Abstract 15633: Inaccuracy of Friedewald-Estimation at Clinically Relevant Very Low Levels of LDL-Cholesterol: The Very Large Database of Lipids Study 1C
Background: Appropriate classification of low-density lipoprotein cholesterol (LDL-C) in the very low range has direct relevance to patient care and clinical trials. Some guidelines recommend targeting an LDL-C <70 mg/dL for high-risk patients or consideration of statin down-titration at levels <40 mg/dL. Moreover, recent lipid-lowering trials have classified many participants with an LDL-C <50 and even <25 mg/dL.
Objectives: Previously, we showed significant underestimation of Friedewald-estimated LDL-C (LDLf-C) compared with directly-measured LDL-C (LDLd-C) at low LDL-C and high triglycerides (TG). We now seek to extend these findings to persons with very low LDLf-C levels.
Methods: We analyzed 191,333 US adults from the Very Large Database of Lipids with LDLf-C <70 mg/dL and TG <400 mg/dL. We calculated the median difference by subtracting LDLd-C from LDLf-C. Upward reclassification by LDLd-C vs LDLf-C was calculated based on several cutpoints: 25, 40, 50, and 70 mg/dL (0.2th, 1.3th, 3.6th, 14.7th percentile, respectively).
Results: Patients were 65±15 years old and 60% men. Greater underestimation was found at lower LDLf-C and higher TG levels. The median difference (25th to 75th percentile) increased from -2 (-7 to 1) at LDLf-C 60-69 mg/dL to -30 (-37 to -22) at LDLf-C <10 mg/dL in the overall population, and from -16 (-22 to -12) to -32 (-39 to -26), respectively, in patients with TG 200-399 mg/dL. Among individuals with LDLf-C lower than 25, 40, 50, and 70 mg/dL cutpoints, 82%, 57%, 43%, and 23% had LDLd-C higher than those levels. Reclassification occurred most commonly in those with high TG (Table).
Conclusion: Extending our prior work on Friedewald LDL-C underestimation at levels <70 mg/dL, we now show even more substantial misclassification at very low LDL-C levels (<25, 40, 50 mg/dL), particularly at high TG. These findings likely have relevance for personalized clinical decision-making and for clinical trial design using more potent lipid lowering agents.
Author Disclosures: R. Quispe: None. A. Hendrani: None. M.B. Elshazly: None. E.D. Michos: None. J.W. McEvoy: None. M.J. Blaha: Consultant/Advisory Board; Modest; Pfizer. P.P. Toth: Speakers Bureau; Modest; for Amarin, AstraZeneca, Genzyme, Kowa, Merck. Consultant/Advisory Board; Modest; Amgen, AstraZeneca, Atherotech, GlaxoSmithKline, Kowa, Liposcience. J. Coresh: None. R.S. Blumenthal: None. S.R. Jones: Consultant/Advisory Board; Modest; Atherotech Diagnostic Lab, Sanofi and Regeneron. S.S. Martin: None.
- © 2015 by American Heart Association, Inc.