Association between carotid intima-media thickness and adiponectin in participants without diabetes or cardiovascular disease of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil)

Abstract

Objective

The study assessed the association of adiponectin concentration with carotid intima-media thickness (CIMT) in middle-aged participants of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil) without diabetes or cardiovascular disease.

Design

Cross-sectional analyses.

Methods

A sample of 687 individuals (35–54 years old) without diabetes or cardiovascular disease was stratified into two categories according to CIMT (< or ≥ 75th percentile). Traditional risk factors, C-reactive protein and adiponectin values were compared between categories by Student’s t-test and frequencies by chi-square test. In linear regression models, associations of CIMT with adiponectin, adjusted for adiposity, blood pressure, C-reactive protein and homeostasis model assessment–insulin resistance were tested. Mean CIMT values were compared across quartiles of adiponectin concentrations using analysis of variance.

Results

Three hundred and forty-one individuals (49.6%) were women and 130 (19.0%) had three traditional cardiovascular risk factors. Those with elevated CIMT (21.8%) had greater mean values of body mass index (26.2(3.8) vs. 27.7(4.0)kg/m², p < 0.001), waist circumference (86.9(10.1) vs. 90.1(10.8) cm, p = 0.001), systolic blood pressure (116.2(13.6) vs.121.2(16.1) mmHg, p < 0.001), homeostasis model assessment index (1.4(0.9–2.4) vs. 1.8(1.1–2.9), p = 0.011), C-reactive protein (1.2 (0.6–2.6) vs. 1.4(0.8–3.2) mg/l, p = 0.054) and adiponectin (9.9 (6.0–14.7) vs. 8.9 (5.3–13.8) µg/ml, p = 0.002) levels than the counterpart, while plasma glucose and lipids were not different between groups. In the adjusted model, blood pressure (directly) and adiponectin (inversely) persisted associated with high CIMT. Mean CIMT was greater in the first quartile of adiponectin when compared with the other three quartiles (p = 0.019).

Conclusions

Lower adiponectin levels together with higher blood pressure were independently associated with elevated CIMT. Adiponectin concentration may be an independent marker of early structural damage in individuals at low-to-moderate cardiovascular risk.

Adiponectin, insulin resistance, IMT, cardiovascular risk factors

Introduction

Cardiovascular disease (CVD) is the leading cause of disability adjusted life years in the developed and developing world.¹ Traditional cardiovascular risk factors, such as obesity, hypertension and disturbances of glucose and lipids metabolism, are commonly clustered, and insulin resistance has been considered a relevant pathophysiological link. However, a number of cardiovascular events occur in the absence of these risk factors,² indicating that there might be a substantial need for the identification of novel risk factors.

Carotid intima-media thickness (CIMT) has been considered as a non-invasive surrogate measurement of subclinical atherosclerosis,³ able to predict major cardiovascular events. Although there is no consensual definition regarding a cutoff for CIMT, values above the 75th percentile for a given age, sex and race have been accepted as abnormal.³ It is relevant to notice that the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil) found that demographic and traditional cardiovascular risk factors did not explain the majority of common carotid artery intima-media thickness, which also reinforces the need to study novel risk factors.⁴

Inflammatory adipose tissue-derived cytokines detected in circulation have been proposed as novel biomarkers of cardiometabolic risk.⁵ These biomarkers are increased in obesity and may play a role in atherogenesis.⁵ When the level of high-sensitivity C-reactive protein was added to the Framingham cardiovascular risk score the prediction of outcomes was shown to improve in women and men.⁶ There is a debate on the role of other circulating biomarkers of inflammation, insulin resistance and endothelium dysfunction as potential early markers of CVD.⁷ Adiponectin – an adipose tissue-produced hormone – participates in energy homeostasis and enhances insulin sensitivity. An inverse association of adiponectinemia with obesity, inflammation, insulin resistance, type 2 diabetes and atherosclerosis has been reported.^8–10 Adiponectin has been associated with CIMT, mainly in high cardiovascular risk populations.¹⁰ However, whether adiponectin concentration could be a potential early marker of atherogenesis in low-to-moderate cardiovascular risk populations is open to question.

The ELSA-Brasil is a prospective cohort study of civil servants in universities in Brazil.^11,12 An objective of the ELSA-Brasil is to examine surrogates of CVD and novel biomarkers of atherosclerosis.¹³ The ELSA-Brasil study group has already reported normal ranges for CIMT according to sex, age and race in a large Brazilian sample and an association of CIMT with body adiposity was found.^12,4 In the present analysis, we hypothesized that, independently of adiposity, the adiponectin concentration could help to identify subclinical atherosclerosis in ELSA-Brasil participants. Therefore, we assessed the association of adiponectin with CIMT in individuals without diabetes or CVD.

Methods

This is a cross-sectional analysis conducted in a subset of the ELSA-Brasil participants from the University of São Paulo, in Sao Paulo city. Details of the main study, registered at clinicaltrials.com as NCT02320461, have been described elsewhere.¹¹ For the purpose of the present analysis, after excluding diabetes and cardiovascular disease, a sample of 1000 individuals, aged 35–54 years, was randomly selected from the participants of São Paulo,¹¹ keeping the same proportions of sexes within the age groups (35–44 and 45–54 years) as found in the main study. Two individuals were excluded from the final sample due to insufficient aliquots frozen for the analysis of novel biomarkers. This subset of 998 participants underwent laboratory determinations and CIMT to address the purpose of the present analysis. To be included individuals should be at a low or moderate cardiovascular risk, defined by the presence of up to three major risk factors as follows: waist circumference ≥ 94 cm for men and ≥ 80 cm for women; systolic blood pressure ≥ 130 mmHg or diastolic blood pressure ≥ 85 mmHg (or antihypertensive treatment); fasting plasma glucose ≥ 100 mg/dl and < 125 mg/dl in the absence of antidiabetic agents; triglyceride ≥ 150 mg/dl (or specific treatment); high-density lipoprotein (HDL) cholesterol < 40 mg/dl for men and < 50 mg/dl for women (or specific treatment). A total of 687 individuals met the inclusion criteria. The institutional ethics committee approved the study and written consent was obtained from participants.

Participants were interviewed using standardized questionnaires and then scheduled for physical examination and laboratory tests.¹¹ Body weight and height were measured using calibrated electronic scales and a fixed rigid stadiometer, while individuals wore light clothing without shoes. Body mass index (BMI) was calculated as weight (kilograms) divided by squared height (meters). Waist circumference was measured with an inextensible tape according to the WHO technique. Blood pressure was taken three times after a 5-min rest in the sitting position. The mean of the second and third measurements defined systolic and diastolic blood pressure and these were used in the analyses. Participants underwent fasting blood sampling and a 2-h 75-g oral glucose tolerance test. American Diabetes Association criteria were used to define categories of glucose tolerance.¹⁴

Plasma glucose was measured by the hexokinase method (ADVIA Chemistry; Siemens, Deerfield, Illinois, USA). Total cholesterol was determined by the cholesterol oxidase method, enzymatic colorimetric, HDL-cholesterol by homogeneous colorimetric, without precipitation, triglycerides by enzymatic colorimetric (ADVIA Chemistry; Siemens, Deerfield, Illinois, USA); low-density lipoprotein (LDL)-cholesterol was calculated by the Friedewald equation. When triglycerides concentration was greater than 400 mg/dl, the LDL-cholesterol concentration was directly measured.¹¹ Aliquots were frozen at −80℃ for further determinations of hormones and inflammatory markers.

Insulin was determined by enzyme-linked immunoenzymatic assay (ELISA; Siemens, Tarrytown, USA). Homeostasis model assessment was used to assess insulin resistance (HOMA-IR).¹⁵ ELISA kits were also used for the determination of high-sensitivity C-reactive protein by immunochemistry (Dade Behring, Siemens, Marburg, Germany) and adiponectin (Enzo Life Sciences, Farmingdale, New York, USA). Intra-assay coefficients of variation ranged from 1.8 to 7.2 and inter-assay coefficients from 0.9 to 9.1.

The CIMT technique, electrocardiogram gated, used in the ELSA-Brasil was previously reported;^11,12 a Toshiba (Aplio XG™) with a 7.5 MHz linear transducer was employed. CIMT was measured in the outer wall of a pre-defined carotid segment of 1 cm in length from 1 cm below carotid bifurcation during three cardiac cycles and considered as valid if the images clearly visualize both the left and right sides of the anatomic guides for the common carotid arteries, interfaces between the lumen and the vessel far wall and 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 as previously described. For the purpose of the present study, CIMT was defined as the average of the mean left CIMT and the mean right CIMT measures. Values above the 75th percentile for a given age, sex and race, according to previously published quantile-regression models, were considered ‘high’ CIMT.^3,12

Statistical analysis

Data were expressed as means and standard deviations (SDs) or as medians and interquartile intervals. When distributions of variables were skewed they were log-transformed before analysis to achieve normality. Continuous data were compared by Student’s t test for independent samples. Categorical variables were defined as follows: normal or high CIMT; weight excess for BMI ≥ 25 kg/m²; central obesity for waist circumference ≥ 94 cm for men and ≥ 80 cm for women; elevated blood pressure (BP) for systolic BP ≥ 130 or diastolic BP ≥ 85 mmHg or antihypertensive treatment; hypertriglyceridemia for triglycerides ≥ 150 mg/dl; low HDL-cholesterol for levels < 40 mg/dl for men and < 50 mg/dl for women; low-grade inflammation for C-reactive protein ≥ 3.0 mg/l. Frequencies were compared by chi-square test and 95% confidence intervals (95% CIs) were shown. In binary logistic regression models, the association between high CIMT and adiponectin concentration, adjusted for major cardiovascular risk factors, HOMA-IR and C-reactive protein was tested. Those variables whose mean values differed between categories with a p-value < 0.20 were included in the model. Analysis of variance was used to test differences in CIMT according to adiponectin quartiles. Sensitivity analysis were performed considering nonsteroidal anti-inflammatory drugs (n = 1), acetylsalicylic acid (n = 7) and lipid-lowering agents (n = 1). As results did not change, data from all the participants were shown. All statistical analyses were performed using the Statistical Package for Social Sciences, V. 19.0 for Windows (SPSS Inc., Chicago, Illinois, USA). A p-value < 0.05 was considered significant.

Results

From 687 individuals 341 (49.6%) were women, 142 (20.7%) had elevated blood pressure, 546 (79.5%) prediabetes, 189 (27.5%) hypertriglyceridemia, 131 (19.0%) low HDL-cholesterol and 157 (22.7%) central obesity. Three hundred and thirty-nine individuals (49.0%) showed only one of these risk factors, 218 (32.0%) showed two and 130 (19.0%) three of them. Those with high CIMT (21.8%) had higher mean values of BMI, waist circumference, systolic and diastolic BP, HOMA-IR index and C-reactive protein and lower levels of adiponectin than the normal CIMT group, but not of lipids and plasma glucose (Table 1). Also, elevated BP, weight excess, central obesity and low-grade inflammation were more frequent in the high CIMT group. On the other hand, frequencies of prediabetes and lipid abnormalities were not different between groups. Frequencies of smoking and family history of CVD or diabetes did not differ either (17.9 vs. 20.0%, p = 0.553 and 51.4 vs. 51.3%, p = 0.989 respectively).

Table 1.

Characteristics of the study sample according to CIMT categories.

	Normal CIMT n = 537	CIMT > P75 n = 150	p-value
Women, %	262 (48.8)	79 (52.7)	0.401
Age, years	45.9 (4.8)	45.4 (5.0)	0.210
Body mass index, kg/m2	26.2 (3.8)	27.7 (4.0)	<0.001
Waist circumference, cm	86.9 (10.1)	90.1 (10.8)	0.001
Systolic blood pressure, mmHg	116.2 (13.6)	121.2 (16.1)	<0.001
Diastolic blood pressure, mmHg	74.7 (9.7)	77.6 (11.5)	0.002
Fasting plasma glucose, mg/dl	103.9 (7.4)	104.2 (7.1)	0.661
Total cholesterol, mg/dl	211.1 (38.0)	214.2 (37.5)	0.376
LDL-cholesterol, mg/dl	131.0 (32.3)	133.9 (32.7)	0.323
HDL-cholesterol, mg/dl	54.5 (12.6)	53.6 (12.5)	0.441
Triglycerides, mg/dl	128.9 (72.9)	135.4 (71.9)	0.329
HOMA-IRa	1.4 (0.9–2.4)	1.8 (1.1–2.9)	0.011
C-reactive proteina, mg/l	1.2 (0.6–2.6)	1.4 (0.8–3.2)	0.054
Adiponectina, µg/ml	9.9 (6.0–14.7)	8.9 (5.3–13.8)	0.002
Frequencies, %
Weight excess	313 (58.4)	115 (76.6)	<0.001
Central obesity	112 (20.9)	45 (30.0)	0.018
Elevated blood pressure	94 (17.5)	48 (21.8)	<0.001
Hypertriglyceridemia	144 (26.8)	48 (30.0)	0.440
Low HDL-cholesterol	105 (19.6)	26 (17.3)	0.541
Pre-diabetes	425 (79.1)	181 (80.7)	0.683
Low-grade inflammation	127 (20.1)	49 (28.0)	0.024

Values are mean (standard deviation) or median (interquartile range). Except for number of women and categorical variables (%): weight excess: body mass index ≥ 25. Central obesity: waist circumference ≥ 94 cm for men and ≥ 80 cm for women. Hypertension: systolic blood pressure ≥ 130 or diastolic blood pressure ≥ 85 mmHg (or antihypertensive treatment). Hypertriglyceridemia: triglycerides ≥ 150 mg/dl. Low HDL-cholesterol: HDL < 40 mg/dl for men and < 50 mg/dl for women. Pre-diabetes: fasting plasma glucose ≥ 100 mg/dl and < 126 mg/dl and 2-h plasma glucose ≥ 140 mg/dl and < 200 mg/dl. Low grade inflammation: C-reactive protein ≥ 3.0 mg/l.-value obtained by chi-square test or Student’s t test.

a

Log-transformed variables for analysis.

CIMT: carotid intima-media thickness; P75: 75th percentile; LDL: low-density lipoprotein; HDL: high-density lipoprotein; HOMA–IR: homeostasis model assessment–insulin resistance

In crude analysis, waist circumference, BP, HOMA, C-reactive protein and adiponectin were significantly associated with abnormal CIMT. In the adjusted model, BP (directly) and adiponectin (inversely) persisted associated with high CIMT. When BMI was included in the model instead of waist circumference, the results did not change (Table 2).

Table 2.

Final model of logistic regression for the association of carotid intima-media thickness > 75th percentile (dependent variable) with adiponectin concentration.

	Crude		Adjusted
	OR	95% CI	OR	95% CI
Body mass index	1.095	1.047–1.146
Waist circumference	1.028	1.012–1.048	1.015	0.992–1.039
Systolic blood pressure	1.024	1.011–1.036	1.020	1.005–1.035
HOMA-IR	1.154	1.030–1.293	1.098	0.942–1.280
C-reactive proteina	1.213	0.997–1.475	1.081	0.870–1.344
Adiponectina	0.785	0.669–0.920	0.782	0.632–0.969

The final adjusted model included the variables presented in this table.

a

Log-transformed values.

OR: odds ratio; CI: confidence interval; HOMA–IR: homeostasis model assessment–insulin resistance

The means of these variables were compared according to quartiles of adiponectin. Mean CIMT (95% CI) was greater in the first quartile of adiponectin when compared with the other three quartiles (0.58 (0.13) vs. 0.55 (0.10) vs. 0.55 (0.09) vs. 0.55 (0.10) mm, p = 0.019, p for trend = 0.036) (Figure 1). There were no differences in values of BP, lipids or plasma glucose across quartiles of adiponectin (data not shown).

Figure 1.

Mean values (95% confidence interval) of carotid intima-media thickness (CIMT) in participants according to adiponectin quartiles. Values of adiponectin quartiles: 1 (< 5.88 µg/ml); 2 (5.88–10.03 µg/ml); 3 (10.04–14.72 µg/ml); 4 (≥ 14.73 µg/ml). p-value obtained by analysis of variance. Bonferroni correction showed difference between the first quartile and the other three quartiles (p = 0.019, p for trend = 0.036).

Discussion

An interesting inverse association of adiponectin with CIMT found in this study may suggest that its determination may be useful as a marker of subclinical atherosclerosis in individuals with low or moderate cardiovascular risk. The strategy was to examine whether biomarkers with a putative role in the early atherogenic process – such as adiponectin – could be associated with high CIMT values in middle-aged adults with a maximum of three cardiovascular risk factors, linked by the insulin resistance.

The impact of body adiposity for the cardiometabolic risk is well established.¹⁶ Hypertrophy of the adipose tissue, particularly the visceral fat, has a central role in the genesis of insulin resistance and cardiometabolic diseases. Adipocytes in association with macrophages produce inflammatory cytokines which deteriorate insulin signaling and predispose to atherosclerosis.^17–19 Our findings of higher mean values of BMI, waist circumference, C-reactive protein and HOMA-IR in individuals with elevated CIMT are coherent with the assumption that fatness could trigger pro-inflammatory pathways and insulin resistance, favoring atherogenesis. The unfavorable risk profile of our participants with high CIMT is also reinforced by their higher frequencies of low-grade inflammation and hypertension compared with normal CIMT. Along the same line, the group with abnormal CIMT group showed lower mean values of adiponectin, which could be contributing to the aggravation of insulin resistance.⁹

It is Important to notice that the proportions of individuals with elevated BP (21.0%) and C-reactive protein > 3.0 mg/l (28.0 %) are small even in the high CIMT group. Also, previous publication of ELSA-Brazil found that, for the majority, the traditional cardiovascular risk factors are unable to predict risk, namely elevated CIMT.⁴ As inflammation, insulin resistance and endothelial dysfunction are underlying mechanisms of type 2 diabetes and atherosclerotic disease, it could be expected that markers of these abnormalities could improve risk prediction.^5,6,20 In fact, C-reactive protein concentration showed to be associated with CIMT in our univariate analysis.

Despite the detection of some differences in variables when comparing normal with high CIMT, after adjustments for waist circumference and insulin resistance they were not significant. However, adiponectin concentration maintained the association with CIMT along with systolic BP. We emphasize that adiponectin was indirectly associated with high CIMT independently of cardiovascular risk factors, such as adiposity, BP, subclinical inflammation (C-reactive protein) and HOMA-IR.

Our findings on association of lower adiponectin levels and subclinical atherosclerosis in low risk individuals support biological plausibility proposed in mechanistic studies. The mechanisms by which adiponectin acts as insulin sensitizer and is anti-atherogenic have been attributed to the effects on specific receptors (AdipoR1, AdipoR2 and T-cadherin) in liver, heart, skeletal muscle and endothelial cells.^21,22 Most beneficial effects on endothelium are mediated by its ability to activate AMPK, increasing eNOS activity and NO production.²³ Adiponectin is shown to inhibit the adhesion of monocytes to endothelium, transformation of macrophages into foam cells and neointimal formation by suppressing proliferation and migration of vascular smooth muscle cells.^21,9 Also, endothelial progenitor cells (EPCs) – involved in endothelial repair following vascular injury – are stimulated by adiponectin.^21,22 This hormone counteracts diabetes-induced damage of EPC function,^22,24 in part due to inhibitory effects on the production of reactive oxygen species has been described.²¹ These effects are coherent with the reports of reduced adiponectin levels in individuals with CVD⁹ and higher CIMT.^25–28 In our sample, we are hypothesizing that in participants with slight endothelial damage, as those with few cardiovascular risk factors, lower circulating adiponectin may be indicating arteries more vulnerable to atherosclerosis.

For preventive purposes, management of obesity, hypertension and other major cardiovascular risk factors are established. As far as novel biomarkers, such as adiponectin, are concerned, the predictive role as well as cutoff values for increased risk of CVD is not defined. Our data provided evidence that adiponectin concentration ≤ 5.1 µg/ml may be indicative of early carotid injury by atherosclerotic process. This value is in agreement with a review of several studies that considered normal values for adiponectin ranging from 5 to 15 µg/ml.²⁹ Therefore, we raise the possibility of a role of adiponectin determination to identify early atherogenesis. Such a hypothesis deserves to be investigated in prospective studies. Early identification of risk is relevant since healthy lifestyle changes in individuals at low risk have been shown to decrease risk for cardiovascular events in middle-aged adults.³⁰

Our study has the strength of a large sample of middle-aged individuals at low or moderate cardiovascular risk. Although this had represented a good opportunity to assess cardiovascular risk in a specific population, extrapolation to the general population is limited. Independent of the characteristics of this population, our results contribute to the knowledge about underlying mechanisms of the diseases. Among those, a potential circulating biomarker – adiponectin – was investigated in association with a structural marker of subclinical atherosclerosis – CIMT. However, our cross-sectional design precludes investigating causality. The ELSA-Brasil cohort will provide the opportunity to examine prospectively the ability of novel biomarkers to predict CVD using accurate methods to detect structural lesions of arteries, such as the CIMT and the coronary computed tomography.^12,18 Regarding detection of structural lesions of arteries, plaques are thought to be a better marker of atherosclerosis than IMT. Unfortunately, we did not have data regarding plaques. The IMT might be expressing an early phase of the atherosclerosis process, maybe more reversible, which can also explain a weaker association with cardiovascular events than plaque. Considering this, IMT can be an interesting test to identify other markers of early atherosclerosis with potential preventive uses.

In conclusion, our results reinforce previous evidence on the benefits of adiponectin on carotid wall, showing its inverse association with CIMT independently of body adiposity. This finding suggests that adiponectin determination may be useful to improve cardiometabolic risk assessment in this subset of individuals at low-to-moderate cardiovascular risk. Further studies are necessary to test the hypothesis raised in our study.

Author contribution

BAP participated in the study design, organization of the data, analysis of novel biomarkers, statistical analysis, interpretation of the results and drafting of the article. FFRF participated in the review of the statistical analysis and of the article. ISS participated in the interpretation of data and manuscript drafting. IMB conceived of the ELSA-Brasil study, participated in interpretation of the results and review of the article. PAL conceived of the ELSA-Brasil study, participated in interpretation of the results and review of the article. SRGF conceived of the actual study, design of the study, participated in interpretation of the results and review of the article. All authors read and approved the final manuscript.

Acknowledgements

The authors would like to acknowledge the participation of the 15,105 individuals recruited for this study without which this study and those based on the ELSA-Brasil cohort would not have been possible.

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 author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The current work was supported by grant from the São Paulo Research Foundation (Fundação de Amparo à Pesquisa do Estado de São Paulo – FAPESP – Protocol 2010/00074-6), São Paulo, SP, Brazil. 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 and CNPq - National Research Council) (grants # 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, 01 06 0071.00 RJ).