BRIEF REPORT

Leptin-to-Adiponectin Ratio as a Potential Atherogenic Index in Obese Type 2 Diabetic Patients

1 3 NORIKO SATOH, MD, PHD KAZUNORI YAMADA, MD, PHD written informed consent. The patients 1 3 MITSUHIDE NARUSE, MD, PHD HIDESHI KUZUYA, MD, PHD had stable and relatively high blood glu- 1 1 TAKESHI USUI, MD, PHD AKIRA SHIMATSU, MD, PHD cose and HbA levels (7.0% Յ HbA Յ 1 2,4 1c 1c TETSUYA TAGAMI, MD, PHD YOSHIHIRO OGAWA, MD, PHD 2 9.0%). They were classified by BMI TAKAYOSHI SUGANAMI, MD, PHD (nonobese BMI Ͻ25.0 and obese BMI Ն25.0 kg/m2, according to the criteria of the Japan Society for the Study of Obesity) (10); 98 and 60 were defined as nonobese besity promotes the progression of vascular remodeling and neointimal for- and obese subjects, respectively. There atherosclerosis by inducing multi- mation are markedly attenuated in leptin- were 39 men and 59 women in the non- O ple cardiovascular and metabolic deficient ob/ob mice and db/db mice with obese group and 30 men and 30 women derangements such as diabetes, hyperten- leptin receptor mutation (6,7), suggesting in the obese group. None were receiving sion, and dyslipidemia, all of which have that leptin may accelerate the develop- insulin, metformin, or thiazolidinedio- high atherogenic potential. Adipose tissue ment of vascular injury. Conversely, stud- nes. Fifty-one nonobese and 25 obese has been considered an important endo- ies with adiponectin-deficient mice have subjects had been treated with sulfonyl- crine organ that secretes many biologi- revealed that adiponectin plays a protec- ureas, whereas 47 and 35, respectively, cally active substances, collectively tive role in the development of atheroscle- had only been treated with diet. known as adipocytokines (1). Two major rosis (8,9). In obese type 2 diabetic Fasting plasma glucose, HbA1c, total adipocytokines, leptin and adiponectin, patients who are susceptible to athero- cholesterol, LDL cholesterol, HDL choles- are thought to play important roles in the sclerosis, plasma concentrations of leptin terol, and triglyceride levels were measured regulation of cardiovascular and meta- are increased, whereas those of adiponec- according to standard procedures. Plasma bolic homeostasis. Leptin acts directly on tin are decreased. We, therefore, hypoth- insulin concentration was measured by the hypothalamus, thereby regulating esize that the leptin-to-adiponectin ratio enzyme immunoassay using a commer- food intake and energy expenditure (2). serves as an atherogenic index superior to cially available kit (Tosoh, Tokyo, Japan). Plasma leptin concentrations are signifi- leptin or adiponectin alone. This study Insulin resistance index was assessed by cantly elevated in obese subjects in pro- was designed to assess the potential of the homeostasis model assessment (11). portion to the degree of adiposity (3), leptin-to-adiponectin ratio as a biomarker Plasma concentrations of adiponectin and suggesting that hyperleptinemia may play for atherosclerosis in obese type 2 diabetic leptin were determined using the respec- a role in the pathogenesis of obesity- patients. tive radioimmunoassay kits (Linco Re- related complications. On the other hand, search, St. Charles, MO). Systolic and adiponectin increases tissue fat oxidation, RESEARCH DESIGN AND diastolic blood pressures were measured leading to reduced levels of fatty acids and METHODS — A total of 158 Japanese twice with an automatic electronic sphyg- tissue triglyceride content, thus increas- type 2 diabetic patients (69 men and 89 momanometer (BP-103i II; Nippon Co- ing insulin sensitivity (4). Paradoxically, women, mean age 59.8 years) were re- lin, Komaki, Japan). Pulse wave velocity plasma adiponectin concentrations are cruited in our outpatient clinics from (PWV) (12) was determined by an auto- decreased in obese subjects (5), suggest- April 2003 to May 2004 (Table 1). The mated multiple pulse wave measurement ing that hypoadiponectinemia is involved study protocol was approved by the ethics (ABI-form model: BP-203RPE; Nippon in the pathophysiology of obesity. Two committee on human research in Kyoto Colin). In this study, PWV was calculated recent studies have demonstrated that Medical Center, and all participants gave as the mean of the left and right brachial- ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● ankle PWVs, as previously described From the 1Clinical Research Institute for Endocrine Metabolic Disease, Kyoto Medical Center, Fushimi-ku, (13). Kyoto, Japan; the 2Department of Molecular Medicine and Metabolism, Medical Research Institute, Tokyo Medical and Dental University, Chiyoda-ku, Tokyo, Japan; the 3Diabetes Center, National Hospital Orga- nization, Kyoto Medical Center, Fushimi-ku, Kyoto, Japan; and the 4Center of Excellence Program for Statistical analysis Frontier Research on Molecular Destruction and Reconstitution of Tooth and Bone, Tokyo Medical and All statistical analyses were performed us- Dental University, Chiyoda-ku, Tokyo, Japan. Address correspondence and reprint requests to Yoshihiro Ogawa, MD, PhD, Department of Molecular ing the StatView program version 5.0 for Medicine and Metabolism, Medical Research Institute, Tokyo Medical and Dental University, 2-3-10 Kanda- Windows (SAS Institute, Cary, NC). Data surugadai, Chiyoda-ku, Tokyo 101-0062, Japan. E-mail: [email protected]. are presented as means Ϯ SE, and P Ͻ Received for publication 20 March 2004 and accepted in revised form 28 June 2004. 0.05 was considered statistically signifi- Abbreviations: PWV, pulse wave velocity. A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion cant. Differences among the nonobese factors for many substances. and obese subjects were assessed by a © 2004 by the American Diabetes Association. two-tailed Student’s t test. Significance of

2488 DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004 Satoh and Associates

Table 1—Baseline characteristics of the study subjects and Spearman’s rank correlation

Nonobese (n ϭ 98) Obese (n ϭ 60) P values Variables Men Women Men Women Men Women n 39 59 30 30 Age (years) 61.7 Ϯ 1.41 60.9 Ϯ 1.04 58.0 Ϯ 2.38 57.7 Ϯ 2.35 NS NS BMI (kg/m2) 22.4 Ϯ 0.27 21.3 Ϯ 0.29 26.6 Ϯ 0.26 27.4 Ϯ 0.35 Ͻ0.0001 Ͻ0.0001 PWV (cm/s) 1667 Ϯ 40 1700 Ϯ 34 1628 Ϯ 37 1634 Ϯ 42 NS NS Leptin (ng/ml) 3.70 Ϯ 0.25 5.83 Ϯ 0.33 8.90 Ϯ 0.35 11.9 Ϯ 0.59 Ͻ0.0001 Ͻ0.0001 Adiponectin (mg/ml) 4.84 Ϯ 0.25 8.25 Ϯ 0.28 3.44 Ϯ 0.16 5.09 Ϯ 0.30 0.0003 Ͻ0.0001 Leptin/adiponectin 0.85 Ϯ 0.08 0.79 Ϯ 0.06 2.81 Ϯ 0.17 2.59 Ϯ 0.20 Ͻ0.0001 Ͻ0.0001 Spearman’s rank correlation (P values) Leptin vs. PWV Ϫ0.060 (NS) 0.053 (NS) 0.281 (NS) 0.315 (NS) Adiponectin vs. PWV 0.079 (NS) 0.090 (NS) Ϫ0.309 (NS) Ϫ0.122 (NS) Leptin/adiponectin vs. PWV Ϫ0.061 (NS) 0.013 (NS) 0.378 (P ϭ 0.0420) 0.390 (P ϭ 0.0356) Data are means Ϯ SE. NS, not significant. Two-tailed Student’s t test was used. correlations for leptin, adiponectin, or 0.0420, female patients: ␳ ϭ 0.390 and by research grants from the Smoking Research leptin-to-adiponectin ratio with PWV was P ϭ 0.0356), whereas leptin or adiponec- Foundation, the Fujisawa Foundation, the assessed by Spearman’s rank correlation tin alone had no correlation in the male Ichiro Kanehara Foundation, the ONO Medi- analysis. and female patients. cal Foundation, AstraZeneca Research (2003), the Mitsui Sumitomo Insurance Welfare Foundation, the Takeda Medical Research RESULTS — There were significant CONCLUSIONS — This study repre- Foundation, and the Takeda Science Founda- differences in BMI, plasma insulin con- sents the first demonstration that in obese tion (to Y.O.); and by Research Grant 13-C5 centration, and the homeostasis model as- type 2 diabetic patients, leptin-to- for Cardiovascular Disease from the Ministry sessment of insulin resistance between of Health, Labor and Welfare (to N.S.). Ͻ adiponectin ratio is more strongly corre- the nonobese and obese subjects (P lated with PWV than leptin or We thank Naoki Akamatsu for suggestions regarding statistical analysis. 0.0001), with no significant differences in adiponectin alone. Since atherosclerosis systolic blood pressure, diastolic blood at necropsy correlates closely with arterial pressure, fasting plasma glucose, HbA1c, stiffness assessed noninvasively in human total cholesterol, LDL cholesterol, HDL References and animal studies (12), the data of this 1. Matsuzawa Y, Funahashi T, Nakamura T: cholesterol, and triglycerides. In all of the study suggest that leptin-to-adiponectin study subjects, leptin, adiponectin, and Molecular mechanism of metabolic syn- ratio may serve as a potential atherogenic drome X: contribution of adipocytokines, leptin-to-adiponectin ratio showed no index in obese type 2 diabetic patients. adipocyte-derived bioactive substances. significant correlation with PWV (leptin: Ann N Y Acad Sci 892:146–154, 1999 ␳ ϭϪ ϭ With no significant correlation between 0.046 and P 0.5676, adiponec- leptin-to-adiponectin ratio and PWV in 2. Friedman JM, Halaas JL: Leptin and the tin: ␳ ϭ 0.094 and P ϭ 0.241, leptin-to- regulation of body weight in mammals. ␳ ϭϪ ϭ the nonobese subjects, we speculate that adiponectin ratio: 0.066 and P the vascular injury seen in these nonobese Nature 395:763–770, 1998 0.4086). In the lean subjects (BMI Ͻ25 3. Considine RV, Subha MK, Heiman ML, 2 ϭ subjects may be related more strongly Kriauciunas A, Stephens TW, Nyce MR, kg/m , n 98), when divided into men with other mechanisms than dysregulated and women, there were no significant cor- Ohannesian JP, Marco CC, McKee LJ, production of adipocytokines. Evidence Bauer TL, Caro JF: Serum immunoreac- relations of leptin, adiponectin, and lep- has suggested that leptin and adiponectin, tin-to-adiponectin ratio with PWV tive-leptin concentrations in normal- both of which occur in the adipose tissue, weight and obese humans. N Engl J Med (leptin: ␳ ϭ 0.062 and P ϭ 0.5399, adi- ␳ ϭ ϭ act directly on vascular cells as proathero- 334:292–295, 1996 ponectin: 0.112 and P 0.2715, 4. Matsuzawa Y, Funahashi T, Kihara S, Shi- ␳ ϭϪ genic and antiatherogenic factors, respec- leptin-to-adiponectin ratio: 0.022 tively (14,15), implying that they are momura I: Adiponectin and metabolic and P ϭ 0.8297). However, in the obese syndrome. Arterioscler Thromb Vasc Biol Ն 2 ϭ important mediators linking adiposity subjects (BMI 25 kg/m , n 60), lep- and atherosclerosis in the adipovascular 24:29–33, 2004 tin-to-adiponectin ratio was positively 5. Arita Y, Kihara S, Ouchi N, Takahashi M, ␳ ϭ ϭ axis (9). Collectively, these observations Maeda K, Miyagawa J, Hotta K, Shimo- correlated with PWV ( 0.308 and P suggest that the leptin-to-adiponectin ra- 0.0182), whereas leptin or adiponectin mura I, Nakamura T, Miyaoka K, ␳ ϭ tio may serve as a potential atherogenic Kuriyama H, Nishida M, Yamashita S, alone had no correlation (leptin: index in obese type 2 diabetic patients. 0.255 and P ϭ 0.0505, adiponectin: ␳ ϭ Okubo K, Matsubara K, Muraguchi M, Ohmoto Y, Funahashi T, Matsuzawa Y: Ϫ0.130 and P ϭ 0.3188). In the male ϭ ϭ Paradoxical decrease of an adipose-spe- (n 30) and female (n 30) patients in Acknowledgments— This work was sup- cific protein, adiponectin, in obesity. Bio- the obese group, leptin-to-adiponectin ported in part by a grant-in-aid for scientific chem Biophys Res Commun 257:79–83, ratios were positively correlated with research from the Ministry of Education, Cul- 1999 PWV (male patients: ␳ ϭ 0.378 and P ϭ ture, Sports, Science and Technology of Japan; 6. Stephenson K, Tunstead J, Tsai A, Gordon

DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004 2489 Leptin-to-adiponectin ratio as atherogenic index

R, Henderson S, Dansky HM: Neointimal Nagai R, Funahashi T, Matsuzawa Y: Role H, Nakao K: Antiatherogenic effect of pio- formation after endovascular arterial in- of adiponectin in preventing vascular glitazone in type 2 diabetic patients irre- jury is markedly attenuated in db/db stenosis: the missing link of adipo-vascu- spective of the responsiveness to its mice. Arterioscler Thromb Vasc Biol 23: lar axis. J Biol Chem 277:37487–37491, antidiabetic effect. Diabetes Care 26: 2027–2033, 2004 2002 2493–2499, 2003 7. Scha¨fer K, Halle M, Goeschen C, Dellas C, 10. The Examination Committee of Criteria 14. Yamagishi SI, Edelstein D, Du XL, Kaneda Pynn M, Loskutoff DJ, Konstantinides S: for “Obesity Disease” in Japan: Japan So- Y, Guzman M, Brownlee M: Leptin in- Leptin promotes vascular remodeling and ciety for the Study of Obesity: new criteria duces mitochondrial superoxide produc- neointimal growth in mice. Arterioscler for “obesity disease” in Japan. Circ J 66: tion and monocyte chemoattractant Thromb Vasc Biol 24:112–117, 2004 987–992, 2002 protein-1 expression in aortic endothelial 8. Kubota N, Terauchi Y, Yamauchi T, Kub- 11. Haffner SM, Kennedy E, Gonzalez C, cells by increasing fatty acid oxidation via ota T, Moroi M, Matsui J, Eto K, Yamashita Stern MP, Miettinen H: A prospective protein kinase A. J Biol Chem 276:25096– T, Kamon J, Satoh H, Yano W, Frougel P, analysis of the HOMA model: the Mexico 25100, 2001 Nagai R, Kimura S, Kadowaki T, Noda T: City Diabetes Study. Diabetes Care 19: 15. Ouchi N, Kihara S, Arita Y, Maeda K, Disruption of adiponectin causes insulin 1138–1141, 1996 Kuriyama H, Okamoto Y, Hotta K, resistance and neointimal formation. J Biol 12. Lehmann ED: Clinical value of aortic Nishida M, Takahashi M, Nakamura T, Chem 277:25863–25866, 2002 pulse-wave velocity measurement. Lancet Yamashita S, Funahashi T, Matsuzawa Y: 9. Matsuda M, Shimomura I, Sata M, Arita Y, 354:528–529, 1999 Novel modulator for endothelial adhesion Nishida M, Maeda N, Kumada M, Oka- 13. Satoh N, Ogawa Y, Usui T, Tagami T, molecules: adipocyte-derived plasma moto Y, Nagaretani H, Nishizawa H, Kohno S, Uesugi H, Sugiyama H, Sug- protein adiponectin. Circulation 100: Kishida K, Komuro R, Ouchi N, Kihara S, awara A, Yamada K, Shimatsu A, Kuzuya 2473–2476, 1999

2490 DIABETES CARE, VOLUME 27, NUMBER 10, OCTOBER 2004