In this study, 61 patients with IGT (aged, 49.8±4.8 years; 7.8–11.1 mmol/l OGTT2h) were selected from 129 individuals who were at a high-risk of DM according to the 2004 guidelines of China diabetes prevention and control (15). Patients were selected from two residential communities (the Cangshan and Gushan districts of Fuzhou, the capital city of Fujian province, China). Twenty-two patients with NGT were chosen as the controls. The procedures of the present study were approved by the Institutional Research Ethics Board of Fujian Normal University and Fujian Medical University (Fuzhou, China). All participants of this study understood the experimental procedures and provided written informed consent.

The OGTT was performed for the diagnosis of NGT and IGT according to the 2004 guidelines of China diabetes prevention and control. Diagnosis was based on the OGTT 2 h following the oral administration of 75 g glucose (OGTT2h) (16). Briefly, all participants underwent 10–16 h of fasting overnight and the OGTT commenced at ~8:00–9:00 a.m. following a 30 min rest at 20–25ºC. Following the extraction of blood samples (5 ml) from patients in a fasted state, the patients orally received 75 g of anhydrous glucose in 250–300 ml warm water over 3–5 min. Second blood samples were collected 2 h following the sugar intake.

Measurements of morphological parameters, including BMI and body fat percentage, were made using a body composition resistance measurement instrument (Biospace InBody 3.0; Biospace Co., Ltd. Seoul, South Korea), according to the national physiological constitution monitoring requirements (15).

The biochemical indices were detected using various kits. Briefly, the blood sugar concentrations were measured immediately following the collection of blood. The biochemical indices, including C-type natriuretic peptide (CNP), endothelin-1 (ET-1), insulin, leptin, resistin, adiponectin, tumor necrosis factor-α (TNF-α) and free fatty acid levels, were detected by radioimmunoassay (RIA). Blood sugar concentrations were analyzed by the glucose oxidase method using a fully automatic biochemical analyzer (Hitachi 7020; Hitachi, Tokyo, Japan). The insulin RIA kit [intra-assay coefficient of variation (cv) <4.30% and inter-assay cv <7.10%], leptin RIA kit (intra-assay cv <10.00% and inter-assay cv <15.00%) and TNF-α RIA kit (intra-assay cv <5.00% and inter-assay cv <8.00%) were purchased from Atomic Gaoke Co., Ltd., Department of Isotope, China Institute of Atomic Energy (Beijing, China). The resistin RIA kit (intra-assay cv <3.59% and inter-assay cv <9.25%) and the adiponectin RIA kit (intra-assay cv <3.59% and inter-assay cv <9.25%) were purchased from Linco Research Inc. (St. Louis, MO, USA). The ET-1 RIA kit (intra-assay cv <10.00% and inter-assay cv <15.00%) and CNP RIA kit (intra-assay cv <10.00% and inter-assay cv < 5.00%) were purchased from the General Hospital of PLA from the Institute of Science and Technology Development Center (Haidian, China). The free fatty acid enzyme-linked immunosorbent assay kit was provided by Randox Laboratories Ltd., (Antrim, UK). All measurements were performed in strict accordance with the kit instructions. Homeostasis model assessment-insulin resistance (HOMA-IR) was calculated with the following formula: HOMA-IR = fasting plasma insulin concentrations (mIU/l) × fasting blood glucose (mmol/l)/22.5 (17).

Evaluation of the carotid artery structure and function was conducted using the Celermajer method (17,18). Briefly, endothelium-dependent vasodilation of the brachial artery was measured using color Doppler supersonic diagnostic equipment (iU-22; Phillips, Amsterdam, The Netherlands) with a L17–5 MHz transducer. Endothelium-independent vasodilation was calculated using the following equation: ΔDia-N (%) = (artery diameter with 0.4 mg sublingual nitroglycerin 1 min - basic diameter)/basic diameter ×100. Brachial artery endothelium-dependent vasodilation was calculated as follows: ΔDia-P (%) = (maximum diameter inside brachial artery with 1 min decompression - basic diameter)/basic diameter ×100 (17).

All experimental data are expressed as the mean ± standard deviation. Paired t-test was used to evaluate statistically significant differences. SPSS software, version 11.5 (SPSS, Inc., Chicago, IL, USA) was used for Pearson correlation analysis and regression analysis of vascular endothelium functions with multivariate step-by-step regression analysis. P<0.05 was considered to indicate a statistically significant difference and P<0.01 was considered to indicate a marked statistically significant difference.

To determine whether vascular endothelial functions were impaired in the IGT group, physiological and biochemical indices were analyzed. The CNP and ET-1 concentrations were detected. The carotid artery intima-media thickness (IMT) and basic diameter of the carotid artery were measured. Endothelium-dependent vasodilation (ΔDia-P) and endothelium-independent vasodilation (ΔDia-N) were calculated in the NGT and IGT groups. The results showed a significant difference in endothelium-dependent vasodilation, but not in the endothelium-independent vasodilation between the NGT and IGT groups (Table I). The increase in the carotid artery IMT in the IGT group confirmed our speculation regarding the effects of IGT on endothelium-dependent vasodilation. To further understand the changes of vasodilation that may result from endothelial dysfunction, the endothelium-derived factors CNP and ET-1 were tested, and the basic diameter of the carotid artery was measured. CNP and ET-1 were identified to be significantly impacted in the IGT group (Table I), indicating endothelium dysfunction. These results suggest that vascular endothelial functions are impaired in patients with IGT.

To investigate the molecular mechanisms involved in the changes of endothelium-dependent vasodilation in patients with IGT, fat and glucose metabolism were also measured in the NGT and IGT groups. The results showed significantly higher expression levels of leptin, resistin and free fatty acids in the IGT group compared with those in the NGT group, and a reduction in the adiponectin level in the IGT group compared with that of the NGT group (Table I), implying the possible involvement of these fat metabolism factors in vascular endothelial function. Concentrations of TNF-α did not show a significant difference between the two groups (Table I).

Analysis of HOMA-IR and glucose metabolism showed that there were also significant differences in fasting blood glucose, OGTT2h, fasting insulin and HOMA-IR levels between the NGT and IGT groups (Table I). Analysis of body composition in the NGT and IGT groups also showed significant differences in BMI, waist circumference, waist to hip ratio and body fat percentage (Table I). These results suggest that adipocytokines, fat and glucose metabolism and body composition were abnormal in the IGT group compared with those in the NGT group.

To further investigate the factors affecting vascular endothelial functions in patients with IGT, correlation analysis of adipocytokines and fat metabolism was conducted. This showed a significant correlation of CNP with leptin and resistin levels (Table II), and a significant correlation between ET-1 and adiponectin concentrations (Table II) in the IGT group, demonstrating that fat metabolism factors may also affect vascular endothelial functions in IGT patients. Furthermore, Pearson regression analysis was performed and the results indicated that adiponectin and OGTT2h were the major factors in the CNP regression equation (Table III). CNP regression equation was as follows: Y_CNP = 3.856 × X_adiponectin − 2.273 × X_OGTT2h + 2.263.

Correlation analysis of HOMA-IR and glucose metabolism showed that there was a significant correlation between ET-1 and fasting insulin, OGTT2h and HOMA-IR (Table II), and carotid artery IMT was significantly correlated with HOMA-IR in the IGT group (Table II). The results of the Pearson regression analysis of ΔDia-P showed that OGTT2h and adiponectin were the major factors affecting ΔDia-P in the IGT group (Table III). ΔDia-P regression equation was as follows: Y_ΔDia-P = 0.981 × X_OGTT2h – 1.036 × X_adiponectin + 7 .608

The results of the correlation analysis also showed that there was a significant correlation of body fat percentage with ET-1 and carotid artery IMT in the IGT group (Table II). The results of the Pearson regression analysis of ET-1 showed that OGTT2h, fasting insulin levels and waist circumference were involved in the ET-1 regression equation (Table III). ET-1 regression equation was as follows: Y_ET-1 = 26.617 × X_OGTT2h +2.273 × X_{fasting insulin} − 2.562 × X_{waist circumference} + 65.661.

Pearson regression analysis of carotid artery IMT indicated that body fat percentage was the major factor affecting carotid artery IMT in the IGT group (Table III) and was calculated as follows: Y_{Carotid artery IMT} = 2.059 × X_{body fat percentage} + 2.059. The above results suggest that vascular endothelial functions are affected by adipocytokines, glucose and fat metabolism and body composition.