Vascular Compliance Is Reduced in the Early Stages of Type 1 Diabetes

Subjects with type 1 diabetes were recruited from the Diabetes Clinic at the University of Alberta Hospital, Edmonton, Alberta, Canada. To determine the presence of microvascular complications, patients were examined by an endocrinologist, underwent a yearly ophthalmologic evaluation, and underwent urinalysis for the presence of microalbuminuria. Outpatient charts were reviewed to obtain the medical history and the method of diabetes management. Control subjects were recruited from within the hospital and had no history of diabetes, cardiovascular disease, or hypertension and were taking no medication that could potentially interfere with vascular compliance. Consent was obtained before participation in the study.

A total of 59 patients with type 1 diabetes (31 men, 28 women) were studied; age ranged from 17–61 years, and duration of diabetes was 17.7 ± 10.9 years. Of the 59 patients, 32 had no evidence of diabetic complications, whereas 27 had microvascular complications (defined as presence of retinopathy, nephropathy, or neuropathy). The control group consisted of 57 healthy subjects (27 men, 30 women) ranging in age from 23–79 years.

Large artery compliance, small artery compliance, blood pressure, and systemic vascular resistance were measured noninvasively using an HDI/Pulsewave Cardiovascular Profiling Instrument CR-2000 (Hypertension Diagnostics, Eagan, MN). Measurements were taken by a single operator after the subjects had been resting supine for 5 min in a quiet room and had not consumed any caffeine or smoked for at least 30 min. The theoretical basis and clinical validation of pulse-wave analysis have been described previously (15,17,18).

All statistical analyses were performed using SPSS statistical software (Version 10.0; SPSS, Chicago, IL). Analysis of variance was used to determine differences in characteristics between the control group, the entire diabetic group, and the subgroups of patients with and without microvascular complications. Variables tested and subsequently used in the analysis of covariance (ANCOVA), if a significant difference from the control was found, included age, gender, systolic and diastolic blood pressure, systemic vascular resistance, and smoking, because these factors have the potential to affect vascular compliance. The marginal mean age between the control group and the comparison diabetic group was used in the ANCOVA for vascular compliance when the age between these two groups was statistically different. Data are expressed as the mean ± standard deviation unless otherwise noted. Univariate ANCOVA was used to test for significant differences; P < 0.05 was considered significant. Linear regression was used to determine correlation coefficients.

The demographic data of the study participants are shown in Table 1. The diabetic patients were younger than the control subjects (31.2 ± 12.6 vs. 37.9 ± 12.6 years, P = 0.005) and had statistically higher systolic blood pressure, which was still within the accepted normal range (130.8 ± 12.6 vs. 123.1 ± 13.6 mmHg, P = 0.005).

In the diabetic group, 17 patients had a history of hypertension and 22 were taking antihypertensive or antinephropathy medications; 19 patients were taking ACE inhibitors, 2 were taking angiotensin receptor blockers, 2 were taking calcium channel blockers, 2 were taking β-blockers, 4 were taking thiazide diuretics, and 2 were taking loop diuretics.

In the control group, large artery compliance (C1) and small artery compliance (C2) were inversely proportional to age (r = −0.55 for C1 and −0.50 for C2; P < 0.01 (Fig. 1). In addition, men had greater C1 and C2 compared with women when controlled for age (C1 = 1.65 ± 0.32 vs. 1.30 ± 0.38 ml/mmHg, C2 = 0.094 ± 0.033 vs. 0.057 ± 0.021 ml/mmHg; P < 0.01, despite no differences in blood pressure or systemic vascular resistance.

ANCOVA was used to compare C1 and C2 between the control group and the diabetic group. No difference was observed between C1 or C2 in the two groups when the variables that also differed between the two groups (age, smoking, and systolic blood pressure) were added as covariables (C1, P = 0.39; C2, P = 0.06). When the diabetic patients without microvascular complications were examined, statistical differences with the control subjects were found for C1 and C2 when adjusted for age and systolic blood pressure, despite both groups being normotensive and having normal systemic vascular resistance (Table 2, Fig. 2). Using an age-adjusted model for the two groups, C1 was 1.51 ± 0.04 for the control group and 1.33 ± 0.06 ml/mmHg for the diabetic group (P = 0.03), whereas C2 was 0.080 ± 0.005 and 0.065 ± 0.005 ml/mmHg for the control and diabetic groups, respectively (P = 0.03) (Fig. 2).

No statistical differences were apparent in the diabetic subgroup with microvascular complications (C1 = 1.33 ± 0.41, C2 = 0.059 ± 0.035 ml/mmHg) compared with the control group (C1 = 1.47 ± 0.39, C2 = 0.074 ± 0.033 ml/mmHg) when blood pressure and systemic vascular resistance were added as covariables to the ANCOVA (C1, P = 0.14; C2, P = 0.23) (Table 2). Of note, 20 of the 27 diabetic patients with microvascular complications were on medical therapy for hypertension or nephropathy. Of the 20 medically treated patients, 18 were taking ACE inhibitors and 1 was taking an angiotensin receptor blocker. The seven patients who were not taking medication for hypertension or nephropathy had only peripheral neuropathy and were normotensive.

In this study, we demonstrated by pulse-wave analysis that young, normotensive subjects with type 1 diabetes have diminished large artery and small artery compliance despite having no microvascular or macrovascular complications. These findings provide further evidence that changes in the vascular wall occur before clinical complications are clinically apparent in seemingly healthy diabetic individuals. No such changes in compliance were found in patients with established microvascular complications. However, 20 of these 27 patients were taking medications that are known to affect compliance.

Several other studies have also found decreased vascular compliance in diabetic patients. Berry et al. (8) used flow velocity Doppler measurements to also show that type 1 diabetic patients without diabetic complications had a 29% reduction in systemic arterial compliance when compared with control subjects. This effect was independent of glycemic control and altered lipid levels. Echo tracking sonography has also been used to show increased aortic stiffness in women with type 1 diabetes (16). Postmortem aortic samples from patients with type 1 diabetes also show reduced extensibility and increased wall thickness (9).

Pulse waveform analysis of diabetic patients has demonstrated similar results. Duprez et al. (19) studied 31 diabetic patients on insulin therapy and found an inverse correlation between small artery compliance and duration of diabetes. No such correlation was seen for large artery compliance. McVeigh et al. (20) used intra-arterial brachial artery waveforms in 28 patients with type 2 diabetes to demonstrate diminished small artery compliance. Of these patients, 12 had retinopathy, 3 had microalbuminuria, and 2 had impaired sensation of vibration. Pulse wave velocity was also higher in type 2 diabetic patients, indicating greater arterial stiffness (21).

The patients with established microvascular disease did not demonstrate a reduction in vascular compliance. The unadjusted data for C1 and C2 seem to show lower compliance values in this diabetic group compared with the control subjects (Table 2). However, when analysis of variance is performed with blood pressure and systemic vascular resistance as covariables, there were no statistical differences in compliance between these two groups. This could represent the beneficial effect of medical therapy on the vasculature, because 74% of these patients were receiving treatment and 70% were using an ACE inhibitor or an angiotensin II receptor blocker. It has recently been shown that in hypertensive individuals on chronic antihypertensive therapy, vascular compliance is indistinguishable from normotensive subjects when measured by pulse-wave analysis (13). Treatment with ACE inhibitors is known to affect smooth muscle proliferation and improve endothelium-dependent vasodilatation by stimulating production of nitric oxide (22). Arcaro et al. (23) found improved vessel wall dilatation and endothelium mediated flow in diabetic patients randomized to treatment with captopril or enalapril. Therefore, it is not surprising that the patients with microvascular complications had vascular compliance similar to the control subjects, because most were taking medications that improve vascular compliance. However, the lack of a difference in compliance between these two groups could still be due to type 2 error, considering the small sample size of this diabetic group and the inherent added variability caused by the treatment of this group of patients with different antihypertensive or renal protective agents. Another study assessing the vascular compliance in a larger group of such subjects before the addition of medical therapy would address this issue.

The diminished vascular compliance found in diabetes is believed to be multifactorial. Hyperglycemia activates the polyol pathway and protein kinase C and causes the formation of advanced glycosylation end products (24). Each of these pathways interferes with endothelial relaxation induced by either nitric oxide or acetylcholine. Johnstone et al. (25) found diminished forearm vasodilation in response to methacholine in diabetic patients compared with control subjects. In addition, diabetes is associated with increased levels of oxygen-derived free radicals, which may inactive nitric oxide or act as a vasoconstrictor (26). Oxidized LDL also inhibits endothelium-dependent relaxation and may be a contributing factor in diabetic patients (14).

The structure of the blood vessels is also altered in patients with diabetes. There is an increase in type IV collagen, fibronectin, and laminin, which leads to changes in the elastic properties of the basement membrane. The extracellular matrix proteins become glycosylated, leading to increased collagen cross-linking (24). Furthermore, insulin is believed to contribute to smooth muscle cell proliferation and increased synthesis of extracellular matrix proteins, which could reduce vessel compliance.

There are several limitations to this study. First, the cross-sectional design did not allow us to follow the patients prospectively to determine whether compliance improved or worsened with further therapeutic intervention. Second, data were not available to assess the degree of glycemic control (HbA_1c); however, others have found that HbA_1c did not correlate with vascular compliance (8,27). Data on the lipid status of the subjects were also not available from the review of the chart. If the lipid profiles of the diabetic group were abnormal, this would only emphasize the need to aggressively control all cardiovascular risk factors, even early in the course of diabetes.

Diabetes is a risk factor for cardiovascular disease (1). We have shown that vascular compliance of both large and small vessels is reduced in patients with type 1 diabetes before any clinical complications from the disease are evident. This study serves to emphasize that vascular changes occur at an early point in the disease and may increase risk of cardiovascular events in diabetic patients. Larger prospective studies are required to confirm this finding and to investigate the efficacy of medical intervention.

We thank Dr. E. Toth for help with patient recruitment and Dr. D. Romney for assistance with the statistical analyses.