Novel Equation Gives More Precise LDL-C, Even Without Fasting

Marlene Busko

January 10, 2018

BALTIMORE, MD — A novel equation based on the traditional Friedewald method provides a truer low-density lipoprotein (LDL) cholesterol (LDL-C) estimate than the original and is just as accurate in nonfasting as in fasting blood samples, researchers report.[1]

Instead of using a fixed 5:1 ratio of triglycerides to very-low-density lipoprotein (VLDL) cholesterol levels, the new equation uses a variable ratio from around 3:1 to 10:1 based on a patient's measured triglycerides and total cholesterol minus high-density lipoprotein (HDL) cholesterol (non-HDL cholesterol).

Laboratories could easily incorporate this update into their lipid profile reporting, say the study authors.

Importantly, the improved precision of the estimates using the new equation was most striking in patients with very low LDL cholesterol or high triglycerides—a growing part of modern clinical practice—and it was just as good in more convenient, nonfasting blood samples, another growing trend.  

"These results may have immediate relevance for guideline committees, laboratory leadership, clinicians, and patients," Dr Vasanth Sathiyakumar (Ciccarone Center for the Prevention of Cardiovascular Disease, Johns Hopkins University, Baltimore, MD) and colleagues report in their study published January 2 in Circulation.

Sathiyakumar told theheart.org | Medscape Cardiology that their analysis showed that "the Friedewald equation is most inaccurate in nonfasting samples when it matters the most—at low LDL-C levels and/or high triglyceride levels." Accurate estimation of low levels of LDL cholesterol is of utmost importance in high-risk patients taking novel therapeutic agents, such as PCSK9 inhibitors, for example, and more and more Americans have elevated triglycerides with the current diabetes and obesity epidemic, he noted.

"We take a more personalized approach [to estimate blood levels of LDL cholesterol], and the result is you get a higher precision, more accurate result as compared with the Friedewald method," senior author, Dr Seth S Martin (Ciccarone Center for the Prevention of Cardiovascular Disease) told theheart.org | Medscape Cardiology.

"Clinicians should be aware that there are inaccuracies" from the Friedewald equation for estimated LDL cholesterol, "and those inaccuracies can be addressed in a better calculation such as the one provided by this novel method," Dr Samia Mora (Brigham and Women's Hospital and Harvard Medical School, Boston, MA) told theheart.org | Medscape Cardiology. Mora is coauthor of an accompanying editorial[2] along with Dr Zareen Farukhi (Brigham and Women's Hospital and Harvard Medical School).

"Our novel equation is already being adopted at-scale by major laboratories in the United States, including Quest Diagnostics," said Sathiyakumar.

It can be easily coded and implemented into any electronic medical record or laboratory reporting system. "In the meantime, interested clinicians may use our smartphone application"[3] to calculate LDL cholesterol in mg/dL or mmol/L, he said.

Improving on the Friedewald Equation

"Lipid testing is one of the major tools available to assess a patient's cardiovascular risk, and nearly 200 million Americans have lipid profiles assessed on a yearly basis, with the Friedewald equation remaining the de facto standard in LDL-C estimation," Sathiyakumar noted.

It is relatively easy and inexpensive to measure total cholesterol, HDL cholesterol, and triglycerides, but not LDL cholesterol or VLDL cholesterol.

In 1972,[4] on the basis of a sample of 448 patients, Friedewald proposed and validated the equation to estimate LDL cholesterol that is used today:  

LDL cholesterol (mg/dL) = total cholesterol – HDL cholesterol – triglycerides/5, where triglycerides/5 represents triglycerides/VLDL cholesterol OR

LDL cholesterol (mmol/L) = total cholesterol – HDL cholesterol –triglycerides/2.2.

This equation works very well in patients such as those in the small cohort in which it was derived, with triglycerides less than 400 mg/dL and LDL cholesterol generally greater than 100 mg/dL, but "it starts to break down" in other patients, Martin noted. 

To improve the accuracy, in 2013,[5] they developed and validated a modified equation to estimate LDL cholesterol:

LDL cholesterol (mg/dL) = total cholesterol – HDL cholesterol – triglycerides/adjustable factor.

The adjustable factor is based on a patient's non-HDL cholesterol level (for six levels, from <100 mg/dL to >220 mg/dL) and triglyceride level (for 30 ranges, from 7–49 mg/dL to 400–13975 mg/dL), in a 180-cell table.

Partial Table of Factors to Calculate Novel Estimated LDL Cholesterola

Triglycerides (mg/dL)

Non-HDL Cholesterol

<100 mg/dL

130–159 mg/dL

>220 mg/dL

7–49

3.5

3.3

3.1

101–105

5.5

5.0

4.5

147–154

6.5

5.7

4.8

202–220

7.6

6.4

5.3

293–399

9.3

7.5

5.9

aNovel estimated LDL cholesterol (mg/dL) = total cholesterol – HDL cholesterol – triglycerides/ factor, where factor = value from table cells.

Nonfasting Blood Samples

The current study compared the accuracy of estimated LDL cholesterol in fasting and nonfasting samples by using the novel equation or the Friedewald equation vs direct measurements, in a new cohort of patients.

The researchers identified 1,545,634 US patients who were part of the Very Large Database of Lipids study, had direct measurements of lipids (using Vertical Auto Profile rapid ultracentrifugation), and had triglyceride levels less than 400 mg/dL.

The patients, mainly adults, had been referred for a lipid profile. About half were women, and they had a mean age of 55 years. More than a third of the patients (38%) had provided nonfasting blood samples, and the rest had provided fasting blood samples. The median ratio of triglycerides to VLDL cholesterol was 4.9 in the fasting group and 5.3 in the nonfasting group.

In both fasting and nonfasting blood samples, the estimated accuracy for LDL cholesterol was higher when it was determined by using the novel equation as opposed to the Friedewald equation, for all LDL cholesterol levels (87% to 94% vs 71% to 93%; P≤0.001).

The accuracy decreased as LDL cholesterol decreased, for both methods.

With the Friedewald estimation of LDL cholesterol, there were more errors of at least 10 mg/dL, and greater misclassification of standard LDL cholesterol clinical cutpoints, in nonfasting vs fasting samples—especially in patients with low LDL cholesterol and high triglycerides.

Specifically, in patients with LDL cholesterol levels less than 70 mg/dL, 19% and 30% of fasting and nonfasting patients, respectively, had at least a 10-mg/dL difference between LDL cholesterol estimated with the Friedewald equation vs a direct measure.

In contrast, only 2% to 3% of fasting and nonfasting patients, respectively, had a similar degree of error with LDL cholesterol estimated with the novel formula.

Similarly, at triglyceride levels of 200 to 399 mg/dL, 73% and 81% of fasting and nonfasting patients, respectively, had at least a 10-mg/dL difference in LDL cholesterol estimated by the Friedewald equation vs a direct measure.  

However, only 25% and 20% of fasting and nonfasting patients, respectively, had this magnitude of error when LDL cholesterol was estimated by the novel equation.  

"There are instances in which fasting lipid profiles may be useful," Sathiyakumar acknowledged, "such as diagnosing hypertriglyceridemic disorders." However, in most cases, nonfasting samples may be used to evaluate initial or on-treatment cardiovascular risk, provided an accurate means of estimating LDL cholesterol is used.  

The novel equation is already in use in South America and Asia, where it has been validated, he noted. The group previously showed that sex and age have little impact on the ratio of triglycerides to VLDL, and this is also likely true for ethnicity.  

The Very Large Database of Lipids study was funded by the David and June Trone Family Foundation. Martin receives research support from the PJ Schafer Cardiovascular Research Fund, the American Heart Association, Aetna Foundation, CASCADE FH, Google, and Apple. Martin and Jones are coinventors on a pending patent filed by Johns Hopkins University for novel LDL cholesterol estimation, and Martin is a paid consultant to Quest Diagnostics. The disclosures of the other authors are listed with the article. Farukhi has no relevant financial disclosures. Mora has received institutional research grant support from Atherotech Diagnostics; served as consultant to Amgen, Lilly, Pfizer, and Quest Diagnostics; and is a coinventor on a patent for a biomarker based prediction of colorectal cancer incidence and mortality.

For more from theheart.org | Medscape Cardiology, follow us on Twitter and Facebook 

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.

processing....