| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Special Articles |
Cardiovascular Research Institute and Department of Medicine University of California–San Francisco San Francisco, California 94143
Address correspondence to: Robert A. Kreisberg, M.D., Dean and Vice President for Health Affairs, University of South Alabama, CSAB 170, Mobile, Alabama 36688-0002.
 |
Introduction |
|---|
 Top Introduction References |
|---|
Evidence from early population studies, animal experiments, and clinical-pathological observations implicated plasma cholesterol as an important and even essential component of atherogenic risk (1). With the knowledge that all plasma cholesterol is carried in plasma lipoproteins came the realization that some species of lipoproteins may be particularly atherogenic and that others may even confer protection against plaque formation. One common monogenic form of hypercholesterolemia that confers greatly increased risk of premature coronary heart disease (heterozygous familial hypercholesterolemia) was found to be associated with striking elevation of LDL, with little or no change in the concentration of VLDL or HDL (3). Separation of cholesterol in VLDL, LDL (or IDL + LDL), and HDL by ultracentrifugation made it practical to classify primary (and presumably genetic) hyperlipoproteinemias by variable elevations of LDL and VLDL and, rarely, of chylomicrons (4). With the addition of lipoprotein electrophoresis, this led to a phenotypic classification of such abnormalities (5). Type IIa was phenotypically (but not genetically) equivalent to familial hypercholesterolemia. With the advent of a simple precipitation method to separate HDL from apo B-containing lipoproteins, the Friedewald formula could be applied to permit phenotypic classification without the need for ultracentrifugation (6). Thus, estimated LDL-cholesterol became a readily accessible analyte for investigators and clinicians. Early evidence had also accrued that at least some species of VLDL are particularly atherogenic, and it was evident that measurement of plasma triglycerides could provide a generally reliable measure of the concentration of triglyceride-rich lipoproteins, mainly VLDL (7). Questions were subsequently raised about the usefulness of plasma triglyceride concentrations as a risk indicator (8) and, thus, inferentially about the contribution of triglyceride-rich lipoproteins. Triglycerides continued to be measured, primarily to permit estimation of LDL-cholesterol in patients who fasted overnight.
It is now increasingly recognized that the dismissal of triglycerides as an independent atherogenic risk factor was incorrect (9). Single measurements of triglycerides may particularly underestimate their association with risk because of the inherently greater variability of triglycerides as compared with LDL-cholesterol and HDL-cholesterol and also because of the strong inverse relationship between the concentration of plasma triglycerides and HDL-cholesterol, which invalidates multivariate analysis involving these two analytes. Prospective studies of lesion progression and clinical outcomes have implicated remnant-like characteristics of triglyceride-rich lipoproteins (such as cholesterol-enrichment) in addition to or, in some cases, instead of LDL as culprit lipoproteins (10, 11). Other evidence suggests that, as with plasma cholesterol, concentrations of triglycerides conferring increased risk may fall well within the range commonly considered as "normal" because they are so prevalent (12). Furthermore, it has been consistently found that multiple lipoprotein abnormalities, particularly those involving all three major lipoprotein classes—moderately elevated LDL-cholesterol and low HDL-cholesterol, accompanied by hypertriglyceridemia (elevated VLDL)—confer a much higher relative risk of atherosclerotic disease than elevated LDL-cholesterol alone (2, 11). These multiple abnormalities are commonly associated with elements of the metabolic syndrome encompassing central obesity, insulin resistance, and hypertension, with or without overt type II diabetes mellitus.
Screening for dyslipoproteinemias as atherogenic risk factors should be simple, precise, and inexpensive. Inclusion of VLDL-cholesterol in screening by measurement of non-HDL-cholesterol fulfills these criteria better than the current NCEP algorithm and gets more directly at the issue of the cholesterol content of triglyceride-rich lipoproteins than measurement of triglycerides per se (13). In the Systolic Hypertension in the Elderly Program, plasma triglyceride concentration was an independent risk factor for coronary heart disease mortality in analyses that included LDL- and HDL-cholesterol, but not in analyses with non-HDL- and HDL-cholesterol (14).
Another approach to routine estimation of apo B-containing
lipoproteins is immunochemical estimation of plasma apo B
concentration. In several studies, apo B concentration has been a
better marker of coronary heart disease than LDL-cholesterol (15). As
might be expected, apo B concentrations are highly correlated with
those of non-HDL-cholesterol. With a chemical approach to apo B
estimation, correlation coefficients exceeded 0.9 in men with normal to
modestly increased total cholesterol and triglyceride concentrations
(16). Although the size and, presumably, the cholesterol content of LDL
falls with increasing plasma triglycerides, the ratio of apo B to total
cholesterol in that study was 0.6, irrespective of plasma triglyceride
concentrations. The ratio of apo B to cholesterol is 
0.6 in LDL and
VLDL from normotriglyceridemic persons, but is usually lower than 0.6
in VLDL of persons with hypertriglyceridemia (17), reflecting
cholesteryl ester-enrichment of hypertriglyceridemic VLDL. Measurement
of non-HDL-cholesterol will include this cholesterol enrichment, but
that of apo B will not. Although apo B can now be measured with
adequate accuracy and precision (15), measurement of total cholesterol
and HDL-cholesterol is already widely available and well
standardized.
Non-HDL-cholesterol may better reflect changes in plasma lipoproteins occurring with lipid-lowering therapy than do changes in LDL-cholesterol alone (13). For example, lipid-lowering drugs (including the statins, fibrates, and nicotinic acid) all lower VLDL, IDL, and LDL concentrations. Furthermore, statins and nicotinic acid may reduce VLDL-cholesterol disproportionately to LDL-cholesterol (i.e. they reduce the cholesterol content of VLDL particles). In the Scandinavian Simvastatin Survival Study, baseline non-HDL-cholesterol predicted cardiovascular events in the placebo group better than baseline LDL-cholesterol, presumably because the former reflected the contribution of triglyceride-rich lipoproteins to events (18). In those treated with simvastatin, percentage changes in non-HDL-cholesterol were equivalent to those in LDL-cholesterol in predicting event reduction, and both were greater than percentage changes in plasma apo B concentration. Available evidence, although sparse, is consistent with the expectation that changes in non-HDL-cholesterol are as good as LDL-cholesterol in predicting clinical benefit of therapeutic interventions directed at plasma lipoproteins.
How should this evidence be used? LDL-cholesterol has become the standard analyte, with HDL in a supporting role for risk assessment, and a nod to plasma triglycerides for those with levels exceeding 200 mg/dL (1). In our experience, the current algorithms are often confusing to practitioners. At least in theory, non-HDL-cholesterol alone (based on total cholesterol and HDL-cholesterol) should suffice, not only for screening, but also for initial risk assessment in primary prevention, obviating the need for a fasting blood specimen.
Measurement of non-HDL-cholesterol clearly becomes more important the
higher the plasma triglyceride concentration. In a group of 548 lipid
clinic patients with plasma triglyceride concentrations below 201 mg/dL
(mean, 123 mg/dL), mean VLDL-cholesterol concentration was 18 mg/dL and
that of LDL-cholesterol was 191 mg/dL (Table 1
). Thus, VLDL-cholesterol constituted
8.6% of non-HDL-cholesterol. However, in those patients with higher
plasma triglyceride concentrations, the percentage of
non-HDL-cholesterol contributed by VLDL-cholesterol increased
rapidly and exceeded 50% at triglyceride concentrations above 800
mg/dL. For individuals with plasma triglycerides below 200 mg/dL,
non-HDL-cholesterol and LDL-cholesterol may be close to equivalent for
risk assessment, and the latter could be retained. This approach,
however, begs the question of why triglycerides should still be
measured and the patient, thus, required to fast.
|
, that increment might be
reduced to 20 mg/dL. This is the average VLDL-cholesterol concentration
equivalent to plasma triglycerides of about 130 mg/dL, the level at
which the "atherogenic lipoprotein phenotype" (characterized by
reduced size of LDL particles and other changes in lipoproteins that
are associated with increased coronary heart disease risk) becomes
increasingly prevalent (Krauss, R. M., personal
communication). In summary, the use of non-HDL-cholesterol in primary prevention recognizes the contribution of triglyceride-rich lipoproteins to atherosclerotic disease and simplifies the physician’s initial assessment of disease risk and the continuing response to therapy. When drugs are indicated, triglycerides should also be measured, however, to help establish a diagnosis and to guide specific therapy. The lipoprotein pattern will, therefore, still need to be assessed in some patients in the primary prevention setting and almost always for secondary prevention. Establishment of the lipoprotein pattern is also needed to evaluate kindred relationships.
|
Footnotes |
|---|
Accepted March 6, 2000.
|
References |
|---|
|
TopIntroductionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. A. Aguilar-Salinas, A. Delgado, F. J. Gomez-Perez, Y. Cui, R. S. Blumenthal, M. K. Whiteman, and J. A. Flaws The Advantages of Using Non-HDL-C in the Diagnosis and Treatment of Dyslipidemia Arch Intern Med, January 14, 2002; 162(1): 108 - 109. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
