Abstract
Background and objectives
Calcium and phosphorus regulatory hormones may contribute to the pathogenesis of calcium nephrolithiasis. However, there has been no prospective study to date of plasma hormone levels and risk of kidney stones. This study aimed to examine independent associations between plasma levels of 1,25-dihydroxyvitamin D (1,25[OH]2D), 25-hydroxyvitamin D, 24,25-dihydroxyvitamin D, fibroblast growth factor 23 (FGF23), parathyroid hormone, calcium, phosphate, and creatinine and the subsequent risk of incident kidney stones.
Design, setting, participants, & measurements
This study was a prospective, nested case-control study of men in the Health Professionals Follow-Up Study who were free of diagnosed nephrolithiasis at blood draw. During 12 years of follow-up, 356 men developed an incident symptomatic kidney stone. Using risk set sampling, controls were selected in a 2:1 ratio (n=712 controls) and matched for age, race, and year, month, and time of day of blood collection.
Results
Baseline plasma levels of 25-hydroxyvitamin D, 24,25-dihydroxyvitamin D, parathyroid hormone, calcium, phosphate, and creatinine were similar in cases and controls. Mean 1,25(OH)2D and median FGF23 levels were higher in cases than controls but differences were small and statistically nonsignificant (45.7 versus 44.2 pg/ml, P=0.07 for 1,25[OH]2D; 47.6 versus 45.1 pg/ml, P=0.08 for FGF23). However, after adjusting for body mass index, diet, plasma factors, and other covariates, the odds ratios of incident symptomatic kidney stones in the highest compared with lowest quartiles were 1.73 (95% confidence interval, 1.11 to 2.71; P for trend 0.01) for 1,25(OH)2D and 1.45 (95% confidence interval, 0.96 to 2.19; P for trend 0.03) for FGF23. There were no significant associations between other plasma factors and kidney stone risk.
Conclusions
Higher plasma 1,25(OH)2D, even in ranges considered normal, is independently associated with higher risk of symptomatic kidney stones. Although of borderline statistical significance, these findings also suggest that higher FGF23 may be associated with risk.
Keywords: kidney stones, risk factors, vitamin D, nephrolithiasis, fibroblast growth factor 23
Introduction
Calcium and phosphorus regulatory hormones may play a central role in the pathogenesis of calcium-containing kidney stones, the most common type of stone (hereafter referred to as “calcium” stones). Idiopathic hypercalciuria (IH) occurs in up to 60% of patients with calcium stones (1), and mean plasma levels of 1,25-dihydroxyvitamin D (1,25[OH]2D) in individuals with nephrolithiasis and IH exceeded that of controls in 11 of 13 cross-sectional studies (2). However, existing IH studies are small and there are well described groups of kidney stone formers with IH who do not have elevated plasma 1,25(OH)2D (3). It is unknown whether higher plasma 25-hydroxyvitamin D (25[OH]D) and/or vitamin D supplementation increases calcium kidney stone risk. Although multiple studies in individuals without kidney stones did not report increased urinary calcium after vitamin D supplementation (4–6), most evaluated urinary calcium as a binary outcome (potentially missing smaller but still clinically important increases in urine calcium). Other vitamin D moieties, including lower levels of the 25(OH)D metabolite 24,25-dihydroxyvitamin D (24,25[OH]2D), are associated with at least some cases of calcium nephrolithiasis (7–9).
Alterations in phosphorus homeostasis also may contribute to calcium kidney stones. In one cross-sectional study of 207 calcium stone formers and 105 controls, mean serum phosphate was 9% lower and fractional excretion of phosphate was 29% higher in stone formers (10). Other investigators also have reported lower values of serum phosphate in smaller series of patients with kidney stones (11–15). More recent data have led to speculation that higher plasma fibroblast growth factor 23 (FGF23), an osteocyte-derived phosphaturic hormone, may contribute to calcium kidney stone formation (16,17).
Studies to date of calcium and phosphorus regulatory hormones and nephrolithiasis have been cross-sectional and focused on urine factors rather than actual kidney stone formation. The difficulties of using urine composition as a proxy for kidney stone risk are manifest by the wide array of well established nonurinary kidney stone risk factors and recent data implicating the renal interstitium as the site of initial stone formation (18–20). To examine independent associations between plasma levels of 1,25(OH)2D, 25(OH)D, 24,25(OH)2D, FGF23, parathyroid hormone (PTH), phosphate, and calcium and the subsequent risk of incident symptomatic kidney stones, we conducted a prospective nested case-control study among 1068 participants of the Health Professionals Follow-Up Study (HPFS) who were free of diagnosed nephrolithiasis at blood draw.
Materials and Methods
Source Population
The HPFS enrolled 51,529 male dentists, optometrists, osteopaths, pharmacists, podiatrists, and veterinarians who were aged 40–75 years in 1986. At baseline, participants filled out a detailed questionnaire about diet, medical history, and medications. Subsequent questionnaires have been mailed every 2 years to update information on health-related behaviors, medications, and medical events. In 1993, blood samples were submitted by 18,225 participants.
Study Population
Based on the cohort that provided blood in 1993, and after exclusion of participants with a history of kidney stones and/or cancer before 1994, we identified 360 participants who had an incident, symptomatic kidney stone before January 31, 2006. Of these 360 participants, 47 had kidney stone composition reports available: 42 stones contained ≥50% calcium oxalate, one contained >50% calcium phosphate, and four were uric acid. Of the 42 stones that were ≥50% calcium oxalate, 12 contained calcium phosphate and none contained uric acid.
We excluded the four participants with uric acid stones. We used risk set sampling (21) to select two controls for each case and matched on age, race, month and year of blood collection, time of day of blood collection (within 2 hours), and fasting status (≥8 hours). Our final study population consisted of 356 kidney stone cases and 712 controls.
The research protocol for this study was reviewed and approved by the institutional review board of Brigham and Women’s Hospital.
Ascertainment of Kidney Stones
The study outcome was an incident kidney stone accompanied by pain or hematuria. HPFS participants reported on the interval diagnosis of kidney stones every 2 years. Participants reporting a new kidney stone were sent an additional questionnaire to determine the date of occurrence and symptoms from the stone. In a previous medical record validation study of self-reported kidney stones in the HPFS, the diagnosis was confirmed in 95% of cases (22). In medical records that contained a stone composition report, 86% had a stone that contained ≥50% calcium oxalate.
Ascertainment of Covariates
Anthropometric data, lifestyle factors, and diet were based on the 1994 questionnaires. Medical history information was obtained from biennial questionnaires. There have been multiple previous reports on the validity and reproducibility of these collected data and measurements (23–28).
The semiquantitative food frequency questionnaire asked about the annual average use of >130 individual foods and 22 beverages. Intakes of dietary factors were computed from the reported frequency of consumption of each specified unit of food and from data on the content of the relevant nutrient in specified portions. Nutrient values were adjusted for total caloric intake to determine the nutrient composition of the diet independent of the total amount of food eaten (29,30).
The intake of mineral and vitamin supplements (e.g., calcium and vitamin D) in multivitamins or isolated form was determined by the brand, type, and frequency of reported use. The reproducibility and validity of the food frequency questionnaire has been documented (28).
Measurement of Biochemical Variables
Blood samples were collected in 10-ml liquid EDTA blood tubes, placed on ice packs, stored in polystyrene containers, and returned via overnight courier. More than 95% of the samples arrived within 24 hours of collection. Blood samples were centrifuged and aliquoted for storage in the vapor phase of liquid nitrogen freezers (−130°C or colder).
Plasma factors were measured in the Department of Laboratory Medicine at the University of Washington School of Medicine. Levels of 1,25(OH)2D, 25(OH)D, and 24,25(OH)2D were measured by immunoaffinity extraction and liquid chromatography–tandem mass spectrometry (31). Intact FGF23 was measured with a two-site ELISA (Kainos Laboratories Inc, Tokyo, Japan). Calcium was measured using atomic absorption (PerkinElmer Inc, Waltham, MA). Intact PTH was measured with an automated sandwich immunoassay (Beckman-Coulter Inc, Brea, CA). Creatinine was measured using the Jaffe rate method and inorganic phosphate by reaction with ammonium molybdate in automated analyzers (Beckman-Coulter Inc). The number of study participants with missing values was as follows: PTH in 12, phosphate in seven, all vitamin D moieties in one, and creatinine in one.
To determine intra-assay variation, we measured plasma factors in 108 quality control samples interspersed among the 1068 samples from our study population. The mean intra-assay coefficients of variation were 14.9% for 1,25(OH)2D, 10.9% for 25(OH)D, 13.6% for 24,25(OH)2D, 19.6% for FGF23, 6.2% for calcium, 7.9% for intact PTH, 8.9% for creatinine, and 7.9% for phosphate.
We assessed within-person variation over time for plasma FGF23, 1,25(OH)2D, and 25(OH)D in previous HPFS pilot studies. In 20 HPFS participants, we measured FGF23 at two time points (1 year apart) and the intraclass correlation was 0.8 (32). In 144 HPFS participants with blood drawn 3 years apart, correlations were 0.5 for 1,25(OH)2D and 0.7 for 25(OH)D (33).
Statistical Analyses
Differences in continuous variables between cases and controls were analyzed using ANOVA (normal distributions) or the Wilcoxon rank sum test (non-normal distributions). Differences in categorical variables were compared using the chi-squared test.
The study design was prospective; blood was collected before the diagnosis of a symptomatic kidney stone. To preserve unmatched cases and controls in our study population after accounting for participants with missing plasma measurements, we used unconditional logistic regression adjusted for matching factors. In secondary analyses, we used conditional logistic regression after restricting the study population to matched cases and controls with complete blood results.
We divided plasma factors into four categories based on quartiles of the control distributions. In multivariable models, we adjusted for matching factors (age, race, month and year of blood collection, time of day, and fasting status) and the following potentially confounding variables: body mass index (BMI) (continuous), history of diabetes mellitus (yes or no), history of hypertension (yes or no), thiazide use (yes or no), calcium supplement use (0, 1–100, 101–500, or >500 mg per day), intakes of caffeine and alcohol (quartiles), dietary calcium, potassium, oxalate, and animal protein (quartiles), and all plasma factors (quartiles). The high correlation (r=0.79) between plasma 25(OH)D and 24,25(OH)2D precluded simultaneous inclusion of these two biomarkers in the same multivariable models. The significance of linear trend across quartiles of each plasma factor was tested by assigning the median value for the quartile to each participant and considering this value as a continuous variable in the multivariable regression models.
Data were analyzed by using SAS software (version 9.3; SAS Institute Inc, Cary, NC).
Results
Characteristics and plasma factors for cases and controls at baseline are shown in Table 1. Participants at baseline who subsequently developed incident nephrolithiasis had lower intakes of dietary calcium and potassium. Plasma levels of 25(OH)D, 24,25(OH)2D, PTH, phosphate, calcium, and creatinine were similar in cases and controls. Compared with participants without kidney stones, participants with an incident kidney stone during follow-up had statistically nonsignificant higher baseline mean 1,25(OH)2D (45.7 pg/ml versus 44.2 pg/ml, P=0.07) and median FGF23 (47.6 pg/ml versus 45.1 pg/ml, P=0.08).
Table 1.
Baseline characteristics of men with incident kidney stones and control participants
| Characteristic | Cases (n=356) | Controls (n=712) | P Value |
|---|---|---|---|
| Age, yra | 57.4 (8.1) | 57.4 (8.1) | — |
| Body mass index, kg/m2 | 26.3 (3.3) | 26.0 (3.3) | 0.11 |
| History of diabetes mellitus, n (%) | 14 (3.9) | 24 (3.4) | 0.64 |
| History of hypertension, n (%) | 85 (23.9) | 169 (23.7) | 0.96 |
| Thiazide use, n (%) | 5 (1.4) | 16 (2.3) | 0.35 |
| Energy-adjusted daily intake | |||
| Dietary calcium, mg | 782 (290) | 826 (303) | 0.03 |
| Animal protein, g | 60.4 (16.8) | 60.2 (15.8) | 0.88 |
| Potassium, mg | 3275 (598) | 3410 (638) | 0.001 |
| Oxalate, mg | 198 (96) | 209 (108) | 0.11 |
| Alcohol consumption, g/d | 5.0 (0–14.7) | 5.5 (0.9–14.9) | 0.31 |
| Caffeine intake, mg/d | 146 (40–367) | 170 (50–371) | 0.10 |
| Supplemental calcium, mg/d | 0 (0–40) | 0 (0–93) | 0.16 |
| Plasma factors and eGFR | |||
| Creatinine, mg/dl | 1.0 (0.2) | 1.0 (0.2) | 0.67 |
| eGFR, ml/min per 1.73 m2 | 86.6 (14.8) | 87.2 (14.6) | 0.53 |
| Calcium, mg/dl | 9.5 (0.6) | 9.5 (0.7) | 0.80 |
| Phosphate, mg/dl | 3.1 (1.0) | 3.1 (0.7) | 0.63 |
| PTH, pg/ml | 34.1 (18.3) | 34.3 (20.0) | 0.39 |
| 25(OH)D, ng/ml | 33.8 (11.1) | 33.5 (9.6) | 0.66 |
| 24,25(OH)2D, ng/ml | 3.7 (2.7–5.0) | 3.8 (2.8–5.1) | 0.32 |
| 1,25(OH)2D, pg/ml | 45.7 (13.4) | 44.2 (13.3) | 0.07 |
| FGF23, pg/ml | 47.6 (38.4–58.7) | 45.1 (37.7–56.7) | 0.08 |
Data are presented as mean (SD) or median (interquartile range) unless otherwise indicated. Laboratory reference ranges are as follows: calcium, 8.9–10.2 mg/dl; phosphate, 2.5–4.5 mg/dl; PTH, 17–65 pg/ml; 25(OH)D, 20.1–50 ng/ml; 24,25(OH)2D, not defined; 1,25(OH)2D, 17–72 pg/ml; and FGF23, not defined. Variables with median (interquartile range) have non-normal distributions. Creatinine values are rounded to the nearest 10th decimal place, but unrounded values were used to calculate eGFR using the Chronic Kidney Disease Epidemiology Collaboration equation. PTH, intact parathyroid hormone; 25(OH)D, 25-hydroxyvitamin D; 24,25(OH)2D, 24,25-dihydroxyvitamin D; 1,25(OH)2D, 1,25-dihydroxyvitamin D; FGF23, intact fibroblast growth factor 23. Dash, no P value was generated.
Matching factor.
Age-standardized baseline characteristics in controls by quartiles of plasma 1,25(OH)2D and FGF23 are shown in Table 2. Participants with higher plasma 1,25(OH)2D had lower BMI, were less likely to use thiazide diuretics, had lower calcium intake, and had higher plasma 25(OH)D and lower plasma FGF23. Participants with higher plasma FGF23 had higher BMI, were more likely to have a history of hypertension, and had lower plasma 1,25(OH)2D.
Table 2.
Baseline characteristics of controls by quartiles of 1,25(OH)2D and FGF23
| Characteristic | Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | P Value (Trend) |
|---|---|---|---|---|---|
| 1,25(OH)2D, pg/ml | ≤34.7 | 34.8–43.2 | 43.3–52.3 | ≥52.4 | |
| No. of controls | 178 | 179 | 177 | 177 | |
| Age, yr | 57.8 (8.2) | 58.4 (8.5) | 57.3 (8.3) | 56.0 (7.3) | 0.02 |
| Caucasian | 95.5 | 94.8 | 96.2 | 94.6 | 0.25 |
| Body mass index, kg/m2 | 26.2 (3.6) | 26.2 (3.3) | 25.6 (3.1) | 25.8 (3.4) | 0.01 |
| History of hypertension | 24.1 | 24.6 | 21.4 | 27.2 | 0.55 |
| History of diabetes | 2.8 | 4.9 | 3.3 | 2.8 | 0.45 |
| Thiazide use | 3.7 | 2.5 | 1.9 | 0.5 | 0.07 |
| Daily intakes | |||||
| Dietary calcium, mg | 869 (312) | 822 (291) | 826 (294) | 791 (320) | 0.02 |
| Animal protein, g | 59.7 (14.3) | 63.0 (17.5) | 58.2 (14.5) | 60.5 (16.3) | 0.67 |
| Potassium, mg | 3439 (634) | 3374 (659) | 3468 (625) | 3389 (657) | 0.85 |
| Plasma factors | |||||
| Creatinine, mg/dl | 1.0 (0.3) | 1.0 (0.2) | 0.9 (0.2) | 0.9 (0.2) | <0.001 |
| Calcium, mg/dl | 9.5 (0.6) | 9.5 (0.7) | 9.5 (0.8) | 9.6 (0.7) | 0.18 |
| Phosphate, mg/dl | 3.2 (0.9) | 3.1 (0.8) | 3.1 (0.6) | 3.1 (0.6) | 0.34 |
| PTH, pg/ml | 35.1 (13.2) | 34.8 (12.7) | 36.3 (13.5) | 38.2 (14.1) | 0.01 |
| 25(OH)D, ng/ml | 30.5 (8.2) | 32.8 (9.3) | 34.3 (9.8) | 36.8 (10.2) | <0.001 |
| 24,25(OH)2D, ng/ml | 3.7 (1.6) | 4.1 (1.8) | 4.1 (1.9) | 4.3 (2.1) | <0.001 |
| FGF23, pg/ml | 56.9 (22.1) | 48.9 (14.0) | 45.9 (15.4) | 42.2 (13.5) | <0.001 |
| FGF23, pg/ml | ≤37.6 | 37.7–45.1 | 45.2–56.6 | ≥ 56.7 | |
| No. of controls | 178 | 178 | 178 | 178 | |
| Age, yr | 56.4 (7.9) | 56.8 (7.9) | 58.4 (8.2) | 58.0 (8.4) | 0.09 |
| Caucasian | 96.1 | 96.0 | 93.2 | 96.3 | 0.80 |
| Body mass index, kg/m2 | 25.6 (3.0) | 25.8 (3.5) | 26.0 (3.5) | 26.3 (3.3) | 0.003 |
| History of hypertension | 21.7 | 19.4 | 26.4 | 28.0 | 0.01 |
| History of diabetes | 1.9 | 4.2 | 2.1 | 5.5 | 0.23 |
| Thiazide use | 2.0 | 0.7 | 3.0 | 3.2 | 0.06 |
| Daily intakes | |||||
| Dietary calcium, mg | 811 (280) | 812 (290) | 857 (331) | 821 (310) | 0.71 |
| Animal protein, g | 60.7 (16.8) | 60.8 (15.7) | 60.1 (16.0) | 59.2 (14.8) | 0.20 |
| Potassium, mg | 3465 (669) | 3379 (641) | 3421 (624) | 3376 (633) | 0.20 |
| Plasma factors | |||||
| Creatinine, mg/dl | 0.9 (0.2) | 0.9 (0.2) | 0.9 (0.2) | 1.0 (0.2) | <0.001 |
| Calcium, mg/dl | 9.4 (0.6) | 9.6 (0.9) | 9.6 (0.6) | 9.5 (0.6) | 0.14 |
| Phosphate, mg/dl | 3.1 (0.7) | 3.1 (0.7) | 3.0 (0.7) | 3.2 (0.8) | 0.16 |
| PTH, pg/ml | 36.3 (14.3) | 34.9 (13.9) | 36.0 (11.9) | 36.7 (13.1) | 0.88 |
| 25(OH)D, ng/ml | 32.1 (9.8) | 33.7 (10.0) | 33.8 (9.2) | 34.5 (9.4) | 0.01 |
| 24,25(OH)2D, ng/ml | 3.7 (2.0) | 4.0 (1.8) | 4.3 (1.8) | 4.3 (1.8) | <0.001 |
| 1,25(OH)2D, pg/ml | 49.4 (14.4) | 46.5 (12.1) | 42.6 (11.6) | 38.0 (11.5) | <0.001 |
Values are means (SD) or percentages and, with the exception of age, are standardized to the age distribution of the study population. Dietary factors are energy adjusted. One control participant had a missing plasma 1,25(OH)2D value.
Spearman correlation coefficients for plasma factors in controls are displayed in Table 3. Plasma levels of 1,25(OH)2D and FGF23 were inversely correlated (r=−0.35, P<0.001). Higher plasma 25(OH)D was highly correlated with higher 24,25(OH)2D (r=0.79, P<0.001). With the exception of the inverse correlation between PTH and 24,25(OH)2D (r=−0.28, P<0.001), correlations between other biomarkers were r<0.25.
Table 3.
Spearman correlation coefficients between plasma factors in control participants (n=712)
| Plasma Factor | PTH | Phosphorus | Creatinine | 25(OH)D | Calcium | 24,25(OH)2D | 1,25(OH)2D | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| r | P Value | r | P Value | r | P Value | r | P Value | r | P Value | r | P Value | r | P Value | |
| FGF23 | 0.06 | 0.13 | ‒0.02 | 0.67 | 0.17 | <0.001 | 0.08 | 0.03 | 0.05 | 0.20 | 0.11 | 0.003 | −0.35 | <0.001 |
| PTH | — | — | 0.03 | 0.45 | 0.01 | 0.81 | −0.21 | <0.001 | −0.11 | 0.004 | −0.28 | <0.001 | 0.08 | 0.04 |
| Phosphate | — | — | — | — | 0.14 | <0.001 | 0.01 | 0.87 | −0.01 | 0.88 | −0.03 | 0.48 | −0.02 | 0.63 |
| Creatinine | — | — | — | — | — | — | 0.17 | <0.001 | 0.03 | 0.40 | 0.04 | 0.33 | −0.11 | 0.01 |
| 25(OH)D | — | — | — | — | — | — | — | — | 0.07 | 0.07 | 0.79 | <0.001 | 0.24 | <0.001 |
| Calcium | — | — | — | — | — | — | — | — | — | — | 0.10 | 0.01 | 0.03 | 0.41 |
| 24,25(OH)2D | — | — | — | — | — | — | — | — | — | — | — | — | 0.11 | 0.004 |
The number of control participants with missing values was as follows: PTH in nine, phosphate in three, and all vitamin D moieties in one. Dash, represents values presented elsewhere in the table.
Multivariable adjusted odds ratios of incident kidney stones across quartiles of each plasma factor during 12 years of follow-up are presented in Table 4. When present, unequal numbers of control participants in quartiles were due to groups of identical plasma values.
Table 4.
Multivariable odds ratios of incident kidney stones by quartiles of baseline 1,25(OH)2D, FGF23, 25(OH)D, 24,25(OH)2D, PTH, phosphate, and calcium
| Biomarker | Quartile 1 | Quartile 2 | Quartile 3 | Quartile 4 | P Value for Trend |
|---|---|---|---|---|---|
| 1,25(OH)2D | |||||
| Median, pg/ml | 30.1 | 39.0 | 47.4 | 58.5 | |
| No. of controls | 178 | 179 | 177 | 177 | |
| No. of cases | 71 | 92 | 98 | 95 | |
| Model 1 | 1.0 reference | 1.29 (0.89 to 1.87) | 1.39 (0.96 to 2.01) | 1.33 (0.92 to 1.93) | 0.15 |
| Model 2 | 1.0 reference | 1.26 (0.86 to 1.87) | 1.47 (1.0 to 2.16) | 1.41 (0.95 to 2.09) | 0.08 |
| Model 3 | 1.0 reference | 1.37 (0.91 to 2.05) | 1.68 (1.11 to 2.54) | 1.73 (1.11 to 2.71) | 0.01 |
| FGF23 | |||||
| Median, pg/ml | 33.0 | 41.4 | 50.4 | 65.1 | |
| No. of controls | 178 | 178 | 178 | 178 | |
| No. of cases | 82 | 73 | 98 | 103 | |
| Model 1 | 1.0 reference | 0.89 (0.61 to 1.30) | 1.20 (0.83 to 1.72) | 1.27 (0.89 to 1.83) | 0.08 |
| Model 2 | 1.0 reference | 0.85 (0.57 to 1.25) | 1.12 (0.77 to 1.63) | 1.23 (0.84 to 1.79) | 0.13 |
| Model 3 | 1.0 reference | 0.85 (0.57 to 1.27) | 1.25 (0.84 to 1.85) | 1.45 (0.96 to 2.19) | 0.03 |
| 25(OH)D | |||||
| Median, ng/ml | 23.2 | 29.7 | 35.7 | 44.5 | |
| No. of controls | 178 | 178 | 178 | 177 | |
| No. of cases | 85 | 101 | 83 | 87 | |
| Model 1 | 1.0 reference | 1.17 (0.82 to 1.68) | 0.97 (0.67 to 1.41) | 1.03 (0.71 to 1.49) | 0.87 |
| Model 2 | 1.0 reference | 1.29 (0.89 to 1.87) | 1.10 (0.74 to 1.62) | 1.17 (0.79 to 1.73) | 0.65 |
| Model 3 | 1.0 reference | 1.17 (0.80 to 1.72) | 0.98 (0.65 to 1.48) | 0.95 (0.62 to 1.46) | 0.55 |
| 24,25(OH)2D | |||||
| Median, ng/ml | 2.2 | 3.3 | 4.4 | 6.2 | |
| No. of controls | 179 | 177 | 180 | 175 | |
| No. of cases | 105 | 89 | 77 | 85 | |
| Model 1 | 1.0 reference | 0.85 (0.60 to 1.22) | 0.72 (0.50 to 1.04) | 0.82 (0.57 to 1.18) | 0.25 |
| Model 2 | 1.0 reference | 0.94 (0.65 to 1.36) | 0.80 (0.55 to 1.18) | 0.93 (0.63 to 1.37) | 0.63 |
| Model 4 | 1.0 reference | 0.87 (0.59 to 1.27) | 0.75 (0.51 to 1.12) | 0.77 (0.51 to 1.16) | 0.21 |
| PTH | |||||
| Median, pg/ml | 22.2 | 30.4 | 38.1 | 49.5 | |
| No. of controls | 176 | 177 | 175 | 175 | |
| No. of cases | 102 | 83 | 75 | 93 | |
| Model 1 | 1.0 reference | 0.80 (0.56 to 1.14) | 0.73 (0.51 to 1.05) | 0.90 (0.63 to 1.28) | 0.57 |
| Model 2 | 1.0 reference | 0.76 (0.52 to 1.11) | 0.69 (0.47 to 1.01) | 0.86 (0.60 to 1.25) | 0.47 |
| Model 3 | 1.0 reference | 0.74 (0.50 to 1.09) | 0.69 (0.47 to 1.02) | 0.83 (0.56 to 1.23) | 0.37 |
| Phosphate | |||||
| Median, mg/dl | 2.5 | 3.0 | 3.3 | 3.8 | |
| No. of controls | 205 | 214 | 127 | 163 | |
| No. of cases | 113 | 102 | 59 | 78 | |
| Model 1 | 1.0 reference | 0.85 (0.61 to 1.19) | 0.82 (0.55 to 1.20) | 0.84 (0.58 to 1.21) | 0.30 |
| Model 2 | 1.0 reference | 0.87 (0.62 to 1.22) | 0.86 (0.57 to 1.29) | 0.86 (0.58 to 1.26) | 0.42 |
| Model 3 | 1.0 reference | 0.88 (0.62 to 1.25) | 0.86 (0.57 to 1.31) | 0.83 (0.56 to 1.24) | 0.32 |
| Calcium | |||||
| Median, mg/dl | 8.8 | 9.4 | 9.7 | 10.2 | |
| No. of controls | 185 | 208 | 168 | 151 | |
| No. of cases | 98 | 78 | 101 | 79 | |
| Model 1 | 1.0 reference | 0.70 (0.49 to 1.01) | 1.12 (0.78 to 1.59) | 0.97 (0.67 to 1.41) | 0.76 |
| Model 2 | 1.0 reference | 0.69 (0.47 to 1.00) | 1.12 (0.78 to 1.62) | 0.98 (0.67 to 1.45) | 0.73 |
| Model 3 | 1.0 reference | 0.67 (0.46 to 0.98) | 1.04 (0.71 to 1.52) | 0.94 (0.63 to 1.39) | 0.96 |
Data are presented as odds ratios with 95% confidence intervals, unless otherwise indicated. Model 1 is adjusted for matching factors (age, race, month and year of blood collection, time of day, and fasting status). Model 2 is adjusted for matching factors and also body mass index (continuous), history of diabetes mellitus (yes or no), history of hypertension (yes or no), thiazide use (yes or no), supplemental calcium (four categories), intakes of caffeine and alcohol (quartiles), and dietary calcium, potassium, oxalate, and animal protein (quartiles). Model 3 is adjusted for matching factors, for all of the variables in model 2, and for all of the plasma factors (quartiles) except 24,25(OH)2D. Model 4 is adjusted for matching factors, for all of the variables in model 2, and for all of the plasma factors (quartiles) except 25(OH)D.
The median plasma 1,25(OH)2D level for participants in the highest quartile was approximately twice that in the lowest. After adjusting for matching factors (age, race, month and year of blood collection, time of day, and fasting status), BMI, history of diabetes mellitus, history of hypertension, thiazide use, supplemental calcium, intakes of caffeine and alcohol, dietary calcium, potassium, oxalate, and animal protein, and other plasma factors, the odds ratio of incident kidney stone formation for participants in the highest compared with lowest quartile of 1,25(OH)2D was 1.73 (95% confidence interval [95% CI], 1.11 to 2.71; P for trend 0.01). Inclusion of BMI in multivariable models increased somewhat the odds ratios for plasma 1,25(OH)2D. However, the large increase in odds ratios for 1,25(OH)2D in the full multivariable analyses (compared with odds ratios adjusted only for matching factors) was due to the inclusion of FGF23 in regression models.
The median plasma FGF23 level for participants in the highest quartile was approximately twice that in the lowest. In the multivariable model, the odds ratio of incident kidney stone formation for participants in the highest compared with lowest quartile of FGF23 was 1.45 (95% CI, 0.96 to 2.19; P trend 0.03). The increase in odds ratios for FGF23 in multivariable analyses was due to the inclusion of 1,25(OH)2D in the regression model.
Plasma levels of 25(OH)D, 24,25(OH)2D, PTH, phosphate, and calcium were not significantly associated with kidney stone risk (Table 4).
We conducted a variety of secondary analyses. Conditional logistic regression yielded similar results to the primary analysis. Excluding participants taking thiazide diuretics did not change the results. We hypothesized that odds ratios for participants in the highest compared with lowest quartile of FGF23 would be higher in those with phosphate levels below the median (3.1 mg/dl) or <2.5 mg/dl. However, there was not a statistically significant interaction between phosphate, FGF23, and kidney stones (P for interaction 0.58). Odds ratios for quartiles of plasma 24,25(OH)2D and dietary calcium were similar in participants with plasma 1,25(OH)2D above and below the median (P values for interaction ≥0.53). Odds ratios for quartiles of 1,25(OH)2D were similar in participants with plasma 25(OH)D above and below the median, and in participants with plasma phosphate above and below the median (P values for interaction ≥0.49). Odds ratios for quartiles of PTH were similar in participants with plasma 25(OH)D above and below 30 ng/ml (P for interaction 0.65).
Discussion
We report results from the first prospective study to examine associations between calcium and phosphorus regulatory hormones and risk of kidney stones. Higher levels of 1,25(OH)2D, even in ranges considered normal, were associated with a higher odds of incident nephrolithiasis independent of age, body size, diet, and other factors. We also observed a modest association of borderline statistical significance between higher levels of FGF23 and higher odds of incident kidney stones.
Our study was not designed to elucidate mechanism(s) whereby higher plasma 1,25(OH)2D may lead to kidney stone formation. However, the association of 1,25(OH)2D with stone risk was independent of calcium intake and plasma levels of PTH, phosphorus, and 25(OH)D, suggesting that the putative effects of 1,25(OH)2D on risk were not mediated by conventional modulators of 1-α-hydroxylase or by differences in vitamin D body stores. To date, the hypothesis that 1,25(OH)2D may contribute to calcium nephrolithiasis has been based on observations about vitamin D physiology and smaller studies of IH. Plasma 1,25(OH)2D stimulates intestinal calcium absorption and bone resorption, and many stone formers with IH exhibit increased calcium absorption from the gut as well as increased calcium mobilization from bone (34). The administration of calcitriol to normal men replicates characteristic features of IH (35,36), and mean plasma 1,25(OH)2D in stone-forming individuals with IH exceeded that of controls in 11 of 13 cross-sectional studies (2).
Higher levels of FGF23 were previously reported in some calcium stone formers (16,17). In a cross-sectional study comparing 110 individuals with calcium nephrolithiasis to 60 controls, C-terminal FGF23 levels in 22 stone formers with “renal phosphate leak” (predefined as serum phosphate <2.5 mg/dl in conjunction with higher urine phosphate losses) were about three times higher than in stone formers with higher serum phosphate and in controls (17). Stone formers without lower serum phosphate had 31% higher C-terminal FGF23 levels compared with controls, but the difference was statistically nonsignificant. In our study, the magnitude of association between higher intact FG23 and higher stone risk was not greater in individuals with plasma phosphate below the median of 3.1 mg/dl or <2.5 mg/dl.
The mechanisms whereby higher FGF23 might lead to calcium stone formation are unclear. FGF23 reduces proximal tubular phosphorus reabsorption, and higher urine phosphorus increases urine supersaturation of calcium phosphate. However, we would expect the majority of kidney stones in our study to be predominantly calcium oxalate. FGF23 is an osteocyte-derived hormone, and it is possible that higher levels in stone formers are a marker of abnormal bone physiology. IH is associated with lower bone mineral density (37) and individuals with kidney stones may be at higher risk of subsequent bone fracture (38). In the general population, some but not all studies have reported associations between higher FGF23 and higher fracture risk (39–41).
The magnitude of association between 1,25(OH)2D and kidney stone odds increased after adjustment for FGF23, and the association for FGF23 increased after adjustment for 1,25(OH)2D. FGF23 downregulates 1-α-hydroxylase, and there was an inverse relation between FGF23 and 1,25(OH)2D in our study population. One possible interpretation of these data is that 1,25(OH)2D and FGF23 promote kidney stone formation via separate pathways, and that the lower odds ratios in unadjusted compared with adjusted analyses in our study were primarily due to participants with higher plasma 1,25(OH)2D having, on average, lower FGF23 (and vice versa). In this scenario, we would expect higher 1,25(OH)2D to associate with increased stone risk in unadjusted analyses in a population composed of individuals with similar FGF23 levels. Regardless of interpretation, our data suggest that future studies of 1,25(OH)2D and calcium kidney stone disease should account for potentially important interindividual differences in plasma FGF23. Of note, in a previous study of HPFS participants, we reported significant associations between plasma FGF23 and a wide variety of environmental factors, dietary intakes, and comorbidities (42).
In this study, the bounds of the 95% CIs for kidney stone odds comparing highest with lowest quartiles of plasma phosphate, 25(OH)D, PTH, and calcium do not exclude clinically meaningful associations. However, the lack of association between plasma 25(OH)D and kidney stone risk is consistent with a null cross-sectional report of 25(OH)D levels and prevalent nephrolithiasis in the National Health and Nutrition Examination Survey (43) and with previous studies reporting no increase in urine calcium after vitamin D supplementation in nonstone formers and stone formers (4–6,44). As with the other plasma factors in our study, the mean level of 25(OH)D was well within the reference range. It is possible that higher levels of 25(OH)D are associated with higher kidney stone risk.
Recent interest in plasma 24,25(OH)2D and calcium nephrolithiasis has been generated by the identification of stone-forming individuals with loss-of-function CYP24A1 mutations (7–9). The CYP24A1 gene encodes 24-hydroxylase, which metabolizes 25(OH)D to 24,25(OH)2D and catalyzes the degradation pathway of 1,25(OH)2D. It is difficult to draw conclusions about the potential importance of CYP24A1 based on the null plasma 24,25(OH)2D results in our study because the actual frequency of functional CYP24A1 variants is unknown and because our stone formers were not selected for higher 24-hour urine calcium.
Our study has limitations. First, the intra-assay coefficient of variation for FGF23 in our quality control samples was relatively high. Assuming such measurement error was random with respect to kidney stone formation, it is reasonable to speculate that our study may have underestimated the magnitude of risk associated with FGF23. Second, many plasma factors (e.g., phosphorus and PTH) exhibit postprandial and/or diurnal variation that may have contributed to nonsignificant results. However, the majority of study participants were fasting at blood draw, and matching factors included fasting status and time of day at blood draw. Third, we did not have data on kidney stone type or 24-hour urine composition for most of our study participants. It is possible that some of the nonsignificant plasma factors in our study may be associated with risk in individuals with more specific phenotypes. It is also likely that some of the participants in our study had uric acid stones. Finally, our study population was male and predominantly white. We have previously reported sex- and race-specific differences in kidney stone risk factors (45–49).
Our prospective study establishes the important role of 1,25(OH)2D in calcium kidney stone formation and suggests FGF23 as a novel risk factor for nephrolithiasis. Because of the inverse relation between plasma 1,25(OH)2D and FGF23 and the positive, independent associations between each of these plasma factors and kidney stone risk, future studies of 1,25(OH)2D and calcium nephrolithiasis require assessment of FGF23. Additional research is needed to examine associations between 1,25(OH)2D, FG23, and kidney stone risk in women, to elucidate the mechanism(s) of increased 1,25(OH)2D in calcium nephrolithiasis, and to determine how FGF23 may promote kidney stone formation.
Disclosures
G.C.C. is Editor-in-Chief of the Clinical Journal of the American Society of Nephrology. G.C.C. and E.N.T. contribute to UpToDate. A.N.H. receives research support from Waters Corporation (Milford, MA) and Thermo Fisher Scientific (Waltham, MA).
Acknowledgments
Because Dr. Curhan is the Editor-in-Chief of CJASN, he was not involved in the peer-review process for this manuscript. Another editor oversaw the peer-review and decision-making process for this manuscript.
Funding for this study was provided by the National Institutes of Health (Grants DK94910, DK70756, HL35464, and CA55075).
These data were presented in part at the Research on Calculus Kinetics Society Meeting, held February 7–8, 2014, in Ft. Lauderdale, Florida.
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
References
- 1.Lemann J, Jr: Pathogenesis of idiopathic hypercalciuria and nephrolithiasis. In: Disorders of Bone and Mineral Metabolism, edited by Coe FL, Favus MJ, New York, Raven Press, 1992, pp 685–706 [Google Scholar]
- 2.Coe F, Parks J, Evan A, Worcester E: Pathogenesis and treatment of nephrolithiasis. In: Seldin and Giebisch’s The Kidney, edited by Alpern R, Hebert SC, Academic Press, 2007, pp 1945–1977 [Google Scholar]
- 3.Coe FL, Favus MJ, Crockett T, Strauss AL, Parks JH, Porat A, Gantt CL, Sherwood LM: Effects of low-calcium diet on urine calcium excretion, parathyroid function and serum 1,25(OH)2D3 levels in patients with idiopathic hypercalciuria and in normal subjects. Am J Med 72: 25–32, 1982 [DOI] [PubMed] [Google Scholar]
- 4.Aloia JF, Patel M, Dimaano R, Li-Ng M, Talwar SA, Mikhail M, Pollack S, Yeh JK: Vitamin D intake to attain a desired serum 25-hydroxyvitamin D concentration. Am J Clin Nutr 87: 1952–1958, 2008 [DOI] [PubMed] [Google Scholar]
- 5.Tjellesen L, Hummer L, Christiansen C, Rødbro P: Serum concentration of vitamin D metabolites during treatment with vitamin D2 and D3 in normal premenopausal women. Bone Miner 1: 407–413, 1986 [PubMed] [Google Scholar]
- 6.Vieth R, Chan PC, MacFarlane GD: Efficacy and safety of vitamin D3 intake exceeding the lowest observed adverse effect level. Am J Clin Nutr 73: 288–294, 2001 [DOI] [PubMed] [Google Scholar]
- 7.Dinour D, Beckerman P, Ganon L, Tordjman K, Eisenstein Z, Holtzman EJ: Loss-of-function mutations of CYP24A1, the vitamin D 24-hydroxylase gene, cause long-standing hypercalciuric nephrolithiasis and nephrocalcinosis. J Urol 190: 552–557, 2013 [DOI] [PubMed] [Google Scholar]
- 8.Nesterova G, Malicdan MC, Yasuda K, Sakaki T, Vilboux T, Ciccone C, Horst R, Huang Y, Golas G, Introne W, Huizing M, Adams D, Boerkoel CF, Collins MT, Gahl WA: 1,25-(OH)2D-24 hydroxylase (CYP24A1) deficiency as a cause of nephrolithiasis. Clin J Am Soc Nephrol 8: 649–657, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Tebben PJ, Milliner DS, Horst RL, Harris PC, Singh RJ, Wu Y, Foreman JW, Chelminski PR, Kumar R: Hypercalcemia, hypercalciuria, and elevated calcitriol concentrations with autosomal dominant transmission due to CYP24A1 mutations: Effects of ketoconazole therapy. J Clin Endocrinol Metab 97: E423–E427, 2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Prié D, Ravery V, Boccon-Gibod L, Friedlander G: Frequency of renal phosphate leak among patients with calcium nephrolithiasis. Kidney Int 60: 272–276, 2001 [DOI] [PubMed] [Google Scholar]
- 11.Jaeger P, Portmann L, Ginalski JM, Jacquet AF, Temler E, Burckhardt P: Tubulopathy in nephrolithiasis: Consequence rather than cause. Kidney Int 29: 563–571, 1986 [DOI] [PubMed] [Google Scholar]
- 12.Lau YK, Wasserstein A, Westby GR, Bosanac P, Grabie M, Mitnick P, Slatopolsky E, Goldfarb S, Agus ZS: Proximal tubular defects in idiopathic hypercalciuria: Resistance to phosphate administration. Miner Electrolyte Metab 7: 237–249, 1982 [PubMed] [Google Scholar]
- 13.Pabico RC, McKenna BA, Freeman RB: Renal threshold phosphate concentration in patients with idiopathic nephrolithiasis: Correlations with tubular functions, serum parathyroid hormone and 1,25(OH)2D3. Proc Eur Dial Transplant Assoc 20: 450–454, 1983 [PubMed] [Google Scholar]
- 14.Williams CP, Child DF, Hudson PR, Soysa LD, Davies GK, Davies MG, De Bolla AR: Inappropriate phosphate excretion in idiopathic hypercalciuria: The key to a common cause and future treatment? J Clin Pathol 49: 881–888, 1996 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Worcester EM, Gillen DL, Evan AP, Parks JH, Wright K, Trumbore L, Nakagawa Y, Coe FL: Evidence that postprandial reduction of renal calcium reabsorption mediates hypercalciuria of patients with calcium nephrolithiasis. Am J Physiol Renal Physiol 292: F66–F75, 2007 [DOI] [PubMed] [Google Scholar]
- 16.Rendina D, Esposito T, Mossetti G, De Filippo G, Gianfrancesco F, Perfetti A, Magliocca S, Formisano P, Prié D, Strazzullo P: A functional allelic variant of the FGF23 gene is associated with renal phosphate leak in calcium nephrolithiasis. J Clin Endocrinol Metab 97: E840–E844, 2012 [DOI] [PubMed] [Google Scholar]
- 17.Rendina D, Mossetti G, De Filippo G, Cioffi M, Strazzullo P: Fibroblast growth factor 23 is increased in calcium nephrolithiasis with hypophosphatemia and renal phosphate leak. J Clin Endocrinol Metab 91: 959–963, 2006 [DOI] [PubMed] [Google Scholar]
- 18.Evan A, Lingeman J, Coe FL, Worcester E: Randall’s plaque: Pathogenesis and role in calcium oxalate nephrolithiasis. Kidney Int 69: 1313–1318, 2006 [DOI] [PubMed] [Google Scholar]
- 19.Evan AP, Lingeman JE, Coe FL, Parks JH, Bledsoe SB, Shao Y, Sommer AJ, Paterson RF, Kuo RL, Grynpas M: Randall’s plaque of patients with nephrolithiasis begins in basement membranes of thin loops of Henle. J Clin Invest 111: 607–616, 2003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Miller NL, Gillen DL, Williams JC, Jr, Evan AP, Bledsoe SB, Coe FL, Worcester EM, Matlaga BR, Munch LC, Lingeman JE: A formal test of the hypothesis that idiopathic calcium oxalate stones grow on Randall’s plaque. BJU Int 103: 966–971, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Prentice RL, Breslow NE: Retrospective studies and failure time models. Biometrika 65: 153–158, 1978 [Google Scholar]
- 22.Taylor EN, Curhan GC: Oxalate intake and the risk for nephrolithiasis. J Am Soc Nephrol 18: 2198–2204, 2007 [DOI] [PubMed] [Google Scholar]
- 23.Ascherio A, Rimm EB, Giovannucci EL, Colditz GA, Rosner B, Willett WC, Sacks F, Stampfer MJ: A prospective study of nutritional factors and hypertension among US men. Circulation 86: 1475–1484, 1992 [DOI] [PubMed] [Google Scholar]
- 24.Chasan-Taber S, Rimm EB, Stampfer MJ, Spiegelman D, Colditz GA, Giovannucci E, Ascherio A, Willett WC: Reproducibility and validity of a self-administered physical activity questionnaire for male health professionals. Epidemiology 7: 81–86, 1996 [DOI] [PubMed] [Google Scholar]
- 25.Giovannucci E, Colditz G, Stampfer MJ, Rimm EB, Litin L, Sampson L, Willett WC: The assessment of alcohol consumption by a simple self-administered questionnaire. Am J Epidemiol 133: 810–817, 1991 [DOI] [PubMed] [Google Scholar]
- 26.Hu FB, Leitzmann MF, Stampfer MJ, Colditz GA, Willett WC, Rimm EB: Physical activity and television watching in relation to risk for type 2 diabetes mellitus in men. Arch Intern Med 161: 1542–1548, 2001 [DOI] [PubMed] [Google Scholar]
- 27.Rimm EB, Stampfer MJ, Colditz GA, Chute CG, Litin LB, Willett WC: Validity of self-reported waist and hip circumferences in men and women. Epidemiology 1: 466–473, 1990 [DOI] [PubMed] [Google Scholar]
- 28.Rimm EB, Giovannucci EL, Stampfer MJ, Colditz GA, Litin LB, Willett WC: Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. Am J Epidemiol 135: 1114–1126, discussion 1127–1136, 1992 [DOI] [PubMed] [Google Scholar]
- 29.Willett W, Stampfer MJ: Total energy intake: Implications for epidemiologic analyses. Am J Epidemiol 124: 17–27, 1986 [DOI] [PubMed] [Google Scholar]
- 30.Willett WC: Nutritional Epidemiology, New York, Oxford University Press, 1998 [Google Scholar]
- 31.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 58: 1711–1716, 2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Taylor EN, Rimm EB, Stampfer MJ, Curhan GC: Plasma fibroblast growth factor 23, parathyroid hormone, phosphorus, and risk of coronary heart disease. Am Heart J 161: 956–962, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Platz EA, Leitzmann MF, Hollis BW, Willett WC, Giovannucci E: Plasma 1,25-dihydroxy- and 25-hydroxyvitamin D and subsequent risk of prostate cancer. Cancer Causes Control 15: 255–265, 2004 [DOI] [PubMed] [Google Scholar]
- 34.Worcester EM, Coe FL: New insights into the pathogenesis of idiopathic hypercalciuria. Semin Nephrol 28: 120–132, 2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Adams ND, Gray RW, Lemann J, Jr, Cheung HS: Effects of calcitriol administration on calcium metabolism healthy men. Kidney Int 21: 90–97, 1982 [DOI] [PubMed] [Google Scholar]
- 36.Maierhofer WJ, Gray RW, Cheung HS, Lemann J, Jr: Bone resorption stimulated by elevated serum 1,25-(OH)2-vitamin D concentrations in healthy men. Kidney Int 24: 555–560, 1983 [DOI] [PubMed] [Google Scholar]
- 37.Asplin JR, Donahue S, Kinder J, Coe FL: Urine calcium excretion predicts bone loss in idiopathic hypercalciuria. Kidney Int 70: 1463–1467, 2006 [DOI] [PubMed] [Google Scholar]
- 38.Melton LJ, 3rd, Crowson CS, Khosla S, Wilson DM, O’Fallon WM: Fracture risk among patients with urolithiasis: A population-based cohort study. Kidney Int 53: 459–464, 1998 [DOI] [PubMed] [Google Scholar]
- 39.Jovanovich A, Bùzková P, Chonchol M, Robbins J, Fink HA, de Boer IH, Kestenbaum B, Katz R, Carbone L, Lee J, Laughlin GA, Mukamal KJ, Fried LF, Shlipak MG, Ix JH: Fibroblast growth factor 23, bone mineral density, and risk of hip fracture among older adults: The cardiovascular health study. J Clin Endocrinol Metab 98: 3323–3331, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Lane NE, Parimi N, Corr M, Yao W, Cauley JA, Nielson CM, Ix JH, Kado D, Orwoll E, Osteoporotic Fractures in Men (MrOS) Study Group : Association of serum fibroblast growth factor 23 (FGF23) and incident fractures in older men: The Osteoporotic Fractures in Men (MrOS) study. J Bone Miner Res 28: 2325–2332, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Mirza MA, Karlsson MK, Mellström D, Orwoll E, Ohlsson C, Ljunggren O, Larsson TE: Serum fibroblast growth factor-23 (FGF-23) and fracture risk in elderly men. J Bone Miner Res 26: 857–864, 2011 [DOI] [PubMed] [Google Scholar]
- 42.Gutiérrez OM, Wolf M, Taylor EN: Fibroblast growth factor 23, cardiovascular disease risk factors, and phosphorus intake in the Health Professionals Follow-Up Study. Clin J Am Soc Nephrol 6: 2871–2878, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Tang J, McFann KK, Chonchol MB: Association between serum 25-hydroxyvitamin D and nephrolithiasis: The National Health and Nutrition Examination Survey III, 1988-94. Nephrol Dial Transplant 27: 4385–4389, 2012 [DOI] [PubMed] [Google Scholar]
- 44.Leaf DE, Korets R, Taylor EN, Tang J, Asplin JR, Goldfarb DS, Gupta M, Curhan GC: Effect of vitamin D repletion on urinary calcium excretion among kidney stone formers. Clin J Am Soc Nephrol 7: 829–834, 2012 [DOI] [PubMed] [Google Scholar]
- 45.Curhan GC, Willett WC, Speizer FE, Spiegelman D, Stampfer MJ: Comparison of dietary calcium with supplemental calcium and other nutrients as factors affecting the risk for kidney stones in women. Ann Intern Med 126: 497–504, 1997 [DOI] [PubMed] [Google Scholar]
- 46.Curhan GC, Willett WC, Knight EL, Stampfer MJ: Dietary factors and the risk of incident kidney stones in younger women: Nurses’ Health Study II. Arch Intern Med 164: 885–891, 2004 [DOI] [PubMed] [Google Scholar]
- 47.Taylor EN, Curhan GC: Differences in 24-hour urine composition between black and white women. J Am Soc Nephrol 18: 654–659, 2007 [DOI] [PubMed] [Google Scholar]
- 48.Taylor EN, Stampfer MJ, Curhan GC: Dietary factors and the risk of incident kidney stones in men: New insights after 14 years of follow-up. J Am Soc Nephrol 15: 3225–3232, 2004 [DOI] [PubMed] [Google Scholar]
- 49.Taylor EN, Stampfer MJ, Curhan GC: Obesity, weight gain, and the risk of kidney stones. JAMA 293: 455–462, 2005 [DOI] [PubMed] [Google Scholar]