UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
UvA-DARE (Digital Academic Repository)
The puzzle of high-density lipoprotein in cardiovascular prevention
El-Harchaoui, A.
Publication date
2009
Link to publication
Citation for published version (APA):
El-Harchaoui, A. (2009). The puzzle of high-density lipoprotein in cardiovascular prevention.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
high density lipoprotein size and particle concentration and
coronary risk
A Karim El Harchaoui, Benoit J Arsenault, Remco Franssen, Jean-Pierre Després, G Kees Hovingh, Erik SG Stroes, James D Otvos, Nicholas J Wareham, John JP Kastelein, Kay-Tee Khaw, S Matthijs Boekholdt.
Chapt
er 3
32
absTracT
background: High-density lipoprotein (HDL) cholesterol levels are inversely related to risk of
coronary artery disease (CAD). Since HDL particles are heterogeneous in size and composition, they may be differentially associated with other cardiovascular risk factors and cardiovascular risk per se.
objective: To study the independent relationships of HDL size and HDL particle concentration
to risk of future CAD.
design: Nested case-control study within the EPIC-Norfolk cohort. Baseline survey between
1993 and 1997, follow-up until November 2003.
setting: Norfolk, United Kingdom.
participants: Cases were 822 apparently healthy men and women who developed CAD during
follow-up. Controls were 1401 participants who remained free of CAD, and were matched to cases by sex, age and enrollment time.
measurements: First CAD events leading to either hospitalization or death.
results: Nuclear magnetic resonance spectroscopy-measured HDL particle concentration
(33.9±5 vs 32.9±6 µmol/l; p<0.001) and HDL size (8.9±0.5 vs 8.8±0.5 nm; p < 0.001) as well as gradient gel electrophoresis-measured HDL size (88.6±4.2 vs 88.1±4.3 nm; p=0.005) were lower in cases compared to controls. HDL size and HDL particle concentration were only weakly correlated (for nuclear magnetic resonance spectroscopy-measured r= 0.08, for gradient gel electrophoresis-measured r=0.10). HDL size was strongly associated with risk factors character-istic of the metabolic syndrome, including waist-to-hip ratio, triglycerides, and apolipoprotein B, whereas HDL particle concentration was not. HDL size and HDL particle concentration were independently associated with CAD risk. The association between HDL size and CAD risk was abolished upon adjustment for apolipoprotein B and triglycerides (adjusted odds ratio,1.00 [95% CI 0.71 to 1.39] for top versus bottom quartile), whereas HDL particle concentration remained independently associated with CAD risk (adjusted odds ratio, 0.50 [95% CI 0.37 to 0.66]).
limitations: Measurements were performed in non-fasting blood samples and residual
con-founding cannot be excluded.
conclusions: HDL size and HDL particle concentration were independently associated with
other cardiovascular risk factors and with the risk of developing CAD. The relationship between HDL size and CAD risk was completely explained by markers associated with the metabolic syndrome, indicating that part of the relationship between HDL cholesterol and CAD risk is merely a reflection of this constellation of metabolic risk.
inTroducTion
The strong inverse relationship between high-density lipoprotein (HDL) cholesterol levels and risk of coronary artery disease (CAD) is well established (1). However, at any given level of HDL cholesterol, HDL particle concentration and HDL size distribution (see glossary table for definitions) may differ substantially between individuals. The atheroprotective role of HDL cholesterol is believed to be mediated mainly by its role in reverse cholesterol transport, the biological pathway that facilitates removal of cholesterol from macrophages in the arterial wall back to the liver (2). Substantial evidence suggests that HDL particles are heterogeneous in their efficacy to facilitate ATP-binding cassette A1 (ABCA1)-mediated cholesterol efflux from macrophages and scavenger receptor class BI (SR-BI)-mediated hepatic uptake of cholesterol from HDL particles. In particular, this heterogeneity has been associated with HDL particle size (3-5). In addition to its role in reverse cholesterol transport, accumulating evidence suggests that HDL particles have anti-coagulant, anti-oxidative and anti-inflammatory properties which contribute to the anti-atherogenic capacity of HDL. In fact, HDL lipoproteins were recently shown to carry a wide array of proteins that mediate these properties (6). The binding affinity of these proteins to the surface of HDL lipoproteins may depend on their size (7;8). Under certain conditions, HDL lipoproteins may lose their anti-atherogenic capacity and become dysfunc-tional (9;10). As a consequence, various HDL lipoprotein subpopulations may differ substan-tially in their capacity to play an atheroprotective role. Despite this functional heterogeneity, HDL has thus far been regarded as a single entity in epidemiological studies and the amount of cholesterol transported by HDL particles is traditionally being used for this purpose. However, we have recently shown that higher HDL cholesterol levels may not necessarily be associated with lower cardiovascular risk (11) and the HDL cholesterol raising drug torcetrapib has recently been shown to have detrimental effects (12). These findings emphasize that a broader perspec-tive on HDL metabolism is warranted.
We hypothesized that HDL particle concentration and HDL size distribution are differen-tially associated with cardiovascular risk factors and with cardiovascular risk. We tested these hypotheses in a case-control study nested in the EPIC-Norfolk cohort.
meThods
We performed a nested case-control study among participants of the European Prospective Investigation into Cancer and Nutrition (EPIC)-Norfolk study, a prospective population study of 25,663 men and women aged between 45 and 79 years, resident in Norfolk, United Kingdom, who completed a baseline questionnaire survey and attended a clinic visit (Figure). Participants were recruited from age-sex registers of general practices in Norfolk as part of the ten-country collaborative EPIC study designed to investigate dietary and other determinants of cancer.
Chapt
er 3
34
Additional data were obtained in EPIC-Norfolk to enable the assessment of determinants of other diseases.
The design and methods of the study have been described in detail (13). In short, eligible participants were recruited by mail. At the baseline survey between 1993 and 1997, partici-pants completed a detailed health and lifestyle questionnaire. Non-fasting blood samples were obtained by venipuncture into plain and citrate bottles. Blood samples were processed for assay at the Department of Clinical Biochemistry, University of Cambridge, or stored at –80˚C. All individuals have been flagged for death certification at the United Kingdom Office of National Statistics, with vital status ascertained for the entire cohort. In addition, participants admitted to hospital were identified using their unique National Health Service number by data linkage with the East Norfolk Health Authority (ENCORE) database, which identifies all hospital contacts throughout England and Wales for Norfolk residents. CAD was defined as codes 410-414 according to the International Classification of Diseases 9th revision. Participants were identified as having CAD during follow-up if they had a hospital admission and/or died with CAD as underlying cause. Previous validation studies in our cohort indicate high specificity of such case ascertainment (14). We report results with follow-up up to November 2003, an aver-age of about 6 years. The study was approved by the Norwich District Health Authority Ethics Committee and all participants gave signed informed consent.
Glossary
Adenosine triphospate-binding cassette A1 (ABCA1)
A cell membrane transporter that facilitates the delivery of cholesterol from cells to lipid-poor apolipoprotein A-I in the extracellular space High density lipoproteins (HDL) The smallest (7.0 nm and 13 nm) and most dense (between 1.036 g/
mL and 1.25 g/mL) of the plasma lipoproteins. They are heterogenous, comprising several subpopulations that vary in shape, size, composition and surface charge
HDL cholesterol The cholesterol content carried in all HDL particles
HDL particle concentration Number of HDL particles per plasma volume, expressed in µmol/L. HDL size The mass-weighted average diameter of the HDL particles in a particular
plasma sample
Large HDLs HDL particles with diameter between 7.3-8.2 nm Medium HDLs HDL particles with diameter between 8.2-8.8 nm
Myeloperoxidase Leukocyte-derived enzyme with a role in the immune system. Paraoxonase-1 Enzyme located on the surface of HDL is believed to protect against the
oxidation of low-density lipoprotein (LDL) and therefore to affect the risk of coronary artery disease
Small HDL HDL particles with diameter between 8.8-13 nm
Scavenger receptor class B1 (SR-BI) HDL receptor, mainly present in the liver, that promotes the selectively uptake of HDL cholesterol
participants
We have previously described other analyses within this prospective nested case-control study (14;15). Briefly, we excluded all individuals who reported a history of heart attack/stroke or use of lipid lowering drugs at the baseline clinic visit. Cases were individuals who developed fatal or non-fatal CAD during follow-up until November 2003. Controls were study participants who remained free of any cardiovascular disease during follow-up. We matched two controls
Cases (n=1138) Controls (n=2237) EPIC-Norfolk Prospective Population Study (n=25.663)
Nested case-control study (n=3375) Controls (n=1807) Excluded (total n =270)* •no LDL cholesterol (n=75) •no HDL cholesterol (n=77) •no Triglycerides (n=15) •no blood pressure (n=2) •no ApoA1 level (n=161) •no ApoB level (n=109) •no CRP level (n=30) •no NMR data (n=10) •no Waist hip ratio (n=2)
Excluded (total n =430)*
•no LDL cholesterol (n=112) •no HDL cholesterol (n=111) •no Triglycerides (n=31) •no blood pressure (n=5) •no ApoA1 level (n=289) •no ApoB level (n=171) •no CRP level (n=27) •no NMR data (n=7) •no Waist hip ratio (n=2)
Cases (n=868)
Excluded for lack of matching
control participants (n=46) Excluded for lack of matching case patients (n=406)
Analyzed Analyzed (n=822) (n=1401)
figure. Study flow diagram
ApoAI = apolipoprotein A-I; ApoB = apolipopotein B; CRP = C-reactive protein; EPIC = European Prospective Investigation into Cancer and Nutrition; HDL = high-density lipoprotein; LDL = low-density lipoprotein; NMR = nuclear magnetic resonance. *Individual case patients and control participants may
Chapt
er 3
36
to each case by age (within 5 years), sex and time of enrolment (within 3 months). The current analysis was performed on all participants who had a complete dataset available for baseline characteristics, apolipoproteins A-I and B, gradient gel electrophoresis-measured HDL size and lipoprotein nuclear magnetic resonance (NMR) spectroscopy.
We selected 1138 people who were apparently healthy at baseline but did develop CAD during follow-up. We aimed to select 2 controls for every case who were healthy at baseline and remained free of cardiovascular disease during follow-up. From the original dataset of 1138 cases and 2237 controls, 270 cases and 430 control patients were excluded because at least one value was missing for any of the parameters mentioned above. A total of 46 cases were excluded because they had no matching control patients, as were 406 control patients because they had no matching cases. The analyses are therefore based on a dataset containing 822 cases and 1401 control patients (243 cases with 1 matching control patient, 579 cases were matched with 2 control patients).
measurements
Data on smoking and alcohol consumption were obtained by health questionnaires at the baseline clinic visit. Physical activity was obtained with a Physical Activity questionnaire (EPAQ2) (13). In the original cohort smoking was classified into current cigarette smokers, former smok-ers and never smoksmok-ers. In the present study smoking was recoded into yes/no. Physical activity was classified in four categories: inactive, moderately active, moderately inactive and active. Use of large amounts of alcohol was defined as more than 21 units of alcohol/week. Diabetes mellitus and use of hormone replacement therapy was self-reported. Participants were asked about medical history with the question “Has a doctor ever told you that you have any of the following?”, followed by a number of choices including diabetes. Blood pressure was recorded by taking two measurements of diastolic and systolic blood pressure using the Accutor Sphyg-momanometer (Datascope, UK), after 3 min of resting.
Serum levels of total cholesterol, HDL cholesterol and triglycerides were measured on fresh samples with the RA 1000 (Bayer Diagnostics, Basingstoke, United Kingdom). Low-density lipo-protein (LDL) cholesterol levels were calculated with the Friedewald formula to closely approach current clinical procedures. Plasma concentrations of C-reactive protein were measured with a sandwich-type enzyme linked immuno sorbent assay as previously described (16). Serum levels of apolipoprotein A-I and apolipoprotein B were measured by rate immunonephelometry (Behring Nephelometer BNII, Marburg, Germany) with calibration traceable to the International Federation of Clinical Chemistry primary standards (17). Serum concentration of myeloperoxi-dase was measured by use of a commercially available enzyme-linked immuno sorbent assay (CardioMPO Test, Prognostix, Cleveland, Ohio, USA) (15). Paraoxonase-1 activity was measured as previously described (18). HDL size was measured by 4-30% nondenaturing polyacrylamide gradient gel electrophoresis (GGE) as previously described (19). Lipoprotein subclass particle concentrations and average size of particles were also measured by proton NMR spectroscopy
(LipoScience, Inc., NC, USA) as previously described (20). In brief, particle concentrations of lipoprotein subclasses of different size were obtained directly from the measured amplitudes of their spectroscopically distinct lipid methyl group NMR signals. HDL particle concentrations are expressed in µmoles of particles per liter (µmol/L). Summation of the HDL subclass levels provides total HDL particle concentration. The following 5 HDL subclasses were defined: H5 (10-13 nm, mean 11.5 nm), H4 (8.8-10 nm, mean 9.4 nm), H3 (8.2-8.8 nm, mean 8.5 nm), H2 (7.8-8.2 nm, mean 8.0 nm), H1 (7.3-7.7 nm, mean 7.5 nm). These diameter range estimates were based on size measurements of the isolated HDL subclass reference standards by GGE. The HDL subclasses H5, H4, H3, H2 and H1 are closely related to the GGE subclass designations 2b, 2a, 3a, 3b and 3c, respectively (21). To simplify data analysis and interpretation, we grouped these HDL subclasses as follows: Small HDL (H1 + H2), Medium HDL (H3), Large HDL (H4 + H5).(22) NMR-measured HDL size was calculated as the mass-weighted average diameter of the HDL particles in a particular plasma sample. The average HDL particle size is computed as the sum of the diameter of each subclass multiplied by its relative mass percentage as estimated from the amplitude of its measured NMR signal. It has been shown that the reproducibility of NMR lipoprotein profile assessments is very good, storing at -70°C and thawing has no significant influence on the quality of the HDL associated measurements (20). Samples were analyzed in random order to avoid systemic bias. Researchers and laboratory personnel were blinded to identifiable information, and could identify samples by number only.
statistical analysis
The hypotheses of the present study were generated prior to data analysis. Baseline character-istics were compared between cases and controls taking into account the matching. A mixed effect model was used for continuous variables and conditional logistic regression was used for categorical variables. Pearson’s correlation coefficients and corresponding p-values were cal-culated to assess associations between HDL-related parameters, metabolic and inflammatory covariates. Mean values of risk factors were calculated per quartile of HDL particle concentration and HDL size. Differences between categories were calculated by analysis of variance (ANOVA). In order to assess the strength of the associations between HDL particle concentration or HDL size and the risk of future CAD, we calculated odds ratios and corresponding 95% confidence intervals (95% CI) using conditional logistic regression, taking into account matching for sex, age and enrollment time and additionally adjusting for smoking. Regression analyses were also performed with additional adjustment for the inflammatory covariates myeloperoxidase, para-oxonase and C-reactive protein levels and for the metabolic covariates apolipoprotein B, and log-transformed triglycerides. The first quartile was used as reference group. P-values represent significance for linearity across the quartiles. Statistical analyses were performed using SPSS software (version 12.0.1, Chicago, Illinois). A P-value < 0.05 was considered to indicate statistical significance.
Chapt
er 3
38
role of the funding source
EPIC-Norfolk is supported by program grants from the Medical Research Council United King-dom and Cancer Research United KingKing-dom and with additional support from the European Union, Stroke Association, British Heart Foundation, Department of Health, Food Standards Agency and the Wellcome Trust. Part of the lipid and apolipoprotein measurements described in this article were funded by an educational grant from the Future Forum. LipoScience Inc. performed all nuclear magnetic resonance spectroscopy measurements. The funding sources had no role in study design, conduct, collection, management, analysis, and interpretation of the data, and preparation, review, or approval of the manuscript.
resulTs
A full dataset was available for 822 cases and 1401 controls; 579 cases were matched to two controls and 243 cases were matched to one control only. A total of 680 cases had a hospital admission during which 104 cases subsequently died. A total of 246 cases had a fatal event without a hospital admission. As expected, cases were more likely than controls to be smokers and have diabetes mellitus (Table 1). Cases had higher levels of LDL cholesterol, higher systolic blood pressure, higher body mass index and waist hip ratio, and lower HDL cholesterol levels. HDL particle concentration was higher in controls compared to cases (p<0.001). Concentra-tions of medium and small HDL particles did not differ between cases and controls. However, controls had an increased number of large HDL particles, suggesting that the reduction in HDL cholesterol levels observed in cases is attributable to a lower number of large HDL particles. Consistently, both NMR-measured and GGE-measured HDL size were lower in cases than con-trols. Controls were more physically active than cases. Use of hormone replacement therapy and intake of large amounts of alcohol (defined as more than 21 units of alcohol/week) did not differ between cases and controls.
relationships between hdl size, hdl particle concentration and cardiovascular
risk factors
NMR-measured HDL size and GGE-measured HDL size were strongly correlated (r= 0.78, p<0.001) (Appendix Table 1). In contrast, NMR-measured HDL size and GGE-measured HDL size were only weakly correlated with HDL particle concentration (r=0.08 and r=0.10, respectively). Both NMR-measured HDL size and GGE-measured HDL size were much more strongly correlated with large HDL particles than with medium or small HDL particles. As expected, triglycerides were inversely correlated with HDL cholesterol (r= -0.38, p<0.001). Apolipoprotein A-I was also strongly correlated with HDL particle concentration but this was entirely explained by a strong correlation with large HDL, whereas the correlation with medium and small HDL particles was weak.
Consistent with the low correlation between HDL particle concentration and HDL size, rela-tionships with other cardiovascular risk factors differed substantially between NMR-measured HDL size and HDL particle concentration (Table 2). As expected, HDL cholesterol levels differed significantly between HDL particle concentration quartiles and between HDL size quartiles (p<0.001 for each). Consistently, apolipoprotein A-I differed significantly between HDL particle
Table 1. Baseline characteristics
Controls Cases
(n = 1401) (n = 822) P Value
Men, % (n) 64 (523) 63 (878) Matched
Mean age (SD), y 65 ± 8 65 ± 8 Matched
Diabetes, % (n) 1.6 (23) 6.1 (49) < 0.001
Smoking, % (n) 8.1 (114) 16.4 (135) < 0.001
Mean waist-to-hip ratio (SD) 0.88 ± 0.08 0.90 ± 0.08 < 0.001 Hormone replacement therapy, % (n) 14.0 (73) 12.0 (36) 0.7 Alcohol intake > 21 units/week, % (n) 8.4 (117) 8.3 (68) 0.9 Systolic blood pressure (SD) mmHg 139 ± 18 144 ± 19 < 0.001 Diastolic blood pressure (SD) mmHg 84 ± 11 86 ± 12 < 0.001 Physical activity, % (n)
- inactive 31.4 (440) 43.1 (354) < 0.001
- moderately inactive 27.4 (384) 25.2 (207) - moderately active 23.6 (331) 17.9 (147)
- active 17.6 (246) 13.9 (114)
Total cholesterol (SD) mmolL 6.2 ± 1.1 6.4 ± 1.2 < 0.001
(mg/dL) (239.4±42.5) (247.1 ±46.3)
HDL cholesterol (SD) mmol/L 1.4 ± 0.4 1.3 ± 0.4 < 0.001
(mg/dL) (54.1±15.4) (50.1±15.4)
LDL cholesterol (SD) mmol/L 4.1 ± 1.0 4.3 ± 1.1 < 0.001
(mg/dL) (158.3±38.7) (166±42.5)
Triglycerides (IQR) mmol/L 1.6 [1.1-2.2] 1.8 [1.3-2.6] < 0.001
(mg/dL) (141.6 [97.3-194.6]) (159.3 [115.0-230.1]) Apolipoprotein A-I (SD) g/L 1.62 ± 0.29 1.55 ± 0.30 < 0.001 Apolipoprotein B (SD) g/L 1.28 ± 0.29 1.38 ± 0.32 < 0.001 C-reactive protein (SD) mg/L 3.3 ± 5.5 4.6 ± 7.4 < 0.001 Myeloperoxidase (SD) pmol/L 744 ± 560 822 ± 676 < 0.001 Paraoxonase (SD) U/L 62.9 ± 46 59.9 ± 44 0.1 HDL size (SD) nm (GGE-measured) 88.6 ± 4.2 88.1 ± 4.3 0.005 HDL size (SD) nm (NMR-measured) 8.9 ± 0.5 8.8 ± 0.5 < 0.001 HDL particle concentration (SD) µmol/L 33.9 ± 5 32.9 ±6 < 0.001 Large HDL (SD) µmol/L 6.2 ± 3.6 5.2 ± 3.4 < 0.001
Medium HDL (SD) µmol/L 3.2 ± 3.0 3.2 ± 3.0 0.9
Small HDL (SD) µmol/L 24.5 ± 4.7 24.5 ± 4.9 0.8
Data are presented as mean ± SD, percentage (n), or median [interquartile range]. Values may be based on fewer observations than the indicated number of subjects. HDL indicates high-density lipoprotein; LDL indicates low-density lipoprotein; Triglyceride levels were log-transformed before analysis. Ddifferences between cases and controls were calculated taking into account the matching procedure. A mixed effect model was used for continuous variables and conditional logistic regression was used for categorical variables
Chapt
er 3
40
concentration quartiles and between HDL size quartiles (p<0.001 for each). Triglycerides were inversely associated with HDL size (p<0.001), whereas they were positively associated with HDL particle concentration. Apolipoprotein B was inversely associated with HDL-size (p<0.001) but levels did not differ between HDL particle concentration quartiles. Waist-to-hip ratio had an inverse relation with HDL-size (p <0.001) but was not associated with HDL particle concentra-tion. Paraoxonase levels were associated with HDL particle concentration such that levels were higher in higher HDL particle concentration quartiles. Paraoxonase levels were not associated with HDL size quartiles. For analyses according to HDL size quartiles, adjustment for sex, age, smoking and BMI resulted in loss of statistical significance for waist-to-hip ratio and apolipo-protein B whereas for paraoxonase it resulted in statistically significant differences between quartiles. For HDL particle concentration, adjustment resulted in the opposite effects such that statistical significance was lost for paraoxonase whereas differences became statistically significant for apolipoprotein B and waist-to-hip ratio. In summary, participants with low HDL size had higher levels of metabolic syndrome features (triglycerides, apolipoprotein B, LDL size and waist-to-hip ratio) than those with higher HDL size. Similar analyses were performed after using GGE-measured HDL size or NMR-measured HDL size instead of NMR-measured HDL size. These analyses showed similar results (data not shown).
Table 2. Relationships between quartiles of HDL size and HDL particle concentration and cardiovascular
risk factors
HDL size quartile HDL particle concentration quartile, by
cardiovascular risk factor 1 2 3 4 P*
1 0.89 ± 0.14 1.00 ± 0.15 1.12 ± 0.17 1.55 ± 0.31 < 0.001 HDL cholesterol 2 0.97 ± 0.15 1.10 ± 0.18 1.28 ± 0.20 1.64 ± 0.34 < 0.001 (mmol/L) 3 1.10 ± 0.17 1.16 ± 0.17 1.42 ± 0.23 1.78 ± 0.31 < 0.001 4 1.22 ± 0.21 1.32 ± 0.25 1.56 ± 0.26 2.02 ± 0.38 < 0.001 P† < 0.001 < 0.001 < 0.001 < 0.001 Triglycerides 1 1.93 ± 0.71 1.78 ± 0.68 1.42 ± 0.60 1.05 ± 0.42 < 0.001 (mmol/L) 2 2.23 ± 0.74 1.95 ± 0.76 1.62 ± 0.64 1.26 ± 0.51 < 0.001 3 2.36 ± 0.85 2.11 ± 0.76 1.58 ± 0.62 1.26 ± 0.54 < 0.001 4 2.51 ± 0.89 2.39 ± 0.99 1.82 ± 0.71 1.42 ± 0.60 < 0.001 P† < 0.001 < 0.001 0.001 < 0.001 Apolipoprotein A-I 1 1.29 ± 0.19 1.34 ± 0.23 1.42 ± 0.16 1.67 ± 0.27 < 0.001 (g/L) 2 1.36 ± 0.13 1.44 ± 0.17 1.56± 0.19 1.78 ± 0.22 < 0.001 3 1.49 ± 0.17 1.54 ± 0.14 1.66 ± 0.18 1.88 ± 0.20 < 0.001 4 1.61 ± 0.17 1.69 ± 0.18 1.82 ± 0.20 2.09 ± 0.26 < 0.001 P† < 0.001 < 0.001 < 0.001 < 0.001
HDL size quartile HDL particle concentration quartile, by
cardiovascular risk factor 1 2 3 4 P*
Apolipoprotein B 1 1.47 ± 0.30 1.22 ± 0.23 1.21 ± 0.24 1.15 ± 0.29 < 0.001 (g/L) 2 1.46 ± 0.23 1.27 ± 0.30 1.29 ± 0.27 1.13 ± 0.26 < 0.001 3 1.46 ± 0.34 1.34 ± 0.27 1.26 ± 0.27 1.16 ± 0.25 < 0.001 4 1.47 ± 0.33 1.38 ± 0.26 1.25 ± 0.24 1.16 ± 0.25 < 0.001 P† 0.9 0.001 0.4 0.7 Small dense LDL 1 1362 ± 431 1020 ± 374 816 ± 360 581 ± 239 < 0.001 (nmol/L) 2 1417 ± 430 1022 ± 354 824 ± 295 599 ± 265 < 0.001 3 1377 ± 454 1132 ± 376 817 ± 287 617 ± 233 < 0.001 4 1350 ± 423 1165 ± 377 864 ± 285 698 ± 264 < 0.001 P† 0.8 0.02 0.6 0.007 C-reactive protein 1 4.60 ± 7.44 4.48 ± 5.57 3.89 ± 5.40 3.47 ± 6.45 0.006 (mg/L) 2 4.18 ± 5.53 3.51 ± 6.55 3.27 ± 4.40 2.36 ± 4.03 0.006 3 3.87 ± 6.80 3.05 ± 3.95 2.69 ± 3.99 2.73 ± 4.71 0.06 4 3.13 ± 5.50 2.98 ± 4.19 2.74 ± 3.45 2.52 ± 4.40 0.05 P† 0.4 0.4 0.5 0.9 Myeloperoxidase 1 874 ± 647 795 ± 546 898 ± 921 746 ± 530 0.5 (pmol/L) 2 805 ± 573 687 ± 630 684 ± 384 644 ± 415 0.2 3 743 ± 654 719 ± 586 702 ± 432 698 ± 597 0.9 4 564 ± 437 666 ± 444 686 ± 634 893 ± 396 0.3 P† 0.03 0.5 0.1 0.1 Paraoxonase 1 53 ± 40 48 ± 32 48 ± 32 50 ± 37 0.9 (u/L) 2 53 ± 39 58 ± 38 57 ± 38 62 ± 44 0.3 3 60 ± 44 66 ± 43 65 ± 45 76 ± 55 0.1 4 81 ± 64 64 ± 48 80 ± 55 76 ± 58 0.2 P† 0.007 0.06 0.001 0.001 Waist-to-hip ratio 1 0.93 ± 0.06 0.93 ± 0.07 0.89 ± 0.09 0.83 ± 0.07 < 0.001 2 0.93 ± 0.06 0.92 ± 0.07 0.89 ± 0.08 0.84 ± 0.08 < 0.001 3 0.92 ± 0.06 0.92 ± 0.07 0.87 ± 0.08 0.82 ± 0.08 < 0.001 4 0.91 ± 0.07 0.91 ± 0.08 0.87 ± 0.08 0.81 ± 0.08 < 0.001 P† 0.1 0.3 0.2 0.1
Data are given as mean ± SD per quartile. *P value for analysis of variance across HDL size quartiles; †P
value for analysis of variance across HDL particle concentration quartiles. The riskfactors are distributed on the basis of HDL size quartiles (horizontal) and HDL particle concentration quartiles (vertical). To convert HDL cholesterol values to mg/dL, divide by 0.0259. To convert triglycerides values to mg/dL divide by 0.0113.
Chapt
er 3
42
hdl size, hdl particle concentration and risk of future coronary heart disease
The unadjusted odds ratio for future CAD for people in the top quartile compared to those in the bottom quartile was 0.40 (95%CI 0.31 to 0.54, p<0.001) for HDL cholesterol, 0.60 (95%CI 0.47 to 0.79, p<0.001) for HDL particle concentration, 0.51 (95%CI 0.39 to 0.67, p<0.001) for NMR-measured HDL size, and 0.65 (95%CI 0.50 to 0.84, p<0.001) for GGE-measured HDL size. Table 3 shows the odds ratios for future CAD associated with increasing quartiles of HDL particle concentration and HDL size or large HDL entered in the same regression model. HDL particle concentration and NMR-measured HDL size were independently associated with CAD risk, with an odds ratio for individuals in the highest quartile compared with those in the lowest quartile of 0.60 (95%CI 0.46 to 0.79) and 0.52 (95%CI 0.40 to 0.69), respectively (p for linearity <0.001 for both) (model 1b). Analyses with HDL particle concentration and GGE-measured HDL size showed similar results, with respective odds ratios of 0.63 (95%CI 0.48 to 0.82) and 0.67 (95%CI 0.52 to 0.87) ( model 1). HDL particle concentration and large HDL particle concentra-tion were also independently associated with CAD risk, with odds ratios of 0.72 (95%CI 0.55 to 0.94, p=0.002) and 0.45 (95%CI 0.33 to 0.59, p<0.001) for individuals in the highest quartile compared with those in the lowest quartile (model 1). Adjustment for myeloperoxidase, para-oxonase and C-reactive protein levels (model 2) did not affect the risk estimates for HDL size or large HDL particle concentration. However, this adjustment did attenuate the risk estimate in the top HDL particle concentration quartile substantially (for the models with HDL size) or abolished the association completely (for the model with large HDL). By contrast, adjustment for the “metabolic” parameters triglycerides and apolipoprotein B had limited effect on the risk estimate for the HDL particle concentration top quartile but abolished the association for the top quartiles of HDL size or large HDL completely (model 3). Analyses with additional adjust-ment for alcohol intake, hormone replaceadjust-ment therapy and physical activity showed similar results (Appendix Table 2).
Table 3. Odds ratios for risk for future coronary heart disease
Model Parameter Odds Ratio (95% CI), by Quartile of HDL Parameter P Value
1 2 3 4 Model 1 HDL particle concentration 1.00 0.59 (0.46-0.76) 0.61 (0.47-0.79) 0.72 (0.55-0.94) 0.002 Large HDL 1.00 0.78 (0.61-1.01) 0.65 (0.50-0.85) 0.45 (0.33-0.59) < 0.001 Model 2 HDL particle concentration 1.00 0.61 (0.47-0.79) 0.66 (0.50-0.86) 0.78 (0.59-1.03) 0.12 Large HDL 1.00 0.78 (0.61-1.01) 0.66 (0.50-0.86) 0.46 (0.35-0.62) < 0.001 Model 3 HDL particle concentration 1.00 0.52 (0.41-0.69) 0.50 (0.38-0.66) 0.53 (0.40-0.72) < 0.001 Large HDL 1.00 0.99 (0.76-1.30) 0.99 (0.74-1.33) 0.85 (0.60-1.20) 0.38
discussion
Traditionally, the HDL fraction is quantified by measuring its cholesterol content. However, individuals with similar HDL cholesterol levels can differ substantially in terms of HDL size distribution and particle concentration. In this study, we measured HDL size and HDL par-ticle concentration among apparently healthy men and women and observed that these parameters were differentially associated with other cardiovascular risk factors, and were independently associated with the risk of future CAD. HDL size was strongly associated with well-known components of the metabolic syndrome, whereas HDL particle concentration was not. As a consequence, the relationship between HDL size and risk of future CAD risk was virtually abolished upon adjustment for metabolic parameters. The relationship between HDL particle concentration and CAD risk was independent of metabolic parameters. These findings indicate that HDL is a heterogeneous lipid fraction and that various HDL subpopulations are differentially associated with other cardiovascular risk factors and with cardiovascular risk.
Model Parameter Odds Ratio (95% CI), by Quartile of HDL Parameter P Value Model 1 HDL particle concentration 1.00 0.56 (0.44-0.73) 0.54 (0.42-0.70) 0.60 (0.46-0.79) < 0.001 NMR spectrometry-measured HDL size 1.00 0.89 (0.68-1.15) 0.76 (0.58-0.98) 0.52 (0.40-0.69) < 0.001 Model 2 HDL particle concentration 1.00 0.59 (0.46-0.76) 0.68 (0.55-0.86) 0.76 (0.60-0.98) 0.043 NMR spectrometry-measured HDL size 1.00 0.91 (0.70-1.18) 0.77 (0.59-1.00) 0.56 (0.42-0.74) < 0.001 Model 3 HDL particle concentration 1.00 0.52 (0.40-0.67) 0.48 (0.37-0.63) 0.50 (0.37-0.66) < 0.001 NMR spectrometry-measured HDL size 1.00 1.09 (0.83-1.43) 1.15 (0.86-1.53) 1.00 (0.71-1.39) 0.92 Model 1 HDL particle concentration 1.00 0.58 (0.45-0.74) 0.56 (0.44-0.73) 0.63 (0.48-0.82) < 0.001 GGE-measured HDL size 1.00 0.70 (0.54-0.90) 0.68 (0.52-0.88) 0.67 (0.52-0.87) 0.003 Model 2 HDL particle concentration 1.00 0.69 (0.53-0.90) 0.61 (0.47-0.79) 0.69 (0.53-0.91) 0.006 GGE-measured HDL size 1.00 0.69 (0.54-0.89) 0.68 (0.52-0.88) 0.60 (0.47-0.78) 0.002 Model 3 HDL particle concentration 1.00 0.53 (0.41-0.69) 0.50 (0.39-0.66) 0.51 (0.39-0.68) < 0.001 GGE-measured HDL size 1.00 0.84 (0.65-1.09) 0.92 (0.70-1.22) 1.04 (0.78-1.38) 0.66
GGE= gradient gel electrophoresis; HDL = high-density lipoprotein; NMR = nuclear magnetic resonance. Odds ratios were calculated by conditional logistic regression, taking into account matching for sex and age, and adjusting for smoking (model 1). Model 2: as model 1 with additional adjustment for myeloperoxidase, paraoxonase and C-reactive protein levels. Model 3: as model 1 with additional adjustment for apolipoprotein B, and log-transformed triglycerides. †P for linear trend.
Chapt er 3 44 a pp endix Table 1. P earson c or rela tion c oefficien ts Risk F ac tor HDL -C TG A poA -I A poB CRP MPO PON HDL siz e (GGE) HDL siz e (NMR) HDL -P Lar ge HDL M edium HDL Small HDL HDL -C 1.00 TG -0.38* 1.00 A poA -I 0.82* -0.22* 1.00 A poB -0.22* 0.43* -0.09 1.00 CRP -0.07 § 0.07 § -0.07 § 0.03 1.00 MPO -0.11* 0.02 -0.12* 0.00 0.26* 1.00 PON-1 0.16* -0.03 0.12* 0.008 -0.03 -0.05 1.00 HDL siz e ( GGE) 0.70* -0.34* 0.58* -0.29* -0.06 § -0.05 0.07 § 1.00 HDL siz e (NMR) 0.76* -0.47* 0.63* -0.38* -0.09 ¶ -0.07 § 0.07 § 0.78* 1.00 HDL -P 0.44* 0.17* 0.54* 0.01 -0.07 § -0.12* 0.23* 0.10* 0.08 § 1.00 Lar ge HDL 0.80* -0.43* 0.69* -0.38* -0.11* -0.07 § 0.12* 0.71* 0.88* 0.33* 1.00 M edium HDL 0.09* 0.25* 0.11* -0.09* 0.13* -0.04 0.11* -0.06 ¶ -0.11* 0.39* -0.12* 1.00 Small HDL -0.17* 0.36* 0.008 0.36* -0.08 § -0.05 0.09* -0.39* -0.52* 0.62* -0.32* -0.10* 1.00 WHR -0.45* 0.27* -0.43* 0.08 § 0.07 § 0.10* -0.07 § -0.40* -0.47* -0.14* -0.45* -0.02 0.20* HDL -C = high densit y lipopr ot ein cholest er ol; TG = tr igly cer ides; A poA -I = apolipopr ot ein A -I; A poB = apolipopr ot ein B; CRP = C-r eac tiv e pr ot
ein; PON = par
ao xonase; MPO = m yeloper oxidase; HDL -P = HDL par ticle c onc en tr ation; WHR = w aist hip r atio . T rigly cer ide lev els w er e log-tr ansf or med bef or e analy sis . * indica tes p<0.001, ¶ indica tes p=0.001; § indica tes p<0.05; † indica tes p=0.05.
hdl size, hdl particle concentration and cardiovascular risk
The published literature about the relationship between HDL subclasses and cardiovascular risk is limited to cross-sectional studies that have used different methods to quantify HDL subclasses and used different measures of cardiovascular risk (23). Several studies using GGE to measure HDL size have reported that CAD patients tend to have more smaller HDL particles and that large HDL particles may protect against the development of atherosclerosis (4;24;25) whereas others have reported that only small HDL3 particle were atheroprotective (26).
Studies that used NMR to measure HDL subclasses have usually analyzed HDL subclasses in concert, which makes the results difficult to interpret because HDL subclasses are highly
appendix Table 2. Odds ratios for risk for Coronary artery disease in patients who did not smoke, use
hormone replacement therapy, or drink large amounts of alcohol
Model Parameter Odds Ratio (95% CI), by Quartile P Value
1 2 3 4
Model 1 HDL particle concentration 1.00 0.58 (0.42-0.74) 0.62 (0.45-0.85) 0.65 (0.46-0.91) 0.03 Large HDL 1.00 0.87 (0.63-1.19) 0.64 (0.46-0.90) 0.49 (0.35-0.70) < 0.001 Model 2 HDL particle concentration 1.00 0.59 (0.43-0.81) 0.63 (0.46-0.88) 0.68 (0.48-0.96) 0.06 Large HDL 1.00 0.86 (0.63-1.18) 0.64 (0.45-0.90) 0.49 (0.34-0.70) < 0.001 Model 3 HDL particle concentration 1.00 0.52 (0.38-0.72) 0.50 (0.36-0.70) 0.46 (0.31-0.66) < 0.001 Large HDL 1.00 1.12 (0.80-1.56) 1.03 (0.71-1.51) 0.97 (0.64-1.48) 0.8 Model 1 HDL particle concentration 1.00 0.55 (0.40-0.75) 0.55 (0.40-0.76) 0.55 (0.39-0.77) 0.001
NMR spectrometry-measured
HDL size 1.00 0.89 (0.64-1.25) 0.69 (0.49-0.96) 0.55 (0.39-0.77) < 0.001 Model 2 HDL particle concentration 1.00 0.56 (0.41-0.77) 0.57 (0.41-0.79) 0.58 (0.41-0.82) 0.004
NMR spectrometry-measured
HDL size 1.00 0.91 (0.65-1.27) 0.69 (0.49-0.98) 0.56 (0.40-0.79) < 0.001 Model 3 HDL particle concentration 1.00 0.52 (0.38-0.72) 0.50 (0.36-0.70) 0.44 (0.31-0.64) < 0.001
NMR spectrometry-measured
HDL size 1.00 1.10 (0.78-1.57) 1.06 (0.73-1.54) 1.03 (0.69-1.55) 0.9 Model 1 HDL particle concentration 1.00 0.55 (0.41-0.75) 0.55 (0.41-0.76) 0.58 (0.42-0.81) 0.001 GGE-measured HDL size 1.00 0.79 (0.57-1.08) 0.63 (0.45-0.87) 0.71 (0.52-0.98) 0.02 Model 2 HDL particle concentration 1.00 0.57 (0.42-0.77) 0.57 (0.47-0.79) 0.61 (0.43-0.85) 0.004 GGE-measured HDL size 1.00 0.79 (0.57-1.08) 0.63 (0.46-0.88) 0.72 (0.52-0.99) 0.03 Model 3 HDL particle concentration 1.00 0.53 (0.38-0.73) 0.51 (0.37-0.71) 0.46 (0.32-0.57) < 0.001 GGE-measured HDL size 1.00 0.95 (0.68-1.31) 0.88 (0.62-1.25) 1.08 (0.76-1.55) 0.7
GGE= gradient gel electrophoresis; HDL = high-density lipoprotein; NMR = nuclear magnetic resonance. Odds ratios were calculated by conditional logistic regression, taking into account matching for sex and age, and adjusting for smoking (model 1). Model 2: as model 1 with additional adjustment for myeloperoxidase, paraoxonase and C-reactive protein levels. Model 3: as model 1 with additional adjustment for apolipoprotein B, and log-transformed triglycerides. †P for linear trend.
Chapt
er 3
46
correlated. In a study among patients undergoing coronary angiography, higher levels of small HDL particles were associated with more severe atherosclerosis(27), whereas a study among patients with type 1 diabetes observed no relationship between HDL subclasses and carotid intima-media thickness (28). The only study investigating a relationship with clinical events was a substudy of the Veterans Affairs High-Density Lipoprotein Intervention Trial, which investi-gated the effect of gemfibrozil on CAD risk (29). In this trial HDL particle concentrations were associated with cardiovascular risk independent of traditional lipid levels. In the current study we decided not to analyze these highly related HDL subclasses, but focused our analyses on HDL particle concentration and either HDL size or large HDL. We observed that HDL cholesterol, HDL size and HDL particle concentration were each individually associated with decreased risk of CAD. When HDL size and HDL particle concentration were analyzed conjointly, both were independently and inversely associated with cardiovascular risk. For HDL size, this relationship was confounded by characteristics of the metabolic syndrome because adjustment for these variables resulted in abolition of the relationship between HDL size and CAD risk.
hdl size and hdl particle concentration versus metabolic dysregulation
We observed that HDL size was significantly associated with several features of the metabolic syndrome including high plasma levels of triglycerides, apolipoprotein B, small LDL particles and higher waist-to-hip ratio, whereas HDL particle concentration was completely independent of these variables. Our results are consistent with previous studies which have shown reduced HDL size in people with diabetes mellitus (30), insulin resistance (31;32), hypertriglyceridemia (33), and parameters associated with the metabolic syndrome (34). In other words, the low HDL cholesterol levels observed among people with metabolic features was explained by reduced HDL size, not by a low HDL particle concentration. In addition, we observed that both reduced HDL size and low HDL particle concentration were associated with cardiovascular risk. How-ever, upon correction for metabolic parameters (apolipoprotein B and triglyceride levels), HDL size was no longer associated with risk of future CAD. In contrast, HDL particle concentration, which is numerically unaffected in the metabolic syndrome, does retain its predictive value after correction for metabolic parameters. These data suggest that the association between HDL size and cardiovascular risk is confounded by metabolic dysregulation.
hdl size and hdl particle concentration versus inflammation
Whereas HDL particle concentration was only weakly correlated with metabolic parameters, it did show an association with inflammatory and oxidative markers, including myeloperoxidase and paraoxonase. These surface-bound enzymes may explain part of the atheroprotective effect of HDL particles. The association between paraoxonase levels and HDL particle con-centration is interesting and suggests that higher particle concon-centration may facilitate more surface-bound molecules. This observation is not consistent with in vitro studies showing that the difference in shape and size of several HDL subpopulations is critical for binding of these
anti-oxidative enzymes (7;8). In contrast we did not observe a strong association between HDL size and paraoxonase levels. In fact, several lines of evidence have supported a direct anti-inflammatory effect of apolipoprotein A-I. Infusion of minute dosages of apolipoprotein A-I had potent anti-atherogenic effects in rabbit models,(35;36) apolipoprotein A-I or its Milano variant had potent anti-inflammatory effects in monocytic cell lines(37;38), and the anti-inflammatory effect of HDL was shown to be due to its apolipoproteins, not to protein-free phospholipid (39).
limitations
A number of issues have to be taken into account when interpreting the results of our study. Measurements were performed in non-fasting blood samples which could have affected the relation of triglycerides with the various HDL subclasses. However, HDL subclass concentra-tions are not significantly altered by freezing or in the postprandial state (20). Diurnal variation, variation over time, and differences in the time since the last meal could have affected our results. However, the relation of non-fasting triglycerides with other lipids and CAD risk in the EPIC-Norfolk cohort is not different than expected for fasting triglycerides. Moreover, in the Western world, people live under constant postprandial conditions, and studies on the associa-tions between lipids, lipoproteins, and CAD risk may be more physiologically relevant under non-fasting conditions, as has recently been demonstrated (40-42). Second, CAD events were scored through death certification and hospital admission data, which may have resulted in under-ascertainment or misclassification. Previous validation studies in this cohort, however, indicate high specificity of such case ascertainment (14). In addition, any misclassification leads to underestimation of true associations and therefore does not negate our results. Third, because only non-fasting samples were obtained in the EPIC-Norfolk cohort, we were not able to define the metabolic syndrome, which requires the measurements of parameters in the fasting state. As a consequence, we were not able to analyze our results in the context of the metabolic syndrome, but only in relation to separate metabolic parameters. Finally, we adjusted for confounding factors, but residual confounding by imperfectly measured or unmeasured confounders cannot be excluded. However this is a common limitation of a non-randomized study.
conclusion
In summary, we observed that HDL size and HDL particle concentration were independently associated with other cardiovascular risk factors and with cardiovascular risk. For HDL size, these associations appeared to be explained by features of the metabolic syndrome. By contrast, the relationships for HDL particle concentration were not affected by adjustment for metabolic parameters. These results suggest that part of the relationship between HDL cholesterol levels and CAD risk is explained by metabolic parameters, and that this part can be
Chapt
er 3
48
quantified by measuring HDL size or the concentration of large HDL particles. The other part of the association between HDL cholesterol levels and CAD risk is independent of metabolic parameters, is weakly associated with inflammatory parameters, and can be quantified by the concentration of HDL particles. Future research into the relationship between HDL cholesterol and cardiovascular risk should take these aspects into account. In addition, these findings may have implications for the development of therapeutic strategies aimed at modifying HDL metabolism. Our results suggest that HDL-raising strategies that primarily affect HDL-size may have different effects on cardiovascular risk than those affecting HDL particle concentration.
references
1 Wilson PW, Garrison RJ, Castelli WP, Feinleib M, McNamara PM, Kannel WB. Prevalence of coronary heart disease in the Framingham Offspring Study: role of lipoprotein cholesterols. Am J Cardiol. 1980;46:649-54. 2 Rader DJ. Molecular regulation of HDL metabolism and function: implications for novel therapies. J Clin
Invest. 2006;116:3090-3100.
3 Thuahnai ST, Lund-Katz S, Dhanasekaran P, de la Llera-Moya M, Connelly MA, Williams DL et al. Scavenger Receptor Class B Type I-mediated Cholesteryl Ester-selective Uptake and Efflux of Unesterified Cholesterol: influence of high density lipoprotein size and structure. J Biol Chem. 2004;279:12448-55.
4 Watanabe H, Soderlund S, Soro-Paavonen A, Hiukka A, Leinonen E, Alagona C et al. Decreased High-Density Lipoprotein (HDL) Particle Size, Pre{beta}-, and Large HDL Subspecies Concentration in Finnish Low-HDL Families: Relationship With Intima-Media Thickness. Arterioscler Thromb Vasc Biol. 2006;26:897-902. 5 Yancey PG, Kawashiri Ma, Moore R, Glick JM, Williams DL, Connelly MA et al. In vivo modulation of HDL
phos-pholipid has opposing effects on SR-BI- and ABCA1-mediated cholesterol efflux. J Lipid Res. 2004;45:337-46. 6 Vaisar T, Pennathur S, Green PS, Gharib SA, Hoofnagle AN, Cheung MC et al. Shotgun proteomics implicates
protease inhibition and complement activation in the antiinflammatory properties of HDL. J Clin Invest. 2007;117:746-56.
7 Josse D, Ebel C, Stroebel D, Fontaine A, Borges F, Echalier A et al. Oligomeric States of the Detergent-solubilized Human Serum Paraoxonase (PON1). J Biol Chem. 2002;277:33386-97.
8 Kontush A, Chantepie S, Chapman MJ. Small, Dense HDL Particles Exert Potent Protection of Atherogenic LDL Against Oxidative Stress. Arterioscler Thromb Vasc Biol. 2003;23:1881-88.
9 Navab M, Anantharamaiah GM, Reddy ST, Van Lenten BJ, Ansell BJ, Fogelman AM. - Mechanisms of disease: proatherogenic HDL--an evolving field. Nat Clin Pract Endocrinol Metab. 2006;2:504-11.
10 Pussinen PJ, Jauhiainen M, Vilkuna-Rautiainen T, Sundvall J, Vesanen M, Mattila K et al. Periodontitis decreases the antiatherogenic potency of high density lipoprotein. J Lipid Res. 2004;45:139-47.
11 van der Steeg WA, Holme I, Boekholdt SM, Larsen ML, Lindahl C, Stroes ESG et al. High-Density Lipoprotein Cholesterol, High-Density Lipoprotein Particle Size, and Apolipoprotein A-I: Significance for Cardiovascular Risk: The IDEAL and EPIC-Norfolk Studies. Journal of the American College of Cardiology. 2008;51:634-42. 12 Barter PJ, Caulfield M, Eriksson M, Grundy SM, Kastelein JJP, Komajda M et al. Effects of Torcetrapib in Patients
at High Risk for Coronary Events. N Engl J Med. 2007;357:2109-22.
13 Day N, Oakes S, Luben R, Khaw KT, Bingham S, Welch A et al. - EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br J Cancer. 1999;80 Suppl. 1:95-103. 14 Boekholdt SM, Peters RJ, Day NE, Luben R, Bingham SA, Wareham NJ et al. - Macrophage migration inhibitory
factor and the risk of myocardial infarction or death due to coronary artery disease in adults without prior myocardial infarction or stroke: the EPIC-Norfolk Prospective Population study. Am J Med. 2004;117:390-7. 15 Meuwese MC, Stroes ESG, Hazen SL, van Miert JN, Kuivenhoven JA, Schaub RG et al. Serum Myeloperoxidase
Levels Are Associated With the Future Risk of Coronary Artery Disease in Apparently Healthy Individuals: The EPIC-Norfolk Prospective Population Study. Journal of the American College of Cardiology. 2007;50:159-65. 16 Bruins P, Velthuis Ht, Yazdanbakhsh AP, Jansen PGM, van Hardevelt FWJ, de Beaumont EMFH et al. Activation
of the Complement System During and After Cardiopulmonary Bypass Surgery : Postsurgery Activation Involves C-Reactive Protein and Is Associated With Postoperative Arrhythmia. Circulation. 1997;96:3542-48. 17 Albers JJ, Marcovina SM, Kennedy H. International Federation of Clinical Chemistry standardization project
for measurements of apolipoproteins A-I and B. II. Evaluation and selection of candidate reference materials. Clin Chem. 1992;38:658-62.
18 Mackness B, Durrington P, McElduff P, Yarnell J, Azam N, Watt M et al. Low Paraoxonase Activity Predicts Coronary Events in the Caerphilly Prospective Study. Circulation. 2003;107:2775-79.
19 Perusse M, Pascot A, Despres JP, Couillard C, Lamarche B. A new method for HDL particle sizing by polyacryl-amide gradient gel electrophoresis using whole plasma. J Lipid Res. 2001;42:1331-34.
20 Jeyarajah EJ, Cromwell WC, Otvos JD. Lipoprotein Particle Analysis by Nuclear Magnetic Resonance Spec-troscopy. Clinics in Laboratory Medicine. 2006;26:847-70.
Chapt
er 3
50
21 Grundy SM, Vega GL, Otvos JD, Rainwater DL, Cohen JC. Hepatic lipase activity influences high density lipoprotein subclass distribution in normotriglyceridemic men: genetic and pharmacological evidence. J Lipid Res. 1999;40:229-34.
22 Colhoun HM, Otvos JD, Rubens MB, Taskinen MR, Underwood SR, Fuller JH. Lipoprotein Subclasses and Par-ticle Sizes and Their Relationship With Coronary Artery Calcification in Men and Women With and Without Type 1 Diabetes. Diabetes. 2002;51:1949-56.
23 Florentin M, Liberopoulos EN, Wierzbicki AS, Mikhailidis DP. Multiple actions of high-density lipoprotein. Curr Opin Cardiol. 2008;23:370-378.
24 Cheung MC, Brown BG, Wolf AC, Albers JJ. Altered particle size distribution of apolipoprotein A-I-containing lipoproteins in subjects with coronary artery disease. J Lipid Res. 1991;32:383-94.
25 Johansson J, Carlson LA, Landou C, Hamsten A. High density lipoproteins and coronary atherosclerosis. A strong inverse relation with the largest particles is confined to normotriglyceridemic patients. Arterioscler Thromb. 1991;11:174-82.
26 Syvanne M, Nieminen MS, Frick MH, Kauma H, Majahalme S, Virtanen V et al. Associations Between Lipopro-teins and the Progression of Coronary and Vein-Graft Atherosclerosis in a Controlled Trial With Gemfibrozil in Men With Low Baseline Levels of HDL Cholesterol. Circulation. 1998;98:1993-99.
27 Freedman DS, Otvos JD, Jeyarajah EJ, Barboriak JJ, Anderson AJ, Walker JA. Relation of Lipoprotein Sub-classes as Measured by Proton Nuclear Magnetic Resonance Spectroscopy to Coronary Artery Disease. Arterioscler Thromb Vasc Biol. 1998;18:1046-53.
28 Lyons TJ, Jenkins AJ, Zheng D, Klein RL, Otvos JD, Yu Y et al. - Nuclear magnetic resonance-determined lipo-protein subclass profile in the DCCT/EDIC cohort: associations with carotid intima-media thickness. Diabet med. 2006;23:955-66.
29 Otvos JD, Collins D, Freedman DS, Shalaurova I, Schaefer EJ, McNamara JR et al. Low-Density Lipoprotein and High-Density Lipoprotein Particle Subclasses Predict Coronary Events and Are Favorably Changed by Gemfibrozil Therapy in the Veterans Affairs High-Density Lipoprotein Intervention Trial. Circulation. 2006;113:1556-63.
30 Festa A, Williams K, Hanley AJG, Otvos JD, Goff DC, Wagenknecht LE et al. Nuclear Magnetic Resonance Lipoprotein Abnormalities in Prediabetic Subjects in the Insulin Resistance Atherosclerosis Study. Circula-tion. 2005;111:3465-72.
31 Garvey WT, Kwon S, Zheng D, Shaughnessy S, Wallace P, Hutto A et al. Effects of Insulin Resistance and Type 2 Diabetes on Lipoprotein Subclass Particle Size and Concentration Determined by Nuclear Magnetic Resonance. Diabetes. 2003;52:453-62.
32 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;49:285-92.
33 Syvanne M, Ahola M, Lahdenpera S, Kahri J, Kuusi T, Virtanen KS et al. High density lipoprotein subfractions in non-insulin-dependent diabetes mellitus and coronary artery disease. J Lipid Res. 1995;36:573-82. 34 Pascot A, Lemieux I, Prud’homme D, Tremblay A, Nadeau A, Couillard C et al. Reduced HDL particle size as an
additional feature of the atherogenic dyslipidemia of abdominal obesity. J Lipid Res. 2001;42:2007-14. 35 Nicholls SJ, Cutri B, Worthley SG, Kee P, Rye KA, Bao S et al. Impact of Short-Term Administration of
High-Density Lipoproteins and Atorvastatin on Atherosclerosis in Rabbits. Arterioscler Thromb Vasc Biol. 2005;25:2416-21.
36 Nicholls SJ, Dusting GJ, Cutri B, Bao S, Drummond GR, Rye KA et al. Reconstituted High-Density Lipoproteins Inhibit the Acute Pro-Oxidant and Proinflammatory Vascular Changes Induced by a Periarterial Collar in Normocholesterolemic Rabbits. Circulation. 2005;111:1543-50.
37 Shah PK, Yano J, Reyes O, Chyu KY, Kaul S, Bisgaier CL et al. High-Dose Recombinant Apolipoprotein A-IMilano Mobilizes Tissue Cholesterol and Rapidly Reduces Plaque Lipid and Macrophage Content in Apolipoprotein E-Deficient Mice : Potential Implications for Acute Plaque Stabilization. Circulation. 2001;103:3047-50.
38 Wang L, Sharifi BG, Pan T, Song L, Yukht A, Shah PK. Bone Marrow Transplantation Shows Superior Atheropro-tective Effects of Gene Therapy With Apolipoprotein A-I Milano Compared With Wild-Type Apolipoprotein A-I in Hyperlipidemic Mice. Journal of the American College of Cardiology. 2006;48:1459-68.
39 van der Poll T, Braxton CC, Coyle SM, Boermeester MA, Wang JC, Jansen PM et al. Effect of hypertriglyceride-mia on endotoxin responsiveness in humans. Infect Immun. 1995;63:3396-400.
40 Bansal S, Buring JE, Rifai N, Mora S, Sacks FM, Ridker PM. Fasting Compared With Nonfasting Triglycerides and Risk of Cardiovascular Events in Women. JAMA. 2007;298:309-16.
41 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.
42 Nordestgaard BG, Benn M, Schnohr P, Tybjaerg-Hansen A. Nonfasting Triglycerides and Risk of Myocardial Infarction, Ischemic Heart Disease, and Death in Men and Women. JAMA. 2007;298:299-308.
