The accuracy in using modified Friedewald equation to calculate LDL from non-fast triglyceride: a pilot study

J Med Assoc Thai. 2009 Feb;92(2):182-7.


Background: Total cholesterol, HDL (high-density lipoprotein) and LDL (low-density lipoprotein) are important risk factors of coronary heart disease. It is costly to perform the LDL test for follow-up cardiovascular diseases (CVD) especially for Gold Card Holders (Thirty Bahts Universal Coverage). Hypertriglyceridemia is also important as it is associated with uncontrolled type 2 Diabetes mellitus, low HDL, and metabolic syndrome. Because the serum triglyceride level changes with time after meal consumption, blood test for triglyceride level should be taken after fasting 12 hours. However this causes hunger and inconvenience in many patients.

Objective: To find out the optimal time to take blood for triglyceride measurement and using it for calculation of LDL with the original Friedewald Formula and the new Modified Friedewald Formula.

Material and method: Patients were asked the approximate time of last meal/eating, drinking soft drink, milk. Additionally, the time of blood drawn from the patients was recorded The blood samples were drawn as usual amounts and the tests were done as the physicians ordered If enough sera were left, it would be analyzed for lipid profiles. LDL was also calculated by using standard Friedewald equation (sfLDL) and Modified Friedewald equation (mfLDL = total cholesterol - HDL - 1/6 triglyceride). Comparison between direct measured LDL (dmLDL), sfLDL, and mfLDL with time interval of last food, drink intake was done.

Results: There were 999 serum tubes left to be analyzed for lipid profiles and 919 sera (92.0%) left having triglyceride less than 300 mg/dl. Of those, 381, 84, and 454 samples came after fasting (nothing per oral = NPO) approximately less than 8 hours (h), 8-11.9 h, and 12 h or more respectively with sfLDL to dmLDL +/- 10 mg, comparison of 64.0%, 65.5% and 68.3% respectively. In contrast, comparing mfLDL to dmLDL +/- 10 mg being of 82.7%, 83.3% and 84.8% from the same samples and time intervals respectively thus, statistical significant (p-value < 0.001, odd ratios (OR) 2.59- 2.68). If blood drawn regardless of time from last food intake with triglyceride less than 300 mg/dl and with the above condition mfLDL, it gave 83.8% related to dmLDL while sfLDL gave only 66.3% p < 0.0001 and OR = 2.63.

Conclusion: The present pilot study showed 919 of 999 sera (92.0%) with serum triglyceride less than 300 mg/ dl, regardless of the time of the last food intake. The authors used the new Modified Friedewald equation to calculate that the LDL had 83.8% accuracy when compared to direct measured LDL +/- 10 mg. This equation is more accurate than the standard (original) Friedewald equation with OR of 2.63. The authors offer that to save the cost, the new Modified Friedewald equation should be used to calculate LDL. Then, direct LDL measurement could be reserved for patients with hypertriglyceridemia, in the treatment of LDL in high-risk CVD.

MeSH terms

  • Cholesterol, LDL / blood*
  • Humans
  • Models, Theoretical
  • Triglycerides / blood*


  • Cholesterol, LDL
  • Triglycerides