To evaluate the potential of proteomics for improving heart failure risk prediction in patients with dysglycaemia, emphasizing its clinical significance.
Key Findings:
Four proteins (NT-proBNP, LTBP2, REN, GDF-15) were independently associated with heart failure incidence, suggesting new avenues for risk assessment.
The PPCF model outperformed traditional clinical risk factor models, indicating a potential shift in predictive strategies.
Model performance was consistent across different glucose metabolism profiles, highlighting its broad applicability.
Interpretation:
While the PPCF model shows promise for heart failure prediction, its clinical utility is tempered by low positive predictive value and limitations in generalizability and outcome definition, necessitating further validation.
Limitations:
Cohorts are predominantly White and healthy, limiting generalizability to diverse populations.
Low positive predictive value (16.8%) may lead to high false positive rates, challenging its standalone utility.
Outcome definition does not differentiate between HFpEF and HFrEF, which may obscure distinct biological pathways.
Single baseline protein measurements may not capture long-term risk dynamics, necessitating longitudinal studies.
Conclusion:
The study highlights the potential of proteomics in heart failure risk stratification but underscores the need for further validation and exploration of clinical applicability, particularly addressing the identified limitations.
