Abstract
Context
Abnormalities in calcium metabolism are common in chronic kidney disease (CKD). Diminished urinary calcium excretion may promote vascular calcification and increased urinary calcium excretion may lead to nephrolithiasis and nephrocalcinosis, conditions associated with CKD.
Objective
To study predictors of urinary calcium excretion and its association with adverse clinical outcomes in CKD.
Design, Setting and Patients
This study assessed 3768 nondialysis participants in the Chronic Renal Insufficiency Cohort study from April 2003 to September 2008. Participants were followed up to October 2018.
Exposure
Clinically plausible predictors of urinary calcium excretion and 24-h urinary calcium excretion at baseline.
Main Outcome Measures
Urinary calcium excretion; incident end stage kidney disease (ESKD), CKD progression [50% estimated glomerular filtration rate (eGFR) decline or incident ESKD], all-cause mortality, and atherosclerotic cardiovascular disease events.
Results
eGFR was positive correlated with 24-h urinary calcium excretion. The variables most strongly associated with 24-h urinary calcium excretion in males and females were 24-h urinary sodium (β = 0.19 and 0.28, respectively), serum parathyroid hormone (β = −0.22 and −0.20, respectively), loop diuretics (β = 0.36 and 0.26, respectively), thiazide diuretics (β = −0.49 and −0.53, respectively), and self-identified black race (β = −0.23 and −0.27, respectively). Lower urinary calcium excretion was associated with greater risks of adverse outcomes, but these associations were greatly attenuated or nullified after adjustment for baseline eGFR.
Conclusion
Urinary calcium excretion is markedly lower in individuals with CKD compared to the general population. Determinants of urinary calcium excretion differed between sexes and levels of CKD. Associations between urinary calcium excretion and adverse clinical events were substantially confounded by eGFR.
Keywords: urinary calcium excretion, determinants, chronic kidney disease, renal outcomes, all-cause mortality, cardiovascular events
Calcium homeostasis is regulated by an interplay of intestinal absorption, reabsorption in the kidneys, and exchange from the bone (1). The kidneys play a critical role in this process by modulating urinary calcium excretion. Urinary calcium excretion is determined by a variety of factors including diet, medications, extracellular volume and acid-base status, and calciotropic hormones such as parathyroid hormone (PTH), 1,25-dihydroxyvitamin D [1,25-(OH)2D], and sex hormones (2). Abnormalities in calcium metabolism are common in the setting of chronic kidney disease (CKD), a condition that affects over 9% of the population worldwide (3). Calcium homeostasis in the setting of CKD is complex. Serum calcium concentrations are generally maintained in the normal range until advanced stages of CKD, when levels decrease slightly as 1,25-D concentrations decrease. Typical 24-h urinary calcium excretion ranges between 100 to 300 mg/day in healthy individuals and declines markedly in advanced CKD, with reports of <80 mg/day in CKD stages 3 and 4 (4). Greatly diminished urinary calcium excretion may promote vascular calcification, which is a major comorbid condition in CKD. Conversely, higher urinary calcium excretion may lead to nephrolithiasis and nephrocalcinosis and could increase the risk of CKD development or progression. Increased urinary calcium excretion in certain tubular disorders such as Dent’s disease and familial hypomagnesemia is associated with increased risk of CKD (5,6), whereas in other conditions such as type 1 renal tubular acidosis, there appears to be no significantly increased risk of CKD (7).
To explore the factors that regulate urinary calcium excretion and test whether higher or lower urinary calcium excretion is associated with adverse kidney or vascular events in CKD, we studied participants enrolled in the Chronic Renal Insufficiency Cohort (CRIC) study, a prospective observational cohort study.
Materials and Methods
Study Cohort
The CRIC study is a multicenter, prospective, observational cohort study designed to investigate the risk factors for CKD progression, cardiovascular disease, and death among individuals with CKD (8). From April 8, 2003 through September 3, 2008 (Phase I), 3939 participants, 21 to 74 years old, were enrolled across 7 clinical centers in the United States, with an estimated glomerular filtration rate (eGFR) ranging from 20 to 70 mL/min/1.73 m2 (9). Exclusion criteria included the inability to provide written consent, institutionalization, enrollment in other research studies, pregnancy, New York Heart Association class III to IV heart failure, human immunodeficiency virus infection, cirrhosis, multiple myeloma, polycystic kidney disease, renal cancer, recent chemotherapy or immunosuppressive therapy, organ transplant, or prior treatment with dialysis for at least 1 month (8-11). The study protocol was approved by the institutional review board at each of the recruiting sites, and all participants provided written informed consent. Data and documentation files for these analyses were obtained from the National Institute of Diabetes and Digestive and Kidney Diseases Central Repository in December 2019, including study documentation, data collection forms, study data, and the data set integrity check for the recruitment and follow-up phases.
Exposure and Outcomes
The exposure of interest was 24-h urinary calcium excretion. Urinary calcium measurements from 24-h samples were available in 3768 participants at baseline. Completed timed urine collections with total volume less than 500 mL or total collection time less than 22 h or greater than 26 h were discarded and collections were performed again. All 24-h urinary measurements were calculated by multiplying the concentration by the urinary volume and then adjusting for the number of hours collected. CRIC study participants were followed up by telephone contact every 6 months and annual in-person visits. We selected the following outcomes of interest: incident end-stage kidney disease (ESKD), defined as the initiation of long-term kidney replacement therapy or kidney transplantation; CKD progression, defined as a composite of incident ESKD or 50% decline in eGFR; all-cause mortality; and atherosclerotic cardiovascular disease (ASCVD) events, including definite or probable myocardial infarction, stroke, or peripheral arterial disease events. ESKD and death were ascertained by cross-linkage with the US Renal Data System and the Social Security Death Index, respectively (12). Time to 50% decline in eGFR was imputed based on the assumption that eGFR changed linearly over time between visit intervals (13). Patient follow-up was censored at first occurrence of death (except for all-cause mortality), loss to follow-up, withdrawal, last clinic visit (only for CKD progression outcome), or cutoff date of October 2018, whichever came first. We also tested for associations of urinary calcium excretion with the presence of mitral annulus calcification (MAC), defined as MAC score > 0 (14), and low ankle brachial index (ABI), defined as ABI < 0.9 (15).
Covariates
Data collected at baseline included self-reported sociodemographic characteristics, medical history, and current intake of medications. Height, body weight, and blood pressure were measured using standard protocols. Twenty-four-hour urine parameters, including urine sodium, potassium, phosphate, calcium, albumin, and creatinine, were collected at baseline. The eGFR was calculated using the CRIC-derived estimating equation (16). We studied factors plausibly related to calcium metabolism, including calcification measurements (MAC and ABI), plasma and urine levels of calcium and phosphate, serum PTH, use of active vitamin D, phosphate binder and diuretics, and total dietary intakes of calcium, vitamin D and caffeine. Comprehensive metabolic panels and 24-h urinary analytes (calcium, phosphate, potassium, sodium and creatinine) were measured using standardized assays (17). Plasma PTH was measured using an intact PTH assay (18). For participants who underwent CT scans at year 1, cardiac calcific lesions were identified by an attenuation threshold of 130 Hounsfield units and a minimum of 3 contiguous pixels, and then MAC was scored using Agatston’s algorithm (14). ABI was obtained according to standard protocol as previously described (19). Dietary data were available for 2739 patients and determined by scoring individual food items from the National Cancer Institute Diet History Questionnaire (20).
Statistical Analysis
For continuous variables, we tested normality using Shapiro-Wilk test and expressed them as means [standard deviation (SD)] or medians [interquartile range (IQR)]. Nonnormally distributed variables were logarithmically transformed for parametric tests. Categorical variables were presented as count with percentages. Baseline characteristics were compared across quintiles of 24-h urinary calcium excretion using chi-square test for categorical variables and analysis of variance and Kruskal-Wallis tests for normally and nonnormally distributed continuous variables, respectively. For skewed data distributions, we performed natural logarithmic transformation.
To test the associations between 24-h urinary calcium excretion and its determinants, we used Pearson or Spearman correlation coefficients as appropriate. We then fitted univariable and multivariable-adjusted linear regression models using natural log-transformed urinary calcium excretion as the dependent variable. We first included all potential predictors into the model, followed by automated variable selection procedures. Continuous independent variables were scaled to SD. The final model was chosen by backward selection based on the smallest Akaike information criterion value. We used the variance inflation factor to test for multicollinearity. Because sex and eGFR significantly interacted with other covariates, we conducted stratified analyses by eGFR (<45 vs ≥45 mL/min/1.73 m2) and by males and females separately. To take model selection uncertainty into account, we additionally performed Bayesian model averaging and estimated posterior effect probabilities (PEP) for all the initially included covariates to ensure robustness of the effect estimates of independent variables (21). We finally kept those covariates identified by backward selection and with PEP ≥85%, as significant predictors with very strong evidence for association. Since 2282 patients had complete data on variables for multilinear analysis, we performed multiple imputation for missing covariates using fully conditional specification (22).The mean (SD) of imputed variables were used to check data convergence and validity.
We performed time-to-event analyses to examine the risk of the outcomes according to 24-h urinary calcium excretion as both a continuous variable (natural log-transformed) and as quintiles. We used complete case analysis as there were less than 5% missing data on variables for time-to-event analyses. Due to violation of proportional hazard assumption, the associations of 24-h urinary calcium excretion with time-to-event outcomes were assessed using weighted Cox regression (23) and accelerated failure time (AFT) models, an alternative strategy for survival data analysis that estimates median survival time ratios (24). The R packages used to perform weighted Cox regression and AFT regression were “coxphw” (25) and “flexsurv” (26), respectively. We confirmed no significant violations of the linearity assumption using restricted cubic splines (27).
The adjustment strategy for weighted Cox and AFT regression models was based on biological and clinical plausibility of covariates as potential confounders of the association between 24-h urinary calcium excretion and outcomes. Model 1 is adjusted for age, sex, self-reported race (black vs nonblack), clinical sites, body mass index (BMI), systolic blood pressure, diabetes, prior cardiovascular diseases (CVD; any history of coronary artery disease, prior revascularization, heart failure, stroke, or peripheral artery disease), and 24-h urinary creatinine. Model 2 includes Model 1 and further adjusts for eGFR. Model 3 includes Model 2 and further adjusts for medication use (loop diuretics, thiazide diuretics, phosphate binder, active Vitamin D, angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker, beta blocker, statins and antiplatelet), hemoglobin, serum albumin, and total PTH. We further performed exploratory subgroup analyses based on biological plausibility and findings from linear regression analyses. For sensitivity analyses, we additionally adjusted models for 24-h urinary albumin, hypothesizing it could also function as a mediator or collider of CKD progression (28). A 2-sided P-value < 0.05 was considered statistically significant. Statistical analyses were performed using R software version 3.6.2.
Results
Baseline Characteristics
Baseline characteristics of the study population are listed in Table 1 by quintiles of 24-h urinary calcium excretion. Median urinary calcium excretion per 24 h was 37.9 mg (IQR 16.3-84.6 mg) in women and 39.4 mg (IQR 18.0-85.3 mg) in men. Urinary calcium excretion was substantially lower in those with more advanced CKD (Fig. 1A) and was positively correlated with eGFR in both women (r = 0.42, P < 0.001) and men (r = 0.41, P < 0.001) (Fig. 1B). By CKD stage, median urinary calcium excretion was 172.8 mg (IQR 108.7-229.5 mg) in stage 1 (eGFR ≥ 90 mL/min/1.73 m2), 86.4 mg (IQR 43.5-153.9 mg) in stage 2 (eGFR 60-89 mL/min/1.73 m2), 49.1 mg (IQR 23.0-100.9 mg) in stage 3a (eGFR 45 to 59 mL/min/1.73 m2), 31.5 mg (IQR, 14.5-62.6 mg) in stage 3b (eGFR 30 to 44 mL/min/1.73 m2), and 23.4 mg (IQR 11.9-41.5 mg) in stages 4 and 5 CKD (eGFR < 30 mL/min/1.73 m2).
Table 1.
Characteristics of Chronic Renal Insufficiency Cohort study participants by quintiles of 24-h urinary calcium excretiona
| Characteristics | All (n = 3768) | Quintile 1 (n = 754) | Quntile 2 (n = 754) | Quintile 3 (n = 753) | Quintile 4 (n = 755) | Quintile 5 (n = 752) | P for trend |
|---|---|---|---|---|---|---|---|
| Urinary calcium excretion, mg/24 h | 38.8 (17.2, 84.8) | 0.56-14.1 | 14.2-29.9 | 30.0-52.8 | 52.9-102.1 | 102.4-696.7 | — |
| Age, years | 58.3 (10.9) | 57.7 (11.8) | 58.8 (11.2) | 59.2 (10.4) | 58.9 (10.4) | 57.0 (10.6) | <0.001 |
| Female, n (%) | 1704 (45.2) | 363 (48.1) | 340 (45.1) | 329 (43.7) | 345 (45.7) | 327 (43.5) | 0.367 |
| Race | |||||||
| White, n (%) | 1600 (42.5) | 182 (24.1) | 255 (33.8) | 327 (43.4) | 376 (49.8) | 460 (61.2) | <0.001 |
| Black, n (%) | 1565 (41.5) | 453 (60.1) | 330 (43.8) | 297 (39.4) | 258 (34.2) | 227 (30.2) | |
| Hispanic, n (%) | 455 (12.1) | 100 (13.3) | 129 (17.1) | 100 (13.3) | 86 (11.4) | 40 (5.3) | |
| Other, n (%) | 148 (3.9) | 19 (2.5) | 40 (5.3) | 29 (3.9) | 35 (4.6) | 25 (3.3) | |
| BMI, kg/m2 | 32.1 (7.8) | 32.0 (8.4) | 31.9 (7.9) | 31.9 (7.4) | 32.2 (7.6) | 32.5 (7.7) | 0.554 |
| Systolic BP, mmHg | 128.2 (21.9) | 131.1 (22.2) | 130.7 (22.8) | 128.5 (21.7) | 127.2 (21.1) | 123.6 (21.1) | <0.001 |
| Diabetes, n (%) | 1820 (48.3) | 426 (56.5) | 426 (56.5) | 383 (50.9) | 327 (43.3) | 258 (34.3) | <0.001 |
| Any CVD, n (%) | 1262 (33.5) | 253 (33.6) | 267 (35.4) | 284 (37.7) | 246 (32.6) | 212 (28.2) | 0.002 |
| MI or revascularization, n (%) | 835 (22.2) | 147 (19.5) | 180 (23.9) | 187 (24.8) | 169 (22.4) | 152 (20.2) | 0.057 |
| CHF, n (%) | 364 (9.7) | 60 (8.0) | 81 (10.7) | 86 (11.4) | 73 (9.7) | 64 (8.5) | 0.119 |
| PVD, n (%) | 256 (6.8) | 56 (7.4) | 55 (7.3) | 68 (9.0) | 52 (6.9) | 25 (3.3) | <0.001 |
| Stroke, n (%) | 374 (9.9) | 90 (11.9) | 86 (11.4) | 69 (9.2) | 76 (10.1) | 53 (7.0) | 0.013 |
| Current Smoking, n (%) | 490 (13) | 117 (15.5) | 93 (12.3) | 119 (15.8) | 92 (12.2) | 69 (9.2 | <0.001 |
| Drinker (≥once/week), n (%) | 762 (20.2) | 100 (13.3) | 132 (17.5) | 134 (17.8) | 169 (22.4) | 227 (30.2) | <0.001 |
| ABI < 0.9, n (%) | 597 (16.1) | 136 (18.3) | 151 (20.4) | 137 (18.6) | 111 (14.9) | 62 (8.3) | <0.001 |
| MAC, n (%) | 484 (16.5) | 79 (13.7) | 89 (15.7) | 91 (16.2) | 114 (18.6) | 111 (17.8) | 0.029 |
| SCr, mg/dL | 1.8 (0.6) | 2.1 (0.7) | 2.0 (0.6) | 1.9 (0.6) | 1.7 (0.6) | 1.5 (0.5) | <0.001 |
| eGFR, mL/min/1.73 m2 | 44.9 (16.7) | 37.2 (13.0) | 39.5 (13.6) | 42.5 (14.4) | 48.3 (16.2) | 57.1 (17.9) | <0.001 |
| ≥60, n (%) | 662 (17.6) | 38 (5.0) | 67 (8.9) | 78 (10.4) | 178 (23.6) | 301 (40.0) | <0.001 |
| 45-59, n (%) | 1059 (28.1) | 153 (20.3) | 172 (22.8) | 235 (31.2) | 240 (31.8) | 259 (34.4) | |
| 30-44, n (%) | 1286 (34.1) | 313 (41.5) | 310 (41.1) | 273 (36.3) | 233 (30.9) | 157 (20.9) | |
| <30, n (%) | 761 (20.2) | 250 (33.1) | 205 (27.2) | 167 (22.2) | 104 (13.8) | 35 (4.7) | |
| Serum hemoglobin, g/dL | 12.6 (1.8) | 11.9 (1.7) | 12.2 (1.7) | 12.5 (1.7) | 12.9 (1.6) | 13.6 (1.6) | <0.001 |
| Serum albumin, g/dL | 4.0 (0.5) | 3.8 (0.5) | 3.9 (0.5) | 3.9 (0.5) | 4.0 (0.4) | 4.1 (0.4) | <0.001 |
| Serum calcium, mg/dL | 9.2 (0.5) | 9.2 (0.5) | 9.3 (0.4) | 9.2 (0.5) | 9.2 (0.5) | 9.2 (0.5) | 0.229 |
| Serum phosphate, mg/dL | 3.7 (0.7) | 3.9 (0.7) | 3.8 (0.7) | 3.8 (0.7) | 3.7 (0.6) | 3.5 (0.6) | <0.001 |
| Serum PTH, pg/mL | 54.0 (34.9, 89.0) | 78.9 (50.1, 123.8) | 62.0 (41.0, 103.0) | 53.7 (36.1, 84.0) | 45.0 (30.6, 72.0) | 37.7 (26.3, 60.3) | <0.001 |
| Urinary creatinine, g/24 h | 1.3 (0.6) | 1.2 (0.6) | 1.3 (0.6) | 1.3 (0.5) | 1.4 (0.6) | 1.6 (0.6) | <0.001 |
| Urinary albumin, mg/24 h | 64.0 (10.1, 560.4) | 204.3 (25.1, 1077.5) | 137.9 (18.7, 819.8) | 78.9 (11.2, 634.1) | 34.0 (7.5, 311.8) | 19.4 (7.0, 131.8) | <0.001 |
| Urinary sodium, mg/24 h | 161.1 (77.5) | 132.3 (63.4) | 146.8 (69.2) | 159.4 (67.9) | 174.2 (76.6) | 192.7 (92.3) | <0.001 |
| Urinary potassium, mg/24 h | 51.1 (37.3, 68.5) | 41.6 (29.5, 58.4) | 47.3 (34.8, 62.5) | 50.4 (37.0, 67.0) | 55.5 (41.2, 71.9) | 62.6 (46.3, 80.4) | <0.001 |
| Urinary phosphate, mg/24 h | 755.0 (351.3) | 612.1 (286.4) | 697.3 (302.8) | 748.4 (322.0) | 785.7 (343.5) | 949.9 (396.6) | <0.001 |
| Medications | |||||||
| ACEi/ARB, n (%) | 1171 (31.3) | 565 (75.6) | 562 (74.9) | 524 (70.0) | 494 (65.8) | 428 (57.3) | <0.001 |
| β blocker, n (%) | 1843 (49.2) | 409 (54.8) | 373 (49.7) | 373 (49.8) | 371 (49.4) | 317 (42.4) | <0.001 |
| Loop diuretics, n (%) | 1419 (37.9) | 281 (37.6) | 303 (40.4) | 312 (41.7) | 278 (37.0) | 245 (32.8) | 0.005 |
| Thiazide diuretics, n (%) | 2230 (59.6) | 516 (69.1) | 470 (62.7) | 469 (62.6) | 414 (55.1) | 361 (48.3) | <0.001 |
| Statins, n (%) | 2071 (55.3) | 437 (58.5) | 445 (59.3) | 429 (57.3) | 412 (54.9) | 348 (46.6) | <0.001 |
| Antiplatelets, n (%) | 1735 (46.3) | 327 (43.8) | 382 (50.9) | 346 (46.2) | 338 (45.0) | 342 (45.8) | 0.062 |
| Phosphate binder, n (%) | 271 (7.2) | 56 (7.5) | 52 (6.9) | 52 (6.9) | 57 (7.6) | 54 (7.2) | 0.981 |
| Ca-containing, n (%) | 259 (95.6) | 50 (89.3) | 49 (94.2) | 49 (94.2) | 57 (100.0) | 54 (100.0) | 0.026 |
| Non-Ca-containing, n (%) | 14 (4.4) | 6 (10.7) | 3 (5.8) | 3 (5.8) | 0 (0.0) | 0 (0.0) | |
| Active vitamin D supplements, n (%) | 119 (3.2) | 35 (4.7) | 25 (3.3) | 19 (2.5) | 22 (2.9) | 18 (2.4) | 0.085 |
| Diet intakes (n = 2739) | |||||||
| Total calcium, mg/day | 751.0 (495.5, 1117.6) | 682.4 (429.7, 992.6) | 686.9 (483.2, 1019.6) | 732.0 (485.1, 1102.1) | 810.6 (510.3, 1208.5) | 836.7 (559.8, 1286.33) | <0.001 |
| Total vitamin D, IU/day | 270.7 (118.4, 507.8) | 188.7 (95.1, 464.4) | 254.8 (112.4, 496.0) | 241.2 (114.4, 502.7) | 329.6 (126.0, 523.8) | 408.3 (149.0, 535.6) | <0.001 |
| Caffeine, mg/day | 162.0 (30.6, 407.7) | 131.6 (28.7, 334.8) | 136.2 (24.2, 403.8) | 147.7 (28.7, 364.2) | 193.9 (33.7, 536.6) | 207.1 (45.7, 477.1) | 0.005 |
aValues for continuous variables are presented as mean (SD), median (IQR, or range). All baseline variable values were from year 0, except for MAC, which is from year 1. Total calcium and vitamin D intake include daily dietary calcium/vitamin D intake plus calcium/vitamin D from supplements. To convert creatinine in milligram per deciliter to micromoles per liter, multiply by 88.4. To convert urinary albumin-to-creatinine ratio in milligram per gram to milligrams per millimoles, divide by 8.84. To convert albumin in gram per deciliter to gram per liter, multiply by 10. To convert phosphate in milligram per deciliter to millimoles per liter, multiply by 0.323. To convert calcium in milligram per deciliter to millimoles per liter, multiply by 0.25.
Abbreviations: ABI, ankle brachial index; ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; BP, blood pressure; CHF, chronic heart failure; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; IQR, interquartile range; MAC, mitral annulus calcification; MI, myocardial infarction; PTH, parathyroid hormone; PVD, peripheral vascular disease; SCr, serum creatinine; SD, standard deviation.
Figure 1.
Correlation between 24-h urinary calcium excretion and estimated glomerular filtration rate (eGFR). (A) 24-h urinary calcium excretion distributions across different chronic kidney disease stages. (B) Correlations between eGFR and natural log of 24-h urinary calcium excretion in males and females, respectively.
Predictors of Urinary Calcium Excretion
Because sex and eGFR potentially modify associations between covariates and 24-h urinary calcium excretion (29), we tested for statistical interactions of sex and eGFR with other covariates respectively. Sex significantly interacted with BMI (P = 0.003), systolic blood pressure (P = 0.030), beta-blocker use (P = 0.030), and total dietary vitamin D intake (P = 0.017). eGFR significantly interacted with PTH (P < 0.001), 24-h urinary sodium (P = 0.009), and serum phosphate (P = 0.043). Results from stratified (by sex and eGFR) multivariable adjusted linear regression models are shown in Table 2. In both men and women, in both strata of eGFR (<45 and >45 mL/min/1.73 m2, respectively), the following variables were significantly associated with higher urinary calcium excretion (Table 2): 24-h urinary sodium excretion, loop diuretic use, thiazide diuretic (negatively associated), PTH levels, and self-reported black race. Other variables that were significant in at least 3 of 4 subgroups included 24-h urinary phosphate excretion, angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker use (negatively), and serum albumin concentration. BMI was associated with higher urinary calcium excretion only in men. In both backward selection models and Bayesian model averaging, neither dietary calcium nor vitamin D intake was associated with 24-h urinary calcium excretion.
Table 2.
Significant predictors of 24-h urinary calcium excretion with strong evidencea
| Male | Female | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variables | All | eGFR < 45 | eGFR ≥ 45 | All | eGFR < 45 | eGFR ≥ 45 | ||||||||||||
| β | % change | P-values | β | % change | P-values | β | % change | P-values | β | % change | P-values | β | % change | P-values | β | % change | P-values | |
| 24-h urinary sodium | 0.19 | 20.4 | <0.001 | 0.21 | 23.9 | <0.001 | 0.16 | 16.8 | <0.001 | 0.28 | 31.8 | <0.001 | 0.38 | 46.2 | <0.001 | 0.16 | 17.9 | <0.001 |
| Loop diuretic | 0.36 | 43.1 | <0.001 | 0.31 | 35.9 | <0.001 | 0.46 | 58.9 | <0.001 | 0.26 | 29.5 | <0.001 | 0.22 | 24.2 | 0.001 | 0.38 | 46.7 | <0.001 |
| Thiazide diuretic | –0.49 | –38.7 | <0.001 | –0.55 | –42.6 | <0.001 | –0.45 | –36.0 | <0.001 | –0.53 | –40.9 | <0.001 | –0.50 | –39.2 | <0.001 | –0.54 | –41.6 | <0.001 |
| ln (PTH) | –0.22 | –19.7 | <0.001 | –0.17 | –15.3 | <0.001 | –0.35 | –29.6 | <0.001 | –0.20 | –18.3 | <0.001 | –0.23 | –20.2 | <0.001 | –0.32 | –27.1 | <0.001 |
| Self-identified black | –0.23 | –20.2 | <0.001 | –0.20 | –18.1 | <0.001 | –0.26 | –23.2 | <0.001 | –0.27 | –23.9 | <0.001 | –0.25 | –22.1 | <0.001 | –0.33 | –28.1 | <0.001 |
| 24-h urinary phosphate | 0.14 | 15.4 | <0.001 | 0.15 | 15.6 | <0.001 | 0.14 | 14.8 | <0.001 | 0.14 | 14.6 | <0.001 | — | 0.21 | 23.5 | <0.001 | ||
| ACEi/ARB | –0.27 | –24.0 | <0.001 | –0.30 | –26.0 | <0.001 | –0.26 | –22.9 | <0.001 | –0.22 | –19.5 | <0.001 | –0.24 | –21.5 | <0.001 | — | ||
| Serum albumin | 0.09 | 9.4 | <0.001 | – | 0.11 | 11.9 | <0.001 | 0.14 | 15.4 | <0.001 | 0.12 | 12.9 | 0.001 | 0.17 | 18.0 | <0.001 | ||
| eGFR | 0.29 | 34.2 | <0.001 | – | 0.28 | 32.5 | <0.001 | 0.28 | 32.8 | <0.001 | — | 0.30 | 35.0 | <0.001 | ||||
| ln (24-h urinary albumin) | — | –0.08 | –8.1 | 0.012 | — | — | –0.11 | –10.7 | 0.002 | — | ||||||||
| ln (24-h urinary potassium) | — | 0.09 | 9.7 | 0.006 | — | — | — | — | ||||||||||
| BMI | 0.11 | 11.4 | <0.001 | — | 0.13 | 14.1 | <0.001 | — | — | — | ||||||||
| Diabetes | –0.17 | –15.7 | <0.001 | — | –0.25 | –22.4 | <0.001 | — | — | — | ||||||||
| Systolic blood pressure | — | — | 0.11 | 11.9 | 0.001 | — | — | — | ||||||||||
| Serum calcium | — | — | – | 0.10 | 10.1 | <0.001 | — | 0.17 | 18.5 | <0.001 |
aTwenty-four-hour urinary calcium excretion and other nonnormally distributed variables were natural log-transformed. β coefficients and P-values were from multilinear regression where 24-h urinary calcium as dependent variable was natural log-transformed. Exp(β)-1 can be interpreted as the percentage difference in urinary calcium excretion. To account for model uncertainty, we used Bayesian model averaging to provide evidence degrees for covariate effects on urinary calcium excretion. PEP value of covariate Xn can be interpreted as the probability of β n not being 0, based on the averaged posterior distribution associated with Xn. PEP ≥ 85% is considered as very strong evidence for an effect. Only the covariates identified by backward selection and with PEP ≥ 85% are presented here.
Abbreviations: ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; PEP, posterior effect probability; eGFR, estimated glomerular filtration rate; PEP, posterior effect probability; PTH, parathyroid hormone.
Calcium Excretion and Evidence of Cardiovascular Calcification
Measurements of cardiovascular calcification in this study were ABI and MAC. The proportion of individuals with low ABI was higher in higher CKD stages, from 2.0% in CKD stage 1 to 26.7% in CKD stages 4 and 5. The proportion of individuals with MAC was relatively stable across CKD stages (Fig. 2A). We found an inverse association between 24-h urinary calcium excretion and ABI and the opposite trend for MAC (Fig. 2B). In multivariable linear regression for predictors of urinary calcium excretion, both ABI and MAC were excluded by backward selection, with low PEP of a nonzero coefficient.
Figure 2.
Features of calcification measurements across chronic kidney disease (CKD) stages and quintiles of 24-h urinary calcium excretion. (A) Percentage of mitral annulus calcification (MAC) and ankle brachial index (ABI) <0.9 across CKD stages. (B) Percentage of MAC and ABI <0.9 across quintiles of 24-h urinary calcium excretion.
Associations of Calcium Excretion With Clinical Outcomes
A total of 1057 ESKD events (3.48/100 person-years), 1334 CKD progression events (5.73/100 person-years), 1268 all-cause death events (3.55/100 person-years), and 714 ASCVD events (2.33/100 person-years) were observed during follow-up. Generally, we found lower crude event rates in higher quintiles of 24-h urinary calcium excretion (Table 3). Unadjusted and multivariable-adjusted hazard ratios (HRs) with 95% CIs for all the outcomes are presented in Table 3 according to baseline 24-h urinary calcium excretion modeled as a continuous variable and by quintiles. In unadjusted analysis, every 1-unit increase in natural log of 24-h urinary calcium excretion was significantly associated with lower risk of ESKD [HR 0.65, 95% CI (0.62, 0.69); P < 0.001], CKD progression [HR 0.69, 95% CI (0.66, 0.73); P < 0.001], all-cause death [HR 0.83, 95% CI (0.79, 0.88); P < 0.001], and ASCVD [HR 0.83, 95% CI (0.78, 0.89); P < 0.001]. Adjustment for demographics, clinic centers, diabetes, systolic blood pressure, BMI, CVD and 24-h urinary creatinine (Model 1) mildly attenuated the associations between urinary calcium excretion and all the outcomes. eGFR (Model 2) significantly confounded the associations between urinary calcium excretion and each outcome by attenuating associations for ESKD and CKD progression and nullifying associations for all-cause death (P = 0.378) and ASCVD (P = 0.915). After further adjustments for laboratory data and medication usage (Model 3), all associations between urinary calcium excretion and outcomes became statistically nonsignificant. Results from AFT and competing risk analysis were similar. Adjusting for 24-h urinary albumin or subgroup analysis did not reveal significant associations (data not shown).
Table 3.
Associations of 24-h urinary calcium with risk of end-stage kidney disease, chronic kidney disease progression, all-cause mortality, and atherosclerotic cardiovascular disease
| Outcomes | Continuous | Quintile 1 | Quintile 2 | Quintile 3 | Quintile 4 | Quintile 5 |
|---|---|---|---|---|---|---|
| ESKD | ||||||
| Events, n | 1057 | 317 | 287 | 209 | 154 | 89 |
| Event rates (per 100 person-years) | 3.48 | 6.31 | 5.19 | 3.55 | 2.32 | 1.22 |
| Crude | 0.65 (0.62, 0.69) | 1 (Reference) | 0.80 (0.68, 0.95) | 0.56 (0.46, 0.67) | 0.38 (0.31, 0.46) | 0.20 (0.16, 0.26) |
| Multivariable Model 1a | 0.71 (0.67, 0.76) | 1 (Reference) | 0.79 (0.66, 0.95) | 0.63 (0.52, 0.76) | 0.45 (0.37, 0.56) | 0.29 (0.22, 0.37) |
| Multivariable Model 2b | 0.90 (0.84, 0.97) | 1 (Reference) | 0.90 (0.75, 1.08) | 0.82 (0.67, 1.00) | 0.80 (0.64, 1.00) | 0.70 (0.54, 0.92) |
| Multivariable Model 3c | 0.96 (0.89,1.04) | 1 (Reference) | 0.94 (0.78, 1.14) | 0.87 (0.70, 1.09) | 0.94 (0.74, 1.19) | 0.83 (0.62, 1.11) |
| CKD progression (ESRD or 50% eGFR decline) | ||||||
| Events, n | 1334 | 365 | 342 | 268 | 215 | 143 |
| Event rates (per 100 person-years) | 5.73 | 9.55 | 8.34 | 6.02 | 4.15 | 2.49 |
| Crude | 0.69 (0.66, 0.73) | 1 (Reference) | 0.87 (0.74, 1.01) | 0.62 (0.52, 0.73) | 0.44 (0.37, 0.53) | 0.27 (0.22, 0.33) |
| Multivariable Model 1a | 0.76 (0.72, 0.80) | 1 (Reference) | 0.89 (0.75, 1.06) | 0.70 (0.58, 0.84) | 0.52 (0.43, 0.64) | 0.40 (0.32, 0.49) |
| Multivariable Model 2b | 0.94 (0.88, 1.00) | 1 (Reference) | 0.96 (0.81, 1.14) | 0.87 (0.72, 1.05) | 0.84 (0.69, 1.02) | 0.83 (0.66, 1.04) |
| Multivariable Model 3c | 0.99 (0.93, 1.06) | 1 (Reference) | 1.02 (0.85, 1.21) | 0.96 (0.80, 1.17) | 0.97 (0.79, 1.19) | 0.94 (0.73, 1.20) |
| All-cause mortality | ||||||
| Events, n | 1268 | 304 | 285 | 289 | 213 | 177 |
| Event rates (per 100 person-years) | 3.55 | 4.45 | 4.07 | 4.23 | 2.91 | 2.31 |
| Crude | 0.83 (0.79, 0.88) | 1 (Reference) | 0.94 (0.79, 1.12) | 0.95 (0.80, 1.12) | 0.64 (0.53, 0.76) | 0.57 (0.44, 0.73) |
| Multivariable Model 1a | 0.92 (0.86, 0.99) | 1 (Reference) | 0.88 (0.74, 1.06) | 0.95 (0.79, 1.13) | 0.69 (0.57, 0.85) | 0.80 (0.60, 1.07) |
| Multivariable Model 2b | 1.03 (0.96, 1.11) | 1 (Reference) | 0.91 (0.77, 1.08) | 1.08 (0.91, 1.27) | 0.88 (0.72, 1.06) | 1.19 (0.93, 1.54) |
| Multivariable Model 3c | 1.05 (0.98, 1.13) | 1 (Reference) | 0.94 (0.78, 1.12) | 1.11 (0.93, 1.32) | 0.92 (0.75, 1.13) | 1.27 (0.98, 1.66) |
| ASCVD (including MI, stroke, or PAD) | ||||||
| Events, n | 714 | 157 | 176 | 147 | 137 | 97 |
| Event rates (per 100 person-years) | 2.33 | 2.68 | 3.04 | 2.52 | 2.19 | 1.42 |
| Crude | 0.83 (0.78, 0.89) | 1 (Reference) | 1.14 (0.91, 1.44) | 0.99 (0.78, 1.26) | 0.83 (0.65, 1.06) | 0.49 (0.38, 0.65) |
| Multivariable Model 1a | 0.93 (0.86, 1.00) | 1 (Reference) | 1.05 (0.83, 1.34) | 0.98 (0.77, 1.26) | 0.93 (0.72, 1.20) | 0.71 (0.53, 0.95) |
| Multivariable Model 2b | 1.01 (0.93, 1.09) | 1 (Reference) | 1.08 (0.85, 1.37) | 1.08 (0.84, 1.38) | 1.09 (0.84, 1.42) | 0.91 (0.67, 1.25) |
| Multivariable Model 3c | 1.03 (0.95, 1.13) | 1 (Reference) | 1.13 (0.89, 1.45) | 1.14 (0.88, 1.47) | 1.18 (0.89, 1.56) | 0.99 (0.71, 1.38) |
Abbreviations: ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; ASCVD, atherosclerotic cardiovascular disease; BMI, body mass index; BP, blood pressure; CCID, clinical center identification; CKD, chronic kidney disease; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; ESKD, end-stage kidney disease; MI, myocardial infarction; PAD, peripheral arterial disease; PTH, parathyroid hormone.
aModel 1 = stratified by CCID, adjusted for age, gender, race, CCID, BMI, diabetes, systolic BP, CVD, and 24-h urinary creatinine.
bModel 2 = Model 1 + adjusted for eGFR at baseline.
cModel 3 = Model 2 + adjusted for lab tests at baseline including hemoglobin, serum albumin, ln (PTH), and medication use at baseline including loop diuretics, thiazide diuretics, phosphate binder, active vitamin D, ACEi/ARB, beta blocker, statins, and antiplatelet.
Discussion
Our findings confirm and expand upon the finding of markedly lower urinary calcium excretion in individuals with CKD compared to those with normal kidney function (17,30,31). Our cross-sectional results show the magnitude of the progressive decrease in urinary calcium excretion across progressive stages of CKD, from median of 172.8 mg in CKD stage 1 to 23.4 mg in CKD stages 4 and 5.
In a randomized controlled trial of 55 adults with CKD stages 3 or 4, Coburn et al recorded mean baseline levels of calcium excretion of <30 mg/24 h (31). The mechanisms for disordered calcium metabolism in CKD include diminished calcitriol production, leading to decreased gastrointestinal absorption and bone reabsorption, and higher PTH, leading to increased calcium reabsorption. The detailed clinical, laboratory, and dietary data available in CRIC allow us to make several observations about urinary calcium excretion in the setting of CKD.
We found that higher urinary sodium excretion was positively associated with higher urinary calcium excretion in the setting of CKD, as has been reported in studies in the general population (32,33). Restricted sodium intake is a key treatment for calcium-containing kidney stones to decrease urinary calcium excretion. Sodium and calcium reabsorption are linked in the proximal tubule, which may account for our findings. Vaidya et al conducted mechanistic and epidemiologic studies to clarify the associations between higher urinary sodium excretion (a proxy for higher sodium intake) and urinary calcium excretion (34). They found that sodium-induced volume expansion decreased renin-angiotensin-aldosterone system (RAAS) activity and stimulated higher calcium excretion, and they provided additional evidence on the relationship between RAAS and calcium physiology. In their patient-oriented physiology study, going from a restricted to ad lib sodium intake suppressed renin activity and plasma aldosterone but increased calcium excretion (34). The investigators hypothesized that excess sodium intake can lead to increased calcium excretion both by increasing sodium and calcium delivery to the proximal tubule and by suppressing the RAAS, which may have a direct effect on lowering PTH levels.
In addition to the consistency of our finding on the relationship between urinary sodium and calcium excretion, several other findings have biological plausibility. We found that loop diuretics were associated with higher calcium excretion and thiazide diuretics were associated with lower calcium excretion. This is consistent with the known effects of loop vs distal diuretics on calcium handling by the kidney (35,36). We also found that higher PTH levels were associated with lower 24-h urinary excretion of calcium, again consistent with the known effects of PTH on calcium metabolism (37). We found no differences in calcium excretion according to treatment with vitamin D, calcium intake, or phosphate binders.
The findings of lower calcium excretion in CKD raises the question of whether individuals with CKD are in positive calcium balance or whether net intestinal absorption is lower and thereby accounts for lower calcium excretion. There are limited studies addressing calcium balance in individuals with CKD. In 1 study involving 6 healthy controls and 6 individuals with CKD stage 4 or 5, Spiegel and Brady found that subjects with CKD were in negative-to-neutral calcium balance on a daily intake of 800 mg calcium but in marked positive balance on 2000 mg/day (4). A similar finding was also reported by Hill et al that 1500 mg elemental calcium from supplements per day leads to positive calcium balance in CKD (38). These 2 studies also reported that urinary calcium excretion remained low or nearly unchanged after calcium intake more than doubled. CRIC participants had a median daily calcium intake of 751 mg, but we lacked information on fecal excretion. Within the limitations of a prospective cohort study not designed primarily to test calcium metabolism, we found no association between calcium intake and 24-h urinary calcium excretion.
Calcification is a common feature of advanced CKD. A prior CRIC study reported that decreased eGFR was associated with higher risk of peripheral artery disease (39) and the presence of MAC (14). Although eGFR and 24-h urinary calcium excretion were highly correlated, we found that neither ABI nor MAC was associated with urinary calcium excretion after multivariable adjustment. The mechanisms involving calcification of blood vessels and tissues in CKD are more complex than net calcium balance.
We found strong associations between lower urinary calcium excretion and adverse clinical events, but these associations were entirely attenuated when adjusting for eGFR and other confounders such as medications and laboratory results. The nonsignificant findings of associations between 24-h urinary calcium excretion and outcomes of CKD patients are in concordance with other published results. Taylor et al reported that lower 24-h urinary calcium excretion was associated with CKD incidence among the general population but no significant association for all-cause mortality or CVD events (40). In patients with stable coronary artery disease, Welles et al. reported that 24-h urinary calcium excretion was associated with neither CVD event nor mortality (41).
To the best of our knowledge, this is the first study exploring determinants of 24-h urinary calcium excretion and its prospective associations with adverse clinical events in individuals with CKD. We utilized a large prospective cohort study with adjudicated follow-up outcomes and a rich repository of clinical details to explore determinants and outcomes associated with 24-h urinary calcium excretion in different strata based on adjustment for several calcium metabolism related measurements. Our study has several limitations. First, the CRIC data set does not include repeated measures of 24-h urine calcium, which would improve estimation of calcium excretion. We did not have complete data on dietary intake, and we also lacked information on serum 25(OH)D and 1,25(OH)2D values. We also did not have information on clinical factors such as history of nephrolithiasis, coronary artery calcification, and other clinical outcomes associated with vascular calcification.
In summary, 24-h urinary calcium excretion is markedly lower in individuals with CKD, especially in those with more advanced CKD, compared to those with intact kidney function. The associations of urinary calcium excretion with adverse clinical outcomes were confounded by adjustment of eGFR, suggesting that calcium excretion is a feature of, rather than a risk factor for, CKD progression and CKD-related complications.
Acknowledgments
Financial Support: T.O.I. is funded by the National Institute of Diabetes and Digestive and Kidney Diseases, K23DK119542. I.M.S. is supported by the American Philosophical Society Daland Fellowship in Clinical Investigation. S.S.W.’s support includes National Institute of Diabetes and Digestive and Kidney Diseases R25DK128858, UH3DK114915, R21DK119751, and U01DK085660.
Additional Information
Disclosures: The authors declare that there are no conflicts of interest relevant to this article.
Data Availability
Some or all data generated or analyzed during this study are included in this published article or in the data repositories listed in References.
References
- 1. Peacock M. Calcium metabolism in health and disease. Clin J Am Soc Nephrol. 2010;5(suppl 1):S23-S30. [DOI] [PubMed] [Google Scholar]
- 2. Bonny O, Rubin A, Huang CL, Frawley WH, Pak CY, Moe OW. Mechanism of urinary calcium regulation by urinary magnesium and pH. J Am Soc Nephrol. 2008;19(8):1530-1537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. GBD Chronic Kidney Disease Collaboration. Global, regional, and national burden of chronic kidney disease, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2020;395(10225):709-733. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Spiegel DM, Brady K. Calcium balance in normal individuals and in patients with chronic kidney disease on low- and high-calcium diets. Kidney Int. 2012;81(11):1116-1122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Godron A, Harambat J, Boccio V, et al. Familial hypomagnesemia with hypercalciuria and nephrocalcinosis: phenotype-genotype correlation and outcome in 32 patients with CLDN16 or CLDN19 mutations. Clin J Am Soc Nephrol. 2012;7(5):801-809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Scheinman SJ. X-linked hypercalciuric nephrolithiasis: clinical syndromes and chloride channel mutations. Kidney Int. 1998;53(1):3-17. [DOI] [PubMed] [Google Scholar]
- 7. Vallés PG, Batlle D. Hypokalemic distal renal tubular acidosis. Adv Chronic Kidney Dis. 2018;25(4):303-320. [DOI] [PubMed] [Google Scholar]
- 8. Feldman HI, Appel LJ, Chertow GM, et al. ; Chronic Renal Insufficiency Cohort (CRIC) Study Investigators . The chronic renal insufficiency cohort (CRIC) study: design and methods. J Am Soc Nephrol. 2003;14(7 suppl 2):S148-S153. [DOI] [PubMed] [Google Scholar]
- 9. Lash JP, Go AS, Appel LJ, et al. ; Chronic Renal Insufficiency Cohort (CRIC) Study Group . Chronic Renal Insufficiency Cohort (CRIC) study: baseline characteristics and associations with kidney function. Clin J Am Soc Nephrol. 2009;4(8):1302-1311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Isakova T, Xie H, Yang W, et al. ; Chronic Renal Insufficiency Cohort (CRIC) Study Group . Fibroblast growth factor 23 and risks of mortality and end-stage renal disease in patients with chronic kidney disease. JAMA. 2011;305(23):2432-2439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Dubin RF, Li Y, He J, et al. ; CRIC Study Investigators . Predictors of high sensitivity cardiac troponin T in chronic kidney disease patients: a cross-sectional study in the Chronic Renal Insufficiency Cohort (CRIC). BMC Nephrol. 2013;14:229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. US Renal Data System. USRDS 2013 Annual Data Report: Atlas of Chronic Kidney Disease and End-stage Renal Disease in the United States. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases; 2013. [Google Scholar]
- 13. Yang W, Xie D, Anderson AH, et al. ; CRIC Study Investigators . Association of kidney disease outcomes with risk factors for CKD: findings from the Chronic Renal Insufficiency Cohort (CRIC) study. Am J Kidney Dis. 2014;63(2):236-243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Abd Alamir M, Radulescu V, Goyfman M, Mohler ER 3rd, Gao YL, Budoff MJ; CRIC Study Investigators . Prevalence and correlates of mitral annular calcification in adults with chronic kidney disease: results from CRIC study. Atherosclerosis. 2015;242(1):117-122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Mostaza JM, Suarez C, Manzano L, et al. ; MERITO Study Group . Relationship between ankle-brachial index and chronic kidney disease in hypertensive patients with no known cardiovascular disease. J Am Soc Nephrol. 2006;17(12 suppl 3):S201-S205. [DOI] [PubMed] [Google Scholar]
- 16. Anderson AH, Yang W, Hsu CY, et al. ; CRIC Study Investigators . Estimating GFR among participants in the Chronic Renal Insufficiency Cohort (CRIC) Study. Am J Kidney Dis. 2012;60(2):250-261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Isakova T, Anderson CA, Leonard MB, et al. ; Chronic Renal Insufficiency Cohort (CRIC) Study Group . Diuretics, calciuria and secondary hyperparathyroidism in the Chronic Renal Insufficiency Cohort. Nephrol Dial Transplant. 2011;26(4):1258-1265. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Isakova T, Wahl P, Vargas GS, et al. Fibroblast growth factor 23 is elevated before parathyroid hormone and phosphate in chronic kidney disease. Kidney Int. 2011;79(12):1370-1378. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Chen J, Mohler ER 3rd, Xie D, et al. ; CRIC Investigators . Risk factors for peripheral arterial disease among patients with chronic kidney disease. Am J Cardiol. 2012;110(1):136-141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Scialla JJ, Appel LJ, Wolf M, et al. ; Chronic Renal Insufficiency Cohort (CRIC) Study Group . Plant protein intake is associated with fibroblast growth factor 23 and serum bicarbonate levels in patients with chronic kidney disease: the Chronic Renal Insufficiency Cohort study. J Ren Nutr. 2012;22(4):379-388.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Hoeting JA, Madigan D, Raftery AE, Volinsky CT. Bayesian model averaging: a tutorial. Statist Sci. 1999;14(4):382-417. [Google Scholar]
- 22. Blazek K, van Zwieten A, Saglimbene V, Teixeira-Pinto A. A practical guide to multiple imputation of missing data in nephrology. Kidney Int. 2021;99(1):68-74. [DOI] [PubMed] [Google Scholar]
- 23. Schemper M, Wakounig S, Heinze G. The estimation of average hazard ratios by weighted Cox regression. Stat Med. 2009;28(19):2473-2489. [DOI] [PubMed] [Google Scholar]
- 24. Patel K, Kay R, Rowell L. Comparing proportional hazards and accelerated failure time models: an application in influenza. Pharm Stat. 2006;5(3):213-224. [DOI] [PubMed] [Google Scholar]
- 25. Dunkler D, Ploner M, Schemper M, Heinze G. Weighted cox regression using the R Package coxphw. J Stat Softw. 2018;84(2):i02. [Google Scholar]
- 26. Jackson CH. flexsurv: a platform for parametric survival modeling in R. J Stat Softw. 2016;70:i08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Govindarajulu US, Malloy EJ, Ganguli B, Spiegelman D, Eisen EA. The comparison of alternative smoothing methods for fitting non-linear exposure-response relationships with Cox models in a simulation study. Int J Biostat. 2009;5(1):2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Staplin N, Herrington WG, Judge PK, et al. Use of causal diagrams to inform the design and interpretation of observational studies: an example from the study of heart and renal protection (SHARP). Clin J Am Soc Nephrol. 2017;12(3):546-552. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Rathod A, Bonny O, Guessous I, et al. Association of urinary calcium excretion with serum calcium and vitamin D levels. Clin J Am Soc Nephrol. 2015;10(3):452-462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Coyne D, Acharya M, Qiu P, et al. Paricalcitol capsule for the treatment of secondary hyperparathyroidism in stages 3 and 4 CKD. Am J Kidney Dis. 2006;47(2):263-276. [DOI] [PubMed] [Google Scholar]
- 31. Coburn JW, Maung HM, Elangovan L, et al. Doxercalciferol safely suppresses PTH levels in patients with secondary hyperparathyroidism associated with chronic kidney disease stages 3 and 4. Am J Kidney Dis. 2004;43(5):877-890. [DOI] [PubMed] [Google Scholar]
- 32. Ticinesi A, Nouvenne A, Maalouf NM, Borghi L, Meschi T. Salt and nephrolithiasis. Nephrol Dial Transplant. 2016;31(1):39-45. [DOI] [PubMed] [Google Scholar]
- 33. Sellmeyer DE, Schloetter M, Sebastian A. Potassium citrate prevents increased urine calcium excretion and bone resorption induced by a high sodium chloride diet. J Clin Endocrinol Metab. 2002;87(5):2008-2012. [DOI] [PubMed] [Google Scholar]
- 34. Bayomy O, Zaheer S, Williams JS, Curhan G, Vaidya A. Disentangling the relationships between the renin-angiotensin-aldosterone system, calcium physiology, and risk for kidney stones. J Clin Endocrinol Metab. 2020;105(6):1937-1946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Reilly RF, Huang CL. The mechanism of hypocalciuria with NaCl cotransporter inhibition. Nat Rev Nephrol. 2011;7(11):669-674. [DOI] [PubMed] [Google Scholar]
- 36. Lee CT, Chen HC, Lai LW, Yong KC, Lien YH. Effects of furosemide on renal calcium handling. Am J Physiol Renal Physiol. 2007;293(4):F1231-F1237. [DOI] [PubMed] [Google Scholar]
- 37. van Abel M, Hoenderop JG, van der Kemp AW, Friedlaender MM, van Leeuwen JP, Bindels RJ. Coordinated control of renal Ca(2+) transport proteins by parathyroid hormone. Kidney Int. 2005;68(4):1708-1721. [DOI] [PubMed] [Google Scholar]
- 38. Hill KM, Martin BR, Wastney ME, et al. Oral calcium carbonate affects calcium but not phosphorus balance in stage 3-4 chronic kidney disease. Kidney Int. 2013;83(5):959-966. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Chen J, Mohler ER, Xie D, et al. ; CRIC Study Investigators . Traditional and non-traditional risk factors for incident peripheral arterial disease among patients with chronic kidney disease. Nephrol Dial Transplant. 2016;31(7):1145-1151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Taylor JM, Kieneker LM, de Borst MH, et al. Urinary calcium excretion and risk of chronic kidney disease in the general population. Kidney Int Rep. 2017;2(3):366-379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Welles CC, Schafer AL, Vittinghoff E, Shlipak MG, Whooley MA, Ix JH. Urine calcium excretion, cardiovascular events, and mortality in outpatients with stable coronary artery disease (from the Heart and Soul study). Am J Cardiol. 2012;110(12):1729-1734. [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.
Data Availability Statement
Some or all data generated or analyzed during this study are included in this published article or in the data repositories listed in References.