Letter by Lund et al Regarding Article, “Fasting Compared With Nonfasting Lipids and Apolipoproteins for Predicting Incident Cardiovascular Events”
To the Editor:
We are delighted about the report from Mora et al showing a lack of predictive value for future cardiovascular disease (CVD) of nonfasting low-density lipoprotein cholesterol (LDL-C) contrasting with the well-known predictive value of fasting LDL-C.1
In our opinion, the applied methodology has some pitfalls:
Measurement by a direct assay (Genzyme or Roche) in frozen samples might not produce reliable LDL-C estimates compared with fresh samples.
The evidence for using direct LDL-C assays in nonfasting samples (fresh or frozen) with respect to the agreement with the method recommended by the Centers for Disease Control and Prevention for measurement of LDL-C (ie, the β quantification) is sparse.
In daily clinical practice, when LDL-C is “measured” in nonfasting samples, in our experience this is rarely made by direct assays. Instead, probably because of more favorable costs, estimation of LDL-C in nonfasting samples is usually made using the Friedewald equation (LDL-C=total cholesterol−HDL cholesterol− triglycerides/2.2).2 This seems to be a widespread practice, despite the (well-known) limited validity of the Friedewald equation in nonfasting samples (ie, it cannot take into account the magnitude of postprandial change in triglycerides).2 However, despite such technical controversies, the Friedewald formula might still produce LDL-C estimates that predict future CVD in fasting as well as nonfasting samples. Thus, because of the widespread use of the formula, the degree to which calculated nonfasting LDL-C predicts future CVD remains an important clinical question. Therefore, to improve the clinical utility and, thereby, potentially the future guidelines for measurement of LDL-C, we encourage the authors to report from their data the predictive value for future CVD of calculated LDL-C by the Friedewald formula in fasting and nonfasting samples.
As previously demonstrated and noted by Mora et al, LDL-C is lower in the nonfasting than in the fasting state (also, when measured by β quantification).3,4 Although the mechanism underlying this phenomenon is not fully understood, it most likely involves a postprandial increase in triglycerides that stimulates the cholesteryl-ester transfer protein to promote transfer of cholesterol from LDL particles into very low–density lipoprotein (VLDL) particles. Also, an increase in nonfasting triglycerides has been associated with increased CVD risk (probably because it increases nonfasting VLDL cholesterol).5 If this is correct, the postprandial decrease in LDL-C could be a marker of increased nonfasting (proatherogenic) VLDL cholesterol. Such an effect would weaken any positive association between nonfasting LDL-C and future CVD and could potentially explain the lack of association with future CVD of nonfasting LDL-C observed by Mora et al. Moreover, our group recently used β quantification to demonstrate that in patients with type 2 diabetes mellitus, postprandial LDL-C decreases to a greater extent in women than in men.4 Hence, if postprandial LDL-C decrease is a marker of increased CVD risk as outlined, there could be important gender differences that might contribute to the effect observed by Mora et al investigating only women. Finally, contrary to the authors’ conclusion, we find their data to support the idea that (even minor) differences between fasting and nonfasting lipid levels have potential clinical relevance.
Sources of Funding
Drs Lund and Jensen have received grant support from the Danish Diabetes Association (Odense, Denmark) and the Clinical Development Foundation at Steno Diabetes Center (Gentofte, Denmark).
Drs Lund and Jensen have received equipment for research from Roche Diagnostics (Mannheim, Germany). Dr Vaag reports no conflicts.
Mora S, Rifai N, Buring JE, Ridker PM. Fasting compared with nonfasting lipids and apolipoproteins for predicting incident cardiovascular events. Circulation. 2008; 118: 993–1001.
Friedewald WT, Levy RI, Frederickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifugation. Clin Chem. 1972; 18: 499–502.
Cohn JS, McNamara JR, Schaefer EJ. Lipoprotein cholesterol concentrations in the plasma of human subjects as measured in the fed and fasted states. Clin Chem. 1988; 34: 2456–2459.