LDL cholesterol calculator: Friedewald, Martin/Hopkins, and Sampson

Most labs don’t measure LDL cholesterol. They calculate it from the other three numbers on a lipid panel: total cholesterol, HDL, and triglycerides. Three equations are in common use, and for most people they land within a few points of each other. When triglycerides are high or LDL is low, they can differ by 20 mg/dL or more. Enter the numbers from your report to see all three side by side.

The three equations

All three start from the same idea. Total cholesterol is the cholesterol in HDL, LDL, and VLDL particles added together. HDL is measured. VLDL is not, so each equation estimates it from triglycerides, and LDL is whatever is left. They differ only in how they estimate VLDL.

Friedewald (1972)

LDL = total cholesterol − HDL − (triglycerides ÷ 5)

The Friedewald equation assumes VLDL always carries one fifth as much cholesterol as triglyceride. That holds reasonably well when triglycerides are normal and LDL is moderate. It isn’t used when triglycerides are 400 mg/dL (about 4.5 mmol/L) or more, and the calculator leaves it blank there.

Martin/Hopkins (2013)

LDL = total cholesterol − HDL − (triglycerides ÷ a factor from a table)

Martin and colleagues measured how the ratio of triglycerides to VLDL cholesterol actually varies, using lipid profiles from about 1.3 million people. Instead of a fixed 5, they built a table of 180 factors, from 3.1 to 11.9, picked by your triglyceride and non-HDL cholesterol levels. The calculator uses that table, checked cell by cell against the published figure and a second published copy.

The 2013 table is meant for triglycerides below 400 mg/dL. An extended version for 400 to 799 mg/dL uses a different table, described by Sajja and colleagues, and is not in this calculator.

Sampson, or the NIH equation (2020)

LDL = TC ÷ 0.948 − HDL ÷ 0.971 − (TG ÷ 8.56 + TG × non-HDL ÷ 2140 − TG² ÷ 16100) − 9.44

Sampson and colleagues at the NIH fitted a single equation instead of a table, with a squared triglyceride term that bends the estimate as triglycerides rise. All values are in mg/dL. The authors validated it for triglycerides up to 800 mg/dL (9 mmol/L).

When they disagree

For a typical panel the three equations agree closely. They separate when triglycerides are high, and the gap matters most when LDL is low, because that is where treatment goals sit.

All values in mg/dL.

In the last row, the same blood could be reported as an LDL of 35, 49, or 63 depending on which equation the lab uses. Friedewald divides 300 by 5 and subtracts 60 for VLDL. Martin/Hopkins divides by 9.5 and subtracts about 32. In studies that compared each equation against LDL measured by ultracentrifugation, Friedewald tends to underestimate LDL in exactly this situation, and the newer equations come closer.

Calculated LDL and direct LDL

A lab can also measure LDL directly, and the report usually says so: “LDL Direct” or “Direct LDL”. Labs often switch to a direct measurement when triglycerides are too high for any equation. Direct and calculated results are close in the middle of the range but are not interchangeable at the edges, so a change between them can move your LDL without any change in your blood.

The method is usually named in the test name or a footnote, for example “LDL Calc” or “LDL (NIH Equation)”. If a later report names a different one, any jump on that date may be the method, not you.

Limits of any calculated LDL

• Triglycerides drive the difference. Every equation estimates VLDL from triglycerides, which rise after a meal far more than total cholesterol or HDL do. A nonfasting sample moves calculated LDL more than it moves the measured numbers.
• Very high triglycerides. Above 800 mg/dL no equation here applies, and labs measure LDL directly or don’t report it.
• The inputs have their own variation. Total cholesterol, HDL, and triglycerides each vary from draw to draw. A calculated LDL carries all three variations at once, so a change of a few mg/dL between panels is usually noise.

Which number to track

Track the LDL your lab reported, and note which method it used. If your history spans several labs or methods, you can recalculate older panels with one equation so the whole series is consistent, but keep the lab’s own number too. That’s the one your clinician sees. How to track your cholesterol over time covers the rest: fasting, non-HDL, and charting results alongside medications.

If some of your results are in mmol/L, the lab unit converter has the cholesterol and triglyceride factors, and this calculator accepts either unit.

Calculating LDL in HealthViewer

HealthViewer is a lab results tracker that runs on your own computer. It imports lipid panels from lab report PDFs, MyChart downloads, and direct connections to your health system, and charts each test over time in Trends.

To recalculate LDL across your whole history with one equation, add a derived test in Settings. A derived test is a formula over your existing results, with a letter for each test. With A as total cholesterol, B as HDL, and C as triglycerides, all in mg/dL, and a date window of 0 days so only results from the same draw are combined:

• Friedewald: A - B - C / 5
• Sampson: A / 0.948 - B / 0.971 - (C / 8.56 + C * (A - B) / 2140 - C * C / 16100) - 9.44

Each letter’s results are converted to the unit most of them were in when you created the formula, so set it up once most of your lipid results are in mg/dL. Martin/Hopkins needs its lookup table, so it can’t be written as a derived test.

Everything stays in an encrypted file on your computer, with no account and no cloud service. Getting Started walks through the first import.

This calculator is for understanding your lab report, not for making treatment decisions. Your clinician will use the LDL your lab reported. Ask before acting on a recalculated value.