The subjects included Japanese patients with PAD who were hospitalized in Cardiovascular Hospital of Central Japan during the period from February 1, 2000 and April 30, 2020. This prospective cohort study complies with the principles of the Declaration of Helsinki, and the Medical Ethical Committee approved the study protocol in our hospital (CCJ-EA-006). Patients who gave their written informed consent to partake in this study were chosen as subjects. Patients satisfying the following inclusion criteria were selected: 1) an Ankle Brachial Pressure Index (ABI) of <0.90; 2) clinical symptoms (intermittent claudication or critical limb ischemia); 3) a stenotic lesion of ≥70% in iliac or femoropopliteal artery was specified with angiography or ultrasound. Patients with a treatment history of PAD including amputation of leg in another side were excluded from this study. The patients were also limited to those with PAD due to atherosclerosis. Patients with PAD due to non-atherosclerotic causes, such as vasculitis, Buerger disease, and fibromuscular dysplasia, were excluded. Patients with dementia were also excluded due to the difficulty of checking vital signs and obtaining other information.

Clinical data characteristics for each patient were obtained from the primary analysis for age, Body Mass Index (BMI), ABI, smoking history, hypertension, and Diabetes Mellitus (DM). The concentration of creatinine, albumin, triglyceride, total cholesterol, High Density Lipoprotein Cholesterol (HDL-C), Low Density Lipoprotein Cholesterol (LDL-C), glucose, D-dimer, and C-reactive protein (CRP) was determined with Hitachi 7180 automatic analyzer (Hitachi High-Tech Fielding Co., Tokyo, Japan). DM was defined as a fasting glucose level of >126 mg/dL at least two data or receiving antidiabetic treatment [18]. Hypertension was specified as blood pressure ³140/90 mmHg recorded at least twice or a requirement for antihypertensive treatment. The estimated Glomerular Filtration Rate (eGFR) was calculated using the Modification of Diet in Renal Disease equation with age and serum creatinine level [19].

A brain Computed Tomography (CT) scan was performed in 5- or 10-mm thick sections without contrast agent (Aquilion 64, Toshiba, Tokyo, Japan). The scans were evaluated by two radiologists who were blinded to the statuses of the patients. Infarcts > 1.5 cm in diameter were defined to be cerebral infarction on brain CT. A low density lesion with a diameter ≤ 1.5 cm was defined as a lacunar infarction. Patients with CVD were defined as cerebral infarction and/or lacunar infarction on brain CT or a history of this disease. An electrocardiogram and an echocardiography were performed for each patient. CHD was defined as a documented history of ischemic heart disease (myocardial infarction, percutaneous coronary intervention, composite of angina pectoris, or coronary artery bypass grafting) or a coronary angiogram (≥50% stenosis in at least one coronary artery).

Each patient was followed up at 1, 3 and 4 months after treatment and assessed interval was 4- or 6-month. Vital signs and medical status were assessed with hospital data and written questionnaires for life statuses were assessed by the Foot Care Club [4] [20]. Ischemic strokes were determined as the presence of a new focal neurological deficit, and magnetic resonance imaging or CT was required to confirm the lesions. TIA was specified as the presence of a new neurological deficit lasting < 24 hours. Definition of myocardial infarction was signified previously [20] [21]. Peripheral restenosis during follow-up was determined as a decrease in ABI of ≥0.15% and ≥50% stenosis using angiography or duplex ultrasonography [20] [22], and major amputation was defined as above-the-ankle amputation.

The primary efficacy endpoints were Cardiovascular or Cerebrovascular Related Death (CCVD) and All-Cause Death (ACD). The secondary efficacy end-points were major adverse cardiovascular events (MACE: all-cause death, non-fatal myocardial infarction, non-fatal ischemic stroke, or transient ischemic attack) and cardiovascular and/or limb events (CVLE: CCVD, non-fatal myocardial infarction or cerebral infarction, transient ischemic attack, presence of a new peripheral lesion, repeat revascularization for a limb, or major amputation).

All statistical analyses were calculated with IBM SPSS Statistics ver. 25.0 (IBM Corp, Armonk, NY). Categorical variables are signified as a number (%) and were compared by chi-square test with a Bonferroni correction. Continuous variables are expressed as a median (interquartile range) and were assessed by Kruskal-Wallis test with the Dann-Bonferroni method [23]. In multiple regression analysis, we calculated among all risk factors with simple Pearson correlations. Factors with p < 0.05 in this correlation analysis were calculated using stepwise forward multiple regression analysis to define relationships between the number of affected arteries and individual risk factors. We calculated Odds Ratio (OR) and Confidence Interval (CI) between DoVD or TrVD and risk factors with a univariate logistic analysis. Factors with p < 0.05 in these analyses were used in a multivariate logistic model to define the risk predictors for DoVD and TrVD. The Kaplan-Meier analysis was used to specify CCVD, ACD, MACE, and CVLE among PADa, DoVD, or TrVD and compared using the log-rank test with Bonferroni correction. In Cox univariate regression model, a Hazard Ratio (HR) and 95% CI were estimated for individual factors. Factors with p < 0.05 were used in multivariate regression models to define significant factors associated with these endpoints. A p-value < 0.05 was defined as statistically significant.

Among 1053 patients, follow-up was possible for 1019 subjects. The mean and median follow-up periods were 80.2 ± 65.9 and 67 (29 - 115) months. The median and mean age were 72 (67 - 79) and 72.6 ± 9.9 years. The patients who died were 539 (52.9%) during the follow-up periods. The prevalence of CCVD was 50.5% (n = 272), as cardiac or major vascular disease (n = 193, 35.8%), cerebrovascular disease (n = 79, 14.7%). Other causes of deaths were malignancy (n = 108, 20.0%), pneumonia (n = 88, 16.3%), and other causes (n = 71, 13.2%). The cumulative 5-, 10-, and 15-year overall survival rates in all patients were 71.8%, 45.4%, and 32.7%, respectively. The baseline clinical characteristics and comorbidities in PAD patients with or without other vascular diseases are summarized in Table 1. Patients with PADa had higher ABI, serum albumin, eGFR, and HDL-C, and lower levels of CRP and D-dimer. The prevalences of CLI and DM were also lower in these patients.

The number of affected arteries had significant positive correlations with age, men, CLI, LDL-C, and DM, and negative correlations with ABI, HDL-C, serum albumin, and eGFR in simple Pearson correlation analysis (p < 0.05). Stepwise forward multiple regression analysis of the relationship between the number of affected arteries and these factors was performed (Table 2). The number of affected arteries had significant negative correlations with eGFR, HDL-C, and ABI, and a positive correlation with DM (p < 0.01).

Relationships between risk factors and DoVD and TrVD were analyzed using multiple logistic analyses (Table 3). PAD with CVD was correlated with older age, lower ABI, eGFR, and atrial fibrillation; PAD with CHD showed a correlation with younger age, lower eGFR, HDL-C, higher LDL-C, and DM; and TrVD was correlated with lower ABI, eGFR, HDL-C, and DM.

Cumulative incidence of the 5-, 10-, and 15-year rates for CCVD are demonstrated in Figure 1. There were significant differences among PADa, DoVD, or TrVD (p < 0.001, respectively). In Cox univariate analysis, age, CLI, CVD, CHD, hypertension, higher level of CRP, D-dimer, lower ABI, BMI, serum albumin, eGFR, and statin were associated to CCVD (p < 0.05). In multivariate analysis, higher age, CRP, lower serum albumin, eGFR, BMI, CLI, CVD, and CHD were also associated with CCVD, and statin also decreased CCVD (Table 4, p < 0.05).

The cumulative 5-, 10-, and 15-year incidence rates for ACD are demonstrated in Figure 2. There were significant differences between PADa and DoVD (p < 0.001) or TrVD (p < 0.001), but there was no significant difference between DoVD and TrVD (p = 0.066). In Cox univariate analysis, age, CLI, CVD, DM, higher level of CRP, D-dimer, lower ABI, BMI, serum albumin, and eGFR were associated with ACD (p < 0.05). Treatment with statin or aspirin and revascularization were related to ACD. In multivariate analysis, higher age, CLI, CRP, lower BMI, serum albumin, eGFR, and CVD were also associated with ACD, and statin decreased ACD (Table 4, p < 0.05).

The cumulative 5-, 10-, and 15-year incidence rates for MACE are shown in.

Double VD: PAD with cerebral infarction or coronary heart disease; Triple VD: PAD with cerebral infarction and coronary heart disease; eGFR: estimated glomerular filtration rate; LDL-C: low density lipoprotein cholesterol; HDL-C: high density lipoprotein cholesterol; ACE: angiotensin-converting enzyme; ARB: angiotensin receptor blocker. *: p < 0.05 (PAD alone vs. Double VD); ^§: p < 0.05 (PAD alone vs. Triple VD); ^¶: p < 0.05 (Double VD vs. Triple VD).

R² = 0.059; F for change in R² = 0.009; P = 0.006. β: standardized coefficient; B: non-standardized coefficient; CI: confidence interval for B; eGFR: estimated glomerular filtration rate; ABI: ankle brachial pressure index.

PAD: peripheral arterial disease; CVD: cerebrovascular disease; CHD: coronary heart disease; OR: odds ratio; CI: confidence interval; ABI: ankle brachial pressure index; eGFR: estimated glomerular filtration rate; HDL-C: high density lipoprotein cholesterol; LDL-C: low density lipoprotein cholesterol.

HR: hazard ratio; CI: confidence interval; eGFR: estimated glomerular filtration rate.

Figure 3. There were significant differences among PADa, DoVD, or TrVD (p < 0.001, respectively). In Cox univariate analysis, age, men, CLI, CVD, CHD, DM, higher level of CRP, D-dimer, lower BMI, ABI, serum albumin, eGFR, and statin were associated to MACE (p < 0.05). In multivariate analysis, higher age, CRP, lower serum albumin, eGFR, CLI, CHD, and DM were also associated with MACE, and statin decreased MACE (Table 5, p < 0.05).

The cumulative 5-, 10-, and 15-year incidence rates for CVLE are shown in Figure 4. There were significant differences among PADa, DoVD, or TrVD

HR: hazard ratio, CI: confidence interval, ABI: ankle brachial pressure index, eGFR: estimated glomerular filtration rate.

(p < 0.001, respectively). In Cox univariate analysis, age, men, CLI, CVD, CHD, DM, higher level of CRP, D-dimer, lower BMI, ABI, serum albumin, eGFR, and statin were associated to CVLE (p < 0.05). In multivariate analysis, higher age, CRP, lower ABI, serum albumin, eGFR, CLI, CHD, and DM were also associated with CVLE, and statin also decreased CVLE (Table 5, p < 0.05).