bs_bs_banner
Nephrology 20 (2015) 625–631
Original Article
Suboptimal vitamin K status and its risk factors in a population of Chinese chronic haemodialysis patients YUNLIN FENG,1,2 YIZHE RUAN,3 QIANG HE,1,2 WENSONG ZHANG1,2 and LI WANG1,2 1 Nephrology Division, Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital, 2Nephrology Division, Affiliated Medical School of University of Electronic Science and Technology, and 3Department of Nephrology, 452nd Hospital of Chinese People’s Liberation Army, Chengdu, China
KEY WORDS: carboxylation, haemodialysis, osteocalcin, risk factor, vitamin K deficiency. Correspondence: Dr Li Wang, Nephrology Division, Sichuan Provincial People’s Hospital, No. 32 2nd West First Ring Road, Chengdu, Sichuan, China. E-mail: [email protected] Accepted for publication 19 April 2015. Accepted manuscript online 28 April 2015. doi:10.1111/nep.12494 Conflict of interest: This study was supported by Doctor Foundation of Sichuan Provincial People’s Hospital (2014) and Foundation of Sichuan Provincial Health Department (140078). We declare no other financial conflicts.
SUMMARY AT A GLANCE Vitamin K deficiency is common in dialysis patients. Defining vitamin K deficiency by %ucOC, suboptimal vitamin K levels appear common in Chinese patients. Time on dialysis and LDL cholesterol level predict vitamin K deficiency.
ABSTRACT: Aims: Vitamin K deficiency is known to be common in haemodialysis patients and associates with adverse outcomes in this population, particularly vascular calcification. We aimed to investigate the vitamin K status in a population of Chinese haemodialysis (HD) patients. Methods: We collected demographic and biochemical data from a population of maintenance HD (MHD) patients in our unit and a control group composed of healthy subjects from our outpatient clinic. Fasting serum samples from all subjects were collected for the measurement of known vitamin K-dependent proteins i.e. matrix Gla protein (MGP), osteocalcin (OC) and uncarboxylated osteocalcin (ucOC). We also quantified the fraction of ucOC to total OC (%ucOC). Differences of these parameters between groups were analyzed and risk factors of vitamin K deficiency based on the definition as per %ucOC were investigated. Results: We enrolled 93 MHD patients as a test group and 93 healthy subjects as a control group. There was no significant difference in MGP between groups (4.0 ± 2.8 pg/mL in MHD vs 4.2 ± 1.2 pg/mL in control; P = 0.676). Mean %ucOC was significantly greater in the MHD patients as compared to control subjects (76.4 ± 20.0% in MHD vs 48.56 ± 15.5%; P < 0.001). Time on dialysis and low-density lipoprotein cholesterol level appeared to be indicators of vitamin K deficiency, with the former being an independent risk factor. Conclusions: Defining Vitamin K deficiency by %ucOC, suboptimal vitamin K levels appear common in Chinese MHD patients. Time on dialysis and LDL cholesterol level predict vitamin K deficiency.
Vitamin K is a key cofactor of γ-glutamylcarboxylase (GGCX) in mammals and functions by mediating carboxylation of residues of Gla proteins, converting these proteins into physiologically active forms.1 Gla proteins can be either intra- or extra-hepatic based on their distribution in the body. Intra-hepatic Gla proteins are various coagulation factors synthesized by the liver, of which the carboxylation activation is mainly mediated by vitamin K1.2 One of these proteins that has been extensively studied is plasma abnormal prothrombin (PIVKA II). Extra-hepatic Gla proteins widely distribute in various organs including skin, skeleton, kidney, vascular, pancreas etc. Their carboxylation activation is mainly mediated by vitamin K2. Extra-hepatic proteins © 2015 Asian Pacific Society of Nephrology
function locally and participate in many physiology activities, such as vascular regulation and skeleton metabolism. The most extensively studied extra-hepatic Gla proteins are osteocalcin (OC), matrix Gla-protein (MGP) and growtharrest-specific gene-6 protein (Gas6).3 The schematic diagram of carboxylation activation of Gla-protein is shown in Figure 1. Vitamin K deficiency leads to impaired Gla proteins carboxylation.4 As mentioned above, some extra-hepatic Gla proteins participate actively in vascular tissue. Dysfunction of these proteins may cause impaired vascular metabolism, sometimes leading to accelerated vascular calcification, which is a major cause of cardiovascular death in haemodi625
Y Feng et al.
consent. The protocol has been approved by the Ethics Committee of Sichuan Provincial People’s Hospital and the study was conducted in accordance with the Declaration of Helsinki.
Clinical and biochemical evaluation
Fig. 1 Chematic diagram of carboxylation activation of Gla-protein. GGCX, γ-glutamylcarboxylase; K, vitamin K; KH2, vitamin K in reduction form; KO, vitamin K in oxidized form; VKOR, vitamin-K epoxide reductase; VSMC, vascular smooth muscle cell.
alysis (HD) patients. In fact, the prevalence of vitamin K deficiency in the HD population has been confirmed by multiple studies. Rachel et al.5 reported in 142 HD patients, 29% had very low level of phylloquinone. Kohlmeier et al.6 also reported suboptimal vitamin K status in 68 HD patients. Results from Ralf et al.4 indicate daily vitamin K supplementation helps to decrease inactive MGP level. Unfortunately, there have been no such studies in the Chinese population yet. Vitamin K status can be quantified (SI: pmol/L) using high-performance liquid chromatography (HPLC);7 however, this method requires specific and expensive equipment. It has been suggested that vitamin K-dependent proteins (i.e. PIVKA II, MGP, OC, Gas6) be used to reflect vitamin K status instead. Previous studies show the percentage of ucOC (uncarboxylated osteocalcin) in total OC (%ucOC) reflects vitamin K status in the body and can be used as a surrogate marker to diagnose vitamin K deficiency.8 It is premised, the higher the %ucOC, the poorer the vitamin K status. In this study, we aimed to determine the vitamin K status in Chinese HD patients, using %ucOC as a surrogate index. Another known vitamin K-dependent protein, i.e. MGP was also quantified. Potential factors related to vitamin K deficiency were investigated.
METHODS Study population This cross-sectional study enrolled maintenance haemodialysis (MHD) patients older than 18 years of age receiving regular HD treatment, which included one haemodiafiltration (HDF) session per week in our HD centre and healthy subjects from the clinic. The study population was divided to a MHD group and a control group accordingly. All enrolled subjects provided written informed 626
Demographic data included age at study entry (y, years old), gender, time on dialysis (m, month), and dialysis frequency. Fasting serum samples were collected for laboratory investigations. For MHD subjects, the serum sample was obtained immediately before the first haemodialysis session of the week. Before measurement, serum samples were stored at −80°C within 30 min of sampling. Most of the laboratory investigations were conducted in the central lab in our hospital, including urea nitrogen (BUN; mmol/L), creatinine (Cr; μmol/L), triglyceride (TG; mmol/L), total cholesterol (TC; mmol/L), low-density lipoprotein cholesterol (LDL-C; mmol/L) and albumin (ALB; g/L) for each subject, as well as intact parathyroid hormone (iPTH; pg/mL), ferritin (FER; ng/mL), calcium (Ca; mmol/L) and phosphate (P; mmol/L) for MHD patients. Corrected calcium (cCa, mmol/L) was calculated. Serum MGP, OC and ucOC were qualified using enzyme linked immunosorbent assay (ELISA) kits (R&B). ucOC/OC ratio was expressed as a percentage of ucOC in total OC (i.e. %ucOC).
Statistical analysis The Kolmogorov-Smirnov test was used to analyze data distribution for normality. Descriptive statistics (mean ± SD for normal continuous data, median and range for non-normal continuous data, and frequency for categorical data) were generated for all variables. Variables between the two groups were compared by means of Student t-test for continuous variables and χ2 test for binary variables. Pearson correlation coefficient was used to assess bivariate relationship between %ucOC and other variables. For MGP and %ucOC, if a significant difference between MHD and control groups had been discovered, the value of mean + 2SD in the control group would be used as the cut-off value to differentiate the MHD group into ‘elevation’ and ‘non-elevation’ subgroups. Afterwards, Logistical regression model was used to identify risk factors for the elevation. Firstly, a univariable regression was done to identify factors with P < 0.1, which were then fitted into a multi-variable regression to identify significantly independent risk factors. All data were analyzed using the SPSS software 17.0 (SPSS, Chicago, IL, USA). SPSS 17.0 and GraphPad Prism software 5.0 (GraphPad Software, La Jolla, CA, USA) were both used for graphing. All statistical tests were two-sided, and a P-value of less than 0.05 was considered statistically significant.
RESULTS Demographic and lab investigations Demographic and laboratory investigation results for the study population are listed in Table 1. There were 186 subjects in all, including 93 MHD patients and 93 control healthy subjects. The two groups had no difference in age or gender composition. The median time on dialysis was 52.0 (25th to 75th percentiles, 18–146) months. The majority © 2015 Asian Pacific Society of Nephrology
Vitamin K deficiency in Chinese HD patients
Table 1 Demographic and laboratory investigation results in the study population Variable
Subjects (n = 186)
Demographic Age (years) (mean ± SD) Male (n, %)/Female (n, %) Time on dialysis (m) (median, range) Dialysis frequency Twice/week (n, %) Three times/week (n, %) Lab investigations (serum) BUN (mmol/L) (mean ± SD) Cr (μmol/L) (mean ± SD) TG (mmol/L) (mean ± SD) TC (mmol/L) (mean ± SD) LDL-C (mmol/L) (mean ± SD) ALB (g/L) (mean ± SD) iPTH (pg/mL) (median, range) FER (ng/mL) (median, range) cCa (mmol/L) (mean ± SD) P (mmol/L) (mean ± SD)
P-value
MHD Group (n = 93)
Control Group (n = 93)
53.8 ± 14.5 50(53.8%)/43 (46.2%) 52.0 (6–240)
47.8 ± 14.3 60 (64.5%)/33 (35.5%) –
0.966 0.322 –
– –
– –
4.7 ± 1.2 74.6 ± 15.3 2.2 ± 1.4 4.8 ± 0.9 2.8 ± 0.6 48.3 ± 1.4 – – – –
Nephrology 20 (2015) 625–631
Original Article
Suboptimal vitamin K status and its risk factors in a population of Chinese chronic haemodialysis patients YUNLIN FENG,1,2 YIZHE RUAN,3 QIANG HE,1,2 WENSONG ZHANG1,2 and LI WANG1,2 1 Nephrology Division, Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital, 2Nephrology Division, Affiliated Medical School of University of Electronic Science and Technology, and 3Department of Nephrology, 452nd Hospital of Chinese People’s Liberation Army, Chengdu, China
KEY WORDS: carboxylation, haemodialysis, osteocalcin, risk factor, vitamin K deficiency. Correspondence: Dr Li Wang, Nephrology Division, Sichuan Provincial People’s Hospital, No. 32 2nd West First Ring Road, Chengdu, Sichuan, China. E-mail: [email protected] Accepted for publication 19 April 2015. Accepted manuscript online 28 April 2015. doi:10.1111/nep.12494 Conflict of interest: This study was supported by Doctor Foundation of Sichuan Provincial People’s Hospital (2014) and Foundation of Sichuan Provincial Health Department (140078). We declare no other financial conflicts.
SUMMARY AT A GLANCE Vitamin K deficiency is common in dialysis patients. Defining vitamin K deficiency by %ucOC, suboptimal vitamin K levels appear common in Chinese patients. Time on dialysis and LDL cholesterol level predict vitamin K deficiency.
ABSTRACT: Aims: Vitamin K deficiency is known to be common in haemodialysis patients and associates with adverse outcomes in this population, particularly vascular calcification. We aimed to investigate the vitamin K status in a population of Chinese haemodialysis (HD) patients. Methods: We collected demographic and biochemical data from a population of maintenance HD (MHD) patients in our unit and a control group composed of healthy subjects from our outpatient clinic. Fasting serum samples from all subjects were collected for the measurement of known vitamin K-dependent proteins i.e. matrix Gla protein (MGP), osteocalcin (OC) and uncarboxylated osteocalcin (ucOC). We also quantified the fraction of ucOC to total OC (%ucOC). Differences of these parameters between groups were analyzed and risk factors of vitamin K deficiency based on the definition as per %ucOC were investigated. Results: We enrolled 93 MHD patients as a test group and 93 healthy subjects as a control group. There was no significant difference in MGP between groups (4.0 ± 2.8 pg/mL in MHD vs 4.2 ± 1.2 pg/mL in control; P = 0.676). Mean %ucOC was significantly greater in the MHD patients as compared to control subjects (76.4 ± 20.0% in MHD vs 48.56 ± 15.5%; P < 0.001). Time on dialysis and low-density lipoprotein cholesterol level appeared to be indicators of vitamin K deficiency, with the former being an independent risk factor. Conclusions: Defining Vitamin K deficiency by %ucOC, suboptimal vitamin K levels appear common in Chinese MHD patients. Time on dialysis and LDL cholesterol level predict vitamin K deficiency.
Vitamin K is a key cofactor of γ-glutamylcarboxylase (GGCX) in mammals and functions by mediating carboxylation of residues of Gla proteins, converting these proteins into physiologically active forms.1 Gla proteins can be either intra- or extra-hepatic based on their distribution in the body. Intra-hepatic Gla proteins are various coagulation factors synthesized by the liver, of which the carboxylation activation is mainly mediated by vitamin K1.2 One of these proteins that has been extensively studied is plasma abnormal prothrombin (PIVKA II). Extra-hepatic Gla proteins widely distribute in various organs including skin, skeleton, kidney, vascular, pancreas etc. Their carboxylation activation is mainly mediated by vitamin K2. Extra-hepatic proteins © 2015 Asian Pacific Society of Nephrology
function locally and participate in many physiology activities, such as vascular regulation and skeleton metabolism. The most extensively studied extra-hepatic Gla proteins are osteocalcin (OC), matrix Gla-protein (MGP) and growtharrest-specific gene-6 protein (Gas6).3 The schematic diagram of carboxylation activation of Gla-protein is shown in Figure 1. Vitamin K deficiency leads to impaired Gla proteins carboxylation.4 As mentioned above, some extra-hepatic Gla proteins participate actively in vascular tissue. Dysfunction of these proteins may cause impaired vascular metabolism, sometimes leading to accelerated vascular calcification, which is a major cause of cardiovascular death in haemodi625
Y Feng et al.
consent. The protocol has been approved by the Ethics Committee of Sichuan Provincial People’s Hospital and the study was conducted in accordance with the Declaration of Helsinki.
Clinical and biochemical evaluation
Fig. 1 Chematic diagram of carboxylation activation of Gla-protein. GGCX, γ-glutamylcarboxylase; K, vitamin K; KH2, vitamin K in reduction form; KO, vitamin K in oxidized form; VKOR, vitamin-K epoxide reductase; VSMC, vascular smooth muscle cell.
alysis (HD) patients. In fact, the prevalence of vitamin K deficiency in the HD population has been confirmed by multiple studies. Rachel et al.5 reported in 142 HD patients, 29% had very low level of phylloquinone. Kohlmeier et al.6 also reported suboptimal vitamin K status in 68 HD patients. Results from Ralf et al.4 indicate daily vitamin K supplementation helps to decrease inactive MGP level. Unfortunately, there have been no such studies in the Chinese population yet. Vitamin K status can be quantified (SI: pmol/L) using high-performance liquid chromatography (HPLC);7 however, this method requires specific and expensive equipment. It has been suggested that vitamin K-dependent proteins (i.e. PIVKA II, MGP, OC, Gas6) be used to reflect vitamin K status instead. Previous studies show the percentage of ucOC (uncarboxylated osteocalcin) in total OC (%ucOC) reflects vitamin K status in the body and can be used as a surrogate marker to diagnose vitamin K deficiency.8 It is premised, the higher the %ucOC, the poorer the vitamin K status. In this study, we aimed to determine the vitamin K status in Chinese HD patients, using %ucOC as a surrogate index. Another known vitamin K-dependent protein, i.e. MGP was also quantified. Potential factors related to vitamin K deficiency were investigated.
METHODS Study population This cross-sectional study enrolled maintenance haemodialysis (MHD) patients older than 18 years of age receiving regular HD treatment, which included one haemodiafiltration (HDF) session per week in our HD centre and healthy subjects from the clinic. The study population was divided to a MHD group and a control group accordingly. All enrolled subjects provided written informed 626
Demographic data included age at study entry (y, years old), gender, time on dialysis (m, month), and dialysis frequency. Fasting serum samples were collected for laboratory investigations. For MHD subjects, the serum sample was obtained immediately before the first haemodialysis session of the week. Before measurement, serum samples were stored at −80°C within 30 min of sampling. Most of the laboratory investigations were conducted in the central lab in our hospital, including urea nitrogen (BUN; mmol/L), creatinine (Cr; μmol/L), triglyceride (TG; mmol/L), total cholesterol (TC; mmol/L), low-density lipoprotein cholesterol (LDL-C; mmol/L) and albumin (ALB; g/L) for each subject, as well as intact parathyroid hormone (iPTH; pg/mL), ferritin (FER; ng/mL), calcium (Ca; mmol/L) and phosphate (P; mmol/L) for MHD patients. Corrected calcium (cCa, mmol/L) was calculated. Serum MGP, OC and ucOC were qualified using enzyme linked immunosorbent assay (ELISA) kits (R&B). ucOC/OC ratio was expressed as a percentage of ucOC in total OC (i.e. %ucOC).
Statistical analysis The Kolmogorov-Smirnov test was used to analyze data distribution for normality. Descriptive statistics (mean ± SD for normal continuous data, median and range for non-normal continuous data, and frequency for categorical data) were generated for all variables. Variables between the two groups were compared by means of Student t-test for continuous variables and χ2 test for binary variables. Pearson correlation coefficient was used to assess bivariate relationship between %ucOC and other variables. For MGP and %ucOC, if a significant difference between MHD and control groups had been discovered, the value of mean + 2SD in the control group would be used as the cut-off value to differentiate the MHD group into ‘elevation’ and ‘non-elevation’ subgroups. Afterwards, Logistical regression model was used to identify risk factors for the elevation. Firstly, a univariable regression was done to identify factors with P < 0.1, which were then fitted into a multi-variable regression to identify significantly independent risk factors. All data were analyzed using the SPSS software 17.0 (SPSS, Chicago, IL, USA). SPSS 17.0 and GraphPad Prism software 5.0 (GraphPad Software, La Jolla, CA, USA) were both used for graphing. All statistical tests were two-sided, and a P-value of less than 0.05 was considered statistically significant.
RESULTS Demographic and lab investigations Demographic and laboratory investigation results for the study population are listed in Table 1. There were 186 subjects in all, including 93 MHD patients and 93 control healthy subjects. The two groups had no difference in age or gender composition. The median time on dialysis was 52.0 (25th to 75th percentiles, 18–146) months. The majority © 2015 Asian Pacific Society of Nephrology
Vitamin K deficiency in Chinese HD patients
Table 1 Demographic and laboratory investigation results in the study population Variable
Subjects (n = 186)
Demographic Age (years) (mean ± SD) Male (n, %)/Female (n, %) Time on dialysis (m) (median, range) Dialysis frequency Twice/week (n, %) Three times/week (n, %) Lab investigations (serum) BUN (mmol/L) (mean ± SD) Cr (μmol/L) (mean ± SD) TG (mmol/L) (mean ± SD) TC (mmol/L) (mean ± SD) LDL-C (mmol/L) (mean ± SD) ALB (g/L) (mean ± SD) iPTH (pg/mL) (median, range) FER (ng/mL) (median, range) cCa (mmol/L) (mean ± SD) P (mmol/L) (mean ± SD)
P-value
MHD Group (n = 93)
Control Group (n = 93)
53.8 ± 14.5 50(53.8%)/43 (46.2%) 52.0 (6–240)
47.8 ± 14.3 60 (64.5%)/33 (35.5%) –
0.966 0.322 –
– –
– –
4.7 ± 1.2 74.6 ± 15.3 2.2 ± 1.4 4.8 ± 0.9 2.8 ± 0.6 48.3 ± 1.4 – – – –
