Blood Hemoglobin Concentrations and the Incidence of Lower Extremity Peripheral Arterial Disease in Patients Undergoing Hemodialysis: 10-Year Outcomes of the Q-Cohort Study

Abstract

Background: Lower extremity peripheral arterial disease is a potentially lethal cardiovascular complication in patients undergoing hemodialysis. Anemia is a risk factor for cardiovascular disease among the hemodialysis population. However, whether blood hemoglobin concentration is associated with the risk of peripheral arterial disease progression in this population remains undetermined.

Methods and results: This is an extension of a 4-year multicenter, prospective, observational cohort study to 10 years. A total of 3504 Japanese patients undergoing maintenance hemodialysis were recruited between 2006 and 2007. The primary exposure was blood hemoglobin concentration at baseline. The main outcome was the first-ever incidence of major adverse limb events (MALE), composed of endovascular treatment, bypass surgery, and amputation. Multivariable-adjusted Cox proportional hazards model, Fine-Gray subdistribution hazards model, restricted cubic spline analysis, and restricted mean survival time analysis were used to determine the association of blood hemoglobin concentration with the incidence of MALE. During a median follow-up of 8.0 years, 257 patients experienced MALE. A Cox proportional hazards model showed that the risk of MALE in patients with blood hemoglobin concentrations <10.0 g/dL was significantly higher than in patients with concentrations of 11.0 to 11.9 g/dL, even after adjusting for confounding factors. In contrast, elevated hemoglobin concentration (≥12.0 g/dL) was not significantly associated with increased risk of MALE. Similar associations were observed when the Fine-Gray subdistribution regression model was used by setting all-cause mortality as the competing risk.

Conclusions: A low blood hemoglobin concentration is an independent risk factor for peripheral arterial disease progression in patients undergoing maintenance hemodialysis.

Keywords: anemia; blood hemoglobin concentration; hemodialysis; major adverse limb events; peripheral arterial disease.

Figure 1

Histogram of the baseline blood hemoglobin concentrations.

Figure 2. Kaplan–Meier curves stratified by baseline blood hemoglobin concentration.

The log‐rank test was used to compare differences in the event‐free survival curves among the 4 groups. A 2‐tailed P‐value of <0.05 was considered statistically significant. Hb indicates hemoglobin.

Figure 3. Multivariable‐adjusted spline plots of the HR (95% CI) for the incidence of MALE by baseline blood hemoglobin concentration.

The solid line represents the adjusted HR, and the dotted lines represent the 95% CIs. The horizontal gray line corresponds to the reference HR of 1.0. The overall median blood hemoglobin concentration was 10.6 g/dL; therefore, this was chosen as the reference value. The multivariable model comprised the following parameters as covariates: age, sex, systolic blood pressure, hemodialysis vintage, hemodialysis time per session, body weight, presence of diabetic nephropathy, history of cardiovascular events and stroke, cardiothoracic ratio, serum concentrations of albumin, total cholesterol, creatinine, albumin‐corrected calcium, and phosphate, log serum C‐reactive protein, log serum ferritin, log serum alkaline phosphatase, log serum parathyroid hormone, and use of antihypertensives, iron preparations, and erythropoiesis‐stimulating agents. HR indicates hazard ratio; MALE, major adverse limb events. A 2‐tailed P value of <0.05 was considered statistically significant.

Figure 4. Multivariable‐adjusted HRs for the association between blood hemoglobin concentration and the development of MALE were calculated by the Cox proportional hazard regression models.

Covariates included age, sex, systolic blood pressure, hemodialysis vintage, hemodialysis time per session, body weight, presence of diabetic nephropathy, a history of cardiovascular events and stoke, cardiothoracic ratio, serum concentrations of albumin, total cholesterol, creatinine, albumin‐corrected calcium, and phosphate, log serum alkaline phosphatase, log serum C‐reactive protein, log serum ferritin, and log serum parathyroid hormone, and use of antihypertensives. Patients were divided into subgroups based on the use of iron preparation or ESAs or median values of ERI (7.03 U/g/dL per kg) or serum ferritin (163 ng/mL) concentrations. As for subgroup analysis on ERI, those who were treated with ESA were analyzed, and those who were not treated with ESAs were excluded. A 2‐tailed P value of <0.05 was considered statistically significant in Cox proportional hazard regression analyses. A 2‐tailed P value of <0.1 was considered statistically significant for interaction terms. ERI indicates erythropoietin resistance index; ESAs, erythropoiesis‐stimulating agents; HR, hazard ratio; and MALE, major adverse limb event. Conversion factors for units: blood hemoglobin in g/dL to mmol/L, ×0.621.