Can You Trust Your LDL cholesterol (LDL-C) Numbers?
The most common way LDL-C is determined is to estimate its concentration using a formula derived from direct measurement of total cholesterol, HDL-C, and triglycerides. The equation was developed in 1972 by William Friedewald and colleagues[51] and continues to be routinely used in clinical assessment of cardiovascular risk and research studies, including those involving low carbohydrate diets. LDL-C is calculated as total cholesterol minus the sum of HDL-C and VLDL-C.
LDL-C = Total Cholesterol -[HDL-C + (Triglycerides/5)]
A major assumption is that the ratio of triglyceride to cholesterol is constant. VLDL is estimated as equal to triglycerides (mg/dL) divided by 5. This presumed 5: 1 ratio is not constant, and the errors from this LDL-C calculation are significant[52]. In the original 1972 paper, these researchers noted that the calculation of LDL-C was inaccurate when chylomicrons were present or triglycerides were above 400 mg/dL. These essential limitations under conditions of high plasma triglycerides are widely recognized today.
Less appreciated are the potential errors associated with low plasma triglycerides, a condition that is highly relevant when interpreting the LDL-C response to low carbohydrate diets since they often result in marked reductions in triglycerides.
For example, a published case report describes
a man with plasma triglycerides of 55 mg/dL who had an estimated LDL-C of 172 mg/ dL using the traditional Freidewald equation. But when measured by two separate direct methods, his actual LDL-C proved to be 126 mg/dL (this was also substantiated by a normal apo B level) [53].
In a formal study of 115 volunteers with plasma triglycerides less than 100 mg/dL, use of the Friedewald formula resulted in a statistically significant overestimation of LDL-C by an average of 12 mg/dL[54].
How does this play out if you are on a low carbohydrate diet? Let's assume that a low carbohydrate diet causes a reduction of triglycerides from 200 to 75 mg/dL with no change in total and HDL cholesterol. As a result, the calculated LDL-C from the Friedewald equation would necessarily increase from 100 to 125 mg/dL. How much of this 25% increase is real and how much artifact? That can only be determined by a direct assessment of LDL-C, which most physicians do not bother to do.