ABSTRACT
Background
Fetuin-A inhibits precipitation of calcium-phosphate crystals by forming calciprotein particles (CPP). A novel T50 test, which measures transformation time from primary to secondary CPP, is an index for calcification propensity. Both lower fetuin-A and shorter T50 levels were associated with cardiovascular disease (CVD) risk in patients with chronic kidney disease (CKD). Extremely high risk for CVD death in advanced CKD patients consists of high-incidental CVD event and high mortality after CVD event. To date, it is unclear whether fetuin-A and/or T50 can equally predict each CVD outcome.
Methods
This prospective cohort study examined patients undergoing maintenance hemodialysis. The exposures were fetuin-A and T50. The outcomes of interests were new CVD events and subsequent deaths. The patients were categorized into tertiles of fetuin-A or T50 (T1 to T3).
Results
We identified 190 new CVD events during the 5-year follow-up of the 513 patients and 59 deaths subsequent to the CVD events during 2.5-year follow-up. A lower fetuin-A but not T50 was significantly associated with new CVD events [subdistribution hazard ratio (HR) 1.73, 95% confidence interval (CI) 1.15–2.61, P = .009 for T1 vs T3]. In contrast, a shorter T50 but not fetuin-A was a significant predictor of deaths after CVD events (HR 3.31, 95% CI 1.42–7.74, P = .006 for T1 + T2 vs T3). A lower fetuin-A was predictive of new CVD events, whereas a shorter T50 was more preferentially associated with subsequent death.
Conclusion
These results indicate that fetuin-A and T50 are involved in cardiovascular risk in different manners.
Keywords: calcification propensity, cardiovascular disease (CVD), fetuin-A, hemodialysis, T50
Graphical Abstract
Graphical Abstract.
KEY LEARNING POINTS.
What was known:
Patients undergoing hemodialysis have an extremely high risk of cardiovascular disease (CVD) death, which consists of both high incidence of CVD events and high mortality rates after CVD events.
Lower fetuin-A, an inhibitor of vascular calcification, and shorter T50, an index for calcification propensity, have been associated with CVD risk.
This study adds:
We found that lower fetuin-A, but not T50, was significantly associated with new CVD events.
On the other hand, shorter T50, but not fetuin-A, was a significant predictor of deaths after CVD events. It appears that fetuin-A and T50 are involved in CVD risk in different manners in hemodialysis patients.
Potential impact:
Careful monitoring of both fetuin-A and T50 may lead to improvements in CVD outcomes.
INTRODUCTION
Fetuin-A (α2-Heremans Schmid glycoprotein: AHSG) is known to be a predictor of cardiovascular disease (CVD) and all-cause mortality in patients undergoing dialysis who are prone to vascular calcification [1–3]. In vitro and in vivo findings have suggested that fetuin-A can prevent ectopic calcification by forming colloidal mineral complexes, termed calciprotein particles (CPP), which consist of mainly calcium, phosphate and fetuin-A [4–8]. Since fetuin-A can buffer super saturation of mineral ions, the association of lower fetuin-A and worse clinical outcomes can be explained at least partly by lower protection against vascular calcification. In addition, fetuin-A may affect cardiovascular outcomes through its effect on insulin resistance [8–10].
In the early phase of the calcification process, primary CPP are formed to prevent precipitation of calcium-phosphate crystals. Further calcium and phosphate load accelerate conversion of primary CPP into larger secondary CPP which contain hydroxyapatite crystals [4–7]. Recently, a novel serum test “T50” has been developed, which is the transformation time from primary to secondary CPP in vitro measured by detecting rapid increase of turbidity by nephelometric method [11]. Although fetuin-A is a dominant contributor of T50, phosphate, calcium, magnesium and albumin levels are also known to influence T50. Thus, T50 is an integrated indicator of various factors of chronic kidney disease (CKD)–mineral and bone disorder (MBD).
A shorter T50, which means higher calcification propensity, was associated with higher risk for mortality, CVD events and kidney failure in patients with CKD stage 2–4 [12, 13]. An inverse association between T50 and all-cause or CVD death was found in patients undergoing hemodialysis [14, 15]. In kidney transplant recipients, a shorter T50 was a potent predictor of CVD events [16], as well as all-cause and CVD-related mortality [17, 18].
Patients undergoing hemodialysis suffer from an extremely increased CVD mortality compared with the general population [19]. The patients have higher risk not only for incidence of a CVD event but also for subsequent death after a CVD event [20, 21]. Thus, these two components, i.e. CVD events and subsequent death, synergistically raise the CVD mortality [19]. It has been reported that the factors affecting death after CVD events were different from those affecting new CVD events in patients undergoing hemodialysis [22–24].
Although the previous studies suggested a considerable overlap between fetuin-A and T50 as predictor of CVD events and mortality, it is unknown whether fetuin-A is equivalent to T50. In the present study, we compared fetuin-A and T50 in terms of the associations with new CVD events and death after CVD events in a cohort of hemodialysis patients.
MATERIALS AND METHODS
Study design and participants
This study was performed using data of the ‘DREAM’ (Dialysis-Related Endocrine and Metabolic changes affecting cardiovascular disease) cohort, which is a prospective, observational, and single-center study as previously registered at the UMIN-CTR (unique ID: UMIN000006168, http://www.umin.ac.jp/ctr/index.htm). Some of the results of this cohort have been reported previously [24, 25]. The DREAM cohort was composed of total 518 patients undergoing hemodialysis at Inoue Hospital, Suita, Osaka, Japan. The ethics committee at Inoue Hospital approved the study protocol (Approval No. 121), which adhered to the Declaration of Helsinki. All participants provided written informed consent and were enrolled in December 2004 and followed up until December 2009. Of the 518 patients, 513 patients with available data on both fetuin-A and T50 were analyzed in this study.
Blood sampling and laboratory tests
Blood samples for routine laboratory tests were taken before initiation of hemodialysis session on the first day of the week (Monday or Tuesday). Serum samples at baseline were frozen and kept at −80°C until the time of additional assays.
Serum fetuin-A assay
Serum fetuin-A was measured by a commercially available enzyme-linked immunosorbent assay kit (BioVender Laboratory Medicine, Inc., Czech Republic) as previously reported [9, 26].
Measurement of serum calcification propensity (T50)
Serum calcification propensity (T50) was quantified according to the method originally developed by Pasch et al. [11]. In brief, serum samples were supersaturated by adding calcium (final concentration of 10 mM) and phosphate (final concentration of 6 mM) to trigger formation of primary CPP which mainly contains fetuin-A and amorphous calcium phosphate. Primary CPP undergoes spontaneous transformation into secondary CPP which includes thermodynamically stable hydroxyapatite crystals. To detect the transition from primary to secondary CPP, time to a rapid increase in turbidity during a thermo-constant incubation at 37°C was quantified by a time-resolved laser light scatter nephelometry using an automated laser-based microplate nephelometer (Nephelostar; BMG Labteck, Offenberg, Germany). The one-half transition time (T50) was determined as calcification propensity specific for individual serum. Thus, a shorter T50 means lower resistance to crystallization and higher calcification propensity. All serum samples were measured in duplicate, and the average values were used for analysis. The CV of inter- and intra-assay were 3.20% and 4.47%, respectively, in our laboratory [27, 28].
Definition of CVD events
CVD events in the DREAM cohort were defined as the new occurrence of ischemic heart disease, ischemic stroke, hemorrhagic stroke, peripheral artery disease, pulmonary edema and cardiac valve disease as previously described [24, 25]. Incident and recurrent CVD and interventions for CVD were included as CVD events whereas chest pain or transient ischemic attack with no evidence with images were excluded. The same definition was applied to the pre-existing CVD at baseline and the new onset of CVD during the follow-up. Sudden deaths were considered as new and fatal CVD events. Furthermore, new CVD events were divided into atherosclerotic CVD events (ischemic heart disease, ischemic stroke and peripheral artery disease) and non-atherosclerotic CVD events (hemorrhagic stroke, pulmonary edema and cardiac valve disease). The patients with sudden death were classified into atherosclerotic CVD events if they had history of CVD events.
Study outcomes
The key outcomes of interest were new CVD events and deaths after CVD events in this study. All-cause mortality was also evaluated as an additional outcome to confirm the previously reports [1, 2, 14, 15].
Other variables
As the major demographic factors, we documented age, sex, duration of hemodialysis, clinical diagnosis of diabetic kidney disease (DKD) (not necessarily biopsy proven) and pre-existing CVD. Serum calcium, phosphate, intact parathyroid hormone (PTH) and use of vitamin D receptor activator (VDRA) were recorded as the parameters of CKD-MBD. Calcium-sensing receptor agonists were not used at baseline in this study. As clinical factors of anemia in CKD, we included hematocrit, dose of erythropoiesis-stimulating agent (ESA) and use of intravenous iron injections. At baseline, only recombinant human erythropoietin preparations were available as ESA in Japan. We considered hypertension, current smoking status, and dyslipidemia as traditional risk factors. We defined hypertension as blood pressure of 140/90 mmHg or higher and/or use of any anti-hypertensive medication according to the guideline of Japanese Society of Hypertension [29]. Dyslipidemia was defined as high-density lipoprotein cholesterol (HDL-C) ≤40 mg/dL, non-HDL-C ≥150 mg/dL and/or use of statin, with these lipid levels being derived from the target levels for patients with CKD recommended by the guideline of Japan Atherosclerosis Society [30]. As indicators of inflammation and wasting, body mass index after dialysis, C-reactive protein (CRP) and serum albumin were recorded.
Follow-up and clinical endpoints
The cohort was followed-up for up to 5 years until the end of 2009. At the end of each year, primary physicians made up an annual follow-up sheet to report dates of death (cause of death), CVD events, kidney transplantation, switching to peritoneal dialysis and moving away to other dialysis unit, as previously reported [24, 25].
Statistics
The patients were divided into tertiles of fetuin-A and T50, and their characteristics were summarized as numbers, percentages or medians (interquartile ranges) as indicated. Comparisons in prevalence and median values among the tertiles were performed by χ2 test and Kruskal–Wallis test, respectively. Unadjusted correlations of fetuin-A with other variables were examined by Spearman's rank correlation. CRP values below detection limit (0.01 mg/dL) were handled as 0.01 mg/dL.
Cumulative incidence of new CVD events was evaluated by Gray's test according to the tertiles of fetuin-A and T50, because death would impede the occurrence of CVD events [31, 32]. The associations of fetuin-A and T50 with new CVD events were examined by multivariable-adjusted Fine–Gray models which consider the competitive risk of death. Subdistribution hazard ratios (sdHRs) with 95% confidence intervals (CIs) were presented with highest tertile of fetuin-A (T3) and T50 (T3) as referent. CRP and T50 were transformed to the common log before entering the regression models. We calculated the sdHRs (95% CIs) of tertiles of fetuin-A and T50 with new CVD events with unadjusted Fine–Gray model. First, the model was minimally adjusted for age and pre-existing CVD (Parsimonious Model 1). These two variables were selected by both forward and backward stepwise methods using either Akaike information criteria or Bayesian information criteria and confirmed by the bootstrap resampling. Second, the model was adjusted for the five major demographic factors (Parsimonious Model 2). Further adjustment was performed for factors related to CKD-MBD (Extended Model 1), anemia in CKD (Extended Model 2), traditional risk factors (Extended Model 3), inflammation and wasting (Extended Model 4), log T50 (Extended Model 5) or fetuin-A (Extended Model 6). The linear trends in risks across tertiles were evaluated by entering the median for each categorical level of exposure. Because the associations of fetuin-A or log T50 with the outcomes were deemed linear, we additionally conducted similar analyses with fetuin-A and log T50 as continuous variables, and sdHRs were expressed per 1-SD lower. As sensitivity analyses, the association of fetuin-A or T50 with new CVD events was evaluated using the Kaplan–Meier curve with log-rank test and Cox proportional hazard regression models. HRs with 95% CIs were presented with the highest tertile (T3) of fetuin-A or T50 as referent. Similar statistical analyses were performed by fetuin-A or log T50 (per 1-SD lower) as continuous variable.
Finally, we focused on deaths after new CVD events. The patients who experienced new CVD events were further followed-up for up to 2.5 years. In addition to the limited numbers of death after CVD events, the death rates were similar between T2 and T3. Therefore, T2 and T3 were combined as T2 + T3 and were compared with T1. Unadjusted associations of fetuin-A and T50 with death after CVD events were examined by Kaplan–Meier method with log-rank test. HRs of fetuin-A and T50 with death after CVD events were further analyzed with unadjusted Cox model and adjusted for age and pre-existing CVD (Parsimonious Model), Parsimonious Model + sex (Extended Model 1), Parsimonious Model + duration of hemodialysis (Extended Model 2), Parsimonious Model + DKD (Extended Model 3), Parsimonious Model + log T50 (Extended Model 4) and Parsimonious Model + fetuin-A (Extended Model 5). These associations were also examined by fetuin-A or log T50 as continuous variable.
These statistical calculations were performed with EZR (version 1.55) developed by Prof. Kanda, Saitama Medical Center, Jichi Medical University, Saitama, Japan, which is graphical interface for R (The R Foundation for Statistical Computing, Vienna, Austria) [33]. More precisely, it is a modified version of R commander designed to add statistical functions frequently used in biostatics. Bootstrap resampling and variable selection procedures were performed using R 4.2.2 software (the R Foundation for Statistical Computing, Vienna, Austria). A P-value <.05 by two-sided test was considered statistically significant.
RESULTS
Selection and characteristics of study participants
Of the 518 total patients in the DREAM cohort, 513 patients were analyzed for this study (Supplementary data, Fig. S1). In the total patients, serum fetuin-A was normally distributed whereas serum T50 was not. The median (interquartile range) levels of fetuin-A and T50 were 268 (238–304) μg/mL and 176 (140–219) min, respectively. Fetuin-A level was positively associated with T50 (ρ = 0.598, P < .001). The baseline characteristics of patients are summarized according to tertiles of fetuin-A and T50 (Table 1).
Table 1:
Clinical characteristics of 513 patients.
| Fetuin-A | T50 | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Total | T1 | T2 | T3 | P | T1 | T2 | T3 | P | |
| Number of patients | 513 | 171 | 170 | 172 | 175 | 168 | 170 | ||
| Fetuin-A (μg/mL) | 268 (238–304) | 221 (200–238) | 267 (258–281) | 318 (304–342) | <.001 | 233 (205–261) | 270 (254–300) | 305 (274–339) | <.001 |
| T50 (min) | 176 (140–219) | 138 (118–171) | 171 (148–199) | 214 (189–248) | <.001 | 129 (115–142) | 177 (165–191) | 240 (219–288) | <.001 |
| Age (years) | 61 (54–68) | 65 (59–72) | 62 (55–68) | 57 (51–62) | <.001 | 63 (56–70) | 61 (55–67) | 60 (53–66) | .017 |
| Male sex (%) | 63.2 | 62 | 62.9 | 64.5 | .885 | 63.4 | 64.3 | 61.8 | .887 |
| Duration of hemodialysis (years) | 9.3 (3.8–15.9) | 7.7 (3.2–13.0) | 8.5 (3.4–14.8) | 11.5 (5.4–19.2) | <.001 | 8.6 (3.5–13.1) | 10.5 (3.4–17.0) | 9.5 (4.6–17.6) | .237 |
| DKD (%) | 21.2 | 28.1 | 22.4 | 13.4 | .004 | 22.9 | 20.2 | 20.6 | .812 |
| Pre-existing CVD (%) | 33.3 | 42.7 | 30.6 | 26.7 | .005 | 40.6 | 32.7 | 26.5 | .021 |
| Calcium (mg/dL) | 9.1 (8.6–9.8) | 9.0 (8.5–9.6) | 9.1 (8.5–9.8) | 9.3 (8.7–9.9) | .010 | 9.0 (8.4–9.6) | 9.1 (8.6–10.0) | 9.2 (8.6–9.9) | .012 |
| Phosphate (mg/dL) | 5.8 (5.0–6.6) | 5.8 (4.9–6.) | 5.6 (5.0–6.6) | 5.8 (5.1–6.7) | .506 | 6.1 (5.2–6.8) | 5.7 (5.1–6.7) | 5.6 (4.7–6.2) | <.001 |
| Intact PTH (pg/mL) | 118 (40–214) | 125(58–199) | 120 (34–226) | 113 (37–208) | .478 | 139 (63–244) | 101 (37–182) | 112 (35–204) | .036 |
| Use of VDRA (%) | 36.3 | 57 (33.3 | 40.6 | 34.9 | .341 | 38.3 | 39.9 | 30.6 | .163 |
| Hematocrit (%) | 30.7 (28.6–32.4) | 30.7 (28.4–32.2) | 30.5 (28.7–32.5) | 31.0 (28.8–32.4) | .626 | 30.4 (28.3–32.2) | 30.7 (28.8–32.3) | 31.0 (29.0–32.7) | .140 |
| ESA dose (×1000 U/week) | 9.0 (7.5–9.0) | 9.0 (9.0–9.0) | 9.0 (7.5–9.0) | 9.0 (6.0–9.0) | .016 | 9.0 (9.0–9.0) | 9.0 (7.5–9.0) | 9.0 (6.0–9.0) | .021 |
| Use of IV iron (%) | 57.7 | 63.2 | 60.6 | 49.4 | .024 | 62.9 | 58.3 | 51.8 | .111 |
| Hypertension (%) | 86.2 | 85.4 | 88.2 | 84.9 | .626 | 88 | 89.9 | 80.6 | .032 |
| Smoker (%) | 41.3 | 38 | 42.9 | 43 | .559 | 44 | 41.7 | 38.2 | .551 |
| Dyslipidemia (%) | 49.7 | 42.1 | 54.7 | 52.3 | .047 | 46.3 | 49.4 | 53.5 | .403 |
| BMI (kg/m2) | 21.5 (19.6–23.5) | 20.8 (19.1–22.4) | 21.9 (20.2–23.6) | 22.0 (19.8–24.0) | <.001 | 21.5 (19.6–23.1) | 21.2 (19.5–23.3) | 21.9 (19.7–23.9) | .445 |
| Serum albumin (g/dL) | 3.7 (3.5–3.9) | 3.6 (3.4–3.8) | 3.7 (3.5–3.9) | 3.8 (3.6–4.0) | <.001 | 3.7 (3.4–3.9) | 3.7 (3.6–3.9) | 3.7 (3.6–3.9) | .001 |
| CRP (mg/dL) | 0.14 (0.05–0.41) | 0.25 (0.08–0.86) | 0.12 (0.05–0.39) | 0.08 (0.04–0.20) | <.001 | 0.20 (0.07–0.72) | 0.14 (0.05–0.32) | 0.08 (0.04–0.25) | <.001 |
Data are expressed as numbers, percentages or median (interquartile range).
P-values were by χ2 test or by Kruskal–Wallis test.
P-values across tertiles.
IV iron, intravenous iron preparation; BMI, body mass index.
Fetuin-A but not T50 as an independent predictor of new CVD events
We had 190 new CVD events during follow-up period (Supplementary data, Fig. S1). Figure 1 shows cumulative incidence function for new CVD events according to tertiles of fetuin-A and T50 considering the competing risk of death. A lower fetuin-A and a shorter T50 was associated with a higher risk of new CVD events. We examined the unadjusted and adjusted associations of fetuin-A and T50 with new CVD events (Table 2). Compared with T3 as the referent, T1 and T2 of fetuin-A were significantly associated with increased sdHRs in Fine–Gray models unadjusted and adjusted for age and pre-existing CVD (Parsimonious Model 1), the five major demographic factors (Parsimonious Model 2), and additional variables related to CKD-MBD (Extended Model 1), anemia in CKD (Extended Model 2), traditional risk factors (Extended Model 3), inflammation and wasting (Extended Model 4), and log T50 (T2 but not T1) (Extended Model 5). In contrast, T50 was not associated with new CVD events in any adjusted models (Parsimonious Model and Extended Models 1 to 4 and 6).
Figure 1:
Cumulative incidence of new CVD events in association with fetuin-A (A) and T50 (B).
Table 2:
Association of fetuin-A or T50 with CVD events by Fine–Gray models.
| Tertiles of fetuin-A | Tertiles of T50 | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| T1 | T2 | T3 | Total | T1 | T2 | T3 | Total | ||
| Number of cases | 81 | 70 | 39 | 190 | 76 | 64 | 50 | 190 | |
| Patient-years | 531.7 | 576.6 | 688.7 | 1797.0 | 556.0 | 603.2 | 637.7 | 1797.0 | |
| Crude rate (cases per 1000 patient-years) | 152.3 | 121.4 | 56.6 | 105.7 | 136.7 | 106.1 | 78.4 | 105.7 | |
| Model | Adjustment | Subdistribution HR (95% CI) | P for trend | Subdistribution HR (95% CI) | P for trend | ||||
| Unadjusted | 2.53 (1.74–3.69)*** | 2.12 (1.44–3.13)*** | Ref. | < 0.001 | 1.70 (1.20–2.42)** | 1.38 (0.96–1.98) | Ref. | .002 | |
| Parsimonious 1 | Age and pre-existing CVD | 1.73 (1.15–2.61)** | 189 (1.27–2.77)** | Ref. | 0.012 | 1.39 (0.96–2.00) | 1.32 (0.92–1.90) | Ref. | .069 |
| Parsimonious 2 | Age, sex, duration of hemodialysis, DKD, pre-existing CVD | 1.65 (1.09–2.51)* | 1.82 (1.23–2.69)** | Ref. | 0.027 | 1.40 (0.97–2.02) | 1.36 (0.95–1.96) | Ref. | .063 |
| Extended 1 | Parsimonious Model 2 + Ca, P, intact PTH, use of VDRA | 1.59 (1.04–2.41)* | 1.79 (1.21–2.66)** | Ref. | 0.045 | 1.31 (0.90–1.91) | 1.35 (0.94–1.94) | Ref. | .140 |
| Extended 2 | Parsimonious Model 2 + Ht, dose of ESA, use of IV iron | 1.66 (1.09–2.53)* | 1.82 (1.23–2.70) ** | Ref. | 0.025 | 1.41 (0.98–2.04) | 1.36 (0.95–1.96) | Ref. | .057 |
| Extended 3 | Parsimonious Model 2 + Hypertension, smoking, dyslipidemia | 1.74 (1.14–2.64)** | 1.83 (1.24–2.71)** | Ref. | 0.012 | 1.39 (0.96–2.01) | 1.29 (0.90–1.86) | Ref. | .072 |
| Extended 4 | Parsimonious Model 2 + BMI, log CRP, albumin | 1.57 (1.00–2.44) * | 1.77 (1.18–2.64)** | Ref. | 0.056 | 1.36 (0.92–2.00) | 1.31 (0.90–1.90) | Ref. | .100 |
| Extended 5 | Parsimonious Model 2 + log T50 | 1.46 (0.91–2.34) | 1.70 (1.13–2.57)* | Ref. | 0.170 | ||||
| Extended 6 | Parsimonious Model 2 + fetuin-A | 1.21 (0.77–1.89) | 1.29 (0.88–1.89) | Ref. | .340 | ||||
Death as a competing risk (Fine–Gray models).
***P < .001, **P < .01, *P < .05.
IV iron, intravenous iron preparation; BMI, body mass index.
Next, fetuin-A or log T50 were entered as continuous variable into Fine–Gray models (Supplementary data, Table S1). Although fetuin-A was significantly associated with CVD events in Parsimonious Model 1, these significant associations disappeared in other adjusted models (Parsimonious Model 2 and Extended Models 1–5). However, when CVD events were divided into atherosclerotic (n = 90) and non-atherosclerotic CVD events (n = 100), fetuin-A was significantly associated with non-atherosclerotic CVD events in all models except Extended Model 4. On the other hand, no significant association of fetuin-A with atherosclerotic CVD events was found in any model. In contrast, log T50 was not associated with new CVD events regardless of the causes. These findings suggest that fetuin-A was consistently associated with new CVD events, especially non-atherosclerotic CVD events, whereas T50 was not.
T50 but not fetuin-A as an independent predictor of mortality after CVD events
We focused on deaths after new CVD events. Among the 190 patients who experienced a new CVD event, 59 patients died during the subsequent follow-up for 2.5 years (Supplementary data, Fig. S1). Fetuin-A was not associated with death after CVD events in unadjusted and all adjusted Cox models (Fig. 2A) (Table 3). In contrast, shorter T50 (T1 + T2) were significantly associated with higher mortality after CVD events compared with T3 (Fig. 2B). The inverse association of T50 with mortality after CVD events remained significant after adjustments for any models. Interestingly, the significant association of T50 with mortality after CVD events was independent of fetuin-A (Extended Model 5).
Figure 2:
Cumulative incidence of death after new CVD events in association with fetuin-A (A) and T50 (B).
Table 3:
Association of fetuin-A or T50 with death after CVD events by Cox models.
| Tertiles of fetuin-A | Tertiles of T50 | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| T1 | T2 | T3 | Total | T1 | T2 | T3 | Total | ||
| Number of cases | 28 | 21 | 7 | 56 | 29 | 21 | 6 | 56 | |
| Patients-years | 119.0 | 99.3 | 51.9 | 270.2 | 109.7 | 85.9 | 74.6 | 270.2 | |
| Crude rate (cases per 1000 patient-years) | 235.3 | 211.4 | 134.9 | 207.2 | 264.3 | 244.4 | 80.4 | 207.2 | |
| T1 + T2 | T3 | T1 + T2 | T3 | ||||||
| Model | Adjustment | Hazard ratio (95% CI) | P | HR (95% CI) | P | ||||
| Unadjusted | 1.72 (0.78–3.80) | Ref. | 0.180 | 3.11 (1.33–7.24) ** | Ref. | .009 | |||
| Parsimonious | Age and pre-existing CVD | 1.34 (0.60–3.02) | Ref. | 0.460 | 3.31 (1.42–7.74) ** | Ref. | .007 | ||
| Extended 1 | Parsimonious Model + sex | 1.34 (0.59–3.01) | Ref. | 0.460 | 3.31 (1.42–7.74) ** | Ref. | .007 | ||
| Extended 2 | Parsimonious Model + duration of hemodialysis | 1.42 (0.63–3.21) | Ref. | 0.390 | 3.20 (1.37–7.50) ** | Ref. | .009 | ||
| Extended 3 | Parsimonious Model + DKD | 1.34 (0.59–3.01) | Ref. | 0.470 | 3.42 (1.45–8.04) ** | Ref. | .007 | ||
| Extended 4 | Parsimonious Model + log 50 | 0.90 (0.38–2.16) | Ref. | 0.810 | |||||
| Extended 5 | Parsimonious Model + fetuin-A | 3.24 (1.37–7.65) ** | Ref. | .009 | |||||
**P < .01.
IV iron, intravenous iron preparation; BMI, body mass index.
The preferential association of death after new CVD events with shorter T50 over lower fetuin-A was confirmed in the analyses in which these exposure variables were handled as continuous variables.
Sensitivity analyses
As sensitivity analyses, we examined the association of fetuin-A or T50 with new CVD events by Cox proportional models in a similar way (Supplementary data, Table S3). Fetuin-A showed significant association with new CVD events in all adjusted models (Parsimonious Models and Extended Models 1–4) except for a model adjusted for T50 (Extended Model 5). In contrast, T50 was not associated with new CVD events in any adjusted models (Parsimonious Models and Extended Models 1–4 and 6). These results by Cox models are almost the same as those by Fine–Gray models (Table 2).
Fetuin-A, T50 and all-cause mortality
To confirm that our cohort shows the associations of fetuin-A and T50 with all-cause mortality which were reported in previous reports [1, 2, 12, 15], we conducted additional analysis using all-cause mortality as the outcome. During the follow-up period, 104 patients died. Both fetuin-A and T50 were inversely associated with all-cause mortality (Supplementary data, Fig. S2). Both fetuin-A and T50 showed significant association with all-cause mortality in unadjusted and all adjusted Cox models except for a model adjusted for inflammation and wasting (Extended Model 4) in fetuin-A. The significant association of fetuin-A or T50 with all-cause mortality disappeared when adjusted each other (Extended Models 5 and 6; Supplementary data, Table S4).
DISCUSSION
In this study of hemodialysis patients, we compared the associations of fetuin-A and T50 with two different cardiovascular outcomes, namely the occurrence of new CVD events and the death after the CVD events. Fetuin-A was an independent predictor of new CVD events, especially non-atherosclerotic CVD events, whereas it was not an independent predictor of death after the CVD events. In contrast, T50 was not an independent predictor of new CVD events, whereas it was an independent predictor of death after the CVD events. These results revealed that fetuin-A and T50 are associated with different aspects of cardiovascular risk in this population.
Previous cohort studies examined the prognostic abilities of fetuin-A and T50 in patients with CKD. Ketteler et al. were the first to report that a lower serum fetuin-A level was an independent predictor of all-cause death and cardiovascular death using data of patients undergoing hemodialysis [1], although they did not examine the association of fetuin-A with new onset of cardiovascular events or subsequent death. In the study by Smith et al. which showed the independent association of T50 with all-cause mortality in 184 patients with CKD stages 3–4, they showed a strong positive correlation between fetuin-A and T50 [12]. However, they did not examine the association of either fetuin-A or T50 with cardiovascular endpoints. Bostom et al. analyzed data from a cohort of 685 kidney transplant recipients and reported that both fetuin-A and T50 predicted composite cardiovascular events to a similar extent, although they did not examine the predictive values of fetuin-A and T50 for deaths after CVD events [16]. In this study of hemodialysis patients, we confirmed that fetuin-A and T50 were closely correlated, and that both a lower fetuin-A concentration and a shorter T50 predicted all-cause mortality. Furthermore, we revealed that a lower fetuin-A independently predicted the onset of new CVD events, especially non-atherosclerotic CVD events, whereas T50 did not. In contrast, our results showed a shorter T50 independently predicted deaths after CVD events, whereas fetuin-A did not. Thus, our study is the first that revealed the distinct prognostic abilities of fetuin-A and T50 in predicting the two aspects of cardiovascular outcomes in CKD patients undergoing hemodialysis.
Patients with a lower kidney function have a higher risk of the onset of CVD events [34] and a higher risk of death after CVD events such as myocardial infarction [35]. Patients with kidney failure undergoing hemodialysis have a higher risk of the onset of CVD events and a higher risk of death after CVD events [20, 21]. These elevated risks for the onset of CVD (the first step) and subsequent death (the second step) could synergistically increase the death risk from CVD in these patients [19, 36]. Only a few previous studies reported differential associations of cardiovascular risk factors with the first and second steps of cardiovascular outcomes in patients with CKD. A higher level of non-HDL-C predicted atherosclerotic CVD events such as myocardial and cerebral infarctions but not subsequent death in patients undergoing hemodialysis [22]. Use of VDRAs was associated with lower odds for incident CVD events whereas it was not associated with death after CVD events in hemodialysis patients [23]. Serum level of insulin-like growth factor I (IGF-I) was not independently associated with new CVD events, but a lower IGF-I was an independent predictor of death after CVD events in patients undergoing hemodialysis [24]. Thus, the findings of fetuin-A and T50 in the present study expanded the list of risk factors which predict the onset of CVD events and death after CVD events differently.
The fact that a lower serum fetuin-A predicted the occurrence of new CVD events in this cohort study needs explanations. Since both fetuin-A and T50 are related to toxicity of calcium phosphate crystals, non-mineral actions of fetuin-A may explain the result of this study. First, fetuin-A is profoundly involved in metabolic disorders such as diabetes, dyslipidemia and the metabolic syndrome [8–10, 37]. In addition, hepatic fetuin-A is strongly induced by nutritional factors such as glucose, fructose and fatty acids [38, 39]. Thus, these relationships of fetuin-A with metabolic and nutritional factors may be relevant to its association with the occurrence of new CVD events. Second, lower serum fetuin-A levels may be associated with a higher indoxyl sulfate, one of the major uremic toxins which can strongly produce reactive oxygen species, and potentially cause severe vascular injury. This speculation is based on the report showing that indoxyl sulfate suppressed production of fetuin-A in cultured hepatocytes [40]. Serum indoxyl sulfate and other uremic toxins may be confounding factors that affect the association between fetuin-A and the occurrence of CVD events, but were not adjusted in this analysis. In addition, we found the more preferential association of fetuin-A with non-atherosclerotic CVD events over atherosclerotic CVD events in this study. Fetuin-A-deficient mice showed severe myocardial calcification resulting in impaired cardiac output, diastolic dysfunction and catecholamine resistance [41]. This experimental report supports our clinical findings. What are the mechanisms for the preferential association of T50 with death after CVD events in this cohort of hemodialysis patients? Although fetuin-A is closely associated with T50, T50 is an integrated indicator of calcification stress which is affected not only by fetuin-A but also by calcium, phosphate, magnesium and serum albumin [11]. Thus, these factors may be related to the above-mentioned result of this study. According to a previous cohort study in hemodialysis patients [23], serum levels of calcium and phosphate were significantly predictive of death after CVD events but not predictive of the onset of CVD events. Another cohort study of hemodialysis patients reported that a lower serum albumin was associated not only with incident myocardial infarction and incident cerebral infarction but also death after myocardial infarction and death after cerebral infarction [22]. In vitro and in vivo studies suggest that magnesium inhibits vascular calcification [42, 43]. Hypomagnesemia was associated with mortality in patients undergoing hemodialysis [44]. In addition, administration of magnesium oxide suppressed coronary artery calcification in CKD patients [45]. Thus, it is possible that these associations of calcium, phosphate, magnesium and albumin, in addition to fetuin-A, are integrated in T50, which may be predictive of the death after CVD events.
This study has certain limitations. First, this was a single-center study in Japan. We cannot generalize our conclusions for other ethnicities and countries. Second, we used single measurements of fetuin-A and T50 at baseline. Variations over time by repeated measures may change the association of fetuin-A or T50 with outcomes. Third, the measured samples were stored for a relatively long time, although we carefully kept them at −80°C without repeated freezing and thawing. Fourth, we did not evaluate serum CPP level directly, although we simultaneously measured both fetuin-A and T50. Fifth, we did not check serum magnesium levels. Sixth, it is possible that statistical power is not enough to detect the association of fetuin-A with mortality after CVD events because of the limited numbers of death after CVD events, although we tried adjustment as carefully as possible. Finally, the results do not necessarily indicate causality because of the observational study design.
In conclusion, we showed the differential associations of fetuin-A and T50 with the occurrence of new CVD events and the death after the new CVD events in a cohort of hemodialysis patients. These results indicate that fetuin-A and T50 are associated with different aspects of cardiovascular risk in this population. Further studies are needed to confirm our observations in a larger cohort and to understand the precise mechanisms behind the observed associations.
Supplementary Material
Contributor Information
Katsuhito Mori, Department of Nephrology, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
Tetsuo Shoji, Department of Vascular Medicine, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan; Vascular Science Center for Translational Research, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
Shinya Nakatani, Department of Metabolism, Endocrinology and Molecular Medicine, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
Hideki Uedono, Department of Metabolism, Endocrinology and Molecular Medicine, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
Akinobu Ochi, Department of Metabolism, Endocrinology and Molecular Medicine, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
Hisako Yoshida, Department of Medical Statistics, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
Yasuo Imanishi, Department of Metabolism, Endocrinology and Molecular Medicine, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
Tomoaki Morioka, Department of Metabolism, Endocrinology and Molecular Medicine, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
Yoshihiro Tsujimoto, Division of Internal Medicine, Inoue Hospital, Suita, Osaka, Japan.
Makoto Kuro-o, Center for Molecular Medicine, Jichi Medical University, Shimotsuke, Tochigi, Japan.
Masanori Emoto, Department of Nephrology, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan; Department of Vascular Medicine, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan; Department of Metabolism, Endocrinology and Molecular Medicine, Osaka Metropolitan University Graduate School of Medicine, Osaka, Japan.
FUNDING
This work was supported by a research grant from JSPS KAKENHI (Grant-in-Aid for Scientific Research, Grant Number 15K09294 (to T.S.).
AUTHORS’ CONTRIBUTIONS
K.M. designed the study, analyzed the data and drafted the manuscript. T.S. acquired data, managed the study and revised the manuscript. S.N., H.U. and A.O. measured serum samples. H.Y. analyzed the data. Y.I., T.M., Y.T., M.K. and M.E. critically reviewed and revised the manuscript. Each author contributed important intellectual content during manuscript drafting or revision and agrees to be personally accountable for the individual's own contributions and to ensure that questions pertaining to the accuracy or integrity of any portion of the work, even one in which the author was not directly involved, are appropriately investigated and resolved, including with documentation in the literature if appropriate.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author (K.M.) upon reasonable request.
CONFLICT OF INTEREST STATEMENT
MK received research funds from JSPS KAKENHI (Grant Number 22H00473), AMED-CREST (Grant Number JP23gm1510012h0002), and Cabinet Office, Government of Japan, Moonshot R&D Program for Agriculture, Forestry and Fisheries. The other authors declared no competing interest relevant to this study.
REFERENCES
- 1. Ketteler M, Bongartz P, Westenfeld R et al. Association of low fetuin-A (AHSG) concentrations in serum with cardiovascular mortality in patients on dialysis: a cross-sectional study. Lancet North Am Ed 2003;361:827–33. 10.1016/S0140-6736(03)12710-9 [DOI] [PubMed] [Google Scholar]
- 2. Hermans MM, Brandenburg V, Ketteler M et al. Association of serum fetuin-A levels with mortality in dialysis patients. Kidney Int 2007;72:202–7. 10.1038/sj.ki.5002178 [DOI] [PubMed] [Google Scholar]
- 3. Wang AY, Woo J, Lam CW et al. Associations of serum fetuin-A with malnutrition, inflammation, atherosclerosis and valvular calcification syndrome and outcome in peritoneal dialysis patients. Nephrol Dial Transplant 2005;20:1676–85. 10.1093/ndt/gfh891 [DOI] [PubMed] [Google Scholar]
- 4. Jahnen-Dechent W, Heiss A, Schafer C et al. Fetuin-A regulation of calcified matrix metabolism. Circ Res 2011;108:1494–509. 10.1161/CIRCRESAHA.110.234260 [DOI] [PubMed] [Google Scholar]
- 5. Jahnen-Dechent W, Smith ER. Nature's remedy to phosphate woes: calciprotein particles regulate systemic mineral metabolism. Kidney Int 2020;97:648–51. 10.1016/j.kint.2019.12.018 [DOI] [PubMed] [Google Scholar]
- 6. Jahnen-Dechent W, Pasch A. Solving the insoluble: calciprotein particles mediate bulk mineral transport. Kidney Int 2023;103:663–5. 10.1016/j.kint.2023.01.011 [DOI] [PubMed] [Google Scholar]
- 7. Kuro-o M. Klotho, phosphate and FGF-23 in ageing and disturbed mineral metabolism. Nat Rev Nephrol 2013;9:650–60. [DOI] [PubMed] [Google Scholar]
- 8. Mori K, Emoto M, Inaba M. Fetuin-A: a multifunctional protein. Recent Pat Endocr Metab Immune Drug Discov 2011;5:124–46. 10.2174/187221411799015372 [DOI] [PubMed] [Google Scholar]
- 9. Mori K, Emoto M, Yokoyama H et al. Association of serum fetuin-A with insulin resistance in type 2 diabetic and nondiabetic subjects. Diabetes Care 2006;29:468. 10.2337/diacare.29.02.06.dc05-1484 [DOI] [PubMed] [Google Scholar]
- 10. Pal D, Dasgupta S, Kundu R et al. Fetuin-A acts as an endogenous ligand of TLR4 to promote lipid-induced insulin resistance. Nat Med 2012;18:1279–85. 10.1038/nm.2851 [DOI] [PubMed] [Google Scholar]
- 11. Pasch A, Farese S, Graber S et al. Nanoparticle-based test measures overall propensity for calcification in serum. J Am Soc Nephrol 2012;23:1744–52. 10.1681/ASN.2012030240 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Smith ER, Ford ML, Tomlinson LA et al. Serum calcification propensity predicts all-cause mortality in predialysis CKD. J Am Soc Nephrol 2014;25:339–48. 10.1681/ASN.2013060635 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Bundy JD, Cai X, Scialla JJ et al. Serum calcification propensity and coronary artery calcification among patients with CKD: the CRIC (Chronic Renal Insufficiency Cohort) study. Am J Kidney Dis 2019;73:806–14. 10.1053/j.ajkd.2019.01.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Pasch A, Block GA, Bachtler M et al. Blood calcification propensity, cardiovascular events, and survival in patients receiving hemodialysis in the EVOLVE trial. Clin J Am Soc Nephrol 2017;12:315–22. 10.2215/CJN.04720416 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Zawada AM, Wolf M, Rincon Bello A et al. Assessment of a serum calcification propensity test for the prediction of all-cause mortality among hemodialysis patients. BMC Nephrol 2023;24:35. 10.1186/s12882-023-03069-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Bostom A, Pasch A, Madsen T et al. Serum calcification propensity and fetuin-A: biomarkers of cardiovascular disease in kidney transplant recipients. Am J Nephrol 2018;48:21–31. 10.1159/000491025 [DOI] [PubMed] [Google Scholar]
- 17. Dahle DO, Asberg A, Hartmann A et al. Serum calcification propensity is a strong and independent determinant of cardiac and all-cause mortality in kidney transplant recipients. Am J Transplant 2016;16:204–12. 10.1111/ajt.13443 [DOI] [PubMed] [Google Scholar]
- 18. Keyzer CA, de Borst MH, van den Berg E et al. Calcification propensity and survival among renal transplant recipients. J Am Soc Nephrol 2016;27:239–48. 10.1681/ASN.2014070670 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Wanner C, Amann K, Shoji T. The heart and vascular system in dialysis. Lancet North Am Ed 2016;388:276–84. 10.1016/S0140-6736(16)30508-6 [DOI] [PubMed] [Google Scholar]
- 20. Iseki K, Fukiyama K. Long-term prognosis and incidence of acute myocardial infarction in patients on chronic hemodialysis. The Okinawa Dialysis Study Group. Am J Kidney Dis 2000;36:820–5. 10.1053/ajkd.2000.17676 [DOI] [PubMed] [Google Scholar]
- 21. Iseki K, Fukiyama K; Okawa Dialysis Study Group. Clinical demographics and long-term prognosis after stroke in patients on chronic haemodialysis. The Okinawa Dialysis Study (OKIDS) Group. Nephrol Dial Transplant 2000;15:1808–13. 10.1093/ndt/15.11.1808 [DOI] [PubMed] [Google Scholar]
- 22. Shoji T, Masakane I, Watanabe Y et al. Elevated non-high-density lipoprotein cholesterol (non-HDL-C) predicts atherosclerotic cardiovascular events in hemodialysis patients. Clin J Am Soc Nephrol 2011;6:1112–20. 10.2215/CJN.09961110 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Shoji T, Marubayashi S, Shigematsu T et al. Use of vitamin D receptor activator, incident cardiovascular disease and death in a cohort of hemodialysis patients. Ther Apher Dial 2015;19:235–44. 10.1111/1744-9987.12274 [DOI] [PubMed] [Google Scholar]
- 24. Nakaya R, Shoji T, Nagata Y et al. Associations of serum insulin-like growth factor 1 with new cardiovascular events and subsequent death in hemodialysis patients: the DREAM cohort. J Atheroscler Thromb 2022;29:1153–65. 10.5551/jat.62992 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Kakiya R, Shoji T, Hayashi T et al. Decreased serum adrenal androgen dehydroepiandrosterone sulfate and mortality in hemodialysis patients. Nephrol Dial Transplant 2012;27:3915–22. 10.1093/ndt/gfs162 [DOI] [PubMed] [Google Scholar]
- 26. Mori K, Emoto M, Araki T et al. Effects of pioglitazone on serum fetuin-A levels in patients with type 2 diabetes mellitus. Metabolism 2008;57:1248–52. 10.1016/j.metabol.2008.04.019 [DOI] [PubMed] [Google Scholar]
- 27. Nakatani S, Mori K, Sonoda M et al. Association between serum zinc and calcification propensity (T(50)) in patients with type 2 diabetes mellitus and in vitro effect of exogenous zinc on T(50). Biomedicines 2020;8:337. 10.3390/biomedicines8090337 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Shoji T, Nakatani S, Kabata D et al. Comparative effects of etelcalcetide and maxacalcitol on serum calcification propensity in secondary hyperparathyroidism: a randomized clinical trial. Clin J Am Soc Nephrol 2021;16:599–612. 10.2215/CJN.16601020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Shimamoto K, Ando K, Fujita T et al. The Japanese Society of Hypertension Guidelines for the Management of Hypertension (JSH 2014). Hypertens Res 2014;37:253–390. [DOI] [PubMed] [Google Scholar]
- 30. Kinoshita M, Yokote K, Arai H et al. Japan Atherosclerosis Society (JAS) Guidelines for Prevention of Atherosclerotic Cardiovascular Diseases 2017. J Atheroscler Thromb 2018;25:846–984. 10.5551/jat.GL2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Noordzij M, Leffondre K, van Stralen KJ et al. When do we need competing risks methods for survival analysis in nephrology? Nephrol Dial Transplant 2013;28:2670–7. 10.1093/ndt/gft355 [DOI] [PubMed] [Google Scholar]
- 32. Hsu JY, Roy JA, Xie D et al. Statistical methods for cohort studies of CKD: survival analysis in the setting of competing risks. Clin J Am Soc Nephrol 2017;12:1181–9. 10.2215/CJN.10301016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Kanda Y. Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant 2013;48:452–8. 10.1038/bmt.2012.244 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Go AS, Chertow GM, Fan D et al. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004;351:1296–305. 10.1056/NEJMoa041031 [DOI] [PubMed] [Google Scholar]
- 35. Smith GL, Masoudi FA, Shlipak MG et al. Renal impairment predicts long-term mortality risk after acute myocardial infarction. J Am Soc Nephrol 2008;19:141–50. 10.1681/ASN.2007050554 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Nishizawa Y, Shoji T, Ishimura E et al. Paradox of risk factors for cardiovascular mortality in uremia: is a higher cholesterol level better for atherosclerosis in uremia? Am J Kidney Dis 2001;38:S4–7. 10.1053/ajkd.2001.27380 [DOI] [PubMed] [Google Scholar]
- 37. Ix JH, Shlipak MG, Brandenburg VM et al. Association between human fetuin-A and the metabolic syndrome: data from the Heart and Soul Study. Circulation 2006;113:1760–7. 10.1161/CIRCULATIONAHA.105.588723 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Takata H, Ikeda Y, Suehiro T et al. High glucose induces transactivation of the alpha2-HS glycoprotein gene through the ERK1/2 signaling pathway. J Atheroscler Thromb 2009;16:448–56. 10.5551/jat.No950 [DOI] [PubMed] [Google Scholar]
- 39. Dasgupta S, Bhattacharya S, Biswas A et al. NF-kappaB mediates lipid-induced fetuin-A expression in hepatocytes that impairs adipocyte function effecting insulin resistance. Biochem J 2010;429:451–62. 10.1042/BJ20100330 [DOI] [PubMed] [Google Scholar]
- 40. Ochi A, Mori K, Nakatani S et al. Indoxyl sulfate suppresses hepatic fetuin-A expression via the aryl hydrocarbon receptor in HepG2 cells. Nephrol Dial Transplant 2015;30:1683–92. 10.1093/ndt/gfv250 [DOI] [PubMed] [Google Scholar]
- 41. Merx MW, Schäfer C, Westenfeld R et al. Myocardial stiffness, cardiac remodeling, and diastolic dysfunction in calcification-prone fetuin-A-deficient mice. J Am Soc Nephrol 2005;16:3357–64. 10.1681/ASN.2005040365 [DOI] [PubMed] [Google Scholar]
- 42. Ter Braake AD, Shanahan CM, de Baaij JHF. Magnesium counteracts vascular calcification: passive interference or active modulation? Arterioscler Thromb Vasc Biol 2017;37:1431–45. 10.1161/ATVBAHA.117.309182 [DOI] [PubMed] [Google Scholar]
- 43. Diaz-Tocados JM, Peralta-Ramirez A, Rodríguez-Ortiz ME et al. Dietary magnesium supplementation prevents and reverses vascular and soft tissue calcifications in uremic rats. Kidney Int 2017;92:1084–99. 10.1016/j.kint.2017.04.011 [DOI] [PubMed] [Google Scholar]
- 44. Sakaguchi Y, Fujii N, Shoji T et al. Hypomagnesemia is a significant predictor of cardiovascular and non-cardiovascular mortality in patients undergoing hemodialysis. Kidney Int 2014;85:174–81. 10.1038/ki.2013.327 [DOI] [PubMed] [Google Scholar]
- 45. Sakaguchi Y, Hamano T, Obi Y et al. A randomized trial of magnesium oxide and oral carbon adsorbent for coronary artery calcification in predialysis CKD. J Am Soc Nephrol 2019;30:1073–85. 10.1681/ASN.2018111150 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author (K.M.) upon reasonable request.