A comparative evaluation of low-density lipoprotein cholesterol estimation: Machine learning algorithms versus various equations.

BACKGROUND: Given the critical importance of Low-density lipoprotein cholesterol (LDL-C) levels in determining cardiovascular risk, it is essential to measure LDL-C accurately. Since the Friedewald formula generates incorrect predictions in many circumstances, new equations have been developed to overcome the Friedewald equations' shortcomings. This study aimed to compare estimated LDL-C with directly measured LDL-C (dLDL-C), as well as their performance in predicting LDL-C, utilizing Friedewald, extended Martin-Hopkins, Sampson, de Cordova, and Vujovic formulas and five machine learning (ML) algorithms. METHODS: A total of 29,504 samples from the ISLAB-2 Core Laboratory were included in the study. All statistical analysis was performed using R version 4.1.2. Statistical Language. RESULTS: Bayesian-Regularized Neural Network (BRNN) (r = 0.957) and Random Forest (RF) (r = 0.957) algorithms showed a higher correlation with dLDL-C than the other equations in all-testing dataset. All ML algorithms demonstrated less bias than pre-existing LDL-C equations with dLDL-C and outperformed the LDL-C estimation equations in terms of concordance in all-testing dataset. CONCLUSIONS: The results of our research indicate that when compared to conventional equations, ML algorithms are much more effective in predicting LDL-C. ML algorithms, aided by a vast dataset, could have the capability to predict LDL-C levels even in cases where triglyceride levels are high, unlike the limited usage of Friedewald formula.

A Novel LDLC Equation is Superior to the NIH LDLC Equation and the Friedewald Equation.

Background: LDLC equations have varying levels of underestimation for the calculated LDLC. Therefore, underestimating LDLC should be avoided as much as possible. We need to establish LDLC equations that underestimate LDLC as little as possible. Methods: We established the equations with a healthy cohort from Shuyang Hospital and validated the equations with an unselected patient cohort from The Second People's Hospital of Lianyungang. We established the novel LDLC equations by using the regression equation. The relationship between two markers was analysed using Pearson's approach. The 95% limits of measuring agreement within ±2 SD for the LDLC equations was performed using Bland‒Altman analysis. ROC curve analysis was used to predict LDLC levels and the accuracy of the LDLC equation for determining the direct LDLC levels at LDLC cut-offs was assessed. Results: We obtained two novel LDLC equations (LDL_nonHDLC equation=-0.899+1.195*nonHDLC-0.00347*nonHDLC2 and LDL_TC(total cholesterol) equation=-2.775+1.29*TC -0.00990* TC 2). The correlation coefficient between the novel LDLC equation and the direct LDLC measurements is not lower than that between the LDL_NIH equation and the direct LDLC measurements. The AUCs of our novel LDLC equations were greater than those of the LDL_NIH equation and the LDL_F equation at the LDLC cut-offs for clinical decision-making. The measuring agreement in the methods of the LDL_nonHDL equation is superior to that of the LDL_NIH equation. Conclusion: LDLC calculated by the novel LDL_nonHDL equation exhibited superiority over the LDL_NIH equation. Combining the LDL_NIH equation and our novel LDLC equation may improve accuracy and avoid undertreatment of high LDLC levels.

Prediction of low-density lipoprotein cholesterol levels using machine learning methods.

OBJECTIVE: Low-density lipoprotein cholesterol (LDL-C) has been commonly calculated by equations, but their performance has not been entirely satisfactory. This study aimed to develop a more accurate LDL-C prediction model using machine learning methods. METHODS: The study involved predicting directly measured LDL-C, using individual characteristics, lipid profiles, and other laboratory results as predictors. The models applied to predict LDL-C values were multiple regression, penalized regression, random forest, and XGBoost. Additionally, a novel 2-step prediction model was developed and introduced. The machine learning methods were evaluated against the Friedewald, Martin, and Sampson equations. RESULTS: The Friedewald, Martin, and Sampson equations had root mean squared error (RMSE) values of 12.112, 8.084, and 8.492, respectively, whereas the 2-step prediction model showed the highest accuracy, with an RMSE of 7.015. The LDL-C levels were also classified as a categorical variable according to the diagnostic criteria of the dyslipidemia treatment guideline, and concordance rates were calculated between the predictive values obtained from each method and the directly measured ones. The 2-step prediction model had the highest concordance rate (85.1%). CONCLUSION: The machine learning method can calculate LDL-C more accurately than existing equations. The proposed 2-step prediction model, in particular, outperformed the other machine learning methods.

Correlation of extended Martin/Hopkins equation with a direct homogeneous assay in assessing low-density lipoprotein cholesterol in patients with hypertriglyceridemia.

BACKGROUND: The Friedewald or Martin/Hopkins equation is widely used to estimate low-density lipoprotein cholesterol (LDL-C) at triglyceride (TG) levels <400 mg/dL. In this study, we aimed to validate the recently developed Sampson and extended Martin/Hopkins equations intended for use in patients with TG levels up to 800 mg/dL by comparing them to a direct homogenous assay. METHODS: In total, 8676 participants with serum TG levels <800 mg/dL were enrolled in this study. LDL-C was directly measured using Abbott homogeneous assay (DLDL) and estimated using the Friedewald (FLDL), Martin/Hopkins (MLDL), extended Martin/Hopkins (EMLDL), and Sampson equations (SLDL). The overall concordance between the DLDL and LDL-C estimates was calculated. The performance of the four equations was also compared using Bland-Altman plots and mean absolute difference (MAD). RESULTS: The EMLDL was more accurate than other LDL-C equations particularly for patients with TG≥400 mg/dL (MAD = 10.43; vs. FLDL: MAD = 21.1; vs. SLDL: MAD 11.62). The overall concordance of FLDL, MLDL, EMLDL, and SLDL with DLDL in TG values ranging from 200 to 799 mg/dL were 52.2, 70.5, 71.6, and 65.7%, respectively (p < 0.001), demonstrating the EMLDL as the most optimal estimation method, particularly for high TG levels (≥200 mg/dL). CONCLUSION: Both the original and extended Martin/Hopkins method are optimal in estimating LDL-C levels in clinical laboratories using the Abbott analyzer in patients with TG levels of 200-399 and 400-799 mg/dL, respectively. Meanwhile, caution is need that considerable underestimation of Friedewald and Sampson equation could lead to undertreatment in hypertriglyceridemia.

An increase in calculated small dense low-density lipoprotein cholesterol predicts new onset of hypertension in a Japanese cohort.

A disorder of lipid metabolism is involved in cardiovascular diseases including hypertension. A high level of small dense low-density lipoprotein cholesterol (sdLDL-C) is a strong risk factor for atherosclerotic cardiovascular disease. However, the association between sdLDL-C and hypertension has not been fully investigated. We investigated the associations between the development of hypertension during a 10-year period and levels of total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), non-HDL-C, triglycerides (TG), and LDL-C and sdLDL-C calculated by using the Sampson equations in 28,990 Japanese subjects who received annual health examinations. After exclusion of subjects with missing data, those with hypertension, and those with TG ≥ 800 mg/dL at baseline, a total of 15,177 subjects (men/women: 9374/5803, mean age: 46 years) were recruited. During the 10-year period, 2379 men (25.4%) and 724 women (12.5%) had new onset of hypertension. Multivariable Cox proportional hazard model analyses showed that levels of HDL-C, non-HDL-C, TG and sdLDL-C, but not levels of TC and LDL-C, were independent risk factors for the development of hypertension after adjustment of age, sex, family history of hypertension, systolic blood pressure, obesity, current smoking habit, alcohol drinking habit, estimated glomerular filtration rate, diagnosis of diabetes mellitus and use of lipid-lowering drugs and that the adjusted risk of sdLDL-C (per 1-standard deviation) was highest (hazard ratio [95% confidence interval: 1.09 [1.05-1.13]). The addition of sdLDL-C to traditional risk factors for hypertension significantly improved the discriminatory capability, which was better than that of other lipid fractions. In conclusion, a high level of calculated sdLDL-C predicts the development of hypertension.

Measurement of Serum Low Density Lipoprotein Cholesterol and Triglyceride-Rich Remnant Cholesterol as Independent Predictors of Atherosclerotic Cardiovascular Disease: Possibilities and Limitations.

The serum low density lipoprotein cholesterol (LDL-C) concentration is the dominant clinical parameter to judge a patient's risk of developing cardiovascular disease (CVD). Recent evidence supports the theory that cholesterol in serum triglyceride-rich lipoproteins (TRLs) contributes significantly to the atherogenic risk, independent of LDL-C. Therefore, combined analysis of both targets and adequate treatment may improve prevention of CVD. The validity of TRL-C calculation is solely dependent on the accuracy of the LDL-C measurement. Direct measurement of serum LDL- C is more accurate than established estimation procedures based upon Friedewald, Martin-Hopkins, or Sampson equations. TRL-C can be easily calculated as total C minus high density lipoprotein C (HDL-C) minus LDL-C. Enhanced serum LDL-C or TRL-C concentrations require different therapeutic approaches to lower the atherogenic lipoprotein C. This review describes the different atherogenic lipoproteins and their possible analytical properties and limitations.

Accuracy of 23 Equations for Estimating LDL Cholesterol in a Clinical Laboratory Database of 5,051,467 Patients.

Background: Alternatives to the Friedewald low-density lipoprotein cholesterol (LDL-C) equation have been proposed. Objective: To compare the accuracy of available LDL-C equations with ultracentrifugation measurement. Methods: We used the second harvest of the Very Large Database of Lipids (VLDbL), which is a population-representative convenience sample of adult and pediatric patients (N = 5,051,467) with clinical lipid measurements obtained via the vertical auto profile (VAP) ultracentrifugation method between October 1, 2015 and June 30, 2019. We performed a systematic literature review to identify available LDL-C equations and compared their accuracy according to guideline-based classification. We also compared the equations by their median error versus ultracentrifugation. We evaluated LDL-C equations overall and stratified by age, sex, fasting status, and triglyceride levels, as well as in patients with atherosclerotic cardiovascular disease, hypertension, diabetes, kidney disease, inflammation, and thyroid dysfunction. Results: Analyzing 23 identified LDL-C equations in 5,051,467 patients (mean±SD age, 56±16 years; 53.3% women), the Martin/Hopkins equation most accurately classified LDL-C to the correct category (89.6%), followed by the Sampson (86.3%), Chen (84.4%), Puavilai (84.1%), Delong (83.3%), and Friedewald (83.2%) equations. The other 17 equations were less accurate than Friedewald, with accuracy as low as 35.1%. The median error of equations ranged from -10.8 to 18.7 mg/dL, and was best optimized using the Martin/Hopkins equation (0.3, IQR-1.6 to 2.4 mg/dL). The Martin/Hopkins equation had the highest accuracy after stratifying by age, sex, fasting status, triglyceride levels, and clinical subgroups. In addition, one in five patients who had Friedewald LDL-C <70 mg/dL, and almost half of the patients with Friedewald LDL-C <70 mg/dL and triglyceride levels 150-399 mg/dL, had LDL-C correctly reclassified to >70 mg/dL by the Martin/Hopkins equation. Conclusions: Most proposed alternatives to the Friedewald equation worsen LDL-C accuracy, and their use could introduce unintended disparities in clinical care. The Martin/Hopkins equation demonstrated the highest LDL-C accuracy overall and across subgroups.

Comparison of Newly Proposed LDL-Cholesterol Estimation Equations.

BACKGROUND: Low-density lipoprotein cholesterol is an important marker highly associated with cardiovascular disease. Since the direct measurement of it is inefficient in terms of cost and time, it is common to estimate through the Friedewald equation developed about 50 years ago. However, various limitations exist since the Friedewald equation was not designed for Koreans. This study proposes a new low-density lipoprotein cholesterol estimation equation for South Koreans using nationally approved statistical data. METHODS: This study used data from the Korean National Health and Nutrition Examination Survey from 2009 to 2019. The 18,837 subjects were used to develop the equation for estimating low-density lipoprotein cholesterol. The subjects included individuals with low-density lipoprotein cholesterol levels directly measured among those with high-density lipoprotein cholesterol, triglycerides, and total cholesterol measured. We compared twelve equations developed in the previous studies and the newly proposed equation (model 1) developed in this study with the actual low-density lipoprotein cholesterol value in various ways. RESULTS: The low-density lipoprotein cholesterol value estimated using the estimation formula and the actual low-density lipoprotein cholesterol value were compared using the root mean squared error. When the triglyceride level was less than 400 mg/dL, the root mean squared of the model 1 was 7.96, the lowest compared to other equations, and the model 2 was 7.82. The degree of misclassification was checked according to the NECP ATP III 6 categories. As a result, the misclassification rate of the model 1 was the lowest at 18.9%, and Weighted Kappa was the highest at 0.919 (0.003), which means it significantly reduced the underestimation rate shown in other existing estimation equations. Root mean square error was also compared according to the change in triglycerides level. As the triglycerides level increased, the root mean square error showed an increasing trend in all equations, but it was confirmed that the model 1 was the lowest compared to other equations. CONCLUSION: The newly proposed low-density lipoprotein cholesterol estimation equation showed significantly improved performance compared to the 12 existing estimation equations. The use of representative samples and external verification is required for more sophisticated estimates in the future.

Comparison of low-density lipoprotein cholesterol estimation methods in individuals with insulin resistance: A cross-sectional study.

BACKGROUND AND AIMS: The Friedewald, Sampson, and Martin equations were developed to estimate low-density lipoprotein cholesterol levels; however, the validation data of these equations with and without insulin resistance are insufficient. MATERIALS AND METHODS: We collected data on low-density lipoprotein cholesterol and lipid profiles from the Korea National Health and Nutrition Examination Survey. Using the data on insulin requirement, insulin resistance was calculated for 4,351 participants (median age, 48 [36-59] years; 49.9% male) using the homeostatic model assessment for insulin resistance (n = 2,713) and quantitative insulin-sensitivity check index (n = 2,400). RESULTS: According to the mean and median absolute deviation, the Martin equation yielded more accurate estimates than other equations when the triglyceride level was < 400 mg/dL with insulin resistance; the Sampson equation yielded lower estimates when the direct low-density lipoprotein cholesterol level was < 70 mg/dL and triglyceride level was < 400 mg/dL without insulin resistance. However, the three equations yielded similar estimates when the triglyceride level was < 150 mg/dL with and without insulin resistance. CONCLUSION: The Martin equation yielded more appropriate estimates than the Friedewald and Sampson equations for triglyceride levels < 400 mg/dL with and without insulin resistance. If the triglyceride level was < 150 mg, the Friedewald equation could also be considered.

Evaluation of low-density lipoprotein cholesterol equations by cross-platform assessment of accuracy-based EQA data against SI-traceable reference value.

OBJECTIVES: Low-density lipoprotein cholesterol (LDLC) is the primary cholesterol target for the diagnosis and treatment of cardiovascular disease (CVD). Although beta-quantitation (BQ) is the gold standard to determine LDLC levels accurately, many clinical laboratories apply the Friedewald equation to calculate LDLC. As LDLC is an important risk factor for CVD, we evaluated the accuracy of Friedewald and alternative equations (Martin/Hopkins and Sampson) for LDLC. METHODS: We calculated LDLC based on three equations (Friedewald, Martin/Hopkins and Sampson) using the total cholesterol (TC), triglycerides (TG), and high-density lipoprotein cholesterol (HDLC) in commutable serum samples measured by clinical laboratories participating in the Health Sciences Authority (HSA) external quality assessment (EQA) programme over a 5 years period (number of datasets, n=345). LDLC calculated from the equations were comparatively evaluated against the reference values, determined from BQ-isotope dilution mass spectrometry (IDMS) with traceability to the International System of Units (SI). RESULTS: Among the three equations, Martin/Hopkins equation derived LDLC had the best linearity against direct measured (y=1.141x - 14.403; R2=0.8626) and traceable LDLC (y=1.1692x - 22.137; R2=0.9638). Martin/Hopkins equation (R2=0.9638) had the strongest R2 in association with traceable LDLC compared with the Friedewald (R2=0.9262) and Sampson (R2=0.9447) equation. The discordance with traceable LDLC was the lowest in Martin/Hopkins (median=-0.725%, IQR=6.914%) as compared to Friedewald (median=-4.094%, IQR=10.305%) and Sampson equation (median=-1.389%, IQR=9.972%). Martin/Hopkins was found to result in the lowest number of misclassifications, whereas Friedewald had the most numbers of misclassification. Samples with high TG, low HDLC and high LDLC had no misclassification by Martin/Hopkins equation, but Friedewald equation resulted in ∼50% misclassification in these samples. CONCLUSIONS: The Martin/Hopkins equation was found to achieve better agreement with the LDLC reference values as compared to Friedewald and Sampson equations, especially in samples with high TG and low HDLC. Martin/Hopkins derived LDLC also enabled a more accurate classification of LDLC levels.

Comparison of low-density lipoprotein cholesterol equations in patients with dyslipidaemia receiving cholesterol ester transfer protein inhibition.

AIMS: Low-density lipoprotein (LDL-C) lowering is imperative in cardiovascular disease prevention. We aimed to compare accuracy of three clinically-implemented LDL-C equations in a clinical trial of cholesterol ester transfer protein (CETP) inhibition. METHODS AND RESULTS: Men and women aged 18-75 years with dyslipidaemia were recruited from 17 sites in the Netherlands and Denmark. Patients were randomly assigned to one of nine groups using various combinations of the CETP inhibitor TA-8995 (obicetrapib), statin therapy, and placebo. In pooled measurements over 12 weeks, we calculated LDL-C by the Friedewald, Martin/Hopkins, and Sampson equations, and compared values with preparative ultracentrifugation (PUC) LDL-C overall and with a special interest in the low LDL-C/high triglycerides subgroup. There were 242 patients contributing 921 observations. Overall median LDL-C differences between estimates and PUC were small: Friedewald, 0.00 (25th, 75th: -0.10, 0.08) mmol/L [0 (-4, 3) mg/dL]; Martin/Hopkins, 0.02 (-0.08, 0.10) mmol/L [1 (-3, 4) mg/dL]; and Sampson, 0.05 (-0.03, 0.15) mmol/L [2 (-1, 6) mg/dL]. In the subgroup with estimated LDL-C <1.8 mmol/L (<70 mg/dL) and triglycerides 1.7-4.5 mmol/L (150-399 mg/dL), the Friedewald equation underestimated LDL-C with a median difference versus PUC of -0.25 (-0.33, -0.10) mmol/L [-10 (-13, -4) mg/dL], whereas the median difference by Martin/Hopkins was 0.00 (-0.08, 0.10) mmol/L [0 (-3, 4) mg/dL] and by Sampson was -0.06 (-0.13, 0.00) mmol/L [-2 (-5, 0) mg/dL]. In this subgroup, the proportion of LDL-C observations <1.8 mmol/L (<70 mg/dL) that were correctly classified compared with PUC was 71.4% by Friedewald vs. 100.0% by Martin/Hopkins and 93.1% by Sampson. CONCLUSION: In European patients with dyslipidaemia receiving a CETP inhibitor, we found improved LDL-C accuracy using contemporary equations vs. the Friedewald equation, and the greatest accuracy was observed with the Martin/Hopkins equation. REGISTRATION: ClinicalTrials.gov, NCT01970215.

Effect of lipoprotein (a) on the analytical determination of low-density lipoprotein cholesterol (LDLc) and its influence on pharmacological treatment with atorvastatin.

Lipoprotein(a) (Lp(a)) and Low-density lipoprotein cholesterol (LDLc) is an important risk factor for atherosclerotic cardiovascular disease. The objective of this study was to determine the impact of Lp(a) concentration both on the indirect analytical measurement of LDLc and on the efficacy of dyslipidaemia treatment using the atorvastatin statin. Two retrospective studies were conducted, one with 340 patients and another with 107 patients treated with atorvastatin. Lp(a) concentrations were measured by turbidimetry with an assay independent of the size of the apo(a) isoform. LDLc was calculated using the Friedewald equation and the corrected LDLc was calculated using the Dahlén equation. A strong positive correlation was observed between the serum Lp(a) concentration and the LDLc-overestimation percentage (r = 0.960, p < .001). It was also observed that as the Lp(a) concentration rose there was no significant variation in the percentage decrease in corrected LDLc during atorvastatin treatment (r = 0.186, p > .05). The concentration of LDLc obtained by using the Friedewald equation included Lp(a) cholesterol. The lowering of LDLc in patients treated with atorvastatin depended solely on accessible LDL cholesterol and not on Lp(a) cholesterol.

Indirect calculation of LDL using thirteen equations in Pakistani population.

BACKGROUND: Owing to the atherogenic properties, low density lipoprotein cholesterol (LDL-C) is the primary target for treatment and diagnosis of cardiovascular diseases (CVDs), hence accurate measurement of LDL-C is critical. Despite the availability of direct measurement assays for LDL-C, it is routinely calculated by Friedewald equation in clinical settings in Pakistan mostly due to financial constraints. However, the validity of this equation is impacted by several factors, therefore several other equations have been developed for the calculation of LDL-C. MATERIALS AND METHODS: LDL-C of 39,385 individuals measured directly by homogenous assays (dLDL) was compared with LDL-C calculated by thirteen equations (cLDL-C). Stratifications based on different lipids i.e., triglycerides (TG), total cholesterol (TC), high-density lipoprotein (HDL) were made to check the validity of these equations across all ranges of lipid profile. The correlation and median difference between dLDL and cLDL-C was statistically analyzed. RESULTS: Overall Teerakanchana equation displayed a strong positive correlation (ρ = 0.967) and least median difference (-8.81) with dLDL, followed by Martin equation (ρ = 0.967). For higher TG ranges (>500 mg/dL), Teerakanchana equation had the least median difference (1.31) and a strong correlation (ρ = 0.800). CONCLUSION: Our data suggest that Teerakanchana equation may be employed as an alternative to Friedewald equation for Pakistani population.

Discordance Between Standard Equations for Determination of LDL Cholesterol in Patients With Atherosclerosis.

BACKGROUND: Accurate estimation of low-density lipoprotein cholesterol (LDL-C) is important for guiding cholesterol-lowering therapy. Different methods currently exist to estimate LDL-C. OBJECTIVES: This study sought to assess discordance of estimated LDL-C using the Friedewald, Sampson, and Martin/Hopkins equations. METHODS: Electronic health record data from patients with atherosclerotic cardiovascular disease and triglyceride (TG) levels of <400 mg/dL between October 1, 2015, and June 30, 2019, were retrospectively analyzed. LDL-C was estimated using the Friedewald, Sampson, and Martin/Hopkins equations. Patients were categorized as concordant if LDL-C was <70 mg/dL with each pairwise comparison of equations and as discordant if LDL-C was <70 mg/dL for the index equation and ≥70 mg/dL for the comparator. RESULTS: The study included 146,106 patients with atherosclerotic cardiovascular disease (mean age: 68 years; 56% male; 91% White). The Martin/Hopkins equation consistently estimated higher LDL-C values than the Friedewald and Sampson equations. Discordance rates were 15% for the Friedewald vs Martin/Hopkins comparison, 9% for the Friedewald vs Sampson comparison, and 7% for the Sampson vs Martin/Hopkins comparison. Discordance increased at lower LDL-C cutpoints and in those with elevated TG levels. Among patients with TG levels of ≥150 mg/dL, a >10 mg/dL difference in LDL-C was present in 67%, 27%, and 23% of patients when comparing the Friedewald vs Martin/Hopkins, Friedewald vs Sampson, and Sampson vs Martin/Hopkins equations, respectively. CONCLUSIONS: Clinically meaningful differences in estimated LDL-C exist among equations, particularly at TG levels of ≥150 mg/dL and/or lower LDL-C levels. Reliance on the Friedewald and Sampson equations may result in the underestimation and undertreatment of LDL-C in those at increased risk.

Comparison of LDL-C calculation by Martin, Sampson and old Friedewald methods in real data and synthetic data set.

OBJECTIVE: LDL-cholesterol (LDL-C) is determined by methods whose accuracy is significantly affected in various clinical or analytical situations. Two computational methods have recently been described, the Martin equation and the Sampson equation, validity of which we compare with the Friedewald equation. METHODS: LDL-C comparisons determined by the 3 equations were performed on 4 real sets of lipid data, generated in various previous studies, ranging from n = 140 to n = 7 393. We have created an artificial set of data on the extent of 900 members with equally distributed values of TC, HDL-C and TG troughout the commonly found range. Such a data set is independent of the phrase "we performed the calculations on our file". Comparisons were also made on this artificial file. RESULTS: The difference between the LDL-C values determined by the different equations gradually increases with decreasing LDL-C levels, both in the subgroup of low TG values and in the subgroups of medium and higher TG values. This applies to all 4 real files as well as to the artificial file. These differences are more visible the larger the file size. For the artificial set, the overall agreement between the LDL-C categories was lowest when comparing the Friedewald and Martin equations (83.1%), higher between the Sampson and Martin equations (88.9%) and highest when comparing the Friedewald and Sampson equations (90.9%). In all 4 real sets, the trends of overestimation and underestimation between the equations were exactly the same as in the artificial set. CONCLUSION: The results of clinical and epidemiological studies are significantly influenced by the method used to determine LDL-C. When comparing the calculation methods for determining LDL-C, it is possible to preferably use the described artificial set.

Is Machine Learning-derived Low-Density Lipoprotein Cholesterol estimation more reliable than standard closed form equations? Insights from a laboratory database by comparison with a direct homogeneous assay.

BACKGROUND: There is no consensus on the best method to estimate Low Density Lipoprotein-Cholesterol (LDL-C) in routine laboratories. METHODS: We conducted a retrospective study to compare the performances of a Machine Learning (ML) algorithm using the K-Nearest Neighbors (LDL-KNN) method with that of the Friedewald formula (LDL-F), the Martin-Hopkins equation (LDL-NF), the de Cordova equation (LDL-CO) and the Sampson equation (LDL-SA) against direct homogeneous LDL-C assay (LDL-D) in patients who presented to the Laboratories of Hôtel Dieu de France university hospital in Beirut, Lebanon, from September 2017 to July 2020. Agreements between methods were analyzed using Intraclass Correlation Coefficients (ICC) and the Bland-Altman method of agreement. RESULTS: 31,922 observations from 19,279 subjects were included, with a mean age of 52 ± 18 years and 10,075 (52.3%) females. All methods except LDL-F and LDL-CO exhibited an overall ICC beyond the 0.9 cut-off. LDL-SA, LDL-NF and LDL-KNN were less susceptible to triglyceridemia than LDL-F and LDL-CO, with LDL-KNN resulting in the lesser fraction of points beyond the Bland-Altman limits of agreement. CONCLUSION: An ML algorithm using LDL-KNN is promising for the estimation of LDL-C as it agrees better with LDL-D than closed form equations, especially in mild and severe hypertriglyceridemia.

Comparative assessment of LDL-C and VLDL-C estimation in familial combined hyperlipidemia using Sampson's, Martin's and Friedewald's equations.

BACKGROUND: Sampson et al. developed a novel method to estimate very low-density lipoprotein cholesterol (VLDL-C) and low-density lipoprotein cholesterol (LDL-C) in the setting of hypertriglyceridemia. Familial Combined Hyperlipidemia (FCHL) is a common primary dyslipidemia in which lipoprotein composition interferes with LDL-C estimation. This study aimed to evaluate performance of LDL-C using this new method (LDL-S) compared with LDL-C estimated by Friedewald's and Martin eq. (LDL-F, LDL-M) in FCHL. METHODS: Data were collected from 340 subjects with confirmed FCHL. Concordance for VLDL-C measured by ultracentrifugation and LDL-C estimated using these measures compared to Sampson's, Martin's and Friedewald's equations was performed using correlation coefficients, root mean squared error (RMSE) and bias. Also, concordance of misclassified metrics according to LDL-C (< 70 and < 100 mg/dL) and Apo B (< 80 and < 65 mg/dL) thresholds were assessed. RESULTS: Sampson's equation was more accurate (RMSE 11.21 mg/dL; R2 = 0.88) compared to Martin's (RMSE 13.15 mg/dL; R2 = 0.875) and the Friedewald's equation (RMSE 13.7 mg/dL; R2 = 0.869). When assessing performance according to LDL-C, Sampson's had highest correlation and lowest RMSE compared to other equations (RMSE 19.99 mg/dL; R2 = 0.840). Comparing performance strength across triglyceride levels, Sampson's showed consistently improved correlations compared to Martin's and Friedewald's formulas for increasing triglycerides and for the FCHL phenotype of mixed dyslipidemia. Sampson's also had improved concordance with treatment goals. CONCLUSIONS: In FCHL, VLDL-C and LDL-C estimation using Sampson's formula showed higher concordance with lipid targets assessed using VLDL-C obtained by ultracentrifugation compared with Friedewald's and Martin's equations. Implementation of Sampson's formula could improve treatment monitoring in FCHL.

Plasma Low-Density Lipoprotein Cholesterol Estimated by Friedewald Compared to Martin-Hopkins Equation in Nigerian Population.

BACKGROUND: Friedewald equation for estimation of plasma low-density lipoprotein cholesterol (LDL-C) has recently been the subject of controversies. We investigated the agreement between LDL-C calculated with the Friedewald equation (LDLC F) and novel Martin-Hopkins formula (LDL-CMH), and the influence of sex, age, and triglyceride stratification on the level of biases. METHODS: We used convenience sample of data from records of 7151 adults who underwent test for plasma lipid profile from 2014 to 2017 at a tertiary Hospital in Nigeria. During the period automated standard enzymatic methods were used for determination of plasma lipids. The Bland-Altman plot was used to evaluate the agreement between the two equations. RESULTS: Participants were 2953 males and 4198 females. The age of the subjects ranged from 21 to 91 years with overall mean age of 54.2±12.1 years. The discrepancy between LDLC MH and LDL-CF ranged from -0.05 to 0.93 mmol/L (median = 0.16) with a mean value of 0.172 ±0.094 mmol/L. The BlandAltman analysis showed an estimated bias of 6.38% (95% CI = -5.02, 20.0). The bias in males and females was 8.3% (95% CI = -5.6, 22.2) and 6.9% (95% CI = -4.4, 18.3), respectively. At an average LDL-C less than 1.81 mmol/L, estimated bias became increased to 16.6% (95% CI = -6.1, 39.2). The calculated LDL-C MH were significantly higher than LDL-CF irrespective of the level of triglyceride. CONCLUSION: Although both showed excellent reliability, the Friedewald equation resulted in a clinically lower LDL-C than the Martin-Hopkins formula. It may be necessary to pay attention to biological sex differences.

CONTEXTE: L'équation de Friedewald pour l'estimation du cholestérol plasmatique des lipoprotéines de basse densité (LDL-C) a récemment fait l'objet de controverses. Nous avons étudié l'accord entre le LDL-C calculé avec l'équation de Friedewald (LDL-CF) et la nouvelle formule de Martin-Hopkins (LDL-CMH), et l'influence du sexe, de l'âge et de la stratification des triglycérides. MÉTHODES: Nous avons utilisé un échantillon de commodité de données provenant d'enregistrements de 8926 adultes qui ont subi un test de profil lipidique plasmatique de 2014 à 2017 dans un hôpital tertiaire au Nigeria. Au cours de la période, des méthodes enzymatiques standard automatisées ont été utilisées pour la détermination des lipides plasmatiques. Le graphique de Bland-Altman a été utilisé pour évaluer la concordance entre les deux équations. RÉSULTATS: Les participants étaient 2953 hommes et 4198 femmes. L'âge des sujets variait de 21 à 91 ans avec un âge moyen global de 54,2 ± 12,1 ans. L'écart entre le LDL-CMH et le LDL-CF variait de -0,05 à 0,93 mmol / L (médiane = 0,16) avec une valeur moyenne de 0,172 ± 0,094 mmol / L. L'analyse de Bland-Altman a montré un biais estimé de 6,38% (IC à 95% = - 5,02, 20,0). Le biais chez les hommes et les femmes était de 8,3% (IC à 95% = -5,6, 22,2) et 6,9% (IC à 95% = -4,4, 18,3), respectivement. À un LDL-C moyen inférieur à 1,81 mmol / L, le biais estimé est passé à 16,6% (IC à 95% = -6,1, 39,2). Le LDLCMH calculé était significativement plus élevé que le LDL-CF quel que soit le niveau de triglycéride. CONCLUSION: Bien que les deux aient montré une excellente fiabilité, l'équation de Friedewald a abouti à un LDL-C inférieur cliniquement inacceptable que la formule de Martin-Hopkins. Il peut être nécessaire de prêter attention aux différences biologiques entre les sexes. MOTS CLÉS: Formule de Martin-Hopkins, lipoprotéine de basse densité, équation de Friedewald, Bland-Altman.