PURPOSE: This study was conducted to determine if the direct low-density lipoprotein (LDL) cholesterol assay would provide a more valid measure of LDL cholesterol in diabetic patients compared with the Friedewald equation.

PATIENTS AND METHODS: Fasting plasma from 148 diabetic patients, 40 with insulin-dependent diabetes (IDDM) and 108 with non-insulin-dependent diabetes (NIDDM) with triglyceride levels <400 mg/dL, were analyzed for LDL cholesterol using the Friedewald equation, the direct LDL assay, and beta-quantification. Forty-six diabetic patients with triglyceride levels ≥400 mg/dL were also studied to determine the validity of the direct LDL cholesterol assay with hypertriglyceridemia.

RESULTS: The Friedewald equation and the direct LDL cholesterol assay correlated well with beta-quantification (r = 0.969 and r = 0.971, respectively) for LDL cholesterol determination in diabetic patients. Although the Friedewald equation in comparison with beta-quantification underestimated (8%) LDL cholesterol values in diabetic patients, the direct LDL cholesterol assay had a mean bias of <1%. Also, the underestimation by the Friedewald equation exceeded 10% for the triglyceride subgroup of 200 to 400 mg/dL. Furthermore, the accuracy of the direct LDL cholesterol assay was superior to the Friedewald equation since LDL cholesterol levels determined by the two methods coincided within ±10% of beta-quantification in 85% and 68% of diabetic patients, respectively (P = 0.0005). Similar results for both the Friedewald equation and the direct LDL cholesterol assay in comparison with beta-quantification were seen when diabetic patients were subgrouped into IDDM and NIDDM. Also, the direct LDL cholesterol assay appeared to provide a reliable estimate in patients with triglycerides ≥400 mg/dL.

CONCLUSION: The results of our studies indicate that the direct LDL cholesterol assay is a more reliable and accurate method than the Friedewald formula for LDL cholesterol determination in diabetic patients and is more rapid and cost effective than the reference method.