RESEARCH DESIGN AND METHODS

A total of 158 Japanese type 2 diabetic patients (69 men and 89 women, mean age 59.8 years) were recruited in our outpatient clinics from April 2003 to May 2004 (Table 1). The study protocol was approved by the ethics committee on human research in Kyoto Medical Center, and all participants gave written informed consent. The patients had stable and relatively high blood glucose and HbA_1c levels (7.0% ≤ HbA_1c ≤ 9.0%). They were classified by BMI (nonobese BMI <25.0 and obese BMI ≥25.0 kg/m², according to the criteria of the Japan Society for the Study of Obesity) (10); 98 and 60 were defined as nonobese and obese subjects, respectively. There were 39 men and 59 women in the nonobese group and 30 men and 30 women in the obese group. None were receiving insulin, metformin, or thiazolidinediones. Fifty-one nonobese and 25 obese subjects had been treated with sulfonylureas, whereas 47 and 35, respectively, had only been treated with diet.

Fasting plasma glucose, HbA_1c, total cholesterol, LDL cholesterol, HDL cholesterol, and triglyceride levels were measured according to standard procedures. Plasma insulin concentration was measured by enzyme immunoassay using a commercially available kit (Tosoh, Tokyo, Japan). Insulin resistance index was assessed by homeostasis model assessment (11). Plasma concentrations of adiponectin and leptin were determined using the respective radioimmunoassay kits (Linco Research, St. Charles, MO). Systolic and diastolic blood pressures were measured twice with an automatic electronic sphygmomanometer (BP-103i II; Nippon Colin, Komaki, Japan). Pulse wave velocity (PWV) (12) was determined by an automated multiple pulse wave measurement (ABI-form model: BP-203RPE; Nippon Colin). In this study, PWV was calculated as the mean of the left and right brachial-ankle PWVs, as previously described (13).

Statistical analysis

All statistical analyses were performed using the StatView program version 5.0 for Windows (SAS Institute, Cary, NC). Data are presented as means ± SE, and P < 0.05 was considered statistically significant. Differences among the nonobese and obese subjects were assessed by a two-tailed Student’s t test. Significance of correlations for leptin, adiponectin, or leptin-to-adiponectin ratio with PWV was assessed by Spearman’s rank correlation analysis.

RESULTS

There were significant differences in BMI, plasma insulin concentration, and the homeostasis model assessment of insulin resistance between the nonobese and obese subjects (P < 0.0001), with no significant differences in systolic blood pressure, diastolic blood pressure, fasting plasma glucose, HbA_1c, total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides. In all of the study subjects, leptin, adiponectin, and leptin-to-adiponectin ratio showed no significant correlation with PWV (leptin: ρ = −0.046 and P = 0.5676, adiponectin: ρ = 0.094 and P = 0.241, leptin-to-adiponectin ratio: ρ = −0.066 and P = 0.4086). In the lean subjects (BMI <25 kg/m², n = 98), when divided into men and women, there were no significant correlations of leptin, adiponectin, and leptin-to-adiponectin ratio with PWV (leptin: ρ = 0.062 and P = 0.5399, adiponectin: ρ = 0.112 and P = 0.2715, leptin-to-adiponectin ratio: ρ = −0.022 and P = 0.8297). However, in the obese subjects (BMI ≥25 kg/m², n = 60), leptin-to-adiponectin ratio was positively correlated with PWV (ρ = 0.308 and P = 0.0182), whereas leptin or adiponectin alone had no correlation (leptin: ρ = 0.255 and P = 0.0505, adiponectin: ρ = −0.130 and P = 0.3188). In the male (n = 30) and female (n = 30) patients in the obese group, leptin-to-adiponectin ratios were positively correlated with PWV (male patients: ρ = 0.378 and P = 0.0420, female patients: ρ = 0.390 and P = 0.0356), whereas leptin or adiponectin alone had no correlation in the male and female patients.

CONCLUSIONS

This study represents the first demonstration that in obese type 2 diabetic patients, leptin-to-adiponectin ratio is more strongly correlated with PWV than leptin or adiponectin alone. Since atherosclerosis at necropsy correlates closely with arterial stiffness assessed noninvasively in human and animal studies (12), the data of this study suggest that leptin-to-adiponectin ratio may serve as a potential atherogenic index in obese type 2 diabetic patients. With no significant correlation between leptin-to-adiponectin ratio and PWV in the nonobese subjects, we speculate that the vascular injury seen in these nonobese subjects may be related more strongly with other mechanisms than dysregulated production of adipocytokines. Evidence has suggested that leptin and adiponectin, both of which occur in the adipose tissue, act directly on vascular cells as proatherogenic and antiatherogenic factors, respectively (14,15), implying that they are important mediators linking adiposity and atherosclerosis in the adipovascular axis (9). Collectively, these observations suggest that the leptin-to-adiponectin ratio may serve as a potential atherogenic index in obese type 2 diabetic patients.