It's a classic scenario that patients have come to know. The patient sees their provider, needs a lipid profile as part of their clinical evaluation, but it's the afternoon and they just had lunch. So the patient makes a return trip to the lab after an overnight fast and is greeted by a sea of patients for the morning rush of fasting bloodwork.

One reason that we have long asked patients to fast for lipid testing is the Friedewald equation. The Friedewald equation was published in 1972 as a method to calculate low-density lipoprotein cholesterol (LDL-C) in fasting patients with triglycerides <400 mg/dL via subtraction of estimated very-low-density lipoprotein cholesterol (VLDL-C) as estimated by the ratio of triglycerides divided by five.¹ The Friedewald equation was needed because direct measurement of LDL-C by the gold standard technique of analytical ultracentrifugation (beta quantification) is slow, costly and really only fit for research settings. The Friedewald equation provided a quick, inexpensive alternative that could be scaled for clinical purposes and has served as a global standard in lipid analysis over the past four decades. However, in response to recent clinical science, practices are changing with respect to recommendations for fasting before lipid assessment and the optimal way to estimate LDL-C.

Recent guidelines have shifted to recommend non-fasting lipid analysis for certain populations given its convenience, and because several studies have supported non-fasting lipids in cardiovascular risk assessment.^2-5 Clinical decisions are of course not only guided by global risk assessment, but also by LDL-C levels. Although the Friedewald LDL-C equation was originally derived in fasting patients, it has been increasingly utilized in the non-fasting setting to guide management to lower LDL-C. This is of concern given that Friedewald and colleagues stated in their original manuscript that simply dividing triglycerides by five does not give an accurate estimate of VLDL-C; they only viewed it as reasonable because the LDL-C concentrations were high, and thus VLDL-C was a relatively small component of the equation.

Despite treatment to low LDL-C in current times, some have advocated for non-fasting testing with the Friedewald equation in all patients without qualification.⁶ A key limitation of such expert opinion is that it does not account for several high-risk sub-populations in which the Friedewald equation is prone to inaccuracy.⁷ A recent analysis by Sathiyakumar et al.⁸ showed that the equation leads to sizable errors more commonly in non-fasting samples versus fasting ones.

The most sizable errors occur in the range of greatest clinical relevance; that is, at low LDL-C levels <70 mg/dL, the zone that we shoot for in the highest risk patients. The Friedewald equation underestimates true LDL-C particularly in non-fasting patients and when triglycerides are moderate to high.^9,10 In this setting, non-fasting values were only accurate 37% of the time in clinical classification and errors of 10 mg/dL were observed 81% of the time. In those with Friedewald LDL-C <70 mg/dL, one in 12 non-fasting patients had 20-29 mg/dL errors as compared with measured LDL-C and one in 28 patients had errors >30 mg/dL.

With proprotein convertase subtilisin/kexin type (PCSK9) inhibitors in clinical use already and other new drugs in the pipeline it is now possible to treat patients to lower and lower LDL levels, and it will be increasingly important to ascertain accurate levels in these populations. Given that 200 million Americans undergo lipid testing per year, taking an unqualified non-fasting approach with the Friedewald equation has the potential to misinform many clinical decisions. This could waste significant time and money by creating a need for re-testing, as well as lead to worse outcomes in patients who are not treated with proven therapies due to falsely estimated LDL-C values.

Since ultracentrifugation is primarily limited to research settings (and newer direct chemical assays for LDL-C add cost without reliable accuracy), we believe the current best solution lies in improved LDL-C estimation. A variety of groups have attempted to improve on the Friedewald equation over the years. The approaches have usually taken the form of choosing a different fixed factor than five or deriving a new regression equation. Though different, these approaches were still one-size-fits-all solutions, and have not proved to work substantially better than Friedewald estimation when applied to external populations, nor have they solved the non-fasting LDL-C assessment problem.

In contrast, the novel Martin-Hopkins LDL-C method is the first to transform the Friedewald equation from a one-size-fits-all to an individualized approach. It replaces the fixed factor of five used for the triglyceride to VLDL-C ratio with one of 180 patient-specific variables, which is calculated based on serum triglyceride and non-high-density lipoprotein cholesterol (HDL-C) concentrations.¹¹ Thus, no additional testing is required; this modernized algorithm is run using the same inputs from the standard lipid profile. Its accuracy and superiority to Friedewald estimation has been validated in the US and internationally in countries such as Brazil, Japan, Korea and Taiwan.

Cross-sectional analysis of over 1.5 million patients showed that LDL-C accuracy remains high with this new Martin-Hopkins method, regardless of fasting versus non-fasting.⁸ For example, as compared with the statistics above in those with non-fasting Friedewald LDL-C <70 mg/dL, where one in 12 and one in 28 patients had errors of 20-29 or >30 mg/dL, respectively, the comparable numbers were one in 500 and no patients when using the modernized method. In fact, >97% of patients have errors <10 mg/dL, even in the non-fasting state. The implication is that recommendations could be liberalized in most clinical circumstances to either fasting or non-fasting LDL-C assessment, which means in turn greater convenience for clinicians and their patients.

In the age of precision medicine, there is increasing focus on precision in diagnostics and tailoring management to the level of the individual. The move from the Friedewald equation to the novel Martin-Hopkins LDL-C method is an early example of this, arising from availability of big data, a fresh approach to data processing and recognition of the clinical importance of accuracy at low LDL-C. Modernizing LDL-C estimation brings the advantage of preserved LDL-C accuracy in the non-fasting state. We are encouraged to see innovative clinical laboratories already embracing this change.

References

1. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499-502.
2. Catapano AL, Graham I, De Backer G, et al. 2016 ESC/EAS guidelines for the management of dyslipidaemias. Eur Heart J 2016;37:2999-3058.
3. Mora S, Rifai N, Buring JE, Ridker PM. Fasting compared with nonfasting lipids and apolipoproteins for predicting incident cardiovascular events. Circulation 2008;118:993-1001.
4. Ridker PM, Rifai N, Cook NR, Bradwin G, Buring JE. Non-HDL cholesterol, apolipoproteins A-I and B100, standard lipid measures, lipid ratios, and CRP as risk factors for cardiovascular disease in women. JAMA 2005;294:326-33.
5. Jacobson TA, Ito MK, Maki KC, et al. National lipid association recommendations for patient-centered management of dyslipidemia: part 1--full report. J Clin Lipidol 2015;9:129-69.
6. Nordestgaard BG. A test in context: lipid profile, fasting versus nonfasting. J Am Coll Cardiol 2017;70:1637-46.
7. Driver SL, Martin SS, Gluckman TJ, Clary JM, Blumenthal RS, Stone NJ. Fasting or nonfasting lipid measurements: it depends on the question. J Am Coll Cardiol 206;67:1227-34.
8. Sathiyakumar V, Park J, Golozar A, et al. Fasting versus nonfasting and low-density lipoprotein cholesterol accuracy. Circulation 2018;137:10-9.
9. Scharnagl H, Nauck M, Wieland H, Marz W. The Friedewald formula underestimates LDL cholesterol at low concentrations. Clin Chem Lab Med 2001;39:426-31.
10. Jun KR, Park HI, Chun S, Park H, Min WK. Effects of total cholesterol and triglyceride on the percentage difference between the low-density lipoprotein cholesterol concentration measured directly and calculated using the Friedewald formula. Clin Chem Lab Med 2008;46:371-5.
11. Martin SS, Blaha MJ, Elshazly MB, et al. Comparison of a novel method vs the Friedewald equation for estimating low-density lipoprotein cholesterol levels from the standard lipid profile. JAMA 2013;310:2061-8.

Keywords: Cholesterol, VLDL, Cholesterol, LDL, Triglycerides, Subtilisin, Cross-Sectional Studies, Expert Testimony, Lipids, Lunch, Cardiovascular Diseases, Risk Factors, Lipoproteins, Proprotein Convertases, Risk Assessment, Saccharomyces cerevisiae Proteins, Ultracentrifugation, Algorithms, Dyslipidemias

< Back to Listings