Effects of Vitamin D₂ Supplementation on Vitamin D₃ Metabolism in Health and CKD

Abstract

Background and objectives

Vitamin D supplements are prescribed to correct low circulating concentrations of 25-hydroxyvitamin D. In CKD, vitamin D metabolism is complicated by decreased conversion of 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D by CYP27B1 and possibly decreased conversion of 25-hydroxyvitamin D to 24,25-dihydroxyvitamin D by CYP24A1. The aim of this study was to determine the effects of vitamin D₂ supplementation on vitamin D metabolism in health and CKD.

Design, setting, participants, & measurements

We conducted a treatment-only intervention study of 25 individuals with CKD (eGFR<60 ml/min per 1.73 m²) and 44 individuals without CKD from three academic centers, all with screening 25-hydroxyvitamin D <30 ng/ml. Each participant was prescribed vitamin D₂ (ergocalciferol) 50,000 IU orally twice weekly for 5 weeks. We tested whether changes in plasma concentrations of vitamin D metabolites and vitamin D metabolic ratios differed by CKD status. Plasma 1,25-dihydroxyvitamin D₃-to-25-hydroxyvitamin D₃ ratio and 24,25-dihydroxyvitamin D₃-to-25-hydroxyvitamin D₃ ratio were calculated as estimates of CYP27B1 and CYP24A1 function, respectively.

Results

With treatment, plasma 25-hydroxyvitamin D₂ and total 25-hydroxyvitamin D concentrations increased similarly for participants with and without CKD. For participants without CKD, 1,25-dihydroxyvitamin D₂ increased (2.8±1.3–32.9±1.4 pg/ml), whereas 1,25-dihydroxyvitamin D₃ decreased (45.6±1.9–14.6±1.9 pg/ml), resulting in no significant change in total 1,25-dihydroxyvitamin D; 1,25-dihydroxyvitamin D₃-to-25-hydroxyvitamin D₃ ratio decreased (3.0±0.2–1.7±0.2 pg/ng), and 24,25-dihydroxyvitamin D₃-to-25-hydroxyvitamin D₃ ratio increased (115.7±7.8–195.2±7.9 pg/ng). Individuals with CKD had lower baseline levels and smaller changes in magnitude for 1,25-dihydroxyvitamin D₂ (2.1±1.6–24.4±1.6 pg/ml; P interaction =0.01), 1,25-dihydroxyvitamin D₃-to-25-hydroxyvitamin D₃ ratio (1.8±0.2–1.1±0.2 pg/ng; P interaction =0.05), and 24,25-dihydroxyvitamin D₃-to-25-hydroxyvitamin D₃ ratio (72.0±9.1–110.3±9.3 pg/ng; P interaction <0.001). Fibroblast growth factor-23 and parathyroid hormone were not significantly changed in either group.

Conclusions

Vitamin D₂ supplementation decreases conversion of 25-hydroxyvitamin D₃ to 1,25-dihydroxyvitamin D₃ and induces vitamin D₃ catabolism as evidenced by changes in D₃ metabolites and vitamin D metabolic ratios. These effects occur without significant changes in fibroblast growth factor-23 or parathyroid hormone and are blunted in CKD.

Podcast

This article contains a podcast at https://www.asn-online.org/media/podcast/CJASN/2017_08_02_CJASNPodcast_17_09.mp3

Introduction

Vitamin D supplements are used widely in people with and without CKD, because low circulating concentrations of 25-hydroxyvitamin D [25(OH)D] are common and deficiency in 25(OH)D may contribute to adverse health outcomes (1,2). In 2003, the National Kidney Foundation-Kidney Disease Outcomes Quality Initiative guidelines recommended vitamin D supplementation for patients with stages 3 and 4 CKD and 25(OH)D<30 ng/ml, specifically using high-dose treatment with vitamin D₂ (ergocalciferol) (3). In 2009, Kidney Disease Improving Global Outcomes guidelines recommended using vitamin D in patients with stage 3 to predialysis stage 5 CKD with low circulating concentrations of 25(OH)D and elevated concentrations of parathyroid hormone (PTH) (4). Despite widespread use, the effects of vitamin D supplements on downstream vitamin D metabolism are unclear.

In health, vitamin D₃ is hydroxylated to 25-hydroxyvitamin D₃ [25(OH)D₃] by the hepatic cytochrome P450 enzyme CYP2R1; 25(OH)D₃ may then be converted by CYP27B1 to the active hormone 1,25-dihydroxyvitamin D₃ [1,25(OH)₂D₃; calcitriol]; by CYP24A1 to 24,25-dihydroxyvitamin D₃ [24,25(OH)₂D₃], which undergoes further metabolism and elimination; or to other metabolites that are less well described (5,6). Vitamin D₂ undergoes parallel metabolism through the same pathways. PTH induces CYP27B1 transcription and downregulates CYP24A1 (6,7). Fibroblast growth factor-23 (FGF-23) decreases CYP27B1 while inducing CYP24A1 (8–10). In addition, 1,25(OH)₂D [i.e., 1,25(OH)₂D₃ and 1,25(OH)₂D₂] regulates its own metabolism by inhibiting CYP27B1 and stimulating CYP24A1. Although these metabolic pathways are well defined, the fate of vitamin D supplements and their effects on this system are not well described.

CKD is known to strongly affect vitamin D metabolism (5). Specifically, CYP27B1-mediated production of 1,25(OH)₂D is markedly reduced (11–13), CYP24A1-mediated vitamin D clearance may also be reduced, and circulating concentrations of PTH and FGF-23 rise markedly, exerting opposing effects on CYP27B1 and CYP24A1 activity (5). The goal of this study was to analyze the physiologic effect of vitamin D supplementation on vitamin D metabolism in persons with and without CKD.

Materials and Methods

Study Participants

Between 2007 and 2009, participants with and without CKD were recruited from the Massachusetts General Hospital (MGH) outpatient nephrology clinics, the MGH campus-wide email solicitation, the University of Miami Hospital, the St. Louis University Hospital, and the affiliated John Cochran Veteran Affairs Medical Center. All potential candidates indicating interest in the study were invited to attend a screening visit, at which eligibility for inclusion was assessed and informed consent was obtained. Inclusion criteria included screening 25(OH)D<30 ng/ml at the local clinical laboratory and age >18 years old. Exclusion criteria are listed in Supplemental Material.

Of 192 participants screened, 69 participants consented, qualified, completed treatment with D₂ supplementation, attended end of study visits, and were included in the final analysis. Baseline plasma samples were used to measure creatinine and cystatin C traceable to isotope dilution mass spectrometry and ERM-DA471 (from the International Federation of Clinical Chemistry). Interassay coefficients were 1.5%–3.0%. The combined creatinine-cystatin C Chronic Kidney Disease Epidemiology Collaboration equation was used to calculate baseline eGFR. Persons with baseline eGFR <60 ml/min per 1.73m² were classified as having CKD. In the primary analysis, 15 participants with screening eGFR <60 ml/min per 1.73 m² (using the Modification of Diet in Renal Disease formula) but baseline eGFR≥60 ml/min per 1.73m² were analyzed as the group without CKD, and one participant with borderline screening eGFR ≥60 ml/min per 1.73 m² and eGFR<60 ml/min per 1.73m² was analyzed as having CKD. These discordant participants were excluded in a sensitivity analysis.

Intervention

All participants were prescribed a fixed dose of oral vitamin D₂ (ergocalciferol) filled at local clinical pharmacies at 50,000 IU twice weekly for 5 weeks. This treatment period was used to evaluate the short-term effects of vitamin D₂ supplementation. Subjects were asked to take no other vitamin D–containing supplements or products, including multivitamins, during the study period.

Vitamin D Metabolites

Vitamin D metabolites were measured from plasma samples collected before the first dose of vitamin D₂ (baseline) and 3–5 days after completion of the last dose of vitamin D₂. For two participants with unavailable plasma, serum was used instead. Samples were stored at −80°C and thawed before analysis in 2015 at the University of Washington Nutrition Obesity Research Center. Immunoaffinity enrichment-liquid chromatography-tandem mass spectrometry on a Xevo TQ-MS triple quadrupole mass spectrometer (Waters, Milford, CT) was used to measure 25(OH)D₂; 25(OH)D₃ [summed to calculate total 25(OH)D]; 1,25(OH)₂D₂; 1,25(OH)₂D₃ [summed to calculate total 1,25(OH)₂D]; and 24,25(OH)₂D₃ (14). Interassay coefficients of variation (CVs) were 8.9%–9.9% for 1,25(OH)D₂; 8.0%–10.4% for 1,25(OH)D₃; 6.0%–6.7% for 25(OH)D₂; 3.5%–4.4% for 25(OH)D₃; and 5.2%–7.4% for 24,25(OH)₂D₃. Calibration of serum 25(OH)D concentrations was verified using SRM 972 from the National Institutes of Standards and Technology. Vitamin D metabolic ratios were calculated as functional estimates of CYP24A1 and CYP27B1 activity: 1,25(OH)₂D₃-to-25(OH)D₃ ratio estimated CYP27B1-mediated production of 1,25(OH)₂D₃, and 24,25(OH)₂D₃-to-25(OH)D₃ ratio estimated CY24A1-mediated clearance of 1,25(OH)₂D₃. Estimation of bioavailable 25(OH)D is described in Supplemental Material.

Other Mineral Metabolism Biomarkers

Intact PTH was measured using an automated two-site immunoassay (Beckman-Coulter, Inc., Brea, CA; interassay CV=6.1%–3.4%). FGF-23 was measured using the Kainos immunoassay, which detects the full-length biologically intact FGF-23 molecule via distal and midmolecule epitopes (interassay CV=6.7%–12.4%). Vitamin D binding protein was quantified by proteolysis and liquid chromatography-tandem mass spectrometry on a Xevo TQ MS triple quadrupole mass spectrometer (Waters) (15).

Statistical Analyses

Descriptive statistics were used to compare baseline participant characteristics according to CKD status.

Spaghetti plots were constructed to reflect trajectories of mineral metabolism biomarkers over time and according to study group. We used generalized estimating equations, accounting for within-participant clustering across study time, to determine whether change in circulating mineral metabolism markers differed with respect to CKD status. A Wald test of the multiplicative CKD by time interaction term was used to evaluate the statistical significance of the difference in change in biomarker by CKD status.

We selected covariates a priori on the basis of biologic plausibility that they could confound or mediate associations of kidney function with change in biomarkers. The first model included age, sex, race, recruitment site, and body mass index. A subsequent model was constructed to evaluate potential mediation effects of PTH and FGF-23 on the estimates. Missing values for covariates were multiply imputed (n=10) with chained equations, and the resulting analyses were combined using the rules of Rubin (16,17).

Institutional Review Board Approval

This study was approved by the institutional review boards of the MGH, the University of Miami, and St. Louis University in 2007. The study was initiated before the adoption of the expanded definition of clinical trial by the International Committee of Medical Journal Editors in 2008 and considered at the time to be a detailed physiologic study rather than a clinical trial. For this reason, it was not prospectively entered into a clinical trials registry.

Results

Characteristics of the Participants

Of 69 participants included in analyses, 25 had CKD mean (SD) eGFR of 40 (13) ml/min per 1.73 m² (Table 1). Participants with and without CKD had mean ages of 62 and 52 years old, respectively; 24% and 50% were women, respectively, and 48% and 41% were nonwhite, respectively. Mean (SD) baseline total 25(OH)D concentrations were 18.9 (1.8) and 20.4 (1.6) ng/ml for participants with and without CKD, respectively. Participants with CKD had more diabetes mellitus, hypertension, and prevalent cardiovascular disease and higher body mass index at baseline compared with participants without CKD.

Table 1.

Baseline characteristics of study participants

Characteristic	eGFR
	<60 ml/min per 1.73 m2, n=25	≥60 ml/min per 1.73 m2, n=44
Age, yrs	62±13	52±14
Sex
Men	19 (76)	22 (50)
Women	6 (24)	22 (50)
Race/ethnicity
White	13 (52)	26 (59)
Black	9 (36)	9 (21)
Hispanic	3 (12)	7 (16)
Asian	0	2 (5)
Study site
Massachusetts General Hospital	15 (60)	31 (70.5)
University of Miami	3 (12)	11 (25)
St. Louis University Hospital	7 (28)	2 (5)
Prevalent cardiovascular disease, %	25	12
Hypertension, %	83	44
Diabetes mellitus, %	42	19
Multivitamin use, %	33	11
Calcium supplement use, %	0	7
BMI, kg/m2	33.1±6.9	31.7±5.3
Creatinine, mg/dl	1.96±0.88	0.91±0.28
Cystatin C, mg/L	0.78±0.24	1.77±0.58
eGFR, ml/min per 1.73 m2	40±13	118±45
eGFR category, ml/min per 1.73 m2
≥60	0	44 (100)
45–59	9 (36)	0
30–44	11 (44)	0
15–29	4 (16)	0
<15	1 (4)	0

Cell contents are presented as n (%) or mean (±SDs). Prevalent cardiovascular disease refers to a history of coronary artery disease and/or myocardial infarction and/or coronary artery bypass grafting and/or congestive heart failure. Data were unavailable for 31 participants regarding diabetes status, 29 participants for multivitamin and calcium supplement use, and 32 participants regarding hypertension and prevalent cardiovascular disease status. Data were missing for BMI for 45 participants. BMI, body mass index.

Vitamin D Metabolites and Vitamin D Metabolic Ratios

Among participants without CKD, there were significant increases in plasma 25(OH)D₂ and total 25(OH)D with treatment (Figure 1, Table 2). Plasma 1,25(OH)₂D₂ also increased, whereas 1,25(OH)₂D₃ decreased, resulting in no significant change in total 1,25(OH)₂D. The 1,25(OH)₂D₃-to-25(OH)D₃ ratio decreased (3.0–1.70 pg/ng), whereas the 24,25(OH)₂D₃-to-25(OH)D₃ ratio increased (115.7–195.2 pg/ng) (Figure 2, Table 2).

Figure 1.

Changes in 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D metabolites with vitamin D₂ supplementation. Spaghetti plots show individual changes in metabolite concentrations from baseline to end of treatment (blue lines). The means are represented by the bold red lines. Mean changes with treatment and 95% confidence intervals are represented within each graph. (A) 25-Hydroxyvitamin D₂ [25(OH)D₂]. (B) 1,25-Dihydroxyvitamin D₂ [1,25(OH)₂D₂]. (C) 25-Hydroxyvitamin D₃ [25(OH)D₃]. (D) 1,25-Dihydroxyvitamin D₃ [1,25(OH)₂D₃]. (E) 25-Hydroxyvitamin D₂ and 25-hydroxyvitamin D₃ [total 25(OH)D]. (F) 1,25-Dihydroxyvitamin D₂ and 1,25-dihydroxyvitamin D₃ [total 1,25(OH)₂D].

Table 2.

Mineral metabolism biomarkers by CKD status before and after treatment

Biomarker	CKD (eGFR<60 ml/min per 1.73 m2)				Non-CKD (eGFR≥60 ml/min per 1.73 m2)				Difference in Change by CKD Status
	Baseline	Post-Treatment	Change (95% CI)	P Value	Baseline	Post-Treatment	Change (95% CI)	P Value	Difference in Change (95% CI)	P Value Interaction
25(OH)D total, ng/ml	18.9 (1.8)	41.1 (1.9)	22.3 (18.3 to 26.2)	<0.001	20.4 (1.6)	39.3 (1.6)	18.9 (15.6 to 22.3)	<0.001	3.4 (−1.8 to 8.5)	0.21
25(OH)D3, ng/ml	16.9 (1.1)	9.6 (1.2)	−7.3 (−9.0 to −5.7)	<0.001	18.1 (1.0)	9.2 (1.0)	−8.9 (−10.3 to −7.5)	<0.001	1.6 (−0.6 to 3.7)	0.15
25(OH)D2, ng/ml	2.0 (1.5)	31.7 (1.6)	29.7 (25.7 to 33.6)	<0.001	2.3 (1.3)	30.1 (1.3)	27.8 (24.4 to 31.2)	<0.001	1.9 (−3.3 to 7.1)	0.48
1,25(OH)2D total, pg/ml	30.8 (2.5)	34.2 (2.6)	3.4 (−1.3 to 8.1)	0.15	48.4 (2.2)	47.5 (2.2)	−0.9 (−4.8 to 3.1)	0.67	4.3 (−1.8 to 10.4)	0.17
1,25(OH)2D3, pg/ml	28.7 (2.2)	9.9 (2.3)	−18.8 (−23.8 to −13.7)	<0.001	45.6 (1.9)	14.6 (1.9)	−31.0 (−35.3 to −26.7)	<0.001	12.2 (5.6 to 18.9)	<0.001
1,25(OH)2D2, pg/ml	2.1 (1.6)	24.4 (1.6)	22.3 (18.2 to 26.4)	<0.001	2.8 (1.3)	32.9 (1.4)	30.1 (26.6 to 33.6)	<0.001	−7.8 (−13.2 to −2.4)	0.01
24,25(OH)2D3, ng/ml	1.3 (0.2)	1.0 (0.2)	−0.2 (−0.5 to −0.0)	0.05	2.2 (0.2)	1.7 (0.2)	−0.5 (−0.7 to −0.3)	<0.001	0.3 (−0.1 to 0.6)	0.10
PTH, pg/ml	103.1 (9.2)	94.6 (9.3)	−8.6 (−20.3 to 3.1)	0.15	60.4 (8.0)	50.6 (8.0)	−9.8 (−19.7 to 0.1)	0.05	1.2 (−14.1 to 16.5)	0.88
FGF-23, pg/ml	81.6 (6.1)	83.6 (6.2)	2.0 (−5.7 to 9.8)	0.61	47.0 (5.3)	50.5 (5.3)	3.4 (−3.1 to 10.0)	0.30	−1.4 (−11.6,8.7)	0.78
Calcium, mg/dl	8.9 (0.2)	8.9 (0.2)	0.0 (−0.3 to 0.4)	0.86	8.7 (0.2)	8.7 (0.2)	−0.0 (−0.4 to 0.3)	0.83	0.1 (−0.4 to 0.6)	0.78
Phosphorus, mg/dl	3.4 (0.2)	3.6 (0.2)	0.2 (−0.3 to 0.6)	0.45	3.2 (0.2)	3.5 (0.2)	0.3 (−0.1 to 0.7)	0.11	−0.1 (−0.7 to 0.4)	0.62
24,25(OH)2D3 to 25(OH)D3, pg/ng	72.0 (9.1)	110.3 (9.3)	38.3 (20.9 to 55.7)	<0.001	115.7 (7.8)	195.2 (7.9)	79.5 (64.7 to 94.2)	<0.001	−41.2 (−64.0 to −18.3)	<0.001
1,25(OH)2D3 to 25(OH)D3, pg/ng	1.8 (0.2)	1.1 (0.2)	−0.7 (−1.2 to −0.3)	0.001	3.0 (0.2)	1.7 (0.2)	−1.3 (−1.7 to −0.9)	<0.001	0.6 (0.0 to 1.2)	0.05
Vitamin D binding protein, μg/ml	264.0 (7.0)	253.8 (7.2)	−10.1 (−22.1 to 1.8)	0.10	255.4 (6.1)	260.8 (6.1)	5.5 (−4.6 to 15.6)	<0.001	−15.6 (−31.2 to 0.0)	0.05
Estimated bioavailable 25(OH)D	2.2 (0.3)	5.2 (0.3)	2.9 (2.3 to 3.6)	<0.001	2.3 (0.3)	4.4 (0.3)	2.1 (1.5 to 2.7)	<0.001	0.8 (−0.0 to 1.7)	0.06

Cell contents are mean (SD), mean change (95% CI), difference in mean changes (95% CI), P value within group, or P value for interaction comparing change among participants with CKD with change among participants without CKD. 95% CI, 95% confidence interval; 25(OH)D, 25-hydroxyvitamin D; 25(OH)D₃, 25-hydroxyvitamin D₃; 25(OH)D₂, 25-hydroxyvitamin D₂; 1,25(OH)₂D, 1,25-dihydroxyvitamin D; 1,25(OH)₂D₂, 1,25-dihydroxyvitamin D₂; 1,25(OH)₂D₃, 1,25-dihydroxyvitamin D₃; 24,25(OH)₂D₃, 24,25-dihydroxyvitamin D₃; PTH, parathyroid hormone; FGF-23, fibroblast growth factor-23.

Figure 2.

Changes in the 1,25-dihydroxyvitamin D₃ [1,25(OH)₂D₃]-to-25-hydroxyvitamin D₃ [25(OH)D₃] and 24,25-dihydroxyvitamin D₃ [24,25(OH)₂D₃]-to-25-hydroxyvitamin D₃ [25(OH)D₃] ratios, parathyroid hormone (PTH), and fibroblast growth factor-23 (FGF-23) with vitamin D₂ supplementation. Spaghetti plots show individual changes in vitamin D metabolic ratio concentrations from baseline to end of treatment (blue lines). The means are represented by the bold red lines. Mean changes with treatment and 95% confidence intervals are represented within each graph. (A) The 1,25(OH)₂D₃-to-25(OH)D₃ ratio. (B) PTH. (C) The 24,25(OH)₂D₃-to-25(OH)D₃ ratio. (D) FGF-23.

In the CKD group, baseline concentrations and changes in 25(OH)D₂ and total 25(OH)D (18.9–41.1 ng/ml) were similar to those in the control group (treatment group interaction, P>0.05). However, comparing the CKD group with the control group, baseline levels were lower and changes were smaller in magnitude for 1,25(OH)₂D₂, 1,25(OH)₂D₃-to-25(OH)D₃ ratio, and 24,25(OH)₂D₃-to-25(OH)D₃ ratio when evaluated on the absolute scale (each interaction P<0.05). Comparing participants with and without CKD, differences in changes in 1,25(OH)₂D₂, 1,25(OH)₂D₃-to-25(OH)D₃ ratio, and 24,25(OH)₂D₃-to-25(OH)D₃ ratio remained significant after adjusting for demographic variables (Table 3). In a sensitivity analysis excluding 15 participants who were recruited as subjects with CKD but later found to have study eGFR ≥60 ml/min per 1.73 m² and one participant with borderline screening eGFR ≥60 ml/min per 1.73 m² and baseline eGFR <60 ml/min per 1.73 m², differences comparing participants with CKD with the control group were of larger magnitude (Supplemental Tables 1 and 2). When changes in mineral metabolism markers were evaluated on the relative scale (as percentage change), there were no significant differences comparing participants with and without CKD (Supplemental Table 3).

Table 3.

Associations of CKD status with change in vitamin D metabolic ratios

Covariate Adjustments	Difference in Change in the 24,25(OH)2D3-to-25(OH)D3 Ratio, pg/ng		Difference in Change in the 1,25(OH)2D3-to-25(OH)D3 Ratio, pg/ng
	Difference (95% Confidence Interval)	P Value	Difference (95% Confidence Interval)	P Value
None (unadjusted)	−41.29 (−64.09 to −18.39)	<0.001	0.59 (0.01 to 1.18)	0.05
Age/sex/race, site, BMI	−30.40 (−53.9 to −6.88)	0.01	0.7 (−0.10 to 1.30)	0.02
Age/sex/race, site, BMI + FGF-23 + PTH	−22.40 (−49.0 to 4.30)	0.10	0.8 (0.13 to 1.47)	0.02

Cell contents represent the difference in change in vitamin D metabolic ratios comparing participants with CKD with those in non-CKD group [(change in vitamin D metabolic ratio from baseline to post-treatment in CKD group) − (change in vitamin D metabolic ratio from baseline to post-treatment in the non-CKD group)], with 95% confidence intervals and P values. Model 1: age, sex, race, site of enrollment, and BMI were assessed as potential confounding factors in the differences in change noted in the vitamin D metabolic ratios between the participants with CKD and the participants without CKD. Model 2 was adjusted for all covariates in model 1 in addition to baseline PTH and FGF-23 to assess if PTH and/or FGF-23 were potential mediators in the differences observed in the vitamin D metabolic ratios by CKD status. 24,25(OH)₂D₃-to-25(OH)D₃ ratio, the ratio of 24,25-dihydroxyvitamin D₃ to 25-hydroxyvitamin D₃; 1,25(OH)₂D₃-to-25(OH)D₃ ratio, the ratio of 1,25-dihydroxyvitamin D₃ to 25-hydroxyvitamin D₃; BMI, body mass index; FGF-23, fibroblast growth factor-23; PTH, parathyroid hormone.

Plasma PTH, FGF-23, and Vitamin D Binding Protein

As expected at baseline, the CKD group was observed to have a higher mean PTH concentration of 103.1 pg/ml compared with the control group: 60.4 pg/ml (Table 2). With treatment, small decreases in mean PTH concentration that did not meet statistical significance were noted in each treatment group, with no difference in change by group (P interaction =0.88).

Mean FGF-23 was also higher in the CKD group before treatment: 81.6 (6.1) compared with 47.0 (5.3) pg/ml. There was no significant change in FGF-23 concentration in either treatment group. At baseline, FGF-23 was inversely correlated with 1,25(OH)₂D, and PTH was inversely correlated with 24,25(OH)₂D₃. However, baseline PTH and baseline FGF-23 were not associated with change in any of the metabolites (Supplemental Figures 1–4). Adjustment for PTH and FGF-23 attenuated the association of CKD with 24,25(OH)₂D₃-to-25(OH)D₃ ratio (Table 3). Mean vitamin D binding protein concentration did not differ by CKD status, and there were no changes in vitamin D binding protein with treatment. Changes in bioavailable 25(OH)D paralleled changes in total 25(OH)D.

Discussion

In this detailed physiologic study, intensive vitamin D₂ supplementation effectively raised plasma concentrations of 25(OH)D₂ and total 25(OH)D among participants with and without CKD. Vitamin D₂ supplementation also raised plasma 1,25(OH)₂D₂ concentrations, with a reciprocal fall in 1,25(OH)₂D₃ concentration, resulting in no significant change in total 1,25(OH)₂D. There was a decrease in the 1,25(OH)₂D₃-to-25(OH)D₃ ratio, suggesting downregulation of CYP27B1, and an increase in the 24,25(OH)₂D₃-to-25(OH)D₃ ratio, suggested upregulation of CYP24A1. Comparing participants with CKD with those in the non-CKD group, baseline levels were lower and changes were smaller in magnitude for 1,25(OH)₂D₂, 1,25(OH)₂D₃-to-25(OH)D₃ ratio, and 24,25(OH)₂D₃-to-25(OH)D₃ ratio when evaluated on the absolute scale but not when evaluated on the relative scale. No significant change in FGF-23, PTH, or vitamin D binding protein was observed with vitamin D₂ supplementation.

Numerous other studies, together spanning a large range of kidney function from normal to CKD to ESRD, have shown that vitamin D₂ or D₃ supplementation increases circulating 25(OH)D concentration (18–20). The conversion of vitamin D₂ or D₃ to 25(OH)D₂ or 25(OH)D₃ is catalyzed by the hepatic enzyme CYP2R1, which is not expected to be altered in CKD. However, it has been hypothesized that increased CYP24A1 activity induced by FGF-23 may contribute to 25(OH)D deficiency and blunt the 25(OH)D response to supplements in CKD. A prior study found that catabolism of 25(OH)D may contribute to low circulating vitamin D levels in CKD, because elevations of FGF-23 in CKD were associated with increased 24,25(OH)₂D₃ after cholecalciferol administration (21). Our findings suggest that stages 3 and 4 CKD do not significantly modify the effect of vitamin D supplementation on circulating 25(OH)D concentration.

Our results are consistent with recent studies reporting little or no effect of vitamin D supplementation on circulating total 1,25(OH)₂D concentrations (22,23). We found that increased plasma 1,25(OH)₂D₂ (from the vitamin D₂ supplement) was completely compensated by a reciprocal decrease in 1,25(OH)₂D₃ (derived from other sources), resulting in no significant change in total 1,25(OH)₂D concentration. Two prior studies assessed the effects of supplementation on circulating vitamin D metabolites measured by mass spectrometry among people with normal kidney function and found similar results (22,23). Older studies reporting increased 1,25(OH)₂D concentrations with vitamin D supplementation used less specific assays for dihydroxyvitamin D metabolites and may have detected increased concentrations of 25(OH)D or 24,25(OH)₂D rather than 1,25(OH)₂D. Our data support the prevailing theory that total circulating 1,25(OH)₂D concentration is tightly regulated, even during vitamin D supplementation, and suggest that this regulation is maintained in moderate-severe CKD, albeit at a lower 1,25(OH)₂D concentration.

Examination of vitamin D metabolic ratios and circulating hormones provided new insight into underlying mechanisms regulating circulating 1,25(OH)₂D concentration. With supplementation, the 1,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio decreased, and the 24,25(OH)₂D₃-to-25(OH)D₃ ratio increased. These changes in vitamin D metabolic ratios could be influenced by reduced availability of substrate [25(OH)D₃] or competition by supplemented vitamin D₂. However, similar changes in vitamin D metabolic ratios were reported in a prior study of vitamin D₃ supplementation, in which 25(OH)D₃ concentrations increased substantially with treatment and 25(OH)D₂ concentrations were presumably very low (24). Together, these studies suggest that circulating 1,25(OH)₂D concentration is maintained during vitamin D supplementation through downregulation of CYP27B1 and upregulation of CYP24A1.

Circulating 1,25(OH)₂D and 24,25(OH)₂D are largely produced in the kidneys, and therefore, these data suggest a shift in renal vitamin D metabolism during supplementation. Interestingly, these changes cannot be explained by changes in circulating PTH or FGF-23, hormones that are known to regulate renal CYP27B1 and CYP24A1, because changes in PTH and FGF-23 were small or null during this study. This observation suggests that the kidney may contribute to vitamin D homeostasis through mechanisms that are both dependent on and independent of systemic hormones. For example, vitamin D metabolites within kidney tubular cells, particularly 1,25(OH)₂D, may directly regulate CYP27B1 and CYP24A1 (5,25).

It has long been known that 1,25(OH)₂D production is decreased in CKD (11–13). Our findings that baseline 1,25(OH)₂D concentrations and the 1,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio are lower in participants with versus without CKD are consistent with this. With supplementation, CKD was associated with blunted absolute changes but similar relative (percentage) changes in 1,25(OH)₂D and the 1,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio. This suggests that the same defects contributing to impaired 1,25(OH)₂D production in the unsupplemented state extend to vitamin D supplementation. Plasma FGF-23 concentration inversely correlated with plasma 1,25(OH)₂D concentration at baseline, and adjusting for PTH and FGF-23 concentrations attenuated associations of CKD with change in the 24,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio, suggesting that PTH and FGF-23 may mediate, in part, effects of CKD on vitamin D metabolism. However, it is hard to distinguish PTH and FGF-23 concentrations from CKD itself given large differences by CKD status, and estimates of association of CKD change in vitamin D metabolic ratios were not reduced to zero with adjustment for PTH and FGF-23.

Moreover, our data examining plasma 24,25(OH)₂D₃ and the 24,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio support the concept that vitamin D catabolism is also reduced in CKD, consistent with a “stagnant” state of vitamin D metabolism (5). Lower eGFR has been associated with lower plasma 24,25(OH)₂D₃ concentrations in published studies (26–28). Our findings of lower baseline 24,25(OH)₂D₃ concentration and lower 24,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio among participants with versus without CKD are consistent with these reports. Furthermore, among participants with CKD, 24,25(OH)₂D₃-to-25(OH)D₃ ratio increased substantially with supplementation to levels near those of the non-CKD group without supplementation. This is important, because low 24,25(OH)₂D₃ concentration has been associated with mortality and proposed as a biomarker of insufficient 1,25(OH)₂D₃ activity in CKD (5,26,29). Our data show that the 24,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio is modifiable with a simple intervention.

As for 1,25(OH)₂D and the 1,25(OH)₂D₃-to-25(OH)D₃ ratio, absolute changes in the 24,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio with vitamin D₂ supplementation were blunted among participants with versus without CKD, but relative (percentage) changes were similar. These results are consistent with those of a published study, in which vitamin D₃ supplementation increased plasma 24,25(OH)₂D₃ concentration and the 24,25(OH)₂D₃-to-25(OH)D₃ ratio, with significantly smaller changes among participants with versus without CKD (24). In comparison, our study used a vitamin D₂ intervention to more clearly discriminate vitamin D metabolites arising from study medication versus usual sources and evaluate changes in the 24,25(OH)₂D₃-to-25(OH)D₃ ratio without substantial increases in 25(OH)D₃. Together, the two studies offer a compelling argument that CYP24A1-mediated clearance of vitamin D supplements is reduced in CKD, perhaps due to decreased 25(OH)D delivery to kidney tubules, impaired 25(OH)D uptake by megalin and cubulin, or reduced CYP24A1 enzyme abundance or activity.

Recently, there has been increasing interest in bioavailable 25(OH)D [i.e., the sum of free 25(OH)D and 25(OH)D loosely bound to albumin, excluding 25(OH)D tightly bound to vitamin D binding protein] (30,31). In this context, it is important to note that we observed no effect of vitamin D₂ supplementation on the plasma concentration of vitamin D binding protein measured using a precise assay unbiased with regard to vitamin D binding protein isoform (15,32). Therefore, effects of vitamin D₂ supplementation on bioavailable 25(OH)D and 1,25(OH)₂D paralleled effects on total 25(OH)D and 1,25(OH)₂D.

Strengths of this study include the ability to compare changes in vitamin D metabolites and mineral metabolism biomarkers with treatment in persons with CKD with those in healthy individuals. In addition, the use of vitamin D₂ allowed rough differentiation of vitamin D metabolites arising from the study drug versus other sources and the evaluation of vitamin D metabolic ratios. The use of specific and sensitive mass spectrometry allowed for accurate and reliable measurements of the major vitamin D metabolites, eliminating the problem of variability or crossreactivity with immunoassay measurements (32). Our study population was derived from four nephrology clinic sites from three different parts of the country, increasing the generalizability of our findings to the prevalent CKD population compared with single-center studies.

Limitations of this study include the relatively small sample size, which prohibits us from excluding small differences in biomarkers for which null or insignificant results were observed, such as in FGF-23 and PTH; they have been shown to decrease with vitamin D interventions in prior studies (33). This study cannot determine the responsible site(s) of vitamin D metabolism. Circulating levels of vitamin D metabolites were measured and analyzed but may not reflect those in tissues. Although participants received the same dose of vitamin D₂, a specific formulation was not used. Participants with and without CKD were not matched, although differences in vitamin D metabolic ratios by CKD status persisted with adjustment, and the large observed differences are difficult to explain by other potential confounders. Misclassification of CKD status for participants with screening and baseline eGFRs near 60 ml/min per 1.73 m² may have biased associations of CKD with changes in vitamin D metabolites toward the null. The short duration of treatment time may have obscured the full long-term effects of vitamin D₂ supplementation. Reliability of vitamin D metabolite assays may have been overestimated using internal reference standards.

We conclude that vitamin D₂ supplementation decreases conversion of 25(OH)D₃ to 1,25(OH)₂D₃ and induces vitamin D₃ catabolism as evidenced by changes in plasma vitamin D₃ metabolites and vitamin D metabolic ratios. These effects do not seem to be fully attributable to changes in circulating PTH or FGF-23 levels. In CKD, production of 1,25(OH)₂D₃ and 24,25(OH)₂D₃ is reduced in the unsupplemented stated and with vitamin D₂ supplementation. Nonetheless, vitamin D₂ supplementation does increase the 24,25(OH)₂D₃-to-25(OH)D₃ metabolic ratio to levels comparable with those observed among unsupplemented participants without CKD, showing that the 24,25(OH)₂D₃-to-25(OH)D₃ ratio is a modifiable biomarker of impaired vitamin D metabolism in CKD.

Disclosures

T.I. is a consultant for Kyowa Hakko Kirin. K.J.M. is a consultant for Amgen, Diasorin, and OPKO. M.S.W. is a consultant or has received honoraria from Amag, Amgen, Ardelyx, Diasorin, Incyte, Keryx, Lilly, Pfizer, Sanofi, Ultragenyx, and ZS and has received grant support from Shire. Z.B., C.R.-C., A.N.H., B.K., and I.H.d.B. have nothing to disclose.