Adipokines are highly active biopeptides produced by adipose tissue (AT), involved in regulating glucose metabolism, insulin function and the development and progression of obesity and its associated diseases (1). Thus, AT also exerts its function in an endocrine manner, which is widely accepted (2). Adipokines have been classified as hormones, cytokines and growth factors and includes, among others, adiponectin, visfatin and tumor necrosis factor alpha (TNFα) (3).

AT may be divided into visceral adipose tissue (VAT) and abdominal subcutaneous AT (SAT), further subdivided into superficial (sSAT) and deep (dSAT). Both SAT and especially VAT have been associated with the metabolic syndrome (MetS) and its risk factors (4). dSAT and VAT are reported to have similar and a more pro-inflammatory profile than sSAT, both associated with insulin resistance and atherosclerosis (5, 6). The mechanism behind this differentiation remains unclear, but it has been speculated whether VAT is a more bioactive compartment than SAT irrespective of BMI (7).

Adiponectin is produced almost exclusively by adipocytes (8). It is known to have insulin sensitizing, anti-atherogenic and anti-inflammatory properties, and hypoadiponectemia is associated with obesity, insulin-resistance and type 2 diabetes mellitus (T2DM) (9, 10). Adiponectin is believed to exert its actions by inhibiting transformation of macrophages into foam cells and through inhibition of TNFα (11). Adiponectin seems furthermore to play a key role in determining metabolic health, independent of body mass index (BMI), as shown in healthy individuals (12).

Visfatin, predominantly expressed in VAT, is elevated in obesity and T2DM (13, 14). Visfatin is particularly interesting to examine in order to understand the biological difference between VAT and SAT, as previous studies have reported associations to the amount of VAT, but not to SAT (15).

TNFα was initially assumed to be produced by adipocytes in adipose tissue, due to its correlation to BMI (16). However, the stromovascular fraction of AT, including endothelial cells, macrophages and leukocytes, seems to have higher TNFα production than adipocytes (17). Fain et al. also demonstrated that it was primarily the non-fat cells in AT that were responsible for TNFα production (18). TNFα is thought to increase insulin resistance and cause dyslipidemia through suppression of adiponectin production (19). Furthermore, TNFα modulates insulin receptors through inhibition of tyrosine kinase activity, decreasing their insulin sensitivity (20).

The relationship between AT distribution and adipokines has not been adequately investigated and we therefore aimed to examine whether gene expression levels in SAT of selected adipokines and their corresponding circulating levels, associates with the amount of abdominal AT in sSAT, dSAT and VAT in a cohort of middle-aged, healthy men. We hoped to gain a better understanding of the source of circulating adipokines. Any associations with glucometabolic variables that have previously been linked to adipokines were further explored.

Characteristics of the study population, the amount of abdominal AT and levels of the measured circulating variables are shown in Table 1 . The study participants were all male, with an average BMI of 26 kg/m² and an average age of 39.5 years when the blood samples and biopsies were obtained. Sixty-one patients had a BMI above 25 kg/m². Of the 103 participants included in the study, CT scan was available for 92. AT biopsies for adiponectin, visfatin and TNFα were available for 80, 77 and 64 participants, respectively. The excluded samples were mainly due reluctance of some participants to have the procedure performed and due to inadequate tissue sample. Serum sample was available for 102 participants due to 2 lost to follow-up. The glucometabolic variables were in the normal range and 20% had a family history of diabetes. Median GDR was 6.27 mg/kg/min.

One of the main findings in our study on healthy middle-aged men was an inverse association between circulating adiponectin levels and the amount of abdominal AT, particularly in dSAT and VAT. There seems to be a unanimous agreement on the beneficial effects of adiponectin on metabolic conditions such as insulin resistance, diabetes and MetS, due to its insulin-sensitizing effect (27). In accordance with this, we could demonstrate circulating adiponectin to be inversely correlated to insulin and C-peptide and positively correlated to insulin sensitivity, assessed by GDR. Previous studies have shown a similar pattern, and considering adiponectin’s beneficial mechanism of action, the findings fits well with the literature (28), and it seems that these mechanisms are similar in healthy subjects.

The deep layers of the AT, in particular the visceral component, has consistently showed an association to dyslipidemia and MetS (29). However, the quantity of VAT and SAT varies between individuals and it seems like their biological effect also are related to their anatomical sites (30). Furthermore, although adiponectin is almost exclusively, produced by AT, the contribution of VAT vs. SAT and dSAT to the circulating adiponectin levels has not been fully understood and data here seems to be somewhat conflicting. From our study, it seems that dSAT is even as important as SAT and VAT. However, in-vitro gene expression of adiponectin was shown to be lower in VAT than in SAT, suggesting SAT to be a more important source for adiponectin (31). In a study performed on 783 non-obese, healthy men aged 20-29 years, circulating adiponectin was shown to correlate significantly and inversely to BMI, waist-circumference and to the amount of both VAT and SAT in a univariate analysis, but in a multiple regression analysis only the inverse correlations to SAT remained significant (31). The VAT and its resident macrophages are believed to be more active in the production of pro-inflammatory cytokines, such as TNFα, IL-18 and MCP-1, and less active with adiponectin production (30). These pro-inflammatory cytokines are discussed to be involved in the pathogenesis of metabolic disorders and endothelial dysfunction, partly through inhibition of adiponectin (32). Based on this, one may speculate whether the benefits of reduction in VAT are primarily due to reduced pro-inflammatory cytokine production, whereas the benefits of SAT or dSAT reduction are derived from increased production of beneficial adipokines such as adiponectin. On the other hand, a study on 916 overweight but not affected by obesity (BMI 28.8kg/m2) participants conducted by Jain et al. showed a stronger inverse correlation between adiponectin and the amount of VAT compared to other AT compartments (33).

Low levels of adiponectin in combination with high VAT mass has been associated with insulin resistance, decreased blood glucose control and metabolic disorders (34). When we divided our study population into lean and over-weight, based on BMI below or above 25, respectively, we found that limited correlations between the AT compartments to circulating adiponectin and TNFα RNA expression in either groups, unlike the findings in the total population. The limited correlations found in the subgroups was unexpected, but could possibly be explained by the smaller number of samples in each group, hence under-powering the study. In another study, it was shown that reduction in both SAT and VAT after one year of lifestyle changes with body weight reduction correlated to changes in circulating adiponectin levels and more so for VAT than for SAT (35). Lastly, studies have reported that visceral fat accumulation contributes to low levels of adiponectin (36). The underlying mechanism could be reduced production directly from VAT or, more likely, increased production of inhibiting factors such as TNFα (36).

Another main finding in our study was that TNFα gene expression correlated to the AT amount in all compartments, and particularly in SAT. The known associations between TNFα and reduced insulin sensitivity and MetS are probably mediated through multiple mechanisms, including suppression of insulin receptor and GLUT4, responsible for glucose uptake, and through inhibition of adiponectin (17). Contrary to our results, there seems to be more consistent data on a stronger association between visceral TNFα expression and VAT mass compared to TNFα and other AT compartments, however, unlike our population these study populations were either affected by obesity or had other underlying metabolic diseases (37, 38). This poses the question of whether there is a shift from SAT to VAT as the main source for TNFα expression as the metabolic disease transpires. However, even though TNFα expression has been shown to associate with VAT mass, net secretion of TNFα from VAT into the circulation has not yet been documented.

Gene expression and circulating levels of visfatin was limited associated with the amount of AT in sSAT, dSAT and VAT in our population. It has been suggested that visfatin is primarily produced by VAT and associates with intra-abdominal fat mass, but not with SAT (15). Visfatin seems to play an important role in obesity induced insulin-resistance (15). We did however, observe an inverse relationship between visfatin gene expression to insulin and C-peptide. It has been proposed that visfatin production might potentially be low in the lean state and that intra-abdominal obesity leads to increased visfatin production, which simultaneously increases with obesity, but also changes the properties of visfatin, giving it more insulin-mimetic qualities (15). This could to some extent, explain our findings with inverse correlations to insulin and C-peptide.

In this population of healthy, slightly overweight men, circulating levels of adiponectin, associated inversely with the amount of AT in all abdominal AT-compartments and to some glucometabolic variables. This may indicate that high amount of AT with simultaneous low levels of circulating adiponectin, could potentially be early signs of increased risk of developing metabolic disorders. The observed association between TNFα gene expression and the amount of abdominal AT, particularly in sSAT and dSAT, could indicate that SAT from the gluteal region also contributes to an increased pro-inflammatory state.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Regional Committees for Medical and Health Research Ethics. The patients/participants provided their written informed consent to participate in this study.

HZ conducted the study and was responsible for the statistical analyses and drafting the manuscript. TA was responsible for the main randomized trial involving recruitment, contributed to the study protocol, acquired data and contributed to the intellectual content of the manuscript. SÅ was responsible for laboratory analyses and contributed in the discussion of the manuscript. HE contributed to the manuscript. RB contributed to the manuscript. IS contributed to the intellectual content, interpretation of the results and drafting the manuscript. TB was involved in drafting of the manuscript, interpretation of the results and contributed to the intellectual content. All authors read and approved the final version.

The work was supported by Stein Erik Hagens Foundation for Clinical Heart Research, Oslo, Norway.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.