Abnormal Calcium Metabolism Mediated Increased Risk of Cardiovascular Events Estimated by High Ankle-Brachial Index in Patients on Peritoneal Dialysis

Abstract

Cardiovascular disease (CVD) is the leading cause of death in peritoneal dialysis (PD) patients. But the relationship between regular PD and the risk of major adverse cardiovascular events (MACE) remains controversial. The possible risk factors are not fully elucidated. This study aims to investigate the possible factors affecting the risk of MACE estimated by high ankle-brachial index (ABI) in PD patients. A total of 243 patients were enrolled and divided into chronic kidney diseases (CKD) stage 1, non-dialyzed CKD stages 2-5, and PD groups. The prevalence of high ABI, indicating increased MACE, was elevated with CKD progression but not further increased in PD patients. Systolic blood pressure was closely correlated with high ABI in non-dialyzed CKD patients (β = 0.059, P = 0.001). But in PD patients, serum calcium had a crucial effect on high ABI (β = -9.853, P < 0.001). Additionally, PD patients with high ABI tended to dialyze inadequately (Kt/V <1.7) compared to those with normal ABI (29.0 vs. 13.3%, P = 0.031). Further mediation analysis revealed that ~86.2% of the relationship between Kt/V and high ABI was mediated by serum calcium in PD patients (mediation effect = 86.2%, ab = -0.220, 95% CI: -0.381 to -0.059, P = 0.008), especially in those starting PD before 55 years of age and with normal body mass index. This present study indicated that improvement of PD adequacy by maintaining calcium balance might be a promising method to reduce the risk of MACE estimated by high ABI for PD patients.

Keywords: Kt/V; ankle-brachial index; calcium; peritoneal dialysis; vascular calcification.

Figure 1

The prevalence of high ABI in various CKD stages. (A) The rates of high ABI were similarly increased in non-dialyzed CDK stage 2–5 and PD groups compared with CKD stage 1. *P < 0.05 vs. CKD stage 1 group. (B) Similar upgraded trend of high ABI was shown in the decline of e-GFR. However, the prevalence of high ABI in PD patients was much lower than that in non-dialyzed CKD patients with e-GFR <30 ml/min per 1.73m². *P < 0.05 vs. the group with e-GFR ≥ 90 ml/min per 1.73 m²; # P < 0.05 vs. the group with 60 ≤ e-GFR <90 ml/min per 1.73 m²; and & P < 0.05 vs. the group with 30 ≤ e-GFR <60 ml/min per 1.73 m². ABI, ankle-brachial index; CKD, chronic kidney disease; e-GFR, estimated-glomerular filtration rate; PD, peritoneal dialysis.

Figure 2

Comparison of the history of adverse cardiovascular events between the normal and high ABI groups in PD patients. The prevalence of cardiac death, CHD, and CHF (A–C), but not AMI and ACT (D,E), was significantly different between patients with normal and high ABI. A significant difference was also observed in the rate of MACE (F) between the two groups. *P < 0.05 vs. normal ABI group. ABI, ankle-brachial index; ACI, acute cerebral infarction; AMI, acute myocardial infarction; CHD, coronary heart disease; CHF, congestive heart failure; MACE, major adverse cardiac events.

Figure 3

Mediation analysis on the association between e-GFR or Kt/V and ABI. (A) The mediation effect of serum calcium did not exist in the association between e-GFR and high ABI in non-dialyzed CKD patients. (B) Serum calcium mediated the association between Kt/V and high ABI in PD patients. ABI, ankle-brachial index; Ca, calcium; e-GFR, estimated-glomerular filtration rate; ME, mediation effect; SE, standard error.

Figure 4

The relationship among adequacy dialysis and high ABI. (A) PD patient with high ABI tended to have Kt/V lower than 1.7 compared with those with normal value. *P < 0.05 vs. normal ABI group. (B) The level of serum calcium in PD patients with Kt/V ≥ 1.7 were relatively higher but within normal range. *P < 0.05 vs. Kt/V <1.7. ABI, ankle-brachial index; Ca, calcium.

Figure 5

Subgroup analysis of assessing the mediation effect of serum calcium. The mediation effect of serum calcium on the association between Kt/V and high ABI only existed in PD patients starting dialysis before 55 years of age (C) and with normal BMI (G). No mediation effect of serum calcium was observed in other subgroups (A,B,D–F,H–N). The number of patients in subgroups: age ≤ 55 y: n = 88, Age > 55 y: n = 26; 20 y ≤ start PD ≤ 55 y: n = 93, start PD > 20 y OR <50 y: n = 17; male: n = 73, female: n = 41; 18 kg/m² ≤ BMI ≤ 25 kg/m²: n = 79, BMI <18 kg/m² OR > 25 kg/m²: n = 31; smoking: n = 45, non-smoking: n = 69; DM: n = 16, non-DM: n = 98; hypertension: n = 100, non-hypertension: n = 14. ABI, ankle-brachial index; BMI, body mass index; Ca, calcium; DM, diabetes mellitus; e-GFR, estimated-glomerular filtration rate; ME, mediation effect; PD, peritoneal dialysis; SE, standard error.