Evaluation of calculated low-density lipoprotein against a direct assay

To evalate compare the calculated LDL determined by the Friedewald formula when Tg < 200 mg/dL, 200-399 mg/dL, and > or =400 mg/dL against a direct method.

Samples from 202 participants (122 males, 80 females, aged 20-87 years old) were determined for cholesterol, triglyceride (Tg), high-density lipoprotein (HDL), and low-density lipoprotein (LDL) at Department of Laboratory Medicine, King Chulalongkorn Memorial Hospital (KCMH). LDL was determined by Friedewald formula and a direct method.

Intra-assay and inter-assay precisions at Tg <200 mg/dL, 200-399 mg/dL, > or =400 mg/dL of calculated LDL and direct LDL were 4.80%, 3.29%, 20.37%, 4.86, 8.42%, 8.32%, 2.11%, 1.79%, 3.99%, 2.36%, 2.41% and 6.16%, respectively. The mean absolute biases calculated for calculated LDL against direct LDL at Tg <200 mg/dL, 200-399 mg/dL, > or =400 mg/dL and for total samples were 4.70%, 11.73%, 63.65%, and 7.46%, respectively. Linear regression analysis for calculated LDL vs direct LDL for total samples and grouped as Tg <200 mg/dL, 200-399 mg/dL, and > or =400 mg/dL were 0.9190, 0.9796, 0.9440, and 0.7910, respectively. Intraclass correlation coefficient (ICC) at 95% confidence interval of calculated LDL against direct LDL at Tg <200 mg/dL, 200-399 mg/dL, > or =400 mg/dL and for total samples were 0.963, 0.930, 0.767, and 0.889, respectively.

The present data suggested that direct LDL is superior over calculated LDL in terms of precision and accuracy. The present study supported that at Tg > or =400 mg/dL calculated LDL should not be used and the traditional cutoff of Tg <400 mg/dL for using Friedewald formula should be revised. In addition, regarding patient convenience, financial reason, and precision and accuracy of analytical method, direct LDL is recommended when Tg > or =200 mg/dL.