Diabetes alters the association between high-density lipoprotein subfractions and carotid intima-media thickness: The Brazilian Longitudinal Study of Adult Health (ELSA-Brasil)

Abstract

Introduction:

High-density lipoprotein cholesterol comprises a group of heterogeneous subfractions that might have differential effects on atherosclerosis. Moreover, prior investigations suggest that the presence of diabetes (T2D) modifies the impact of some subfractions on atherosclerosis. In this study, we aimed to evaluate the association between high-density lipoprotein cholesterol subfractions and carotid intima-media thickness in the baseline assessment of the Brazilian Longitudinal Study of Adult Health participants from the São Paulo investigation centre.

Methods:

We evaluated 3930 individuals between 35 and 74 years without previous cardiovascular disease not using lipid-lowering drugs. High-density lipoprotein cholesterol subfractions (HDL₂-C and HDL₃-C) were measured by vertical ultracentrifugation (vertical auto profile). The relationship between each high-density lipoprotein cholesterol subfraction and carotid intima-media thickness was analysed by multiple linear regression models.

Results:

Total high-density lipoprotein cholesterol, as well as HDL₂-C and HDL₃-C, was negatively associated with carotid intima-media thickness after adjustment for demographic data (all p < 0.001) and traditional risk factors (all p < 0.05). When stratified by T2D status, the HDL₂-C/HDL₃-C ratio showed a negative association with carotid intima-media thickness in participants with T2D (p = 0.032), even after fully controlling for confounding variables, including total high-density lipoprotein cholesterol.

Conclusion:

HDL₂-C, HDL₃-C and HDL₂/HDL₃-C ratio are inversely associated with carotid intima-media thickness after adjustment for traditional risk factors. Association of the HDL₂-C/HDL₃-C ratio is modified by the presence of diabetes, being more pronounced in diabetic individuals.

Keywords

Cholesterol/metabolism high-density lipoprotein high-density lipoprotein/metabolism diabetes atherosclerosis vertical auto profile zonal ultracentrifugation

Introduction

High-density lipoproteins (HDLs) comprise a group of heterogeneous particles, and high-density lipoprotein cholesterol (HDL-C) has a well-established inverse association with the risk of cardiovascular (CV) events in prospective longitudinal cohorts followed throughout the world.^1,2 However, genetic polymorphisms that increase HDL-C do not correlate with CV risk when excluded variants additionally linked to low-density lipoprotein cholesterol (LDL-C) or triglycerides.^3–5 Also, pharmacologic interventions with cholesteryl ester transfer protein (CETP) inhibitors to raise HDL-C levels did not reduce CV outcomes and even might be harmful.^6–9 This lack of association has motivated studies of HDL-C subfractions to more comprehensively evaluate the mechanisms linking HDL-related pathways with atherogenesis. Among the methods used to separate those subfractions, the vertical auto profile (VAP) separates HDL-C into two subfractions: HDL₂-C (larger and less dense) and HDL₃-C (smaller and denser).¹⁰

Carotid intima-media thickness (cIMT) is an important marker for subclinical vascular disease and is considered an atherosclerosis surrogate.¹¹ Increases in cIMT are accompanied by a higher incidence of CV morbidity including coronary artery disease, myocardial infarction^12–14 and stroke.¹⁵ Although low HDL-C has been associated with increased cIMT,¹⁶ this relationship is not always observed,¹⁷ suggesting that total HDL may not entirely explain HDL’s antiatherogenic properties. Also, the association of HDL-C subfractions and cIMT remains disputed since different results are obtained with the use of different separation methods.^18–20 This relationship is even more intricate because diabetic individuals have lower HDL-C levels and a lower HDL₂-C/HDL₃-C ratio, which seems to modify the association of HDL-C with cIMT in comparison with the non-diabetic population. It is also well documented that the functionality of HDL species from patients with T2D is impaired and even dysfunctional.^21,22

In this study, we evaluated the association between HDL-C subfractions and cIMT in a Brazilian sample without clinical manifestations of any cerebrovascular disease. In addition, we evaluated whether the presence of diabetes modifies this association.

Methods

Sample

The Brazilian Longitudinal Study of Adult Health (ELSA-Brasil) design has been previously described.^23,24 Briefly, we enrolled 15,015 civil servants from six Brazilian cities, between August 2008 and December 2010. We included those who underwent HDL-C measurement by the VAP method and cIMT assessment – that were all participants from São Paulo centre. Exclusion criteria for this analysis were the cohort restrictions (pregnancy or recent childbirth; cognitive deficit or impaired communication; in retired servants, living outside of a study centre’s area) in addition to individuals with prior CV disease (myocardial infarction, stroke, heart failure and coronary revascularization) and participants using lipid-lowering drugs.

HDL-C and subfractions

Blood collection was performed in patients after nocturnal fasting. The samples were centrifuged at the sites and stored in tubes at −80°C.²⁵ The HDL-C, HDL₂-C and HDL₃-C subfractions were obtained by the VAP method through an apparatus, the vertical rotor of which holds the tube perpendicular to the base of the equipment.¹⁰

cIMT measurement

The measurement protocol was published previously.^26–28 cIMT measurements were performed using a Toshiba (Aplio XG^™) with a 7.5-MHz linear transducer. Intima-media thickness (IMT) was measured in the outer wall of a predefined carotid segment of 1 cm in length from 1 cm below carotid bifurcation, during three cardiac cycles. All participating centres obtained the carotid images, and the images were forwarded to the centralized core reading centre in São Paulo. Images were classified as valid if the following was obtained for the left and right sides: (1) the anatomic guides for the common carotid arteries, (2) interfaces between the lumen and the far vessel wall and (3) interfaces between the media and the adventitia layers of the far vessel wall. We used MIA^™ software to standardize the reading and interpretation of carotid scans.

Statistical analysis

We presented continuous variables as descriptive statistics with mean and standard deviation for normal distribution and, for non-normal distributions, with medians and quartiles. Categorical variables were shown in absolute and relative frequency. Baseline characteristics table was exposed according to total HDL-C quartiles to show different HDL-C subfraction levels across the total HDL-C groups. For correlation measurement, we used Pearson. The association between HDL-C subfractions and variables was determined by linear regression. We standardized HDL-C, HDL₂-C, HDL₃-C and HDL₂-C/HDL₃-C ratio, which were included as the main explanatory variables in separate multiple linear models. Models including HDL₂-C, HDL₃-C or HDL₂-C/HDL₃-C ratio were also further adjusted for total HDL-C (as exposed in Table 2 – model 4 and Table 3 – model 2).

As we observed significant interaction terms between HDL₂-C/HDL₃-C ratio and T2D, we performed additional analyses stratified by diabetes status. The models were adjusted for race/ethnicity, age and sex, smoking, alcohol, physical activity, LDL-C, systolic blood pressure, waist circumference, fasting glucose, log-transformed triglycerides, estimated glomerular filtration rate and antihypertensive use. Statistical significance was defined as p < 0.05. Analyses were performed with Stata version 14.0 (StataCorp, USA).

Results

We included 3930 participants (1048 exclusions from São Paulo centre) with a mean age of 50.1 ± 8.4 years and 1793 (45.6%) males. The mean HDL-C was 1.42 ± 0.37 mmol/L and mean cIMT was 0.796 ± 0.195 mm (Table 1).

Table 1.

Demographic and clinical characteristics of the study sample according to HDL-C quartiles.

	Mean (±SD or IQR) or n (%)
	n = 3930	Q1	Q2	Q3	Q4
Age (years)	50.1 (±8.4)	49.5 (±8.3)	49.7 (±8.4)	49.8 (±8.4)	51.7 (±8.4)
Race (%)
White	2269 (57.7)	636 (59.1)	515 (55.5)	539 (56.4)	579 (59.5)
Brown	860 (21.9)	253 (23.5)	233 (25.1)	197 (20.6)	177 (18.2)
Black	554 (14.1)	130 (12.1)	124 (13.4)	151 (15.8)	149 (15.3)
Asian	152 (3.9)	22 (2)	33 (3.6)	45 (4.7)	52 (5.3)
Indigenous	46 (1.2)	20 (1.8)	13 (1.4)	8 (0.8)	5 (0.5)
Other	49 (1.3)	14 (1.3)	9 (1)	15 (1.6)	11 (1.1)
Men (%)	1793 (45.6)	762 (70.9)	463 (50)	341 (35.7)	227 (23.3)
Hypertension (%)	1018 (25.9)	324 (30.1)	262 (28.2)	219 (23)	213 (21.9)
BMI (kg/m²)	27 (±4.75)	28.3 (±4.64)	27.4 (±4.7)	26.6 (±4.8)	25.6 (±4.5)
Waist circumference (cm)	89 (±12.3)	94 (±11.2)	90 (±11.5)	86 (±11.8)	83 (±11.6)
Smoking (%)
Current	657 (16.7)	214 (19.9)	157 (16.9)	162 (17)	124 (12.7)
Former smokers	1155 (29.4)	328 (30.5)	284 (30.6)	253 (26.5)	290 (29.8)
Never	2118 (53.9)	533 (49.6)	486 (52.4)	540 (56.5)	559 (57.5)
SBP (mmHg)	118.9 (±16.2)	122 (±16.3)	119.4 (±16)	117.4 (±16.1)	116.4 (±15.9)
DBP (mmHg)	75 (±10.8)	77.3 (±10.7)	75.6 (±10.5)	73.9 (±10.5)	73 (±10.8)
Diabetes (%)	641 (16.3)	233 (21.7)	161 (17.4)	129 (13.5)	118 (12.1)
Dyslipidaemia (%)*	1989 (50.6)	538 (50)	508 (54.8)	498 (52.1)	445 (45.7)
Triglycerides (mmol/L)	1.24 (0.88–1.78)	1.77 (1.24–2.40)	1.35 (0.98–1.86)	1.10 (0.84–1.50)	0.95 (0.72–1.23)
Fasting plasma glucose (mmol/L)	6.06 (±1.48)	6.36 (±1.8)	6.1 (±1.5)	5.9 (±1.3)	5.8 (±1)
Total cholesterol (mmol/L)	5.68 (±1.08)	5.37 (±1.05)	5.63 (±1.07)	5.72 (±1.04)	6 (±1.08)
LDL-C (mmol/L)**	3.42 (±0.88)	3.41 (±0.88)	3.51 (±.90)	3.46 (±0.85)	3.32 (±0.89)
HDL-C (mmol/L)	1.42(±0.37)	1.02 (±0.11)	1.28 (±0.06)	1.50 (±0.07)	1.93 (±0.27)
HDL₂ -C (mmol/L)	0.39 (±0.17)	0.22 (±0.05)	0.31 (±0.05)	0.41 (±0.06)	0.61 (±0.16)
HDL₃ -C (mmol/L)	1.04 (±0.22)	0.80 (±0.08)	0.96 (±0.05)	1.09 (±0.07)	1.32 (±0.16)
Mean cIMT (mm)	0.796 (±0.195)	0.83 (±0.21)	0.80 (±0.20)	0.78 (±0.18)	0.77 (±0.17)

SD: standard deviation; IQR: interquartile range; cIMT: carotid artery intima-media thickness; HDL-C: high-density lipoprotein cholesterol; SBP: systolic blood pressure; DBP: diastolic blood pressure; LDL-C: low-density lipoprotein cholesterol; BMI: body mass index.

Dyslipidaemia was defined as high triglycerides (TGs; ⩾150 mg/dL), low HDL-C [<40 (men) and <50 (women) mg/dL] or high LDL-C (⩾130 mg/dL or ever taking lipid-lowering agents).

All p < 0.001, except **p = 0.006 and *p = 0.03.

As previous publications analysing HDL-C subfractions,^18,29,30 we observed a high correlation between total HDL-C and both subfractions HDL₂-C (p = 0.94) and HDL₃-C (p = 0.96) and a lower correlation with the HDL₂-C/HDL₃-C ratio (p = 0.73). In crude models, an inverse association of HDL-C, HDL₂-C, HDL₃-C and HDL₂-C/HDL₃-C ratio with mean cIMT was noted. These results remained essentially unchanged after adjusting for demographic characteristics (model 1) and CV risk factors (except in HDL₂-C/HDL₃-C ratio – model 3). However, when subfractions were adjusted for HDL-C, they were no longer associated with lower mean cIMT (Table 2).

Table 2.

Association of total HDL-C, HDL₂-C and HDL₃-C subfractions and HDL₂-C/HDL₃-C ratio with cIMT.

	Model	β	95% CI	p
HDL-C	Univariate	−0.021	(–0.0276, −0.0155)	<0.001
	Model 1	−0.026	(–0.0320, −0.0204)	<0.001
	Model 2	−0.024	(–0.0298, −0.0180)	<0.001
	Model 3	−0.008	(–0.0142, −0.0016)	0.014
HDL₂-C	Univariate	−0.024	(–0.0301, −0.0180)	<0.001
	Model 1	−0.028	(–0.0336, −0.02205)	<0.001
	Model 2	−0.025	(–0.0316, −0.0198)	<0.001
	Model 3	−0.006	(–0.0128, −0.0001)	0.04
	Model 4	0.005	(–0.0107, +0.0215)	0.51
HDL₃-C	Univariate	−0.018	(–0.024, −0.0118)	<0.001
	Model 1	−0.022	(–0.0282, −0.0168)	<0.001
	Model 2	−0.02	(–0.0260, −0.0143)	<0.001
	Model 3	−0.01	(–0.0151, −0.0032)	0.003
	Model 4	−0.008	(–0.0281, +0.0120)	0.43
HDL₂-C/HDL₃-C ratio	Univariate	−0.026	(–0.0325, −0.0204)	<0.001
	Model 1	−0.027	(–0.0328, −0.0213)	<0.001
	Model 2	−0.026	(–0.0315, −0.0198)	<0.001
	Model 3	−0.006	(–0.0119, +0.0004)	0.069
	Model 4	−0.014	(–0.0093, +0.0065)	0.728

HDL-C: high-density lipoprotein cholesterol; CI: confidence interval; cIMT: carotid intima-media thickness.

Linear regression model: model 1 – adjusted for race/ethnicity, age and sex; model 2 – adjusted for variables in model 1 and also smoking, alcohol and physical activity; model 3 – adjusted for variables in model 2 and also low-density lipoprotein cholesterol, systolic blood pressure, waist circumference, fasting glucose, body mass index, log-transformed triglycerides, EGFR and antihypertensive use; and model 4 – adjusted for variables in model 3 and also total HDL-C.

We found an interaction between T2D status and the HDL₂-C/HDL₃-C ratio. Furthermore, when stratified by the presence of T2D, the HDL₂-C/HDL₃-C ratio was inversely associated with cIMT in participants with diabetes in full models (Figure 1), but not in participants without diabetes. We also found that HDL₃-C was negatively associated with cIMT in non-T2D individuals and not in participants with T2D. In this case, a p value of borderline significance (p = 0.051) was observed. After further adjustment for HDL-C levels, only the association between the HDL₂-C/HDL₃-C ratio and cIMT in individuals with diabetes remained significant (p = 0.032; Figure 2 and Table 3).

Figure 1.

Detailed enrolment flow diagram.

Figure 2.

Association between the HDL₂-C/HDL₃-C ratio and cIMT in univariate analysis.

Table 3.

Association of total HDL-C, HDL₂-C and HDL₃-C subfractions and the HDL₂-C/HDL₃-C ratio with cIMT when stratified by diabetes mellitus status.

	Non-diabetes			Diabetes			Interaction
	β	CI	p	β	CI	p	p
Model 1
HDL-C	−0.007	(–0.0138, −0.0005)	0.03	−0.011	(–0.0302, +0.0071)	0.22	0.77
HDL₂-C	−0.005	(–0.0116, +0.0017)	0.11	−0.016	(–0.0355, +0.0034)	0.11	0.35
HDL₃-C	−0.008	(–0.0144, −0.0015)	0.01	−0.007	(–0.0248, +0.0103)	0.41	0.051
HDL₂-C/HDL₃-C ratio	−0.003	(–0.0091, +0.0039)	0.44	−0.023	(–0.0409, −0.0046)	0.01	0.036
Model 2
HDL₂-C	0.011	(–0.0063, +0.0274)	0.22	−0.032	(–0.0814, +0.0176)	0.21	0.38
HDL₃-C	−0.015	(–0.0360, +0.0059)	0.16	0.038	(–0.0230, +0.0992)	0.22	0.77
HDL₂-C/HDL₃-C ratio	0.003	(–0.0052, +0.0115)	0.46	−0.025	(–0.0474, −0.0022)	0.03	0.032

cIMT: carotid intima-media thickness; HDL-C: high-density lipoprotein cholesterol; CI: confidence interval.

Linear regression adjusted for race/ethnicity, age and sex, smoking, alcohol, physical activity, low-density lipoprotein cholesterol, systolic blood pressure, waist circumference, fasting glucose, body mass index, log-transformed triglycerides, antihypertensive use, EGFR and HDL-C (except the own variable HDL-C).

Discussion

Our results demonstrate that both HDL₂-C and HDL₃-C are inversely associated with cIMT even after controlling for demographic and CV risk factors, but they are not independent predictors. This association is particularly robust with HDL₂-C/HDL₃-C ratio for diabetics, whereas such findings were noted in non-diabetics only with HDL₃-C.

Previous studies have observed controversial results in the relationship between HDL-C subfractions and cIMT. While some studies have found no association of HDL₂-C or HDL₃-C with cIMT,^19,20 a recent publication has shown that, in addition to HDL-C, the HDL₂-C subfraction is associated with a lower cIMT, suggesting that this subfraction is atheroprotective.¹⁸ In contrast to these data, a Finnish study³¹ demonstrated that HDL-C, HDL₂-C and HDL₃-C subfractions and also the HDL₂-C/HDL₃-C ratio are negatively associated with cIMT, a finding that was reproduced in our study. These discordant results might be explained by ethnic differences in the population, differences in baseline characteristics (younger, more obese people), and participants not being treated with lipid-lowering therapies.

The lipid panel is known to be altered in about 50% of the individuals with diabetes, either due to high triglycerides or LDL-C levels or by low HDL-C.³² Diabetic dyslipidaemia is related to the impact of insulin resistance on metabolic pathways and systemic changes related to diabetes, such as renal disease. HDL-C was the only lipid parameter to be independently associated with carotid IMT in a cohort of young adults with type 2 diabetes³³ and, in another trial, low-HDL-C elevation secondary to statin use was related to the IMT decrease.^34,35

Diabetes leads to impaired reverse cholesterol transport through both reduced HDL concentrations and HDL dysfunction. Thus, a more pronounced inverse association of the HDL₂-C/HDL₃-C ratio with cIMT could be explained in some pathways: inhibition of lipoprotein lipase, where less apolipoprotein A-I (apo A-I) is liberated from chylomicrons, decreasing HDL formation in serum; ABCA1 down-regulation on the surface of visceral adipocytes, also reducing HDL formation;³⁶ lecithin–cholesterol acyltransferase (LCAT) dysfunction,³⁷ which lowers cholesterol esterification – decreases the maturation of HDL₃ to HDL₂; and finally, increases in CETP activity, which promote triglyceride enrichment of HDL particles, rendering them better substrates for lipolysis and catabolism by hepatic lipase.

Indeed, studies have shown that, in patients with diabetes, raising HDL-C levels by drugs other than statin is not accompanied by improvement or reverse cholesterol transport.³⁸

Thus, considering that the infusion of reconstituted HDL in patients with coronary heart disease (CHD) led to changes in inflammatory markers, cholesterol efflux and plaque conformation,³⁹ perhaps this fact calls attention to need to clinically evaluate the functionality of HDL instead of its level individually, especially in the diabetic milieu. Moreover, although CETP inhibitors are known to increase asymmetrically HDL-C subfractions,⁴⁰ as well as statin use, which alters the HDL composition and HDL₃-C levels,⁴¹ it is unclear whether the variability of the HDL₂-C/HDL₃-C ratio may result in CV benefit.

Strengths of this study include a large cohort, a method of direct subfraction measurement and the use of a computer-aided protocol for cIMT measurements that provided reliable and accurate cIMT data. Our study also has some limitations. As a cross-sectional study, the described associations do not confirm causation. Also, cIMT was used instead of plaque burden, and total plaque area may be more representative of atherosclerosis than cIMT. However, the cIMT may be more reversible, which may explain the weaker association with CV disease events than the plaque.²⁸ We are also unable to evaluate the antiatherogenic effects of HDL and its subfractions directly mechanistically.

In conclusion, in a unicentric, cross-sectional study of a Brazilian sample, HDL-C and its subfractions HDL₂-C and HDL₃-C are inversely associated with cIMT. In participants with diabetes, the negative association is limited to the HDL₂-C/HDL₃-C ratio, which has a statistically significant association even when adjusted for HDL-C.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The ELSA-Brasil baseline study was supported by the Brazilian Ministry of Health (Science and Technology Department) and the Brazilian Ministry of Science and Technology (Financiadora de Estudos e Projetos and CNPq National Research Council; Grant Nos 01 06 0010.00 RS, 01 06 0212.00 BA, 01 06 0300.00 ES, 01 06 0278.00 MG, 01 06 0115.00 SP and 01 06 0071.00 RJ).

ORCID iDs

Giuliano Generoso

Itamar S Santos

References

Lewington

Whitlock

Clarke

et al . Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55,000 vascular deaths. Lancet 2007; 370: 1829–1839.

Di Angelantonio

Sarwar

Perry

et al . Major lipids, apolipoproteins, and risk of vascular disease. JAMA 2009; 302: 1993–2000.

Holmes

Asselbergs

Palmer

et al . Mendelian randomization of blood lipids for coronary heart disease. Eur Heart J 2015; 36: 539–550.

Laurinavicius

Santos

et al . Extremely elevated HDL-cholesterol levels are not associated with increased carotid intima-media thickness: data from ELSA Brasil. J Clin Lipidol 2016; 10: 898–904.e1.

Nordestgaard

Tybjærg-Hansen

Genetic determinants of LDL, lipoprotein(a), triglyceride-rich lipoproteins and HDL: concordance and discordance with cardiovascular disease risk. Curr Opin Lipidol 2011; 22: 113–122.

The AIM-HIGH Investigators. Niacin in patients with low HDL cholesterol levels receiving intensive statin therapy. N Engl J Med 2011; 365: 2255–2267.

Kaur

Pandey

Negi

et al . Effect of HDL-raising drugs on cardiovascular outcomes: a systematic review and meta-regression. PLoS ONE 2014; 9: e94585.

Barter

Caulfield

Eriksson

et al . Effects of torcetrapib in patients at high risk for coronary events. N Engl J Med 2007; 357: 2109–2122.

The HPS2-THRIVE Collaborative Group. Effects of extended-release niacin with laropiprant in high-risk patients. N Engl J Med 2014; 371: 203–212.

10.

Kulkarni

KR.

Cholesterol profile measurement by vertical auto profile method. Clin Lab Med 2006; 26: 787–802.

11.

Peters

SAE

den Ruijter

Bots

et al . Improvements in risk stratification for the occurrence of cardiovascular disease by imaging subclinical atherosclerosis: a systematic review. Heart 2012; 98: 177–184.

12.

Plichart

Celermajer

Zureik

et al . Carotid intima-media thickness in plaque-free site, carotid plaques and coronary heart disease risk prediction in older adults. The Three-City Study. Atherosclerosis 2011; 219: 917–924.

13.

Lorenz

Polak

Kavousi

et al . Carotid intima-media thickness progression to predict cardiovascular events in the general population (the PROG-IMT collaborative project): a meta-analysis of individual participant data. Lancet 2012; 379: 2053–2062.

14.

Lorenz

Markus

Bots

et al . Prediction of clinical cardiovascular events with carotid intima-media thickness: a systematic review and meta-analysis. Circulation 2007; 115: 459–467.

15.

O’Leary

Polak

Kronmal

et al . Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. Cardiovascular Health Study Collaborative Research Group. N Engl J Med 1999; 340: 14–22.

16.

Touboul

P-J

Labreuche

Bruckert

et al . HDL-C, triglycerides and carotid IMT: a meta-analysis of 21,000 patients with automated edge detection IMT measurement. Atherosclerosis 2014; 232: 65–71.

17.

Yan

Tian

et al . Association of lipoprotein subclasses and carotid intima-media thickness in the Chinese population. Clin Lipidol 2014; 9: 407–415.

18.

Tiozzo

Gardener

Hudson

et al . High-density lipoprotein subfractions and carotid plaque: the Northern Manhattan Study. Atherosclerosis 2014; 237: 163–168.

19.

Notsu

Yano

Takeda

et al . Association of high-density lipoprotein subclasses with carotid intima-media thickness: Shimane CoHRE study. J Atheroscler Thromb 2017; 25, https://www.jstage.jst.go.jp/article/jat/advpub/0/advpub_38844/_article

20.

Maeda

Nakanishi

Yoneda

et al . Associations between small dense LDL, HDL subfractions (HDL2, HDL3) and risk of atherosclerosis in Japanese-Americans. J Atheroscler Thromb 2012; 19: 444–452.

21.

Srivastava

RAK

. Dysfunctional HDL in diabetes mellitus and its role in the pathogenesis of cardiovascular disease. Mol Cell Biochem 2018; 440: 167–187.

22.

Farbstein

Levy

AP.

HDL dysfunction in diabetes: causes and possible treatments. Expert Rev Cardiovasc Ther 2012; 10: 353–361.

23.

Aquino

EML

Barreto

Bensenor

et al . Brazilian Longitudinal Study of Adult Health (ELSA-Brasil): objectives and design. Am J Epidemiol 2012; 175: 315–324.

24.

Schmidt

Duncan

Mill

et al . Cohort profile: Longitudinal Study of Adult Health (ELSA-Brasil). Int J Epidemiol 2015; 44: 68–75.

25.

Fedeli

Vidigal

Leite

et al . Logistica de coleta e transporte de material biologico e organizacao do laboratorio central no ELSA-Brasil. Rev Saúde Pública 2013; 47: 63–71.

26.

Mill

Pinto

Griep

et al . Afericoes e exames clinicos realizados nos participantes do ELSA-Brasil. Rev Saúde Pública 2013; 47: 54–62.

27.

Santos

Bittencourt

Oliveira

IRS

et al . Carotid intima–media thickness value distributions in The Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). Atherosclerosis 2014; 237: 227–235.

28.

Santos

Alencar

Rundek

et al . Low impact of traditional risk factors on carotid intima-media thickness: the ELSA-Brasil cohort. Arterioscler Thromb Vasc Biol 2015; 35(9): 2054–2059.

29.

Kim

Burt

Rosenthal

et al . HDL-3 is a superior predictor of carotid artery disease in a case-control cohort of 1725 participants. J Am Heart Assoc 2014; 3: e000902.

30.

Lamarche

Moorjani

Cantin

et al . Associations of HDL2 and HDL3 subfractions with ischemic heart disease in men: prospective results from the Quebec Cardiovascular Study. Arterioscler Thromb Vasc Biol 1997; 17: 1098–1105.

31.

Alagona

Soro

Ylitalo

et al . A low high density lipoprotein (HDL) level is associated with carotid artery intima-media thickness in asymptomatic members of low HDL families. Atherosclerosis 2002; 165: 309–316.

32.

Schofield

Liu

Rao-Balakrishna

et al . Diabetes dyslipidemia. Diabetes Ther 2016; 7: 203–219.

33.

Shah

Urbina

Khoury

et al . Lipids and lipoprotein ratios: contribution to carotid intima media thickness in adolescents and young adults with type 2 diabetes mellitus. J Clin Lipidol 2013; 7: 441–445.

34.

Ishigaki

Kono

Katagiri

et al . Elevation of HDL-C in response to statin treatment is involved in the regression of carotid atherosclerosis. J Atheroscler Thromb 2014; 21: 1055–1065.

35.

Davidson

Meyer

Haffner

et al . Increased high-density lipoprotein cholesterol predicts the pioglitazone-mediated reduction of carotid intima-media thickness progression in patients with type 2 diabetes mellitus. Circulation 2008; 117: 2123–2130.

36.

McGillicuddy

Reilly

Rader

DJ.

Adipose modulation of high-density lipoprotein cholesterol: implications for obesity, high-density lipoprotein metabolism, and cardiovascular disease. Circulation 2011; 124: 1602–1605.

37.

Nakhjavani

Esteghamati

Esfahanian

et al . HbA1c negatively correlates with LCAT activity in type 2 diabetes. Diabetes Res Clin Pract 2008; 81: 38–41.

38.

Fadini

Iori

Marescotti

et al . Insulin-induced glucose control improves HDL cholesterol levels but not reverse cholesterol transport in type 2 diabetic patients. Atherosclerosis 2014; 235: 415–417.

39.

Shaw

Bobik

Murphy

et al . Infusion of reconstituted high-density lipoprotein leads to acute changes in human atherosclerotic plaque. Circ Res 2008; 103: 1084–1091.

40.

Mabuchi

Nohara

Inazu

Cholesteryl ester transfer protein (CETP) deficiency and CETP inhibitors. Mol Cells 2014; 37: 777–784.

41.

de Souza Zago

Tanus-Santos

Gardin Danelon

et al . Chemical modification of high density lipoprotein subfractions – HDL2 and HDL3 – after use of atorvastatin. Int J Clin Pharmacol Ther 2014; 52: 277–283.