Identification of key metabolic indicators associated with the comorbidity of ischemic stroke and diabetes mellitus using an optimal interpretable clinlabomics model - Report - MDSpire
Advertisement
Identification of key metabolic indicators associated with the comorbidity of ischemic stroke and diabetes mellitus using an optimal interpretable clinlabomics model
Clinical Report: Metabolic Markers Linked to Ischemic Stroke and Diabetes
Overview
This study identifies nine metabolic biomarkers associated with the coexistence of ischemic stroke (IS) and diabetes mellitus (DM).
Background
Ischemic stroke is a leading cause of disability and mortality globally, with diabetes mellitus being a significant risk factor that exacerbates stroke outcomes. The prevalence of IS-DM comorbidity is rising.
Data Highlights
Indicator
Odds Ratio (OR)
95% Confidence Interval (CI)
P-value
TyG index
4.76
4.01–5.64
<0.001
Atherogenic index of plasma (AIP)
3.60
2.74–4.72
<0.001
Key Findings
12 metabolic indicators were significantly associated with IS-DM comorbidity.
The TyG index and AIP were identified as major risk factors for IS-DM comorbidity.
The rpart algorithm model achieved an accuracy of 88.5% and an AUC of 0.910 in the training set.
SHAP analysis indicated that HbA1c/HDL-C was the most important feature in the model.
9 candidate metabolic biomarkers were identified, including TyG, uric acid, and total cholesterol.
Clinical Implications
The identification of specific metabolic biomarkers can facilitate early screening and risk assessment for patients with IS-DM comorbidity.
Conclusion
The study identifies key metabolic indicators for IS-DM comorbidity.
