net reclassification improvement (NRI) and integrated discrimination improvement (IDI).
RESULTS
Of the 143 circulating metabolic biomarkers analyzed, 122 and 90 were significantly associated with all-cause and cardiovascular mortality, respectively, in patients with CKD (all FDR adjusted P value < 0.05). Very-low-density lipoprotein (VLDL) and low-density lipoprotein (LDL) particles, small high-density lipoprotein (HDL) sub-classes, polyunsaturated fatty acids, and branched-chain amino acids were negatively associated with mortality risk. Conversely, larger HDL particles, specific triglycerides, and the inflammatory marker glycoprotein acetylation (GlycA) were positively associated with mortality risk. Harrell's C-index of the conventional prediction model was 0.729 (95% CI: 0.707, 0.750) for all-cause mortality; after adding the circulating NMR metabolic biomarkers, the C-index increased to 0.750 (95% CI: 0.728, 0.771). For cardiovascular mortality, the C-index improved from 0.761 (95% CI: 0.724, 0.798) to 0.770 (95% CI: 0.733, 0.807). The continuous NRI and IDI were 0.097 (95% CI: 0.032, 0.158) and 0.043 (95% CI: 0.034, 0.051) for all-cause mortality, and 0.125 (95% CI: 0.007, 0.234) and 0.011 (95% CI: 0.005, 0.018) for cardiovascular mortality, respectively.
CONCLUSION
This study identified multiple plasma metabolic biomarkers associated with all-cause and cardiovascular mortality in patients with CKD. Incorporating these biomarkers into conventional risk models significantly enhanced prediction accuracy.
Chen Jiang, Chengmiao Qiu, Dengren Li et al.· Journal of renal nutrition· 0 citations
Background It remains unclear whether cardiovascular-kidney-metabolic (CKM) syndrome and genetic susceptibility are associated with cancer incidence. This study aims to evaluate the associations between CKM health status, genetic susceptibility, and cancer incidence, and to develop a machine learning-based prediction model for cancer risk in patients with advanced CKM syndrome. Methods This study included 399,034 UK Biobank participants (389,289 with genetic data). CKM health was categorized into stages 0–4. Genetic susceptibility was assessed via a polygenic risk score (PRS). Associations were evaluated using Cox proportional hazard models. For participants with advanced CKM, candidate predictors were screened via the Boruta algorithm, LASSO regression, and multivariable logistic regression. Eight machine learning models were constructed and validated, and their predictive discrimination, calibration performance, and clinical practical value were further assessed. Additionally, Shapley Additive Explanations (SHAP) were adopted to interpret the internal mechanism of the optimal model. The dose–response relationship was assessed using restricted cubic splines (RCS), and mediation analysis was further performed to explore the underlying mechanism of the observed associations. Results During a median follow-up of 13.7 years, 48,247 incident cancer cases were identified. Compared to CKM stage 0, multivariable-adjusted hazard ratios (HRs) were 1.01 (95% CI: 0.91-1.12) for stage 1, 1.21 (1.10–1.33) for stage 2, and 2.17 (1.95–2.42) for advanced CKM (stage 3–4). Participants with high PRS and advanced CKM faced the highest cancer risk (HR 3.24, 95% CI 2.68–3.93). A total of 14 predictors were retained after screening. The GBM model achieved the best predictive performance (test AUC = 0.801). A web-based calculator was developed to realize individualized cancer risk prediction. RCS analyses indicated that diastolic blood pressure (Pnonlinearity<0.05), neutrophil count(Pnonlinearity<0.05), alkaline phosphatase (Pnonlinearity=0.001), and TyG-BMI (Pnonlinearity<0.05) were significantly and nonlinearly associated with cancer risk. Total bilirubin the strongest positive mediating effect, explaining 44.45% of the total association (P < 0.001). Conclusions Advanced CKM syndrome independently increases incident cancer risk, and this elevated risk is significantly amplified by high genetic susceptibility. The validated GBM model provides an interpretable tool for individualized risk estimation, which serves as a valuable complementary tool for clinical risk stratification and cancer patients with, cancer.
Xue He, J. An, Chun-Jia Yan et al.· Frontiers in Endocrinology· 0 citations
OBJECTIVE
To develop a protein risk score (ProRS) for predicting liver-related events (LREs) in patients with diabetes and compare its predictive performance with the Fibrosis-4 Index (FIB-4) and an established polygenic risk score.
RESEARCH DESIGN AND METHODS
This prospective cohort study included 13 516 individuals with prediabetes and type 2 diabetes (T2D) from the UK Biobank. Cox proportional hazards models and LASSO regression were applied to identify proteins associated with incident LREs and construct the ProRS. Predictive performance was assessed using Harrell's C-index, time-dependent area under the receiver operating characteristic curve, net reclassification improvement and integrated discrimination improvement.
RESULTS
Over a median follow-up of 13.5 years, 171 (1.3%) incident LREs occurred. We identified 877 proteins associated with LRE risk, primarily enriched in inflammatory signalling, extracellular matrix remodelling and complement/coagulation cascades. In the training set, we developed a 24-protein ProRS (C-index, 0.842; 95% CI 0.797-0.884) that stratified individuals into low-, medium- and high-risk groups, with 10-year cumulative incidences of LREs of 0.2%, 1.2% and 14.2%, respectively. Compared with the low-risk group, the hazard ratio for LREs was 57.1 (95% CI 31.9-102) in the high-risk group. In the internal validation set, the ProRS model (C-index, 0.876; 95% CI 0.827-0.920) accurately predicted both short- and long-term LREs and outperformed FIB-4 index (C-index, 0.733; 95% CI 0.657-0.807) and polygenic risk score (C-index, 0.636; 95% CI 0.564-0.706).
CONCLUSIONS
The protein risk score demonstrated superior performance compared with the FIB-4 index and the polygenic risk score in predicting incident LREs among individuals with prediabetes and T2D. The score allows stratification of individuals according to liver-related risk, though external validation in multi-ethnic cohorts is warranted.
Chenhao Ye, Zi-Qi Zhang, Yi Ding et al.· Diabetes, obesity and metabo...· 0 citations
BACKGROUND
Individuals with type 2 diabetes (T2D) have increased risk of acute myocardial infarction (AMI). Cardiovascular risk within T2D is heterogeneous and not fully explained by conventional risk factors, highlighting the need for improved biological risk stratification.
OBJECTIVES
The authors aimed to identify circulating proteins associated with incident AMI in T2D, evaluate their incremental predictive value, and explore potential causal relationships.
METHODS
In 1,830 adults with T2D from the prospective SMART2D cohort, 1,457 plasma proteins were quantified at baseline using Olink proximity extension assays. Participants were followed for incident AMI through linkage with the Singapore Myocardial Infarction Registry over a median of 10.2 years (IQR: 9.1-10.7). Multivariable Cox regression assessed protein-AMI association. Incremental predictive performance was evaluated using Harrell's C-statistic, likelihood ratio tests (LRTs), net reclassification index (NRI), and integrated discrimination index (IDI). Two-sample Mendelian randomization, using cis-protein qualitative-trait loci from SMART2D and outcome data from Biobank Japan (14,992 cases; 146,214 controls), examined causal associations.
RESULTS
During follow-up, 223 participants (12.2%) developed AMI. Seventeen proteins were independently associated with AMI after Bonferroni correction (P < 3.43 × 10-5), with the strongest associations observed for NEFL (HR: 1.70, 95% CI: 1.43-2.01), EDA2R (HR: 2.25, 95% CI: 1.69-2.97), and CHRDL1 (HR: 2.69, 95% CI: 1.83-3.94). An 8-protein panel improved risk classification beyond clinical variables (LRT P = 0.002; IDI: 0.049, 95% CI: 0.022-0.128, P = 0.002; and NRI: 0.318, 95% CI: 0.112-0.411, P = 0.008). Genetically predicted plasma FGF5 levels were associated with myocardial infarction risk.
CONCLUSIONS
Plasma proteomics improves AMI risk stratification in T2D, and FGF5 may play a role in cardiovascular risk in Asian populations.
R. L. Gurung, Huili Zheng, J. Tan et al.· JACC: Asia· 0 citations
Traditional risk factors do not fully account for the residual cardiometabolic risk of major adverse cardiovascular events (MACE) in coronary artery disease (CAD). We aimed to identify circulating metabolic signatures associated with MACE susceptibility and uncover potential pathobiological mechanisms underlying the gut-liver-heart axis. In this retrospective case-control study, untargeted high-performance liquid chromatography-mass spectrometry (HPLC-MS) was performed on fasting serum from 200 patients with CAD and MACE, 200 with CAD without MACE, and 400 matched non-CAD controls. Metabolomics data were processed using univariate analysis, multivariate analysis, and eXtreme Gradient Boosting (XGBoost) machine learning. Pathway enrichment was conducted using metabolite set enrichment analysis. Circulating fibroblast growth factor 19 (FGF19) was quantified via enzyme-linked immunosorbent assay to validate enterohepatic endocrine disruption. The MACE cohort exhibited a pronounced cardiometabolic phenotype, characterized by significantly highest rates of diabetes, hypertension, and dyslipidemia (p < 0.01). The XGBoost model robustly discriminated patients with CAD and MACE from non-CAD controls (area under the curve [AUC] = 0.984) and from patients with CAD without MACE (AUC = 0.932). Pathway analysis revealed marked dysregulation of linoleic acid metabolism and peroxisome proliferator-activated receptor (PPAR) signaling (p < 0.05). Specifically, pro-inflammatory oxidized linoleic acid metabolites, including 9- and 13-hydroxyoctadecadienoic acid (HODE)—which drive plaque instability—were significantly elevated in the MACE cohort. Furthermore, atheroprotective primary bile acids were significantly depleted in patients with CAD (p < 0.001). This depletion was accompanied by an elevated serum FGF19 level (p = 0.003), reflecting a potential disruption of the gut-liver-heart endocrine axis. In conclusion, dysregulated linoleic acid oxidation, altered PPAR signaling, and disturbed primary bile acid-FGF19 metabolism may represent key metabolic pathways associated with MACE susceptibility. Integrating these gut-liver-heart axis signatures into machine learning models holds significant promise for refining cardiovascular risk stratification and guiding targeted preventive interventions.
Min-Qing Lin, Chun-Ka Wong, K. Au et al.· Cardiovascular Diabetology· 0 citations
BACKGROUND AND AIMS
Atherothrombotic cardiovascular disease (CVD) risk prediction in older adults remains suboptimal. The relative contributions of circulating protein biomarkers and polygenic scores (PGS) are uncertain.
METHODS
In 10,433 older individuals aged ≥70 years without prior CVD events, we evaluated traditional risk factors alongside three circulating biomarkers (high-sensitivity C-reactive protein, hsCRP; N-terminal pro-b-type natriuretic peptide, NT proBNP; and high-sensitivity troponin I, hsTnI), and two PGSs (coronary artery disease, ischemic stroke) for prediction of major adverse cardiovascular events (MACE). Associations were assessed using Cox proportional hazards models. Model performance was evaluated using the C-index, calibration, and continuous net reclassification improvement (NRI).
RESULTS
Over a median follow-up of 4.5 years (interquartile range 3.4-5.5), 359 MACE occurred. Each biomarker and both PGSs were independently associated with MACE, with NT-proBNP showing the strongest association (adjusted HR per SD 1.50, 95% CI 1.36-1.65). Compared with the base clinical model, the addition of the three circulating biomarkers (hsCRP, NT-proBNP, and hsTnI) resulted in a greater improvement in discrimination than the addition of the two PGSs (ΔC-index +0.041 vs. +0.016). The fully combined model achieved the highest discrimination (C-index 0.734) with good calibration. Circulating biomarkers improved reclassification primarily through correct identification of non-cases (NRI 0.34), whereas PGSs contributed relatively more to identification of cases (NRI 0.31).
CONCLUSIONS
In older adults, circulating biomarkers and PGSs provide cumulative information for CVD risk prediction, contributing differently to discrimination and risk reclassification. Integrating biomarkers and genetic risk may improve CVD risk prediction in older people beyond traditional risk factors.
Cheng-Long Yu, C. Tran, J. Neumann et al.· Atherosclerosis· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
