Which LDL Formula Is Most Accurate When Triglycerides Are High?

Most lipid panels do not measure LDL directly. They calculate it from total cholesterol, HDL and triglycerides, and the most common calculation, Friedewald, understates LDL when triglycerides are high. If you have high triglycerides, the most accurate formula is Sampson, with Martin-Hopkins a close second. Friedewald is the one to distrust once triglycerides pass roughly 200 mg/dL and should be retired above 400.

Which LDL formula is most accurate when triglycerides are high? For triglyceride values below about 150 mg/dL, all three formulas produce nearly the same LDL. The problem begins with high triglycerides. Above 200 mg/dL, Friedewald starts to drift. Above 400 mg/dL, many labs will not report a Friedewald LDL at all because the formula is known to fail. Sampson is the most accurate published formula for high triglycerides, validated up to 800 mg/dL. Martin-Hopkins is much better than Friedewald and stays close in most high-TG results, but Sampson was built specifically to handle the high-TG pattern.

When triglycerides are high, Friedewald can make your LDL look lower than it really is. Sampson and Martin-Hopkins are the formulas that stay close to the measured truth.

Why your LDL is usually calculated, not measured Most routine lipid panels actually measure total cholesterol, HDL cholesterol and triglycerides directly. LDL is then estimated. The reference method for LDL, ultracentrifugation, is too slow and expensive for routine labs. The estimate hinges on VLDL cholesterol, the triglyceride-rich particle that carries triglycerides in blood. Since VLDL-C is rarely measured directly, formulas use triglycerides as a stand-in.

Friedewald, published in 1972, assumes that for every 5 mg/dL of triglycerides there is 1 mg/dL of VLDL cholesterol. That 5:1 ratio is roughly true in normal fasting blood. It becomes false when triglycerides are high because VLDL particles then carry more triglyceride per particle. When the formula divides by 5, it overestimates VLDL-C and subtracts too much from LDL. The LDL number drops, not because your LDL dropped, but because the VLDL estimate was wrong.

What Friedewald gets wrong when triglycerides rise When triglycerides are high, VLDL particles are larger and carry more triglyceride per cholesterol molecule. The true triglyceride-to-VLDL-C ratio can shift from 5:1 to 8:1 or higher. Friedewald still divides by 5, overestimating VLDL-C and understating LDL. The error is not trivial. In people with triglycerides between 200 and 400 mg/dL, Friedewald can produce an LDL 10 to 20 mg/dL lower than the more accurate equations. In the 400 to 800 mg/dL range, the gap can be much larger. That matters if your LDL is near a threshold like 70 or 100 mg/dL. Above 800 mg/dL, the evidence thins for every calculated formula.

Martin-Hopkins, from 2013, replaced the fixed 5:1 ratio with a lookup table based on triglycerides and non-HDL cholesterol. That lets the formula adjust as VLDL composition changes.

Sampson, from 2020, uses a nonlinear equation with both triglycerides and non-HDL cholesterol. It was built from a large set of directly measured LDL values and validated in high-triglyceride samples up to 800 mg/dL. When triglycerides are high, Sampson is the one to trust.

What a lab report looks like when this matters You may see 'LDL cholesterol (calculated)' or 'LDL-C (calc)' on a report. Labs rarely print which formula they use. If triglycerides are above 400 mg/dL, some reports say 'unable to calculate' or leave LDL blank. If triglycerides are 200 to 400 mg/dL, the lab may still report a Friedewald LDL with no warning. That is the range where the number can look better than it is.

A practical check: run the same numbers through all three formulas. If Friedewald is much lower than Sampson and Martin-Hopkins, the low LDL was partly a mathematical artifact. Do not anchor on it.

What about direct LDL measurement? People sometimes assume a directly measured LDL would settle it. It can help, but it is not automatically better. Direct LDL assays also disagree with the reference method in high-triglyceride samples, and they have their own between-laboratory variability. The Sampson equation was tested against direct measurements and performed well. If your triglycerides are high and a direct LDL is ordered, that is reasonable. But the calculated Sampson number is a valid comparison, not a guess.

Does the formula matter when triglycerides are normal? No, not in practice. Below 150 mg/dL, Friedewald, Martin-Hopkins and Sampson usually agree within a few mg/dL. If your triglycerides are normal and you are not near a treatment threshold, switching formulas will not change much. If your LDL is right at a threshold, running all three is still worth a check because even a small difference can change a label.

Check your own LDL in all three formulas You need three numbers from your lipid panel: total cholesterol, HDL cholesterol and triglycerides. Enter them in the free LDL calculator and it runs Friedewald, Martin-Hopkins and Sampson side by side, along with non-HDL and remnant cholesterol. If your triglycerides are high, compare Friedewald against Sampson. That gap is the size of the understatement.

Open the LDL calculator at https://healthosx.com/blood/calculators/ldl-cholesterol and run all three formulas on your own result.