ABSTRACT
BACKGROUND: LDL cholesterol (LDL-C) is regarded as a significant therapeutic target and a known risk factor for atherosclerosis. It can be calculated using the results of the other lipid tests or tested directly. Despite its shortcomings, the Friedewald formula is most frequently utilized since it is simple and practical. Until now, several formulae have been proposed for calculating LDL-C; however, their accuracy has not been evaluated across different populations. We sought to evaluate the validity of calculated LDL-C by comparing the findings with values acquired by the direct homogeneous technique, utilizing 13 distinct formulae from the literature. METHODS: This study was a retrospective observational study conducted for a year at SRIHER, Chennai, Tamil Nadu, India. From the total 25 043 patients who had their serum lipid profile tested, 16 314 participants had their fasting blood sugar and fasting lipid profile measured simultaneously, and they were chosen for the research. RESULTS: The de Cordova, Chen, Martin/Hopkins (initial), and Teerakanchana equations correlated well with the direct LDL-C assay. When the dataset was stratified according to triglycerides, the Chen and Martin/Hopkins initial equations had the better measurement of agreement compared to other equations. The Martin/Hopkins initial equation outperformed all the other equations when the whole dataset irrespective of the triglyceride population was considered. CONCLUSIONS: Our study suggests that the Martin/Hopkins initial equation outperformed all the other equations and can be used as an alternative to direct LDL-C measurement in a South Indian population.
