Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Feb 1.
Published in final edited form as: Am J Kidney Dis. 2015 Oct 23;67(2):235–242. doi: 10.1053/j.ajkd.2015.08.026

Biomarkers of Vitamin D Status and Risk of ESRD

Casey M Rebholz 1, Morgan E Grams 1,2, Pamela L Lutsey 3, Andrew N Hoofnagle 4, Jeffrey R Misialek 3, Lesley A Inker 5, Andrew S Levey 5, Elizabeth Selvin 1, Chi-yuan Hsu 6, Paul L Kimmel 7, Ramachandran S Vasan 8, John H Eckfeldt 9, Josef Coresh 1,10, on behalf of the Chronic Kidney Disease Biomarkers Consortium
PMCID: PMC4724452  NIHMSID: NIHMS725529  PMID: 26475393

Abstract

Background

Disordered mineral metabolism is characteristic of decreased kidney function. However, the prospective associations between circulating levels of vitamin D binding protein, vitamin D, and end-stage renal disease (ESRD) have not been extensively evaluated in epidemiologic studies.

Study Design

Nested case-control study

Setting & Participants

Middle-aged, black and white men and women from four US communities.

Predictors

Baseline levels of vitamin D binding protein, 25-hydroxyvitamin D (25(OH)D), and 1,25-dihydroxyvitamin D (1,25(OH)2D) were measured in blood samples collected at study visit 4 (1996–1998) of the Atherosclerosis Risk in Communities (ARIC) study.

Outcome

ESRD cases (n=184) were identified through hospitalization diagnostic codes from 1996–2008 and were frequency matched to controls (n=251) on categories of estimated glomerular filtration rate, albuminuria, diabetes mellitus, sex, and race.

Measurements

Logistic regression was used to estimate the association between mineral metabolism biomarkers (vitamin D binding protein, 25(OH)D, 1,25(OH)2D) and incident ESRD, adjusting for age, sex, race, estimated glomerular filtration rate, albuminuria, diabetes mellitus, hypertension, education, specimen type, and serum levels of calcium, phosphate, and parathyroid hormone.

Results

Higher vitamin D binding protein levels were associated with elevated risk of incident ESRD (OR, 1.76; 95% CI, 1.22–2.54; p=0.003). Higher free and bioavailable 25(OH)D levels were associated with reduced risk of incident ESRD (ORs of 0.65 [95% CI, 0.46–0.92; p=0.02] and 0.63 [95% CI, 0.43–0.91; p=0.02] for free and bioavailable 25(OH)D, respectively). There was no association between ESRD and overall 25(OH)D (OR, 0.83; 95% CI, 0.58–1.19; p=0.3) or 1,25(OH)2D (OR, 0.73; 95% CI, 0.48–1.13; p=0.2).

Limitations

Lack of direct measurement of free and bioavailable vitamin D.

Conclusions

In the general population, blood levels of vitamin D binding protein were positively associated and blood levels of free and bioavailable 25(OH)D were inversely associated with new-onset ESRD during follow-up.

Index words: biological markers, chronic renal failure, end-stage renal disease (ESRD), risk factors, vitamin D-binding protein, vitamin D, mineral metabolism biomarker, vitamin D insufficiency

Disordered mineral metabolism is one of the earliest complications of chronic kidney disease (CKD).1–3 Decreased glomerular filtration rate (GFR) is associated with lower 1α-hydroxylase activity, which results in decreased activation of 25-hydroxyvitamin D (25(OH)D) to 1,25-dihydroxyvitamin D (1,25(OH)2D), leading to hypocalcemia and hyperparathyroidism.1,4–6 Adequate vitamin D levels are a therapeutic goal for kidney disease patients.7 However, clinical recommendation to use vitamin D supplementation in CKD patients for correcting vitamin D deficiency is opinion-based and needs additional empirical evidence.8 Detecting deficient vitamin D levels may be useful in estimating future risk of kidney disease progression.2,9,10

In clinical and research settings, levels of 25(OH)D are usually reported as an assessment of vitamin D status. The majority (approximately 85%–90%) of circulating 25(OH)D, like 1,25(OH)2D, is tightly bound to vitamin D binding protein and is thought to be biologically inactive.11 Bioavailable 25(OH)D consists of a smaller amount (10%–15%) that is loosely bound to albumin and <1% that is circulating in a free and unbound form.11–13 Binding affinities for 25(OH)D vary by isoforms of vitamin D binding protein; these isoforms differ by racial group, and they explain some of the variability in circulating levels of vitamin D binding protein and 25(OH)D.14–16 Compared to previous studies of 25(OH)D, current efforts can now more completely evaluate the association of 25(OH)D with adverse outcomes using blood levels and isoforms of vitamin D binding protein to estimate free and bioavailable 25(OH)D.

The objective of this study was to assess the relationship between vitamin D-related biomarkers with incident end-stage renal disease (ESRD) in a community-based population, the Atherosclerosis Risk in Communities (ARIC) study in collaboration with the Chronic Kidney Disease Biomarkers Consortium. We hypothesized that blood levels of vitamin D binding protein would be positively and independently associated with ESRD risk, after accounting for demographics, kidney measures, and known kidney disease risk factors. Furthermore, we hypothesized that blood levels of 25(OH)D, and, secondarily, alternative measures of vitamin D status including 1,25(OH)2D, free vitamin D, and bioavailable vitamin D, would be inversely and independently associated with ESRD risk. We explored the distribution of 3-epi-25(OH)D3, a form of vitamin D that is not well characterized, according to incident ESRD case status.17

METHODS

Study Design

The ARIC study is a prospective cohort study of 15,792 predominantly black and white men and women, 45–64 years of age at enrollment, from four US communities: Forsyth County, North Carolina; Jackson, Mississippi; suburbs of Minneapolis, Minnesota; and Washington County, Maryland. The ARIC study participants were recruited and enrolled in 1987–1989, and four follow-up study visits were conducted: 1990–1992, 1993–1995, 1996–1998, and 2011–2013. The ARIC Study is described in detail elsewhere.18 In the present nested case-control study, ARIC study visit 4 (1996–1998) was the baseline visit. The main reason for using study visit 4 as baseline for the present analysis was that urinary albumin-creatinine ratio (UACR) was measured at this time point, which is an important indicator of kidney damage.

Study Participants

A total of 11,656 ARIC study participants (73.8% of the original ARIC cohort) completed the baseline (visit 4) examination. Study participants were excluded from the present nested case-control study if they were missing information on the factors making up the frequency matching categories (estimated GFR [eGFR], UACR, diabetes mellitus, sex, race) or developed ESRD prior to baseline. Incident ESRD cases (n=184) were defined using diagnostic codes for hospitalizations and deaths identified through active surveillance from baseline (1996–1998) through December 31, 2008. The ESRD case status was defined by: 1) International Classification of Disease (ICD) codes for hospitalizations related to kidney transplantation, dialysis, or procedural code indicating dialysis, excluding hospitalizations with concomitant ICD codes for traumatic anuria (958.5) or acute kidney injury (586, 788.9); or 2) death certificates with kidney failure-related ICD codes (584–584.9, 586, N17.0) as an underlying cause of death and history of CKD. This outcome definition is described in detail elsewhere.19 Frequency matching was used to identify controls (n=251) based on eGFR category (<45, 45–59, 60–74, 75–89, 90–104, ≥105 mL/min/1.73 m2), UACR category (<30, 30–300, >300 mg/g), diabetes mellitus, sex, and race. Controls were selected to match the frequency of cases within each stratum (approximately 1–2 controls per case within each stratum). These strong risk factors for ESRD (diabetes mellitus, sex, race) and indicators of kidney function (eGFR) and kidney damage (UACR) were selected as matching factors in order to evaluate the ability of novel biomarkers to predict ESRD risk beyond established factors. The study protocol was approved by the institutional review boards of all participating institutions, and written documentation of informed consent was obtained from all study participants.

Measurement of Covariates

At visit 4 (baseline for the present study), a questionnaire was administered, blood pressure and anthropometrics were measured, and blood specimens and spot urine specimens were collected. The questionnaire was administered by trained staff to assess demographic characteristics (age, sex, race), socioeconomic status (education level), health behaviors, medical history, and medication use. Participants brought current medications to the study visit, and medications were transcribed and coded. Blood pressure measurements were taken by a certified technician using a random-zero sphygmomanometer after the participant was seated and resting for five minutes. Body mass index was calculated as weight (in kilograms) divided the the square of height (in meters) from measurements taken while participants wore light clothing without shoes. Glucose was measured by the modified hexokinase/glucose-6-phosphate dehydrogenase method. Creatinine was measured in the plasma and urine by the modified kinetic Jaffe method, and values were calibrated to the National Institute of Standards and Technology standard. Albumin was measured in urine specimens by a nephelometric method on a Dade Behring BN100 and a Beckman Image Nephelometer. Serum calcium (coefficient of variation [CV], 1.3%) and phosphorus (CV, 1.9%–2.1%) were measured by colorimetric methods and serum intact parathyroid hormone (CV, 1.4%–5.8%) was measured by a sandwich immunoassay and quantified using the Roche Cobas 6000 Chemistry Analyzer (Roche Diagnostics Corporation, Indianapolis, IN). Serum albumin was measured from specimens collected at baseline (study visit 4, 1996–1998) with bromocresol purple dye-binding assay on a Roche Cobas 6000 Chemistry Analyzer using Roche reagents (Roche Diagnostics Corporation, Indianapolis, IN) (CV, 2.2%–2.6%).

Diabetes mellitus was defined as fasting glucose >126 mg/dL, non-fasting glucose >200 mg/dL, current medication use for diabetes mellitus, or self-report of physician-diagnosed diabetes mellitus. Hypertension status was defined as the mean of two measurements of systolic blood pressure ≥140 mm Hg, the mean of two measurements of diastolic blood pressure ≥90 mm Hg, or self-reported anti-hypertensive medication use in the past two weeks. The eGFR was calculated with serum creatinine using the CKD-EPI (CKD Epidemiology Collaboration) 2009 equation.20

Measurement of Vitamin D Binding Protein and Vitamin D

Blood specimens were collected at baseline (ARIC study visit 4, 1996–1998) and stored at −70°C until laboratory analysis. Vitamin D-related biomarkers (concentrations of 25(OH)D2, 25(OH)D3, 1,25(OH)2D2, 1,25(OH)2D3, 3-epi-25(OH)D3, and vitamin D binding protein; isoform of vitamin D binding protein) were quantified by liquid chromatography-tandem mass spectrometry with high-sensitivity spectrometers at the Advanced Research and Diagnostic Laboratory at the University of Minnesota in 2012–2015 (AB Sciex 5500) and the Department of Laboratory Medicine at the University of Washington in 2015 (Waters Xevo TQ MS).21 Immuno-affinity enrichment was used to measure levels of 1,25(OH)2D. After trypsin digestion of blood specimens, vitamin D binding protein was measured using isotopically-labeled peptides as internal standards.22 Laboratory technicians were masked to case status and participant characteristics. Laboratory measurement of mineral metabolism biomarkers were conducted in serum specimens, except for 18% which were conducted in citrated plasma specimens. Specimen type (plasma or serum) was unrelated to levels of mineral metabolism biomarkers except for 25(OH)D, which was lower in plasma specimens. We adjusted for specimen type in regression models.

Genotyping for vitamin D binding protein single-nucleotide polymorphisms (SNPs; reference SNP identification number [rs]4588 and rs7041) was obtained from the ITMAT-Broad-CARe Chip at the Broad Institute of Massachusetts Institute of Technology and Harvard University.23

Calculation of Overall, Free, and Bioavailable Vitamin D

Overall 25(OH)D and 1,25(OH)2D levels were calculated as the sum of the D2 and D3 isoforms. As a secondary analysis, alternative measures of vitamin D status (free and bioavailable 25(OH)D) were estimated using previously reported equations and affinity constants for albumin and vitamin D binding protein (isoform-specific; Item S1, available as online supplementary material).12–16,24,25

Statistical Analysis

Cases and controls were described and compared with respect to the frequency matching factors (eGFR, UACR, diabetes mellitus, sex, race), baseline characteristics, and concentrations of mineral metabolism biomarkers using descriptive statistics and χ2 tests for categorical variables and t tests for means. We tested for difference in log-transformed means of mineral metabolism biomarkers using linear regression and reported geometric means after race-adjustment to the overall study population. Differences in vitamin D binding protein concentrations by vitamin D binding protein SNPs (rs4588, rs7041) and vitamin D binding protein isoforms were tested using analysis of variance. In exploratory analyses, we assessed the concentrations of vitamin D markers according to isoforms of vitamin D binding protein. The cross-sectional relationship between the vitamin D-related biomarkers and eGFR was described using Spearman’s rank correlation coefficients.

Multivariable logistic regression was used to estimate the association between mineral metabolism biomarkers and risk of ESRD, accounting for matching factors. The continuous form of eGFR and UACR were used in regression models to account for residual confounding after matching on categories of eGFR and UACR. Regression models were additionally adjusted for age and hypertension status due to their known, strong associations with ESRD, for education level as an indicator of socioeconomic status, and for other markers of mineral metabolism (serum levels of calcium, phosphate, and parathyroid hormone). We also adjusted for specimen type (plasma or serum). Odds ratios (ORs) with 95% confidence intervals (CIs) were estimated per one interquartile range (IQR) higher level of biomarkers and, for overall 25(OH)D concentrations, according to clinical categories (assessing two cut-points: 20 and 30 ng/mL). Due to the non-normal distribution, log-transformed 25(OH)D was used in regression models. Stata version 13.1 statistical software was used for analysis (StataCorp LP, College Station, TX, USA).

RESULTS

Baseline Characteristics

In the overall study population, 57% of participants had diabetes mellitus, 48% were women, and 44% were black. Cases and controls were similar with respect to demographic characteristics, level of education, and kidney disease risk factors, with the exception of eGFR (16.3% of cases and 2.8% of controls had eGFR <30 mL/min/1.73 m2; p<0.001) and UACR (36.4% of cases and 17.1% of controls had UACR >300 mg/g; p<0.001) (Table 1).

Table 1.

Baseline demographic, clinical characteristics, and biochemical measurements of study participants by incident ESRD case status

Incident ESRD Case Status
P-valueb
Case (n=184) Control (n=251)
Female sexc 87 (47.3) 47.8% (120) 0.9
Black racec 45.7% (84) 41.8% (105) 0.5
Diabetes mellitusc 60.3% (111) 54.0% (136) 0.2
eGFRc
 ≥60 mL/min/1.73 m2 46.7% (86) 63.1% (159) <0.001
 30–59 mL/min/1.73 m2 37.0% (68) 34.1% (86)
 <30 mL/min/1.73 m2 16.3% (30) 2.8% (7)
UACRc
 <30 mg/g 37.0% (68) 52.8% (133) <0.001
 30–300 mg/g 26.6% (49) 30.2% (76)
 >300 mg/g 36.4% (67) 17.1% (43)
Age, y 64.5 (5.6) 64.7 (5.8) 0.7
Hypertension status 75.3% (137) 68.6% (170) 0.1
Body mass index, kg/m2 30.5 ±6.2 30.2 ±6.0 0.7
Education level
 <High school 37.7% (69) 28.9% (72) 0.1
 High school 37.2% (68) 39.0% (97)
 >High school 25.1% (46) 32.1% (80)
Serum albumin, g/dL 3.6 ±0.4 3.7 ±0.3 0.005
Serum calcium, mg/dL 9.2 ±0.5 9.2 ±0.4 0.5
Serum phosphorus, mg/dL 3.5 ±0.8 3.4 ±0.5 0.08
Serum PTH, pg/mL 77.3 ±106.4 44.2 ±27.0 <0.001
Open in a new tab

eGFR, estimated glomerular filtration rate; ESRD, end-stage renal disease; PTH, parathyroid hormone; UACR, urine albumin-creatinine ratio

Note: Values for categorical variables are given as number (percentage); for continuous variables, as mean ± standard deviation.

b

P-value from χ2 test for categorical variables and t test for continuous variables

c

Sex, race, diabetes mellitus, eGFR category, and UACR category were matching factors and therefore would not be expected to differ by case status. Due to the differences in eGFR and UACR categories by case status, the continuous form of eGFR and UACR were used in multivariable regression models to account for potential residual confounding.

Levels of Mineral Metabolism Biomarkers

Blood levels of vitamin D binding protein were higher among cases than controls and 25(OH)D and 1,25(OH)2D were lower among cases than controls after race adjustment (Table 2). Overall, levels of the mineral metabolism biomarkers were correlated with each other and baseline eGFR (p ≤0.01 for all; Table 3). In contrast, 1,25(OH)2D was not correlated with overall, free, and bioavailable 25(OH)D. Free and bioavailable 25(OH)D were not correlated with eGFR.

Table 2.

Vitamin D-related biomarkers by incident ESRD case status

Marker Incident ESRD Case Status
P-valueb
Cases (n=184) Controls (n=251)
Vitamin D binding protein, μg/mL 260 (233, 288) 250 (225, 277) 0.02
Detectable 3-epi-25(OH)D3 30 (16.3) 53 (21.1) 0.2
25(OH)D, ng/mL 18.6 (13.5, 25.7) 20.3 (15.0, 27.3) 0.04
Deficient 25(OH)Dc 155 (84.2) 204 (81.0) 0.4
Free 25(OH)D, pg/mL 6.79 (4.91, 9.44) 8.08 (5.66, 11.21) 0.003
Bioavailable 25(OH)D, ng/mL 2.21 (1.63, 3.15) 2.69 (1.83, 3.79) 0.001
1,25(OH)2D, pg/mL 34.0 (26.0, 45.6) 39.0 (30.4, 50.4) 0.001
Open in a new tab

Note: Values for categorical variables are given as number (percentage); for continuous variables, as geometric mean (25th, 75th percentile). Total cases and total controls were race-adjusted to the overall population. Conversion factors for units: 25(OH)D, free 25(OH)D, and bioavailable 25(OH)D in ng/mL to nmol/L, ×2.496; 1,25(OH)2D in pg/mL to pmol/L, ×2.6.

ESRD, end-stage renal disease; 25(OH)D, 25-hydroxyvitamin D; 1,25(OH)2D, 1,25-dihydroxyvitamin D

b

P-value from χ2 test or linear regression testing the difference in proportions or means

c

Deficient 25(OH)D: ≤30 ng/mL

Table 3.

Cross-sectional relationship between vitamin D-related biomarkers and kidney function

VDBP 25(OH)D Free 25(OH)D Bioavailable 25(OH)D 1,25(OH)2D eGFR
Vitamin D binding protein 1
25(OH)D r=0.14; p=0.01 1
Free 25(OH)D r= −0.18; p=0.002 r=0.58; p<0.001 1
Bioavailable 25(OH)D r= −0.18; p=0.002 r=0.60; p<0.001 r=0.98; p<0.001 1
1,25(OH)2D r=0.20; p<0.001 r= −0.08; p=0.1 r=0.01; p=0.9 r=0.03; p=0.6 1
eGFR r= −0.16; p=0.002 r= −0.18; p<0.001 r=0.001; p=0.9 r=0.03; p=0.6 r=0.44; p<0.001 1
Open in a new tab

eGFR, estimated glomerular filtration rate; 25(OH)D, 25-hydroxyvitamin D; 1,25(OH)2D, 1,25-dihydroxyvitamin D

For the SNP rs4588, vitamin D binding protein levels were highest for the CC genotype and lowest for the AA genotype (p<0.001; Figure S1A). For rs7041, vitamin D binding protein levels were highest for the GG genotype and lowest for the TT genotype (p<0.001; Figure S1B). Blood levels of vitamin D binding protein varied by isoform of vitamin D binding protein (p<0.001; Figure S1C). In addition, blood levels of free and bioavailable 25(OH)D and 1,25(OH)2D differed according to vitamin D binding protein isoforms (p <0.009 for all; Table S1).

Association Between Mineral Metabolism Biomarkers and ESRD

After adjusting for eGFR, UACR, age, sex, race, diabetes mellitus, hypertension, specimen type, education, and serum levels of calcium, phosphate, and parathyroid hormone, one IQR higher in vitamin D binding protein was associated with higher odds of incident ESRD (OR, 1.76; 95% CI, 1.22–2.54; p = 0.003; Table 4). In the fully adjusted model, higher levels of 25(OH)D were not associated with ESRD when expressed continuously (OR per 1 IQR higher, 0.83; 95% CI, 0.58–1.19; p = 0.3) or according to clinical categories (ORs of 1.11 [95% CI, 0.59–2.07; p=0.7] and 0.95 [95% CI, 0.57–1.56; p=0.8) for ≤30 versus >30 ng/mL and ≤20 versus >20 ng/mL, respectively).

Table 4.

Adjusted ORs for incident ESRD per 1 IQR higher concentration of mineral metabolism biomarkers

Biomarker and Model OR (95% CI) P-value
Vitamin D binding protein
 Model 1 1.48 (1.08, 2.02) 0.02
 Model 2 1.41 (1.02, 1.95) 0.04
 Model 3 1.76 (1.22, 2.54) 0.003
25(OH)D
 Model 1 0.70 (0.51, 0.97) 0.03
 Model 2 0.73 (0.53, 1.00) 0.05
 Model 3 0.83 (0.58, 1.19) 0.3
Free 25(OH)D
 Model 1 0.64 (0.47, 0.89) 0.007
 Model 2 0.64 (0.46, 0.89) 0.009
 Model 3 0.65 (0.46, 0.92) 0.02
Bioavailable 25(OH)D
 Model 1 0.63 (0.44, 0.88) 0.007
 Model 2 0.63 (0.44, 0.90) 0.01
 Model 3 0.63 (0.43, 0.91) 0.02
1,25(OH)2D
 Model 1 0.87 (0.61, 1.26) 0.5
 Model 2 0.82 (0.56, 1.20) 0.3
 Model 3 0.73 (0.48, 1.13) 0.2
Open in a new tab

CI, confidence interval; OR, odds ratio

Note: Interquartile ranges of vitamin D binding protein, log-transformed 25(OH)D,, free 25(OH)D, and bioavailable 25(OH)D are 54.24, 0.25, 4.91, and 1.74.respectively. Model 1: sex, race, diabetes mellitus, estimated glomerular filtration rate, log(urine albumin-creatinine ratio), specimen type; Model 2: variables in model 1 plus age, hypertension status, education; Model 3: variables in model 2 plus calcium, phosphorus, parathyroid hormone.

Higher levels of estimated free and bioavailable 25(OH)D were significantly associated with lower odds of ESRD (ORs of 0.65 [95% CI, 0.46–0.92; p=0.02] and 0.63 [95% CI, 0.43–0.91; p=0.02] for free and bioavailable 25(OH)D, respectively). Levels of 1,25(OH)2D were not associated with ESRD risk.

When vitamin D binding protein, 25(OH)D, and 1,25(OH)2D were included in the same model, the results were similar to the results for the biomarkers modelled separately. Vitamin D binding protein was more strongly associated with ESRD than the other biomarkers (for the fully adjusted model, ORs were 2.08 [95% CI, 1.40–3.08; p<0.001], 0.71 [95% CI, 0.44–1.12; p=0.1], and 0.64 [95% CI, 0.39–1.04; p=0.1] for vitamin D binding protein, 25(OH)D, and 1,25(OH)2D, respectively).

DISCUSSION

In this community-based study population, blood levels of vitamin D binding protein were positively associated with incident ESRD during follow-up even after complete multivariable adjustment including other mineral metabolism biomarkers. In addition, higher levels of estimated free and bioavailable 25(OH)D, which were calculated by incorporating concentrations of binding proteins for 25(OH)D (vitamin D binding protein and albumin) and vitamin D binding protein isoforms as a proxy for binding affinity, were associated with lower ESRD risk.

We did not find a significant association between 25(OH)D levels and ESRD risk in the fully adjusted model. The existing literature on this topic is inconsistent. For example, vitamin D insufficiency (<30 ng/mL) was not independently associated with ESRD in the Chronic Renal Insufficiency Cohort (CRIC) Study.2 In contrast, among 13,328 NHANES (National Health and Nutrition Examination Survey) participants, those with lower 25(OH)D levels (<15 ng/mL) had higher ESRD risk (adjusted incidence rate ratio, 2.64; 95% CI, 1.00–7.05).10 In a small CKD clinic study of 168 patients, 25(OH)D levels ≥15 ng/mL were significantly associated with lower risk of initiating dialysis.9 Inconsistencies in the vitamin D literature may be due to differences in study populations, ESRD ascertainment, or assays or due to the failure to measure the physiologically-relevant component of vitamin D.

Vitamin D binding protein binds tightly to the majority of 25(OH)D in the circulation and may be important to measure to accurately assess 25(OH)D status.14,16 When levels of vitamin D binding protein are elevated, more 25(OH)D is bound to vitamin D binding protein, and there is less 25(OH)D in the circulation able to interact with other receptors.26,27 Free and bioavailable 25(OH)D represent the physiologically-active portion of vitamin D levels, accounting for blood levels of vitamin D binding protein and albumin and binding affinities of albumin and vitamin D binding protein to 25(OH)D using isoform-specific information.13–16,24 No studies have previously related free and bioavailable 25(OH)D to kidney disease risk. Our findings extend recent studies documenting that bioavailable vitamin D levels are related to other markers of mineral metabolism and that the vitamin D binding protein isoforms explain racial differences in overall 25(OH)D levels.13,16 Blood levels of free and bioavailable 25(OH)D may better characterize vitamin D status than overall levels of 25(OH)D.

We did not find that 1,25(OH)2D levels were related to ESRD risk in the present study. This is consistent with the recommendation not to measure this active form of vitamin D for the evaluation of vitamin D status since it is typically normal or elevated in individuals with vitamin D deficiency.28 Levels of 1,25(OH)2D are tightly regulated by parathyroid hormone in the kidney, and low levels are detected only in late stages of kidney disease.29 Since extrarenal tissues express 1α-hydroxylase and vitamin D receptors, 1,25(OH)2D levels may represent other disease states.30

This study adds to a broader body of literature on mineral metabolism biomarkers and kidney disease risk. In particular, fibroblast growth factor 23 (FGF-23), a bone-derived hormone, has many endocrine functions in the kidney, such as urinary phosphorus excretion and suppression of parathyroid hormone synthesis.31,32 Studies have also shown a bidirectional relationship between FGF-23 and vitamin D: FGF-23 inhibits vitamin D production by regulating key enzymes and vitamin D supplementation influences blood FGF-23 levels.33–35 The associations between mineral metabolism markers (FGF-23, vitamin D-related biomarkers) and ESRD from our present and previous studies remained significant after adjusting for calcium, phosphate, and parathyroid hormone; thus, there may be additional pathways by which markers of mineral metabolism confer risk.36 Future research should focus on identifying these pathways. Further, the use of multiple biomarkers of mineral metabolism may improve risk prediction and evaluation of pharmacologic and dietary interventions to reduce kidney disease risk.37,38

There are some limitations of this study that should be considered when interpreting these findings. We were unable to find exact matches for ESRD cases and controls based on kidney function measures. To account for these differences, we adjusted for the continuous forms of eGFR and UACR in multivariable regression models. The blood specimens were stored for approximately 15 years before laboratory measurement of vitamin D binding protein, 25(OH)D, and 1,25(OH)2D. Previous research has documented the stability of blood levels of vitamin D with a long duration of freezer storage.39 The ESRD cases were identified based on diagnostic codes for hospitalizations and deaths. Use of hospitalization data has been shown to be a valid method for defining ESRD and is more representative of the overall burden of kidney disease by incorporating treated and untreated kidney disease including deaths due to kidney disease.40,41 We previously reported that, compared to medical record review, our composite definition including ICD-9/10 codes for hospitalizations has a sensitivity of 88% and specificity of 97%.41 Assessment of free and bioavailable 25(OH)D was based on calculations rather than direct measurement. Future studies that directly measure free and bioavailable 25(OH)D may provide a better assessment of vitamin D status.

The present study has several strengths. This case-control study was nested within a well-characterized prospective cohort. As such, it was possible to match on several important factors due to the large sample size of the overall ARIC study. Strong confounders were selected as matching factors, resulting in an efficient study design, which mitigated confounding and maximized statistical power to detect the association between mineral metabolism biomarkers and ESRD. The findings in our study were robust to adjustment for demographics, kidney function, established risk factors, and other mineral metabolism markers (serum calcium, phosphate, and parathyroid hormone). Multiple methods to evaluate vitamin D status, as well as the isoform of vitamin D binding protein, were assessed in the present study, providing a comprehensive analysis of these important indicators of decreased kidney function.

In conclusion, the present study demonstrates that vitamin D binding protein is positively associated with ESRD risk in a community-based population, independent of demographics, baseline kidney function, established kidney disease risk factors, and other mineral metabolism markers (calcium, phosphate, parathyroid hormone). Furthermore, estimated free and bioavailable 25(OH)D were inversely associated with ESRD risk. Directly measuring the free and bioavailable fractions of overall 25(OH)D concentrations may be worthwhile for clinical and research purposes. Further research is warranted to determine whether these markers can be intervened upon to reduce ESRD risk.

Supplementary Material

1

Table S1. Case status, racial group, and levels of mineral metabolism biomarkersa according to vitamin D binding protein isoform

2

Figure S1. Blood concentrations of vitamin D binding protein according to (A) vitamin D binding protein single nucleotide polymorphism rs4588, (B) vitamin D binding protein single nucleotide polymorphism rs7041, and (C) vitamin D binding protein isoforms

The middle horizontal line represents the median value, the outer horizontal lines represent the 25th and 75th percentiles, the whiskers on either side of the box represent the most extreme value within 1.5 times the interquartile range, and the dots represent any data points outside the whiskers.

3

Item S1. Calculation of free and bioavailable 25-hydroxyvitamin D

Acknowledgments

Support: The ARIC Study is carried out as a collaborative study supported by National Heart, Lung and Blood Institute (NHLBI) contracts (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN268201100009C, HHSN268201100010C, HHSN268201100011C, and HHSN268201100012C). The authors thank the staff and participants of the ARIC study for their important contributions. This work was supported in part by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) grant number R01DK089174 (Principle Investigator: E. Selvin) and NHLBI grant number R01HL103706 (principle invesitgaor: P. Lutsey). Additional support is provided by U01DK085689 from the NIDDK (Chronic Kidney Disease Biomarkers Consortium). The study sponsors had no role in study design; collection, analysis, and interpretation of data; writing the report; and the decision to submit the report for publication.

Footnotes

Financial Disclosure: The authors declare that they have no other relevant financial interests.

Contributions: Research idea and study design: LAI, ASL, C-yH, JC; data acquisition: PLL, ANH, ES, PLK, RSV, JHE, JC; data analysis/interpretation: CMR, MEG, PLL, JRM, JC; statistical analysis: CMR, JC; supervision and mentorship: JC. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. JC takes responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

Supplementary Material

Table S1: Case status, racial group, and levels of mineral metabolism biomarkers according to vitamin D binding protein isoform.

Figure S1: Blood concentrations of vitamin D binding protein according to rs4588, rs7041, and isoforms.

Item S1: Calculation of free and bioavailable 25(OH)D.

Note: The supplementary material accompanying this article (doi:_______) is available at www.ajkd.org

Supplementary Material Descriptive Text for Online Delivery

Supplementary Table S1 (PDF). Case status, racial group, and levels of mineral metabolism biomarkers according to vitamin D binding protein isoform.

Supplementary Figure S1 (PDF). Blood concentrations of vitamin D binding protein according to rs4588, rs7041, and isoforms.

Supplementary Item S1 (PDF). Calculation of free and bioavailable 25(OH)D.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Craver L, Marco MP, Martinez I, et al. Mineral metabolism parameters throughout chronic kidney disease stages 1–5--achievement of K/DOQI target ranges. Nephrol Dial Transplant. 2007;22(4):1171–1176. doi: 10.1093/ndt/gfl718. [DOI] [PubMed] [Google Scholar]
  • 2.Scialla JJ, Astor BC, Isakova T, Xie H, Appel LJ, Wolf M. Mineral metabolites and CKD progression in African Americans. J Am Soc Nephrol. 2013;24(1):125–135. doi: 10.1681/ASN.2012070713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Levin A, Bakris GL, Molitch M, et al. Prevalence of abnormal serum vitamin D, PTH, calcium, and phosphorus in patients with chronic kidney disease: results of the study to evaluate early kidney disease. Kidney Int. 2007;71(1):31–38. doi: 10.1038/sj.ki.5002009. [DOI] [PubMed] [Google Scholar]
  • 4.Doorenbos CR, van den Born J, Navis G, de Borst MH. Possible renoprotection by vitamin D in chronic renal disease: beyond mineral metabolism. Nat Rev Nephrol. 2009;5(12):691–700. doi: 10.1038/nrneph.2009.185. [DOI] [PubMed] [Google Scholar]
  • 5.Koizumi M, Komaba H, Fukagawa M. Parathyroid function in chronic kidney disease: role of FGF23-Klotho axis. Contrib Nephrol. 2013;180:110–123. doi: 10.1159/000346791. [DOI] [PubMed] [Google Scholar]
  • 6.Reinhardt W, Dolff S, John P, Kribben A, Witzke O. Impact of renal failure on metabolic bone parameters in a vitamin D-deficient patient cohort. Clin Nephrol. 2011;75(5):403–409. doi: 10.5414/cnp75403. [DOI] [PubMed] [Google Scholar]
  • 7.National Kidney Foundation. KDOQI Clinical Practice Guideline for Diabetes and CKD: 2012 Update. Am J Kidney Dis. 2012;60(5):850–886. doi: 10.1053/j.ajkd.2012.07.005. [DOI] [PubMed] [Google Scholar]
  • 8.Kandula P, Dobre M, Schold JD, Schreiber MJ, Jr, Mehrotra R, Navaneethan SD. Vitamin D supplementation in chronic kidney disease: a systematic review and meta-analysis of observational studies and randomized controlled trials. Clin J Am Soc Nephrol. 2011;6(1):50–62. doi: 10.2215/CJN.03940510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Ravani P, Malberti F, Tripepi G, et al. Vitamin D levels and patient outcome in chronic kidney disease. Kidney Int. 2009;75(1):88–95. doi: 10.1038/ki.2008.501. [DOI] [PubMed] [Google Scholar]
  • 10.Melamed ML, Astor B, Michos ED, Hostetter TH, Powe NR, Muntner P. 25-hydroxyvitamin D levels, race, and the progression of kidney disease. J Am Soc Nephrol. 2009;20(12):2631–2639. doi: 10.1681/ASN.2009030283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Bikle DD, Siiteri PK, Ryzen E, Haddad JG. Serum protein binding of 1,25-dihydroxyvitamin D: a reevaluation by direct measurement of free metabolite levels. J Clin Endocrinol Metab. 1985;61(5):969–975. doi: 10.1210/jcem-61-5-969. [DOI] [PubMed] [Google Scholar]
  • 12.Mendel CM. The free hormone hypothesis: a physiologically based mathematical model. Endocr Rev. 1989;10(3):232–274. doi: 10.1210/edrv-10-3-232. [DOI] [PubMed] [Google Scholar]
  • 13.Bhan I, Powe CE, Berg AH, et al. Bioavailable vitamin D is more tightly linked to mineral metabolism than total vitamin D in incident hemodialysis patients. Kidney Int. 2012;82(1):84–89. doi: 10.1038/ki.2012.19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Bikle DD, Gee E, Halloran B, Kowalski MA, Ryzen E, Haddad JG. Assessment of the free fraction of 25-hydroxyvitamin D in serum and its regulation by albumin and the vitamin D-binding protein. J Clin Endocrinol Metab. 1986;63(4):954–959. doi: 10.1210/jcem-63-4-954. [DOI] [PubMed] [Google Scholar]
  • 15.Arnaud J, Constans J. Affinity differences for vitamin D metabolites associated with the genetic isoforms of the human serum carrier protein (DBP) Hum Genet. 1993;92(2):183–188. doi: 10.1007/BF00219689. [DOI] [PubMed] [Google Scholar]
  • 16.Powe CE, Evans MK, Wenger J, et al. Vitamin D-binding protein and vitamin D status of black Americans and white Americans. N Engl J Med. 2013;369(21):1991–2000. doi: 10.1056/NEJMoa1306357. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Lutsey PL, Eckfeldt JH, Ogagarue ER, Folsom AR, Michos ED, Gross M. The 25-hydroxyvitamin D3 C-3 epimer: distribution, correlates, and reclassification of 25-hydroxyvitamin D status in the population-based Atherosclerosis Risk in Communities Study (ARIC) Clin Chim Acta. 2015;442:75–81. doi: 10.1016/j.cca.2014.12.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.The ARIC investigators. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. Am J Epidemiol. 1989;129(4):687–702. [PubMed] [Google Scholar]
  • 19.Bash LD, Coresh J, Kottgen A, et al. Defining incident chronic kidney disease in the research setting: The ARIC Study. Am J Epidemiol. 2009;170(4):414–424. doi: 10.1093/aje/kwp151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604–612. doi: 10.7326/0003-4819-150-9-200905050-00006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Laha TJ, Strathmann FG, Wang Z, de Boer IH, Thummel KE, Hoofnagle AN. Characterizing antibody cross-reactivity for immunoaffinity purification of analytes prior to multiplexed liquid chromatography-tandem mass spectrometry. Clin Chem. 2012;58(12):1711–1716. doi: 10.1373/clinchem.2012.185827. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Henderson CM, Lutsey PL, Misialek JR, et al. Measurement by LC-MS/MS reveals similar plasma concentrations of vitamin D binding protein in blacks and whites. Clin Chem. 2015 doi: 10.1373/clinchem.2015.244541. under revision. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Musunuru K, Lettre G, Young T, et al. Candidate gene association resource (CARe): design, methods, and proof of concept. Circ Cardiovasc Genet. 2010;3(3):267–275. doi: 10.1161/CIRCGENETICS.109.882696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab. 1999;84(10):3666–3672. doi: 10.1210/jcem.84.10.6079. [DOI] [PubMed] [Google Scholar]
  • 25.Braun A, Bichlmaier R, Cleve H. Molecular analysis of the gene for the human vitamin-D-binding protein (group-specific component): allelic differences of the common genetic GC types. Hum Genet. 1992;89(4):401–406. doi: 10.1007/BF00194311. [DOI] [PubMed] [Google Scholar]
  • 26.Safadi FF, Thornton P, Magiera H, et al. Osteopathy and resistance to vitamin D toxicity in mice null for vitamin D binding protein. J Clin Invest. 1999;103(2):239–251. doi: 10.1172/JCI5244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Bikle DD, Gee E. Free, and not total, 1,25-dihydroxyvitamin D regulates 25-hydroxyvitamin D metabolism by keratinocytes. Endocrinology. 1989;124(2):649–654. doi: 10.1210/endo-124-2-649. [DOI] [PubMed] [Google Scholar]
  • 28.Holick MF, Binkley NC, Bischoff-Ferrari HA, et al. Evaluation, treatment, and prevention of vitamin D deficiency: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2011;96(7):1911–1930. doi: 10.1210/jc.2011-0385. [DOI] [PubMed] [Google Scholar]
  • 29.Al-Badr W, Martin KJ. Vitamin D and kidney disease. Clin J Am Soc Nephrol. 2008;3(5):1555–1560. doi: 10.2215/CJN.01150308. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Holick MF. Vitamin D deficiency. N Engl J Med. 2007;357(3):266–281. doi: 10.1056/NEJMra070553. [DOI] [PubMed] [Google Scholar]
  • 31.Dominguez JR, Shlipak MG, Whooley MA, Ix JH. Fractional excretion of phosphorus modifies the association between fibroblast growth factor-23 and outcomes. J Am Soc Nephrol. 2013;24(4):647–654. doi: 10.1681/ASN.2012090894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Nakanishi S, Kazama JJ, Nii-Kono T, et al. Serum fibroblast growth factor-23 levels predict the future refractory hyperparathyroidism in dialysis patients. Kidney Int. 2005;67(3):1171–1178. doi: 10.1111/j.1523-1755.2005.00184.x. [DOI] [PubMed] [Google Scholar]
  • 33.Shimada T, Hasegawa H, Yamazaki Y, et al. FGF-23 is a potent regulator of vitamin D metabolism and phosphate homeostasis. J Bone Miner Res. 2004;19(3):429–435. doi: 10.1359/JBMR.0301264. [DOI] [PubMed] [Google Scholar]
  • 34.Uzum AK, Salman S, Telci A, et al. Effects of vitamin D replacement therapy on serum FGF23 concentrations in vitamin D-deficient women in short term. Eur J Endocrinol. 2010;163(5):825–831. doi: 10.1530/EJE-10-0591. [DOI] [PubMed] [Google Scholar]
  • 35.Turner C, Dalton N, Inaoui R, Fogelman I, Fraser WD, Hampson G. Effect of a 300 000-IU loading dose of ergocalciferol (Vitamin D2) on circulating 1,25(OH)2-vitamin D and fibroblast growth factor-23 (FGF-23) in vitamin D insufficiency. J Clin Endocrinol Metab. 2013;98(2):550–556. doi: 10.1210/jc.2012-2790. [DOI] [PubMed] [Google Scholar]
  • 36.Rebholz CM, Grams ME, Coresh J, et al. Serum fibroblast growth factor-23 is associated with incident kidney disease. J Am Soc Nephrol. 2015;26(1):192–200. doi: 10.1681/ASN.2014020218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Adema AY, de Borst MH, Ter Wee PM, Vervloet MG Consortium N. Dietary and pharmacological modification of fibroblast growth factor-23 in chronic kidney disease. J Ren Nutr. 2014;24(3):143–150. doi: 10.1053/j.jrn.2013.09.001. [DOI] [PubMed] [Google Scholar]
  • 38.Taal MW, Thurston V, McIntyre NJ, Fluck RJ, McIntyre CW. The impact of vitamin D status on the relative increase in fibroblast growth factor 23 and parathyroid hormone in chronic kidney disease. Kidney Int. 2014;86(2):407–413. doi: 10.1038/ki.2013.537. [DOI] [PubMed] [Google Scholar]
  • 39.Agborsangaya C, Toriola AT, Grankvist K, et al. The effects of storage time and sampling season on the stability of serum 25-hydroxy vitamin D and androstenedione. Nutr Cancer. 2010;62(1):51–57. doi: 10.1080/01635580903191460. [DOI] [PubMed] [Google Scholar]
  • 40.Bash LD, Astor BC, Coresh J. Risk of incident ESRD: a comprehensive look at cardiovascular risk factors and 17 years of follow-up in the Atherosclerosis Risk in Communities (ARIC) Study. Am J Kidney Dis. 2010;55(1):31–41. doi: 10.1053/j.ajkd.2009.09.006. [DOI] [PubMed] [Google Scholar]
  • 41.Rebholz CM, Coresh J, Ballew SH, et al. Kidney Failure and ESRD in the Atherosclerosis Risk in Communities (ARIC) Study: Comparing Ascertainment of Treated and Untreated Kidney Failure in a Cohort Study. Am J Kidney Dis. 2015;66(2):231–239. doi: 10.1053/j.ajkd.2015.01.016. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1

Table S1. Case status, racial group, and levels of mineral metabolism biomarkersa according to vitamin D binding protein isoform

NIHMS725529-supplement-1.pdf (156KB, pdf)
2

Figure S1. Blood concentrations of vitamin D binding protein according to (A) vitamin D binding protein single nucleotide polymorphism rs4588, (B) vitamin D binding protein single nucleotide polymorphism rs7041, and (C) vitamin D binding protein isoforms

The middle horizontal line represents the median value, the outer horizontal lines represent the 25th and 75th percentiles, the whiskers on either side of the box represent the most extreme value within 1.5 times the interquartile range, and the dots represent any data points outside the whiskers.

NIHMS725529-supplement-2.pdf (282.7KB, pdf)
3

Item S1. Calculation of free and bioavailable 25-hydroxyvitamin D

NIHMS725529-supplement-3.pdf (174.6KB, pdf)

RESOURCES