Article
Letter to the Editor
Ann Lab Med 2024; 44(4): 363-366
Published online January 19, 2024 https://doi.org/10.3343/alm.2023.0387
Copyright © Korean Society for Laboratory Medicine.
Direct LDL Cholesterol Assay vs. Estimated Equations in Patients With Hypertriglyceridemia or Low LDL Cholesterol Levels
Jennifer Rodríguez-Domínguez, RD.J.*, Álvaro Piedra-Aguilera, PA.A.*, María Martínez-Bujidos, MB.M. , Susana Malumbres-Serrano , MS.S. , Cristian Morales-Indiano , MI.C. , and Carla Fernández-Prendes, FP.C.
Laboratory Medicine Department, Hospital Universitari Germans Trias i Pujol, Badalona, Spain
Correspondence to: Carla Fernández-Prendes, FP.C.
Laboratory Medicine Department, Hospital Universitari Germans Trias i Pujol, Carretera de Canyet S/N, 08916 Barcelona, Badalona, Spain
E-mail: cfernandezp.germanstrias@gencat.cat
* These authors contributed equally to this study as co-first authors.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Dear Editor,
Low-density lipoprotein–cholesterol (LDL-C) levels strongly correlate with cardiovascular disease risk. Guidelines recommend targeting low LDL-C levels to reduce this risk. Therefore, precise measurement is essential [1].
The LDL-C reference method requires ultracentrifugation by trained personnel and large serum samples [2, 3]. Equations, such as Friedewald, Martin–Hopkins, or Sampson, are often used instead to calculate LDL-C levels [4, 5]. Equations have limitations, particularly for the calculation LDL-C≤70 mg/dL or triglycerides levels (TG) ≥400 mg/dL. Furthermore, bias and imprecision from the different measurements used in the calculation can adversely affect LDL-C calculation accuracy [6].
LDL-C clinical decision limits are based on the Friedewald equation. Despite the additional cost, the direct LDL-C method (dLDL-C) is faster than ultracentrifugation and more accurate than equations. dLDL-C results may vary either among manufacturers or depending on the reagent-calibrator-instrument combination. Therefore, it is advisable to compare the dLDL-C with the reference method [7].
We assessed and compared the dLDL-C and equations for calculating LDL-C against the reference method to identify quick and accurate alternatives. This was a prospective study developed in Spain, from March 2022 to March 2023.
In total, 212 serum samples stored at 4ºC were analyzed and categorized into two groups (IRB; PI-21-036): group 1 (TG≥400 mg/dL, positive lipemic index, N=113) and group 2 (TG≤200 mg/dL, LDL-C≤70 mg/dL, N=99). Group 1 was subdivided into 1a (TG=400–500 mg/dL, N=62) and 1b (TG=500–900 mg/dL, N=51). Group 2 was subdivided into 2a (LDL-C=10–40 mg/dL, N=49) and 2b (LDL-C=40–70 mg/dL, N=50).
dLDL-C level was measured using the enzymatic selective protection method [8] with an AU5800 analyzer (Beckman Coulter, Brea, CA, USA). The dLDL-C calibrator value is traceable to the reference method. LDL-C was calculated using the Friedewald, Martin–Hopkins (180-cell strata), and Sampson equations. Very-low-density lipoprotein–cholesterol (VLDL-C) was isolated from 1.5 mL of serum utilizing sequential density gradient ultracentrifugation at 100,000×g for 18 hrs with an F50L fixed-angle rotor (Thermo Scientific, Basingstoke, UK) using KBr for density adjustment (1,006 g/mL). TC, HDL-C, TG, and VLDL-C contents were measured using the AU5800 analyzer, and LDL-C level was calculated as follows: LDL-C=TC−HDL-C−VLDL-C [2, 3, 4].
Statistical analysis was conducted using MedCalc v19.6 (MedCalc Software, Ostend, Belgium). Means were compared using t-test for paired samples to compare the different methods against the reference method.
Bias between methods was calculated as follows:
where Cn and Cx represent the LDL-C determined using the reference and alternative methods, respectively. Bias was compared with the reference change value (RCV), which was calculated for LDL-C considering a unilateral Z statistic with 95% confidence (Z=1.65), as follows:
where CVa and Cvi are the analytical CV and within-subject biological variation according to the European Federation of Clinical Chemistry and Laboratory Medicine Biological Variation Database, respectively [9]. The RCV was calculated to be 20.0% based on CVa=2.18 and CVi=8.3. Bland–Altman difference plots were used to compare the reference and alternative methods in groups 1 and 2.
Table 1 reveals significant differences in LDL-C between the reference and alternative methods, except for dLDL-C in group 1. However, the bias did not exceed the RCV for dLDL-C and Martin–Hopkins. In group 2, LDL-C differed significantly from those determined using the reference method for all alternative methods; however, the bias did not exceed the RCV for dLDL-C.
-
Table 1 . Comparison of methods for direct LDL-C assay and estimated equations for LDL-C with a reference method
Group Variable Ultracentrifugation* dLDL-C Friedewald Martin–Hopkins Sampson Group 1 (TG≥400 mg/dL) Median (IQR) 148 (116–177) 149 (130–170) 85 (59–112) 121 (101–142) 101 (81–122) P 0.7745 <0.0001 <0.0001 <0.0001 Bias (%) 0.5 −42.4 −18.5 −31.7 Group 1a (TG=400–500 mg/dL) Median (IQR) 151 (115–180) 149 (122–174) 93 (61–124) 126 (101–151) 106 (78–132) P 0.4070 <0.0001 <0.0001 <0.0001 Bias (%) −1.5 −38.4 −16.6 −30.1 Group 1b (TG=500–900 mg/dL) Median (IQR) 148 (124–174) 148 (132–165) 77 (55–98) 114 (102–134) 98 (82–112) P 0.8201 <0.0001 <0.0001 <0.0001 Bias (%) −0.2 −48.3 −22.7 −34.0 Group 2 (LDL-C≤70 mg/dL) Median (IQR) 58 (49–62) 61 (54–68) 41 (34–46) 45 (37–50) 43 (36–48) P <0.0001 <0.0001 <0.0001 <0.0001 Bias (%) 5.3 −29.6 −22.9 −25.4 Group 2a (LDL-C=10–40 mg/dL) Median (IQR) 49 (44–54) 55 (51–63) 34 (29–38) 36 (30–41) 36 (29–40) P <0.0001 <0.0001 <0.0001 <0.0001 Bias (%) 10.9 −31.8 −25.8 −26.7 Group 2b (LDL-C=40–70 mg/dL) Median (IQR) 62 (58–66) 65 (61–69) 46 (43–49) 49 (46–53) 48 (45–51) P <0.0001 <0.0001 <0.0001 <0.0001 Bias (%) 5.8 −25.1 −20.2 −21.9 *Reference method.
Abbreviations: IQR, interquartile range; TG, triglycerides; LDL-C, low-density-lipoprotein cholesterol; dLDL-C, direct LDL-C assay.
In groups 1a and 1b, LDL-C measured using the reference method differed significantly from the values determined using the alternative methods, except for dLDL-C. The bias never exceeded the RCV for dLDL-C in groups 1a and 1b nor for Martin–Hopkins in group 1a. In groups 2a and 2b, all methods differ significantly from the reference method; however, the bias did not exceed the RCV for dLDL-C (Table 1).
Bland–Altman plots demonstrated minimal bias of dLDL-C vs. the reference method in groups 1 and 2 (Fig. 1).
-
Figure 1. Bland–Altman plots for the alternative methods and the ultracentrifugation (reference) method. (A) dLDL-C, (B) Friedewald, (C) Martin–Hopkins, and (D) Sampson methods applied in group 1. (E) dLDL-C, (F) Friedewald, (G) Martin–Hopkins, and (H) Sampson methods applied in group 2.
Abbreviations: TG, triglycerides; LDL-C, low-density lipoprotein–cholesterol; dLDL-C, direct low-density lipoprotein–cholesterol.
Notably, for TG≥400 mg/dL, dLDL-C was the superior method. For TG=500–900 mg/dL, results from all alternative methods differed from ultracentrifugation results, except for dLDL-C. For TG=400–500 mg/dL, both dLDL-C and Martin–Hopkins methods were suitable, with dLDL-C being the most accurate. dLDL-C is applicable for LDL-C<70 mg/dL.
For LDL-C<70 mg/dL or TG≥400 mg/dL, Friedewald, Martin–Hopkins, and Sampson tend to underestimate LDL-C. Martin–Hopkins and Sampson equations reportedly have improved accuracy over Friedewald with TG≥400 mg/dL or LDL-C<70 mg/dL [5, 10]. However, these studies did not compare these methods with ultracentrifugation, but with dLDL-C.
In conclusion, the study showed that both equations and the Friedewald method continue to underestimate LDL-C. The findings support that dLDL-C is an excellent choice for TG≥400 mg/dL or LDL-C<70 mg/dL. These results are specific for the Beckman Coulter dLDL-C; differences among manufacturers have been reported [7]. Similarly, TC, HDL-C, and TG assays used to calculate LDL-C are not perfectly standardized among manufacturers.
ACKNOWLEDGEMENTS
None.
AUTHOR CONTRIBUTIONS
Martínez-Bujidos M, Morales-Indiano C and Fernández-Prendes C contributed to the conception and design of the study; Rodríguez-Domínguez J, Piedra-Aguilera A, and Fernández-Prendes C interpreted the results; Piedra-Aguilera A performed the statistical analysis; Rodríguez-Domínguez J, Piedra-Aguilera A, Malumbres-Serrano S and Fernández-Prendes C drafted the manuscript; and Fernández-Prendes C supervised the study. All authors read and approved the final manuscript.
CONFLICTS OF INTEREST
None declared.
RESEARCH FUNDING
None declared.
References
- Wilson PWF, Polonsky TS, Miedema MD, Khera A, Kosinski AS, Kuvin JT. Systematic review for the 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation 2019;139:e1144-61.
- Havel RJ, Eder HA, Bragdon JH. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest 1955;34:1345-53.
- Ordovas JM. Fast Ultracentrifugation Methods for the Separation of Plasma Lipoproteins. Methods Mold Biol 1998;110:93-103.
- Frost PH, Havel RJ. Rationale for use of non-high-density lipoprotein cholesterol rather than low-density lipoprotein cholesterol as a tool for lipoprotein cholesterol screening and assessment of risk and therapy. Am J Cardiol 1998;81:26B-31B.
- Ertürk Zararsız G, Bolat S, Cephe A, Kochan N, Yerlitaş SI, Doğan HO, Zararsız G. Validation of Friedewald, Martin-Hopkins and Sampson low-density lipoprotein cholesterol equations. PLoS One 2022;17:e0263860.
- Martin SS, Blaha MJ, Elshazly MB, Brinton EA, Toth PP, McEvoy JW, et al. Friedewald-estimated versus directly measured low-density lipoprotein cholesterol and treatment implications. J Am Coll Cardiol 2013;62:732-9.
- Nauck M, Warnick GR, Rifai N. Methods for measurement of LDL-cholesterol: a critical assessment of direct measurement by homogeneous assays versus calculation. Clin Chem 2002;48:236-54.
- Miki Y. A homogeneous assay for the selective measurement of LDL-cholesterol in serum. Enzymatic selective protection method. Clin Lab 1999;45:398-401.
- EFLM Biological Variation Database. https://biologicalvariation.eu (Updated on April 2023).
- Azimi V, Farnsworth CW, Roper SM. Comparison of the Friedewald equation with Martin and Sampson equations for estimating LDL cholesterol in hypertriglyceridemic adults. Clin Biochem 2022;108:1-4.