Friedewald equation underestimates low-density lipoprotein cholesterol at low concentrations in young people with and without Type 1 diabetes

Aims Although the limitations of the Friedewald-calculated serum low-density lipoprotein cholesterol (LDL-C) are well recognized, many diabetes and lipid guidelines propose LDL-C as a therapeutic target. The validity of calculated LDL-C in people with Type 1 diabetes (T1DM) is uncertain and the use of alternatives such as non-high-density lipoprotein cholesterol (non-HDL-C) or apolipoprotein measurement unexplored. We have therefore measured LDL-C with the designated reference method and examined some of the potential sources of such bias, including plasma concentrations of other lipids and apolipoproteins. Methods Seventy-four people with T1DM and 80 healthy control subjects were recruited. Fasting samples were collected for analysis of lipid profiles by a beta-quantification (BQ) reference method and by routine laboratory methods including direct HDL-C and calculation of LDL-C using the Friedewald formula. Results Overall, Friedewald LDL-C was 0.29 +/- 0.02 (mean +/- SE) mmol/l (P < 0.001) lower in the two groups than by the BQ method. This resulted in misclassification of approximately 50% of people with a calculated LDL-C < 2.0 mmol/l. Overestimation of HDL-C by the routine assay [0.08 +/- 0.01 mmol/l (P < 0.001)] accounted for similar to 28% of the error in calculation of LDL-C and the remainder appeared to be as a result of triglyceride in lipoprotein particles other than very-low-density lipoprotein (VLDL). Correlation of non-HDL-C with apolipoprotein B was better than LDL-C with apolipoprotein B for both assays in both diabetic and non-diabetic populations. Conclusions Calculated LDL-C is unsuitable as a therapeutic target in T1DM. Consideration should be give to the greater use of apolipoprotein B or non-HDL-C in clinical practice.