Results of Blood Inflammatory Markers Are Associated More Strongly With Toe-Brachial Index Than With Ankle-Brachial Index in Patients With Type 2 Diabetes

We studied 103 type 2 diabetic patients (47 women and 56 men). The diagnosis of type 2 diabetes was made according to the criteria of the World Health Organization. All patients who fulfilled the following inclusion criteria were considered for study: no episodes of ketoacidosis, diagnosis of diabetes after age >30 years, and insulin therapy, if any, started after at least 5 years of known disease.

CVD was defined as coronary artery disease and/or stroke and/or PVD. Coronary artery disease was defined as a history of myocardial infarction, coronary artery bypass grafting, or an abnormal coronary angiogram. Stroke was defined as a history of ischemic stroke confirmed by cerebral computed tomography or nuclear magnetic resonance imaging. Although PVD was defined as symptoms of intermittent claudication, a history of peripheral artery reconstruction, or the amputation of a foot, patients with lower extremity vascular surgery were excluded. Twenty-six patients had CVD.

Forty-nine of the diabetic patients had hypertension, which was defined as systolic blood pressure >140 mmHg and/or diastolic blood pressure >90 mmHg or alternatively as treatment with antihypertensive agents. These medications included an ACE inhibitor (15 patients), a calcium channel blocker (24 patients), and/or an angiotensin-receptor blocker (7 patients). All patients gave informed consent. The study was approved by the local ethics committee.

All measurements were performed in a room maintained at a temperature of 23–25°C after at least 30 min of acclimatization, with the subject in a supine position. The ABI was determined as the ratio of ankle systolic blood pressure to the brachial systolic blood pressure, with both determined using an automatic device (form PWV/ABI; AT, Komaki, Aichi, Japan). This automatic device simultaneously measured bilateral brachial and ankle blood pressure with the modified oscillometric pressure sensor method. The ABI of this device is highly correlated with that of the Doppler method (16). As brachial pressure, the higher of left and right pressures was used for calculating ABI and TBI.

To measure toe pressure, a 25-mm digital cuff was placed around the base of the great toe and a photo-transducer was taped to the pulp of the toe. The cuff was inflated to a maximum of 200 mmHg and then slowly deflated. Using a plethysmograph (EC-5R plethysmograph; DE Hokanson, Bellevue, WA), the point at which the arterial waveform reappeared was defined as the toe systolic blood pressure. The toe pressure was reported as the mean of three measurements. Similarly to ABI, TBI was calculated as the ratio of toe systolic blood pressure to brachial systolic blood pressure. All measurements were performed by an appropriately trained physician (K.O.). We evaluated the reproducibility of toe pressure measurements. The measurements were performed on two different days in a group of 10 healthy control subjects. We found a highly linear correlation between the results of the first day and those of the repeat day (r = 0.914).

Venous blood was obtained between 6:00 and 7:00 a.m. after an overnight fast. Serum concentrations of hsCRP were determined by an immunonephelometric assay (N-High-Sensitivity CRP; Dade Behring, Marburg, Germany), and the intra- and interassay coefficients of variation were 1.72 and 2.80%, respectively. Plasma concentrations of IL-6 were measured by a chemiluminescent enzyme assay (Human IL-6 CLEIA; Fujirebio, Tokyo, Japan), and the intra- and interassay coefficients of variation were 5.24 and 6.83%, respectively. Plasma concentrations of fibrinogen were determined by the Claus method. To evaluate the relationship between insulin resistance and atherosclerosis, fasting plasma C-peptide, an index of insulin resistance, was determined by radioimmunoassay. Urinary albumin excretion (concentrations in a 24-h collection specimen) was measured with an immunoturbidimetric assay.

Data are expressed as the mean ± SD or the median and interquartile ranges. Differences between groups were analyzed by a Student’s paired t test or an unpaired t test. For nonparametric data, differences between groups were analyzed by the Mann-Whitney U test. Correlation was determined by linear regression analysis. A P value <0.05 was accepted as indicating statistical significance. Statistical analyses were carried out using SPSS 8.0 J software (SPSS, Tokyo, Japan).

We divided the diabetic patients into two groups according to presence of CVD. As shown in Table 1, we found no difference in duration of diabetes, glycemic control, or blood lipid profile between diabetic patients with and without CVD. Plasma concentrations of fibrinogen were significantly higher in patients with than in those without CVD (P < 0.01). Furthermore, hsCRP and IL-6 were significantly higher in patients with than in those without CVD (P < 0.01, respectively). Although ABI was significantly lower in patients with than in those without CVD (P < 0.05), TBI was profoundly lower in patients with than in those without CVD (P < 0.001). Hypertension was more prevalent in patients with CVD (P < 0.01).

In all patients with type 2 diabetes, TBI was negatively correlated with plasma fibrinogen (r = −0.224, P = 0.0462) (Table 2). TBI also tended to correlate negatively with age. We found no significant correlation between plasma IL-6 concentrations and ABI or TBI (Table 2). Serum fasting C-peptide, a possible index of insulin resistance, showed no significant correlation with ABI or TBI. Right ABI was positively correlated with left ABI (r = 0.670, P < 0.0001) (Fig. 1A). Right TBI showed a stronger positive correlation with left TBI (r = 0.724, P < 0.0001) (Fig. 1B).

We found no significant correlation between serum hsCRP and right, left, or lower ABI (r = −0.141, P = 0.2041; r = −0.165, P = 0.1433; and r = −0.128, P = 0.2574, respectively) (Fig. 2). On the other hand, serum concentrations of hsCRP correlated negatively with right and lower TBI values (r = −0.372, P = 0.0010 and r = −0.315, P = 0.0055) (Fig. 3). Left TBI tended to correlate negatively with serum hsCRP (r = −0.215, P = 0.0633).

In diabetic patients with CVD, serum hsCRP showed no significant correlation with ABI (r = 0.012, P = 0.9602) or TBI (r = −0.188, P = 0.4278). In diabetic patients without CVD, serum hsCRP was negatively correlated with TBI (r = −0.276, P = 0.0415) but not with ABI (r = −0.146, P = 0.2672).

To determine independent influences on TBI, we performed a stepwise multivariate analysis controlling for age, diabetes duration, blood pressure, blood lipid concentrations, glycemic control, and renal function. In a model that explained 51.1% of the variation of TBI, ABI (β = 0.464, P = 0.0005) and hsCRP (β = −0.273, P = 0.020) were independent determinants of TBI in patients with type 2 diabetes (data not shown).

The present study demonstrated that in patients with type 2 diabetes, serum hsCRP concentrations show significant negative correlation with TBI but not with ABI. We also found a negative correlation between TBI and plasma fibrinogen, an acute-phase reactant, in diabetic patients. No significant correlation was evident between ABI and plasma fibrinogen. Several markers of inflammation are available for predicting the development of CVD because chronic low-grade inflammation plays an important role in the pathogenesis of atherosclerosis. Among these, hsCRP is the most promising so far because several studies have demonstrated that baseline CRP is associated with future cardiovascular events (17,18). Elevated CRP thus may be a strong independent predictor of coronary episodes. Furthermore, a prospective study (15) demonstrated that in apparently healthy men, elevated hsCRP predicts risk of subsequent symptomatic PVD. Plasma fibrinogen is also an established predictor of CVD (13). A previous study (19) identified plasma fibrinogen as an independent determinant of PVD in the elderly. The present study also indicated that both hsCRP and fibrinogen concentrations were significantly higher in patients with CVD than in those without CVD, suggesting that chronic inflammation may be an important component of atherosclerosis. On the basis of our findings, TBI may be more strongly associated than ABI with the development of atherosclerotic disease, including PVD, as estimated by markers of vascular inflammation in patients with type 2 diabetes.

We found a significant difference in age between diabetic patients with and without CVD. The patients with CVD were older than those without CVD. However, there were no significant differences in duration of diabetes and metabolic profiles between the patients with and without CVD. These results suggest that advancing age, a nonreversible factor, is a major risk factor for atherosclerotic vascular disease, irrespective of whether the subject has diabetes. In fact, there is a growing body of evidence (20) that age-associated vascular changes, including increased large artery thickening and stiffness and endothelial dysfunction, predict a higher risk for developing clinical atherosclerosis.

A low ABI (≤0.9), which is a useful diagnostic tool for detecting PVD, is also considered a strong predictor of cardiovascular morbidity and mortality in older adults (4). In short, a low ABI may reflect systemic atherosclerosis. However, because 5–10% of diabetic patients have medial arterial calcification, these patients have stiffened, noncompressible peripheral vessels (21). This yields a falsely high ABI value. Most patients with an ABI >1.3 have been shown to have such calcification (7). Instead of ABI, assessment of TBI should be used to diagnose PVD in diabetic patients, especially when the ABI is ≥1.3 (22). Because the present study showed no significant correlation between ABI values and serum hsCRP or fibrinogen when all diabetic subjects were considered, ABI does not have a straightforward association with PVD or systemic vascular inflammation in type 2 diabetes. Unfortunately, we did not obtain radiographs of the feet, so we could not assess numbers of patients with medial arterial calcification among diabetic patients. However, falsely high ABI values may account for the absence of a significant correlation between hsCRP and ABI in our diabetic patients. In these patients, TBI is a better marker of and for generalized atherosclerosis than ABI and should be substituted.

In the present study, we found no significant correlation between plasma IL-6 concentration and TBI or ABI in patients with type 2 diabetes. IL-6 is the main proinflammatory cytokine that induces acute-phase responses, including CRP and fibrinogen (10). An increased plasma IL-6 concentration is a powerful predictor of myocardial infarction in healthy individuals (11). The present study showed that plasma IL-6 was higher in patients with CVD than in those without CVD, supporting an association of plasma IL-6 with atherosclerosis. However, TBI showed a significant negative correlation with serum hsCRP but not with plasma IL-6. This suggests that CRP may be a more promising marker for atherosclerosis. Ridker et al. (17) demonstrated that in a recent analysis from the Women’s Health Study, CRP was the strongest predictor of cardiovascular events of all of the inflammatory markers, including IL-6. One possible explanation for the lack of association of IL-6 with ABI or TBI is that, unlike IL-6, CRP has a long half-life, affording stability levels with no circadian variation (23). Another possibility is that CRP per se has several direct effects on the pathogenesis of atherosclerosis (9), resulting in a significant correlation between TBI and hsCRP.

Certain limitations of the present study should be noted. As mentioned, one major limitation is the lack of radiographs of the feet to confirm the presence of medial arterial calcification in our diabetic patients. Another major limitation is the cross-sectional nature of the study design. The causal relationship cannot be proven by cross-sectional data. A prospective study should be undertaken to confirm causality between high ABI and the development of atherosclerotic disease.

We thank S. Moriyama of SRL for his kind assistance.