Concordance between LDL-C estimated by various formulas and directly measured LDL-C

Area of Science:

• Cardiovascular Diagnostics
• Clinical Chemistry

Background:

• Direct measurement of low-density lipoprotein cholesterol (LDL-C) is possible, but estimation formulas are widely used.
• The performance of these LDL-C estimation formulas has not been previously evaluated in Colombia.

Purpose of the Study:

• To assess the concordance between directly measured LDL-C and calculated LDL-C using various estimation formulas.
• To evaluate the performance of different LDL-C estimation methods in a Colombian population.

Main Methods:

• A diagnostic test study analyzing 2144 complete lipid profiles from 2014-2022.
• Compared direct LDL-C measurements with estimations from DeLong, Sampson, Friedewald, extended Martin/Hopkins, Anandaraja, and Cordova formulas.
• Utilized Lin's concordance correlation coefficient (CCC), Bland-Altman plots, and Kappa coefficients for risk category agreement.

Main Results:

• DeLong (CCC=0.971) and Sampson (CCC=0.969) formulas showed the highest concordance with direct LDL-C measurement.
• Extended Martin/Hopkins and Friedewald formulas also demonstrated good performance (CCC=0.964).
• Anandaraja (CCC=0.921) and Cordova (CCC=0.881) formulas exhibited lower concordance; performance decreased with high triglycerides or low LDL-C levels.

Conclusions:

• DeLong, Sampson, extended Martin/Hopkins, and Friedewald formulas are reliable for LDL-C estimation in Colombia.
• These formulas demonstrate interchangeability with direct LDL-C measurements in most clinical scenarios.
• Anandaraja and Cordova formulas are not recommended due to inferior performance.