Elevated plasma total homocysteine (tHcy) levels are a powerful risk factor for atherosclerotic vascular disease (1), but it is still unclear by which pathophysiological mechanisms tHcy may promote atherothrombosis. In both experimental animal and cell culture studies (2,3), acute hyperhomocysteinemia induces endothelial dysfunction, leading to a low-grade inflammatory state that results in increased leukocyte adherence by upregulation of cell adhesion molecules. Accordingly, an impaired endothelium-dependent flow-mediated dilation was found in nondiabetic subjects with high tHcy when compared with subjects with low tHcy levels (4). In a recent cross-sectional study (5) of both nondiabetic individuals and type 2 diabetic subjects, tHcy was significantly associated with endothelial dysfunction, as estimated from plasma von Willebrand factor (vWF), and with leukocyte adhesion, as estimated from plasma vascular cell adhesion molecule-1. To our knowledge, there is a lack of available data regarding the relationships of tHcy levels with plasma markers of endothelial dysfunction and inflammation in young type 1 diabetic adults. It has been reported in only one previous study (6) that type 1 diabetic patients with higher tHcy levels compared with patients with lower tHcy levels had significantly elevated soluble thrombomodulin, a marker of endothelial function. However, because the group of patients with higher tHcy also had a significantly increased prevalence of microvascular and macrovascular complications, these results should be interpreted with some degree of caution. We have previously demonstrated that young type 1 diabetic patients have significantly higher tHcy levels than healthy control subjects and that smoking itself may be one of the major lifestyle determinants of tHcy (7). In this study, we endeavored to evaluate a selected group of 36 (16 men and 20 women) lean (BMI 23.6 ± 0.5 kg/m²), nonsmoking, normotensive (systolic/diastolic blood pressure 125 ± 2/80 ± 1 mmHg), normolipidemic (total cholesterol and triglycerides 4.6 ± 0.1 and 0.92 ± 0.1 mmol/l, respectively), young (age 31 ± 1 year) type 1 diabetic adults who were without any clinical evidence of macrovascular complications. Their average glycometabolic control was good (HbA_1c 6.6 ± 0.2%), and their average duration of diabetes was 15 ± 1 years. To exclude the presence of clinical macroangiopathy, a 12-lead resting electrocardiogram, a measurement of the ankle brachial pressure index, and carotid ultrasonography were performed in all of the diabetic patients. We measured plasma levels of tHcy (by an automated high-performance liquid chromatography analyzer with fluorescence detection) (7) and fibrinogen (IL-test-PT-fibrinogen HS; Instrumentation Laboratory, Lexington, KY). By using commercially available enzyme-linked immunosorbent assay kits, we measured interleukin-6 (IL-6), vWF, soluble intercellular adhesion molecule-1 (sICAM-1), P-selectin, and soluble tumor necrosis factor (TNF)-α receptors (i.e., sTNF-R1 and sTNF-R2), which reflect the degree of TNF-α activation more accurately than the measurement of TNF-α itself. The tHcy levels were significantly associated with sICAM-1 (r = 0.34, P < 0.05), vWF (r = 0.45, P < 0.01), and sTNF-R1 (r = 0.56, P < 0.001). The adjustment for potential confounders did not modify these results. The tHcy levels did not significantly correlate with fibrinogen, IL-6, P-selectin, or sTNF-R2 levels. Similarly, when diabetic patients were subdivided into groups according to the median value of the distribution of tHcy, the two groups were comparable for age, sex, BMI, lipids, creatinine, blood pressure, glycometabolic control, diabetes duration, and microvascular complications (i.e., retinopathy and/or microalbuminuria). Nevertheless, plasma levels of sICAM-1 (273 ± 11 vs. 241 ± 7 ng/ml), vWF (122 ± 8 vs. 91 ± 8%), and sTNF-R1 (2.0 ± 0.2 vs. 1.5 ± 0.1 ng/ml) were markedly elevated (P < 0.05 or less) in patients with higher tHcy (n = 18, 12.7 ± 0.8 μmol/l) versus lower tHcy levels (n = 18, 8.8 ± 0.2 μmol/l). Fibrinogen concentration tended to be higher in patients with higher tHcy (3.53 ± 0.3 vs. 3.06 ± 0.1 g/l, P = 0.08) but did not achieve statistical significance. No significant differences were found in IL-6, P-selectin, and sTNF-R2 levels between the two groups.

Overall, therefore, these results indicate that in nonsmoking, normotensive, normolipidemic young type 1 diabetic adults with good glycometabolic control and without any clinical evidence of macrovascular complications, there is a significant relationship between tHcy and plasma markers of endothelial dysfunction and inflammation. Although this study is cross-sectional and therefore cannot prove a direct cause-and-effect relationship, our results extend previous observations in nondiabetic and type 2 diabetic individuals, supporting the hypothesis that the pathophysiological link between tHcy and atherothrombosis can, at least in part, be explained by endothelial dysfunction, which leads to an increased endothelial adherence of leukocytes and a low-level chronic inflammatory state.