Increased plasma interleukin-1 receptor antagonist levels in men with visceral obesity

Abstract

Background. Adipose tissue (AT) is an important source of interleukin-1 receptor antagonist (IL-1Ra). Its expression is markedly increased in obesity.

Aims and methods. To quantify the associations of IL-1Ra with body fat distribution as well as to examine the respective contributions of IL-1Ra and visceral adiposity to the variation in some cardiometabolic risk (CMR) markers. Plasma IL-1Ra levels were measured in a sample of 117 healthy non-diabetic men (age: 44.9±10.1 years; body mass index (BMI): 28.8±4.5 kg/m²).

Results. Plasma IL-1Ra levels correlated positively with BMI, waist girth, and visceral and subcutaneous AT (0.39 ≤ r<0.48; P < 0.0001). Multiple regression analyses revealed that visceral AT was the best independent predictor of IL-1Ra levels, explaining 22% (P < 0.0001) of its variance. IL-1Ra (P < 0.05) was an independent predictor of several CMR markers including triglyceride and high-density lipoprotein (HDL) cholesterol levels, cholesterol/HDL cholesterol ratio, glucose and insulin concentrations in response to a 75 g oral glucose load, and fasting insulin levels, in addition to the expected contribution of visceral AT (P < 0.05).

Conclusions. These results suggest that elevated IL-1Ra concentrations are influenced to a greater extent by visceral than subcutaneous adiposity and that IL-1Ra is independently related to some features of CMR beyond the known contribution of visceral adiposity.

• Cardiometabolic risk markers
• IL-1Ra
• visceral obesity

Introduction

Many proinflammatory cytokines such as tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6) are secreted by the adipose tissue (AT) in the context of abdominal obesity and are key mediators of obesity-induced insulin resistance [1], [2]. The anti-inflammatory cytokine IL-1 receptor antagonist (IL-1Ra) is a natural antagonist to the proinflammatory cytokine IL-1 and acts by binding to IL-1 receptors without inducing a cellular response [3]. It has been recently reported that plasma IL-1Ra levels are elevated in human obesity and are reduced after bariatric surgery [4]. The same authors have also shown that human AT is a major source of plasma IL-1Ra and that its expression in this tissue is markedly increased in obesity [5]. Moreover, the production of IL-1Ra is much higher compared to that of IL-6, TNF-α, and IL-1. In fact, plasma IL-1Ra levels are elevated in human morbid obesity to a similar extent as in systemic inflammation encountered in inflammatory autoimmune diseases [5].

Key messages

• Interleukin-1 receptor antagonist (IL-1Ra) concentrations are influenced to a greater extent by visceral than subcutaneous adiposity.

• IL-1Ra levels are independently related to several features of cardiometabolic risk beyond the known contribution of visceral adiposity.

As IL-1Ra knock-out mice are more susceptible to septic shock and inflammatory disorders, IL-1Ra probably plays an important role in the regulation of inflammatory responses [6]. Elevated circulating IL-1Ra concentrations appear to be a response to increased IL-1 production, and this phenomenon may represent a preventive mechanism in prolonged inflammatory response. IL-1Ra acts also as an acute-phase protein because its expression is regulated by proinflammatory cytokines in hepatocytes [7]. It appears that it is the imbalance between IL-1Ra and IL-1 which predisposes to the development of many chronic diseases such as diabetes [8]. As such, in clinical practice, IL-1Ra has been used as a therapeutic agent in human patients for treating rheumatoid arthritis [9].

There are limited data on the association between plasma levels of IL-1Ra and abdominal obesity measured by computed tomography in middle-aged men. In patients with Cushing's syndrome, elevated circulating IL-1Ra has been associated with central obesity, a characteristic feature in this condition [10]. Given the well known metabolic complications associated with visceral adiposity, the objectives of the present study were: 1) to investigate the associations of plasma IL-1Ra with specific abdominal AT depots (subcutaneous and visceral AT measured by computed tomography); and 2) to examine the respective contributions of plasma IL-1Ra and visceral adiposity to the variation in cardiometabolic risk variables.

Material and methods

Subjects

This cross-sectional study was performed in a sample of 117 apparently healthy adult men (age 44.9 ± 10.1 (mean±SD) years) from the Québec city metropolitan area, recruited through the media and selected to cover a wide range of body fatness values (body mass index (BMI) ranged from 25.5 (25th percentile) to 32.1 (75th percentile) kg/m²). All subjects gave written consent to participate in the present study, which was approved by the Medical Ethics Committee of Université Laval. Men with diabetes or coronary heart disease as well as smokers and morbidly obese individuals were excluded. None of the subjects were on medication known to affect insulin action or plasma lipoprotein-lipid levels. No subjects were on anti-inflammatory drugs either before or at the time of the study. Individuals using aspirin as a chronic medication were excluded from the study. Subjects were not allowed to take any medication or to perform physical activity for at least 24 hours before any metabolic investigation.

Abbreviations

Table

AT	adipose tissue
ANOVA	analysis of variance
BMI	body mass index
CRP	C-reactive protein
HDL	high-density lipoprotein
IDF	International Diabetes Federation
IFN-β	interferon-beta
IL-1Ra	interleukin-1 receptor antagonist
IL-6	interleukin-6
NCEP	National Cholesterol Education Program
OGTT	oral glucose tolerance test
TNF-α	tumor necrosis factor-alpha
SD	Standard deviation

Anthropometric measurements

Height, body-weight [11], and waist circumference [12] were measured following standardized procedures.

Computed tomography

Cross-sectional areas of abdominal subcutaneous and visceral AT were performed by computed tomography at L4-L5, using previously described procedures [13]. Briefly, each subject was examined in the supine position with both arms stretched above the head. Visceral AT area was determined by drawing a line in the middle of the muscle wall surrounding the abdominal cavity. The subcutaneous AT area was calculated by subtracting visceral AT area from the total abdominal AT area. Cross-sectional AT areas (cm²) were computed using an attenuation range of −190 to −30 Hounsfield units as previously described [13].

Oral glucose tolerance test (OGTT)

A 75 g OGTT was performed in the morning after an overnight fast. Blood samples were collected in ethylenediamine tetraaceticacid (EDTA)-containing tubes (Miles Pharmaceuticals, Rexdale, Ontario, Canada) through a venous catheter placed in an antecubital vein at −15, 0, 15, 30, 45, 60, 90, 120, 150, and 180 minutes for the determination of plasma glucose and insulin concentrations. Plasma glucose was measured enzymatically [14], whereas plasma insulin was measured by radioimmunoassay with polyethylene glycol separation [15]. The total glucose and insulin areas under the curve (AUC) during the OGTT were also calculated.

Plasma lipoprotein-lipid variables

Blood samples were collected from an antecubital vein into vacutainer tubes containing EDTA (Miles Pharmaceuticals, Rexdale, Ontario, Canada) after a 12-hour overnight fast for the measurement of plasma lipid and lipoprotein levels. Cholesterol and triglyceride levels were determined in plasma and lipoprotein fractions by the use of a Technicon RA-500 (Bayer Corporation, Tarrytown, New York), and enzymatic reagents were obtained from Randox (Crumlin, UK). The high-density lipoprotein (HDL) fraction was obtained after precipitation of low-density lipoprotein (LDL) in the infranatant (density >1.006 g/mL) with heparin and MnCl₂[16]. Cholesterol and triglyceride concentrations in the infranatant were measured before and after the precipitation step. Total apolipoprotein A1 and B concentrations were measured in plasma using the rocket immunoelectrophoretic method of Laurell [17] or by a nephelometric method using polyclonal antibodies on the BN ProSpec (Dade-Behring, Germany).

Determination of plasma IL-1Ra, TNF-α, IL-6, leptin, adiponectin, and C-reactive protein concentrations

Fasting plasma IL-6, TNF-α, and IL-1Ra concentrations were assessed on deeply frozen plasma samples (−80°C) and were measured by a high-sensitivity enzyme-linked immuno sorbent assay (ELISA) (R&D Systems Inc., Minneapolis, USA). Fasting plasma leptin and adiponectin concentrations were determined by an ELISA (B-Bridge International, Inc., San Jose, CA) on deeply frozen plasma samples (−80°C). C-reactive protein (CRP) levels were measured with a highly sensitive immunoassay that used a monoclonal antibody coated with polystyrene particles; the assay was performed with a BN ProSpec nephelometer (Dade-Behring, Germany) according to the methods described by the manufacturer [18]. The intra- and interassay coefficients of variation were <10% and <20% for all ELISA cytokine measurements and <5% for automated CRP measurements.

Statistical analyses

Data presented in Table I are means±SD, or median values with interquartile range for skewed variables. The normality of distribution of each variable was verified using the Shapiro-Wilk test, and logarithmic transformations were performed if necessary. Spearman correlations were used to compute relationships among variables. A cardiometabolic risk score based on stratification of men according to the number of metabolic abnormalities was also calculated. One point was attributed to men each time a metabolic risk marker was within the top 75th percentile of the distribution of cholesterol/HDL cholesterol ratio, triglycerides, apolipoprotein B, fasting insulin, post-OGTT glucose level, plasma CRP, TNF-α, and systolic blood pressure, or within the bottom 25th percentile for HDL cholesterol and adiponectin levels. Mean levels of IL-1Ra were calculated for each cardiometabolic risk score obtained. Group differences were analyzed using one-way or two-way ANOVA with interactive effect. A multiple regression analysis was performed to evaluate the independent contribution of adiposity variables and leptin to the variance in IL-1Ra concentrations. Multiple regression analyses were also computed to sort out the independent and respective contributions of adiposity and circulating IL-1Ra concentrations to the variance in cardiometabolic risk variables. Results were considered significant at the P ≤ 0.05 level. All analyses were performed with the SAS statistical package version 9.1 (SAS Institute, Cary, NC).

Table I.  Anthropometric and metabolic characteristics of the study.

Variables	Values	25th–75th percentile
n	117
Age (years)	44.9±10.1	(39–52.5)
Body mass index (kg/m^2)	28.8±4.5	(25.5–32.1)
Weight (kg)	87.5±14.7	(76.8–97.4)
Waist girth (cm)	100.6±13.2	(91.0–109.6)
Adipose tissue accumulation (cm^2):
visceral	199±95	(128–270)
subcutaneous	263±115	(180–318)
total	460±176	(325–561)
Cholesterol (mmol/L)	4.92±0.80	(4.5–5.5)
LDL cholesterol (mmol/L)	3.04±0.67	(2.6–3.5)
HDL cholesterol (mmol/L)	0.99±0.21	(0.9–1.1)
Cholesterol/HDL cholesterol	5.13±1.12	(4.5–5.8)
Triglycerides (mmol/L)	2.06 (1.40, 2.69)
Apolipoprotein B (g/L)	1.03±0.20	(0.92–1.16)
Apolipoprotein A1 (g/L)	1.15±0.15	(1.04–1.24)
Fasting insulin (pmol/L)	102 (62, 177)
Fasting glucose (mmol/L)	5.75±0.49	(5.5–6.1)
Insulin area ((pmol/L/min)×10^−3)	114 (60.8, 179)
Glucose area ((mmol/L/min)×10^−3)	1.35 (1.18, 1.48)
HOMA	26.9 (15.5, 45.8)
C-reactive protein (mg/L)	1.34 (0.66, 2.58)
Leptin (ng/mL)	7.89 (4.56, 12.21)
Adiponectin (µg/mL)	3.94 (2.99, 5.59)
Interleukin-1Ra (pg/mL)	218 (160, 299)
Interleukin-6 (pg/mL)	0.94 (0.65, 1.27)
Tumor necrosis factor-α (pg/mL)	0.99 (0.77, 1.22)

Data are presented as mean±SD, or as median (interquartile range) for skewed variables.

HDL = high-density lipoprotein; LDL = low-density lipoprotein.

HOMA = fasting insulin (mmol/L)×fasting glucose (µmol/L)/22.5.

Results

Table I presents the characteristics of the study sample. Men were in their mid-40s and covered a wide range of adiposity values. The average BMI was 28.8±4.5 kg/m² and corresponded to a waist girth of 100.6±13.2 cm. The median values and interquartile ranges of plasma IL-1Ra levels was 218 (160, 299) pg/mL.

Table II shows that circulating IL-1Ra was significantly and strongly correlated with all indices of adiposity as well as with all markers of cardiometabolic risk, with the exception of total cholesterol and LDL cholesterol levels.

Table II.  Spearman correlation coefficients between concentrations of IL-1Ra and indices of adiposity and cardiometabolic risk variables.

	IL-1Ra	P-value
Body mass index (kg/m^2)	0.47	≤0.0001
Weight (kg)	0.41	≤0.0001
Waist girth (cm)	0.48	≤0.0001
Adipose tissue accumulation (cm^2):
visceral	0.47	≤0.0001
subcutaneous	0.39	≤0.0001
total	0.50	≤0.0001
Cholesterol (mmol/L)	0.08
LDL cholesterol (mmol/L)	0.00
HDL cholesterol (mmol/L)	−0.42	≤0.0001
Cholesterol/HDL cholesterol	0.43	≤0.0001
Triglycerides (mmol/L)	0.46	≤0.0001
Apolipoprotein B (g/L)	0.24	<0.05
Apolipoprotein A1 (g/L)	−0.28	<0.05
Fasting insulin (pmol/L)	0.49	≤0.0001
Fasting glucose (mmol/L)	0.22	<0.05
Insulin area ((pmol/L/min)×10^−3)	0.46	≤0.0001
Glucose area ((mmol/L/min)×10^−3)	0.47	≤0.0001
C-reactive protein (mg/L)	0.40	≤0.0001
Leptin (ng/mL)	0.37	≤0.0001
Adiponectin (µg/mL)	−0.35	≤0.0001
Interleukin-6 (pg/mL)	0.34	<0.001
Tumor necrosis factor-α (pg/mL)	0.27	<0.001

HDL = high-density lipoprotein; IL-1Ra = interleukin-1 receptor antagonist; LDL = low-density lipoprotein.

In order to examine the potentially additional contribution of visceral AT and circulating IL-1Ra to the variance in metabolic risk variables, 2×2 analyses of variance were conducted where subjects were classified into those having either low or elevated visceral adiposity (using the median of the distribution corresponding to < or ≥187 cm²) and further classified on the basis of low or high levels of circulating IL-1Ra (< or ≥218 pg/mL). Figure 1 shows that although visceral AT significantly contributed (P < 0.01) to the variation in fasting insulin (A), in the post-OGTT plasma insulin levels (B), in triglyceride levels (C), and in the cholesterol/HDL cholesterol ratio (D), circulating IL-1Ra levels also significantly and independently contributed to the variation in fasting insulin (P < 0.001), and the area under the curve of plasma insulin levels measured during the OGTT (P < 0.01), as well as to the variance in triglycerides (P < 0.05) and in the cholesterol/HDL cholesterol ratio (P < 0.05).

Figure 1.  Respective contributions of plasma IL-1Ra levels and visceral AT to the variance in metabolic risk variables (fasting insulin (A), area under the curve (AUC) of plasma insulin (B), triglycerides (C), and cholesterol/HDL cholesterol ratio (D)) in the different subgroups (1, 2, 3, 4) classified on the basis of the 50th percentile of visceral AT and IL-1Ra. Because variables were skewed, P-values of the log-transformed values are presented in panels A, B, and C (1, 2, 3 = significantly different from the corresponding subgroups; P < 0.05). IL-1Ra = interleukin-1 receptor antagonist; AT = adipose tissue; HDL = high-density lipoprotein.

Figure 2 shows plasma IL-1Ra levels in men classified on the basis of the number of metabolic abnormalities predictive of an increased cardiometabolic risk. Plasma IL-1Ra levels increased gradually with increasing numbers of metabolic abnormalities.

Figure 2.  Plasma IL-1Ra levels among subgroups of men classified on the basis of their number of cardiometabolic risk variables. a = significantly different from score 0: P < 0.05 on log-transformed values; b = significantly different from score 1: P < 0.05 on log-transformed values; c = significantly different from score 2: P < 0.05 on log-transformed values. HDL = high-density lipoprotein; AUC = area under the curve; IL-1Ra = interleukin-1 receptor antagonist.

Finally, multiple regression analyses were also conducted to sort out the independent contribution of leptin, visceral and subcutaneous AT to the variation in circulating IL-1Ra levels. This analysis revealed that visceral AT (22%; P < 0.0001) explained most of the variance in IL-1Ra followed by subcutaneous AT (4.9%; P < 0.05) for a total of 26.9%. A second model was analyzed to sort out the independent contribution of visceral AT, subcutaneous AT and circulating IL-1Ra levels to the variance in indices of plasma glucose-insulin homeostasis and lipoprotein-lipid variables (Table III). This analysis revealed that visceral AT explained the largest percentage of the variation in fasting insulin levels (partial R²: 24.5%; P < 0.0001), the area under the curve of plasma insulin (partial R²: 28.8%; P < 0.0001), and glucose (partial R²: 16.6%; P < 0.0001) levels measured during the OGTT, triglyceride (partial R²: 31.8%; P < 0.0001), and apolipoprotein B (total R²: 13.2%; P < 0.0001) levels. IL-1Ra contributed independently to the variation in HDL cholesterol (total R²: 28.1%; P < 0.0001), the cholesterol/HDL cholesterol ratio (partial R²: 25.8%; P < 0.0001), and apolipoprotein A1 levels (total R²: 9.3%; P < 0.0001).

Table III.  Multiple regression analyses for relative contributions of fat distribution and IL-1Ra to the variance in indices of plasma glucose-insulin homeostasis and lipids.

Dependent variable	Independent variable(s) (partial R^2)	Total R^2 (%)	P-value
Fasting insulin	Visceral AT (24.5%) Subcutaneous AT (5.3%) IL-1Ra (2.8%)	32.6	<0.05
Insulin area	Visceral AT (28.8%) IL-1Ra (5.2%) Subcutaneous AT (2.8%)	36.7	<0.05
Glucose area	Visceral AT (16.6%) IL-1Ra (5.1%)	21.6	<0.01
Triglycerides	Visceral AT (31.8%) IL-1Ra (7.9%) Subcutaneous AT (3.2%)	42.8	<0.05
HDL cholesterol	IL-1Ra (28.1%)	28.1	<0.0001
Cholestrol/HDL cholesterol	IL-1Ra (25.8%) Visceral AT (5.8%)	31.7	<0.01
Apolipoprotein B	Visceral AT (13.2%)	13.2	<0.0001
Apolipoprotein A1	IL-1Ra (9.3%)	9.3	<0.001

Visceral AT, subcutaneous AT and circulating IL-1Ra levels were the variables included in the model.

AT = adipose tissue; HDL = high-density lipoprotein; IL-1Ra = interleukin-1 receptor antagonist.

Discussion

Results of the present study are consistent with previous studies that had reported increased plasma IL-1Ra levels in obesity. To the best of our knowledge, our study provides for the first time evidence that circulating IL-1Ra concentrations are influenced to a greater extent by visceral than subcutaneous adiposity. Visceral AT explained 22% of the variation in plasma IL-1Ra levels compared to only 5% for subcutaneous AT. Moreover, plasma IL-1Ra showed highly significant correlations with several adiposity indices and cardiometabolic risk variables compared to previously studied cytokines. We also found that plasma IL-1Ra levels increased with the number of metabolic abnormalities. Furthermore, we found an association between IL-1Ra levels and several cardiometabolic risk variables which appeared to be partly independent from the variation in visceral adiposity.

Meier et al. [4] were the first to observe increased plasma levels of IL-1Ra in morbidly obese women. Then they showed in vitro that IL-1Ra protein and mRNA were clearly present in human AT and markedly increased in obesity and that adipocytes were the main source of IL-1Ra [5]. In contrast, a study by Fain et al. [19] suggested that IL-Ra release by human AT could be attributed to other cell types rather than to adipocytes such as macrophages. This study is the first to suggest that visceral AT contributes to the variation in circulating IL-1Ra in men. As for IL-1Ra protein and mRNA, no statistically significant differences were found between visceral and subcutaneous AT in morbidly obese patients [5]. A recent study [20] reported an association between the relative change of BMI between childhood and adulthood and the levels of IL-1Ra and CRP. The elevated levels of inflammatory markers predicted weight gain retrospectively, both in men and women.

It has also been reported that IL-1Ra knock-out mice are significantly lighter because of a defect in lipid accumulation in AT [21]. Moreover, Matsuki et al. [21] have shown that IL-1Ra-/- mice exhibit defects in postprandial insulin secretion and lipid metabolism. Insulin is a major regulator of lipid metabolism in adipocytes. Thus, low insulin levels may cause reduced fat accumulation in AT of IL-1Ra-/- mice. In addition, it has been recently shown in rats that elevated IL-1Ra reduces the sensitivity to insulin by impairing glucose uptake in skeletal muscle [22]. In humans, plasma IL-1Ra levels are strongly correlated with insulin resistance [4]. Accordingly, our results demonstrate that circulating IL-1Ra levels are strongly correlated to indices of plasma glucose and insulin homeostasis and contribute independently to the variation in fasting and the area under the curve of plasma insulin levels measured during the OGTT beyond the well known contribution of visceral AT. It has also been reported that IL-1Ra was the most sensitive marker of cytokine response in the prediabetic state in offspring of patients with type 2 diabetes [23]. However, according to recent work [24], the causes of the lean phenotype in IL-1Ra knock-out mice may be more complex than just the consequence of a hypoinsulinemic state. For instance, in addition to impaired energy storage in AT, IL-1Ra knock-out mice were found to be characterized by reduced adipocyte differentiation and by small adipocyte sizes. These observations suggest that the locally increased ratio of IL-1Ra to IL-1 in AT of obese patients [5] might contribute to obesity by limiting the auto/paracrine antilipogenic action of endogenous IL-1 [25]. In vivo studies also showed that IL-1Ra-deficient mice exhibit an elevated metabolic rate and a reduced caloric intake following a high-fat diet [24]. Thus, endogenous IL-1Ra is suggested to be a pro-adipogenic factor. Furthermore, leptin induces IL-1Ra in monocytes [26], and the elevated plasma IL-1Ra in human obesity is positively associated with leptin [4]. It has been suggested that IL-1Ra may contribute to central leptin resistance by antagonizing the anorexic effects of IL-1 on food intake [27], [28].

To the best of our knowledge, our study is the first to investigate the association between IL-1Ra and components of the atherogenic dyslipidemia, such as high triglycerides, low HDL cholesterol, and an elevated cholesterol/HDL cholesterol ratio. Similarly to what was observed for indices of plasma glucose-insulin homeostasis, circulating IL-1Ra levels were found to be significantly and independently associated with triglyceride levels and with the cholesterol/HDL cholesterol ratio. Moreover, IL-1Ra was the best predictor of HDL cholesterol levels (28%), cholesterol/HDL cholesterol ratio (25.8%), and apolipoprotein A1 (9.3%). It has been suggested that the IL-1/IL-1Ra system may control lipid and lipoprotein metabolism through direct actions on AT [29], [30]. However, the mechanisms involved remain unknown.

The known stimulators of IL-1Ra secretion are cytokines (IL-1, TNF-α, interferon-β (IFN- β)) [25]. It has been speculated that the increase of IL-1Ra in obese patients is due to its related inflammatory state [4]. Our results show significant correlations between IL-1Ra and TNF-α (r=0.27), IL-6 (r=0.34), and CRP (r=0.40) levels, proinflammatory markers associated with hypertrophied AT and produced mainly by the infiltrated macrophages [31]. In human AT explants, IL-1Ra secretion has been shown to be induced by IL-1 and IFN-β through a local paracrine/autocrine action since these molecules are produced by stromal cells or by the adipocytes themselves [25]. More recently, hypoxia-induced leptin in white AT was reported to cause an increase in plasma IL-1Ra in mice [32]. Although we found that leptinemia was indeed positively correlated with IL-1Ra levels (r=0.37, P < 0.0001), multivariate analysis revealed that leptinemia was not an independent contributor to the variance in IL-1Ra after inclusion of visceral and subcutaneous adipose tissue in the model.

Increased circulating Il-1Ra levels in obese subjects might contribute to obesity-associated metabolic complications, including insulin resistance and atherogenic dyslipidemia. Circulating IL-1Ra is elevated in subjects with the metabolic syndrome [33] and linearly correlated with the number of components of the metabolic syndrome according to the International Diabetes Federation (IDF) and National Cholesterol Education Program (NCEP) clinical criteria [34]. Using a score that takes into account ten common cardiometabolic risk markers associated with abdominal obesity, our results further reinforce the notion that individuals with a deteriorated cardiometabolic risk profile have higher plasma IL-1Ra levels.

Results of the present study reinforce the notion that circulating IL-1Ra concentrations are elevated in obesity. More specifically, we found that visceral AT was associated with increased IL-1Ra independently from total or subcutaneous adiposity. Furthermore, we also showed that the inflammatory cytokines associated with abdominal obesity were related to IL-1Ra levels. Our results add to the evidence that plasma IL-1Ra levels in obese humans might contribute to obesity-associated abnormalities. Indeed our results indicate that IL-1Ra levels are independently related to some features of cardiometabolic risk beyond the known contribution of visceral adiposity. More studies will be necessary to clarify the mechanism responsible for such independent associations. Nevertheless, we believe that IL-1Ra can be considered as another important marker of the dysmetabolic/inflammatory state of visceral obesity.

Acknowledgements

This study was supported by the Canadian Institutes of Health Research (MT-14014) and by the Heart and Stroke Foundation of Canada. Jean-Pierre Després is the Scientific Director of the International Chair on Cardiometabolic Risk which is supported by an unrestricted grant awarded to Université Laval by Sanofi Aventis. Angelo Tremblay is partly funded by the Canada Research Chair in Environment and Energy Balance. Amélie Cartier is a recipient of a scholarship from the Merck Frosst/CIHR Research Chair in Obesity. We would like to express our gratitude to the study subjects for their contribution to the present study and to our staff for their dedicated work. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.