What is the clinical evidence from this study?

Study design: Cohort. Population: Cardiovascular-kidney-liver-metabolic (CKLM) disease (n=49312). Intervention: Clinical biomarkers and plasma proteomics. Primary outcome: First occurrence of chronic kidney disease, cardiovascular disease, type 2 diabetes, or metabolic dysfunction-associated steatotic liver disease.

June 7, 2026Diabetes

2247-P: Uncertainty-Calibrated Prediction of Cardiovascular–Kidney–Liver–Metabolic Disease Using Clinical Biomarkers and Plasma Proteomics

Q: What does this research mean for the field?

Integrating clinical biomarkers and plasma proteomics using conformal inference provides accurate, uncertainty-calibrated risk prediction for cardiovascular-kidney-liver-metabolic (CKLM) disease. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

View Full Paper

Ask AI

Bookmark

Key result

Clinical biomarkers and plasma proteomics predict incident cardiovascular-kidney-liver-metabolic disease with 0.78 AUC.

n=49,312

Why the study?

Traditional risk models for cardiovascular-kidney-liver-metabolic disease often yield miscalibrated or unstable predictions, particularly in data-sparse regions.

Does integrating clinical biomarkers and plasma proteomics improve uncertainty-calibrated risk prediction for cardiovascular-kidney-liver-metabolic disease?

Population

49,312 UK Biobank participants with proteomic profiling

Comparison

Machine learning prediction framework integrating clinical biomarkers and plasma proteomics

Design

Cohort study

Follow-up

Median 12.3-year

Authors

MXManrong XuShanghai Sixth People's Hospital LGLuqin GanBaton Rouge Clinic WCWENTAO CAOBaton Rouge Clinic

Discussion

Member takes

Implication

May refine observational CKLM risk estimates; leaves open prospective validation before clinical use.

Key Points

The study aims to develop a risk prediction framework for cardiovascular-kidney-liver-metabolic disease using clinical biomarkers and plasma proteomics, focusing on uncertainty calibration.
Analyzed 49,312 UK Biobank participants with proteomic profiling
Used an 80/20 stratified split and cross-validated random forest to select predictors
Employed J+aB conformal inference for uncertainty-calibrated risk prediction and Cox models for hazard ratios.
9,787 participants developed CKLM over a median follow-up of 12.3 years.
Final model achieved AUC of 0.78 and average precision of 0.52, with consistent performance across demographics.
Key clinical predictors included HbA1c and cystatin C; top proteomic predictors included GDF15 and HAVCR1.

Study Design

Type

Cohort (n=49,312)

Structured PICO

Does integrating clinical biomarkers and plasma proteomics improve uncertainty-calibrated risk prediction for cardiovascular-kidney-liver-metabolic disease?

Population

49,312 UK Biobank participants with proteomic profiling followed for a median of 12.3 years to predict incident cardiovascular-kidney-liver-metabolic disease.

Exposure

Risk prediction framework integrating clinical biomarkers and large-scale plasma proteomics using machine learning with conformal inference

Outcome

First occurrence of chronic kidney disease, cardiovascular disease, type 2 diabetes, or metabolic dysfunction-associated steatotic liver diseasecomposite

Integrating clinical and proteomic data with machine learning provides accurate, uncertainty-calibrated risk estimates for cardiovascular-kidney-liver-metabolic disease.

Cite This Study

Xu et al. (2026) conducted a cohort in Cardiovascular-kidney-liver-metabolic (CKLM) disease (n=49,312). Clinical biomarkers and plasma proteomics was evaluated on First occurrence of chronic kidney disease, cardiovascular disease, type 2 diabetes, or metabolic dysfunction-associated steatotic liver disease. Integrating clinical biomarkers and plasma proteomics achieved an area under the ROC curve of 0.78 for predicting incident cardiovascular-kidney-liver-metabolic disease.

synapsesocial.com/papers/6a250bca7def13d035e1bd10 https://doi.org/10.2337/db26-2247-p

View Full Paper

Ask AI

Bookmark

June 7, 2026Diabetes

2247-P: Uncertainty-Calibrated Prediction of Cardiovascular–Kidney–Liver–Metabolic Disease Using Clinical Biomarkers and Plasma Proteomics

View Full Paper

Ask AI

Bookmark

Key result

Clinical biomarkers and plasma proteomics predict incident cardiovascular-kidney-liver-metabolic disease with 0.78 AUC.

n=49,312

Why the study?

Traditional risk models for cardiovascular-kidney-liver-metabolic disease often yield miscalibrated or unstable predictions, particularly in data-sparse regions.

Does integrating clinical biomarkers and plasma proteomics improve uncertainty-calibrated risk prediction for cardiovascular-kidney-liver-metabolic disease?

Population

49,312 UK Biobank participants with proteomic profiling

Comparison

Machine learning prediction framework integrating clinical biomarkers and plasma proteomics

Design

Cohort study

Follow-up

Median 12.3-year

Authors

MXManrong XuShanghai Sixth People's Hospital LGLuqin GanBaton Rouge Clinic WCWENTAO CAOBaton Rouge Clinic

Discussion

Member takes

Implication

May refine observational CKLM risk estimates; leaves open prospective validation before clinical use.

Key Points

The study aims to develop a risk prediction framework for cardiovascular-kidney-liver-metabolic disease using clinical biomarkers and plasma proteomics, focusing on uncertainty calibration.
Analyzed 49,312 UK Biobank participants with proteomic profiling
Used an 80/20 stratified split and cross-validated random forest to select predictors
Employed J+aB conformal inference for uncertainty-calibrated risk prediction and Cox models for hazard ratios.
9,787 participants developed CKLM over a median follow-up of 12.3 years.
Final model achieved AUC of 0.78 and average precision of 0.52, with consistent performance across demographics.
Key clinical predictors included HbA1c and cystatin C; top proteomic predictors included GDF15 and HAVCR1.

Study Design

Type

Cohort (n=49,312)

Structured PICO

Does integrating clinical biomarkers and plasma proteomics improve uncertainty-calibrated risk prediction for cardiovascular-kidney-liver-metabolic disease?

Population

49,312 UK Biobank participants with proteomic profiling followed for a median of 12.3 years to predict incident cardiovascular-kidney-liver-metabolic disease.

Exposure

Risk prediction framework integrating clinical biomarkers and large-scale plasma proteomics using machine learning with conformal inference

Outcome

First occurrence of chronic kidney disease, cardiovascular disease, type 2 diabetes, or metabolic dysfunction-associated steatotic liver diseasecomposite

Integrating clinical and proteomic data with machine learning provides accurate, uncertainty-calibrated risk estimates for cardiovascular-kidney-liver-metabolic disease.

Cite This Study

synapsesocial.com/papers/6a250bca7def13d035e1bd10 https://doi.org/10.2337/db26-2247-p

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark