Background
Omentin, also referred as intelectin-1, was first identified as a galactofuranose-binding lectin [
1], and is abundantly expressed in human omental adipose tissue [
2]. It enhances insulin action in human adipocytes in vitro [
2] and its circulating levels are decreased in individuals with obesity [
3,
4] and/or patients with type 2 diabetes (T2D) [
3,
5,
6]. Omentin also has beneficial effects on cardiovascular system in mice through its actions on endothelial cells, smooth muscle cells, macrophages, and cardiomyocytes [
7,
8]. In line with the experimental evidence, decreased circulating omentin levels were observed in patients with coronary artery disease [
9], peripheral artery disease [
10], and established carotid atherosclerosis [
11] compared to those without. An inverse relationship was also observed in recent clinical studies between circulating omentin levels and surrogate makers of cardiovascular risk, including arterial stiffness [
6], endothelial dysfunction [
12], carotid intima-media thickness (IMT) [
13,
14], and carotid plaque instability [
15]. More recently, we reported that plasma omentin levels were higher in T2D patients than healthy controls and were positively associated with flow-mediated vasodilatation, a marker of endothelial function, in high-risk subgroups of T2D patients with older age, reduced kidney function, or preexisting cardiovascular diseases (CVDs) [
16]. Those prior studies [
6,
9‐
14], including ours [
16], suggest a protective role of human omentin on atherosclerotic CVDs.
Adiponectin is well-recognized as an anti-inflammatory adipocytokine which exerts beneficial actions on metabolic and cardiovascular disorders [
8,
17,
18]. Previous reports consistently demonstrated that circulating adiponectin levels were decreased in obesity, T2D, and coronary artery disease [
8,
17,
18]. However, despite its anti-inflammatory and cardioprotective properties in cellular and animal models, a number of prospective human studies have indicated a positive, rather than the expected negative, association between circulating adiponectin levels and all-cause and cardiovascular mortality across many clinical conditions including diabetes, CVDs, and chronic kidney disease [
19,
20]. Regarding subclinical atherosclerosis, conflicting results were reported on the association between hypoadiponectinemia and severity of carotid IMT, especially in subjects with metabolic disorders or inflammatory diseases [
21]. Namely, most studies in relatively healthy subjects or subjects with wide range of glucose tolerance significantly associated lower adiponectin levels with greater IMT, while the majority of studies in subjects with T2D alone failed to show a significant association between adiponectin and IMT [
21]. These clinical findings collectively suggest that adiponectin fails to exert its protective effect against atherosclerotic CVDs in individuals with T2D.
Both omentin and adiponectin are among the adipokines with anti-inflammatory and cardioprotective effects in experimental models [
8]. However, no prior study has investigated a direct molecular link between omentin and adiponectin. Evidence from clinical studies indicates that circulating omentin levels are increased, while adiponectin levels are decreased in proinflammatory states such as type 1 diabetes, Crohn’s disease, and nonalcoholic fatty liver disease [
7]. The reciprocal regulation observed between omentin and adiponectin may point towards a more important role of omentin as a regulator of inflammatory process. In addition, our recent study showed that omentin, but not adiponectin, was positively associated with vascular endothelial function in high CVD risk subgroups of T2D patients [
16], suggesting a dominant role of omentin, rather than adiponectin, against endothelial dysfunction in patients at high CVD risk status. In the present study, we hypothesized that omentin plays an anti-atherogenic effect, while adiponectin levels are increased but lose its effect, against atherosclerosis in patients with high-risk status for CVDs. To this end, we investigated the clinical features of patients with T2D and increased plasma adiponectin levels, and the relationship between plasma omentin levels and carotid IMT in those patients.
Discussion
In the present study, we investigated the clinical features of T2D patients showing increased plasma adiponectin levels and the relationship between plasma omentin levels and carotid IMT in those patients. Our results demonstrated that the high-adiponectin group had increased omentin levels, multiple CVD risk factors, and non-significantly greater IMT, as compared to the low-adiponectin group of T2D patients. Moreover, in the high-adiponectin group, plasma omentin levels were found to be inversely associated with IMT, independently of adiponectin levels and traditional CVD risk factors, but not in the low-adiponectin group. To our knowledge, this study is the first to demonstrate the association between omentin and IMT which is modified by adiponectin level in patients with T2D.
This study first demonstrated that T2D patients with higher adiponectin levels (6.2 μg/mL or greater) had older age, higher systolic blood pressure, lower eGFR, as well as lower BMI, HbA1c, insulin resistance indices, and C-reactive protein, than those with lower adiponectin levels. The finding indicates the presence of multiple CVD risk factors in high-adiponectin group of T2D patients. A clustering of CVD risk factors in high-adiponectin group was not unexpected because previous studies have consistently associated higher adiponectin levels with older age [
28] and chronic kidney disease [
29]. Higher systolic blood pressure may also be explained by older age and lower eGFR in high-adiponectin group than in low-adiponectin group. Although inverse association between adiponectin and blood pressure is evidently shown in the general population [
30], the adiponectin levels in hypertensive patients could be elevated depending on kidney function, as previously reported [
31].
In addition to the clustering of CVD risk factors, carotid IMT tended to be greater in high-adiponectin group than low-adiponectin group. The finding contradicts the anti-atherogenic and anti-inflammatory properties of adiponectin [
8]. However, recent cross-sectional studies have indicated that an inverse association between adiponectin level and IMT was significant only in healthy subjects or the general population, but not in patients with chronic disorders such as diabetes and inflammatory diseases [
21]. Moreover, recent prospective studies suggest a positive, rather than the expected negative, relationship between adiponectin levels and cardiovascular and all-cause mortality in many clinical settings including type 2 diabetes [
19,
20]. Several explanations have been proposed for the paradoxical relationship between adiponectin and mortality, including inflammation, weight loss, and the confounding role of natriuretic peptides [
19,
20]. The inverse relationship between BMI and IMT only in high-adiponectin group (Table
3), which is contrary to what is generally shown in obese or pre-diabetic population [
32], could also be understood in the context of paradoxical relationship between increased adiponectin levels and advanced atherosclerotic CVDs [
19,
20]. Because plasma adiponectin level was not an independent determinant of IMT (Table
3), our data indicate that some factors related to older age and reduced BMI, rather than adiponectin itself, are implicated in the advanced atherosclerosis in high-adiponectin group.
This study clearly demonstrated an inverse relationship between omentin and IMT in T2D patients showing increased plasma adiponectin levels and multiple CVD risk factors. The inverse association between omentin and IMT is in line with clinical [
6,
10,
11,
13‐
15] and experimental [
33‐
35] studies reported previously. A number of studies demonstrated that omentin levels were inversely associated with carotid atherosclerosis, as evaluated by IMT [
13,
14], carotid stenosis [
11], plaque presence [
6], or plaque instability [
15], in non-diabetic population [
11,
13‐
15] and in patients with T2D [
6]. Experimental studies also consistently indicated a protective effect of omentin on atherosclerosis in mice [
33‐
35] through attenuating inflammatory macrophages [
33,
35] and migration/proliferation of smooth muscle cells [
34,
35]. Because no prior study has assessed the relationship between omentin and atherosclerosis by using IMT only in T2D patients, this study provides additional evidence supporting an anti-atherogenic effect of omentin on atherosclerosis in humans.
It should be mentioned that plasma omentin levels were higher in T2D patients with higher adiponectin levels. Because omentin levels are known to be inversely correlated to parameters of obesity and the metabolic syndrome [
4,
13,
36,
37], plasma omentin levels could be elevated in association with lower BMI, lower insulin resistance index, and higher HDL-cholesterol levels in high-adiponectin group of our T2D patients. The higher omentin levels could also be related to older age and lower eGFR in high-adiponectin group because circulating omentin levels are known to be elevated in patients with advanced chronic kidney disease [
38]. Indeed, in the multivariate analysis (Table
2), lower BMI, higher HDL-cholesterol, older age, and lower eGFR were independent determinants of plasma omentin levels. Moreover, our data clearly showed, for the first time to our knowledge, a positive association between omentin and adiponectin levels which is independent of potential confounders. The data may raise a possibility that adiponectin directly upregulates omentin levels in humans. However, no experimental evidence has not been available on the molecular mechanism through which adiponectin regulates omentin production or secretion in human adipose tissue, which needs to be investigated in future studies.
This study demonstrated for the first time that the relationship between omentin levels and IMT varies depending on plasma adiponectin levels. An inverse association between omentin and IMT was found in high-adiponectin group but not in low-adiponectin group. As discussed above, omentin levels were increased in association with several CVD risk factors and possibly with increased adiponectin levels in high-adiponectin group. While studies in individuals without CVDs indicated an inverse relationship between omentin and subclinical atherosclerosis [
6,
11,
13,
14], studies in patients with established CVDs demonstrated that higher omentin levels were cross-sectionally associated with the severity of coronary artery disease [
35,
39] and were longitudinally associated with adverse CVD outcome [
40,
41]. Increased plasma omentin levels were also demonstrated after acute cardiac ischemia in patients undergoing cardiac surgery [
42]. Importantly, in patients with coronary artery disease and ≥ 90% coronary occlusion, plasma omentin levels were not only increased with the disease severity, but also were independently associated with better collateral circulation, or better endothelial function [
39]. Besides, we recently found that plasma omentin levels were increased and independently associated with better endothelial function in high CVD risk subgroups of patients with T2D, but not in low-risk patients [
16]. Taken together, our data suggest that omentin levels are increased to compensate for vascular damage and/or atherogenesis due to accumulation of CVD risk factors in T2D patients with increased adiponectin levels.
Because both omentin and adiponectin are considered to be the adipokines which exert protective action against atherosclerosis [
8], it is noteworthy that only omentin, but not adiponectin, showed a significant inverse association with atherosclerosis. It is evidently known that hyperglycemic status, or diabetes, affects the pro-inflammatory/oxidative properties and the pro-thrombotic properties in the arterial plaque lesion [
43‐
45]. Evidence from experimental studies has established the anti-inflammatory, anti-oxidative, and anti-thrombotic properties of adiponectin [
8]. However, recent meta-analyses have indicated that a direction of the relationship between circulating adiponectin level and IMT or carotid plaque presence is dependent on the severity of underlying disease, while increased adiponectin levels are associated with an increased risk of ischemic stroke [
21,
46]. These reports could associate increased adiponectin levels with carotid plaque instability. A recent human study also demonstrated an overall abundance of adiponectin with a decreased adiponectin receptor 2 expression and activity in unstable plaques in patients which underwent a carotid endarterectomy [
47], suggesting a failure of adiponectin action in unstable carotid plaque. Our results are in accordance with previous reports [
21,
46] including ours [
25] that failed to associate adiponectin levels with carotid IMT in patients with T2D and suggested a loss of anti-atherogenic effect of adiponectin in T2D. In contrast to adiponectin, the anti-atherogenic effect of omentin was evidently shown in our T2D patients even with elevated CVD risk. While both adipokines were correlated with lower eGFR, omentin levels, but not adiponectin levels, were also independently associated with older age and higher HbA1c levels (Table
2). Taken together, we speculate that omentin levels were upregulated by CVD risk factors, such as aging, renal dysfunction, and chronic hyperglycemia, more potently than adiponectin to play a role against atherosclerosis in patients with T2D.
It needs to be mentioned that insulin resistance [
48] and/or hyperglycemic status [
49] could also affect IMT in patients with T2D. Our data do not exclude the possible involvement of HOMA-R in the link between omentin and IMT, because the sample size was not enough in the high-adiponectin group. In our results, HbA1c levels were lower in high-adiponectin group than low-adiponectin group, suggesting that hyperglycemia is not involved in the advanced atherosclerosis in high-adiponectin group. Moreover, results from the subgroup analysis by HbA1c level suggest that glycemic control does not affect the relationship between omentin and IMT in T2D patients. Taken together with the results of Table
3, it is suggested that some factors related to older age and/or lower BMI, rather than hyperglycemia over the last few months, are implicated in the advanced atherosclerosis in high-adiponectin group.
A potential implication of inflammation in IMT in patients with T2D [
32,
43] also needs to be considered. The additional data for the limited subjects suggest that inflammation as assessed by C-reactive protein is involved in atherosclerosis in low-adiponectin group, but not in high-adiponectin group (Additional file
1: Table S3). Considering that omentin levels were inversely associated with IMT in high-adiponectin group, as was observed in the full sample, it is speculated that increased omentin level in high-adiponectin group plays a role against atherosclerosis through its anti-inflammatory effect.
With respect to the anti-atherogenic effect of antidiabetic agents, 33.2% of the participants were receiving incretin-related drugs, i.e. dipeptidyl peptidase-4 inhibitors or glucagon-like peptide-1 receptor agonists, which were recently shown to have beneficial effects on CVD outcomes [
50,
51] and on the progression of carotid IMT [
52] in patients with T2D. We found that the IMT was lower in users of these drugs than in non-users (median IMT, 0.78 vs. 0.83 mm,
p = 0.015). However, the use of incretin-related drugs was not an independent determinant of IMT after adjustment for potential confounders (
β = − 0.078,
p = 0.094). Because this study was not primarily designed to evaluate the effect of these drugs on IMT, our data never exclude the possible effect of incretin-related drugs on IMT in our study participants.
This study has several limitations. First, neither adipokines other than adiponectin nor proinflammatory cytokines were evaluated, with which the effect of omentin on atherosclerosis would have been characterized better than with adiponectin alone. Second, because over 40% of the participants were receiving insulin therapy, we could not fully assess the possible effect of HOMA-R on IMT. Third, this study did not utilize atherectomy specimens, with which we could apply histological and/or molecular approaches. Forth, we did not measure waist circumference or visceral fat area using computed tomography, which would be more closely correlated with adiponectin levels than BMI and could affect the results obtained. Fifth, the participants were receiving RAS inhibitors, statins, and/or anti-hyperglycemic agents including which could have affected omentin levels, adiponectin levels, IMT, and/or CVD risk factors. To minimize the effect of at least RAS inhibitors and statins on IMT, use of these drugs was adjusted for in multivariate analyses. Sixth, the main outcome of this study was IMT, which was shown to have limited value for predicting CVD outcomes [
53] and is not recommended for assessing CVD risk in recent guideline [
54]. Finally, because of the consecutive manner of subject inclusion in the university hospital, our participants with T2D had long disease duration, poor glycemic control, and high prevalence of macrovascular complications. Thus, the present results may not be generalized to the entire population of patients with T2D.