Measuring LDL cholesterol: for old and new calculations, is there an optimal formula?

Over the last 30 years, since release of the first National Cholesterol Education Program’s Adult Treatment Panel (NCEP-ATP) Guidelines, LDL cholesterol (LDL-C) has been the focus of, and basis for, therapeutic decisions, both for when to start drug therapy and for what goals to achieve (1 ). Although the recent American Heart Association and American College of Cardiology (AHA/ACC) guidelines (2 ) minimize LDL-C treatment goals, the majority of guidelines throughout the world continue to recommend therapeutic targets similar to those of NCEP-ATP III (3–5 ). Furthermore, given the substantial controversy and debate since the release of the AHA/ACC guidelines, it is unclear how successfully they will ultimately influence physician practice regarding the use of LDL-C treatment targets (6, 7 ). Thus, measurement of LDL-C remains a vital component of our decision making in terms of even the new “nontarget” guidelines. From the outset, the NCEP recognized the importance of accurate and reproducible assessment for LDL-C and convened an expert panel of laboratory scientists to address the standardization of the lipid measurements included in the guidelines (8, 9 ). It was recognized that there was no simple and accurate method for actual LDL-C measurement, and thus calculation of LDL-C using the Friedewald formula, which was widely accepted as an accurate and cost-effective alternative to the reference method, preparative ultracentrifugation (PUC), was recommended for routine clinical purposes in patients with triglycerides (TG) 400 mg/dL (4.52 mmol/L) (8 –10 ). The formula described and validated against PUC in 1972 calculated LDL-C using total cholesterol (TC), TG, and HDL-C (10 ). However, over the last 20 years, as clinical outcome trials provided additional data indicating that lower LDL-C is associated with further reductions in the risk of cardiovascular disease, treatment goals for LDL-C have been adjusted downward (3–5 ). In comparing the performance of the Friedewald formula, it is important to consider that measurements of TC and HDL-C are common components of, and were therefore identical to, both the Friedewald formula and PUC. Thus, any differences between them relate solely to the calculation of very-low-density lipoprotein cholesterol (VLDL-C), which the Friedewald formula assumes to be one-fifth, or 0.2, of the TG concentration if values are in milligrams per deciliter (10 ). Two large studies over the next 2 decades reexamined the relationship between LDL-C calculated by Friedewald and measured by PUC (11, 12 ). The study by De Long et al. (11 ) compared 10483 participants in the Lipid Research Clinics prevalence studies and concluded that VLDL-C was better represented by a factor of TG/6 or 0.16 TG, when reported in milligrams per deciliter (0.37 when reported in millimoles per liter). McNamara et al. (12 ) assessed 4797 samples from a wide range of patients using various formulas for calculating VLDL-C and the effect on calculated LDL-C to PUC. They found that the best representation of VLDL-C varied by TG concentration and suggested different formulas based on the TG value [TG/4 (ratios were not provided for TG in mmol/L) for concentrations 50 mg/dL (0.56 mmol/L); TG/4.5 for TG 50 – 200 mg/dL (0.56 –2.26 mmol/L); and TG/5 for TG 200 – 400 mg/dL (2.26 – 4.52 mmol/L)] (12 ). However, the differences in the actual calculated LDL-C with the various factors were small and likely clinically insignificant. Despite finding values slightly more consistent with PUC, neither study resulted in adoption of a new formula and Friedewald endured. The Friedewald formula was originally validated based on a relatively small number of samples from 3 lipid patient populations: 96 subjects considered “normal” in whom LDL-C ranged from 62 to 185 mg/dL (1.6 to 4.78 mmol/L); 204 subjects with Fredrickson type II hyperlipidemia (HLP) with LDL-C ranging from 173 to 840 mg/dL (4.47 to 21.7 mmol/L); and 111 Fredrickson type IV HLP patients with TG 400 1 Medpace Reference Laboratories and Metabolic & Atherosclerosis Research Center, Cincinnati, OH. * Address correspondence to the author at: Medpace Reference Laboratories and Metabolic & Atherosclerosis Research Center, 5355 Medpace Way, Cincinnati, OH 45227. E-mail [email protected]. Received September 28, 2014; accepted September 30, 2014. © 2014 American Association for Clinical Chemistry Previously published online at DOI: 10.1373/clinchem.2014.232793 2 Nonstandard abbreviations: NCEP-ATP, National Cholesterol Education Program’s Adult Treatment Panel; LDL-C, LDL cholesterol; AHA/ACC, American Heart Association and American College of Cardiology; PUC, preparative ultracentrifugation; TG, triglycerides; TC, total cholesterol; HLP, hyperlipidemia; VAP, vertical auto profile. Clinical Chemistry 60:12 1466–1468 (2014) Editorials