Methods and results Male patients with primary hypertriglyceridemia and without the metabolic syndrome ( n = 22) were compared with normotriglyceridemic individuals ( n = 25). Plasma adiponectin concentration was measured by standardised enzyme-linked immunosorbent assay. Body mass index, waist circumference, glucose, insulin and non-esterified fatty acid levels, lipoprotein profile, and CETP activity were evaluated. Adiponectin levels were significantly decreased in hypertriglyceridemic patients in comparison with normotriglyceridemic subjects (4292 ± 1717 vs. 6939 ± 3249 ng/ml, p < 0.005, respectively). Adiponectin was negatively associated with glucose ( r = − 0.44, p < 0.01), insulin ( r = − 0.37, p < 0.01), HOMA ( r = − 0.40, p < 0.01), triglycerides ( r = − 0.36, p < 0.01), VLDL-C ( r = − 0.34, p < 0.05), and CETP ( r = − 0.47, p < 0.001). Positive and significant correlations were observed with QUICKI ( r = 0.49, p < 0.001) and HDL-C ( r = 0.33, p < 0.05). In the multiple linear regression analysis, considering waist circumference, QUICKI, Log-triglycerides, HDL-C, and CETP as independent variables, Log-adiponectin showed a positive correlation with QUICKI, with an r 2 = 0.229 and p < 0.001. Therefore, the independent variable QUICKI explained the 23% of the variance in Log-adiponectin concentration. Conclusions We found low adiponectin levels in a population of primary hypertriglyceridemic men without the metabolic syndrome and an independent relationship between adiponectin concentration and insulin resistance. A reduction in insulin sensitivity and its impact on adiponectin concentration could be linked to high non-esterified fatty acid levels, increased triglyceride synthesis in the liver and impaired catabolism of triglyceride-rich lipoproteins. Keywords Hypertriglyceridemia Adiponectin Atherosclerosis Lipoproteins Introduction Adiponectin is an adipokine highly and specifically expressed by adipose cells. Like tumor necrosis factor-α, leptin and resistin, adiponectin is considered an adipose regulatory peptide. It has been shown to exert antiatherogenic and antiinflammatory activities with a direct effect on the endothelium and on foam cell formation [1] . Regarding its mechanism, adiponectin mediates fatty acid metabolism, stimulates glucose uptake [2] and modulates glucose metabolism by acting as an insulin sensitizer [3] . Moreover, peroxisome proliferator-activated receptor γ (PPAR γ) agonists have been described to increase adiponectin expression [1] . Low adiponectin levels have been described in some insulin-resistant conditions such as obesity [4] , type 2 diabetes [5] , lipodystrophy [6] , and metabolic syndrome [7] , as well as in essential hypertension [8] and coronary artery disease [9] . Even if the relationship between adiponectin expression and insulin resistance has already been reported, other studies have also shown that the inverse correlations observed between adiponectin and triglyceride, high density lipoprotein-cholesterol (HDL-C) or small and dense low density lipoprotein (LDL) levels were independent of the amount of intraabdominal fat and the degree of insulin resistance [10,11] . Moreover, Chan et al. [12] proposed that adiponectin levels would exert an independent role in regulating the metabolism of triglyceride-rich lipoproteins, evaluated through the measurement of plasma concentrations of apolipoprotein (apo) B48, apo C-III, and the cholesterol transported in remnant-like particles. The aim of the present study was to evaluate adiponectin levels in male patients with primary hypertriglyceridemia not due to the presence of acute and chronic pathologies or to the metabolic syndrome. Associations with clinical parameters, markers of insulin-resistance and lipid/lipoprotein status were also explored. Methods Subjects We studied 47 male subjects who did not present with the metabolic syndrome diagnosed employing ATPIII criteria [13] , diabetes, cardiovascular disease, or altered thyroid and renal functions, evaluated by clinical and biochemical parameters. Subjects did not present with fatty livers and hepatic function was also normal. All participants were consecutively recruited during a period of about twelve months at the Institute of Cardiology and Cardiovascular Surgery of the Favaloro Foundation, Buenos Aires, Argentina. None of them showed additional causes of dyslipidemia such as excessive tobacco (>10 cigarettes/day) or ethanol consumption (>30 g/day), therapy with drugs that could affect lipoprotein or carbohydrate metabolism and a familial history of diabetes mellitus. Subjects were classified according to their plasma triglyceride (TG) levels into two groups: subjects with high plasma TG (≥150 mg/dl) ( n = 22) and subjects with normal plasma TG (<150 mg/dl) ( n = 25). Informed consent was obtained from all participants and the protocol was approved by the Ethical Committees from Hospital de Clínicas José de San Martín, University of Buenos Aires and from the Institute of Cardiology and Cardiovascular Surgery of the Favaloro Foundation, Buenos Aires, Argentina. Study protocol and samples After a 12-h overnight fast, venous blood was drawn from the antecubital vein. Aliquots were collected in clean tubes. Samples were centrifuged at 1500 × g , for 15 min, at 4 °C, and serum was stored at 4 °C and −70 °C. Analytical procedures Glucose, triglycerides and total cholesterol were quantified by standardized methods (Roche Diagnostics, Mannheim, Germany) in a Hitachi 727 autoanalyzer. LDL-C and HDL-C levels were determined by selective precipitation methods employing polyvinylsulphate [14] and phosphotungstic acid in the presence of magnesium ions [15] , respectively. Very low density lipoprotein-cholesterol (VLDL-C) and non-HDL-C were also calculated. Apo B and apo A–I were evaluated by immunoturbidimetry (Roche Diagnostics Mannheim, Germany). Non-esterified fatty acid plasma concentration was measured by an enzymatic-colorimetric assay (Randox Laboratory Ltd., United Kingdom). Cholesteryl ester transfer protein activity Cholesteryl ester transfer protein (CETP) activity was determined in serum samples according to the general procedure previously described [16] . 3 H-cholesteryl esters-HDL 3 was employed as substrate and radioactivity was measured in the supernatant obtained after selective precipitation of apo B containing lipoproteins using a Liquid Scintillation Analyzer (Packard 210TR; Packard Instruments, Meridian, CT). Adiponectin and insulin levels Adiponectin levels were measured by monoclonal antibody-based enzyme-linked immunosorbent assay (ELISA) (R & D Systems, USA) and insulin concentration was determined by microparticle enzyme immunoassay (MEIA) (ABBOTT, Japan). HOMA (Homeostasis Model Assessment) was calculated by [Glucose (mmol l −1 ) · Insulin (uU ml −1 )]/22.5 and QUICKI (Quantitative Sensitivity Check Index) by 1/[ln Glucose (mmol l −1 ) + ln Insulin (mU l −1 )]. Data and statistical analysis Data following the Gaussian distribution were presented as the mean ± standard deviation and the Student parametric test ( T test) was used to compare the different groups, while the median (range) expression and the Mann–Whitney non-parametric test ( U test) were employed for data that did not follow the Gaussian distribution. Correlations between all variables were carried out by Pearson or Spearman tests depending on data distribution. Multiple stepwise regression analysis was performed between adiponectin as the dependent variable and some of the parameters related to insulin sensitivity and lipoprotein metabolism, which were significantly associated with adiponectin in the simple correlation analysis and/or because of their biological importance, as the independent ones. Log-transformed data were used for non-parametric variables. Differences were considered significant at p < 0.05 in the bilateral situation. Results Table 1 shows general characteristics and biomarkers of insulin resistance from primary hypertriglyceridemic patients and normotriglyceridemic controls. Age, body mass index (BMI), waist circumference and glucose levels were similar in both groups, while fasting insulin concentrations, HOMA and QUICKI were altered in patients with high triglyceride levels. Hypertriglyceridemic patients exhibited an atherogenic lipid and lipoprotein profile consisting of high VLDL-C, non-HDL-C and apo B levels, accompanied by decreased HDL-C concentration ( Table 2 ). Furthermore, patients showed higher non-esterified fatty acid levels ( Table 2 ), which were in turn directly associated with triglyceride concentration ( r = 0.56, p < 0.0001). On the other hand, CETP, a protein responsible for modulating lipoprotein composition, showed higher activity in patients than in control individuals ( Table 2 ). Adiponectin levels were significantly decreased in the group of patients with hypertriglyceridemia in comparison with normotriglyceridemic subjects (4292 ± 1717 vs. 6939 ± 3249 ng/ml, respectively; p < 0.005). Moreover, adiponectin was negatively associated with glucose ( r = − 0.44, p < 0.01), insulin ( r = − 0.37, p < 0.01), HOMA ( r = − 0.40, p < 0.01), triglycerides ( r = − 0.36, p < 0.01), VLDL-C ( r = − 0.34, p < 0.05), and CETP ( r = − 0.47, p < 0.001). Positive and significant correlations were observed with QUICKI ( r = 0.49, p < 0.001) and HDL-C ( r = 0.33, p < 0.05). No statistically significant correlation was detected between adiponectin and waist circumference ( r = − 0.12, p > 0.05) or BMI ( r = − 0.06, p > 0.05). Single correlations were corrected for multiple testing considering that 10 analyses had been carried out and, then, only CETP and QUICKI remained significant ( p < 0.005). Moreover, multiple stepwise regression analyses were performed considering adiponectin levels as the dependent variable and waist circumference, QUICKI, Log-triglycerides, HDL-C, and CETP as selected independent variables related to insulin sensitivity and lipoprotein metabolism. This analysis was also carried out alternatively excluding Log-triglycerides, HDL-C or both of them. In the first evaluation, Log-adiponectin showed a positive and significant correlation with QUICKI ( r 2 = 0.229, p < 0.001). When Log-triglycerides, HDL-C and both parameters were alternatively excluded, exactly the same results were obtained with QUICKI being the only parameter which was positively and significantly associated with Log-adiponectin ( r 2 = 0.229, p < 0.001; in the three cases). Therefore, in this regression model, the independent variable QUICKI explained the 23% of the variance in Log-adiponectin concentrations. Discussion In this study, we showed that adiponectin levels were significantly decreased in male patients with primary hypertriglyceridemia. Given that adiponectin is an antiatherogenic and antiinflammatory adipokine [17] , the alteration described here adds to the different modifications already reported in hypertriglyceridemic patients that contribute to understanding the relationship between high triglyceride levels and atherosclerosis [18–20] . Most studies reported in the literature have been carried out in patients with obesity [4] , the metabolic syndrome [7] , type 2 diabetes [5] and in HIV patients under treatment with highly active antiretrovirals (HAART) [21] , and the reduction observed in adiponectin concentration was associated with insulin resistance, hypertriglyceridemia and fat distribution. Accordingly, in this study, single correlation analysis showed that adiponectin was negatively associated with glucose, insulin, HOMA, triglycerides, VLDL-C and CETP, and directly related to QUICKI and HDL-C. When multiple testing correction was applied, only CETP and QUICKI remained significant. The inverse relationship described above, between adiponectin and CETP, disappeared in multivariate analysis taking plasma lipids into account. These findings are in agreement with the results obtained and the conclusions drawn by Dullaart RPF et al. [22] . Adiponectin could be affecting the physiological process of triglyceride–cholesteryl ester interchange via CETP, but for this measure of CETP action the level of endogenous (triglyceride-rich) plasma lipoproteins would be a more important determinant than the adiponectin level. The lack of association between adiponectin and BMI has been also reported in other studies [23] . On the other hand, even if most authors showed negative correlations between adiponectin and waist circumference, the lack of relationship observed in our study, which is in agreement with the report by Staiger et al. [24] , could be partially attributed to the fact that hypertrophic visceral tissue would downregulate adiponectin expression [25] , and waist circumference represents not only this visceral tissue but also subcutaneous adipose tissue. It must be taken into account that in our study, and in the one by Staiger et al. [24] , no obese subjects (BMI >30 kg/m 2 ) were included. The direct association between adiponectin and QUICKI was confirmed in multiple regression analysis, which showed that QUICKI independently explained the 23% of the variance in adiponectin concentration. Accordingly, Engeli et al. [26] studied a group of non-diabetic obese women and showed that insulin resistance (as indicated by HOMA) was a strong negative predictor of the expression of adiponectin mRNA, independent of other confounding obesity-associated variables. As speculated by the authors, insulin could possibly play a role in the downregulation of adiponectin expression. Nevertheless, the hypothesis that decreased adiponectin levels could lead to insulin resistance and hyperinsulinemia was not discarded by the authors [26] . In fact, Combs et al. [27] proposed that an increase in adiponectin precedes the increment in insulin sensitivity. Therefore, insulin resistance could be both a consequence and a determining factor for adiponectin expression. In addition, other factors could also be regulating adiponectin expression and explaining the remnant variance of adiponectin that was not accounted for by QUICKI, such as leptin, cortisol [28] , prolactin, growth hormone [29] and androgens [30] . In primary hypertriglyceridemic patients, low insulin sensitivity, as evidenced by low QUICKI, could be affecting the reduction in adiponectin levels [27] , which would be responsible for an increment in non-esterified fatty acid concentrations [2] . In physiologic conditions, adiponectin stimulates fatty acid oxidation by skeletal muscle through the AMP-kinase pathway, and hence serves to decrease circulating free fatty acids. High non-esterified fatty acids are known to increase triglyceride synthesis in the liver and the consequent increase in VLDL synthesis and secretion into plasma [31] . As adiponectin also stimulates lipoprotein lipase activity and reduces apo C-III expression [32,33] , a reduction in its concentration could contribute to generate the hypertriglyceridemic syndrome made up of accumulation of VLDL particles and low HDL concentration. In conclusion, we found an independent relationship between low adiponectin levels and insulin resistance in a population of primary hypertriglyceridemic male patients without the metabolic syndrome. A reduction in insulin sensitivity and its impact on adiponectin concentration could be linked to high free fatty acid levels, increased triglyceride synthesis in the liver and impaired catabolism of triglyceride-rich lipoproteins. Acknowledgements This work was supported in part by grants from Agencia Nacional de Promoción Científica y Tecnológica (PICT 14299), CONICET (PIP 6111) and the University of Buenos Aires (UBACYT B069). References [1] K.R. Rabin Y. Kamari I. Avni E. Grossman Y. Sharabi Adiponectin: linking the metabolic syndrome to its cardiovascular consequences Expert Rev Cardiovasc Ther 3 2005 465 471 [2] T. Yamauchi J. Kamon Y. Minokoshi Y. Ito H. Waki S. Uchida Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase Nat Med 8 2002 1288 1295 [3] A.H. Berg T.P. Combs X. Du M. Brownlee P.E. Scherer The adipocyte-secreted protein Acrp30 enhances hepatic insulin action Nat Med 7 2001 947 953 [4] Y. Arita S. Kihara N. Ouchi M. Takahashi K. Maeda J. Miyagawa Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity Biochem Biophys Res Commun 257 1999 79 83 [5] K. Hotta T. Funahashi Y. Arita M. Takahashi M. Matsuda Y. Okamoto Plasma concentrations of a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients Arterioscler Thromb Vasc Biol 20 2000 1595 1599 [6] W.A. Haque I. Shimomura Y. Matsuzawa A. Garg Serum adiponectin and leptin levels in patients with lipodystrophies J Clin Endocrinol Metab 87 2002 2395 2398 [7] K. Kishida H. Nagaretani H. Kondo H. Kobayashi S. Tanaka N. Maeda Disturbed secretion of mutant adiponectin associated with the metabolic syndrome Biochem Biophys Res Commun 306 2003 286 292 [8] M. Adamczak A. Wiecek T. Funahashi J. Chudek F. Kokot Y. Matsuzawa Decreased plasma adiponectin concentration in patients with essential hypertension Am J Hypertens 16 2003 72 75 [9] M. Kumada S. Kihara S. Sumitsuji S. Kawamoto S. Matsumoto N. Ouchi Osaka CAD Study Group Coronary artery disease. Association of hypoadiponectinemia with coronary artery disease in men Arterioscler Thromb Vasc Biol 23 2003 85 89 [10] M. Cnop P.J. Havel K.M. Utzschneider D.B. Carr M.K. Sinha E.J. Boyko Relationship of adiponectin to body fat distribution, insulin sensitivity and plasma lipoproteins: evidence for independent roles of age and sex Diabetologia 46 2003 459 469 [11] R. Baratta S. Amato C. Degano M.G. Farina G. Patane R. Vigneri Adiponectin relationship with lipid metabolism is independent of body fat mass: evidence from both cross-sectional and intervention studies J Clin Endocrinol Metab 89 2004 2665 2671 [12] D.C. Chan G.F. Watts T. Ng Y. Uchida N. Sakai S. Yamashita Adiponectin and other adipokines as predictors of markers of triglyceride-rich lipoprotein metabolism Clin Chem 51 2005 578 585 [13] Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) JAMA 285 2001 2486 2497 [14] G. Assmann H.U. Jabs U. Kohnert W. Nolte H. Schriewer LDL (low-density-lipoprotein) cholesterol determination in blood serum following precipitation of LDL with polyvinylsulfate Clin Chim Acta 140 1984 77 83 [15] G.R. Warnick J. Benderson J.J. Albers Dextran sulfate-Mg2+ precipitation procedure for quantitation of high-density-lipoprotein cholesterol Clin Chem 28 1982 1379 1388 [16] L. Lagrost H. Gandjini A. Athias V. Guyard-Dangremont C. Lallemant P. Gambert Influence of plasma cholesteryl ester transfer activity on the LDL and HDL distribution profiles in normolipidemic subjects Arterioscler Thromb 13 1993 815 825 [17] A.H. Berg T. Combs P.E. Scherer ACRP30/adiponectin: an adipokine regulating glucose and lipid metabolism Trends Endocrinol Metab 13 2002 84 89 [18] F. Brites C. Bonavita M. Cloës M. Yael J.C. Fruchart G. Castro VLDL compositional changes and plasma levels of triglycerides and high density lipoprotein Clin Chim Acta 269 1998 107 124 [19] F. Brites C. Bonavita C. De Geitere M. Cloës B. Delfly M. Yael Alterations in the main steps of reverse cholesterol transport in male patients with primary hypertriglyceridemia and low HDL-cholesterol levels Atherosclerosis 152 2000 181 192 [20] F. Brites J. Verona L. Schreier J.C. Fruchart G. Castro R. Wikinski Paraoxonase 1 and platelet-activating factor acetylhydrolase activities in patients with low HDL-cholesterol levels with or without primary hypertrigliceridemia Arch Med Res 35 2004 235 240 [21] C. Addy A. Gavrila S. Tsiodras K. Brodovicz A. Karchmer C. Mantzoros Hypoadiponectinemia is associated with insulin resistance, hypertriglyceridemia, and fat redistribution in human immunodeficiency virus-infected patients treated with highly active antiretroviral therapy J Clin Endocrinol Metab 88 2003 627 636 [22] R.P. Dullaart R. de Vries G.M. Dallinga-Thie A. van Tol W.J. Sluiter Plasma cholesteryl ester transfer protein mass and phospholipid transfer protein activity are associated with leptin in type 2 diabetes mellitus Biochim Biophys Acta 1771 2007 113 118 [23] E. Sobngwi V. Effoe P. Boudou D. Njamen J.F. Gautier J.C. Mbanya Waist circumference does not predict circulating adiponectin levels in sub-Saharan women Cardiovasc Diabetol 6 2007 31 [24] H. Staiger O. Tschritter J. Machann C. Thamer A. Fritsche E. Maerker Relationship of serum adiponectin and leptin concentrations with body fat distribution in humans Obes Res 11 2003 368 372 [25] C.M. Halleux M. Takahashi M.L. Delporte R. Detry T. Funahashi Y. Matsuzawa Secretion of adiponectin and regulation of apM1 gene expression in human visceral adipose tissue Biochem Biophys Res Commun 288 2001 1102 1107 [26] S. Engeli M. Feldpausch K. Gorzelniak F. Hartwig U. Heintze J. Janke Association between adiponectin and mediators of inflammation in obese women Diabetes 52 2003 942 947 [27] T.P. Combs J.A. Wagner J. Berger T. Doebber W.J. Wang B.B. Zhang Induction of adipocyte complement-related protein of 30 kilodaltons by PPAR-γ agonists: a potential mechanism of insulin sensitization Endocrinology 143 2002 998 1007 [28] A. Gavrila C.K. Peng J.L. Chan J.E. Mietus A.L. Goldberger C.S. Mantzoros Diurnal and ultradian dynamics of serum adiponectin in healthy men: comparison with leptin, circulating soluble leptin receptor, and cortisol patterns J Clin Endocrinol Metab 88 2003 2838 2843 [29] L. Nilsson N. Binart Y.M. Bohlooly M. Bramnert E. Egecioglu J. Kindblom Prolactin and growth hormone regulate adiponectin secretion and receptor expression in adipose tissue Biochem Biophys Res Commun 331 2005 1120 1126 [30] H. Nishizawa I. Shimomura K. Kishida N. Maeda H. Kuriyama H. Nageratani Androgens decrease plasma adiponectin, an insulin-sensitizing adipocyte-derived protein Diabetes 51 2002 2734 2741 [31] U. Julius Influence of plasma free fatty acids on lipoprotein synthesis and diabetic dyslipidemia Exp Clin Endocrinol Diabetes 111 2003 246 250 [32] T.P. Combs U.B. Pajvani A.H. Berg Y. Lin L.A. Jelicks M. Laplante A transgenic mouse with a deletion in the collagenous domain of adiponectin displays elevated circulating adiponectin and improved insulin sensitivity Endocrinology 145 2004 367 383 [33] T. Yamauchi J. Kamon H. Waki Y. Terauchi N. Kubota K. Hara The fat-derived hormone adiponectin reverses insulin resistance associated with both lipoatrophy and obesity Nat Med 7 2001 941 946

A Somewhat Unexpected Concavity