Results
Subject Characteristics
Subject characteristics are detailed in Table 1. Controls and subjects with diabetes were well matched in terms of age, sex, height, smoking status, alcohol consumption, pulse rate, diastolic blood pressure, mean arterial pressure, total cholesterol, HDL-cholesterol, LDL-cholesterol, total cholesterol:HDL ratio, triglyceride and renal function (p>0.05 for all comparisons). Subjects with diabetes were heavier (mean weight 81.49 (SE 2.52) kg vs. 70.79 (2.15) kg, p=0.02), with a higher BMI (26.41 (0.73) kgmvs. 23.87 (0.66) kgm, p=0.011) than controls. More subjects with diabetes were taking aspirin (17.9 vs. 0 %, p=0.012) or an HMG-CoA reductase inhibitor (25.6 vs. 0 %, p=0.001). Subjects with diabetes had a higher systolic blood pressure (123.3 (1.7) mmHg vs. 115.3 (1.7) mmHg, p=0.002), than control subjects but this was well within what would currently be considered 'normal' limits.
Common Carotid Artery Waveform Comparisons
The mean amplitude was higher for subjects with type 1 diabetes than controls in band 7 (median 0.42 (IQR 0.28–0.62) cms vs. 0.38 (0.26–0.63) cms, p=0.032), and lower in band 11 (mean 46.61 (SE 1.48) cmsvs. 52.88 (1.73) cms, p=0.007). The mean velocity was lower in those with diabetes (49.15 (1.39) cmsvs. 55.00 (1.52) cms, p=0.006). There was a trend towards lower amplitude in band 6 for subjects with diabetes (p=0.088). The RI was higher in those with diabetes (0.70 (0.66–0.73) vs. 0.68 (0.62–0.70), p=0.020). These data are presented in Figure 2 and Table 3.
(Enlarge Image)
Figure 2.
Significant differences between blood velocity waveforms from subjects with type 1 diabetes and controls in (a) common carotid artery, (b) ophthalmic artery, and (c) central retinal artery. Centre of box = median; extremities of box = interquartile range; whiskers = range. * p<0.05, **p<0.01
Ophthalmic Artery Waveform Comparisons
The mean amplitude was higher for subjects with type 1 diabetes than controls in bands 1 (median 0.24 (IQR 0.20–0.29) cmsvs. 0.19 (0.16–0.24) cms, p=0.007), 2 (0.62 (0.45–0.74) cmsvs. 0.53 (0.40–0.62) cms, p=0.025) and 4 (2.30 (1.55–2.88) cmsvs. 1.76 (1.38–2.18) cms, p=0.022). It was lower in band 7 (0.55 (0.36–0.79) cmsvs. 0.73 (0.54–1.13) cms, p=0.020). There was a trend towards lower amplitude for subjects with diabetes in band 3 (p=0.091). No differences were apparent when mean velocity or time domain analyses were compared. These data are presented in Figure 2 and Table 3.
Central Retinal Artery Waveform Comparisons
The mean amplitude was higher for subjects with type 1 diabetes than controls in band 1 (median 0.18 (IQR 0.16–0.25) cmsvs. 0.16 (0.13–0.21) cms, p=0.020), band 2 (0.38 (0.30–0.51) cmsvs. 0.31 (0.26–0.41) cms, p=0.012), band 3 (0.70 (0.56–0.84) cmsvs. 0.49 (0.45–0.61) cms, p=0.000) and band 5 (1.37 (1.08–1.89) cmsvs. 1.25 (=(0.97–1.40) cms, p=0.028). There was a trend towards higher amplitude for subjects with diabetes in band 4 (p=0.066). No differences were apparent when mean velocity or time domain analyses were compared. These data are presented in Figure 2 and Table 3.
Effects of Confounding
After taking into account the effect of the between group differences in systolic blood pressure and BMI, differences in the following parameters lost significance when compared between groups: CCA amplitude in bands 7 (p=0.159) and 11 (p=0.067), CCA RI (p=0.079), OA amplitude in bands 1 (p=0.120), 2 (p=0.061) and 7 (p=0.068), and CRA amplitude in band 1 (p=0.065). All other parameters remained significantly different between groups. Since no control subjects were taking aspirin or a HMG-CoA reductase inhibitor, it was impossible to meaningfully control for these confounders. Any differences seen between groups should therefore be interpreted in the knowledge that 18% of those with diabetes were on aspirin and 26% were on an HMG-CoA reductase inhibitor.