Comparison of the Accuracy of Calculated Low-Density Lipoprotein Cholesterol Versus Direct Measurement: A Cross-Sectional Study

Abstract

Background: Accurate assessment of lipid profiles particularly low-density lipoprotein cholesterol (LDL-C) is essential for the effective diagnosis and management of cardiovascular disease (CVD). While the Friedewald’s formula remains the most commonly used method for estimating LDL-C, it has known limitations, especially in certain clinical scenarios. This has led to increased interest in alternative calculation methods.
Aim: To evaluate the accuracy of 12 calculated low-density lipoprotein cholesterol (LDL-C) estimation formulas in comparison with directly measured LDL-C levels.
Materials and Methods: Lipid profile records from 93 patients aged 18 years and older, with triglyceride levels below 400 mg/dL, were analysed. LDL-C levels were measured using the enzymatic method (CHE/CHO/POD) and also calculated using the following formulas: Friedewald, Anandaraja, De Cordova, Vujovic, Ahmadi, Puavillai, Chen, Hattori, Martin–Hopkins, Rao, DeLong, and Teerakanchana
Results: When comparing LDL-C values obtained by direct and calculated methods using the paired t-test, most calculated methods showed statistically significant differences from the direct method. However, the Chen and Martin methods demonstrated no significant difference, indicating close agreement with the direct assay. Overall, the calculated methods exhibited strong correlations with the direct method (r = 0.9142 to 0.996), except for the Ahmedi method, which showed a weak correlation (r = 0.450).
Conclusion: In conclusion, while most calculated LDL-C methods differed significantly from the direct assay, they generally showed strong agreement. The Chen and Martin formulas demonstrated the highest accuracy, with no significant differences and strong correlations, making them the most reliable alternatives. In contrast, the Ahmedi method showed poor correlation and limited clinical utility.