Friedewald LDL Calculator

A calculated LDL is total cholesterol, minus HDL, minus an estimate of the cholesterol carried in VLDL particles. Friedewald and Martin-Hopkins differ only in that last estimate; Sampson's equation also re-weights total and HDL cholesterol and adds a constant. Friedewald, Levy and Fredrickson published the first in Clinical Chemistry in 1972; their abstract describes measuring fasting total cholesterol, triglyceride and HDL cholesterol, and reports correlation coefficients of .94 to .99 against the ultracentrifuge.

In mg/dL the VLDL estimate is triglycerides ÷ 5. In mmol/L, Singapore General Hospital's laboratory page and Malaysia's 2017 dyslipidaemia guideline both print it as triglycerides ÷ 2.2, and that is what this calculator uses in mmol/L. Converting the mg/dL formula exactly gives ÷ 2.183 instead; at a triglyceride level of 4.5 mmol/L the two differ by 0.016 mmol/L.

Where dividing by five drifts

Martin and colleagues (JAMA 2013) checked the shortcut against 1,350,908 US lipid profiles measured by ultracentrifugation. The median ratio of triglycerides to VLDL cholesterol in that data was 5.2, but the full range ran from 0.4 to 145. Their method divides triglycerides by a factor read from a 180-cell table indexed by triglycerides and non-HDL cholesterol; the factors run from 3.1 to 11.9. From triglycerides of 174 mg/dL (1.97 mmol/L) upward every factor in the table is 5.1 or more, so there Martin-Hopkins cannot read below Friedewald in either unit; at lower triglycerides it can. For triglycerides under 400 mg/dL, it placed 91.7% of results in the same guideline category as the measured LDL, against 85.4% for Friedewald.

“Friedewald and colleagues noted that simply dividing triglyceride values by 5 does not give an accurate estimate of VLDL-C.” — Martin SS et al., JAMA 2013 (full text on PubMed Central)

The gap matters most at low LDL with raised triglycerides. In Martin's data, among results that Friedewald put under 70 mg/dL with triglycerides of 200 to 399 mg/dL, the measured LDL was also under 70 in 40.3% of cases, against 84.0% for Martin-Hopkins. That is why the calculator flags a Friedewald result under 70 mg/dL (1.8 mmol/L): it is the line below which the 2018 AHA/ACC guideline said a direct measurement or a modified estimate was reasonable.

Sampson's equation, and its two sets of constants

Sampson and colleagues (JAMA Cardiology 2020) derived a third estimate from 8,656 NIH Clinical Center patients measured by β-quantification, and describe it as usable for triglycerides up to 800 mg/dL. It is published as two equations — one for VLDL, one for LDL — and their constants are not the same: the VLDL equation divides triglycerides by 8.59, the LDL equation by 8.56, and the other two terms differ as well. This calculator computes LDL with the LDL equation as printed, rather than subtracting the VLDL one. In the paper's comparison, the root-mean-square error against β-quantification was 15.2 mg/dL for Sampson, 25.7 for Martin-Hopkins and 32 for Friedewald.

None of the three is a measurement. A laboratory measures LDL with a direct assay or by β-quantification; these are estimates from the same three numbers, and they disagree with each other as well as with Friedewald — the table under the result shows by how much for your values.

What SGH's lab page and Malaysia's guideline print

SGH's laboratory page describes its LDL as a calculated parameter using the Friedewald equation in mmol/L, with triglycerides ÷ 2.2, and calls it invalid when triglycerides are above 4.5 mmol/L. Malaysia's 2017 Clinical Practice Guideline on dyslipidaemia (5th edition) prints the same formula and limit, and says that above it LDL-C will have to be measured directly.

In Singapore, a study from the Health Sciences Authority's external quality assessment programme (Tan et al., Clin Chem Lab Med 2023) compared the three equations on 345 external quality assessment datasets collected over five years — laboratory control samples, not patients — against reference-method values. Its abstract reports the lowest median discordance for Martin-Hopkins (−0.725%), then Sampson (−1.389%), then Friedewald (−4.094%), and says that samples with high triglycerides, low HDL and high LDL had no misclassification by Martin-Hopkins but about 50% by Friedewald.

Which guideline is current, and the lines it draws

The 2018 AHA/ACC cholesterol guideline was retired and replaced by the 2026 ACC/AHA dyslipidemia guideline, published on 13 March 2026. The 2018 text used LDL thresholds such as 70 mg/dL rather than goals. The American Heart Association's announcement of the 2026 guideline says LDL goals are back: less than 100 mg/dL at borderline or intermediate risk, less than 70 at high risk, and less than 55 for secondary prevention at very high risk. The announcement does not say which LDL formula to use, and this page does not attribute a preference to either guideline.

Singapore's ACE clinical guideline on lipid management (December 2023, amended June 2025) works in mmol/L. It aims for LDL-C below 1.4 mmol/L in patients with a history of acute coronary syndrome and below 1.8 mmol/L in other patients with atherosclerotic cardiovascular disease, and says LDL-C above 4.9 mmol/L suggests a possible diagnosis of familial hypercholesterolaemia. It does not name a calculation method. Malaysia's 2017 guideline sets LDL-C targets by cardiovascular risk group in its Table 5, from below 3.0 mmol/L at low risk to below 1.8 mmol/L at very high risk. These are lines a clinician applies to a whole risk assessment; the calculator does not place a result against any of them.