Prevalence of vitamin D deficiency in Thai patients receiving various modalities of renal replacement therapy

The present study was a retrospective, observational, single tertiary center, cross-sectional study performed during June to July 2014 and was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University (IRB No. 449/57). The study included data from 111 consecutive adult Thai patients who received either HD, PD, or KT at King Chulalongkorn Memorial Hospital, Bangkok, Thailand, located at a latitude of just 13°43′N (suggesting strong exposure to sunlight). Demographic data including age, sex, comorbidity, dialysis vintage, and post-transplantation time were abstracted from a retrospective review of electronic medical records. No participants ever received oral ergocalciferol or cholecalciferol supplements.

Blood samples had been drawn midweek to obtain predialysis serum [25(OH)D], complete blood count, and calcium, phosphate, intact parathyroid hormone (iPTH), alkaline phosphatase, blood urea nitrogen (BUN), creatinine, uric acid, and albumin levels, lipid profile, and iron studies. Samples were immediately analyzed using standard automated methods. Serum [25(OH)D] was determined using a commercially available chemiluminescent immunoassay (DiaSorin Liaison, Stillwater, MN, USA). Interassay coefficient of variation (functional sensitivity) was 20% and the limit of detection was <4 ng/mL. According to the NKF/K/DOQI guidelines 25(OH)D deficiency is defined as <15 ng/mL, and insufficiency as 15–30 ng/mL, and sufficiency as >30 ng/mL. Total calcium (Ca) level was corrected using the equation: total Ca = serum Ca + [(4.0 – serum albumin) × 0.8]. Adequacy of dialysis was determined by using a urea kinetic model to achieve delivered Kt/V for HD patients. Total (renal and peritoneal) weekly Kt/V and total weekly creatinine clearance were assessed in patients receiving PD. Normalized protein nitrogen appearance (nPNA), representing daily protein intake, was derived from calculation of Kt/V, 24 h dialysate, and 24 h urine collection during a steady state. The renal dietitian routinely followed and monitored dialysis and KT recipients for body weight, height, body mass index (BMI) in conjunction with nutritional status assessment by the Malnutrition Inflammation Score.

All data are expressed as mean ± standard deviation (SD) while non-normally distributed variables are expressed as median with interquartile range (IQR). The statistical analysis was performed using SPSS Statistics for Windows, version 17.0 (SPSS Inc, Chicago, IL, USA). Differences between variables were compared using a one-way analysis of variance (ANOVA) and unpaired Student t tests for normally distributed variables. Kruskal–Wallis and Mann–Whitney U tests (Wilcoxon rank-sum test) were used to compare groups with non-normally distributed variables. Selected variables were analyzed by binary logistic regression with stepwise forward selection (data presented as [odds ratio (OR); 95% confidence interval (CI)]). P < 0.05 was considered significant.

We analyzed data from 111 patients who received on-line hemodiafiltration (OL-HDF, n = 32), PD (n = 37; continuous ambulatory PD, CAPD =13 and automated PD, APD = 24) and KT (n = 42; KT deceased donor, DDKT = 23 and KT living, related donor, LRKT = 19). Table 2 details baseline characteristics of the 3 RRT groups. Low vitamin D status was noted in 100% of patients receiving PD (81% deficient, 19% insufficient), 100% of KT recipients (55% deficient, 45% insufficient), and 94% of patients receiving OL-HDF (50% deficient, 44% insufficient). Only 2% of the patients had normal serum [25(OH)D] (30–80 ng/mL) and were in OL-HDF group.

Table 2

Baseline characteristics of patients whose data were included in the study

Patients in the PD group had a significantly lower level of serum 25(OH)D than those in the OL-HDF and KT groups. All patients with severe vitamin D deficiency, that is, serum [25(OH)D] –5 ng/mL, were maintained on PD. Patients receiving PD were significantly older and had less dialysis vintage (years of treatment) than patients in the OL-HDF and KT groups. In the PD group, there were no significant differences in age, sex, history of diabetes, BMI, adequacy of dialysis, or serum albumin, calcium, or phosphate between those with CAPD and APD. Patients with APD tended to have a lower mean serum [25(OH)D] than those with CAPD, but the difference was not quite significant.

Vitamin D-deficient (<15 ng/mL) OL-HDF patients had significantly lower vascular access flow, lower hemoglobin levels, and significantly higher BMI than OL-HDF patients with 25(OH)D ≥15 ng/mL (Table 3).

Table 3

Laboratory variables in patients receiving dialysis

DDKT recipients had significantly lower mean [25(OH)D] than patients with LRKT (Table 4).

Table 4

Laboratory variables in kidney transplant (KT) recipients

Notably, LRKT recipients had significantly higher serum albumin and lower 24-hour urine protein loss than those in the DDKT group. There was no significant difference in age, renal function, BMI, immunosuppressive regimen, or prednisolone dose between KT recipient subgroups.

There were no significant differences in serum [25(OH)D] between patients of either sex, with or without diabetes, or high or low BMI (data not shown). Multiple logistic regression analysis showed treatment with renin angiotensin system (RAS) blockade, serum triglyceride, and iPTH were significantly associated with 25(OH)D deficiency in patients receiving RRT, and that serum uric acid and phosphate had a significant inverse association with vitamin D deficiency (Table 5).

Table 5

Multiple logistic regression analysis of 25-hydroxyvitamin D (calcifediol) deficiency (<15 ng/mL) in 111 patients with renal replacement therapy

Serum calcium, albumin, and female sex were not found significantly associated with serum 25(OH)D deficiency by multivariate analysis. Among patients receiving dialysis, treatment with RAS blockade and serum triglyceride level were significantly associated with 25(OH)D deficiency. In KT recipients, treatment with RAS blockade and older age were significantly associated with 25(OH)D deficiency after adjustment for confounding factors (data not shown).