Intended for healthcare professionals


Endothelium and inelastic arteries: an early marker of vascular dysfunction in non-insulin dependent diabetes

BMJ 1996; 312 doi: (Published 23 March 1996) Cite this as: BMJ 1996;312:744
  1. Jonathan Goodfellow, research fellowa,
  2. Mark W Ramsey, lecturera,
  3. Lisa A Luddington, research nursea,
  4. Christopher J H Jones, lecturera,
  5. Phillip A Coates, senior registrar (medicine)b,
  6. Frank Dunstan, lecturerc,
  7. Malcolm J Lewis, professora,
  8. David R Owens, senior lecturerb,
  9. Andrew H Henderson, professora
  1. a Cardiovascular Sciences Research Group, University of Wales College of Medicine, Cardiff CF4 4XN
  2. b Diabetes Research Unit, University of Wales College of Medicine
  3. c Department of Medical Statistics and Computing, University of Wales College of Medicine
  1. Correspondence to: Dr Goodfellow.
  • Accepted 9 January 1996

Atheroma is the main cause of mortality and morbidity in patients with non-insulin dependent diabetes mellitus. Endothelial dysfunction underlies atheroma formation. Flow mediated vasodilatation is a measure of endothelial function as well as a determinant of arterial distensibility,1 measurements of which have not previously been reported in patients with non-insulin dependent diabetes.

Subjects, methods, and results

We studied 12 symptom free, non-smoking, healthy patients (six men) with well controlled non-insulin dependent diabetes diabnosed 1-7 (mean 3.8) years previously (mean age 50 (SD 9) years, body mass index 26.9 (4.8) kg/m2, serum cholesterol concentration 5.8 (0.5) mmol/l). Patients with known causes of endothelial dysfunction were excluded. All patients had insulin resistance with basal hyperinsulinaemia,2 no biochemical evidence of renal impairment, and no microalbuminuria. A group of 12 normal non-smoking subjects was matched for sex, age, body mass index, blood pressure, and serum cholesterol concentration and served as a control group.

Artery diameter (D) and distensibility were measured by high resolution ultrasonic vessel wall tracking1; blood pressure (P) by finger photo-plethysmography; blood flow by ultrasound as the product of mean velocity corrected for Doppler angle and internal brachial artery diameter. Distensibility (δD/D.δ}P) was derived from systolic distension (systolic minus diastolic intravascular diameter, δD) and simultaneous pulse pressure ({δ}P).

Fasting subjects were studied in a temperature controlled room (21-23°C) after 15 minutes of supine rest. Measurements were made at baseline, during reactive hyperaemia 60 seconds after deflation of a wrist cuff inflated to suprasystolic pressure for five minutes (flow related, endothelium dependent vasodilatation) three minutes after 10 µg (submaximal dose) sublingual glyceryl trinitrate and three minutes after 400 µg (supramaximal dose) sublingual glyceryl trinitrate (endothelium independent vasodilatation). Haemodynamic variables returned to baseline 15 minutes after each intervention.

Groups were compared with unpaired t tests, apart from the glyceryl trinitrate distensibility data, which failed a test for normality and were analysed with the Mann-Whitney test. Paired t tests were used within groups to assess changes from baseline. P < 0.05 was regarded as significant.

Baseline characteristics and resting brachial artery measurements were similar between diabetic and normal subjects (heart rate 61 (SD 4) v 56 (8) per minute; blood pressure 127 (12)/69 (13) v 124 (15)/63 (9) mm Hg; blood flow 45 (28) v 49 (41) ml/min; diastolic diameter 4.82 (0.60) v 4.47 (1.05) mm; distensibility 4.8 (1.4) v 5.9 (1.6) per kPa). During reactive hyperaemia, heart rate, blood pressure, and the increase in brachial artery blood flow (554.8% (99.5%) v 590% (106.6%)) were similar in diabetic and normal subjects (figure 1) but the increases in brachial artery diastolic diameter (1.2% (2.7%) v 9.1% (4.4%); difference 6.9% (95% confidence interval 4.9% to 11%)) and distensibility (-15.3% (10.4%) v 31% (31.7%); difference 46.3% (25.5% to 67.1%)) were significantly less in diabetic subjects (both P<0.001). After 10 µg glyceryl trinitrate or 400 µg glyceryl trinitrate the increase in flow, diameter, and distensibility was similar in diabetic and normal subjects (figure 1).

Fig 1
Fig 1

Mean changes in blood flow, diastolic diameter, and distensibility of brachial artery in 12 diabetic subjects and 12 controls during reactive hyperaemia, 10 µg glyceryl trinitrate, and 400 µg glyceryl trinitrate. Bars show 95% confidence intervals


Endothelium dependent, flow related dilatation and increase in distensibility of the brachial artery are greatly impaired in patients with non-insulin dependent diabetes, but endothelium independent responses induced by glyceryl trinitrate are similar to those in normal subjects. Loss of flow related increase in arterial distensibility will augment systolic pressure, myocardial wall stress, and heart work relative to stroke output, potentially promoting left ventricular hypertrophy and lowering ischaemic threshold. Late systolic pressure will be further augmented by early wave reflection from the periphery3 because pulse wave velocity is increased when distensibility is reduced. Loss of flow mediated vasodilatation reflects endothelial dysfunction and may thus also provide a marker of atherogenetic susceptibility. Our data provide evidence of vascular dysfunction in non-insulin dependent diabetes before the appearance of microalbuminuria, previously regarded as its earliest marker.4

We thank Dr J R Peters for allowing us to study patients under his care and Wendy Simons and Julie-Ann Davies for their secretarial assistance.


  • Funding British Heart Foundation.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
View Abstract