Letter by Ala-Korpela et al Regarding Article, “Lipoprotein Particle Profiles by Nuclear Magnetic Resonance Compared With Standard Lipids and Apolipoproteins in Predicting Incident Cardiovascular Disease in Women”
To the Editor:
There has been increasing interest in the applications of proton (1H) nuclear magnetic resonance (NMR) spectroscopy of serum in clinical and epidemiological research.1 The article by Mora et al2 made an important contribution by demonstrating that the standard lipid or apolipoprotein measurements and the 1H NMR measures on the number and size of lipoprotein particles are comparable relative to cardiovascular disease risk prediction. However, the authors did not discuss that 1H NMR can also quantify serum and lipoprotein lipids, as independently demonstrated by Otvos et al3 and by us.4
The 1H NMR from lipoprotein particles originate from all the lipid constituents of the particles, the signal position being dependent on the size of the entire particle.1 Thus, the true measure recorded by 1H NMR is the total NMR-visible lipid amount that, via various mathematical and spectral analysis methods, can be converted into information on lipoprotein subclasses. Thus, the spectral analysis is the primary issue in defining the capability of 1H NMR-based lipoprotein analytics.1 Mora et al2 adopted a curve-fitting approach3 to estimate the total NMR-visible lipid amounts and then transformed these data to lipoprotein particle concentrations by approximating the lipoprotein particle diameter and core lipid volume and mass.2,3 We have previously used a similar curve-fitting method4 or various regression modeling approaches1 to isolate the different lipoprotein categories. However, our convention has been to transform the mathematically isolated total NMR-visible lipid amounts to lipoprotein fraction–specific triglyceride or cholesterol concentrations by assuming an averaged lipid composition of the lipoprotein particles at a certain NMR chemical shift range. Nevertheless, the key is to recognize that both lipoprotein lipid and particle concentrations are only different approximations based on the actual 1H NMR measure (ie, the total NMR-visible lipid amount).
1H NMR-based lipoprotein lipid quantification has been shown to be analytically good for the main lipoprotein fractions and their major lipid constituents, namely, very-low–density lipoprotein triglycerides, low-density and high-density lipoprotein cholesterol, and serum triglyceride and cholesterol concentrations.1,3,4 These earlier analytical demonstrations that 1H NMR spectroscopy can also be used to quantify standard lipids put the conclusion by Mora et al2 into a new perspective: If the use of standard lipids were recommended, those can indeed be also obtained by a single 1H NMR spectroscopic measurement. Furthermore, we have recently shown that serum triglyceride, cholesterol, and high-density lipoprotein cholesterol enable estimation of apolipoprotein B and A-I concentrations.5 Thus, 1H NMR spectroscopic measurements of standard lipids together with the computational estimation of apolipoprotein B and A-I can thus be an appealing cost-effective alternative for the tedious separate measurements of several lipids and apolipoproteins. In addition, the use of 1H NMR in a metabonomics fashion1 is able to provide quantitative molecular data not only on lipoproteins but on many additional and clinically important metabolites.1 Therefore, it is tempting to envision that 1H NMR of serum might eventually replace the standard lipid measurements with holistic multimetabolic risk phenotyping.
Mora S, Otvos JD, Rifai N, Rosenson RS, Buring JE, Ridker PM. Lipoprotein particle profiles by nuclear magnetic resonance compared with standard lipids and apolipoproteins in predicting incident cardiovascular disease in women. Circulation. 2009; 119: 931–939.
Ala-Korpela M, Korhonen A, Keisala J, Hörkkö S, Korpi P, Ingman LP, Jokisaari J, Savolainen MJ, Kesäniemi YA. 1H NMR-based absolute quantitation of human lipoproteins and their lipid contents directly from plasma. J Lipid Res. 1994; 35: 2292–2304.
Niemi J, Mäkinen VP, Heikkonen J, Tenkanen L, Hiltunen Y, Hannuksela ML, Jauhiainen M, Forsblom C, Taskinen MR, Kesäniemi YA, Savolainen MJ, Kaski K, Groop PH, Kovanen PT, Ala-Korpela M. Estimation of VLDL, IDL, LDL, HDL2, apoA-I and apoB from the Friedewald inputs– apoB and IDL, but not LDL, are associated with mortality in type 1 diabetes. Ann Med. 2009; 41: 451–461.