Association of plasma osteoprotegerin and adiponectin with arterial function, cardiac function and metabolism in asymptomatic type 2 diabetic men

Type 2 diabetes mellitus (T2DM) patients are at high risk of developing cardiovascular disease [1]. Even in the absence of hypertension and coronary artery disease, left ventricular (LV) diastolic dysfunction occurs in a large proportion of asymptomatic T2DM patients [2, 3]. Identification of the individual risk is important for the prevention and treatment of cardiovascular disease. Appropriate biomarkers could be important for risk stratification.

Osteoprotegerin (OPG) was first described as a key factor in bone remodeling [4]. OPG is a member of the tumor necrosis factor (TNF) receptor family, and a decoy receptor for the receptor activator of nuclear factor-κB ligand (RANKL) and TNF-α related apoptosis-inducing ligand. It prevents RANKL from binding to its receptor on osteoclasts, thereby inhibiting osteoclastogenesis [4]. Furthermore, OPG has been implicated in human atherogenesis [5, 6]. Previously, an association between OPG and LV dysfunction in the general population was described [7]. In addition, elevated OPG levels were present in patients with heart failure [8]. Recently, circulating OPG levels were reported as independent predictor of cardiovascular mortality in patients with stable coronary artery disease [9]. In asymptomatic T2DM patients OPG levels predicted subclinical atherosclerosis and cardiovascular events [10]. A 17-year prospective observational study in T2DM patients showed a strong predictive value of OPG for all-cause mortality, independent of conventional risk for cardiovascular disease, including renal function [11].

The adipocytokine adiponectin (ADPN) is downregulated in T2DM [12]. This hypoadiponectinemia in T2DM is associated with impaired insulin sensitivity, and to a lesser extent with adiposity and glycemia [13]. Furthermore, lower adiponectin levels have been found in T2DM patients with diabetic foot, a complication characterized by neuropathy, microvascular defects and inflammation compared to T2DM patients without diabetic foot [14]. Among men with T2DM, ADPN was negatively associated with the risk of coronary artery disease, which seemed to be partly mediated by high density lipoprotein (HDL)-cholesterol, but independent of other blood lipids, glycated hemoglobin (HbA1c), or inflammatory markers [15].

The present study was based on the hypotheses that circulating OPG and ADPN levels may differ between asymptomatic men with uncomplicated T2DM and age matched normoglycemic controls and that these biomarkers may be associated with arterial function, cardiac function and metabolism in these subjects.

As expected, T2DM patients versus healthy men had increased BMI, waist, and HbA1c values, but lower HDL-cholesterol, and M-value. Although within the normal range, systolic and diastolic blood pressure and heart rate were significantly higher in T2DM patients relative to controls (Table 1). T2DM patients had a lower aortic distensibility as compared to controls. Furthermore, T2DM patients and controls had comparable LV mass and systolic function parameters. Parameters related to LV volume and diastolic function were, however, significantly decreased in patients (Table 2). Insulin stimulated MMRglu was significantly impaired in patients compared with controls, in the presence of comparable myocardial blood flow [2, 17] (Table 2).

Plasma OPG levels were significantly elevated in T2DM men compared to controls (Figure 1A), whereas ADPN levels were lower (Figure 1B). PET examinations were successful in 58 T2DM patients and 9 controls. In T2DM patients, plasma OPG levels correlated inversely with aortic distensibility, LV endsystolic volume, E/A peak ratio and MMRglu, and positively with age, and HbA1c and LV mass/volume ratio. The correlations of plasma OPG with systolic blood pressure and LV enddiastolic volume were only significant in pooled analyses (Table 3). In multivariable regression analysis, plasma OPG was negatively associated with aortic distensibility, even after adjustment for systolic blood pressure and age (β = -0.268, P = 0.007; Figure 2A). Likewise, the association of plasma OPG and LV mass/volume ratio was independent of systolic blood pressure and age (β = 0.216, P = 0.05; Figure 2B). However, age and blood pressure significantly affected the association between OPG and LV enddiastolic volume in the T2DM patients (β = -0.104, P = 0.32). In addition, age and HbA1c were contributors of the association between OPG and E/A peak ratio (β = -0.071, P = 0.47; Figure 2C).

Plasma ADPN levels were inversely correlated with BMI and waist, and positively with HDL-cholesterol and M-value, in T2DM patients. Correlations of ADPN and HbA1c, rate pressure product and MMRglu were only significant in pooled analyses. ADPN was not significantly correlated with LV functional parameters (Table 4). In multiple regression analysis, M-value, HbA1c and waist were contributors of the association between ADPN and MMRglu (β = 0.055, P = 0.66; Figure 3).

The present study confirmed the previously reported increase in circulating OPG and decrease in ADPN in patients with T2DM versus health controls. We extended these data by showing inverse associations of OPG and cardiac function as well as aortic distensibility, whereas ADPN was positively related to cardiac metabolism.

These correlations were seen both in pooled analyses and the analyses of T2DM men separately, but not in the analyses of the controls alone.

Higher levels of OPG have documented in T2DM patients with asymptomatic coronary artery disease [28, 29], but at the same time also in experimental and clinical heart failure, even in the absence of ischemic cardiomyopathy. Similarly, in our population inducible ischemia was excluded. OPG has been associated with progression of atherosclerosis in symptomatic cardiovascular disease. For example, in a population based cohort, serum OPG was significantly associated with myocardial infarction, ischemic stroke, and total mortality [30]. Although, some indicate a beneficial role of OPG [8, 31], the higher levels of OPG have been interpreted as a counter regulatory protective response to atherosclerosis. For instance, selective deletion of OPG in mice results in significant medial calcification of the aorta and renal arteries [31]. Futhermore, Ueland et al. [8] showed enhanced systemic expression of RANKL and increased expression of the receptor RANK in cardiomyocytes, vascular smooth muscle cells and endothelial cells of the failing myocardium. OPG can act as a decoy receptor, binding to RANKL, preventing the interaction of RANK and RANKL. Thereby, OPG can potentially protect against the negative effects of RANKL. Besides, in patients with subclinical atherosclerosis serum OPG levels decreased linearly with increased calcification expressed as an increased echogenicity of the atherosclerotic plaques [32]. In addition to LV diastolic function, in our study, plasma levels OPG were significantly associated with age and HbA1c in T2DM patients, showing close correlation between OPG and several cardiovascular risk factors, as demonstrated previously [10, 33]. This is the first study showing an inverse association between OPG levels and arterial- and LV diastolic function in asymptomatic T2DM patients in the absence of inducible myocardial ischemia. The multivariate analysis indicated that plasma OPG may contribute to the effect of aging and hyperglycemia on arterial- and LV diastolic function in T2DM. These results indicate a positive relationship between OPG and cardiovascular disease, instead of a protective role of OPG in cardiovascular disease.

The association between ADPN and systemic glucose metabolism has been documented before by others [12, 13]. In the present study, we found a positive association of ADPN and myocardial glucose metabolism, but not with cardiac function. This finding seems to be at odds with previous data, which have described an association of circulating levels of ADPN with parameters of cardiac structure and function in healthy adults [34], elderly [35] and patients with coronary artery disease and/or heart failure [36, 37]. In addition, in healthy subjects, circulating ADPN levels were inversely related to LV wall thickness and mass [34] and, in patients with coronary artery disease and/or heart failure, they were positively related to NT-proBNP and LV systolic dysfunction [36, 37]. In chronic heart failure, the paradoxical (positive) association of ADPN and cardiac function was ascribed to its role in wasting in cardiac cachexia [38]. Differences in study population, and in methodology used to assess cardiac function and ADPN analysis, respectively, all could have accounted for the differential findings. Indeed, we studied men with well-controlled uncomplicated T2DM and with stress-echocardiography confirmed absence of inducible ischemia, whereas others included healthy adults [34], elderly [35] and patients with coronary artery disease and/or heart failure [36, 37].

Although in line with previous investigations, we have determined total circulating ADPN levels in our study, but other data suggest that high-molecular weight ADPN may have stronger associations with insulin resistance and coronary heart disease in T2DM patients [39].

Limitations of this study include the cross-sectional design, which precludes conferring a causal relationship between OPG, ADPN and the cardiac variables studied. Furthermore, the correlations of OPG with cardiac function and ADPN with myocardial glucose uptake were not observed in the healthy controls alone. This may be due to the relatively small number of a homogeneous control population.

In addition, the inclusion of only males with uncomplicated T2DM limits generalizations. On the other hand, a strength of this investigation is the use of state-of-the-art cardiac phenotyping methods in a relatively large group of well-characterized T2DM patients.

The findings of this study indicate that in asymptomatic men with uncomplicated T2DM, but without inducible ischemia, OPG and ADPN may be markers of underlying mechanisms linking the diabetic state to cardiac abnormalities. The use of OPG and ADPN as markers might add to the risk stratification of (future) cardiac disease in asymptomatic men with less-advanced T2DM prior to the onset of coronary artery disease.