Metabolic Syndrome Accompanied by Hypercholesterolemia Is Strongly Associated With Proinflammatory State and Impairment of Fibrinolysis in Patients With Type 2 Diabetes

RESEARCH DESIGN AND METHODS

We studied 136 type 2 diabetic patients (70 female, 66 male). The patients had been referred to the diabetes outpatient clinic at the Dokkyo University Hospital to optimize glycemic control. Excluded from the study were patients with known liver disease, because TAFI is produced mainly by the liver (14). Patients with medications that could affect the coagulation or fibrinolytic systems (such as anticoagulants and antiplatelet agents) were excluded from study.

Diagnosis of MS was based on criteria in the recent Third Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) (15). MS was defined as an alteration in three or more of the following five components: obesity (BMI >25.0 kg/m²), triglycerides >150 mg/dl and/or treatment with fibrates, HDL cholesterol <50 mg/dl for women and <40 mg/dl for men, systolic blood pressure >130 mmHg, diastolic blood pressure >85 mmHg and/or antihypertensive medication, and fasting plasma glucose >110 mg/dl. We used a BMI cutoff value >25.0 kg/m² to define obesity, since the waist circumference is not suitable for detection of obesity in the Japanese population.

HC was defined as serum LDL cholesterol >140 mg/dl (3.6 mmol/l) or alternatively as treatment with a hydroxymethylglutaryl coenzyme A reductase inhibitor (statin). We used LDL cholesterol cutoff value >140 mg/dl for HC based on criteria in the Japanese Atherosclerosis Society guidelines for diagnosis and treatment of atherosclerotic CVD.

On the basis of the above two definitions, we divided the diabetic patients into four groups: group A, those with no MS and no HC (n = 38); group B, those with MS but no HC (n = 39); group C, those with no MS but with HC (n = 26); and group D, those with both MS and HC (n = 33).

CVD was defined as coronary artery disease, stroke, and peripheral vascular disease. Coronary artery disease was defined as a history of myocardial infarction, coronary artery bypass grafting, or an abnormal coronary angiography. Stroke was defined as a history of ischemic stroke confirmed by cerebral computed tomography or nuclear magnetic resonance imaging. Peripheral vascular disease was defined as a history of peripheral artery reconstruction or amputation of foot. Twelve of the diabetic patients had CVD.

Measurements

Venous blood was obtained between 6 and 7 a.m. after an overnight fast and collected in a tube containing 3.8% sodium citrate for plasma separation. Serum and plasma samples were centrifuged at 2,500 rpm for 15 min, and the supernatant was stored at −70°C until use. Plasma concentrations of TAFI were measured by a commercially available sandwich enzyme-linked immunosorbent assay (TAFI [ProCPR] ELISA kit; Institute for Protein Science, Nagoya, Japan) (16). Intra- and interassay coefficients of variation (CV) were 3.23 and 4.50%, respectively. Plasma PAI-1 was measured by a latex photometric immunoassay (LPIA tPAI-1 test; IATRON Laboratories, Tokyo, Japan). Intra- and interassay CV were 2.01 and 2.38%, respectively.

Plasma concentrations of plasmin-α2-antiplasmin complex (PAP) were determined by sandwich enzyme immunoassay (Enzygnost F1 + 2 micro, Enzygnost PAP micro; Dade Behring, Marburg, Germany). Serum concentrations of high-sensitivity C-reactive protein (hs-CRP) were determined by an immunonephelometric assay (N-high-sensitivity CRP; Dade Behring); intra- and interassay CV were 1.72 and 2.80%, respectively.

Serum total and HDL cholesterol, aspartate aminotransferase, alanine aminotransferase, and γ-glutamyltransferase were measured enzymatically using an automated analyzer. LDL cholesterol was measured directly by an enzymatic method (Cholesterol LDL; Daiichi Pure Chemical, Tokyo, Japan).

Plasma insulin concentrations were determined by radioimmunoassay. Insulin resistance was evaluated by homeostasis model assessment, calculated as fasting plasma insulin (μU/ml) × fasting plasma glucose/22.5.

Statistical analysis

Data are presented as the means ± SD or median and interquartile range. Differences in normally distributed data were assessed by a one-way ANOVA, using the Newman-Keuls multiple comparison test. For not-normally distributed data, differences between groups were analyzed by the Kruskal-Wallis with Dunn’s multiple comparison test. Correlation was determined by linear regression analysis or multivariate analysis. Logarithmic transformation of hs-CRP and urinary albumin excretion was used to render the distribution normal for parametric tests. A P value <0.05 was accepted as indicating statistical significance.

RESULTS

As shown in Table 1, linear regression analysis in 105 patients with type 2 diabetes demonstrated that plasma TAFI correlated positively with total and LDL cholesterol (P = 0.0002 and P = 0.0004, respectively), triglyceride (P = 0.0078), and hs-CRP (P = 0.0026), while plasma PAI-1 correlated positively with BMI (P < 0.0001), triglyceride (P < 0.0001), aspartate aminotransferase (P = 0.0207), alanine aminotransferase (P = 0.0003), γ-glutamyltransferase (P = 0.0012), homeostasis model assessment of insulin resistance (P < 0.0001), and hs-CRP (P < 0.0001) but negatively with plasma PAP (P < 0.0001). We also found a significant positive correlation between plasma TAFI and PAI-1 concentrations in patients with type 2 diabetes (r = 0.237, P = 0.0055).

To determine factors independently influencing plasma concentrations of TAFI or PAI-1, we performed multivariate analysis including selected significant variables. In a model that explained 46.7% of variation of plasma TAFI, only LDL cholesterol was an independent determinant of plasma TAFI in patients with type 2 diabetes (β = 0.417, P = 0.000; Table 2). In a model that explained 58.3% of variation of plasma PAI-1, only BMI was an independent determinant of plasma PAI-1 in patients with type 2 diabetes (β = 0.382, P = 0.000; Table 2). To investigate whether PAI-1 or TAFI was more strongly associated with ongoing fibrinolytic activity in vivo, we performed stepwise multivariate analysis for plasma PAP, a marker for fibrinolysis, considering BMI, renal function, PAI-1, and TAFI. PAI-1 was the strongest determinant for PAP (β = −0.420, P = 0.000), while creatinine clearance was also an independent factor of plasma PAP (β = −0.257, P = 0.0008). However, we found no significant independent relationship between PAP and TAFI.

To further investigate the relationship between TAFI and PAI-1, we then divided subjects into four groups described in research design and methods on the basis of presence or absence of MS and presence or absence of HC. We summarized patient characteristics and laboratory data of diabetic subgroups in Table 3. The prevalence of CVD was highest in patients who had both MS and HC.

Plasma concentrations of PAI-1 were significantly higher in diabetic patients with MS but no HC than in those with no MS and no HC (P < 0.001) or those with no MS but with HC (P < 0.05; Fig. 1A). Plasma concentrations of PAI were also significantly higher in patients with MS and HC than in those with no MS and no HC (P < 0.001) or those with no MS but with HC (P < 0.01). Plasma concentrations of TAFI were significantly higher in diabetic patients with no MS but with HC than in those with no MS and no HC (P < 0.01) or those with MS but no HC (P < 0.05; Fig. 1B). Plasma concentrations of TAFI also were significantly higher in those with MS and HC than in those with no MS but no HC (P < 0.001) or those with MS but no HC (P < 0.05).

Serum concentrations of hs-CRP were significantly higher in diabetic patients with MS and HC than in those with no MS and no HC (P < 0.001) or those with MS but no HC (P < 0.05) or those with no MS but with HC (P < 0.01; Fig. 2A). Thus, serum hs-CRP was highest in diabetic patients who had both MS and HC. Plasma concentrations of PAP were significantly lower in diabetic patients with MS and HC than in those with no MS and no HC (P < 0.05; Fig. 2B). Thus, plasma PAP was lowest in patients who had both MS and HC.

CONCLUSIONS

The present study demonstrated that plasma concentrations of TAFI correlated positively with serum total and LDL cholesterol in patients with type 2 diabetes. Furthermore, multivariate analysis after adjustment for triglyceride and insulin resistance showed serum LDL cholesterol to have an independent influence on plasma TAFI in type 2 diabetic patients. In a glucose clamp study (12), plasma TAFI was reported to be independently associated with glucose infusion rate, a sensitive index of insulin resistance, in type 2 diabetes. This suggested that like PAI-1, TAFI may be a component of the MS. However, another study (13) found only nonindependent association between plasma TAFI and components of MS in obese subjects. In the present study, we also could not find any significant associations between plasma TAFI and components of the MS, such as BMI, triglyceride, and HDL cholesterol, in type 2 diabetic patients. Plasma TAFI concentrations vary widely between individuals and therefore have been believed to be mainly regulated by genotype (17). The mechanism underlying association between elevated plasma TAFI and LDL cholesterol in type 2 diabetic patients remains to be determined. Malyszko et al. (18) reported that in hyperlipidemic patients undergoing renal transplantation, treatment with a statin brought about significant decreases in both plasma TAFI and LDL cholesterol, supporting present findings of a positive correlation between TAFI and LDL cholesterol.

The present study also confirmed that PAI-1 is closely associated with components of MS, in agreement with many previous studies (5,19). Multivariate analysis showed a clear relationship between plasma PAI-1 and BMI, suggesting that obesity is a main determinant of plasma PAI-1 concentrations (20). Although PAI-1 is synthesized in many tissues, the main production site of PAI-1 is the increased adipose tissue mass in subjects with MS. A previous study (21) reported that the degree of adipose tissue PAI-1 expression is related to plasma PAI-1 concentrations in human subjects. Furthermore, we found a significant positive correlation between plasma PAI-1 and circulating liver enzymes, especially alanine aminotransferase, in diabetic patients. Many recent studies (22–24) have shown a strong association between components of MS and nonalcoholic steatohepatitis. Coexistence of nonalcoholic steatohepatitis may have contributed to the association between elevated plasma PAI-1 and serum liver enzyme elevations in patients with type 2 diabetes.

We found a positive but only weak, although statistically significant, correlation between plasma TAFI and PAI-1 in diabetic patients. One possible explanation for this weak correlation between two fibrinolysis inhibitors is a difference in production sites of each fibrinolysis inhibitor. PAI is produced mainly by adipose tissues and vascular endothelial cells (21), while TAFI is produced exclusively by the liver (14). Another possibility is that plasma TAFI concentrations may be regulated more strongly by genotype rather than metabolic factors as mentioned above (17).

The present study demonstrated for the first time that diabetic patients with both MS and HC may be in a more intense inflammatory state than those with either factor alone, since these patients had the highest serum hs-CRP concentrations among the four diabetic subgroups. Ridker et al. (25) demonstrated a significant association between serum concentrations of hs-CRP and number of components of MS as defined by the National Cholesterol Education Program Adult Treatment Panel III, suggesting an increase in serum hs-CRP in subjects with MS. Our finding that diabetic patients with MS have elevated serum concentrations of hs-CRP confirms the findings of Ridker’s report.

Because of its critical importance in atherogenesis, LDL cholesterol is a focus of current guidelines for assessment of CVD. However, CVD often occurs in the absence of HC, so other preventable risk factors would appear to be at work. Several recent studies (3,4) provide clear evidence that the MS is another critical risk factor for CVD. An epidemiologic study found a much higher prevalence of coronary heart disease in type 2 diabetic patients with MS than in those without MS. MS also is associated closely with hs-CRP (26), while LDL cholesterol shows only minimal correlation with hs-CRP. Considering that chronic inflammation plays a significant role in the pathogenesis of MS as well as atherosclerosis, elevated serum CRP may contribute independently to development of CVD irrespective of LDL cholesterol levels, being associated instead with the MS. We also found a significant correlation between plasma TAFI and serum hs-CRP in diabetic patients, suggesting that elevated TAFI also may be associated with proinflammatory state. Associations between plasma TAFI and both serum hs-CRP and LDL cholesterol could account for the particularly intense proinflammatory state in our diabetic patients with both MS and HC. However, it remains unclear whether plasma TAFI is associated independently with serum hs-CRP because of the cross-sectional nature of our study.

Diabetic patients with both MS and HC have more marked impairment of fibrinolysis, since plasma PAP concentrations, a measure of ongoing fibrinolysis, were lowest in this diabetic patient subgroup. These patients therefore may tend toward a prothromobotic state. Several studies (27,28) reported that plasma PAP correlates negatively with plasma PAI-1 concentrations in human subjects. Thus, PAI-1 is the most potent inhibitor of fibrinolysis in vivo. In a previous study (29), we also found that in patients with type 2 diabetes, weight reduction by intensive metabolic control causes a decrease in plasma PAI-1, resulting in a reciprocal increase in plasma PAP. TAFI is thought to be another potent inhibitor of fibrinolysis, acting by a removal of COOH-terminal lysine and arginine residues from partially degraded fibrin. This reduces plasminogen binding to the surface of fibrin and impedes plasmin generation (7–9). However, since we found no significant correlation between plasma TAFI and PAP in diabetic patients, PAI-1 appears to be the more important determinant of fibrinolytic activity rather than TAFI. Further, our multivariate analysis identified PAI-1 but not TAFI as negatively associated with PAP in these diabetic patients. Thus, PAI-1 may be a more potent inhibitor of fibrinolysis in vivo than TAFI.

The present study clearly has limitations. A major limitation is the cross-sectional nature of the design. As a causal relationship cannot be proven by cross-sectional data, a prospective study should be undertaken to confirm causality between development of cardiovascular disease and high plasma TAFI concentrations or MS with HC.

In conclusion, LDL cholesterol proved to be a main determinant of plasma TAFI in patients with type 2 diabetes. Coexistence of MS and HC synergistically accelerates inflammation and impairs fibrinolysis in these patients via elevated concentrations of two fibrinolysis inhibitors, TAFI and PAI-1.