This was a single center, retrospective study performed at Huh's Diabetes Center in Seoul, Republic of Korea. Among the patients who visited this clinic from January 2003 to December 2006, we analyzed 324 T2DM patients over 50 years of age, and their vascular stenosis was concurrently evaluated by the ABI, color Doppler ultrasonography, and CIMT. Subjects who had a previous history of ketoacidosis, serum fasting C-peptide level ≤0.5 ng/mL, history of cardiovascular disease (CVD) including stroke, myocardial infarction, any coronary revascularization or surgery due to chest pain and hospitalization due to heart failure, history of PAD, or a diagnosis of an orthopedic disease in the lumbar spine were excluded. All subjects were categorized into three groups according to their leg symptoms: the asymptomatic, typical, and atypical symptomatic groups. The typical symptomatic group included patients with symptomatic claudication, resting pain, and necrosis or gangrene, which are compatible with typical symptoms in the Fontaine staging system [16]. The atypical symptomatic group included patients with vague pain, which was difficult to distinguish from symptoms of peripheral neuropathy. This study was approved by the Institutional Review Board of Yonsei University (IRB No. 4-2017-0288). Due to the retrospective nature of this study, the requirement for written informed consent was waived.

Height (cm) and body weight (kg) were measured to the nearest 0.1 cm and 0.1 kg, respectively, in barefoot subjects wearing light clothing. Body mass index (BMI) was calculated as weight divided by height squared (kg/m²). Waist circumference was measured to the nearest 0.1 cm at the narrowest point between the lower limit of the ribcage and the iliac crest. Blood pressure (BP) was recorded using a mercury sphygmomanometer on the right arm with the subject in the sitting position. Two readings were taken 5 minutes apart, and the mean value was used. All patients were asked about their smoking history and were classified as never-smokers, ex-smokers, or current smokers using a structured questionnaire. Subjects were also asked to refrain from smoking for 24 hours and from consuming alcohol for 7 days before the blood sampling.

After a 14-hour overnight fast, venous blood samples were drawn from the antecubital vein between 7:00 AM and 9:00 AM. Plasma was separated immediately by centrifugation (2,000 rpm, 20 minutes, at 4℃), and biochemical measurements were determined immediately. Fasting plasma concentrations of glucose, total cholesterol, triglycerides, and high density lipoprotein cholesterol and low density lipoprotein cholesterol were measured enzymatically using the Hitachi 747 chemical analyzer (Hitachi, Tokyo, Japan). The glycosylated hemoglobin (HbA1c) value was determined using a high-performance liquid chromatography method (Varient II; Green-cross, Seoul, Korea). Fasting and 2-hour postprandial C-peptide levels were measured by double-antibody radioimmunoassay (DiaSorin, Stillwater, MN, USA).

A non-invasive screening device (VP-1000; Colin Medical Technology Corp., Komaki, Japan) was used to measure the ABI. All patients were allowed to rest for at least 5 minutes before the measurements. The American College of Cardiology/American Heart Association (ACC/AHA) guidelines [17] recommend selecting the higher of the two arm pressures (brachial) and the higher of the two ankle pressures (anterior tibial/dorsalis pedis or posterior tibial) to calculate the ABI. The lower of the two ABI values was used for the analysis. According to the ACC/AHA recommendations the ABI value was classified as follows: normal, 1.00 to 1.40; borderline, 0.91 to 0.99; abnormal, ≤0.90 (mild, 0.70 to 0.90; moderate, 0.40 to 0.69; severe, <0.40); and poorly compressible >1.40 [17].

All Doppler studies were performed by a single radiologist using color Doppler ultrasonography (Logic 7; General Electronics, Milwaukee, WI, USA). Information on the presence and grade of stenosis, post-stenotic turbulence, peak systolic velocity, end diastolic velocity, and the presence of collaterals and their flow characteristics were assessed on color Doppler imaging. The degree of stenosis was graded according to Jager's criteria. Normal, triphasic waveform with thin spectral band. Grade I, 1% to 19% stenosis: normal triphasic flow with normal peak systolic velocity with spectral broadening. Grade II, 20% to 49% stenosis: triphasic waveform with an increase in peak systolic velocity ≥30% with respect to the proximal recording site. Marked spectral broadening. Grade III, 50% to 99% stenosis: monophasic waveform with an increase in peak systolic velocity ≥100% and marked spectral broadening. Distal waveform is abnormal. Grade IV, 100% stenosis: no forward flow detected with altered flow patterns both proximal and distal to the stenosis [18]. Grade III and IV stenoses were defined as PAD in this study.

Both carotid arteries were scanned by a single experienced examiner using high-resolution real-time B mode ultrasonography with a 10 MHZ probe (LOGIQ 7; General Electronics), as previously described [19]. The CIMT was defined as the distance between the lumen-intima and media-adventitia interfaces [20]. CIMT was measured at three points on the far wall in sections (1 cm long) of the common carotid artery in the proximity of the carotid bulb. The mean of six measurements from the right and left carotid arteries was used as the mean CIMT [21].

Normally distributed data are expressed as mean±standard deviation, and categorical variables are expressed as percentages. One-way analysis of variance or Student t-test was used to compare continuous variables, and the chi-square test was used to compare categorical variables. Logistic regression analyses were performed using the enter method to identify the risk factors for PAD in patients with ABI 0.91 to 1.40. Receiver operating characteristic (ROC) analysis was performed to assess the utility of CIMT in predicting PAD in symptomatic and asymptomatic patients with ABI 0.91 to 1.40. All statistical tests were two-sided, and analyses were executed using SPSS for Windows version 23.0 (IBM Co., Armonk, NY, USA). A P value <0.05 was considered significant.

A total of 324 patients were analyzed in this study (male: female=192:132). The mean age of all subjects was 62.5±7.3 years, and BMI was 24.5±2.9 kg/m². Mean DM duration was 11.4±7.3 years, and the mean HbA1c was 8.4%±1.8%.

Table 1 shows the clinical characteristics of the subjects with and without PAD. The PAD group was older, more likely male, and lower in BMI. The PAD group had more current smokers and a longer duration of DM. Other metabolic parameters, including fasting plasma glucose levels, lipid levels, and waist circumference, did not differ between the groups. Systolic and diastolic BP was higher, and leg symptoms were more common, in the PAD group. The proportions of abnormal or borderline ABI were higher in PAD group, compared with non-PAD group. The mean CIMT was also significantly higher in the PAD group, compared with non-PAD group.

The clinical presentation of PAD according to the ABI and color Doppler ultrasonography results is detailed in Fig. 1. Among the 324 patients, only 18 (5.6%) had an abnormally low ABI (≤0.90) and the other 306 (94.4%) had ABI 0.91 to 1.40. However, color Doppler ultrasonography demonstrated that 93 (28.7%) had PAD. Among the 93 patients diagnosed with PAD by Doppler ultrasonography, only 16 (17.2%) had an abnormally low ABI, and 77 (82.8%) had ABI 0.91 to 1.40. Among the 306 patients with ABI 0.91 to 1.40, 34 (11.1%) had grade III (50% to 99% stenosis) lesions, and 43 (14.1%) had grade IV (100% stenosis) lesions on color Doppler ultrasonography (Fig. 2). These results suggest that the ABI has the potential to underestimate the presence of PAD.

Next, we compared the characteristics of the stenotic lesions between ABI 0.91 to 1.40 and ABI ≤0.90 among the patients with PAD (Table 2). Multiple stenotic lesions were less frequent in ABI 0.91 to 1.40 group than in ABI ≤0.90 group. Although stenotic lesions occurred most frequently in the infrapopliteal arteries, such as the anterior and posterior tibial arteries, in both groups, stenosis of the common femoral (14.3% vs. 37.5%, P=0.029) or superficial femoral (22.1% vs. 32.0%, P=0.022) arteries was less frequent in ABI 0.91 to 1.40 group than in ABI ≤0.90 group. Bilateral stenosis was also less common in ABI 0.91 to 1.40 group than in ABI ≤0.90 group (64.9% vs. 93.8%, P=0.022). Collectively, multiple, proximal, and bilateral arterial stenoses were less common in ABI 0.91 to 1.40 group than in ABI ≤0.90 group, suggesting that the ABI alone potentially underestimates PAD if the stenotic lesions are less extensive, distal, or unilateral.

We compared the clinical characteristics between patients with ABI 0.91 to 1.40 versus ≤0.90 among patients with PAD (Table 3). No significant differences in age, sex, or BMI were observed between the groups. In addition, no significant differences in BP, HbA1c, lipid profile, or smoking history were observed between the groups. However, ABI ≤0.90 group had a longer duration of DM, more frequent insulin use, and more frequent leg symptoms compared with those in ABI 0.91 to 1.40 group. Although the mean CIMT was significantly higher in patients with PAD than in those without PAD (1.03±0.27 mm vs. 0.90±0.19 mm, P<0.001) (Table 1), no difference in CIMT was observed between the patients with ABI 0.91 to 1.40 and ≤0.90 among the patients with PAD (Table 3).

The ABI is a more accessible tool for screening PAD in the clinical setting than is color Doppler ultrasonography. However, it has the potential to underestimate the presence of PAD, which may delay proper management. If we could find out the characteristics of the patients with PAD who had ABI 0.91 to 1.40, further vascular evaluation could be recommended for them. Thus, we investigated the risk factors associated with the presence of PAD in the 306 patients with ABI 0.91 to 1.40. Univariate analysis demonstrated that older age, a longer duration of DM, current smoking status, leg symptoms, and a higher mean CIMT were associated with the presence of PAD in patients with ABI 0.91 to 1.40 (Table 4). In the multivariate analysis, we adjusted for age, sex, BMI, systolic BP, DM duration, current smoking status, HbA1c, total cholesterol, presence of leg symptoms, and mean CIMT (Table 4). The results demonstrated that age per 10 years (odds ratio [OR], 1.86; 95% confidence interval [CI], 1.18 to 2.93; P=0.008), current smoking status (OR, 2.62; 95% CI, 1.10 to 6.21; P=0.029), leg symptoms (OR, 7.35; 95% CI; 3.71 to 14.57; P<0.001), and mean CIMT (OR, 4.73; 95% CI, 1.14 to 19.57; P=0.032) were significantly associated with the presence of PAD in these patients (Table 4). After adjusting for age as a categorical variable in the multivariate analysis, age ≥70 years was significantly associated with a higher risk of PAD (OR, 3.08; 95% CI, 1.23 to 7.69; P=0.016) compared with age <60 years. There was no difference in the risk of PAD between the 60 to 69 and <60-year age groups (OR, 2.04; 95% CI, 0.97 to 4.30; P=0.061) (data not shown).

The cut-off CIMT value to predict the presence of PAD was 0.92 mm, with a 62.3% sensitivity and 59.0% specificity. ROC analysis demonstrated that the area under the curve was 0.69 (95% CI, 0.56 to 0.83; P=0.010) for the symptomatic patients (Fig. 3A) and 0.68 (95% CI, 0.61 to 0.76; P=0.001) for the asymptomatic patients (Fig. 3B). Collectively, these results suggest that older age (≥70 years), current smoking status, presence of leg symptoms, high mean CIMT (>0.92 mm) were factors associated with the presence of PAD, even in patients with ABI 0.91 to 1.40.