Chronic Kidney Disease and Diabetes Mellitus Predict Resistance to Vitamin D Replacement Therapy

Abstract

Background:

25-Hydroxyvitamin D [25(OH)D] is a marker of nutritional status; however, chronic kidney disease (CKD) results in alterations in vitamin D metabolism, including the loss of vitamin D-binding proteins and alterations in CYP27B1 and CYP24 enzymes that metabolize 25(OH)D. This study was designed to determine the predictors of responsiveness to correction of vitamin D deficiency with oral vitamin D₂ (ergocalciferol) in adults.

Methods:

A retrospective study of 183 veterans with 25(OH)D level <30 ng/mL, who were treated with 50,000 IU per week of vitamin D₂, was performed. Logistic regression models were developed to determine the factors predicting the response to treatment, defined as either the change in serum 25(OH)D level/1000 IU of vitamin D₂ or the number of vitamin D₂ doses (50,000 IU per dose) administered.

Results:

The mean age of the patients was 63 ± 12 years. About 87% were men and 51% diabetic, and 29% had an estimated glomerular filtration rate of <60 mL/min/1.73 m². The average number of vitamin D₂ doses was 10.91 ± 5.95; the average increase in 25(OH)D level was 18 ± 10.80 ng/mL. 25(OH) D levels remained <30 ng/mL in 61 patients after treatment. A low estimated glomerular filtration rate and the presence of diabetes mellitus were significant independent predictors for inadequate response to vita-min D₂ treatment in logistic regression models. Patients with CKD required greater amounts of vitamin D₂ to achieve similar increases in 25(OH)D levels, versus non-CKD patients.

Conclusions:

The presence of CKD and diabetes mellitus is associated with resistance to correction of 25(OH)D deficiency with vitamin D₂ therapy. The underlying mechanism needs to be evaluated in prospective studies.

Native vitamin D is a secosteroid that is available in the diet from either animal sources as cholecalciferol (vitamin D₃) or plants, as ergocalciferol (vitamin D₂).¹ With ultraviolet light exposure, 7-dehydrocholesterol is endogenously converted in the skin to vitamin D₃.² Vitamin D₃ is then hydroxylated in the liver by 25-hydroxylase to 25-hydroxyvitamin D [25(OH)D],¹ that subsequently is converted to 1,25-dihydroxyvitamin D₂ [1,25(OH)₂D] by 1α-hydroxylase in the kidneys.^3,4 This step is upregulated by parathyroid hormone (PTH) and downregulated by phosphate and fibroblast growth factor-23 (FGF23).⁵ 25(OH) D and 1,25(OH)₂D are metabolized by 24-hydroxylase to inactive metabolites, a step that is upregulated by FGF23 and downregulated by PTH.

Measurement of the serum level of 25(OH)D is considered to be the best biochemical index of vitamin D nutritional stores.⁶ Vitamin D deficiency is typically defined as a level <30 ng/mL.⁷ 25(OH)D deficiency is a common problem in the general population, with a prevalence of 36% to 57%.⁸ Low circulating concentrations of 25(OH)D is even more prevalent in patients with chronic kidney disease (CKD), with a prevalence of 50% to 86%.⁹ Factors contributing to reduced levels of 25(OH)D in CKD include associated chronic illnesses, inadequate nutrition and lack of adequate sun exposure, similar to the general population. In addition, patients with CKD have more complex abnormalities of vitamin D metabolism that may contribute to reductions in 25(OH)D level that do not represent a true vitamin D deficiency. In this regard, patients with CKD may have nephrotic-range proteinuria that is associated with the loss of vitamin D-binding proteins (VDBPs) with subsequent 25(OH)D deficiency and resistance to vitamin D replacement therapy.¹⁰ In addition, elevated levels of FGF23 act as a vitamin D counterregulatory hormone by decreasing 1,25(OH)₂D production through the downregulation of 1-α hydroxylase (encoded by CYP27B1) and the upregulation of 24-hydroxylase (encoded by CYP24).^11,12 FGF23-mediated increments in CYP24 may also lead to reduced 25(OH)D levels and resistance to vitamin D replacement therapy by increasing the catabolism of 25(OH)D. This study was designed to evaluate the predictors of resistance to vitamin D₂ (ergocalciferol) replacement therapy in a cohort of vitamin D-deficient patients who had a wide range of underlying kidney function, including patients with both CKD and normal kidney function.

METHODS

Study Population

This is a retrospective study of veterans diagnosed with vitamin D deficiency [serum 25(OH)D level <30 ng/mL], between April 2009 and July 2010, who received vitamin D supplementation with vitamin D₂ (ergocalciferol). The study was approved by the Veterans Affairs Medical Center Institutional Review Board (Memphis, TN). The inclusion criteria were as follows: (1) men and women over the age of 18 years, (2) serum 25(OH)D levels <30 ng/mL and (3) patients who received vitamin D₂ replacement therapy. The exclusion criteria were as follows: (1) an estimated glomerular filtration rate (eGFR) of <15 mL/min/1.73 m² and chronic dialysis; (2) cirrhosis, sarcoidosis, lymphoma, malabsorption syndrome, solid organ transplant or hyperparathyroidism; and (3) the chronic use of medications known to alter vitamin D metabolism, including rifampin, corticosteroids, antiepileptics, phosphate binders, active forms of vitamin D and calcimimetics. Initially, 598 patients with vitamin D deficiency were identified. Of these, 183 patients were included in the final analysis (Figure 1). The following information was obtained from the medical records: demographic data (age, sex and race), clinical characteristics (body weight, height, presence or absence of CKD, hypertension and diabetes mellitus [DM]), and the number of vitamin D₂ doses. Treatment with vitamin D₂ was confirmed by reviewing the pharmacy records for prescribing and releasing the medication to the patients. All patients received weekly doses of 50,000 IU of vitamin D₂. The patient’s primary care provider determined the total number of vitamin D₂ doses. Baseline laboratory data included the following: concentrations of serum creatinine, albumin, calcium, and 25 (OH)D, and urinary protein excretion. Serum 25(OH)D level was measured at baseline and at the end of the vitamin D₂ treatment. Urinary protein excretion was measured as a ratio of spot urine protein-creatinine (UPC) ratio or as spot urine albumin-creatinine ratio (ACR). UPC ratios were converted to urine ACR by the following equation: ACR = UPC^(1.054) ×0.596.¹³ All data used in the study were collected as part of usual patient care.

FIGURE 1.

Flow diagram for derivation of study cohort. 25(OH)D, 25-hydroxyvitaminD; eGFR, estimated glomerular filtration rate.

Procedures, Assays and Calculations

Response to vitamin D₂ treatment was assessed as (1) the ratio of change in serum 25(OH)D level and the total amount of vitamin D₂ received [final serum 25(OH)D concentration − baseline 25(OH)D concentration/1000 IU ergocalciferol], and (2) the total number of vitamin D₂ doses (50,000 IU per dose) received by each patient. Serum 25 (OH)D levels were measured by immunochemiluminometric assay (ICMA) using the DiaSorin (Stillwater, MN) Liaison instrument in all samples, serum albumin levels by bromocresol green assay and serum creatinine levels by isotope dilution mass spectrometry traceable Jaffe method from Roche (Stillwater, MN). The eGFR was calculated according to the 4-variable Modified Diet in Renal Disease formula.¹⁴ CKD was defined as an eGFR <60 mL/min/1.73m². Body mass index (BMI) was calculated by person’s weight in kilograms divided by his or her height in square meters.

Statistical Analysis

Continuous variables were presented as means and standard deviations, and categorical variables as percentages, unless otherwise specified. In this cohort, the median number of vitamin D₂ doses was 10, which was used as the cutoff value between the high (>10) and low (≤10) number of vitamin D₂ doses. The median number of vitamin D₂ dose was used as the dependent variable in the first logistic regression model. The median value of the ratio of change in serum 25(OH)D level/1000 IU ergocalciferol dose was 0.03. This value was used as the cutoff value between the high (≥0.03) and low (<0.03) ratio, and was used as the dependent variable in the second logistic regression analysis. Impaired response (resistance) to vitamin D₂ treatment was defined as the requirement for higher number of vitamin D₂ dosages (>10 doses) or by a low ratio (<0.03) of change in 25(OH)D/1000 IU of vitamin D₂. The following predictor variables were used for both logistic regression analyses: age; sex; race; BMI; DM; season during which 25(OH)D levels were obtained; baseline concentrations of albumin, calcium and 25(OH)D; and eGFR. A season variable (summer, winter) was created based on the timing of blood collection for baseline 25(OH)D, as sun exposure can affect previtamin D synthesis in human skin. The estimated odds ratio along with the corresponding 95% confidence intervals and P values are reported for all regression covariates. Only variables with a P value <0.10 were included in the multivariable stepwise logistic regression analysis. The final multivariable models were formally assessed for the presence of multicollinearity among the explanatory variables by using the variance inflation factor.¹⁵ Pretreatment and posttreatment laboratory test results were compared using paired t test and Wilcoxon signed-rank test. All tests were 2-sided, and a P value <0.05 was considered significant, unless otherwise stated. Statistical analysis was conducted using SAS version 9.1 (SAS Institute Inc, Cary, NC).

RESULTS

Baseline Characteristics

Characteristics of the study cohort (n = 183) are depicted in Table 1. A total of 159 (87%) were men, 94 (48%) were African American, 93 (51%) were diabetic, 53 (29%) had an eGFR <60 mL/min/1.73 m², 39 (21%) with stage 3 CKD and 14 (8%) with stage 4 CKD. The severity of serum 25(OH)D deficiency varied, with severe reductions (< 5 ng/mL) in 2 (1%), moderate reductions (5–15 ng/mL) in 79 (43%) and mild reductions (16–30 ng/mL) in 102 (56%) patients.

TABLE 1.

Baseline clinical characteristics of the study cohort

	Frequency	Mean ± SD	Minimum	Maximum
Age (yr)	183	63 ± 12	31	98
BMI (kg/m2)	183	31 ± 7	19	69
Initial serum creatinine (mg/dL)	183	1.3 ± 0.6	0.6	3.7
Initial eGFR (mL/min/1.73 m2)	183	71.6 ± 29.5	20	140
Initial serum 25(OH)D (ng/mL)	183	16.8 ± 6.5	4.4	29.8
No. ergocalciferol doses	183	10.8 ± 5.9	2	36
Initial serum calcium (mg/dL)	169	9.3 ± 0.7	4.8	10.7
Initial serum albumin (gm/dL)	149	4.2 ± 0.4	2.4	5.1
Urinary albumin creatinine ratio (g/g)	133	0.2 ± 0.8	0	7.3

25(OH)D, 25-hydroxyvitamin D; BMI, body mass index; eGFR, estimated glomerular filtration rate; SD, standard deviation.

Effects of Ergocalciferol Treatment

The average number of ergocalciferol doses administered was 10.91 ± 5.95, and the average duration of therapy was 2.5 months. Changes in concentrations of serum 25(OH)D, calcium, PTH and phosphorous levels are shown in Figure 2 and Table 2. The average increase in serum 25(OH)D level after vitamin D₂ supplementation was 18.0 ± 10.8 ng/mL. In 61 patients (33%), serum 25(OH)D level remained below normal (,<30 ng/mL) after treatment. Baseline serum 25(OH)D concentrations correlated significantly with the female sex (r = 0.18, P = 0.01) and African American race (r = 0.25, P = 0.0007), but not with age, DM, baseline eGFR, winter season or the number of prescribed vitamin D₂ dosages. Baseline eGFR inversely correlated with the number of vitamin D₂ doses (r = 20.21, P = 0.003) and with patient’s age (r = 0.44, P = 0.0001), and positively correlated with the female sex (r = 0.17, P = 0.02) and serum albumin level (r = 0.16, P = 0.04). Despite similar baseline 25(OH)D concentrations, patients with CKD required significantly greater amounts of vitamin D₂ (640 ± 296 versus 506 ± 291 thousand units, P 5 0.0006) to achieve similar increases in serum 25(OH)D levels, as compared with the non-CKD patients.

FIGURE 2.

Relationship between serum 25(OH)D and serum creatinine level showing pre- (●) and post (Δ) treatment 25(OH) D level. 25(OH)D, 25-hydroxyvitaminD.

TABLE 2.

Changes after treatment with ergocalciferol

	Frequency	Pretreatment	Posttreatment	Pa
Serum 25(OH)D (ng/mL)	183	17 ± 7	35 ± 10	<0.0001
eGFR (mL/min/1.73 m2)	172	72 ± 12	70 ± 25	0.06
Serum calcium (mg/dL)	140	9.3 ± 0.7	9.4 ± 0.5	0.001
Serum PTH (pg/mL)	30	114 ± 86	109 ± 75	0.76
Serum phosphorus (mg/dL)	19	3.8 ± 0.8	3.9 ± 0.8	0.95
Serum albumin (gm/dL)	113	4.2 ± 0.4	4.1 ± 0.4	0.002

Data expressed as mean ± standard deviation.

^aVariables eGFR and serum phosphorous were tested by paired t test; Wilcoxon signed-rank test used for the rest of the variables. 25(OH)D, 25-hydroxyvitamin D; eGFR, estimated glomerular filtration rate; PTH, parathyroid hormone.

Resistance to Ergocalciferol Treatment

The change in serum 25(OH)D level after the ergocalciferol treatment correlated significantly with the eGFR (Figure 3). Univariate logistic regression analysis of the change in serum 25 (OH)D level/1000 IU of vitamin D₂ showed statistically significant association of female sex, baseline eGFR, baseline serum 25(OH)D level and the presence of DM (Table 3). These 4 variables and the interaction terms between DM and baseline eGFR and baseline 25(OH)D were used to construct a multivariable model for the response to vitamin D₂ supplementation. In the final multivariable model, lower baseline eGFR, higher baseline serum 25(OH)D level and the presence of DM without the interaction terms significantly predicted inadequate response to vitamin D₂ supplementation (Table 3).

FIGURE 3.

Relationship between actual change in serum 25 (OH)D level and estimated glomerular filtration rate. 25(OH)D, 25-hydroxyvitaminD.

TABLE 3.

Predictors of an inadequate response to vitamin D supplementation defined by low ratio of change in serum 25(OH)D level/1000 IU ergocalciferol in univariate and multivariable analyses

	Univariate			Multivariablea
Predictor variables	OR	95% CI	P	ORb	95% CI	P
Race (1 = white, 0 = black)	0.63	0.35–1.12	0.12
Age (yr)	1.00	0.98–1.02	0.98
Sex (1 = male, 0 = female)	0.42	0.17–1.05	0.06
BMI (kg/m2)	1.01	0.97–1.06	0.56
HTN (1 = presence, 0 = absence)	1.18	0.60–2.34	0.63
DM (1 = presence, 0 = absence)	1.82	1.01–3.27	0.05	2.18	1.09–4.33	0.027
Season (1 = winter, 0 = summer)	0.91	0.50–1.65	0.75
Baseline serum albumin (gm/dL)	1.44	0.69–3.03	0.33
Baseline serum calcium (mg/dL)	0.99	0.63–1.58	0.98
Baseline serum 25(OH)D (ng/mL)	1.13	1.08–1.19	<0.0001	1.14	1.08–1.21	<0.0001
Baseline eGFR (mL/min/1.73m2)	0.97	0.96–0.99	<0.0001	0.97	0.96–0.99	<0.0001

^aHosmer-Lemeshow goodness-of-fit statistics χ²(8) of 5.32 and P value 0.72 suggesting that the model fits our data well. Variance inflation factor: 1.04 for eGFR, 1.04 for DM and 1.04 for 25(OH)D.

^bVariable excluded (P > 0.10) during stepwise logistic regression modeling. 25(OH)D, 25-hydroxyvitamin D; BMI, body mass index; CI, confidence interval; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate, HTN, hypertension; OR, odds ratio.

The response to vitamin D₂ supplementation in relation to the number of ergocalciferol doses administered was evaluated separately. In univariate logistic regression analysis, the presence of DM, low baseline eGFR and white race significantly predicted >10 weekly doses of ergocalciferol (Table 4). Lower baseline eGFR and the presence of DM were found to be significant predictors of a higher number of vitamin D₂ doses in the final multivariable model (Table 4). To help visualize analysis results, conditional effect plot for outcome prediction of high vitamin D₂ dose was drawn on the basis of the fitted final logistic regression model (Figure 4). It plotted the estimated probability of having a high vitamin D₂ dose (>10) against a chosen continuous covariate, eGFR, with the values of the other discrete and continuous covariates held in constant. The probabilities of resistance increased with worsening eGFR. The plot also showed that for a given value of eGFR, probabilities of resistance to vitamin D₂ supplementation were higher in diabetic patients, as compared with nondiabetic patients.

TABLE 4.

Predictors of an inadequate response to vitamin D supplementation defined by >10 ergocalciferol doses in univariate and multivariable analyses

	Univariate			Multivariablea
Predictor variables	OR	95% CI	P	ORb	95% CI	P
Race (1 = white, 0 = black)	1.85	1.02–3.33	0.042
Age (yr)	1.00	0.98–1.02	0.81
Sex (1 = male, 0 = female)	0.95	0.40–2.26	0.91
HTN (1 = presence, 0 = absence)	1.23	0.62–2.45	0.55
DM (1 = presence, 0 = absence)	2.11	1.16–3.82	0.014	1.93	1.04–3.56	0.04
BMI (kg/m2)	0.99	0.96–1.04	0.97
Season (1 = winter, 0 = summer)	0.73	0.40–1.33	0.30
Serum albumin (gm/dL)	0.69	0.33–1.45	0.33
Serum calcium (mg/dL)	0.71	0.43–1.17	0.18
Baseline serum 25(OH)D (ng/mL)	1.03	0.96–1.08	0.19
Baseline eGFR (mL/min/1.73 m2)	0.98	0.97–0.99	0.0004	0.98	0.97–0.99	0.0009

^aHosmer-Lemeshow goodness-of-fit statistics χ²(8) of 2.98 and P value 0.94 suggesting that the model fits our data well. Variance inflation factor: 1.015 for DM and 1.015 for eGFR.

^bVariable excluded (P >0.10) during stepwise logistic regression modeling.

25(OH)D, 25-hydroxyvitamin D; BMI, body mass index; CI, confidence interval; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate, HTN, hypertension; OR, odds ratio.

FIGURE 4.

Conditional effect plot showing the relationship between the estimated probability of inadequate response to ergocalciferol treatment (ergocalciferol dose >10) and estimated glomerular filtration rate in diabetic and non-diabetic patients.

Urinary ACR was available in a subset of patients (n = 133). Thirteen patients had an ACR of >0.6 g/g and 3 patients had nephrotic-range proteinuria (ACR >.2.2 g/g). Urinary ACR failed to contribute to either of the final multivariable logistic regression models for the response to vitamin D₂ supplementation.

After vitamin D₂ supplementation, serum 25(OH)D level decreased in 9 patients. The eGFR was ≥60 mL/min/1.73 m² in 6 of the 9 patients. A sensitivity analysis performed after excluding these patients showed the same predictors of inadequate response to vitamin D therapy.

DISCUSSION

In the current study, we evaluated the effectiveness of weekly vitamin D₂ treatment to raise serum 25(OH)D levels in a cohort of vitamin D-deficient patients that included patients with both normal and impaired renal function (CKD stages 3 and 4). We found that DM and CKD were significant independent predictors of resistance to vitamin D replacement therapy. Response to vitamin D₂ treatment showed progressive worsening with decreasing eGFR, and the effect was more pronounced in diabetic than in nondiabetic patients. The association of resistance to vitamin D supplementation with CKD and DM was independent of age, sex, ethnicity, seasonal variation, BMI and hypertension.

Despite similar baseline 25(OH)D levels, patients with CKD, as compared with non-CKD patients, required significantly greater amounts of ergocalciferol (640 ± 296 versus 506 ± 291 thousand units, P = 0.0006) to achieve similar increments in serum 25(OH)D levels. In the current study, the change in serum 25(OH)D level correlated significantly with eGFR (r = 0.25, P =0.0004), and baseline eGFR was inversely correlated with the number of vitamin D₂ doses (r = 20.21, P = 0.003). Moreover, in this cohort, the presence of low baseline eGFR and DM was associated with higher probabilities of resistance to vitamin D therapy and for the requirement of high doses of vitamin D₂ to achieve normal vitamin D levels (Figure 4).

To our knowledge this is the first study to identify the contribution of CKD and DM in a general group of patients with vitamin D deficiency. Previous studies that investigated possible resistance to treatment of 25(OH)D deficiency with vitamin D₂, as evidenced by the difficulty in achieving target 25(OH)D levels, only included patients with CKD.^16–18 The response was not compared to a group with normal renal function. Al-Aly et al¹⁷ treated vitamin D-deficient CKD patients with a vita-min D₂ regimen according to the National Kidney Foundation Kidney Disease Outcomes Quality Initiative-recommended protocol (15 doses).¹⁹ After 6 months of follow-up, serum 25(OH) D level increased by an average of 16.6 ng/mL to 27.2 ng/mL. However, the increase in 25(OH)D level was <5 ng/mL in 45% of the patients. In another study by Zisman et al,¹⁶ vitamin D-deficient CKD stage 3 and 4 patients were treated according to the National Kidney Foundation Kidney Disease Outcomes Quality Initiative protocol. Serum 25(OH)D levels increased to a range of 31 to 35 ng/mL in 60% of patients. In contrast, 33% of patients in the current study had serum 25(OH)D levels <30 ng/mL after treatment. This was influenced by the presence of renal disease, in that 17% of the patients in the low eGFR group (eGFR <60 mL/min) and 8% patients in the normal eGFR group (eGFR >60 mL/min) had <5% increase in serum 25(OH)D level (P = 0.09). We observed a somewhat better response in patients with CKD in the current study compared with the prior study, which may be related to the higher number of vitamin D₂ doses, because 40% of the patients with CKD in the current study received >15 vitamin D₂ doses.

There is controversy as to whether vitamin D₂ is less potent and effective than vitamin D₃ in raising serum 25(OH) D levels in the general population and that vitamin D₂ may enhance degradation of 25-hydroxyvitamin D₃.^20–22 The mechanism for these differences is not clear but could be secondary to the lower affinity of vitamin D₂ to VDBPs_, to differences in metabolism between vitamin D₂ and vitaminD₃, and to a shorter shelf life of vitamin D₂.²⁰To our knowledge, there have been no direct comparison studies of vita-min D₂ (ergocalciferol) versus vitamin D₃ (cholecalciferol) therapy for correction of vitamin D deficiency in patients with CKD.

The mechanism underlying the resistance to vitamin D₂ in diabetics and CKD is not known. Inadequate intake, hypoalbuminemia and albuminuria have been implicated as possible reasons for low 25(OH)D levels in diabetic patients.^23,24 There is no evidence that gastrointestinal absorption of 25(OH)D is altered in patients with CKD or diabetes.²⁵ Vitamin D metabolites are transported in the blood bound to VDBPs (85%–88%) and albumin (12%–15%), with very little circulating in the free form.^3,5,10 Because the liver produces VDBPs and albumin, and these proteins can be lost in the setting of nephrotic syndrome, these conditions can result in low levels of the transport proteins.^26,27 This results in low total levels of vitamin D metabolites [25(OH)D and 1,25(OH)₂D] without necessarily changing the free circulating levels.^10,27

Albuminuria in diabetics has been associated with higher urinary VDBP excretion, as compared with the normal subjects,²⁶ suggesting that higher requirements for vitamin D replacement therapy in albuminuric vitamin D-deficient CKD and diabetic patients may be related to the loss of VDBPs. In the current study, however, proteinuria per se was not associated with a diminished response to vitamin D replacement therapy in the subset of 133 of the 182 subjects in whom quantitative measurements for proteinuria were available. Lee et al²⁸ reported a significant association between resistance to vitamin D supplementation and higher BMI. This study included a small number of patients (n 5 17) with an average BMI of 25 kg/m² and the duration of cholecalciferol supplementation was also short. Although DM is strongly associated with obesity, we did not find any significant association between decreased response to ergocalciferol therapy and high BMI in the current study.

Other explanations for the increased vitamin D₂ requirements may be related to complex derangements in vitamin D metabolism that occur in CKD, including PTH stimulation of CYP27B1,^11,29,30 and FGF23 inhibition of CYP27B1^31–38 and stimulation of CYP24 expression.^31–37,39 In an adenine-induced animal model of CKD, mice with CKD had elevated FGF23 and PTH levels and decreased 1,25(OH)₂D levels with a 5-fold increase in the messenger RNA expression of CYP24 and a 2-fold increase in the messenger RNA expression of CYP27B1, as compared with control animals.³⁸ The effects of FGF23 on CYP27B1/CYP24 in CKD, with subsequent enhanced degradation of 25 (OH)D by the CYP24 pathway, could potentially explain the decreased responsiveness to vitamin D replacement associated with low eGFR. In advanced CKD, extrarenal CYP24 may play a role in the degradation of 25(OH)D.²⁴ At present, we have no data on either CYP24 activity in the kidney or catabolism of 25 (OH)D in diabetic patients with CKD. CYP24 expression in peripheral blood monocytes, however, was reported to be similar among patients with type 1 DM and healthy individuals.⁴⁰ Further prospective studies are needed to evaluate whether increased vita-min D catabolism is associated with resistance to treatment with vitamin D therapy in CKD.

Limitations of the current study include the observational and cross-sectional nature of the study. Most patients included in this study were men and >60 years of age, which may limit generalizability. Measurements of quantitative proteinuria were available only in a subset of patients (133 of the 183 patients) and direct measurements of 1,25(OH)₂D, intact PTH and FGF23 were not available. Vitamin D₂ dosing was standardized (50,000 IU per week) in the study with pharmacy documentation that the medication was dispensed to the patient. Adherence to the medication, however, could not be further verified because of the retrospective study design. The total duration of ergocalciferol therapy was determined by the patients’ primary care providers and, therefore, was not standardized.

This retrospective cohort study provides evidence for resistance to vitamin D₂ replacement therapy in vitamin D-deficient patients with CKD and DM. This may reflect underlying altered metabolism of vitamin D associated with these conditions. Prospective studies of vitamin D metabolism in vitamin D-deficient patients treated with nutritional vitamin D supplements are needed to define the underlying altered mechanism.