evaluation of performance of the newly developed de cordova's formula for calculation of low-density lipoprotein cholesterol without use of triglycerides

Background: Various formulas are available to estimate serum low-density lipoprotein (LDL) cholesterol. All of these are serum triglycerides (TG) dependent. But very recently de Cordova et al developed a simple formula (CF) to calculate LDL cholesterol without using serum TG and claimed it to be more accurate than Friedewald.s formula (FF). Objective: The objective of the present study was to evaluate the performance of the CF for the calculation of LDL cholesterol in a Bangladeshi population. Materials and Methods: Three hundred and sixty adult Bangladeshi subjects were purposively included in this study. Serum total cholesterol (TC), TG, high-density lipoprotein (HDL) cholesterol and LDL cholesterol were measured by direct automated methods. LDL cholesterol was also calculated by CF and FF. Results were expressed in conventional unit as mean ± SD and compared by two-tailed paired t test, bias against measured LDL cholesterol, Pearson's correlation coefficient (r), Passing & Bablok regression and accuracy within ±10% of the measured LDL cholesterol. Results: The mean values of directly measured LDL cholesterol, LDL cholesterol calculated by CF and FF were 117.7 ± 31.0, 111.8 ± 31.0 and 108.9 ± 39.7 mg/dL respectively. Bias of calculated LDL cholesterol against measured LDL cholesterol was -5.2% for CF and -9.6% for FF. The correlation coefficients of measured LDL cholesterol were 0.9796 (p<0.001) for CF and 0.9525 (p<0.001) for FF. Passing & Bablok regression yielded the equation y = 0.9938x - 6.2 for CF and y = 1.2774x - 40.9 for FF. Accuracy within ±10% of measured LDL cholesterol was 81% for CF and 49% for FF. Conclusion: This study revealed better performance of the de Cordova's formula than Friedewald's formula for approximate calculation of LDL cholesterol without using serum triglycerides.