Fed versus Fasting

When LDL Cholesterol values are estimated by the Friedewald formula, they are subject to the cumulative error of 3 independent measurements: Cholesterol (Total error 9%); Triglycerides (Total error 15%); and HDL (Total error 12%). The NCEP Lipoprotein Measurement working group recommends that the Friedewald formula only be applied when direct total cholesterol, HDL, and triglyceride measurements can be made on 12-hour fasting samples.

In a study of lipoprotein concentrations in fed and fasted human volunteers, it observed that the Friedewald formula erroneously registers a decrease in LDL Cholesterol when compared with the reference method. At 3 hours after a meal, the LDL cholesterol as estimated by the Friedewald formula was decreased by 22% from the more accurate measured LDL Cholesterol reference method.

Researchers state that CHD risk will be underestimated when based on the Friedewald formula LDL cholesterol values from non fasting samples.

In an assessment of the accuracy of LDL cholesterol values for patients with controlled type 1 and type 2 diabetes, researchers found that the Friedewald calculated LDL cholesterol values came within 10% of the measured LDL cholesterol values only half the time.