Index > 6VCC > Epidemiology and Cardiovascular Prevention Brief Communication Accuracy of the Friedewald Formula Being in Significant Relation to Total Cholesterol Rudolf Gaško, Caio Mauricio Mendes de Cordova*, Slávka Geletková Railway Hospital and Policlinic Košice, Biostatistic Unit and Dept. of Clinical Biochemistry, Košice, Slovak Republic, *Departamento de Ciéncias Farmacéuticas, Fundacao Universidade Regional De Blumenau, Blumenau, Santa Catarina, Brazil Abstract Introduction Numerous epidemiological and clinical studies reported an independent relationship between increases in LDL-cholesterol (LDL-C)* concentrations and the risk of development of a coronary heart disease. European Fourth Joint Task Force on Cardiovascular Disease Prevention in Clinical Practice [1] and the US NCEP ATP III also [2] recommends reducing LDL-C to below certain cut-off points as the primary target of therapy. The reference method for measuring LDL-C is the ultracentrifugation with beta quantification procedure (BQ), which is not routinely available in clinical laboratories. The most common approach to determining LDL-C in the clinical laboratory worldwide is a calculation based on the Friedewald formula [3]. However, there are several well-established limitations to the Friedewald calculation, which led an expert panel convened by the NCEP to recommending development of accurate methods for direct determination of LDL-C [4, 5]. Recently, a new generation of homogeneous assays has been introduced as direct methods, which have been certified by the US CRMLN [4, 5]. Many studies have previously reported that the LDL-C calculated by the Friedewald formula (FLDL-C) differs from the LDL-C assessed by direct measurement (DLDL-C) or by the reference method [6]. However, those data showed that differences occurred only over parts of the concentration ranges for TG, LDL-C and VLDL-C [7-12]. Furthermore, only one study shows a significant relationship to TC [13]. This study compares the effects of both TC and TG concentrations on the difference between the FLDL-C and DLDL-C in two other populations from Slovakia and Brazil, under conditions allowing for application of the Friedewald formula. *Non-standard abbreviations: LDL-C, HDL-C, and VLDL-C, LDL-, HDL-, and VLDL-cholesterol; TG, triglyceride; TC, total cholesterol; NCEP, National Cholesterol Education Program; ATP III, Adult Treatment Panel III; BQ, beta-quantification; CRMNL, Cholesterol Reference Methods Laboratory Network; Methods Košice dataset For this part of study the blood samples were obtained from 11,331 inpatients and outpatients who visited the Railway Hospital and Policlinic in Košice, Slovakia, from January 2008 to March 2009. Age range: 18-97 years, mean age: 59 years, males: 53.3%, females: 46.7%. The Institutional Ethics Board approved this examination. All the persons fasted overnight (for 12 h) before blood collection, and the levels of TC, TG, high-density lipoprotein cholesterol (HDL-C) and DLDL-C in serum samples were measured on the day of blood collection. Sera with TG level >4.5 mmol/l (1,007 samples, i.e. 8.89% out of 11,331 samples), were excluded. We assessed the total number of 10,324 samples. TC, TG and HDL-C were measured using enzymatic methods (Pliva Lachema Diagnostika, Brno, Czech Republic). DLDL-C were measured using a LDL Cholesterol Direct Liquid assay (Pliva Lachema Diagnostika, Brno, Czech Republic), which is based on the Wako Chemicals method. All measurements were performed on a Olympus AU400 automated clinical chemistry analyzer (Olympus, Hamburg, Germany). As regards samples with triglyceride levels < 4,5 mmol/l, the LDL-C level was estimated using the Friedewald formula: FLDL-C = TC HDL-C (TG/2.2). Blumenau dataset. This part of the study assessed blood samples of 10,664 patients who sought treatment at Laboratório Santa Isabel de Análises Clínicas, Blumenau, SC, Brazil, to undergo TC, LDL-C, HDL-C, and TG measurements from January 2000 to December 2002. Their age ranged from 14 to 93 years (females: 54.8%, males: 45.2%). Blood samples were collected after a 12- to 14-hour period of fasting, incubated in a hot-water bath for 15 minutes for coagulation, and centrifuged at 2,000g for 5 minutes. The serum was separated and the assays were performed on the day of sample collection. Sera with TG level >400 mg/dl (340 samples, 3.18% of out 10,664 samples) were excluded. We assessed 10,324 samples. The measurements of the TG and TC were performed with the following reagents: Triglycerides FS (DiaSys Diagnostic Systems GmbH & Co. KG, Holzheim, Germany), and Cholesterol (BioSystems S.A., Barcelona, Spain), respectively according to the specifications of the manufacturers, in a Spectrum CCX II device (Abbott Diagnostics, Abbott Park, IL, USA). The HDL-C measurement was performed using a homogeneous method without precipitation with the HDL-C Immuno FS reagent (DiaSys). The tests were calibrated with the CCX Multicalibrator Set (Abbott), with curves of 3 points. The LDL-C measurement with the homogeneous method was performed with the reagent LDL-C Select FS (DiaSys), which
is based on the Wako Chemicals method. For samples with triglyceride levels < 400 mg/dl, the LDL-C level was estimated using the Friedewald formula: FLDL-C = TC HDL-C (TG/5). The patients data had been used in study (8) originally. The % DLDL was calculated using the following formula: % DLDL=[(FLDL-C DLDL-C)/DLDL-C]x100, for both datasets. Statistical analysis Descriptive statistics, means and standard deviations were calculated, and charts were drawn using Microsoft Office Excel 2007. Paired t-test and correlation analysis was performed using MedCalc, Ver.10.3 (MedCalc Software, Mariakerke, Belgium). Results Košice dataset The correlation coefficient between the FLDL-C and DLDL-C was 0.94 (p<0.01), and there was a significant difference between the means of the two groups (p<0.01, paired t-test). The mean ±SD of % LDL was 13.3±10.2%, and the quantile distribution is shown in Figure 1a. The % LDL values differed by more than ±5% in case of 79.6% of patients and by more than ±10% in case of 56.1% of patients. The FLDL-C value was lower than the DLDL-C value in case of 89.9% of patients. Figure 1. Distribution of the percentage differences between the LDL-C measured directly using a homogeneous assay and that calculated using Friedewald formula, data from Kosice (10,324 patients), facsimile from ref. [13]. The mean % LDL values for the lowest and highest TG concentration groups were 9.5±9.3% and -31.6±15.1% respectively, i.e. a 3.3-fold difference (Table 1, Figure 2). The TC concentration also affected the % LDL, resulting in mean % LDL values for the lowest and highest TC groups of -19.4±9.6% and 2.1±14.2% respectively, i.e. a 9.2-fold difference. The group with the lowest TG and highest TC concentrations did not contain the lowest mean % LDL (3.4±15.0%) it was in the group with the lowest TG and second highest TC (1.5±14.6). The mean % LDL in the group with the highest TG and lowest TC concentrations was 39.4-fold the lowest mean ( 59.1±24.1%). Tabla 1. Summary of the percentage differences between the LDL-C measured directly using a homogeneous assay and that calculated using the Friedewald formula with regard to the triglyceride and total cholesterol concentrations; data from Kosice, Slovakia.
Figue 2. Concentrations of triglyceride and total cholesterol affect the percentage difference between the LDL-C measured directly using a homogeneous assay and that calculated using the Friedewald formula, data from Kosice. Blumenau dataset The correlation coefficient between the FLDL-C and DLDL-C was 0.93 (p<0.001), and there was a significant difference between the means of the two groups (p<0.01, paired t-test). The mean±sd of % LDL was 8.5±14.6%, and the quantile distribution is shown in Figure 1b. The % LDL values differed by more than ±5% in case of 82.9% of patients and by more than ±10% in case of 61.2% of patients. The FLDL-C value was lower than the DLDL-C value in case of 25.4% of patients. The mean % LDL values for the lowest and highest TG concentration groups were 12.8±12.2% and -6.5±11.6% respectively, i.e. a 3.3-fold difference (Table 2, Figure 3). The TC concentration also affected the % LDL, resulting in mean % LDL values for the lowest and highest TC groups of 11.4±19.3% and 6.5±14.6%, respectively, i.e. a 1.8-fold difference. The group with the lowest TG and highest TC concentrations did not contain the lowest mean % LDL. The mean % LDL in the group with the highest TG and lowest TC concentrations was 35.4-fold the lowest mean ( 35.4±0.2%). Tabla 2. Summary of the percentage differences between the LDL-C measured directly using a homogeneous assay and that calculated using the Friedewald formula with regard to the triglyceride and total cholesterol concentrations; data from Blumenau, Santa Catarina, Brazil.
Figure 3. Concentrations of triglyceride and total cholesterol affect the percentage difference between the LDL-C measured directly using a homogeneous assay and that calculated using the Friedewald formula, data from Blumenau. For comparison, a facsimile of plots based on the data from Korea [13] is shown in Figure 1c, and Figure 4. % LDL from all the three study populations are summarized in Figure 5. Figure 4. Concentrations of triglyceride and total cholesterol affect the percentage difference between the LDL-C measured directly using a homogeneous assay and that calculated using the Friedewald formula, data from Seoul, facsimile from ref. [13].
Figure 5. Comparison of percentage differences, data from Kosice, Blumenau and Seoul. Discussion The European and US guidelines are in agreement as regards the fact that LDL-C lowering comprises the principal target of lipid treatment. The problem is to achieve a reliable determination of LDL-C (reducing variability of Friedewald LDL-C.) A major disadvantage related to calculating LDL-C is that the variability is a product of combined variabilities in the three underlying measurements. The NCEP Expert Panel observed in experienced and well-standardized lipid laboratories that total analytical variability in calculated LDL-C averaged 4.0%, ranging between 2.7% and 6.8% for LDL-C concentrations between 2.59-5.83 mmol/l [14]. In routine laboratories, variability appeared to be much higher, e.g. eight survey samples of the College of American Pathologists (CAP) Comprehensive Chemistry Survey analyzed in more than 1,150 laboratories gave overall CVs averaging 12% [14]. This CV reflects not only imprecision within laboratories, but also method-to-method biases from the many different assays used in TC, TG, and HDL-C determinations. The Panel, concluding that many routine laboratories would not be able to achieve the requisite analytical performance using the Friedewald calculation, recommended development of more precise direct methods. The TC determinations have the most significant effect on variability in the calculation [4]. TG values are divided by 2.2 (or 5, for non SI Units) and HDL-C concentrations are relatively lower, diminishing their impact. Observations from the lipid laboratories as well as from the CAP survey of routine laboratories suggested that CVs for LDL-C were approximately a double of those for TC [14]. However, increasing TGs contribute progressively to a higher variability the calculated LDL-C. Adoption of fully automated homogeneous methods for HDL-C is reality in routine laboratories (including Slovakia and Brazil) in the last years, and addresses the matter of imprecision, including the contribution to calculated LDL-C; nevertheless, the conclusions of the NCEP Expert Panel are likely to be still valid. Priority findings on the direct proportionality of total cholesterol and size % LDL, authors Jun et al [13], were confirmed in both datasets. Both TC and TG represent significant variables affecting the difference between the directly measured and the calculated LDL-C over the entire range of TC and TG values; the %DLDL increased as TC decreased and TG increased. Differences in the average value percentage may be due to - Analytical uncertainty. Analytical uncertainty can not be excluded despite the described good value CV and external quality assessment. We compared two index methods, not a reference method with an index test. - Different composition of patients in the sets. It was demonstrated that DM, hepatopathy, nephropathy is substantially different in FF bias [6]. The divergence structure of both the sets suggests that 8.89% of patients in Kosice had TG higher then 4.5 mmol/l, but only 3.18% of patients in Blumenau, and TC higher than 6.48 mmol/l the ratios were 23.84% and 8.04% respectively. Quite on the contrary, TC less than 3.89 mmol/l the ratios were 4.44% and 13.71% respectively. - Racial genetic differences [15, 16]. A percentage difference of more than 10% is not surprising as it has already been described. Branchi et al [17] describe 34% of samples from diabetic patients and 26% of samples from non-diabetic patients, but once TG level exceeded 2.26 mmol/l, 45% of samples from diabetic patients and 34% of samples from others differed from measured LDL-C by >10%. Recent data demonstrate that apolipoprotein B is a better means of measurement of the circulating LDL particle number concentration and it is also a more reliable indicator of risk than LDL-C, and there is growing support for the idea that addition of an apo B measurement to the routine lipid panel for assessing and monitoring patients at risk for cardiovascular disease would enhance patient management [18, 19].
Conclusions It is to be stated that we observed a poor concordance between the calculated and measured LDL cholesterol in three large populations from Brazil, Slovakia and South Korea, despite a good correlation between the two methods. The accuracy of the Friedewald formula proves to be in a significant relation to total cholesterol. Further clinical evaluation is required to verify this finding. Acknowledgement This study has been supported by a research grant from the Slovak Society of Cardiology 2008. References 1.Graham I, Atar D, Borch-Johnsen K, et al. European guidelines on cardiovascular disease prevention in clinical practice: executive summary. Eur Heart J 2007;28:2375-2414. 2. Third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) final report. Circulation 2002;106:3143-3421. 3. 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. 4. 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;48:236 254. 5. Bachorik PS, Ross JW. National Cholesterol Education Program recommendations for measurement of low-density lipoprotein cholesterol: executive summary. The National Cholesterol Education Program Working Group on Lipoprotein Measurement. Clin Chem 1995;41:1414 1420. 6. Gaško R, Sánchez-Meca J. LDL-cholesterol: A critical assessment of analytical accuracy Friedewald`s formula. Metaanalysis. Cardiol, 2008, accepted, paper in press. 7. Ahmadi S, Boroumand M, Gohan-Moghaddam K, et al. The impact of low serum triglyceride on LDL-cholesterol estimation. Arch Iran Med 2008;11:318-321. 8. Cordova CM, Schneider CR, Juttel ID, Cordova MM. Comparison of LDL-cholesterol direct measurement with the estimate using the Friedewald formula in a sample of 10,664 patients. Arq Bras Cardiol 2004;83:482 487. 9. Scharnagl H, Nauck M, Wieland H, Marz W. The Friedewald formula underestimates LDL cholesterol at low concentrations. Clin Chem Lab Med 2001;39:426 431. 10. Sniderman AD, Blank D, Zakarian R, et al. Triglycerides and small dense LDL: the twin Achilles heels of the Friedewald formula. Clin Biochem 2003;36:499 504. 11. Tremblay AJ, Morrissette H, Gagne JM, et al. Validation of the Friedewald formula for the determination of low-density lipoprotein cholesterol compared with beta-quantification in a large population. Clin Biochem 2004;37:785 790. 12. Wang TY, Haddad M, Wang TS. Low triglyceride levels affect calculation of low-density lipoprotein cholesterol values. Arch Pathol Lab Med 2001;125:404 405. 13. Jun KR, Park H, Chun S, et al. Effects of total cholesterol and triglyceride on the percentage difference between the lowdensity lipoprotein cholesterol concentration measured directly and calculated using the Friedewald formula. Clin Chem Lab Med 2008;46:371-375. 14. National Cholesterol Education Program. Recommendations on lipoprotein measurement. Working Group on Lipoprotein Measurement. NIH Publication No. 95-3044. 1995 National Institutes of Health, National Heart, Lung, and Blood Institute Bethesda, MD. 15. McQeen MJ, Hawken S, Wang X, et al. Lipids, lipoproteins, and apolipoproteins as risk markers of myocardial infarction in 52 countries (the INTERHEART study): a case-control study. Lancet 2008;372:224-233. 16. Sandhu MS, Waterworth DM, Deberham SI, et al. LDL-cholesterol concentrations: a genome wide association study. Lancet 2008;371:483-491. 17. Branchi A, Rovellini A, Torri A, Sommariva D. Accuracy of calculated serum low-density lipoprotein cholesterol for assessment of coronary heart disease risk in NIDDM patients. Diabetes Care 1998;21:1397-1402. 18. Contois JH, McConnell JP, Sethi AA, et al. Apolipoprotein B and cardiovascular disease risk: Position statement from the AACC Lipoproteins and Vascular Diseases Division Working Group on best practices. Clin Chem 2009;55:407-419. 19. Stanley I, Lichtenstein AH, Chung M, et al. Association of low-density lipoprotein subfractions with cardiovascular outcomes (Systematic review). Ann Intern Med 2009;15:474-484.
Publication: September - November/2009 Your questions, contributions and commentaries will be answered by the lecturer or experts on the subject in the Epidemiology and Cardiovascular Prevention list. Please fill in de form and Press the "Send" button. Question, contribution or commentary: Name and Surname: Country: Argentina E-Mail address: Re-type Email address: Send Erase 1994-2009 2009 CETIFAC - Bioengineering UNER - 6VCC ISBN 978-987-22746-1-0 Updated: 10/06/2009 - DHTML JavaScript Menu By Milonic.com - Webmaster - HonCode - pwmc