Prevalence, clinical characteristics, and predictors of peripheral arterial disease in hemodialysis patients: a cross-sectional study

Peripheral arterial disease (PAD) of the lower extremities is widespread, and it affects over 200 million people worldwide [1]. PAD is also highly prevalent among patients with end-stage renal disease, and it has serious consequences that influence patient outcomes [2, 3]. The prevalence of PAD has increased over the last decade, particularly in low-income countries [4]; further, it increases with age and is 4.5–14.5% more prevalent among individuals aged ≥65 years [5]. Among patients on hemodialysis (HD), the prevalence of PAD is much higher, ranging from 17 to 48% [6, 7].

The findings from the Dialysis Outcomes and Practice Patterns Study (DOPPS) described a PAD prevalence of 25.3% among patients on HD that was accompanied by significant geographic variations. In European countries, PAD prevalence was 17.5–37.8%, while in Japan¸ the prevalence was significantly lower at 11.5% [8]. Similarly, the findings from the Hemodialysis Study (HEMO) showed that PAD was seen in 23% of patients on HD [9]. The findings from the National Health and Nutrition Examination Survey (NHANES) 1999–2000 showed that PAD was present in 24% of patients with renal insufficiency and in 3.7% of participants with normal renal function; these findings were independent of diabetes, age, and hypertension [10]. Japanese investigators have shown that PAD prevalence in patients on HD ranges from 23.8 to 47%, depending on the study population [2, 11]. Although these studies were valuable, they had different methods for PAD diagnosis.

The early diagnosis and management of PAD can improve the prognosis for patients on HD [11]. Therefore, many patients on HD with PAD could avoid or at least delay adverse events, such as amputations, cardiovascular events, and death if PAD was timely diagnosed and adequately treated [12]. The traditional risk factors for PAD, namely, age, diabetes mellitus (DM), smoking, hyperlipidemia, hypertension, and male sex, are well recognized within the general population [7], but these risk factors are supplemented by other risk factors in patients on HD [13, 14]. To our knowledge, no studies have previously investigated the effects of vascular access on PAD prevalence. Thus, this study aimed to examine the prevalence of PAD, the risk factors associated with PAD, and the clinical characteristics of PAD in patients on HD.

There were no statistically significant differences between patients who did and did not participate in the study in terms of age, sex, and HD vintage. Of the 156 patients identified, PAD was present in 55 (35.3%; 95% CI, 27.7–42.8%), of which 40% were women (Table 1). Hypertension was the most common condition that caused the end-stage renal disease (53.2%), and DM was the second most common cause of the end-stage renal disease (12.8%) in our patients. Compared with the non-PAD group, the PAD patients were older, more likely to be male, and had higher incidence rates of anemia and DM. The participants’ mean HD vintage was 71.7 ± 69.5 (median: 55; range: 0–354) months. The groups did not differ in relation to the HD type and daily HD duration. Type of vascular access significantly differed between the PAD and non- PAD groups (p = 0.004). The prevalence of PAD was the highest in patients with Hickman vascular catheter (75%), followed by patients with arteriovenous graft (AVG; 53.8%), and was the lowest in patients with arteriovenous fistula (AVF; 30%). The only statistically significant difference in PAD prevalence between these groups was found when patients with Hickman vascular catheter and AVF were compared (p = 0.003). Although the prevalence of PAD was greater in patients with AVG than in those with AVF, there was no significant difference (p = 0.117).

The prevalence of PAD was associated with the age of patients on HD, and it was most prevalent among patients who were ≥ 80 years old (Fig. 1).

Table 2 presents the participants’ symptoms and clinical findings. Of the patients on HD without PAD, 19.8% had various symptoms in their legs, including coldness and numbness, and 2% had claudications. The PAD and non-PAD groups differed significantly in relation to the clinical findings, including pulse palpation, femoral bruit auscultation, and ABI values (all p < 0.001).

The patients on HD with PAD had significantly lower median ABIs than the patients on HD without PAD (0.90 vs. 1.13; p < 0.001) (Fig. 2).

Table 3 summarizes the laboratory blood test results. The albumin, total cholesterol, and iron levels were lower and the glucose and CRP levels were higher in the PAD group than in the non-PAD group.

Table 4 presents the results of logistic regression analyses using PAD as the dependent variable. The univariate analysis indicated that age, DM, symptoms based on the Fontaine classification, vascular access provided by a Hickman catheter, lower albumin and total cholesterol levels, and a higher CRP level were associated with PAD (all p < 0.05). The multivariate logistic regression analysis showed that the independent factors associated with PAD in patients on HD were a higher CRP level (odds ratio [OR], 1.03; 95% confidence interval [CI], 1.00–1.06; p = 0.030), vascular access provided by a Hickman catheter (OR, 4.66; 95% CI, 1.03–21.0; p = 0.045), and presence of PAD symptoms based on the Fontaine classification (OR, 5.20; 95% CI, 2.60–10.4; p < 0.001).

Approximately 35% of patients on HD in this study had PAD, which concurs with the rates published to date [6, 7]. In this study, PAD was diagnosed based on an ABI of < 0.9, clinical findings, and categorization of symptoms based on the Fontaine classification; 41 patients with PAD had ABI values < 0.9, seven patients had ABI values > 1.3, and seven had ABI values between 0.9 and 1.3, or more specifically, between 0.91 and 1.0. We confirmed PAD in seven patients with normal and in seven patients with high ABI values using color Doppler ultrasonography. There were significant differences between the PAD and non-PAD groups with respect to the clinical findings, including those from pulse palpation and femoral bruit auscultation, and ABIs. The clinical findings were noticeably worse in the PAD group than in the non-PAD group. Often, these patients have DM; chronic kidney disease alone leads to arterial incompatibility as a consequence of pronounced calcification, and arterial stiffness generates false-positive ABIs [2, 6, 16]. An ABI < 0.9 alone should not be used to assess the severity of PAD in patients on HD [2]. An ABI of < 0.9, in conjunction with the clinical findings, namely, absence of pulses and presence of femoral bruits, and PAD symptoms, including claudication, would provide a more robust foundation for a diagnosis of PAD.

Measuring the ABI is simple, reliable, inexpensive, and non-invasive [18]. According to the most recent American Heart Association guidelines, ABIs of 0.9–1.09 are lower normal values [18]; however, some Japanese authors have recommended ABIs of 1.05–1.1 as the limit for a diagnosis of PAD in patients on HD [2, 11].

Intermittent claudication is the classic and most striking symptom associated with PAD, and it is relatively rare in the general population and in patients on HD [5, 19]. Therefore, intermittent claudication alone is not a reliable criterion for a diagnosis of PAD [6]. The findings from preliminary studies have shown that 20–50% of patients with PAD who had ABIs of < 0.9 did not report claudication; hence, these patients belong to the asymptomatic group of patients with PAD [5, 20]. In the UK, Webb et al. [21] studied 325 patients receiving HD and reported that 19% of these patients had claudication. In this study, 37.2% of the patients had claudication. In the DOPPS and HEMO studies of patients on HD, a diagnosis of PAD was based on a previous PAD diagnosis, the presence of intermittent claudication or critical ischemia of the extremities, previous surgical revascularization or amputation as a consequence of PAD, and prior diagnosis of an abdominal aortic aneurysm. Based on these criteria, the rates of PAD in the DOPPS and HEMO studies were 25.3 and 23%, respectively [8, 12]. Likewise, the prevalence of PAD in the NHANES was 24% in the group of patients with chronic kidney disease, and the same criteria were used to diagnose PAD [10].

In studies that used ABI < 0.9 as a criterion for PAD diagnosis in patients on HD, PAD prevalence was slightly higher and it reached 37% [6]. Because ABIs < 0.9 are relevant for a diagnosis of PAD, they are an important prognostic factor associated with cardiovascular disease and mortality in patients on HD [11, 16, 19].

In our study, PAD was associated with older age, DM, and anemia, and a high percentage of patients with PAD had hypertension; however, the rate of left ventricular hypertrophy was not significantly higher in this group of patients, perhaps because it was determined by electrocardiography. Significant differences did not exist with respect to the systolic and diastolic blood pressure values determined before and after HD nor in relation to the pulse pressure and mean arterial pressure values in this study, but other investigators have described significantly higher diastolic and pulse pressure values in patients with PAD on HD [12, 13]. Viazzi et al. [3] suggested that the predictive power associated with ambulatory interdialytic blood pressure monitoring was much greater than the predictive powers of blood pressure measurements attained in the clinic or those determined immediately before and after dialysis. Our patients on HD were closely monitored and their blood pressure was well controlled, regardless of the presence of PAD.

Several authors have described higher frequencies of cardiovascular disease and stroke among patients with PAD [5, 22]. Although the patients in this study had high rates of cardiovascular disease and stroke, the groups did not differ with regard to their frequencies. Anemia was present among patients with PAD, a finding that concurs with the findings from a previously published study that showed that patients with PAD with anemia had a more severe form of the disease and had higher risks of mortality and limb amputation than patients with PAD without anemia [23].

Whereas the traditional risk factors associated with PAD, namely, age, DM, smoking status, hypertension, male sex, and hyperlipidemia [8], are valid within the general population, the risk factors for PAD slightly differ among patients on HD [6, 13]. In addition to the traditional risk factors, renal-specific factors are associated with PAD in patients on HD [6, 12, 13]. A smaller number of studies have investigated correlations between PAD and its risk factors in patients on HD [24]. The findings from the HEMO study that included 936 patients on HD, showed that DM, age, and smoking status are associated with PAD, while the other conventional risk factors, namely, male sex, hypertension, and hypercholesterolemia, are not associated with PAD [9].

The findings from the United States Renal Data System study and the Dialysis Morbidity and Mortality Study, which examined the associations among the variables related to dialysis and the conventional cardiovascular risk factors for PAD in patients on dialysis, showed that a diagnosis of PAD was associated with age, male sex, being Caucasian, DM, smoking status, higher diastolic blood pressure, and left ventricular hypertrophy. In an adjusted analysis, PAD was also associated with low serum parathyroid hormone levels, a longer dialysis vintage, a lower Kt/V, and a low serum albumin level [25]. Compared with the non-PAD group, the HD vintage was shorter in the PAD group in this study, but the difference was not significant, which concurs with the findings from a study carried out in Japan [12]; however, other investigators consider that PAD is related to the dialysis vintage [14]. Lower Kt/V values were observed among patients in the PAD group, which is consistent with the findings from other studies [11, 25]. These findings might help explain why some uremic toxins that are not removed by dialysis, for example, dimethylarginine, accumulates, damages endothelial cells, and causes atherogenesis [19].

Chronic uremia is associated with systemic inflammation that causes hypoalbuminemia and increases the risk of PAD [13, 26], and may play an important role in the pathogenesis and progression of atherosclerosis [13]. Chronic inflammation, malnutrition, and hyperphosphatemia are associated with PAD in patients with end-stage renal disease [6, 7]. The serum albumin and total cholesterol levels indicate not only a patient’s nutritional status but also chronic inflammation [25, 26]. Serum CRP is related not only to vascular inflammatory reactions but also promotes atherogenesis and atherothrombosis [27]. In this study, patients with PAD had significantly higher CRP levels and lower albumin, total cholesterol, and iron levels, which are indirect indicators of malnutrition, inflammation, and atherosclerosis syndrome [28].

Similar to our study, several investigators have shown that hyperlipidemia is not a risk factor for PAD in patients on HD [12, 29], but some authors have described hyperlipidemia as a risk factor for PAD [5, 11]. In contrast, the findings from one study indicated that hyperlipidemia had a protective effect against PAD [29].

The patients with PAD in our study had significantly higher levels of glycemia, which correlated with the higher incidence of diabetes in these patients. Several authors have described DM and poor glycemic control as important risk factors for PAD [5, 8, 13, 26]. Earlier reports have suggested that inflammation plays a key role in atherosclerosis and confirmed that patients with elevated CRP and lower albumin levels have high occurrence rates of PAD [27]. The findings from this study concur with those from previous studies.

A distal radiocephalic AVF is the vascular access of choice for HD at our center because it is associated with the longest survival times and the lowest number of complications. Patients with inadequate blood vessels undergo AVG implantations. Tunneled Hickman vascular catheters are used for permanent vascular access in patients who have poor prognoses and shorter life expectancies and whose blood vessels are inappropriate for an AVF or AVG implantations [30]. Our data showed that compared with the patients without Hickman vascular catheters (25%), a significantly higher number of patients on HD with Hickman vascular catheters had PAD (75%). We also observed that patients with PAD and Hickman vascular catheter had significantly lower hemoglobin (p = 0.001), and albumin (p = 0.022) levels than patients with AVF. Compared to the patients with AVF, patients with implanted Hickman vascular catheter and AVG had significantly higher CRP values (p = 0.05). The presence of a permanent vascular catheter was a predictor of PAD, and the use of the catheter may be associated with silent or overt inflammation. As is already known, there is a strong correlation between the atherosclerosis of blood vessels of the arm and PAD. Moreover, an increased incidence of vascular access failure was noted in patients with AVG compared to those with AVF [16, 31].

The management and treatment of patients with PAD requires a multidisciplinary approach [32]. The National Kidney Foundation’s Kidney Disease Outcomes Quality Initiative guidelines [33] recommend that at the beginning of a dialysis session, all patients should be screened for PAD, be clinically examined, and have their ABIs measured. These patients often have insufficient daily activity and diabetic neuropathy, and they are usually asymptomatic; due to this, they are underdiagnosed and undertreated [12, 20].

Once PAD is diagnosed, all risk factors must be rigorously managed, which includes smoking cessation, administering aspirin, statins, and cilostazol, and undertaking active exercise [19, 22]. Timely diagnosis of PAD and its early treatment can improve patients’ quality of life and avoid or, at least, delay undesirable events, including amputation or death [12, 13].

The prevalence of PAD was high in our group of patients on HD. Independent predictors of PAD were the presence of symptoms based on the Fontaine classification, a higher CRP level, and vascular access provided by a Hickman catheter for HD. An ABI < 0.9 combined with the clinical findings and claudication more accurately signifies clinically significant PAD. An ABI < 0.9 alone, without clinical findings, is not sufficient for the detection of clinically significant PAD in patients on HD. Adequate and timely diagnosis and treatment of PAD could improve the quality of life and postpone complications in these patients.