Low Levels of Vitamin C in Dialysis Patients is Associated With Decreased Prealbumin and Increased C-reactive Protein

Kunying Zhang; Li Liu; Xuyang Cheng; Jie Dong; Qiuming Geng; Li Zuo


BMC Nephrology 

In This Article


Study Patients

The study was designed as a cross-sectional analysis including both MHD patients and CAPD patients in Peking University First Hospital. Patients aged between 18 and 80 years and could provide informed consent were included. Patients who had autoimmune disease, malignancy, hepatitis in active conditions, currently used steroids or immune-suppressants, had positive human immunodeficiency virus(HIV) serology, and had any kind of acute infection within one month were excluded. However, patients having stable cardiovascular disease, or using central venous catheter, using angiotensin converting enzyme inhibitor (ACEI), angiotensin receptor antagonist (ARB), or statins were not excluded.

MHD patients were dialyzed three times per week, 4–4.5 hours per session, among them, 54 patients were treated with conventional hemodialysis (HD), 51 with high-flux hemodialysis (HFD), and 12 with hemodiafiltration (HDF). CAPD patients received 3–4 daily exchanges of 2L of peritoneal dialysis solution with either 1.5% or 2.5% glucose.

MHD and CAPD patients were classified by vitamin C level into three subgroups, respectively, according to previous recommendation:[1,2] group A, or vitamin C deficiency group, had vitamin C < 2 ug/ml; group B, or vitamin C insufficiency group, had vitamin C level between 2–4 ug/ml and group C, or vitamin C normal or above the normal group, had vitamin C > 4 ug/ml.

This study was approved by the local ethics committee and written informed consent was obtained from each participant.

Sample Collection and Laboratory Procedures

Blood samples for vitamin C determination were drawn from the arteriovenous fistula just before dialysis session in MHD patients, and from a forearm vein in CAPD patients. All subjects were in fasting condition. Blood was collected into heparin-containing tube, and then was transported in an ice bath to the laboratory. Plasma was separated by centrifugation (2000 × g, 10 min) at 4°C within 30 minutes, an aliquot of 200 ul plasma was mixed with 200 ul of 10% metaphosphoric acid (MPA) and stored at -80°C until analysis within 2 weeks.

Vitamin C was determined by high-performance liquid chromatography (HPLC) (Agilent 1100 series, Agilent Technologies, USA) on a Diamonsil C18 column(150 mm × 4.6 mm, 3 um) at 25°C. For HPLC analysis, the thawed plasma with MPA was centrifuged at 22000 g for 10 min, 4°C, 20 ul supernatants was injected into the HPLC by an autosampler. The mobile phase consisted of 25 mM sodium dihydrogenphosphate pH 2.6, the flow rate was 0.5 mL/min, and observed wavelength was 244 nm. Intra-assay and inter-assay coefficients of variation were 2.7% and 2.5%, respectively. The reference range of vitamin C level in normal population for this assay was 4–14 mg/L.[1,2]

Prealbumin, high-sensitivity C-reactive protein (hsCRP), albumin, ferritin, calcium, serum bicarbonate, uric acid and hemoglobin were measured by standard techniques (the Clinical Laboratory of Peking University First Hospital, Beijing, China).

Statistical Analysis

The relationships between vitamin C and each of albumin, prealbumin and hsCRP levels were tested using Spearman correlation analysis. Levels of each of vitamin C, hsCRP, albumin, prealbumin, uric acid, and hemoglobin were compared between MHD and CAPD patients using t-test. The above parameters were also compared using one-way analysis of variance (1-way-ANOVA) among groups A, B and C in each of MHD and CAPD patients, respectively. Using each of prealbumin, albumin and log10hsCRP as outcome and vitamin C as predictor, three linear regression model were constructed to explore the relationship between categories of plasma vitamin C level(< 2 ug/ml,2–4 ug/ml and > 4 ug/ml) and each of prealbumin, albumin and log10hsCRP after gender, age, diabetes, modality of dialysis and other confounding effects were adjusted. Numeric parameters were assessed by χ2 test. Median, range, percentiles and Kruskal Wallis Test were used for non-normally distributed variables. Statistical analysis was performed using SPSS version11.5 (SPSS, Inc., Chicago, IL, USA). P values < 0.05 were considered as statistically significant.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.