How Low Is Safe? The Frontier of Very Low (<30 mg/dL) LDL Cholesterol

Angelos D. Karagiannis; Anurag Mehta; Devinder S. Dhindsa; Salim S. Virani; Carl E. Orringer; Roger S. Blumenthal; Neil J. Stone; Laurence S. Sperling


Eur Heart J. 2021;42(22):2154-2169. 

In This Article

Different Methods of Measuring LDL-C

In the lower ranges of LDL-C, attention to measurement is crucial. There are several available methods for measuring LDL-C (Figure 1).[26,27] In routine clinical practice, LDL-C is not measured directly but is calculated based on the measured total cholesterol, HDL-cholesterol and triglycerides using the Friedewald equation. However, it has been shown that Friedewald-calculated LDL-C can be significantly underestimated in patients with low LDL-C (<70 mg/dL) or with elevated triglycerides.[28] Direct LDL-C measurement with preparative ultracentrifugation (beta quantification) is the most accurate approach in these patients (gold standard), but it is costly, time consuming, and therefore, not often performed.[28] Both the FOURIER and ODYSSEY OUTCOMES trials used beta quantification for patients with very low LDL-C levels to increase accuracy.[6,7]

Figure 1.

Different methods of measuring the LDL-C. LDL-C can be measured using various techniques. As recent cholesterol guidelines recommend achieving lower LDL-C, it is crucial to use methods that measure LDL-C accurately in low LDL-C ranges.

The Martin/Hopkins method is a novel, practical, indirect LDL-C calculation equation that is cost and time effective.[29] A post hoc analysis of the FOURIER trial showed that, when compared to the ultracentrifugation standard, 13.3% of Friedewald calculated values differed by >10 mg/dL, as contrasted with only 2.6% by the Martin/Hopkins method.[30] In a nationally representative sample, a hospital laboratory, and a reference laboratory, approximately one-fifth of individuals with Friedewald-estimated LDL-C < 70 mg/dL have a value of ≥70 mg/dL using the Martin/Hopkins LDL-C equation.[31] These individuals usually have significantly higher non-HDL-C and apoB concentrations, conferring an increased risk for cardiovascular disease.[31] Recently, Sampson et al.[32] have reported a new method for improving LDL-C calculation. LDL-C values derived by Sampson's method compare favourably with results obtained by preparative ultracentrifugation, long considered a 'gold standard' for LDL-C measurement, even in patients with very high triglyceride levels. Overall, Sampson's equation extends the accuracy of LDL-C calculation to individuals with triglyceride values up to 800 mg/dL, whereas Martin's equation significantly improves LDL-C estimation in patients with low LDL-C levels (<70 mg/dL).[33]

Finally, direct beta quantification should not be confused with commercially available homogenous assays that use proprietary chemical-based methods to measure LDL-C directly and are not necessarily reliable or standardized.[34–36]

Understanding the importance of more accurate measurement is essential in any discussion of LDL-C levels <70 mg/dL. High-risk patients with the LDL-C levels of 50–70 mg/dL may be considered well managed based upon the Friedewald-calculated LDL-C; however, they may, in fact, benefit from additional LDL-C lowering if a more accurate method such as the Martin/Hopkins method was used. Given evidence of the Martin/Hopkins method's improved accuracy, the 2018 AHA/ACC/Multi-society Cholesterol Guideline supports its use in persons with low LDL-C levels[23] and it is possible that future guidelines will consider the broader use of this method.

Of note, most LDL-C determining methods, including ultracentrifugation and Friedewald-calculated LDL-C, incorrectly count the cholesterol present in Lp(a) particles as 'LDL-C', thus overestimating LDL-C levels.[37] Lp(a)-corrected LDL-C can be estimated with the Dahlen modification, which assumes that 30% of Lp(a) weight consists of cholesterol: LDL-C − [Lp(a) × 0.30].[38,39]

Finally, plasma LDL-C is a measure of cholesterol mass carried by LDL particles (LDL-p) and therefore only an estimate of circulating LDL-p. In most clinical settings, measured LDL-C mirrors LDL-p values. However, there are conditions (e.g. hypertriglyceridaemia, metabolic syndrome) where patients may have normal LDL-C, while having lower cholesterol mass per LDL particle and larger LDL-p numbers, being thus in higher cardiovascular risk than expected on their LDL-C measurements.[37] Even though LDL-p is the most abundant apoB-containing circulating lipoprotein and the lipoprotein for which we have clear evidence of its atherogenicity, other ApoB lipoproteins (VLDL, ILDL) may also be atherogenic. A Mendelian randomization study showed that TG-lowering lipoprotein lipase variants and LDL-C-lowering LDL receptor variants had the same impact on ASCVD risk per unit change of ApoB, suggesting that all ApoB-containing lipoproteins have the same effect on the risk of cardiac heart disease and that the clinical benefit of LDL-C and triglyceride-lowering treatments may be proportional to the absolute difference in ApoB.[40] In a meta-analysis, the mean risk reduction per SD decrease was higher for apoB [24%; 95% confidence interval (CI), 19–29%] compared to 20% (16–24%) for LDL-C and 20% (15–25%) for non-HDL-C across seven major statin trials.[41] ApoB measurement is an easy, fully automated, standardized, fasting-independent test that can accurately detect the numbers of plasma atherogenic lipoproteins as well as elevated LDL-p often missed with LDL-C alone. Per 2019 ESC/EAS guidelines, it should be considered as a complementary marker to LDL-C for ASCVD risk calculation (especially in people with high triglyceride levels, diabetes mellitus, obesity, metabolic syndrome, or very low LDL-C levels) or as an alternative to LDL-C, as the primary measurement for screening, diagnosis, and management.[21]