当前位置: 首页 > 期刊 > 《循环学杂志》 > 2005年第6期 > 正文
编号:11176260
Nuclear Magnetic Resonance Lipoprotein Abnormalities in Prediabetic Subjects in the Insulin Resistance Atherosclerosis Study
http://www.100md.com 《循环学杂志》
     the Department of Medicine, University of Texas Health Science Center, San Antonio (A.F., K.W., S.M.H.)

    Leadership Sinai Centre for Diabetes, Mount Sinai Hospital, University of Toronto, Toronto, Ontario, Canada (A.J.G.H.)

    LipoScience, Inc, Raleigh, NC (J.D.O.)

    Department of Public Health Sciences, Wake Forest University School of Medicine, Winston-Salem, NC (D.C.G., L.E.W.)

    Eli Lilly & Company, Area Medical Center, Vienna, Austria (A.F.).

    Abstract

    Background— Subjects with type 2 diabetes have smaller LDL and HDL particles in addition to higher levels of triglycerides and lower HDL cholesterol. Elevated insulin resistance, blood pressure, and dyslipidemia (including small dense LDL) predicted incident diabetes. In the Insulin Resistance Atherosclerosis Study (IRAS) we studied nuclear magnetic resonance (NMR) lipoprotein particle measures in prediabetic individuals, considering potentially modifying covariates, including insulin resistance, as directly measured using a frequently sampled intravenous glucose tolerance test.

    Methods and Results— Of 830 subjects who were nondiabetic at baseline, 130 (15.7%) developed diabetes after a mean follow-up of 5.2 years. Various lipoprotein abnormalities were found in prediabetic subjects compared with subjects who stayed nondiabetic at follow-up. In logistic regression analyses (demographically adjusted), VLDL particles, large VLDL, LDL particles, small LDL, large HDL, small HDL, VLDL size, LDL size, and HDL size were related to incident diabetes. The relation of VLDL size and small HDL to incident diabetes was independent of waist (odds ratio [OR] [95% CI], 1.43 [1.18 to 1.73] and 1.23 [1.01 to 1.51] for VLDL size and small HDL, respectively) and independent of conventionally (chemically) measured triglycerides and HDL cholesterol (OR [95% CI], 1.45 [1.18 to 1.78] and 1.30 [1.06 to 1.60], respectively). Insulin sensitivity attenuated the relation to incident diabetes of VLDL size (OR [95% CI], 1.25 [1.01 to 1.53]) but not of small HDL particles (OR [95% CI], 1.25 [1.02 to 1.54]).

    Conclusions— We have shown a range of lipoprotein abnormalities in prediabetic individuals, including compositional changes in HDL and VLDL. These findings extend previous work indicating a proatherogenic state in healthy, nondiabetic subjects who subsequently develop diabetes.

    Key Words: diabetes mellitus ; epidemiology ; lipids ; lipoproteins

    Introduction

    Patients with type 2 diabetes typically present with abnormalities not only in lipid concentrations but also in lipoprotein size and subclass composition. Dyslipidemia in individuals with diabetes or insulin resistance is characterized by high levels of triglycerides, low levels of HDL cholesterol, and a preponderance of small dense LDL particles, as well as a relative increase of large VLDL particles and small HDL particles.1–3 Previous work has suggested that elevated insulin resistance, blood pressure, and triglycerides and lower HDL cholesterol precede the onset of type 2 diabetes.4–7 In one report, small dense LDL predicted the incidence of type 2 diabetes.8 However, no report has examined heterogeneity within lipoproteins in a population in which insulin resistance was directly measured.

    Recently, a new procedure for quantifying plasma levels of lipoprotein subclass particles by proton nuclear magnetic resonance (NMR) spectroscopy has been developed.9 NMR spectroscopy provides concentrations of VLDL, LDL, and HDL subspecies simultaneously, and NMR-determined subclasses correspond well with those obtained with established methods.9

    In the Insulin Resistance Atherosclerosis Study (IRAS) we studied NMR lipoprotein measures in prediabetic individuals, considering potentially modifying covariates, including insulin resistance, as directly measured using a frequently sampled intravenous glucose tolerance test. Our goal was to assess whether atherogenic lipoprotein composition and particle size differences precede the onset of actual diabetes, as do other cardiovascular disease risk factor differences.10 To accomplish this goal, we examined the lipoprotein characteristics at baseline according to whether the subjects developed type 2 diabetes at follow-up. In addition, we assessed the ability of baseline lipoproteins and their composition to predict the incidence of type 2 diabetes. These different approaches yielded similar results in this report.

    Methods

    The IRAS is a multicenter, epidemiological study that seeks to explore relationships between insulin resistance, cardiovascular risk factors, and disease across different ethnic groups and varying states of glucose tolerance. A full description of the design and methods of the IRAS has been published.11 The IRAS protocol was approved by local institutional review committees, and all subjects gave informed consent. A total of 1624 individuals participated in the IRAS. This report includes baseline data on 513 diabetic patients, as well as baseline and follow-up data on 830 nondiabetic patients at baseline, in whom plasma levels of lipoprotein subclasses were assessed by proton NMR spectroscopy. Subjects were investigated twice following the same protocol. The mean follow-up duration was 5.2 years (range, 4.5 to 6.6 years). Each of the 2 IRAS examinations required 2 visits. Patients were asked before each visit to fast for 12 hours, to abstain from heavy exercise and alcohol for 24 hours, and to refrain from smoking the morning of the examination. Family history of type 2 diabetes and physical activity were assessed with the use of standard interviewing procedures. A positive family history of diabetes was defined as parents and/or siblings having type 2 diabetes. Race and ethnicity were assessed by self-report. Blood pressure was measured by standard methods, and hypertension was defined as systolic blood pressure 140 mm Hg and/or diastolic blood pressure 90 mm Hg or current use of antihypertensive medication.

    Assessment of Glucose Tolerance and Insulin Sensitivity

    A standard 75-g oral glucose tolerance test was performed, and a frequently sampled intravenous glucose tolerance test12 with minimal model analysis13 was performed to assess insulin sensitivity. Two modifications of the original protocol were used. An injection of regular insulin, rather than tolbutamide, was used to ensure adequate plasma insulin levels for the accurate computation of insulin sensitivity across a broad range of glucose tolerance. In addition, the reduced sampling protocol (which required 12 rather than 30 plasma samples) was used because of the large number of subjects. Insulin sensitivity, expressed as the insulin sensitivity index (SI), was calculated by mathematical modeling methods (MINMOD, version 3.0 [1994]). Height, weight, girth measurements, and laboratory measurements, including chemical analyses of lipids and lipoproteins, were performed by standard methods, as described previously.14

    NMR Spectroscopy

    Lipoprotein subclass particle concentrations and average VLDL, LDL, and HDL particle diameters were measured by NMR spectroscopy at LipoScience, Inc (Raleigh, NC) with a modification of the previously described method.9,15 In brief, the particle concentrations of the different-size lipoprotein subclasses are given by the measured amplitudes of the characteristic lipid methyl group NMR signals they emit. The subclass signal amplitudes were extracted from the composite lipid methyl group signal envelope of each plasma sample with the use of an improved spectral deconvolution algorithm that includes an expanded number (>30) of lipoprotein subpopulations to better represent the continuum of particle subspecies actually present in plasma. Because the NMR signal of each subpopulation differs only slightly in frequency and line shape from the signals of neighboring subpopulations, measurement reproducibility of the individual signal amplitudes is inherently limited. To overcome this limitation, neighboring subpopulations were grouped empirically into a smaller number of subclass categories (large, medium, and small) so that the summed amplitudes of the individual subpopulation signals gave acceptable measurement precision (coefficient of variation <10%). The following 9 subclass categories were investigated: large VLDL (including chylomicrons, if present) (>60 nm), medium VLDL (35 to 60 nm), small VLDL (27 to 35 nm), IDL (23 to 27 nm), large LDL (21.2 to 23 nm), small LDL (18 to 21.2 nm), large HDL (8.8 to 13 nm), medium HDL (8.2 to 8.8 nm), and small HDL (7.3 to 8.2 nm). The small LDL subclass comprises the sum of medium small particles (19.8 to 21.2 nm, previously named intermediate LDL) and very small particles (18 to 19.8 nm), which were found to have virtually identical correlations with lipid levels and insulin sensitivity. VLDL and LDL subclass particle concentrations are given in units of nanomoles per liter and those of HDL subclasses in micromoles per liter. Further summation of the subclass levels also provides total VLDL, LDL (including IDL), and HDL particle concentrations. Weighted-average VLDL, LDL, and HDL particle sizes (in nanometer diameter units) were computed as the sum of the diameter of each subclass multiplied by its relative mass percentage as estimated from the amplitude of its methyl NMR signal. LDL and HDL subclass distributions determined by NMR and gradient gel electrophoresis are highly correlated.16,17 LDL subclass diameters, which are consistent with electron microscopy data,18 are uniformly 5 nm smaller than those estimated by gradient gel electrophoresis. Reproducibility of the NMR-measured lipoprotein particle parameters was determined by replicate analyses of plasma pools. Coefficients of variation <4% were observed for total VLDL, LDL, and HDL particle concentrations; <2% for VLDL size; <0.5% for LDL and HDL size; <10% for large, medium, and small VLDL subclasses; <8% for large and small LDL subclasses; and <5% for large and small HDL subclasses. Higher coefficients of variation for IDL (<20%) and medium HDL (<35%) subclasses reflect their typically low concentrations.

    Statistical Analyses

    To analyze whether changes in lipoprotein composition occur before the onset of diabetes, we mainly used 2 different statistical methods: First, we compared NMR measures in prediabetic individuals with those individuals who remained nondiabetic at follow-up using ANCOVA; second, we analyzed baseline NMR measures in relation to incident diabetes (dependent variable) using logistic regression analyses. Separate ANCOVAs were conducted with the use of the actual measures or log-transformed data and with rank transformation of all measures into percentiles. A demographically adjusted logistic regression was performed for each NMR measure. Stepwise logistic regression analyses were used to assess the relation of NMR measures to incident type 2 diabetes and to identify which measures were most significantly associated with incident diabetes in the IRAS. Additional models were fit for VLDL size and small HDL with consideration of further covariates, including baseline glucose tolerance status (impaired glucose tolerance [IGT]), SI, waist, triglycerides, HDL cholesterol, VLDL size, small HDL, and all of these covariates in a fully adjusted model. To assess whether NMR measures were related to diabetes incidence independent of conventional lipid measures, the stepwise method with P<0.05 to enter and stay in the model was used to select variables after demographic variables and established diabetes risk factors were forced in. Conventional lipid measures were allowed to enter initial "conventional" models. Variables thus selected were forced into "+NMR" models, which allowed significant NMR measures to enter as predictors. Finally, diabetes incidence was assessed across quartiles of VLDL size and small HDL.

    All statistical calculations were performed with SAS version 8.0. Log-transformed values were used in all analyses for all continuous variables, which appeared to be more normally distributed with the transformation than without. Although an unadjusted probability value <0.05 was considered statistically significant, it may be appropriate to view marginally significant (0.005
    Results

    Descriptive Data at Baseline

    Mean levels of total VLDL particles were higher in DM, as were levels of the large VLDL subclass (Table 2). Mean total LDL particles were higher in DM, primarily because of higher levels of small LDL (both medium small and very small LDL particles). Mean levels of total HDL particles were not different between DM and non-DM, but a shift toward small HDL was found in DM (lower levels of large HDL and higher levels of small HDL particles). Lipoprotein particle size was markedly different between DM and non-DM; VLDL size was higher in DM, and both LDL size and HDL size were lower in DM. No differences were seen between DM and non-DM in chemically measured LDL cholesterol and total cholesterol levels. The significance of each rank-transformed analysis (data not shown) was concordant with the significance indicated in Table 2 except for 2 marginally significant (both P=0.044) comparisons: medium VLDL particles (P=0.076) and large LDL particles (P=0.11).

    Overall, the lipid pattern found in DM resembled the pattern found in patients with overt diabetes at baseline (DM-BL).

    Adjustments for gender and/or ethnicity did not materially affect the results as presented.

    Correlations Between NMR Lipoprotein Measurements and Metabolic Variables

    Analyses of Covariance

    Adjustment for triglycerides and HDL cholesterol abolished the differences between DM and non-DM in VLDL particles, large VLDL, small LDL, large HDL, and LDL size. In another metabolic model (Table 5), small HDL and VLDL size remained significantly different between DM and non-DM.

    Logistic Regression Analyses With Incident Diabetes as the Dependent Variable

    The following measures with their odds ratio (OR) for a 1-SD increment and 95% CI were significantly associated with incident diabetes in demographically adjusted logistic regression models: VLDL particles, 1.23 (1.02 to 1.49); large VLDL, 1.46 (1.22 to 1.73); LDL particles, 1.49 (1.24 to 1.78); small LDL, 1.48 (1.24 to 1.77); large HDL, 0.61 (0.48 to 0.77); small HDL, 1.25 (1.03 to 1.52); VLDL size, 1.48 (1.22 to 1.78); LDL size, 0.71 (0.59 to 0.86); and HDL size, 0.52 (0.41 to 0.66). After adjustment for SI, all of these measures continued to be statistically significant except VLDL particles and LDL size. Significant ORs for chemically measured lipid parameters were as follows: HDL cholesterol, 0.52 (0.41 to 0.68); triglycerides, 1.30 (1.10 to 1.55). Neither of these lipid measures remained significant after adjustment for SI.

    In stepwise logistic regression models, VLDL size and small HDL particle concentration emerged as significant NMR contributors to incident diabetes. Among conventional lipid measures, triglycerides and HDL cholesterol were significant predictors of diabetes in a demographically adjusted model (Table 6, model 1: conventional), whereas among NMR measures both VLDL size and small HDL remained significant in this model (model 2: +NMR).

    VLDL Size and Small HDL Particles in Relation to Incident Diabetes

    There was a linear increase in the incidence of diabetes across quartiles of VLDL size and to a lesser extent also across quartiles of small HDL particle concentrations (Figures 1 and 2). A comparable pattern was found in subjects with normal glucose tolerance (NGT) or IGT at baseline (Figures 1 and 2).

    Finally, VLDL size, but not small HDL particle concentration, remained statistically significantly related to incident diabetes in a fully adjusted, comprehensive model including commonly known risk factors for diabetes (model G).

    Stratified Analyses by Glucose Tolerance Status, Ethnicity, and Gender

    There were no significant interactions of glucose tolerance status and ethnicity on the relation of NMR variables (namely, VLDL size and small HDL particles) to incident diabetes, indicating that the relation of these variables of interest to incident diabetes was consistent across glucose tolerance categories and across the 3 ethnic groups of the IRAS (Figure 4). A borderline significance was found for gender, indicating a trend toward a stronger relation in women than in men (P=0.06 for VLDL size and P=0.08 for small HDL in unadjusted models, respectively, and P=0.036 and P=0.048 in demographically adjusted logistic regression models, respectively). The fully adjusted stratified logistic regression analyses (Figure 4) showed a significant interaction of gender on the relation of VLDL size to incident diabetes (P=0.013; Figure 4A) but not on the relation of small HDL to incident diabetes (P=0.36; Figure 4B). No significant interaction of gender, however, was found for VLDL size and small HDL in demographically adjusted ANCOVA (P=0.7 for VLDL size and P=0.2 for small HDL).

    Discussion

    The present report yielded several novel findings: (1) A broad range of abnormalities in lipoprotein composition precedes the onset of diabetes; (2) the associations for small HDL particle concentration and large VLDL size were independent of the greater insulin resistance present in prediabetic subjects; and (3) the findings were consistent across glucose tolerance categories and in the 3 ethnic groups of the IRAS. These results are based on 2 different statistical approaches: (1) logistic regression analyses with incident diabetes as the dependent variable and (2) ANCOVA models with the lipid variables of interest as respective dependent variables. Both approaches yielded similar results, which strengthens the results of the present study and the conclusion we would like to draw from these results. Thus, the present study provides evidence that abnormalities in lipoprotein composition occur before the onset of type 2 diabetes and therefore might contribute to the increased cardiovascular disease risk in prediabetic subjects. Although we believe that the highly significant findings (P<0.001) are robust, given the number of hypotheses tested and our use of stepwise methods, replication in other data sets is warranted.

    In one previous report (in which gradient gel electrophoresis was used), the preponderance of small LDL particles was identified as a risk factor for the development of diabetes in elderly individuals.8 In this study, subjects with small dense LDL particles had a >2-fold increased risk of incident diabetes over 3.5 years, independent of age, sex, glucose intolerance, and body mass index. Adjustment for either fasting insulin or triglycerides, however, rendered the association nonsignificant, suggesting that small dense LDL reflects underlying insulin resistance (as represented by elevated triglycerides and/or elevated fasting insulin levels) rather than a causal relation between this lipoprotein phenotype and the development of diabetes. No measure of insulin resistance is available from this report.

    In the present study, LDL size, as assessed by NMR technology, was also related to incident diabetes, but adjustment for insulin resistance (as directly measured) rendered this association nonsignificant. This indicates that the association of LDL size to incident diabetes is indeed explained by the (inverse) association of LDL size with insulin resistance, as shown in previous cross-sectional studies.19,20

    In the present study, VLDL particle size and small HDL particles, rather than LDL particle size, emerged as significant contributors to incident diabetes. The relation of these 2 lipoprotein particle measures to incident diabetes was independent of metabolic covariates, including insulin resistance. This is potentially important because in previous reports, both VLDL size and small HDL particles21 or HDL size (inversely)3,21 have been related to increased insulin resistance. Additionally, these and other NMR-identified lipoprotein abnormalities, which were prevalent in patients with type 2 diabetes, have been attributed to the underlying increased insulin resistance in these patients rather than the diabetic state itself.21 NMR technology has been used to investigate compositional lipoprotein abnormalities in several previous studies involving diabetic populations, including diabetic mice,22 patients with type 1 diabetes,23,24 patients with type 2 diabetes,21,25 or individuals with insulin resistance.21 In a cross-sectional study, a relation of small HDL to renal dysfunction was shown in patients with type 1 diabetes.23 In a placebo-controlled intervention trial, atorvastatin reduced concentrations of medium and small VLDL, as well as large and medium LDL, and increased concentrations of large HDL.25

    Our findings were consistent in the 3 ethnic groups of the IRAS as well as in subjects with NGT or IGT at baseline, but we found a gender interaction reaching borderline significance in the logistic regression models. Such a finding would be supportive of previous analyses from the San Antonio Heart Study, indicating that prediabetic women have a greater burden of cardiovascular risk factors than men.26 However, because of the relatively small number of cases of incident diabetes in women (n=76) and in men (n=53), respectively, this finding needs to be interpreted with caution.

    How might a relation of lipoprotein size and subclass composition to incident diabetes be explained; First, a causal relationship might exist, ie, alterations in lipoprotein size and subclass composition might directly affect the pathophysiology of type 2 diabetes, namely, increased insulin resistance and/or impaired insulin secretion. Free fatty acids are released from VLDL particles, and an excess of free fatty acids, in turn, has been shown to directly affect insulin resistance, namely, by inhibiting hepatic glycogen synthesis and glucose oxidation.27 Furthermore, we have shown in the present study that the relation of VLDL size to incident diabetes is attenuated after adjustment for insulin resistance, as directly measured. This indicates that increased insulin resistance in prediabetic individuals contributes (to some extent) to the demonstrated associations. Accordingly, in a previous report in which clamp techniques were used to assess insulin resistance, large VLDL particle concentration was suggested as the primary abnormality underlying the complex lipoprotein subclass pattern associated with increased insulin resistance (high serum triglycerides, decreased LDL size, increased overall LDL particle concentration, decreased large and increased small HDL particles).21

    Second, common genetic factors might underlie both the demonstrated lipoprotein abnormalities and diabetes risk. A number of studies have demonstrated that small dense LDL is genetically influenced,28 and a common polymorphism in the cholesteryl ester transfer protein gene affects HDL and LDL particle size.29 Potential candidate genes linking insulin resistance and dyslipidemia include the hepatic lipase gene and the fatty acid binding protein gene 2.30

    Third, lipoprotein abnormalities associated with the risk of developing diabetes might indicate a proatherogenic state in prediabetic individuals. It has indeed been shown in previous reports that the prediabetic state is characterized by an increased prevalence of cardiovascular risk factors.4–7 The present report extends this knowledge; abnormalities of size and composition of both VLDL and HDL particles have been related to cardiovascular disease.24,31–33 In the present study the lipoprotein parameters in prediabetic subjects resembled that in overt diabetic patients rather than that in nonconverters (Table 2). The view of a proatherogenic risk profile in prediabetic individuals is further supported by data from the Nurses’ Health Study, showing that the risk of developing cardiovascular disease in prediabetic women was increased, even before the onset of diabetes.10 The present study provides supportive data that might explain this finding; whether these demonstrated proatherogenic lipoprotein abnormalities in prediabetic individuals are related to increased insulin resistance or rather impairment in insulin secretion is the topic of further ongoing analyses.

    In summary, we have shown a range of lipoprotein abnormalities in prediabetic individuals. These findings extend previous work indicating a proatherogenic state in healthy, nondiabetic subjects who subsequently develop diabetes. This finding has important clinical and public health implications, providing a rationale for an aggressive prevention strategy in individuals at high risk of developing diabetes, such as those with IGT. Such a prevention strategy yields the potential to favorably affect diabetes incidence and also, and equally importantly, the development of cardiovascular disease.

    Acknowledgments

    This work was supported by the National Heart, Lung, and Blood Institute (grants Hl47887, Hl47889, Hl47890, Hl47892, Hl47902, Hl55208, and R01 Hl58329) and the General Clinic Research Centers Program (grants NCRR GCRC, M01 RR431, and M01 RR01346).

    Disclosure

    Dr Festa is an employee of Eli Lilly and Co.

    References

    Haffner SM. Management of dyslipidemia in adults with diabetes. Diabetes Care. 1998; 21: 160–178.

    Malmstr;m R, Packard CJ, Caslake M, Bedford D, Stewart P, Yki-Jarvinen H, Shepherd J, Taskinen MR. Defective regulation of triglyceride metabolism by insulin in the liver in NIDDM. Diabetologia. 1997; 40: 454–462.

    MacLean PS, Vadlamudi S, MacDonald KG, Pories WJ, Houmard JA, Barakat HA. Impact of insulin resistance on lipoprotein subpopulation distribution in lean and morbidly obese nondiabetic women. Metabolism. 2000; 49: 285–292.

    Haffner SM, Stern MP, Hazuda HP, Mitchell BD, Patterson JK. Cardiovascular risk factors in confirmed prediabetic individuals: does the clock for coronary heart disease start ticking before the onset of clinical diabetes; JAMA. 1990; 263: 2893–2898.

    McPhillips JB, Barrett-Connor E, Wingard DL. Cardiovascular disease risk factors prior to the diagnosis of impaired glucose tolerance and non–insulin dependent diabetes mellitus in a community of older adults. Am J Epidemiol. 1990; 131: 443–453.

    Mykkanen L, Kuusisto J, Pyorala K, Laakso M. Cardiovascular disease risk factors as predictors of type 2 (non–insulin dependent) diabetes mellitus in elderly subjects. Diabetologia. 1993; 36: 553–559.

    DeFronzo RA. Pathogenesis of type 2 (non–insulin dependent) diabetes mellitus: a balanced overview. Diabetologia. 1992; 35: 389–397.

    Austin MA, Mykkanen L, Kuusisto J, Edwards KL, Nelson C, Haffner SM, Pyorala K, Laakso M. Prospective study of small LDLs as a risk factor for non–insulin dependent diabetes mellitus in elderly men and women. Circulation. 1995; 92: 1770–1778.

    Otvos JD. Measurement of lipoprotein subclass profiles by nuclear magnetic resonance spectroscopy. Clin Lab. 2002; 48: 171–180.

    Hu FB, Stampfer MJ, Haffner SM, Solomon CG, Willett WC, Manson JE. Elevated risk of cardiovascular disease prior to clinical diagnosis of type 2 diabetes. Diabetes Care. 2002; 25: 1129–1134.

    Wagenknecht LE, Mayer EJ, Rewers M, Haffner SM, Selby J, Borok GM, Henkin L, Howard G, Savage PJ, Saad MF, Bergman RN, Hamman R. The Insulin Resistance Atherosclerosis Study (IRAS): objectives, design and recruitment results. Ann Epidemiol. 1995; 5: 464–471.

    Bergman RN, Finegood DT, Ader M. Assessment of insulin sensitivity in vivo. Endocr Rev. 1985; 6: 45–86.

    Pacini G, Bergman RN. MINMOD: a computer program to calculate insulin sensitivity and pancreatic responsivity from the frequently sampled intravenous glucose tolerance test. Comput Methods Programs Biomed. 1986; 23: 113–122.

    Festa A, D’Agostino R Jr, Mykk;nen L, Tracy RP, Hales CN, Howard BV, Haffner SM. Low density lipoprotein (LDL) particle size in relation to insulin, proinsulin, and insulin sensitivity: the Insulin Resistance Atherosclerosis Study. Diabetes Care. 1999; 22: 1688–1693.

    Freedman DS, Otvos JD, Jeyarajah EJ, Shalaurova I, Cupples LA, Parise H, D’Agostino RB, Wilson PWF, Schaefer EJ. Sex and age differences in lipoprotein subclasses measured by nuclear magnetic resonance spectroscopy: the Framingham Study. Clin Chem. 2004; 50: 1189–1200.

    Grundy SM, Vega GL, Otvos JD, Rainwater DL, Cohen JC. Hepatic lipase influences high density lipoprotein subclass distribution in normotriglyceridemic men: genetic and pharmacological evidence. J Lipid Res. 1999; 40: 229–234.

    Blake GJ, Otvos JD, Rifai N, Ridker PM. LDL particle concentration and size as determined by NMR spectroscopy as predictors of cardiovascular disease in women. Circulation. 2002; 106: 1930–1937.

    Rumsey SC, Galeano NF, Arad Y, Deckelbaum RJ. Cryopreservation with sucrose maintains normal physical and biological properties of human plasma low density lipoproteins. J Lipid Res. 1992; 33: 1551–1561.

    Reaven GM, Chen YDI, Jeppesen J, Maheux P, Krauss RM. Insulin resistance and hyperinsulinaemia in individuals with small dense low density lipoprotein particles. J Clin Invest. 1993; 92: 141–146.

    Mykkanen L, Haffner SM, Rainwater DL, Karhap;; P, Miettinen H, Laakso M. Relationship of LDL size to insulin sensitivity in normoglycemic men. Arterioscler Thromb Vasc Biol. 1997; 17: 1447–1453.

    Garvey WT, Kwon S, Zheng D, Shaughnessy S, Wallace P, Hutto A, Pugh K, Jenkins AJ, Klein RL, Liao Y. Effects of insulin resistance and type 2 diabetes on lipoprotein subclass particle size and concentration determined by nuclear magnetic resonance. Diabetes. 2003; 52: 453–462.

    Hammad SM, Powell-Braxton L, Otvos JD, Elridge L, Won W, Lyons TJ. Lipoprotein subclass profiles of hyperlipidemic diabetic mice measured by nuclear magnetic resonance spectroscopy. Metabolism. 2003; 52: 916–921.

    Jenkins AJ, Lyons TJ, Zheng D, Otvos JD, Lackland DT, McGee D, Garvey WT, Klein RL, and the DCCT/EDIC Research Group. Lipoproteins in the DCCT/EDIC cohort: associations with diabetic nephropathy. Kidney Int. 2003; 64: 817–828.

    Soedamah-Muthu SS, Chang Y-F, Otvos J, Orchard TJ. Lipoprotein subclass measurements by nuclear magnetic resonance spectroscopy improve the prediction of coronary artery disease in type 1 diabetes. Diabetologia. 2003; 46: 674–682.

    Soedamah-Muthu SS, Colhoun HM, Thomason MJ, Betteridge DJ, Durrington PN, Hitman GA, Fuller JH, Julier K, Mackness MI, Neil HAW, for the CARDS Investigators. The effect of atorvastatin on serum lipids, lipoproteins and NMR spectroscopy defined lipoprotein subclasses in type 2 diabetic patients with ischaemic heart disease. Atherosclerosis. 2003; 167: 243–255.

    Haffner SM, Miettinen H, Stern MP. Relatively more atherogenic coronary heart disease risk factors in prediabetic women than in prediabetic men. Diabetologia. 1997; 40: 711–717.

    Roden M, Price TB, Perseghin G, Petersen KF, Rothman DL, Cline GW, Shulman GI. Mechanism of free fatty acid–induced insulin resistance in humans. J Clin Invest. 1996; 97: 2859–2865.

    Austin MA. Genetics of low-density lipoprotein subclasses. Curr Opin Lipidol. 1993; 4: 125–132.

    Ordovas JM, Cupples LA, Corella D, Otvos JD, Osgood D, Martinez A, Lahoz C, Coltell O, Wilson PW, Schaefer EJ. Association of cholesteryl ester transfer protein-Taq1B polymorphism with variations in lipoprotein subclasses and coronary heart disease risk: the Framingham Study. Arterioscler Thromb Vasc Biol. 2000; 20: 1323–1329.

    Laakso M. Gene variants, insulin resistance, and dyslipidemia. Curr Opin Lipidol. 2004; 15: 115–120.

    Freedman DS, Otvos JD, Jeyarajah EJ, Barboriak JJ, Anderson AJ, Walker JA. Relation of lipoprotein subclasses as measured by proton nuclear magnetic resonance spectroscopy to coronary artery disease. Arterioscler Thromb Vasc Biol. 1998; 18: 1046–1053.

    Rosenson RS, Otvos JD, Freedman DS. Relations of lipoprotein subclass levels and low-density lipoprotein size to progression of coronary artery disease in the Pravastatin Limitation of Atherosclerosis in the Coronary Arteries (PLAC-I) trial. Am J Cardiol. 2002; 90: 89–94.

    Kuller L, Arnold A, Tracy R, Otvos J, Burke G, Psaty B, Siscovick D, Freedman DS, Kronmal R. Nuclear magnetic resonance spectroscopy of lipoproteins and risk of coronary heart disease in the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol. 2002; 22: 1175–1180.(Andreas Festa, MD; Ken Wi)