New Calculation Method Better at Estimating LDL-C in Patients Who Don't Fast
Findings from a cross-sectional study of over 1.5 million U.S. patients indicate that a newer method of calculating LDL-C could potentially eliminate the need for routine fasting in most patients before cholesterol testing, making screening more convenient. The full study has been published in Circulation.
The new method, developed by Seth Martin, MD, MHS, and colleagues at Johns Hopkins, has been validated by recent literature and "uses a flexible approach to derive patient-specific triglyceride [TG] to very low-density lipoprotein-cholesterol ratios." While the classic Friedewald method (total cholesterol minus high-density lipoprotein [HDL] minus TG divided by 5) does not account for variations among patients, the newer method (total cholesterol minus HDL minus TG divided by a specific value from a chart developed by Johns Hopkins physicians) includes 180 different factors to better calculate LDL and individualize the patient's assessment.
"Some patients can have sizable changes in triglycerides after eating, and that is what makes the older Friedewald method less accurate for these people because this isn't taken into account and it exaggerates the underestimation problem of LDL levels," said Vasanth Sathiyakumar, M.D., a senior resident at Johns Hopkins.
The study involved a sample of fasting (≥10–12 hours; n=959,153) and non-fasting (n=586,481) adults from the second harvest of the Very Large Database of Lipids study and compared the accuracy of the novel LDL calculation approach versus the Friedewald method when patients fasted or did not fast. Accuracy was defined as the percentage of directly measured LDL-C content (via rapid ultracentrifugation) falling within an estimated LDL-C category by set clinical cutoffs. For LDL-C <70mg/dL, researchers evaluated the accuracy by TG levels.
The data showed greater accuracy with the novel calculation method for all clinical LDL-C categories (87–94%) vs Friedewald estimation (71–93%) in both fasting and non-fasting samples (P≤0.001). For low estimated LDL-C, accuracy of the non-fasting method was greater than the Friedewald method (92% vs 71%; P<0.001). There were differences of ≥10mg/dL between the Friedewald and direct measurement in 19% of fasting and 30% of non-fasting patients compared to 2% and 3% of patients, respectively, with the novel calculation method.
The Friedewald estimation showed decreased accuracy of LDL-C <70mg/dL as TG increased versus the novel method ( range: 37–96% vs 82–94%). The authors reported superior LDL-C <70mg/dL accuracy with the novel method (82%) versus Friedewald estimation (37%) with TG 200–399mg/dL in non-fasting patients. There were differences of ≥10mg/dL between the Friedewald and direct measurement in 73% of fasting and 81% of non-fasting patients compared to 25% and 20% of patients, respectively, with the novel calculation method.
The authors concluded that the novel adaptable LDL-C estimation performs better in non-fasting samples, particularly in settings of low LDL-C and high TG. However, for some patients (those at high-risk of cardiovascular disease where treatment may change based on a small inaccuracy or those with a triglyceride disorder) fasting may still be important. They added,"...these results may have immediate relevance to guideline committees, laboratory leadership, clinicians and patients."
For more information visit circ.ahajournals.org.