Abstract
Purpose:
Prior studies have shown that thiazide diuretics’ ability to decrease symptomatic stone events depends on the extent to which they reduce urine calcium. However, it remains unknown whether the calcium change from thiazide diuretics depends on the baseline urine calcium level.
Materials and Methods:
Among a cohort of Medicare beneficiaries with a 24-hour urine collection for kidney stone disease processed by Labcorp/Litholink between 2011–2018, we identified a subset aged >18 years newly prescribed a thiazide diuretic who performed a second collection between 30 and 180 days after their initial prescription fill. We then fit multivariable linear regression models to estimate the association between dose of thiazide prescribed and change in urine calcium, stratifying by baseline urine calcium. We compared cumulative incidences of clinical stone events stratified by baseline urine calcium groups within a treated and untreated cohort.
Results:
From a total of 634 participants, higher baseline urine calcium was associated with greater 24-hour mean absolute (in mg/d) and percentage urine calcium reductions (Ptrend<0.001 for both). Higher thiazide dose was associated with greater absolute and percentage urine calcium reduction. No statistically significant differences were seen comparing the adjusted incidences of clinical stone events across baseline calcium groups after thiazide prescription. However, among individuals without thiazide exposure, the unadjusted and adjusted cumulative incidences were higher with greater baseline urine calcium (both Ptrend<0.001).
Conclusions:
Greater urine calcium reductions after thiazide treatment are observed among those with higher baseline urine calcium, and higher thiazide dose led to a larger reduction. These data are relevant for individualizing thiazide dose for kidney stone prevention.
Keywords: urolithiasis, secondary prevention, thiazides
Introduction
Thiazide diuretics are recommended for reducing the risk of kidney stone recurrence, among individuals with a history of recurrent calcium-based stones, particularly in the presence of hypercalciuria.1,2 Prior empirical work has shown that the extent to which thiazides reduce future symptomatic stone events depends on the magnitude of the reduction of urinary calcium after initiating treatment.3 However, while this finding may represent the average treatment effect, it is unclear whether the hypocalciuric effect of thiazide diuretics depends on the degree of hypercalciuria at baseline. A recent large trial demonstrated urine calcium reduction with hydrochlorothiazide compared to placebo regardless of baseline urine calcium levels in the treatment groups,4 suggesting their hypocalciuric effect was independent of baseline urine calcium levels. Additionally, several positive thiazide trials did not require elevated urine calcium levels as inclusion criteria,5,6 leading some to advocate for empirical thiazide use in settings when 24-hour urine calcium levels are unknown.7
On the other hand, it is plausible that the effect of thiazides on urine calcium depends on baseline calcium levels. Subgroup analyses of thiazide trials for hypertension have shown their efficacy varies based on baseline blood pressure levels.8 Considering the hypocalciuric effect from thiazides may mirror their anti-hypertensive effect observed in the setting of hypertension treatment, individuals with lower range urine calcium may benefit less from thiazides, which would warrant baseline urine testing before initiating treatment. In current practice, few patients are prescribed thiazide diuretics if urine calcium is in the reference range,9 consistent with two widely cited clinical guideline panels that recommend thiazide diuretics for recurrent kidney stone disease specifically when urine calcium is elevated.1,2
Within this context we sought to compare urine calcium changes by baseline urine calcium after starting thiazide by performing an observational cohort study of individuals with kidney stone disease prescribed thiazide diuretics with baseline and on-treatment urine calcium testing. We hypothesized that greater urine calcium reductions would be observed among individuals with higher baseline urine calcium. We additionally examined the association of thiazide dose with the treatment effects. Finally, to evaluate the impact of thiazide treatment on clinical outcomes by baseline calcium levels, we compared rates of clinically significant stone events by baseline urine calcium groups within a treated population with thiazides and an untreated population.
Materials and Methods
Data source and study populations
In this study we utilized the Medicare-Litholink database,10 comprised of claims in the Medicare Provider and Analysis Review (MedPAR), Outpatient, Carrier, and Part D Event Research Identifiable Files from 141,544 Medicare beneficiaries aged ≥18 years linked to results from 24-hour urine collections by Labcorp’s Litholink between January 1, 2011, and December 31, 2018. To allow for comorbidity adjustment we excluded beneficiaries without continuous enrollment in Medicare Parts A and B during the 1 year prior to the baseline 24- hour urine. We then constructed a primary analytic cohort including those who were treated with thiazides, and an untreated cohort including those who were not treated with thiazides, alkali citrate, or allopurinol to serve as controls when comparing the rate of clinical stone events.
Identifying thiazide users and nonusers
For identifying those treated with thiazides, we used National Drug Codes (see Supplemental File) to identify the subset that had a prescription fill for a thiazide agent, specifically hydrochlorothiazide, chlorthalidone, or indapamide, within 6 months after the initial baseline 24-hour urine test; and a second follow-up urine test 30 to 180 days after the initial prescription for thiazide (see Supplementary File for the cohort diagram). To allow for adequate follow up, we further excluded beneficiaries without continuous enrollment in Medicare Parts A and B during baseline 24-hour urine through 6 months after the initial prescription. To ensure capture of prescription coverage, we excluded beneficiaries without continuous enrollment in Medicare Part D during the 6 months prior to the baseline 24-hour urine through 6 months after the baseline 24-hr urine and 4 months after the initial prescription. We excluded those receiving thiazides or alkali citrate up to 6 months prior to the baseline 24-hour urine. We further excluded those receiving alkali citrate or allopurinol up to 120 days after thiazide initiation, those who changed daily dosing of thiazide between the baseline and follow-up urine collections, and those who had either inadequate baseline or follow-up 24-hour urine collections (accepted range for 24-hour creatinine/kilogram 11.9–24.4 mg/kg males, 8.7 to 20.3 mg/kg females). This yielded 634 beneficiaries in the treated cohort.
For our untreated cohort, we included those who did not have a prescription fill for thiazides, alkali citrate, or allopurinol 6 months prior through 1 year after the baseline 24-hr urine. To ensure adequate follow-up, we further excluded beneficiaries without continuous enrollment in Medicare Parts A and B up to 1 year after the baseline 24- hour urine. To ensure capture of prescription coverage, we excluded beneficiaries without continuous enrollment in Medicare Part D during the 6 months prior through 1 year after the baseline 24-hour urine. Applying these criteria resulted in identifying 26,798 beneficiaries for our untreated cohort.
Exposures
Our main exposure of interest was receipt of a thiazide diuretic. Among those in our treated cohort, we defined separate dosage groups for each medication subclass. Specifically, we classified thiazide doses into low (chlorthalidone <12.5 mg per day, indapamide <0.6125 mg per day, HCTZ <25 mg per day), medium (chlorthalidone 12.5 mg to <25 mg per day, indapamide 0.6125 to <1.25 mg per day, HCTZ 25 mg to <50 mg per day), and high (chlorthalidone 25 mg or higher, indapamide 1.25 mg per day or higher, HCTZ 50 mg per day or higher) dosage groups based on typical starting doses for kidney stone prevention.3 Among treated and untreated cohorts, we categorized all beneficiaries in our study into three baseline urine calcium groups: <250mg/day, 250 to <320mg/day, and 320mg/day or higher based on commonly used laboratory cutoffs.2,11
Outcomes
First, we examined for mean absolute and percentage differences in urinary calcium between the first and follow-up 24-hour urine test among those in the treated cohort. Next, for both the treated and untreated cohorts, we then evaluated for the cumulative occurrence of clinical kidney stone events– specifically a stone-related emergency department (ED) visit, hospitalization, or surgery – using a previously described diagnosis and procedure code-based algorithm.12,13 Specifically we assessed for the occurrence of a clinical stone event starting from six months up to up to 48 months after the first prescription fill for those in the treated cohort. For our untreated cohort, we randomly assigned them to a member of our treated cohort within the same baseline calcium group and then measured clinical stone events within the same time window after their baseline 24-hr urine test as their treated counterpart. We defined the event endpoint as a composite measure, indicating the first stone event of any type that occurred six months after the PPT fill date, regardless of if there was a stone event 6 months before the PPT fill date. We censored patients at loss of health insurance coverage, or on December 31, 2018, whichever came first. We identified ED visits by the presence of outpatient claims with revenue center codes 0450 to 0459 or 0981 or a non-zero ED charge amount in the MedPAR file.14 We identified hospitalizations using claims in the MedPAR file and surgery from ICD procedure and CPT codes (see Appendix 1). We defined ED visits and hospitalizations as stone related if they had an associated primary diagnosis of kidney stone disease during the encounter (see Supplemental File for codes).
Statistical analysis
For our initial analytic step, we compared beneficiaries in the treated cohort by baseline calcium groups across a variety of factors, including days to from baseline 24 hr. urine to prescription fill, age at baseline 24-hour urine test, sex, race/ethnicity, level of comorbid illness (number of hierarchical condition categories),15 region of residence, dual-eligibility status, high-risk status for stone recurrence, and medication adherence. Level of comorbid illness was measured using diagnoses from the year prior to the first 24-hour urine test. High-risk status was determined if there were diagnoses putting the patient at higher risk for kidney stones in the year prior to the first 24-hour urine test (see Supplemental File).16 Medication adherence was defined as ≥80% days of prescription coverage from start of thiazide to 120 days, as previously described.17 We additionally compared baseline 24-hour urine values, including calcium, volume, oxalate, citrate, pH, uric acid, sodium, potassium, magnesium, urea nitrogen, creatinine, net GI alkali absorption,18 and supersaturations of calcium oxalate and calcium phosphate. For these comparisons, we used the Wilcoxon rank sum test to compare the days to from baseline 24 hr. urine to prescription fill, Chi-squared tests to compare the beneficiary factors listed above, and ANOVA tests to compare the baseline 24-hour urine values.
We then examined the mean absolute and percentage differences in urinary calcium between the first and follow-up 24-hour urine test by baseline urine calcium groups, among those in the treated cohort. We performed a multivariable linear regression model for each outcome to test for a linear trend between our ordinal baseline calcium groups and the change in urine calcium and to estimate the marginal mean change in urine calcium, adjusted for age at baseline 24-hour urine test, sex, race/ethnicity,, level of comorbid illness, region of residence, dual-eligibility status, high-risk status for stone recurrence, and medication adherence. Since sodium and animal protein intakes associate with urine calcium excretion, we additionally adjusted for the change in urine sodium and change in urine urea nitrogen between the first and follow-up 24-hour urine tests.
We additionally compared the adjusted mean absolute and percentage differences by thiazide dosage within strata of baseline urine calcium groups. To do this, we calculated linear regression models among those in each baseline urine calcium group, adjusting for thiazide dosage and the same covariates as described previously to test for a linear trend between our ordinal thiazide dose groups and the change in urine calcium and to estimate the marginal mean change in urine calcium for each thiazide dose group within our baseline calcium group. Since these analyses were secondary, p-values were not adjusted for multiple comparisons.
Then, we used the Kaplan-Meier method to generate the unadjusted cumulative incidences of clinical stone events stratified by our baseline urine calcium groups among the treated and untreated cohort, using log rank trend tests to test for a linear trend between our ordinal baseline urine calcium groups and the cumulative incidence of stone events, within our treated and untreated cohort. Finally, we tested for a linear trend between baseline calcium groups and cumulative incidence and calculated the adjusted marginal cumulative incidences by baseline calcium group by fitting multivariable Cox proportional hazards models among our treated and untreated cohorts, adjusting for age at baseline 24-hour urine test, sex, race/ethnicity, level of comorbid illness, region of residence, dual-eligibility status, and high-risk status for stone recurrence. We calculated the unadjusted and adjusted cumulative incidences as the complement of the unadjusted and adjusted survival probabilities, respectively. Since calcium excretion may vary by body mass, a sensitivity analysis was performed stratifying the cohort into tertiles of baseline calcium divided by creatinine, and evaluating the outcome of absolute and percentage change in calcium/creatinine.
We conducted all analyses using SAS software, Version 9.4 (SAS Institute Inc., Cary, NC). We performed two-sided significance testing with alpha set at 0.05. The Institutional Review Board at the University of Michigan Health System deemed that this study was exempt from its oversight.
Results
There were 115, 186, and 333 individuals in the baseline urine calcium groups <250 mg/day, 250 to <320 mg/day, and ≥320 mg/day in the treated cohort (Table 1). Comparing baseline characteristics of the beneficiaries by baseline urine calcium groups, there were statistically significant differences in sex, race/ethnicity, and medication adherence. Individuals with higher baseline urine calcium received higher doses of thiazide and were more adherent to medication (both p<0.05). In addition to differences in baseline 24-hour urine calcium, there were statistically significant differences in 24-hour urine volume, citrate, sodium, potassium, magnesium, net GI alkali absorption, urea nitrogen, creatinine, and supersaturations of calcium oxalate and calcium phosphate. There were also minor differences observed for in urinary uric acid.
Table 1.
Baseline characteristics of beneficiaries by baseline urine calcium groups.
| Characteristic | Baseline calcium <250mg/day (n=115) | Baseline calcium 250 to <320mg/day (n=186) | Baseline calcium ≥320mg/day (n=333) | P-Value |
|---|---|---|---|---|
| Days to PPT fill from baseline 24 hr. urine, median (IQR) | 50 (24–85) | 35 (21–58) | 33 (21–59) | 0.004 |
| Age in years (%) | 0.4 | |||
| 18 to 64 | 21 (18) | 27 (15) | 50 (15) | |
| 65 to 69 | 41 (36) | 73 (39) | 139 (42) | |
| 70 to 74 | 34 (30) | 58 (31) | 111 (33) | |
| 75 or higher | 19 (17) | 28 (15) | 33 (10) | |
| Female (%) | 59 (51) | 106 (57) | 130 (39) | <0.001 |
| White (%) | 103 (90) | 173 (93) | 326 (98) | <0.001 |
| HCCs (%) | 0.2 | |||
| 0 | 37 (32) | 59 (32) | 129 (39) | |
| 1 | 32 (28) | 54 (29) | 104 (31) | |
| 2 | 21 (18) | 41 (22) | 49 (15) | |
| 3 or higher | 25 (22) | 32 (17) | 51 (15) | |
| Region of residence (%) | 0.8 | |||
| Northeast | 26 (23) | 49 (26) | 85 (26) | |
| South | 40 (35) | 54 (29) | 110 (33) | |
| West | 16 (14) | 35 (19) | 52 (16) | |
| Midwest | 33 (29) | 48 (26) | 86 (26) | |
| Dual Eligible (%) | 14 (12) | 19 (10) | 41 (12) | 0.8 |
| High risk16 stone former (%) | 65 (57) | 111 (60) | 177 (53) | 0.4 |
| Adherent to thiazide (%) | 46 (40) | 115 (62) | 262 (79) | <0.001 |
| Thiazide dosage | 0.022 | |||
| Low | 40 (35) | 58 (31) | 72 (22) | |
| Medium | 42 (37) | 62 (33) | 126 (38) | |
| High | 33 (29) | 66 (35) | 135 (41) | |
| Combination HCTZ formulation (%) | 87 (76%) | 133 (72%) | 215 (65%) | 0.05 |
| Baseline 24hr urine values, mean (SD) | ||||
| Urine volume, L/day | 1.8 (0.7) | 2.2 (0.8) | 2.4 (0.8) | <0.001 |
| Calcium, mg/day | 180 (59.7) | 286 (19.6) | 414 (90.4) | <0.001 |
| Oxalate, mg/day | 39.6 (16.7) | 37.4 (12.5) | 40.2 (15.5) | 0.11 |
| Citrate, mg/day | 673 (312) | 778 (281) | 932 (442) | <0.001 |
| pH | 6.2 (0.6) | 6.3 (0.5) | 6.2 (0.5) | 0.035 |
| Uric acid, mg/day | 582 (193) | 590 (174) | 702 (229) | <0.001 |
| Sodium, mmol/day | 153 (63.3) | 161 (56.9) | 197 (71.8) | <0.001 |
| Potassium, mmol/day | 58.0 (21.6) | 63.8 (21.2) | 72.1 (26.3) | <0.001 |
| Magnesium, mg/day | 92.1 (39.4) | 108 (44.4) | 126 (49.4) | <0.001 |
| Urea nitrogen, g/day | 9.1 (3.4) | 9.9 (3.3) | 12.2 (4.1) | <0.001 |
| Net GI alkali absorption | 33.4 (22.5) | 35.9 (20.5) | 43.1 (25.0) | <0.001 |
| Creatinine, mg/day | 1544 (466) | 1501 (455) | 1770 (520) | <0.001 |
| Creatinine per weight (mg/kg/day) | 15.0 (3.2) | 15.3 (3.4) | 16.4 (3.1) | <0.001 |
| SSCaOx | 7.4 (3.5) | 7.8 (2.9) | 8.4 (2.7) | 0.003 |
| SSCaP | 1.3 (1.1) | 1.7 (1.0) | 1.9 (1.1) | <0.001 |
HCC = Hierarchical Condition Categories
SSCaOx = Supersaturation of Calcium Oxalate
SSCaP = Supersaturation of Calcium Phosphate
suppressed due to cell size N<11
In the treated cohort, comparing the urine calcium changes by baseline urine calcium groups <250 mg/day, 250 to <320 mg/day, and ≥320 mg/day; the adjusted mean (95% CI) absolute and percentage calcium changes with thiazide were −32.0 mg (−50.1, −13.9) and −11.4% (−17.0, −5.7), −70.0 mg (−84.2, −55.9) and −24.4% (−28.8, −19.9), −128 mg (−139, −118) and −29.7% (−33.0, −26.4), respectively (p-trend <0.001 for both) (Figure 1). Similar trends were observed by baseline calcium/creatinine groups (see Supplemental Table).
Figure 1:
Adjusted absolute and percent change in urine calcium by baseline urine calcium group. Error bars represent 95% confidence intervals.
The adjusted urine calcium change by thiazide dosage is shown in Table 2. Higher daily dosage was associated with significantly greater absolute and percentage change in urine calcium for the baseline calcium 250 to <320 mg/day and ≥320 mg/day groups.
Table 2.
Adjustedb absolute and percent change in urine calcium by baseline urine calcium group across thiazide dosage.
| Baseline urine calcium group | Thiazide dose groupa | N | Mean absolute change, mg/day (95% CI) | P-trend | Mean percent change, % (95% CI) | P-trend |
|---|---|---|---|---|---|---|
| <250mg/day | Low | 40 | 0.1 (−25.1, 25.4) | 0.13 | 4.2 (−11, 20) | 0.2 |
| Medium | 42 | −41.0 (−65.5, −16.5) | −19 (−34, −3.9) | |||
| High | 33 | −29.7 (−56.1, −3.2) | −11 (−27, 5.3) | |||
| 250-<320mg/day | Low | 58 | −50.5 (−73.1, −28.0) | 0.039 | −17 (−25, −9.3) | 0.029 |
| Medium | 62 | −63.5 (−85.1, −41.9) | −22 (−30, −15) | |||
| High | 66 | −82.7 (−103, −62.2) | −29 (−36, −22) | |||
| ≥320mg/day | Low | 72 | −105 (−131, −79.2) | 0.005 | −25 (−31, −19) | 0.007 |
| Medium | 126 | −131 (−151, −112) | −31 (−36, −27) | |||
| High | 135 | −152 (−171, −133) | −35 (−39, −31) |
Thiazide dose groups are defined as low (chlorthalidone <12.5 mg per day, indapamide <0.6125 mg per day, HCTZ <25 mg per day), medium (chlorthalidone 12.5 mg to <25 mg per day, indapamide 0.6125 to <1.25 mg per day, HCTZ 25 mg to <50 mg per day), and high (chlorthalidone 25 mg or higher, indapamide 1.25 mg per day or higher, HCTZ 50 mg per day or higher)
Estimated marginal means for thiazide dose groups were obtained using linear regression models, adjusted for age at baseline 24-hour urine test, sex, race/ethnicity, level of comorbid illness, region of residence, dual-eligibility status, high-risk status for stone recurrence, medication adherence, the change in urine sodium and change in urine urea nitrogen between the first and follow-up 24-hour urine tests.
The cumulative incidences of clinical stone events comparing baseline urine calcium groups within treated and untreated cohorts are shown in Table 3 and Figure 2. In the untreated cohort, the unadjusted and adjusted cumulative incidences were higher with greater baseline urine calcium (both p-trend <0.001). In the treated cohort, no statistically significant differences were seen comparing the unadjusted (p=0.7) and adjusted incidences across baseline calcium groups (p=0.3)
Table 3.
Crude and adjusteda clinical stone event rates comparing baseline urine calcium groups within treated and untreated cohorts.
| Baseline urine calcium group | Unadjusted cumulative incidence, % (95% CI) | P-trend | Adjusteda cumulative incidence, % (95% CI) | P-trend | |
|---|---|---|---|---|---|
| Treated with thiazides | <250mg/day | 23 (15, 34) | 0.7 | 24 (14, 32) | 0.3 |
| 250 to <320mg/day | 25 (18, 35) | 25 (17, 32) | |||
| ≥320mg/day | 22 (17, 28) | 20 (15, 25) | |||
| Not on a thiazide | <250mg/day | 23 (22, 24) | <0.001 | 23 (22, 24) | <0.001 |
| 250 to <320mg/day | 27 (25, 29) | 25 (24, 27) | |||
| ≥320mg/day | 29 (27, 32) | 28 (26, 30) |
Estimated marginal cumulative incidences for baseline urine calcium groups were obtained using Cox proportional hazards models adjusted for age at baseline 24-hour urine test, sex, race/ethnicity, level of comorbid illness, region of residence, dual-eligibility status, and high-risk status for stone recurrence. We first estimated the marginal survival and then calculated the marginal cumulative incidences as the complement of the survival.
Figure 2:
Cumulative incidence of a clinical stone event among cohorts of (left) untreated and (right) treated with thiazides by baseline urine calcium group. On the x-axis, time is time from prescription fill date in the treated cohort and assigned date in the control cohort, and y is years.
Discussion
There were several important findings from this study. First, we observed greater urine calcium reduction among treated individuals who had higher urine calcium at baseline. Second, there was a dose-dependent effect of thiazide diuretics among individuals with baseline urine calcium ≥250 mg/day, in whom higher doses were associated with greater urine calcium reductions. Finally, higher baseline urine calcium levels are positively associated with subsequent kidney stone events resulting in clinical care episodes among individuals who are not treated with thiazides, whereas no association was observed among those who were treated.
These findings further elucidate the role of thiazide diuretics in kidney stone disease management. Consistent with prior literature,19 hypercalciuria is a major risk factor for calcium stone formation. A reduction in urine calcium and a corresponding reduction in kidney stone events are closely linked.3 Real-world data confirm that patients with hypercalciuria who adhere to thiazide diuretic therapy have a significant reduction in future symptomatic stone events compared to those who are non-adherent.13 The finding of higher thiazide dosing among individuals with higher baseline urine calcium suggests that providers are adjusting the treatment dose to the degree of hypercalciuria. Additionally, the observation that medication adherence was higher among higher baseline urine calcium groups is notable. Perhaps these individuals were more motivated to adhere to treatment from knowledge of their test results or received more intensive counseling. To our knowledge, our study is the first to show that the degree of hypercalciuria and the thiazide dosage play an important role in treatment response. A key finding is that for individuals with medium and high baseline calcium levels, there is an association between thiazide therapy and a risk reduction in symptomatic stone events towards that of low baseline urine calcium. Whereas among those untreated, these patients have substantially higher risk of symptomatic stone events.
Major societies including the American College of Physicians (ACP), European Urologic Association (EUA), and the American Urological Association (AUA) endorse thiazides for individuals with recurrent calcium stones, especially when fluid and dietary interventions are insufficient.1,2,20 The ACP guideline pointed to the need for evidence to establish a correlation between baseline urine calcium and the efficacy of thiazide therapy.20 In this study, our results support the use of 24-hour urine testing to assess urine calcium level prior to initiating thiazides. Additionally, the EAU guidelines recommend thiazides above a urine calcium threshold of 320 mg/day,2 whereas our data would suggest that those below this threshold may have clinical benefit with treatment.
Several limitations of this study should be noted. Limitations inherent in this retrospective cohort study include unmeasured confounding and the potential for miscoding. Unmeasured confounding would likely tend to understate the influence of thiazides, as there may be uncaptured factors about stone recurrence risk driving the decision to use them. The findings may not be generalizable to other populations, as this cohort was primarily among older individuals and excluded individuals who were on allopurinol or on combination therapy. Clinical details were unavailable such as stone composition, frequency of medication dosing per day, use of calcium and vitamin D supplements, radiographic characterization of stone burden, or stone passage not requiring clinical care. However, our focus on emergency department (ED) visits, hospitalization, and surgery represent clinically meaningful outcomes.
Despite these limitations, our findings have important clinical implications. Since preventative interventions for kidney stone disease are long-term, shared-decision making prior to initiating thiazides is warranted, weighing the clinical benefit in reducing future stone events against the harms, including the side effect profile of thiazides21,22 and the need for monitoring during follow-up.1 Finally, since the thiazide dose is an important factor in the treatment effect, our data suggest that individualizing thiazide dose for stone prevention should be considered. Future studies should examine how thiazides should be individualized for treatment. For example, in addition to dose, data is needed on the efficacy of shorter-acting thiazides (e.g. hydrochlorothiazide) versus longer-acting thiazides.23 Additional studies are also needed on the relationship among treatment with thiazides, urine calcium change, and sodium restriction.24
Conclusions
Greater urine calcium reductions after thiazide treatment are observed among those with higher baseline urine calcium. We additionally observed a dose-dependent effect of thiazide dose and urine calcium reduction, particularly among individuals with baseline urine calcium ≥250 mg/day. Our findings also confirm that among individuals who are not treated with thiazides, higher baseline urine calcium levels are associated with subsequent kidney stone events resulting in clinical care episodes.
Supplementary Material
Funding:
Supported by National Institutes of Health Grant R01DK121709.
Contributor Information
Ryan S. Hsi, Department of Urology, Vanderbilt University Medical Center; Department of Urology, University of California, Irvine.
Phyllis L. Yan, Dow Division of Health Services Research, Department of Urology, University of Michigan.
Joseph J. Crivelli, Department of Urology, University of Alabama at Birmingham School of Medicine.
Pietro Manuel Ferraro, Section of Nephrology, Department of Medicine, Università degli studi di Verona, Verona, Italy.
Gary C. Curhan, Channing Division of Network Medicine, Renal Division, Department of Medicine, Brigham and Women’s Hospital.
Sarah Best, Department of Urology, University of Wisconsin, Madison, Wisconsin.; Litholink Corporation, Laboratory Corporation of America Holdings®
John R. Asplin, Litholink Corporation, Laboratory Corporation of America Holdings®.
Vahakn Shahinian, Dow Division of Health Services Research, Department of Urology, University of Michigan; Division of Nephrology, Department of Internal Medicine, University of Michigan.
John M. Hollingsworth, Department of Urology, University of Florida College of Medicine, Gainesville..
References
- 1.Pearle MS, Goldfarb DS, Assimos DG, et al. : Medical management of kidney stones: AUA guideline. J Urol 2014; 192: 316–324. Available at: http://www.ncbi.nlm.nih.gov/pubmed/24857648. [DOI] [PubMed] [Google Scholar]
- 2.Skolarikos A, Somani B, Neisius A, et al. : Metabolic Evaluation and Recurrence Prevention for Urinary Stone Patients: An EAU Guidelines Update. Eur. Urol 2024; 86: 343–363. Available at: https://pubmed.ncbi.nlm.nih.gov/39069389/, accessed August 31, 2025. [DOI] [PubMed] [Google Scholar]
- 3.Hsi RS, Yan PL, Maalouf NM, et al. : Thiazide Dose, Urine Calcium, and Symptomatic Kidney Stone Events. JAMA Netw. Open 2024; 7: e2428953. Available at: https://pubmed.ncbi.nlm.nih.gov/39172453/, accessed September 1, 2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dhayat NA, Bonny O, Roth B, et al. : Hydrochlorothiazide and Prevention of Kidney-Stone Recurrence. N. Engl. J. Med 2023; 388: 781–791. Available at: https://pubmed.ncbi.nlm.nih.gov/36856614/, accessed September 1, 2025. [DOI] [PubMed] [Google Scholar]
- 5.Laerum E and Larsen S: Thiazide prophylaxis of urolithiasis. A double-blind study in general practice. Acta Med Scand 1984; 215: 383–389. Available at: http://www.ncbi.nlm.nih.gov/pubmed/6375276. [PubMed] [Google Scholar]
- 6.Ettinger B, Citron JT, Livermore B, et al. : Chlorthalidone reduces calcium oxalate calculous recurrence but magnesium hydroxide does not. J Urol 1988; 139: 679–684. Available at: http://www.ncbi.nlm.nih.gov/pubmed/3280829. [DOI] [PubMed] [Google Scholar]
- 7.Goldfarb DS: Empiric therapy for kidney stones. Urolithiasis 2018. Available at: https://www.ncbi.nlm.nih.gov/pubmed/30478476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Musini VM, Nazer M, Bassett K, et al. : Blood pressure-lowering efficacy of monotherapy with thiazide diuretics for primary hypertension. Cochrane Database of Systematic Reviews 2014; 2017. Available at: https://pubmed-ncbi-nlm-nih-gov.proxy.library.vanderbilt.edu/24869750/, accessed September 1, 2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Song S, Thomas I-C, Ganesan C, et al. : Twenty-Four Hour Urine Testing and Prescriptions for Urinary Stone Disease-Related Medications in Veterans. Clin. J. Am. Soc. Nephrol 2019; 14: 1773–1780. Available at: files/12120/Song et al. - 2019 - Twenty-Four Hour Urine Testing and Prescriptions f.pdf. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Krampe NA, Oerline MK, Asplin JR, et al. : Potential for Urolithiasis-related Research Using the Novel Medicare-Litholink Database. Urol. Pract 2023; 10: 147–152. Available at: https://pubmed.ncbi.nlm.nih.gov/37103409/, accessed October 23, 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Anon: 24-Hour Urine Testing | Litholink. Available at: https://litholink.labcorp.com/testing/24-hour-urine-testing, accessed September 1, 2025.
- 12.Hsi RS, Yan PL, Goldfarb DS, et al. : Comparison of Selective Versus Empiric Pharmacologic Preventative Therapy With Kidney Stone Recurrence. Urology 2020. Available at: http://www.ncbi.nlm.nih.gov/pubmed/33352163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hollingsworth JM, Oerline MK, Hsi RS, et al. : Real-World Effectiveness of Preventive Pharmacological Therapy in Patients With Urolithiasis: A Retrospective Cohort Study. Am. J. Kidney Dis 2024; 84: 83–93.e1. Available at: https://pubmed.ncbi.nlm.nih.gov/38432593/, accessed June 30, 2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Anon: How to Identify Hospital Claims for Emergency Room Visits in the Medicare Claims Data | ResDAC. Available at: https://resdac.org/articles/how-identify-hospital-claims-emergency-room-visits-medicare-claims-data, accessed September 1, 2025.
- 15.Pope GC, Kautter J, Ellis RP, et al. : Risk Adjustment of Medicare Capitation Payments Using the CMS-HCC Model. Health Care Financ. Rev 2004; 25: 119. Available at: https://pmc.ncbi.nlm.nih.gov/articles/PMC4194896/, accessed November 6, 2025. [PMC free article] [PubMed] [Google Scholar]
- 16.Hsi RS, Yan PL, Crivelli JJ, et al. : Comparison of Selective vs Empiric Pharmacologic Preventive Therapy of Kidney Stone Recurrence With High-Risk Features. Urology 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Dauw CA, Yi Y, Bierlein MJ, et al. : Medication Nonadherence and Effectiveness of Preventive Pharmacological Therapy for Kidney Stones. J Urol 2016; 195: 648–652. Available at: http://www.ncbi.nlm.nih.gov/pubmed/26485048. [DOI] [PubMed] [Google Scholar]
- 18.Oh MS: A new method for estimating G-I absorption of alkali. Kidney Int. 1989; 36: 915–917. Available at: https://pubmed-ncbi-nlm-nih-gov.proxy.library.vanderbilt.edu/2615198/, accessed September 1, 2025. [DOI] [PubMed] [Google Scholar]
- 19.Curhan GC, Willett WC, Speizer FE, et al. : Twenty-four-hour urine chemistries and the risk of kidney stones among women and men. Kidney Int. 2001; 59: 2290–2298. Available at: https://pubmed-ncbi-nlm-nih-gov.proxy.library.vanderbilt.edu/11380833/, accessed September 2, 2025. [DOI] [PubMed] [Google Scholar]
- 20.Qaseem A, Dallas P, Forciea MA, et al. : Dietary and pharmacologic management to prevent recurrent nephrolithiasis in adults: a clinical practice guideline from the American College of Physicians. Ann Intern Med 2014; 161: 659–667. Available at: https://www.ncbi.nlm.nih.gov/pubmed/25364887. [DOI] [PubMed] [Google Scholar]
- 21.Huen SC and Goldfarb DS: Adverse Metabolic Side Effects of Thiazides: Implications for Patients With Calcium Nephrolithiasis. J. Urol 2007; 177: 1238–1243. Available at: https://www-auajournals-org.proxy.library.vanderbilt.edu/doi/10.1016/j.juro.2006.11.040, accessed February 27, 2023. [DOI] [PubMed] [Google Scholar]
- 22.Hsi RS, Crivelli JJ, Yan PL, et al. : Comparison of Class-Specific Side Effects Across Preventative Pharmacologic Therapies for Kidney Stone Disease. Urol. Pract 2024; 11: 172–178. Available at: https://pubmed.ncbi.nlm.nih.gov/38117963/, accessed June 3, 2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Wollin DA, Fattah H, Asplin JR, et al. : Effect of Chlorthalidone and Hydrochlorothiazide on Twelve-Hour versus Twenty-Four Hour Urinary Calcium Excretion. Journal of Urology 2025; 214: 280–288. Available at: https://www.auajournals.org/doi/pdf/10.1097/JU.0000000000004562, accessed September 2, 2025. [DOI] [PubMed] [Google Scholar]
- 24.Borghi L, Schianchi T, Meschi T, et al. : Comparison of two diets for the prevention of recurrent stones in idiopathic hypercalciuria. N Engl J Med 2002; 346: 77–84. Available at: http://www.ncbi.nlm.nih.gov/pubmed/11784873. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.