Heptanoylcarnitine Outperforms Traditional Clinical Markers for CHD Risk Stratification in Type 2 Diabetes
In a 213-participant study (81 healthy controls, 69 T2DM, 63 T2DM with CHD), an artificial neural network model using serum myristic acid, palmitic acid, and heptanoylcarnitine discriminated T2DM patients with and without coronary heart disease with accuracy exceeding 0.95 in both training and testing sets [1]. This performance was explicitly reported as superior to traditional indicators including lipid profiles, blood glucose, and blood pressure measurements [1]. Among the numerous acylcarnitines profiled, acylcarnitines were identified as the most disturbed metabolite class between T2DM patients and healthy controls, yet heptanoylcarnitine was specifically selected by machine learning feature reduction as one of only three metabolites achieving maximal discriminative power [1].
| Evidence Dimension | Diagnostic accuracy for discriminating T2DM patients with vs. without CHD |
|---|---|
| Target Compound Data | ANN model accuracy >0.95 (training and testing sets) using myristic acid + palmitic acid + heptanoylcarnitine panel |
| Comparator Or Baseline | Traditional diagnostic indicators (lipid profile, fasting blood glucose, blood pressure); accuracy not explicitly quantified but stated as inferior |
| Quantified Difference | Accuracy >0.95; explicitly superior to traditional markers (qualitative comparison reported) |
| Conditions | Serum UHPLC-MS/MS metabolomics; 213 participants; ANN model with feature selection |
Why This Matters
Procurement of heptanoylcarnitine as an authentic standard is essential for laboratories developing or validating CHD risk stratification assays in diabetic populations, where this specific C7:0 acylcarnitine—not its C6:0 or C8:0 neighbors—carries independent diagnostic weight.
- [1] Hu T, Zhang W, An Z, et al. Machine learning reveals serum myristic acid, palmitic acid and heptanoylcarnitine as biomarkers of coronary artery disease risk in patients with type 2 diabetes mellitus. Clinica Chimica Acta. 2024;556:117852. DOI: 10.1016/j.cca.2024.117852. View Source
