Lipoprotein subclass profiles in individuals with varying degrees of glucose tolerance: a population-based study of 9399 Finnish men

Authors

  • J. Wang,

    1. From the Department of Medicine, University of Eastern Finland and Kuopio University Hospital, Kuopio
    Search for more papers by this author
  • A. Stančáková,

    1. From the Department of Medicine, University of Eastern Finland and Kuopio University Hospital, Kuopio
    Search for more papers by this author
  • P. Soininen,

    1. Computational Medicine Research Group, Institute of Clinical Medicine, University of Oulu, Oulu
    2. NMR Metabonomics Laboratory, Laboratory of Chemistry, Department of Biosciences, University of Eastern Finland, Kuopio
    Search for more papers by this author
  • A. J. Kangas,

    1. Computational Medicine Research Group, Institute of Clinical Medicine, University of Oulu, Oulu
    Search for more papers by this author
  • J. Paananen,

    1. From the Department of Medicine, University of Eastern Finland and Kuopio University Hospital, Kuopio
    Search for more papers by this author
  • J. Kuusisto,

    1. From the Department of Medicine, University of Eastern Finland and Kuopio University Hospital, Kuopio
    Search for more papers by this author
  • M. Ala-Korpela,

    1. Computational Medicine Research Group, Institute of Clinical Medicine, University of Oulu, Oulu
    2. NMR Metabonomics Laboratory, Laboratory of Chemistry, Department of Biosciences, University of Eastern Finland, Kuopio
    3. Department of Internal Medicine and Biocenter Oulu, Clinical Research Center, University of Oulu, Oulu; Finland
    Search for more papers by this author
  • M. Laakso

    1. From the Department of Medicine, University of Eastern Finland and Kuopio University Hospital, Kuopio
    Search for more papers by this author

Markku Laakso MD, PhD, Academy Professor, Department of Medicine, University of Eastern Finland and Kuopio University Hospital, 70210 Kuopio, Finland.
(fax: +358 17 173993; e-mail: markku.laakso@kuh.fi).

Abstract

Abstract.  Wang J, Stančáková A, Soininen P, Kangas AJ, Paananen J, Kuusisto J, Ala-Korpela M, Laakso M (University of Eastern Finland and Kuopio University Hospital, Kuopio; Institute of Clinical Medicine, University of Oulu, Oulu; University of Eastern Finland, Kuopio; and Clinical Research Center, University of Oulu, Oulu; Finland). Lipoprotein subclass profiles in individuals with varying degrees of glucose tolerance: a population-based study of 9399 Finnish men. J Intern Med 2012; doi: 10.1111/j.1365-2796.2012.02562.x.

Objectives.  We investigated serum concentrations of lipoprotein subclass particles and their lipid components determined by proton nuclear magnetic resonance spectroscopy in a population-based study.

Design and methods.  A total of 9399 Finnish men were included in the study: 3034 men with normal fasting glucose and normal glucose tolerance; 4345 with isolated impaired fasting glucose (IFG); 312 with isolated impaired glucose tolerance (IGT); 1058 with both IFG and IGT; and 650 with newly diagnosed type 2 diabetes (New DM). Lipoprotein subclasses included chylomicrons (CM) and largest VLDL particles, other VLDL particles (five subclasses), intermediate-density lipoprotein (IDL), LDL (three subclasses) and HDL (four subclasses). The phospholipid, triglyceride (TG), cholesterol, free cholesterol and cholesterol ester levels of the lipoprotein particles were measured.

Results.  Abnormal glucose tolerance (especially IGT and New DM) was significantly associated with increased concentrations of VLDL subclass particles and their components (with the exception of very small VLDL particles). After further adjustment for total TGs and HDL cholesterol, increased lipid concentrations in the CM/largest VLDL particles and in most of the other VLDL particles remained significant in individuals with isolated IGT, IFG+IGT and New DM. There was a consistent trend towards a decrease in large and an increase in small HDL particle concentrations in individuals with hyperglycaemia even after adjustment for serum total TGs and HDL cholesterol.

Conclusions.  Abnormal glucose tolerance modifies the concentrations of lipoprotein subclass particles and their lipid components in the circulation and is also related to compositional changes in these particles.

Abbreviations:
2-h PG

2-h postload glucose during an oral glucose tolerance test

CE

cholesterol ester

CM

chylomicron

CVD

cardiovascular disease

FC

free cholesterol

FPG

fasting plasma glucose

IDL

intermediate-density lipoprotein

IIFG

isolated impaired fasting glucose

IIGT

isolated impaired glucose tolerance

METSIM

METabolic Syndrome In Men

New DM

newly diagnosed type 2 diabetes

NFG

normal fasting glucose

NGT

normal glucose tolerance

NMR

proton nuclear magnetic resonance

OGTT

oral glucose tolerance test

PL

phospholipid

TG

triglyceride

VLDL

very low-density lipoprotein

Introduction

Abnormal glucose tolerance increases the risk of cardiovascular disease (CVD) mortality and morbidity [1]. Alterations in lipoproteins and lipids could at least partially explain the fact that atherosclerosis is often accelerated in diabetes. Although previous studies have investigated the association between abnormalities in glucose tolerance and adverse changes in lipids and lipoproteins [2–11], these studies have been small in size, and none has systematically investigated the associations with lipoprotein subclasses [2–7]. In particular, we are not aware of any studies that have assessed the lipoprotein subclass lipid compositions in all categories of glucose tolerance. Therefore, we investigated the association between subclasses of serum lipoprotein particles and their lipid components measured by proton nuclear magnetic resonance (NMR) spectroscopy in a large sample of Finnish men with different degrees of glucose tolerance.

Design and methods

Study subjects

The METSIM (METabolic Syndrome In Men) study [12] is a population-based study including 10 197 Finnish men living in Kuopio town, Eastern Finland, randomly selected from the population register of 95 000 inhabitants. Diagnosis of type 2 diabetes was based on fasting plasma glucose (FPG) and/or 2-h postload glucose (2-h PG) levels in an oral glucose tolerance test (OGTT) conducted at baseline, and the lack of ketoacidosis or other signs and symptoms of type 1 diabetes. Individuals with prior known diabetes (787 with type 2 diabetes and 25 with type 1 diabetes) were excluded from all statistical analyses. A total of 9399 men (8749 without diabetes and 650 with newly diagnosed type 2 diabetes [New DM]) comprised the final study population. The protocol included a 1-day visit to the Clinical Research Unit of the University of Eastern Finland. This study was approved by the Ethics Committee of the University of Eastern Finland and was conducted in accordance with the Declaration of Helsinki.

Diagnosis of glucose tolerance and diabetes

According to the ADA criteria [13], participants were classified into the following categories of glucose tolerance at baseline based on FPG and 2-h PG: both normal fasting glucose (NFG) and normal glucose tolerance (NGT) (NFG+NGT; FPG <5.6 mmol L−1 and 2-h PG <7.8 mmol L−1); isolated impaired fasting glucose (IIFG; FPG between 5.6 and 6.9 mmol L−1 and 2-h PG <7.8 mmol L−1); isolated impaired glucose tolerance (IIGT; 2-h PG between 7.8 and 11.0 mmol L−1 and FPG <5.6 mmol L−1); both impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) (IGT+IFG; FPG between 5.6 and 6.9 mmol L−1 and 2-h PG between 7.8 and 11.0 mmol L−1); and New DM (FPG ≥7.0 mmol L−1 or 2-h PG ≥11.1 mmol L−1). Our study cohort included 3034 men with NFG+NGT, 4345 with IIFG, 312 with IIGT, 1058 with IFG+IGT and 650 with New DM.

Anthropometric measurements and biochemical assays

Family history of diabetes was defined as a history of type 2 diabetes amongst first-, second- or third-degree relatives. Physical activity during leisure time was classified as physically inactive (little and occasional activity) or physically active (regular exercise at least once a week for at least 30 min per session). The use of statins was dichotomized. Because only 32 subjects were receiving fibrates, the use of these drugs was not included in the analyses. Body weight of subjects wearing light clothing without shoes was measured to the nearest 0.1 kg, and height was measured to the nearest 0.5 cm; body mass index (BMI) was calculated as body weight in kilograms divided by height in metres squared. Waist circumference was measured as the average of two measurements taken after inspiration and expiration at the midpoint between the lowest rib and iliac crest. During a 2-h OGTT (75 g glucose), blood samples for the measurement of plasma glucose and insulin were drawn at 0, 30 and 120 min to evaluate the glucose tolerance and the insulin response to an oral glucose load. Plasma glucose was measured using the enzymatic hexokinase photometric assay (Konelab System Reagents; Thermo Fischer Scientific, Vantaa, Finland). Serum insulin was determined by immunoassay (ADVIA Centaur Insulin IRI, no 02230141; Siemens Medical Solutions Diagnostics, Tarrytown, NY, USA). Total cholesterol, LDL and HDL cholesterol, and total triglyceride levels were determined by enzymatic colorimetric tests (Konelab System Reagents), and concentrations of apolipoproteins A1 (ApoA1) and B (ApoB) were measured by immunoturbidometric methods (Konelab Systems Reagents).

Lipoprotein subclass analysis by NMR spectroscopy

Fasting concentrations of lipoprotein subclass particles and their main lipid components were analysed by proton NMR spectroscopy in native serum samples [14–16]. The NMR data were measured at 37 °C using a Bruker AVANCE III spectrometer operating at 500.36 MHz using a new automated platform, as described previously [17]. The following 14 lipoprotein subclasses were calibrated using high-performance liquid chromatography [18, 19]: chylomicrons (CMs) and largest VLDL particles (CM/largest VLDL; average particle diameter ≥75 nm); five different VLDL subclasses, i.e. very large (average particle diameter 64.0 nm), large (53.6 nm), medium (44.5 nm), small (36.8 nm) and very small VLDL (31.3 nm); intermediate-density lipoprotein (IDL; 28.6 nm); three LDL subclasses, i.e. large (25.5 nm), medium (23.0 nm) and small LDL (18.7 nm); and four HDL subclasses, i.e. very large (14.3 nm), large (12.1 nm), medium (10.9 nm) and small HDL (8.7 nm). The following components of the lipoprotein particles were quantified: phospholipids (PLs), triglycerides (TGs), cholesterol, free cholesterol (FC) and cholesterol esters (CEs). Because of resolution of the method, not all of these components were measured for every subclass [18]. The percentage composition of the measured individual lipids with regard to the total lipid content in each lipoprotein subclass was also calculated. The mean size for VLDL, LDL and HDL particles was calculated weighting the subclass diameters with their corresponding particle concentrations. IDL particles were included in the LDL measure.

Calculation of insulin sensitivity

Matsuda index, calculated from glucose and insulin concentrations obtained during the OGTT, was used to estimate insulin sensitivity [20].

Statistical analyses

Because of skewed distributions, concentrations of lipids and lipoproteins were log transformed for statistical analysis. anova adjusted for covariates was applied to compare differences in serum lipids and in lipoprotein subclass particles and their main lipid components across the categories of glucose tolerance. P-values for pair-wise comparisons between the groups were corrected with Bonferroni method for multiple testing. Pearson correlations of Matsuda insulin sensitivity index with lipoprotein subclass particle concentrations and their main lipid components were calculated. All statistical analyses were performed with spss 14.0 software (SPSS, Chicago, IL, USA).

Results

Concentrations of serum lipids and insulin sensitivity in the different glucose tolerance categories

Compared to subjects with NFG+NGT, more subjects with IIFG, IIGT, IFG+IGT and New DM had a family history of diabetes, were taking statins, were obese and had hypertension; in addition, those with IIGT, IFG+IGT and New DM were older and more subjects with IFG+IGT and New DM were physically inactive (Table 1). After adjustments for age, family history of diabetes, physical activity during leisure time, BMI, waist circumference and the use of statins, the IIFG, IIGT, IFG+IGT and New DM groups had significantly increased concentrations of total triglycerides (14%, 36%, 51% and 62%, respectively) and decreased Matsuda index (−31%, −36%, −57% and −63%, respectively), compared to the NFG+NGT group. Abnormal glucose tolerance affected the average size of lipid particles, especially VLDL particles. Individuals in the IIGT, IFG+IGT and New DM groups had significant decreases in concentrations of HDL cholesterol (−6%, −11% and −10%, respectively), compared to the NFG+NGT group; furthermore, increased concentrations of total cholesterol were observed in the IIFG (2%), IFG+IGT (3%) and New DM groups (4%). Increased concentrations of LDL cholesterol were observed only in the IIFG group (2%).

Table 1. Concentrations of serum lipid measures and mean size of lipid particles, and clinical characteristics in comparison with subjects with normal fasting glucose and normal glucose tolerance
 NFG+NGT (= 3034)IIFG (= 4345)IIGT (= 312)IFG+IGT (= 1058)New DM (= 650) P
  1. NFG, normal fasting glucose; NGT, normal glucose tolerance; IIFG, isolated impaired fasting glucose; IIGT, isolated impaired glucose tolerance; New DM, newly diagnosed type 2 diabetes; SBP, systolic blood pressure; VLDL, very low-density lipoprotein.

  2. *P-values for pair-wise comparisons between the NFG+NGT group and the other groups using Bonferroni correction. †P, P‡ for overall differences across the groups (†P, unadjusted; P‡, adjusted for age, family history of diabetes, physical activity during leisure time, BMI, waist circumference and use of statins). aData are percentage. bmg dL · mU L. c< 0.05 for the comparison between the IIGT and IIFG groups using Bonferroni correction.

 Mean ± SDMean ± SD P*Mean ± SD P*Mean ± SD P*Mean ± SD P* 
Age (years)56.8 ± 6.956.7 ± 7.01.00059.8 ± 7.11 × 10−1159.0 ± 7.18 × 10−1759.4 ± 6.82 × 10−161 × 10−41
Family history of diabetes (%)a43483 × 10−4490.037517 × 10−7556 × 10−12 × 10−13
Physically inactive during leisure time (%)a32350.060371.000451 × 10−10461 × 10−131 × 10−18
Use of statins (%)a20254 × 10−4312 × 10−4332 × 10−15321 × 10−92 × 10−20
BMI (kg m−2)25.7 ± 3.327.0 ± 3.61 × 10−4427.0 ± 3.63 × 10−828.9 ± 4.35 × 10−12129.7 ± 4.99 × 10−1268 × 10−196
Waist (cm)94.3 ± 9.598.0 ± 10.33 × 10−5098.8 ± 9.91 × 10−12103.4 ± 11.41 × 10−128105.5 ± 12.41 × 10−1332 × 10−208
SBP (mmHg)134 ± 15137 ± 154 × 10−14141 ± 173 × 10−14143 ± 172 × 10−54147 ± 182 × 10−742 × 10−107
 Mean ± SD (mmol L−1)% change P*% change P*% change P*% change P* P
Total cholesterol5.31 ± 0.95+27 × 10−7+21.000+30.002+47 × 10−62 × 10−8
LDL cholesterol3.35 ± 0.85+20.008+11.000+11.000+20.8470.019
HDL cholesterol1.50 ± 0.39−21.000−60.034c−112 × 10−5−100.0224 × 10−9
Triglycerides1.22 ± 0.66+141 × 10−8+361 × 10−15c+513 × 10−48+622 × 10−485 × 10−80
Matsuda index9.0 ± 4.7b−311 × 10−110−369 × 10−35c−579 × 10−242−635 × 10−2301 × 10−209
 Mean ± SD (nm)% change P*% change P*% change P*% change P* P
Diameter of VLDL36 ± 1+10.001+21 × 10−11+34 × 10−34+32 × 10−214 × 10−49
Diameter of LDL23 ± 0.100.00101 × 10−405 × 10−1209 × 10−108 × 10−17
Diameter of HDL10 ± 0.300.025−10.104−14 × 10−13−10.0062 × 10−12

Lipoprotein subclass particles and apolipoproteins in the different glucose tolerance categories (Table 2)

Table 2. Lipoprotein subclass particle concentrations and apolipoproteins in comparison with subjects with normal fasting glucose and normal glucose tolerance
 NFG+NGT (= 3034)IIFG (= 4345)IIGT (= 312)IFG+IGT (= 1058)New DM (= 650) P P†§
Mean ± SD (nmol L−1)% change P* P% change P* P% change P* P% change P* P
  1. NFG, normal fasting glucose; NGT, normal glucose tolerance; IIFG, isolated impaired fasting glucose; IIGT, isolated impaired glucose tolerance; New DM, newly diagnosed type 2 diabetes; VLDL, very low-density lipoprotein; IDL, intermediate-density lipoprotein; ApoA1/ApoB, apolipoprotein A1 to apolipoprotein B ratio.

  2. *Model 1, P-values for pair-wise comparisons between the NFG+NGT group and the other groups using Bonferroni correction †Model 1, P for overall difference across the groups (anova adjusted for age, family history of diabetes, physical activity during leisure time, BMI, waist circumference and use of statins). §Model 2, total triglycerides and HDL cholesterol were added into Model 1.

CM/largest VLDL0.05 ± 0.09+140.0341.000+1023 × 10−70.928+839 × 10−141.000+1072 × 10−150.0283 × 10−240.018
VLDL
 Very large0.54 ± 0.58+101 × 10−41.000+596 × 10−111.000+545 × 10−261.000+594 × 10−231.0005 × 10−420.443
 Large3.99 ± 3.64+88 × 10−51.000+481 × 10−70.376+445 × 10−291.000+462 × 10−181 × 10−42 × 10−381 × 10−4
 Medium16 ± 10+53 × 10−40.086+283 × 10−80.016+282 × 10−270.010+272 × 10−158 × 10−186 × 10−364 × 10−17
 Small32 ± 11+44 × 10−51.000+122 × 10−52 × 10−5+131 × 10−196 × 10−8+132 × 10−102 × 10−231 × 10−231 × 10−29
 Very small36 ± 10+20.0731.000+21.0000.044+30.0461 × 10−4+30.3282 × 10−50.0164 × 10−9
 IDL96 ± 23+10.6821.000+11.0000.639+11.0000. 022+21.0000.9080.2730.003
LDL
 Large170 ± 41+10.0801.000+11.0000.617+11.0000.072+20.1161.0000.0360.004
 Medium142 ± 36+10.0081.000+11.0000.703+10.1540.164+40.0041.0000.0010.006
 Small159 ± 39+24 × 10−41.000+30.2561.000+30.0010.592+54 × 10−61.0002 × 10−70.041
HDL
 Very large273 ± 174−20.9411.000−30.6401.000−141 × 10−90.075−30.1381.0005 × 10−90.075
 Large873 ± 425−20.0720.060−92 × 10−50.563−134 × 10−150.015−81 × 10−61.0009 × 10−180.004
 Medium1616 ± 360+10.5051.000−11.0001.000+20.2870.161+21.0001.0000.0700.080
 Small4547 ± 502+24 × 10−70.011+11.0001.000+29 × 10−70.097+20.0011.0002 × 10−100.001
ApoA1/ApoB1.51 ± 0.01−11.0001.000−50.0090.442−53 × 10−70.004−55 × 10−70.0621 × 10−114 × 10−4

CM and VLDL.  The concentrations of CM/largest VLDL particles (14–107%) and very large (10–59%), large (8–48%), medium (5–28%) and small VLDL (4–13%) were significantly increased compared to the NFG+NGT group after adjustment for age, family history of diabetes, physical activity during leisure time, BMI, waist circumference and the use of statins (Model 1). After further adjustment for total TGs and HDL cholesterol (Model 2), the percentage changes in the concentrations of CM/largest VLDL and large, medium and small VLDL subclass particles remained statistically significant in the New DM group.

IDL and LDL.  Compared to the NFG+NGT group, small LDL particle concentrations were increased significantly in the IIFG (2%), IFG+IGT (3%) and New DM (5%) groups, but these changes were no longer significant after adjustment for total TGs and HDL cholesterol (Model 2).

HDL.  Large HDL particle concentrations were significantly decreased in the IIGT (−9%), IFG+IGT (−13%) and New DM (−8%) groups, but these changes were also no longer significant after adjustment for serum total TGs and HDL cholesterol.

ApoA1/ApoB ratio.  Individuals with IGT (isolated or combined with IFG) and New DM had significantly lower apoA1/ApoB ratios (by ∼5%) than those with NGT. These differences were not statistically significant (except for the IFG+IGT group) after adjusting for TGs and HDL cholesterol levels.

Concentrations of lipoprotein subclass lipid components in the different glucose tolerance categories

CM and VLDL particles.  Phospholipid and TG concentrations increased significantly in CM/largest VLDL particles (up to 135% and 118%, respectively) and in all other larger VLDL particles (up to 96% and 91%, respectively) as the degree of glucose tolerance decreased (Table 3). Similar changes were also observed for cholesterol (66%), FC (72%) and CEs (60%) in large VLDL particles. After adjustment for confounding variables, including total TGs and HDL cholesterol (Model 2), changes were no longer statistically significant in any groups except for the New DM group. Increasing hyperglycaemia across glucose tolerance categories had substantially less effect on the concentrations of PLs, TGs, cholesterol, FC and CEs (9–50%) in medium and small VLDL subclasses, although the differences remained statistically significant after adjustment for all confounding factors (Model 2) in the IIGT, IFG+IGT and New DM groups (with the exception of PLs in very small VLDL).

Table 3. Lipid concentrations of VLDL subclasses in comparison with subjects with normal fasting glucose and normal glucose tolerance
 NFG+NGT (= 3034)IIFG (= 4345)IIGT (= 312)IFG+IGT (= 1058)New DM (= 650) P P†§
Mean ± SD (μmol L−1)% change P* P% change P* P% change P* P% change P* P
  1. NFG, normal fasting glucose; NGT, normal glucose tolerance; IIFG, isolated impaired fasting glucose; IIGT, impaired glucose tolerance; New DM, newly diagnosed type 2 diabetes; CM, chylomicrons; VLDL, very low-density lipoprotein; PL, phospholipid; TG, triglyceride; FC, free cholesterol; CE, cholesterol ester.

  2. *Model 1, P-values for pair-wise comparisons between the NFG+NGT group and the other groups using Bonferroni correction. †Model 1, P for overall difference across the groups (anova adjusted for age, family history of diabetes, physical activity during leisure time, BMI, waist circumference and use of statins). §Model 2, total triglycerides and HDL cholesterol were added into Model 1.

CM and lar-VLDL
 PLs1.53 ± 2.32+290.0331.000+1008 × 10−140.001+1112 × 10−200.095+1351 × 10−240.0019 × 10−435 × 10−7
 TGs11 ± 14+253 × 10−51.000+898 × 10−173 × 10−5+964 × 10−361 × 10−6+1184 × 10−332 × 10−62 × 10−613 × 10−12
Very large VLDL
 PLs8.49 ± 9.82+220.0031.000+758 × 10−120.233+839 × 10−241.000+961 × 10−221.0006 × 10−420.061
 TGs36 ± 38+224 × 10−51.000+701 × 10−120.675+813 × 10−350.006+913 × 10−251.0006 × 10−520.005
Large VLDL
 PLs41 ± 40+190.0230.032+602 × 10−111.000+697 × 10−271.000+765 × 10−184 × 10−43 × 10−412 × 10−4
 TGs149 ± 134+199 × 10−51.000+573 × 10−81.000+682 × 10−291.000+735 × 10−300.0231 × 10−400.020
 Cholesterol51 ± 47+170.0031.000+542 × 10−101.000+613 × 10−251.000+664 × 10−170.0027 × 10−370.007
 FC26 ± 24+191 × 10−41.000+573 × 10−101.000+662 × 10−281.000+727 × 10−180.0082 × 10−390.015
 CEs25 ± 22+160.0210.367+512 × 10−91.000+565 × 10−170.011+601 × 10−133 × 10−59 × 10−283 × 10−6
Medium VLDL
 PLs111 ± 62+114 × 10−40.051+299 × 10−80.004+372 × 10−253 × 10−4+392 × 10−141 × 10−194 × 10−331 × 10−19
 TGs290 ± 189+149 × 10−51.000+371 × 10−70.027+477 × 10−280.389+502 × 10−149 × 10−158 × 10−352 × 10−14
 Cholesterol159 ± 80+90.0030.001+256 × 10−80.008+292 × 10−235 × 10−8+314 × 10−131 × 10−223 × 10−302 × 10−23
 FC69 ± 40+111 × 10−40.158+319 × 10−70.003+483 × 10−260.001+401 × 10−132 × 10−215 × 10−332 × 10−21
 CEs89 ± 41+70.0495 × 10−5+206 × 10−70.024+231 × 10−169 × 10−11+257 × 10−101 × 10−204 × 10−235 × 10−23
Small VLDL
 PLs160 ± 46+62 × 10−51.000+93 × 10−42 × 10−4+148 × 10−162 × 10−6+144 × 10−84 × 10−162 × 10−183 × 10−23
 TGs262 ± 108+97 × 10−51.000+194 × 10−63 × 10−7+267 × 10−237 × 10−8+281 × 10−222 × 10−274 × 10−285 × 10−34
 Cholesterol285 ± 82+44 × 10−41.000+60.0210.016+86 × 10−83 × 10−5+90.0018 × 10−106 × 10−95 × 10−15
 FC108 ± 32+111 × 10−51.000+310.0010.001+485 × 10−142 × 10−5+406 × 10−72 × 10−134 × 10−162 × 10−20
Very small VLDL
 PLs153 ± 4201.0001.000−31.0000.010−21.0009 × 10−501.0000.0010.2297 × 10−8
 TGs124 ± 36+52 × 10−41.000+80.0044 × 10−4+132 × 10−124 × 10−6+151 × 10−81 × 10−82 × 10−151 × 10−14

IDL and LDL particles.  There were no consistent changes with increasing hyperglycaemia across glucose tolerance categories with regard to lipid concentrations of IDL, large LDL, medium LDL or small LDL particles (Table S1).

HDL particles.  Concentrations of PLs, TGs, cholesterol, FC and CEs in very large HDL particles did not change consistently with increasing hyperglycaemia across the glucose tolerance categories (Table S2). By contrast, these lipid concentrations were decreased consistently in large HDL particles in the IIGT, IFG+IGT and New DM groups, but the results were no longer statistically significant when adjusted for total TGs and HDL cholesterol (Model 2). No consistent changes were observed in medium HDL particles; however, TG levels in small HDL particles were increased significantly in the IFG+IGT and New DM groups even after adjustment for all confounding factors.

Correlation between insulin sensitivity and lipoprotein subclass and lipid concentrations

In all glucose tolerance groups, concentrations of the larger VLDL and the small HDL particles were negatively correlated with insulin sensitivity, whereas insulin sensitivity and very large and large HDL particles were positively correlated (Fig. 1a). All the lipid components of larger VLDL particles as well as TGs in the very small VLDL were negatively correlated with insulin sensitivity (Fig. 1b), whereas FC in IDL particles and insulin sensitivity were positively correlated (Fig. 1c). PL and all cholesterol measures in very large and large HDL particles were positively correlated with insulin sensitivity; by contrast, TG levels in small HDL particles were negatively correlated with insulin sensitivity (Fig. 1d).

Figure 1.

 Correlation between the Matsuda insulin sensitivity index and different lipoprotein particle concentrations (a) and their lipid compositions: VLDL (b); IDL and LDL (c) and HDL (d). The colour of each square represents the strength and direction of correlation (shades of red = positive correlation, shades of blue = negative correlation), and the size of an enclosed circle represents the level of statistical significance (P-value).

Lipid composition of the lipoprotein subclasses in the different glucose tolerance categories

Results for the VLDL, IDL and LDL, and HDL subclasses are shown in Tables S3–S5, respectively. The TG concentrations in the medium and smaller VLDL particles were increased in all hyperglycaemia categories in comparison with the NFG+NGT group; conversely, the cholesterol content showed a concomitant decrease in all hyperglycaemia categories in comparison with the normoglycaemia group. In the larger VLDL subclasses, no consistent compositional differences were seen. The lipid changes in the IDL subclass were similar to those observed for the medium and smaller VLDL particles. There was also a clear tendency towards a decrease in cholesterol in all LDL subclass particles in all hyperglycaemic groups. Measurements of TG levels were not available for the LDL subclasses. The changes in composition with hyperglycaemia varied amongst the HDL subclasses; for example, the total cholesterol content in both the very large and large HDL particles appeared to increase as a result of hyperglycaemia but changes in the relative content of free and esterified cholesterol differed between these particles (Table S5).

Discussion

To the best of our knowledge, the present study is the first to investigate in detail the concentrations of lipoprotein subclasses in individuals with different degrees of glucose tolerance. We found that the concentrations of all lipid components in the VLDL subclasses were increased as glucose tolerance decreased. No consistent changes were found in the IDL and LDL subclasses, whereas the HDL subclass lipids were consistently decreased in the different categories of abnormal glucose tolerance, particularly in the large HDL subclasses. Insulin sensitivity, independently of hyperglycaemia, is likely to play a major role in these changes because the correlations between insulin sensitivity and lipoprotein subclass measures were similar across all glucose tolerance categories.

In previous studies in this area [2–5, 8–11], lipoprotein subclasses and their lipid components have not been systematically examined in subjects with varying degrees of glucose tolerance We found that in comparison with subjects with NFG and NGT, individuals with abnormal glucose tolerance had consistently increased concentrations of larger VLDL particles and decreased concentrations of larger HDL particles. Additionally, the percentage differences in the concentrations of CM/largest VLDL particles and other VLDL particles (except for very small VLDL) and their lipid components, such as PLs and TGs, were independent of the concentrations of serum total TGs and HDL cholesterol in subjects with New DM. Increased concentrations of VLDL particles of larger size and small HDL particles and their lipid components were significantly associated with reduced insulin sensitivity, whereas larger-sized HDL particles were significantly associated with improved insulin sensitivity. These findings suggest that in addition to hyperglycaemia, insulin sensitivity may play an important role in the modification of lipoprotein subclass metabolism.

Lipoprotein particles contain a hydrophobic core, composed mainly of CEs and TGs, and an outer monolayer predominantly consisting of PLs, FC and various apolipoproteins [21]. Each lipoprotein fraction (CMs, VLDL, IDL, LDL and HDL) or subclass varies in size, density and lipid composition. Given that CM/largest VLDL particles are reported to contain up to 85% TGs (up to 75% in the present study, and an increased total TG level is associated with hyperglycaemia [22–24]), the present finding of an association between TGs in CM and VLDL particles and abnormal glucose tolerance was not unexpected. We also found associations between hyperglycaemia and both PLs in CM/largest VLDL particles and FC in large HDL particles. Because previous studies have not investigated the lipoprotein subclass lipid compositions in individuals with abnormal glucose tolerance, further studies are needed to confirm our findings that hyperglycaemia modifies the lipid components in CM/largest VLDL and other VLDL particles as well as in large HDL particles. When interpreting our findings in comparison with those of previous studies, differences in the definitions of lipoprotein subclasses and their size ranges, particularly with regard to the heterogeneous HDL particles, should be taken into account [15, 25, 26].

Our study is the first to provide evidence that, in addition to more unfavourable levels of serum lipids (e.g. increased total triglycerides), low HDL cholesterol and insulin sensitivity, subjects with IIGT also have more adverse changes in lipoprotein subclass profiles, compared to individuals with IIFG. These lipoprotein subclass changes include increased concentrations of larger-sized VLDL subclass particles and their lipid components as well as decreased concentrations of larger-sized HDL particles and their lipid components. In a prospective study of nondiabetic healthy women, the use of lipoprotein profiles evaluated by NMR was comparable but not superior to that of standard lipids for CVD risk prediction [26]. However, our findings suggest that abnormal lipoprotein profiles may explain why IGT is more closely associated with CVD mortality than IFG [27]. Prospective follow-up of this and other cohorts are needed to confirm this conclusion.

Previous studies have provided inconsistent results concerning whether the association between large VLDL particles and diabetes is independent of serum lipids [8–11]. Small sample sizes, different covariates used in statistical adjustments, and differences between ethnic groups and methods used to quantify lipoprotein subclasses may contribute to these inconsistent findings. We used proton NMR spectroscopy which has been previously validated against high-performance liquid chromatography [18, 19] and successfully applied in our recent large-scale genetic study [28]. None of the previous studies has taken into account lipid-lowering medication or several other confounding factors that were included in statistical analyses in the present study. Our results indicate that the increased serum concentrations of TG-rich lipoproteins are clearly related to hyperglycaemia whilst the compositional TG-enrichment of the lipoprotein subclass particles in hyperglycaemia, even though a consistent finding, is likely to be due to hypertriglyceridaemia rather than hyperglycaemia per se. A limitation of our study is that it included only white European men, and therefore, it remains uncertain whether our findings are applicable to women or other ethnic groups.

In conclusion, hyperglycaemia and insulin resistance were associated with increased serum concentrations of essentially all ApoB-containing lipoprotein particles. The serum concentrations of the smallest HDL particles were also slightly increased, whereas those of the largest HDL particles decreased as a result of hyperglycaemia and insulin resistance. Compositional TG-enrichment of some lipoprotein subclass particles was also related to hyperglycaemia, although these differences were due more to hypertriglyceridaemia than to hyperglycaemia.

Conflict of interest statement

No conflicts of interest to declare.

Acknowledgements

This work was supported by grants from the Academy of Finland (A.S., M.A.-K., M.L.), the Finnish Diabetes Research Foundation (M.L.), the Finnish Foundation for Cardiovascular Research (M.A.-K., M.L.), the University of Eastern Finland and the Kuopio University Hospital (EVO grant no. 5207; M.L.).

Ancillary