Evaluation of current indirect methods for measuring LDL-cholesterol

Accurately quantifying low-density lipoprotein cholesterol (LDL-C) is crucial for precise cardiovascular disease risk assessment and treatment decisions. The commonly used Friedewald equation (LDL-C_FW) has faced criticism for its tendency to underestimate LDL-C, particularly at high triglycerides (TG) or low LDL-C, potentially leading to undertreatment. Newer equations, such as those by Martin and Hopkins (LDL-C_MH) or Sampson (LDL-C_SN), have been proposed as alternatives. Our study aimed to assess the validity of LDL-C_FW, LDL-C_MH, and LDL-C_SN compared to ß-quantification (LDL-C_UC), the reference method recommended by the Lipid Research Clinics.

Using data from three studies comprising 5,738 datasets, LDL-C was determined with the four methods in samples with TG up to 5.65 mmol/L. We calculated median and mean differences, correlations, and used the Passing and Bablok regression for comparisons. Concordance/discordance analyses were conducted.

All equations provided generally accurate LDL-C estimations with slight differences among them. At TG<1.69 mmol/L, no clinically significant divergences were observed. As TG values increased, LDL-C_FW offered the most accurate estimation, followed by LDL-C_SN, while LDL-C_MH exhibited increasingly strong positive bias. LDL-C_FW was not inferior to LDL-C_SN and LDL-C_MH in terms of concordance/discordance.

LDL-C_FW generally provided reliable estimates of LDL-C in most samples, showing non-inferiority to LDL-C_SN or LDL-C_MH, thereby confirming its legitimacy for routine use. Since current treatment recommendations are based on studies employing LDL-C_FW, its replacement by alternatives is not justified.