Searchable abstracts of presentations at key conferences in endocrinology
Endocrine Abstracts (2015) 37 EP545 | DOI: 10.1530/endoabs.37.EP545

ECE2015 Eposter Presentations Obesity and cardiovascular endocrinology (108 abstracts)

Calculating serum LDL cholesterol (LDL-C): comparison of LDL-C measured by direct assay with various formulae by combination of ages, genders, and triglycerides

Joana Menezes Nunes 1, , Elisabete Rodrigues 1, , Davide Carvalho 1, & João Tiago Guimarães 2,


1Endocrinology, Diabetes and Metabolism Department, Centro Hospitalar São João, Porto, Portugal; 2Faculty of Medicine, Porto University, Porto, Portugal; 3Pathology Department, Centro Hospitalar São João, Porto, Portugal.


Introduction: LDL cholesterol (LDL-C) is a major risk factor for atherosclerosis. The Friedewald formula (FF) is limited by hypertriglyceridaemia. We aimed to correlate serum LDL-C measured by direct assay with serum LDL-C estimated by several formulas: Friedewald (FF): LDL-C=CT-HDL−TG/5; Tsai (TsaiF): LDL-C=TC-HDL−TG/8 (TG/8 represents very-LDL-C); DeLong (DeLongF): LDL-C=TC-HDL−TG×0.16 and Chen (ChenF): LDL-C=non-HDL-cholesterol×0.9−TG×0.1; and analyse LDL-C estimated values by combination of age, gender and triglycerides (TG).

Methods: Statistical analysis of 52 925 lipid profiles performed in our laboratory (SPSS.v21; Mac). Data are expressed in means±S.D.s. Student’s t-test, Pearson’s correlation and linear regression were used for a statistical significance <0.01.

Results: For TG ≤4.5 mmol/l (n=51 975), LDL-C measured by direct assay was 3.12±0.96 mmol/l (120.8±37.0 mg/dl), by FF 3.07±1.11 mmol/l (118.6±42.8 mg/dl), ChenF 3.26±1.14 mmol/l (126.2±44.2 mg/dl), TsaiF 3.32±1.13 mmol/l (128.3±43.6 mg/dl) and DeLongF 3.20±1.11 mmol/l (123.8±43.1 mg/dl). Linear regression showed a significant (P<0.001) better correlation (ρ=0.947, r2=0.897) between the measured and calculated LDL-C by the ChenF. Regarding age, ChenF revealed more accurate results. For TG >4.5 mmol/l (n=950), LDL-C measured by direct assay was 3.88±1.32 mmol/l (150.1±51.1 mg/dl), by FF 2.39±2.50 mmol/l (92.4±96.8 mg/dl), ChenF 4.13±1.93 mmol/l (159.8±74.8 mg/dl), TsaiF 3.62±2.01 mmol/l (140.2±77.8 mg/dl) and DeLongF 3.05±2.20 mmol/l (117.9±85.1 mg/dl). TsaiF showed the best significant correlation (ρ=0.849, r2=0.720). ChenF revealed significant better correlations for children and elderly and TsaiF for adults from 18 to 64 years. Regarding gender, linear regression showed a significant better correlation for ChenF for both TG categories.

Conclusions: FF underestimates LDL-C and the other formulas overestimate it. For patients with TG≤4.5 mmol/l, ChenF provided more accurate for all studied groups. For TG >4.5 mmol/l, TsaiF showed better correlation but regarding age ChenF revealed to be better for children and elderly and TsaiF for middle age patients. Our results suggest that TsaiF and ChenF may be more suitable for LDL-C estimation, even in the presence of moderate to severe hypertriglyceridaemia.

Article tools

My recent searches

No recent searches.