Cardiovascular risk reduction therapy in diabetic patients reduces vascular events (1) by retarding atherosclerosis and improving vascular reactivity. Matrix metalloproteinases (MMPs) and their inhibitors (TIMPs) are integral to the vascular changes of atheroma, and MMP-9 and TIMP-1 are raised in diabetes (2). We hypothesized that global risk reduction of intensified diabetes and cardiovascular risk management in type 2 diabetes would be associated with reductions in circulating markers of extracellular matrix turnover, namely MMP-9, TIMP-1, and TIMP-2.
RESEARCH DESIGN AND METHODS
Cross-sectional study
Eighty-six patients with type 2 diabetes (3) and treated hypertension with serum creatinine <120 mmol/l were recruited from local clinics (Table 1). Of these, 46% (n = 36) had a clinical history of cardiovascular disease. Data from these patients were compared with 49 healthy control subjects. Ethical approval and informed consent were obtained.
Interventional study
Sixty-five of the recruited diabetic patients (of whom 36 had cardiovascular disease) consented to intensified diabetes and cardiovascular risk management therapy. This consisted of 3 monthly consultations, lifestyle advice, and adjusting oral hypoglycemia therapy to achieve HbA1c <6.5%. Metformin (starting at 500 mg b.d.) was instigated in patients with a BMI >25 kg/m2 or added second-line in lean patients, who received initial treatment with gliclazide MR (30 mg o.d.) and vice versa. Insulin could be started if HbA1c remained >7.0% despite maximal oral hypoglycemic therapy. Blood pressure reduction was targeted at <130/80 mmHg. All patients were commenced on the ACE inhibitor perindopril (2 mg o.d.), with dose increments or other antihypertensive agents added to achieve goal blood pressure. Statin therapy and 75 mg daily aspirin was advised in all patients. HbA1c, blood pressure, lipid profile, and research indexes were measured at baseline and following 1 year of intensified therapy.
Laboratory analyses
Citrated venous blood was obtained and immediately centrifuged at 1,000g and 4°C for 20 min. Plasma was aliquoted and stored at −70°C for a batch analysis of MMP and TIMP by enzyme-linked immunosorbent assay (5).
Power calculations and statistical analysis
We hypothesized a minimum difference of half an SD between case and control subjects. To achieve this at 2p<0.025 and 1-β = 0.85 required a minimum of 44 case and control subjects. We estimated that the treatment would produce a quarter SD difference in research indexes, and at 2p < 0.025 and 1-β = 0.85, calculated a minimum requirement of 60 subjects. We used the Mann-Whitney test to compare the control subjects with the pretreatment group and the Wilcoxon’s rank-sum test for pre- and posttreatment MMP/TIMP comparisons. Correlations were performed using the Spearman test. A P value of <0.05 was taken as statistically significant.
RESULTS
Patients and control subjects were of similar age, sex, smoking status, and BMI (Table 1), but baseline systolic blood pressure and HbA1c were higher while HDL was lower in the diabetic patients.
The circulating plasma levels of MMP-9, TIMP-1, and TIMP-2 were all significantly higher in the diabetic patients (Table 1). Within diabetic patients, current smokers (n = 14) had higher circulating TIMP-1 levels than nonsmokers (510 ng/ml [335–660] vs. 390 [290–475], P = 0.017). There was a weak positive correlation between TIMP-1 and HbA1c (r = 0.214, P = 0.048) only.
Effects of intervention
Mean serum cholesterol (5.0 ± 1.0 vs. 4.3 ± 0.8 mmol/l, P < 0.0001) and HbA1c (7.7 [7.0–8.4%] vs. 7.1 [6.5–7.5], P = 0.0001) levels fell significantly; however, blood pressure or BMI did not. Baseline and 1-year usage of aspirin, metformin statins, ACE inhibitors, and additional antihypertensive treatment (i.e., β-blocker, diuretics, calcium channel blockers, α-blocker) were 42 vs. 62%, 42 vs. 68%, 24 vs. 74%, 100 vs. 100%, and 64 vs. 74%, respectively. Only TIMP-1 levels fell significantly (415 ng/ml [375–515] vs. 399 [306–485], P < 0.0001) with intervention.
CONCLUSIONS
Circulating MMP-9, TIMP-1, and TIMP-2 are raised in treated hypertensive patients with type 2 diabetes compared with normotensive control subjects, which is in keeping with extracellular matrix abnormalities present in this high-risk population.
TIMP-1 is an endogenous MMP inhibitor that may be involved in vascular matrix fibrosis (4) and has a role in left ventricular hypertrophy and diastolic dysfunction by reducing cardiac collagen type I turnover, thereby increasing cardiac mass and stiffness (4,5). Increased central and peripheral artery stiffness (6) occurs with diabetes (7), and higher circulating levels of TIMP-1 complement the hypothesis that altered TIMP-1 activities could be linked to arterial stiffness. Smokers have abnormal arterial compliance (8), and raised TIMP-1 levels could mirror increased vascular stiffness. Glucose and cholesterol reduction have both been shown to improve vascular compliance (9,10). The fall in TIMP-1 seen here by improvement in metabolic control raises the possibility of TIMP-1 as a marker of vascular composition in diabetes.
Recent reports have suggested a decrease in renal (11) and vascular tissue MMP-9 (12,13) and an increase in endothelial cell and macrophage MMP-9 in response to glucose (14). Our findings suggest a total increase in MMP-9 in diabetes and that endothelium and macrophages may be partially responsible for this rise. The failure of MMP-9 (and TIMP-2) to fall may relate to the lack of change in blood pressure during intervention. Treatment with insulin (in obese nondiabetic subjects) can reduce MMP-9 and may also explain why circulating MMP-9 did not fall in our study (15).
Diabetes is associated with abnormal angiogenesis (both impaired and/or excessive) and can impair the process of collateralization (16). TIMP-2 may attenuate angiogenesis (17). Increases in circulating TIMP-2 may reflect abnormal blood vessel formation. Manipulating this process may have some therapeutic potential.
A major limitation of this and many other studies is a lack of tissue-based measures. Unfortunately, obtaining human tissue is difficult, and venesection is often the only practical analysis. Furthermore, our patient group is heterogenous, and therefore, it is difficult to show in our group which particular intervention had the most effect. However, since most patients in routine clinical practice show similar profiles to our cohort, this point is of academic rather than clinical interest, as it is rare for such patients to present, and multifactorial intervention is the rule nowadays. This study suggests that further work in either human (in particular vascular compliance, which we were unable to measure) or animal models to further clarify the role of matrix in diabetes is clearly warranted.
Baseline characteristics of study population
| . | Patients . | Control subjects . | P . |
|---|---|---|---|
| n | 86 | 63 | — |
| Age (years) | 68 ± 6 | 66 ± 10 | 0.115* |
| Male | 67 | 53 | 0.079† |
| Cardiovascular disease | 46 | — | — |
| Current smokers | 16 | 12 | 0.575† |
| Drug therapy | |||
| ACE inhibitor | 100 | — | — |
| Aspirin | 42 | — | — |
| Statin | 24 | — | — |
| Sulphonylurea | 100 | — | — |
| Metformin | 43 | — | — |
| Glitazone | 3 | — | — |
| BMI (kg/m2) | 28 ± 4 | 27 ± 4 | 0.575* |
| Systolic blood pressure (mmHg) | 139 ± 19 | 131 ± 13 | 0.006* |
| Diastolic blood pressure (mmHg) | 77 ± 10 | 76 ± 10 | 0.578* |
| Cholesterol (mmol/l) | 5.0 ± 1.1 | 5.3 ± 1.1 | 0.065* |
| HDL (mmol/l) | 1.3 ± 0.3 | 1.6 ± 0.4 | <0.0001* |
| Triglycerides (mmol/l) | 1.6 (1.2–2.6) | 1.5 (0.8–2.2) | 0.146‡ |
| HbA1c (%) | 7.4 (6.6–8.2) | 5.4 (5.2–5.6) | <0.0001‡ |
| TIMP-2 (ng/ml) | 143 (120–160) | 110 (93–143) | <0.0001‡ |
| MMP-9 (ng/ml) | 65 (48–81) | 56 (45–75) | 0.028‡ |
| TIMP-1 (ng/ml) | 397 (300–496) | 280 (225–305) | <0.0001‡ |
| . | Patients . | Control subjects . | P . |
|---|---|---|---|
| n | 86 | 63 | — |
| Age (years) | 68 ± 6 | 66 ± 10 | 0.115* |
| Male | 67 | 53 | 0.079† |
| Cardiovascular disease | 46 | — | — |
| Current smokers | 16 | 12 | 0.575† |
| Drug therapy | |||
| ACE inhibitor | 100 | — | — |
| Aspirin | 42 | — | — |
| Statin | 24 | — | — |
| Sulphonylurea | 100 | — | — |
| Metformin | 43 | — | — |
| Glitazone | 3 | — | — |
| BMI (kg/m2) | 28 ± 4 | 27 ± 4 | 0.575* |
| Systolic blood pressure (mmHg) | 139 ± 19 | 131 ± 13 | 0.006* |
| Diastolic blood pressure (mmHg) | 77 ± 10 | 76 ± 10 | 0.578* |
| Cholesterol (mmol/l) | 5.0 ± 1.1 | 5.3 ± 1.1 | 0.065* |
| HDL (mmol/l) | 1.3 ± 0.3 | 1.6 ± 0.4 | <0.0001* |
| Triglycerides (mmol/l) | 1.6 (1.2–2.6) | 1.5 (0.8–2.2) | 0.146‡ |
| HbA1c (%) | 7.4 (6.6–8.2) | 5.4 (5.2–5.6) | <0.0001‡ |
| TIMP-2 (ng/ml) | 143 (120–160) | 110 (93–143) | <0.0001‡ |
| MMP-9 (ng/ml) | 65 (48–81) | 56 (45–75) | 0.028‡ |
| TIMP-1 (ng/ml) | 397 (300–496) | 280 (225–305) | <0.0001‡ |
Data are means ± SD, median (interquartile range), or percent.
Student’s t test
χ2 analysis
Mann-Whitney U test. To convert cholesterol and HDL from mmol to mg/dl multiply by 38.6; to convert triglycerides multiply by 88.5
Article Information
The authors acknowledge the support of the Sandwell and West Birmingham Hospitals National Health Service Trust Research and Development Programme for support of the Hemostasis Thrombosis and Vascular Biology Unit.
