Abstract
Key Points
In a cross-sectional study, sodium-glucose cotransporter 2 inhibitors were associated with a significant increase in urine volume and urine citrate.
Glucagon-like peptide-1 receptor agonists were not associated with any significant changes in 24-hour urine parameters that affect stone formation.
Background
Emerging data suggest sodium-glucose cotransporter-2 inhibitors (SGLT2is) and glucagon-like peptide-1 receptor agonists (GLP-1RAs) may lower stone risk.
Methods
We characterized 24-hour urine parameters among patients with kidney stone disease receiving these agents using Medicare claims from beneficiaries with urine collections processed by a central laboratory between January 2010 and December 2019. We identified a cross-sectional cohort with diabetes and a prescription fill for SGLT2is or GLP-1RAs within the 6 months preceding their urine collection and matched controls. We additionally identified a subset of patients who performed two collections and had a prescription fill for SGLT2is or GLP-1RAs before the second collection, but not the first. We compared across 24-hour urinary parameters in both cohorts and adjusted for multiple comparisons.
Results
The cross-sectional cohort included 124 patients with a prescription fill for SGLT2is (and 620 matched controls) and 349 patients with a prescription fill for GLP-1RAs (and 349 matched controls). Compared with controls, patients on SGLT2is had a higher mean urine citrate (838 versus 636 mg; P < 0.01) and volume (2.4 versus 2.0 L; P < 0.01) with improved calcium phosphate supersaturation (P < 0.01). Lower urine pH and higher sulfate and uric acid were observed in the SGLT2i group (P < 0.01 for each). There were no significant differences in urine parameters with GLP-1RA. In the longitudinal analyses of SGLT2is (59 patients) and GLP-1RAs (154 patients), there were no significant differences in urinary parameters.
Conclusions
SGLT2is were associated with higher urine volume and citrate in a cross-sectional cohort. GLP-1RAs were not associated with changes that would reduce stone risk.
Keywords: kidney stones, SGLT2
Visual Abstract
Introduction
Novel pharmacologic agents such as sodium-glucose cotransporter-2 inhibitors (SGLT2is) and glucagon-like peptide-1 receptor agonists (GLP-1RAs) are welcome additions to treat diabetes and its complications.1,2 SGLT2is block the SGLT2 in the early proximal tubule, which results in glucosuria. While initially developed to lower blood glucose, SGLT2is offer profound kidney protection through incompletely understood mechanisms.3–5 In retrospective cohort studies, SGLT2is have been associated with lower urinary stone risk,6,7 and prospective cohort studies have shown changes in 24-hour urine parameters that could improve stone risk in patients with diabetes and healthy controls, including increased urine volume and citrate.8,9 However, data are lacking regarding the effect of SGLT2is on 24-hour urine parameters in patients with preexisting kidney stone disease.
GLP-1RAs are analogs of a gut incretin hormone that increase insulin secretion after a meal; reduce glucagon secretion and slow gastric emptying, which reduces appetite; improves blood glucose control; and causes clinically meaningful weight loss.10 Because obesity and metabolic syndrome are both risk factors of stone formation,11–13 GLP-1RAs could plausibly improve the stone risk profile of stone formers. Although these agents are kidney protective,3,14 the exact mechanism of action in the kidney is unclear.10
To better understand how SGLT2is and GLP-1RAs could potentially influence urinary stone risk, we undertook an analysis among Medicare beneficiaries with kidney stones who performed 24-hour urine collections.
Methods
Cross-Sectional Exposure-Control Study
Data Source and Study Population
We used the Medicare-Litholink database,15 comprised of claims in the Medicare Provider and Analysis Review, Outpatient, Carrier, and Part D Event Research Identifiable Files from approximately 181,657 Medicare beneficiaries linked to results from 24-hour urine collections by Labcorp's Litholink subsidiary between January 1, 2010, and December 31, 2019. To be included in the cohort, all the patients had a stone-related diagnosis or a stone-directed surgery within 6 months before the Litholink collection.15 Among these Medicare beneficiaries, we required that beneficiaries had continuous enrollment in Medicare Parts A and B (which provides insurance coverage for inpatient and outpatient care) in the 12 months preceding the first 24-hour urine collection and Part D enrollment (which provides prescription drug coverage) in the 6 months before the first urine collection. We excluded beneficiaries with missing weight or inadequate urine collections on the basis of creatinine per kilogram weight (normal 24-hour creatinine/kg body weight 11.9–24.4 mg/kg for male patients and 8.7–20.3 mg/kg for female patients) and those with Part D claims for specific medications known to affect urine chemistry in the 6 months preceding the 24-hour urine collection. These medications included thiazolidinediones, fenofibrate,16,17 or any pharmacologic treatment for stone prevention (i.e., thiazides, alkali citrate, or allopurinol).18 This yielded an initial population of 43,920 beneficiaries from which exposures and controls were selected.
Identification of SGLT2i and GLP1-RA Cohorts and Their Respective Controls
Among beneficiaries meeting the above inclusion and exclusion criteria, we identified those with at least one prescription fill for either a SGLT2i or a GLP1-RA agent from Part D claims in the 6 months before the 24-hour urine collection that additionally covered the date of the urine collection. We excluded those who received both an SGLT2i and GLP1-RA in the 6 months before the collection, did not have a diagnosis of diabetes mellitus type 2 (see Supplemental Materials), and for whom we could not identify a suitable control. This yielded 124 and 349 unique beneficiaries in the SGLT2i and GLP1-RA cohorts, respectively (Figure 1).
Figure 1.
Cross-sectional cohort. *Stone-prevention agents—alkali, citrate, allopurinol, or thiazide. Cr, serum creatinine; GLP1-RA, glucagon-like peptide-1 receptor agonist; SGLT2i, sodium-glucose cotransporter-2 inhibitor.
Controls were selected from the 43,920 beneficiaries described above. We additionally excluded potential controls who did not have a diagnosis of diabetes mellitus type 2 and those who had an SGLT2i or GLP1-RA agent in the 6 months before their first 24-hour urine collection. This yielded 12,884 potential controls. For the SGLT2i controls, we matched controls to each exposed patient 5:1 on the basis of age (±1 year), sex, and date of collection (±1 year). Mean weights for the SGLT2i exposure and control groups were similar, so no additional matching for weight was performed. For the GLP1-RA controls, we matched controls to each exposed patient 1:1 on the basis of age (±1 year), sex, date of collection (±1 year), and weight (±3 kg).
Statistical Analysis
For our initial analytic step, we compared beneficiaries receiving SGLT2i and GLP-1RA with their respective controls by factors used for matching, including age at 24-hour urine test, sex, race, and weight, as well as other potential confounders, including level of comorbid illness as indicated by the number of hierarchical condition categories (HCCs), region of residence, dual-eligibility status for Medicare and Medicaid, socioeconomic tertile, and high-risk status for stone recurrence.19 HCCs are medical codes that indicate a patient is at high risk of increased future health care costs, while dual-eligibility status for Medicare and Medicaid is a marker of poverty.19 We assessed socioeconomic tertile on the basis of patient zip code, as has been done previously, with the first tertile representing the least advantaged group and the third tertile representing the most advantaged group.20,21
Level of comorbid illness was measured using diagnoses from the year before the 24-hour urine test. High-risk status for stone recurrence is defined by a set of diagnostic codes for comorbid conditions that increase the risk of stone recurrence, as previously defined.15 High-risk status was determined if there were diagnoses putting the patient at higher risk of kidney stones in the year before the first 24-hour urine test (see Supplemental Materials). For these comparisons, we used parametric (t tests) and nonparametric (quantile regression with no regressors) tests as appropriate.
We compared mean values of urinary volume, calcium, oxalate, citrate, pH, uric acid, sodium, potassium, magnesium, phosphorus, ammonium, chloride, sulfate, urea nitrogen, creatinine, and supersaturations of calcium oxalate, calcium phosphate, and uric acid. To evaluate the contribution of gastrointestinal alkali absorption on urinary values, we compared the net gastrointestinal (GI) alkali absorption22 between the exposure and control groups. Net GI alkali absorption estimates the amount of dietary alkali absorbed from the intestinal tract by computing the total quantity of noncombustible cations (sodium, potassium, calcium, and magnesium) minus the total quantity of noncombustible anions (chloride and phosphate).22 For these comparisons, we used parametric (t tests) and nonparametric (quantile regression with no regressors) tests as appropriate, both with and without adjustment for multiple comparisons using the Bonferroni correction method. To explore whether duration of exposure potentially affected the 24-hour urine parameters, we examined subgroups of patients who had received the drug for <90 or >90 days.
Longitudinal Study
Data Source and Study Population
For this analysis, among Medicare beneficiaries with urine collections from January 1, 2010, and December 31, 2019,15 we identified beneficiaries with two 24-hour urine collections. We required Part D coverage from the first collection date to the second collection date and required the second 24-hour urine creatinine to be within 30% of the first 24-hour urine creatinine to ensure similar duration of collection. We additionally excluded any beneficiaries who received any thiazides, alkali citrate, or allopurinol between the first and second urine collections.
Identification of SGLT2i and GLP1-RA Cohorts
Among beneficiaries meeting the criteria above, we identified separate SGLT2i and GLP1-RA cohorts from Part D claims. For the SGLT2i cohort, we identified those receiving SGLT2i on the second, but not the first, collection. Similarly, for the GLP1-RA cohort, we identified those receiving GLP1-RA on the second, but not the first, collection. This yielded a final analytic cohort of 59 and 154 patients receiving SGLT2i and GLP-1RA agents, respectively (Figure 2).
Figure 2.
Longitudinal cohort. (A) SGLT2i. (B) GLP-1RA. *Stone-prevention agents—alkali, citrate, allopurinol, or thiazide.
Statistical Analyses
Descriptive statistics for the SGLT2i and GLP-1RA longitudinal cohorts were calculated for the same variables as described in the previous cross-sectional study. Similarly, the 24-hour urinary values and net GI alkali absorption were compared between the first and second collections for both cohorts using parametric and nonparametric tests as appropriate. We then adjusted for multiple comparisons with a Bonferroni correction.
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
Cross-Sectional Cohort Study
SGLT2is
The median age of patients receiving SGLT2is was 69.4 years (interquartile range [IQR], 66.8–72.0), 37.9% were female, and 15.3% were non-White (Table 1). The exposure group (n=124) and controls (n=620) were similar for age, sex, weight, number of HCCs, region, nonmetropolitan area, dual eligibility for Medicare and Medicaid, and socioeconomic status, indicating that the participants in the exposure and control groups were similar in terms of demographic factors, number of comorbid conditions, and socioeconomic status. The duration of exposure for SGLT2is was >14 days for 96.8% of the cohort. Exposures differed from the controls on their 24-hour urine parameters in several factors that alter risk of stone formation (Table 1). The urine volume of exposures with SGLT2is was higher than controls (2.4±0.8 versus 2.0±0.8 L/d, adjusted P < 0.01, Figure 3A). Urinary citrate was higher in exposures than controls (838±502 versus 636±440 mg/d, adjusted P < 0.01, Figure 3B). The 24-hour urine pH was lower in exposures than in controls (5.6±0.5 versus 5.8±0.6, adjusted P < 0.01, Figure 3C) without a corresponding change in the ammonium (NH4+) level (Table 1). Moreover, sulfate was higher in exposures than in controls (40.6 ± 15.7 versus 34.5 ± 16.3 mEq/d, adjusted P < 0.01, Figure 3E). Net GI alkali absorption was higher among exposures than in controls (30.7±22.9 versus 24.8±23.6 mEq/d, Figure 3F), but was only significant without adjustment for multiple comparisons (unadjusted P = 0.01). SGLT2i users had lower supersaturation of calcium phosphate (0.2 [IQR, 0.1–0.5] versus 0.4 [IQR, 0.2–1.1], adjusted P < 0.01, Figure 3H) and a trend toward a lower supersaturation of calcium oxalate (5.4±3.1 versus 6.3±3.4, adjusted P = 0.05, Figure 3G). Urine uric acid (0.703 ± 0.213 versus 0.610±0.214 g/d, Figure 3D) was higher in exposures than controls (adjusted P < 0.01), and while uric acid supersaturation was also higher in exposures than controls (1.4 [IQR, 0.8–2.0] versus 0.9 [IQR, 0.4–1.8], Figure 3I), it was only statistically significant without adjustment for multiple comparisons (unadjusted P < 0.01). When we compared the change in urine parameters between patients who had been receiving the drug for <90 days versus those who had been receiving the drug for more than 90 days, there was no statistically significant difference after Bonferroni correction (Supplemental Table 1). Compared with controls, urine volume was significantly greater in patients receiving SGLT2is for >90 days (adjusted P value < 0.01).
Table 1.
Comparison between SGLT2i exposures and controls in the cross-sectional cohort
| Characteristic | SGLT2i | |||
|---|---|---|---|---|
| Exposures (n=124) | Controls (n=620) | P Value | P Value (Bonferroni Adjustment) | |
| Sociodemographic and clinical factors | ||||
| Median age (25th–75th) | 69.4 (66.8–72.0) | 69.2 (67.0–72.3) | 0.90 | |
| Female, No. (%) | 47 (37.9) | 235 (37.9) | 1.00 | |
| Race, No. (%) | 0.27 | |||
| Non-White | 19 (15.3) | 73 (11.8) | ||
| White | 105 (84.7) | 547 (88.2) | ||
| Mean weight, kg (SD) | 88.8 (18.9) | 89.0 (19.6) | 0.93 | |
| Total number of HCCs, No. (%) | 0.63 | |||
| 0 | 21 (16.9) | 101 (16.3) | ||
| 1 | 36 (29.0) | 170 (27.4) | ||
| 2 | 30 (24.2) | 128 (20.6) | ||
| 3 or more | 37 (29.8) | 221 (35.6) | ||
| High-risk, No. (%) | 61 (49.2) | 314 (50.6) | 0.77 | |
| Region, No. (%) | 0.16 | |||
| Midwest | 20 (16.1) | 147 (23.7) | ||
| Northeast | 32 (25.8) | 177 (28.5) | ||
| South | 51 (41.1) | 205 (33.1) | ||
| West | 21 (16.9) | 91 (14.7) | ||
| Nonmetropolitan area, No. (%) | 18 (14.5) | 102 (16.5) | 0.59 | |
| Dual eligible, No. (%) | 14 (11.3) | 75 (12.1) | 0.80 | |
| SES tertile, No. (%) | 0.62 | |||
| Low | 42 (33.9) | 206 (33.2) | ||
| Medium | 37 (29.8) | 211 (34.0) | ||
| High | 45 (36.3) | 203 (32.7) | ||
| Urine factors | ||||
| One or more abnormalitiesa, No. (%) | 116 (93.5) | 539 (86.9) | ||
| Mean chemistry (SD) | ||||
| pH | 5.6 (0.5) | 5.8 (0.6) | <0.01 | <0.01 |
| NH4+ (mmol/d) | 32.3 (12.9) | 31.4 (15.4) | 0.57 | 1.00 |
| Calcium (mg/d) | 198.0 (123.1) | 178.8 (117.0) | 0.10 | 1.00 |
| Chloride (mmol/d) | 189.2 (66.4) | 173.0 (70.6) | 0.02 | 0.35 |
| Citrate (mg/d) | 838 (502) | 636 (440) | <0.01 | <0.01 |
| Creatinine (mg/d) | 1378.7 (384.1) | 1370.2 (395.4) | 0.83 | 1.00 |
| Magnesium (mg/d) | 91.6 (41.2) | 92.3 (44.3) | 0.88 | 1.00 |
| Oxalate (mg/d) | 40.6 (13.6) | 41.5 (19.3) | 0.59 | 1.00 |
| Phosphorus (g/d) | 0.877 (0.335) | 0.828 (0.315) | 0.12 | 1.00 |
| Potassium (mmol/d) | 73.0 (26.8) | 60.8 (24.9) | <0.01 | <0.01 |
| Sodium (mmol/d) | 180.3 (66.8) | 168.4 (69.0) | 0.08 | 1.00 |
| Sulfate (mEq/d) | 40.6 (15.7) | 34.5 (16.3) | <0.01 | <0.01 |
| Urea nitrogen (g/d) | 11.6 (3.8) | 10.2 (3.9) | <0.01 | 0.01 |
| Uric acid (g/d) | 0.703 (0.213) | 0.610 (0.214) | <0.01 | <0.01 |
| Calcium oxalate supersaturation | 5.4 (3.1) | 6.3 (3.4) | 0.00 | 0.05 |
| Median SSCaP (25th–75th) | 0.2 (0.1–0.5) | 0.4 (0.2–1.1) | <0.01 | <0.01 |
| Median SSUA (25th–75th) | 1.4 (0.8–2.0) | 0.9 (0.4–1.8) | 0.01 | 0.13 |
| Net GI alkali absorption | 30.7 (22.9) | 24.8 (23.6) | 0.01 | 0.19 |
| Urine volume (L/d) | 2.4 (0.8) | 2.0 (0.8) | <0.01 | <0.01 |
Matched on age (±1 year), sex, and date of collection (±1 year). GI, gastrointestinal; HCC, hierarchical condition category; NH4+, ammonium; SES, socioeconomic status; SGLT2i, sodium-glucose cotransporter-2 inhibitor; SSCaP, calcium phosphate; SSUA, uric acid.
Measured urine abnormalities were hypercalciuria, hypocitraturia, hyperuricosuria, and low pH.
Figure 3.
Twenty-four–hour urine parameters from a cross-sectional cohort. 24 hour urine parameters in SGLT2-exposed patients vs. controls and GLP-1RA exposed patients vs. controls. (A) Urine volume (B) Citrate (C) pH (D) Uric acid (E) Sulfate (F) Net GI alkali absorption (G) supersaturation of calcium oxalate (H) supersaturation of calcium phosphate (I) supersaturation of uric acid. *Adjusted P value < 0.05; **adjusted P value < 0.01; ***adjusted P value < 0.001. †Unadjusted P value < 0.05; ††Unadjusted P value < 0.01.
GLP-1RA
The median age of exposures receiving GLP-1RA was 69.6 years (IQR, 66.6–72.6), 48.5% were female, and 8.9% were non-White (Table 2). The GLP-1RA users and controls were well matched on all other parameters, except for the number of HCCs (P = 0.04), which was higher in the exposures. The duration of exposure for GLP-1RA was >14 days for 98.3% of the cohort. After adjusting for multiple corrections, there were no differences in any 24-hour urine parameters between GLP-1RA users and controls (Figure 3, A, B, F, and H). Without adjusting for multiple comparisons, the exposures had a trend toward a higher uric acid excretion (0.643±0.222 versus 0.606±0.209 g/d, unadjusted P = 0.02, Figure 3D) and a higher uric acid supersaturation (1.2 [IQR, 0.5–2.1] versus 1.0 [IQR, 0.4–1.8], unadjusted P 0.02, Figure 3I) and a lower urinary sulfate level (33.2±15.4 versus 35.6±17.0 mEq/d, unadjusted P = 0.05, Figure 3E), despite no significant changes in urinary pH (Figure 3C). When we compared the change in urine parameters between patients who had been receiving the drug for <90 days versus those who had been receiving the drug for more than 90 days, there was no statistically significant difference after Bonferroni correction (Supplemental Table 2).
Table 2.
Comparison between GLP-1RA exposures and controls in the cross-sectional cohort
| Characteristic | GLP-1RA | |||
|---|---|---|---|---|
| Exposures (n=349) | Controls (n=349) | P Value | P Value (Bonferroni Adjustment) | |
| Sociodemographic and clinical factors | ||||
| Median age (25th–75th) | 69.6 (66.5–72.6) | 69.8 (66.5–72.7) | 0.61 | |
| Female, No. (%) | 169 (48.4) | 169 (48.4) | 1.00 | |
| Race, No. (%) | 0.89 | |||
| Non-White | 31 (8.9) | 32 (9.2) | ||
| White | 318 (91.1) | 317 (90.8) | ||
| Mean weight, kg (SD) | 94.7 (18.2) | 94.8 (18.2) | 0.96 | |
| Total number of HCCs, No. (%) | 0.04 | |||
| 0 | 39 (11.2) | 64 (18.3) | ||
| 1 | 84 (24.1) | 86 (24.6) | ||
| 2 | 88 (25.2) | 71 (20.3) | ||
| 3 or more | 138 (39.5) | 128 (36.7) | ||
| High-risk, No. (%) | 187 (53.6) | 203 (58.2) | 0.22 | |
| Region, No. (%) | 0.59 | |||
| Midwest | 86 (24.6) | 84 (24.1) | ||
| Northeast | 96 (27.5) | 84 (24.1) | ||
| South | 116 (33.2) | 132 (37.8) | ||
| West | 51 (14.6) | 49 (14.0) | ||
| Nonmetropolitan area, No. (%) | 66 (18.9) | 52 (14.9) | 0.16 | |
| Dual eligible, No. (%) | 63 (18.1) | 51 (14.6) | 0.22 | |
| SES tertile, No. (%) | 0.95 | |||
| Low | 114 (32.7) | 118 (33.8) | ||
| Medium | 118 (33.8) | 115 (33.0) | ||
| High | 117 (33.5) | 116 (33.2) | ||
| Urine factors | ||||
| One or more abnormalitiesa, No. (%) | 310 (88.8) | 302 (86.5) | 0.36 | |
| Mean chemistry (SD) | ||||
| pH | 5.8 (0.6) | 5.8 (0.6) | 0.10 | 1.00 |
| NH4+ (mmol/d) | 32.7 (35.4) | 34.0 (33.0) | 0.61 | 1.00 |
| Calcium (mg/d) | 176.8 (120.2) | 180.6 (122.7) | 0.68 | 1.00 |
| Chloride (mmol/d) | 182.3 (73.1) | 179.1 (78.0) | 0.57 | 1.00 |
| Citrate (mg/d) | 653.8 (433.4) | 600.6 (408.2) | 0.10 | 1.00 |
| Creatinine (mg/d) | 1394.9 (418.2) | 1389.2 (399.8) | 0.85 | 1.00 |
| Magnesium (mg/d) | 93.0 (47.1) | 93.2 (50.4) | 0.96 | 1.00 |
| Oxalate (mg/d) | 44.0 (20.8) | 41.7 (17.3) | 0.13 | 1.00 |
| Phosphorus (g/d) | 0.833 (0.333) | 0.877 (0.360) | 0.10 | 1.00 |
| Potassium (mmol/d) | 63.2 (24.8) | 61.8 (23.3) | 0.44 | 1.00 |
| Sodium (mmol/d) | 176.5 (72.6) | 176.5 (79.6) | 1.00 | 1.00 |
| Sulfate (mEq/d) | 33.2 (15.4) | 35.6 (17.0) | 0.05 | 0.99 |
| Urea nitrogen (g/d) | 10.1 (3.7) | 10.4 (4.0) | 0.38 | 1.00 |
| Uric acid (g/d) | 0.643 (0.222) | 0.606 (0.209) | 0.02 | 0.47 |
| Calcium oxalate supersaturation | 6.7 (3.7) | 6.5 (3.7) | 0.44 | 1.00 |
| Median SSCaP (25th–75th) | 0.4 (0.1–0.9) | 0.4 (0.2–1.0) | 0.88 | 1.00 |
| Median SSUA (25th–75th) | 1.2 (0.5–2.1) | 1.0 (0.4–1.8) | 0.02 | 0.33 |
| Net GI alkali absorption | 25.6 (23.6) | 25.1 (23.6) | 0.78 | 1.00 |
| Urine volume (L/d) | 2.0 (0.8) | 2.0 (0.8) | 0.50 | 1.00 |
Matched on age (±1 year), sex, date of collection (±1 year), and weight (±3 kg). GI, gastrointestinal; GLP-1RA, glucagon-like peptide-1 receptor agonist; HCC, hierarchical condition category; NH4+, ammonium; SES, socioeconomic status; SSCaP, calcium phosphate; SSUA, uric acid.
Measured urine abnormalities were hypercalciuria, hypocitraturia, hyperuricosuria, and low pH.
Longitudinal Study
SGLT2is
Clinical characteristics of patients (n=59) in the longitudinal study were similar to patients in the cross-sectional study (Table 3). The median age was 67.3 years (IQR, 65.6–71.2), and 18.6% were female. Data regarding race are suppressed because of small sample size. Patients were on SGLT2is for an average of 234±265 days before they underwent the second 24-hour urine collection, and 86.4% of the cohort was on SGLT2is for more than 14 days. After adjusting for multiple corrections, there were no statistically significant differences in 24-hour urine parameters (Figure 4, B–I), although urine volume was numerically higher after SGLT2i (2.4 L before SGLT2i and 2.7 L after SGLT2i, Figure 4A).
Table 3.
Comparison before and after SGLT2is in the longitudinal cohort
| Characteristic | SGLT2i | |||
|---|---|---|---|---|
| First Collection (n=59) | Second Collection (n=59) | P Value | P Value (Bonferroni Adjustment) | |
| Sociodemographic and clinical factors | ||||
| Median age (25th–75th) | 67.3(65.6–71.2) | |||
| Female, No. (%) | 11 (18.6) | |||
| Race, No. (%) | ||||
| Non-White | a | |||
| White | a | |||
| Mean weight, kg (SD) | 98.4 (22.5) | |||
| Total number of HCCs, No. (%) | ||||
| 0–1 | 32 (54.2) | |||
| 2 | 13 (22.0) | |||
| 3 or more | 14 (23.7) | |||
| High-risk, No. (%) | 25 (42.4) | |||
| Region, No. (%) | ||||
| Midwest | 11 (18.6) | |||
| Northeast | 16 (27.1) | |||
| South | 21 (35.6) | |||
| West | 11 (18.6) | |||
| Nonmetropolitan area, No. (%) | a | |||
| Dual eligible, No. (%) | a | |||
| Urine factors | ||||
| One or more abnormalitiesb, No. (%) | 49 (83.1) | 54 (91.5) | ||
| Mean chemistry (SD) | ||||
| pH | 5.8 (0.5) | 5.7 (0.6) | 0.51 | 1.00 |
| NH4+ (mmol/d) | 33.8 (13.1) | 32.9 (16.3) | 0.74 | 1.00 |
| Calcium (mg/d) | 218.7 (128.2) | 229.8 (139.9) | 0.65 | 1.00 |
| Chloride (mmol/d) | 223.7 (65.2) | 219.1 (88.6) | 0.75 | 1.00 |
| Citrate (mg/d) | 1075.8 (643.6) | 1016.4 (568.3) | 0.60 | 1.00 |
| Creatinine (mg/d) | 1712.8 (442.7) | 1618.8 (459.2) | 0.26 | 1.00 |
| Magnesium (mg/d) | 118.2 (52.0) | 116.4 (47.5) | 0.85 | 1.00 |
| Oxalate (mg/d) | 46.7 (13.5) | 44.7 (16.2) | 0.45 | 1.00 |
| Phosphorus (g/d) | 1.019 (0.396) | 0.975 (0.382) | 0.54 | 1.00 |
| Potassium (mmol/d) | 80.4 (22.7) | 79.9 (22.3) | 0.91 | 1.00 |
| Sodium (mmol/d) | 224.0 (71.1) | 219.7 (82.6) | 0.76 | 1.00 |
| Sulfate (mEq/d) | 47.5 (18.4) | 48.5 (20.6) | 0.78 | 1.00 |
| Urea nitrogen (g/d) | 13.7 (3.9) | 13.2 (4.8) | 0.53 | 1.00 |
| Uric acid (g/d) | 0.745 (0.263) | 0.743 (0.238) | 0.96 | 1.00 |
| Calcium oxalate supersaturation | 5.8 (3.3) | 5.1 (2.9) | 0.21 | 1.00 |
| Median SSCaP (25th–75th) | 0.3 (0.1–0.8) | 0.2 (0.1–0.6) | 0.32 | 1.00 |
| Median SSUA (25th–75th) | 1.2 (0.4–1.9) | 1.1 (0.5–1.8) | 0.72 | 1.00 |
| Net GI alkali absorption | 42.3 (29.1) | 45.1 (24.4) | 0.57 | 1.00 |
| Urine volume (L/d) | 2.4 (0.8) | 2.7 (0.8) | 0.09 | 1.00 |
Weight at the first collection is missing for N=5 of SGLT2i–exposed patients. GI, gastrointestinal; HCC, hierarchical condition category; NH4+, ammonium; SGLT2i, sodium-glucose cotransporter-2 inhibitor; SSCaP, calcium phosphate; SSUA, uric acid.
Suppressed because of inadequate sample size (N<11).
Measured urine abnormalities were hypercalciuria, hypocitraturia, hyperuricosuria, and low pH.
Figure 4.
Twenty-four–hour urine parameters from a longitudinal cohort. 24-hour urine parameters in patients pre-treatment and on-treatment for SGLT2i and GLP-1RA. (A) Urine volume (B) Citrate (C) pH( D) Uric Acid (E) Sulfate (F) Net GI alkali absorption (G) supersaturation of calcium oxalate (H) supersaturation of calcium phosphate (I) supersaturation of uric acid. ††Unadjusted P value < 0.01.
GLP-1RA
Clinical characteristics of patients in the longitudinal study (n=154) were similar to patients in the cross-sectional study (Table 4). The median age was 67.9 years (IQR, 65.7–70.9), 39% were female, and 9.7% were non-White. Patients were on GLP-1RA for an average of 207±96 days before the second collection, and 91.6% of the cohort was on GLP-1RA for more than 14 days. After adjusting for multiple comparisons, there were no statistically significant differences in 24-hour urine parameters. However, without adjusting for multiple comparisons, uric acid excretion was lower after initiation of GLP-1RA (0.656±0.255 g/d before GLP-1RA versus 0.596±0.225 after GLP-1RA, unadjusted P = 0.03, Figure 4D) with no corresponding change in uric acid supersaturation (Figure 4I).
Table 4.
Comparison before and after GLP-1RAs in the longitudinal cohort
| Characteristic | GLP-1RA | |||
|---|---|---|---|---|
| First Collection (n=154) | Second Collection (n=154) | P Value | P Value (Bonferroni Adjustment) | |
| Sociodemographic and clinical factors | ||||
| Median age (25th–75th) | 67.9 (65.7–70.9) | |||
| Female, No. (%) | 60 (39.0) | |||
| Race, No. (%) | ||||
| Non-White | 15 (9.7) | |||
| White | 139 (90.3) | |||
| Mean weight, kg (SD) | 99.9 (23.5) | |||
| Total number of HCCs, No. (%) | ||||
| 0–1 | 62 (40.3) | |||
| 2 | 38 (24.7) | |||
| 3 or more | 54 (35.1) | |||
| High-risk, No. (%) | 87 (56.5) | |||
| Region, No. (%) | ||||
| Midwest | 32 (20.8) | |||
| Northeast | 49 (31.8) | |||
| South | 44 (28.6) | |||
| West | 29 (18.8) | |||
| Nonmetropolitan area, No. (%) | 23 (14.9) | |||
| Dual eligible, No. (%) | 19 (12.3) | |||
| Urine factors | ||||
| One or more abnormalitiesa, No. (%) | 138 (89.6) | 144 (93.5) | ||
| Mean chemistry (SD) | ||||
| pH | 5.8 (0.7) | 5.8 (0.6) | 0.99 | 1.00 |
| NH4+ (mmol/d) | 36.7 (37.0) | 32.8 (18.6) | 0.25 | 1.00 |
| Calcium (mg/d) | 184.1 (130.1) | 177.7 (138.8) | 0.68 | 1.00 |
| Chloride (mmol/d) | 180.5 (75.1) | 174.7 (79.3) | 0.51 | 1.00 |
| Citrate (mg/d) | 729.9 (543.1) | 671.3 (505.9) | 0.33 | 1.00 |
| Creatinine (mg/d) | 1487.0 (515.4) | 1439.2 (482.6) | 0.40 | 1.00 |
| Magnesium (mg/d) | 104.8 (56.9) | 101.8 (59.3) | 0.65 | 1.00 |
| Oxalate (mg/d) | 47.9 (21.7) | 44.5 (23.7) | 0.20 | 1.00 |
| Phosphorus (g/d) | 0.913 (0.402) | 0.890 (0.412) | 0.62 | 1.00 |
| Potassium (mmol/d) | 66.1 (27.9) | 63.0 (26.5) | 0.33 | 1.00 |
| Sodium (mmol/d) | 179.8 (76.0) | 173.2 (79.9) | 0.46 | 1.00 |
| Sulfate (mEq/d) | 38.3 (18.1) | 35.4 (18.3) | 0.15 | 1.00 |
| Urea nitrogen (g/d) | 11.5 (4.5) | 10.9 (4.6) | 0.21 | 1.00 |
| Uric acid (g/d) | 0.656 (0.255) | 0.596 (0.225) | 0.03 | 0.52 |
| Calcium oxalate supersaturation | 6.5 (3.8) | 5.9 (3.7) | 0.15 | 1.00 |
| Median SSCaP (25th–75th) | 0.3 (0.1–0.8) | 0.3 (0.1–0.8) | 0.64 | 1.00 |
| Median SSUA (25th–75th) | 1.2 (0.4–2.1) | 1.0 (0.4–1.8) | 0.60 | 1.00 |
| Net GI alkali absorption | 30.3 (30.0) | 27.3 (28.1) | 0.36 | 1.00 |
| Urine volume (L/d) | 2.1 (0.8) | 2.1 (0.8) | 0.77 | 1.00 |
Weight at the first collection is missing for N=13 of GLP-1RA–exposed patients. GI, gastrointestinal; GLP-1RA, glucagon-like peptide-1 receptor agonist; HCC, hierarchical condition category; NH4+, ammonium; SSCaP, calcium phosphate; SSUA, uric acid.
Measured urine abnormalities were hypercalciuria, hypocitraturia, hyperuricosuria, and low pH.
Discussion
Our data show that SGLT2is were associated with higher urinary volume and higher urinary citrate in the cross-sectional study, which may beneficially reduce the rate of stone formation. In comparison, there were fewer changes in 24-hour urine parameters that could affect stone formation with GLP-1RA. Apart from providing insight into how these agents may affect kidney stone formation, comparing and contrasting the changes in 24-hour urine parameters with these agents provides the opportunity to better understand the effect of these drugs on renal physiology.
In the cross-sectional study, SGLT2is were associated with a greater urine volume by approximately 300–500 ml/d, higher urinary citrate, and lower urine pH. Correspondingly, SGLT2is were associated with statistically significant, but likely clinically insignificant, lower calcium phosphate supersaturation. While calcium phosphate supersaturation was nominally lower in the cross-sectional analysis with SGLT2is, the supersaturation of calcium phosphate was overall low and indicated that our study population was at low risk of calcium phosphate stones, as would be expected in pstients with diabetes and a low urine pH. In addition, while decreased supersaturation of calcium oxalate is associated with decreased stone formation,23,24 the clinical significance of the small magnitude of this change is unclear. While only detected in the cross-sectional analysis, a higher urine volume of 300–500 ml/d is clinically meaningful and could affect the rate of stone formation.25,26 The directionality of the change was preserved in the longitudinal analysis, but was not statistically significant, which may be due to limited sample size. SGLT2i use for longer than 1 month has been associated with increased urine volume in previous prospective studies of healthy controls and patients with type 2 diabetes and preserved kidney function, despite no significant changes in 24-hour urine sodium excretion.8,9 However, in one study where the participants' were treated with SGLT2is for 14 days and salt intake was controlled, there was no significant change in urine volume,27 which is consistent with our finding that the increase in urine volume was statistically significant in patients who had been taking SGLT2is for more than 90 days.
Higher urinary citrate with SGLT2is has also been reported in several studies, and the magnitude of the change in the cross-sectional study seems clinically meaningful.8,9 There are a few possible reasons for the higher urinary citrate. Our data showed a higher net GI alkali absorption with SGLT2is in the analysis without adjustment for multiple testing,22 which could explain the higher citrate excretion. Estimation of net GI alkali absorption closely correlates the direct measurement of net GI alkali absorption from diet and 24-hour stool collections, but it can be influenced by extrarenal and extragastrointestinal bicarbonate loss.22 However, there is generally no impairment with citrate absorption from the intestinal tract in stone formers.28 Moreover, sodium-glucose cotransporter-2 is highly selective for the kidney and has very low expression in the intestines, so it is not clear how these agents would affect gastrointestinal absorption. An alternate explanation is that citrate is an intermediate in the tricarboxylic acid (TCA) cycle, and diabetes increases TCA cycle flux within the proximal tubule.29 Transcripts of TCA cycle are diminished with SGLT2is in the proximal tubule of human kidney biopsies of patients receiving SGLT2i,30 so it is possible that these drugs decrease TCA flux and there is subsequently decreased uptake of citrate from the tubular lumen by SLC13A2 (NaDC1).31 Alternatively, it may be that the exposure group is consuming a more alkaline diet than the control group. Further experimentation would be needed to understand these mechanisms.
Type 2 diabetes is associated with a greater tendency for uric acid stone formation.32,33 There were several factors that would tend toward higher risk of uric acid stone formation with SGLT2is. Lower urine pH was observed with SGLT2is in our cross-sectional study and there was higher uric acid excretion, although these changes are numerically small, so the clinical significance is unclear. SGLT2is have been shown in other studies to be uricosuric because they affect URAT1 and GLUT9 function and limit uric acid reabsorption in the proximal tubule.34,35 It is not clear why there was a change in urine pH with SGLT2i use, although this has been previously observed in studies of healthy control patients,8 but not in patients with diabetes and kidney disease.9 In our study, exposures had a higher urinary sulfate, so they may have been consuming an increased acid load, which could in part explain the lower urine pH. SGLT2is have been associated with changes in the transcriptome of intercalated cells, so there may be unexpected effects on acid–base handling with these agents.30 Further work is needed to understand whether SGLT2is affect urinary buffers and intercalated cell function.
The pattern of changes observed with GLP-1RA differed from SGLT2is, and there was only a change in uric acid excretion in the analysis of the longitudinal cohort without adjustment for multiple comparisons. While the glucagon-like peptide-1 receptor is expressed in the kidney, the effect of GLP1-RAs on the kidney has been less well studied than SGLT2is, although they have been found to ameliorate metabolic disturbances in murine models of diabetic kidney disease.36 However, the relatively fewer changes in 24-hour urine parameters with GLP-1RA use is consistent with available epidemiologic data demonstrating that stone risk is lower with SGLT2is than with GLP-1RAs.7
Our study has several strengths, including that this is one of the largest available cohorts of 24-hour urine parameters. We used two study designs that accounted for confounding factors. In the cross-sectional analysis, we carefully matched for confounding factors, and our exposure and control groups were similar after matching. In the longitudinal study design, the patients served as their own control, which should have better controlled for confounding. Despite these strengths, our study has limitations. This is a retrospective study, and there are likely confounders that we could not account for, especially changes in diet. We do not have access to important information, such as stone type, changes in weight while on therapy, or hemoglobin A1c. In addition, we do not have information on whether the patients were taking SGLT2is or GLP-1RAs during their index stone event, which may have biased our study results toward the null. Moreover, patients may not have been taking the drug for the entirety of the time interval between the two 24-hour urine collections in the longitudinal study, which would also bias our results toward the null. There were limited numbers of patients on SGLT2is and GLP-1RAs available, particularly for the longitudinal analysis, which may have limited statistical power. The longitudinal study design may have selected a group of patients with more severe kidney stone disease because these patients had two collections, so the results from the longitudinal analysis may not be widely generalizable. Moreover, these results may not be generalizable to other populations, including nondiabetic patients and younger patients with stone disease.
In summary, in a cross-sectional cohort, SGLT2is were associated with higher urine volume and urine citrate, which may be protective against stone formation, particularly for calcium oxalate and calcium phosphate stones. However, there were also small alterations in the urine parameters that could be associated with mildly increased risk of uric acid stones, which is common in patients with metabolic syndrome. GLP-1RAs were not consistently associated with any parameters that would alter the rate of stone formation, despite their profound effect on features of metabolic syndrome. Because both of these drugs are now commonly prescribed to patients with diabetes, prospective clinical trials, particularly for SGLT2is, would be needed to determine whether these agents affect stone formation in a clinically meaningful way.
Supplementary Material
Disclosures
Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/KN9/A906.
Funding
V. Shahinian: National Institute of Diabetes and Digestive and Kidney Diseases (1R01DK121709).
Author Contributions
Conceptualization: John M. Hollingsworth, Ryan S. Hsi.
Formal analysis: Mary K. Oerline, Jennifer A. Schaub, Vahakn Shahinian.
Funding acquisition: John M. Hollingsworth, Ryan S. Hsi, Vahakn Shahinian.
Methodology: Vahakn Shahinian.
Supervision: Ryan S. Hsi, Jennifer A. Schaub.
Writing – original draft: Ryan S. Hsi, Jennifer A. Schaub.
Writing – review & editing: John R. Asplin, Sara L. Best, Joseph J. Crivelli, John M. Hollingsworth, Ryan S. Hsi, Naim M. Maalouf, Vahakn Shahinian.
Data Sharing Statement
Data cannot be shared. The individual level data are proprietary and cannot be shared publicly. The statistical code and readme files are available at https://doi.org/10.7302/55an-x018 to reconstruct the analysis should others wish to obtain access to the individual level data.
Supplemental Material
This article contains the following supplemental material online at http://links.lww.com/KN9/A904, http://links.lww.com/KN9/A905.
Supplemental Table 1. Subgroup analysis for duration of exposure of SGLT2i-exposed patients in the cross-sectional study.
Supplemental Table 2. Subgroup analysis for duration of exposure of GLP-1RA–exposed patients in the cross-sectional study.
References
- 1.Navaneethan SD Zoungas S Caramori ML, et al. Diabetes management in chronic kidney disease: synopsis of the KDIGO 2022 clinical practice guideline update. Ann Intern Med. 2023;176(3):381–387. doi: 10.7326/m22-2904 [DOI] [PubMed] [Google Scholar]
- 2.Adhikari R Jha K Dardari Z, et al. National trends in use of sodium-glucose cotransporter-2 inhibitors and glucagon-like peptide-1 receptor agonists by cardiologists and other specialties, 2015 to 2020. J Am Heart Assoc. 2022;11(9):e023811. doi: 10.1161/jaha.121.023811 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Perkovic V Jardine MJ Neal B, et al.; CREDENCE Trial Investigators. Canagliflozin and renal outcomes in type 2 diabetes and nephropathy. N Engl J Med. 2019;380(24):2295–2306. doi: 10.1056/NEJMoa1811744 [DOI] [PubMed] [Google Scholar]
- 4.Heerspink HJL Stefánsson BV Correa-Rotter R, et al.; DAPA-CKD Trial Committees and Investigators. Dapagliflozin in patients with chronic kidney disease. N Engl J Med. 2020;383(15):1436–1446. doi: 10.1056/NEJMoa2024816 [DOI] [PubMed] [Google Scholar]
- 5.Herrington WG Staplin N Wanner C, et al.; The EMPA-KIDNEY Collaborative Group, Empagliflozin in patients with chronic kidney disease. N Engl J Med. 2023;388(2):117–127. doi: 10.1056/NEJMoa2204233 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Paik JM, Tesfaye H, Curhan GC, Zakoul H, Wexler DJ, Patorno E. Sodium-glucose cotransporter 2 inhibitors and nephrolithiasis risk in patients with type 2 diabetes. JAMA Intern Med. 2024;184(3):265–274. doi: 10.1001/jamainternmed.2023.7660 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kristensen KB, Henriksen DP, Hallas J, Pottegård A, Lund LC. Sodium-glucose cotransporter 2 inhibitors and risk of nephrolithiasis. Diabetologia. 2021;64(7):1563–1571. doi: 10.1007/s00125-021-05424-4 [DOI] [PubMed] [Google Scholar]
- 8.Harmacek D Pruijm M Burnier M, et al. Empagliflozin changes urine supersaturation by decreasing pH and increasing citrate. J Am Soc Nephrol. 2022;33(6):1073–1075. doi: 10.1681/ASN.2021111515 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.van Bommel EJM Geurts F Muskiet MHA, et al. SGLT2 inhibition versus sulfonylurea treatment effects on electrolyte and acid-base balance: secondary analysis of a clinical trial reaching glycemic equipoise: tubular effects of SGLT2 inhibition in type 2 diabetes. Clin Sci (Lond). 2020;134(23):3107–3118. doi: 10.1042/cs20201274 [DOI] [PubMed] [Google Scholar]
- 10.Alicic RZ, Cox EJ, Neumiller JJ, Tuttle KR. Incretin drugs in diabetic kidney disease: biological mechanisms and clinical evidence. Nat Rev Nephrol. 2021;17(4):227–244. doi: 10.1038/s41581-020-00367-2 [DOI] [PubMed] [Google Scholar]
- 11.Taylor EN, Stampfer MJ, Curhan GC. Obesity, weight gain, and the risk of kidney stones. JAMA. 2005;293(4):455–462. doi: 10.1001/jama.293.4.455 [DOI] [PubMed] [Google Scholar]
- 12.Jeong IG Kang T Bang JK, et al. Association between metabolic syndrome and the presence of kidney stones in a screened population. Am J Kidney Dis. 2011;58(3):383–388. doi: 10.1053/j.ajkd.2011.03.021 [DOI] [PubMed] [Google Scholar]
- 13.West B, Luke A, Durazo-Arvizu RA, Cao G, Shoham D, Kramer H. Metabolic syndrome and self-reported history of kidney stones: the National Health and Nutrition Examination Survey (NHANES III) 1988-1994. Am J Kidney Dis. 2008;51(5):741–747. doi: 10.1053/j.ajkd.2007.12.030 [DOI] [PubMed] [Google Scholar]
- 14.Sattar N Lee MMY Kristensen SL, et al. Cardiovascular, mortality, and kidney outcomes with GLP-1 receptor agonists in patients with type 2 diabetes: a systematic review and meta-analysis of randomised trials. Lancet Diabetes Endocrinol. 2021;9(10):653–662. doi: 10.1016/s2213-8587(21)00203-5 [DOI] [PubMed] [Google Scholar]
- 15.Krampe NA Oerline MK Asplin JR, et al. Potential for urolithiasis-related research using the novel medicare-litholink database. Urol Pract. 2023;10(2):147–152. doi: 10.1097/upj.0000000000000378 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Maalouf NM, Poindexter JR, Adams-Huet B, Moe OW, Sakhaee K. Increased production and reduced urinary buffering of acid in uric acid stone formers is ameliorated by pioglitazone. Kidney Int. 2019;95(5):1262–1268. doi: 10.1016/j.kint.2018.11.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Uetake D Ohno I Ichida K, et al. Effect of fenofibrate on uric acid metabolism and urate transporter 1. Intern Med. 2010;49(2):89–94. doi: 10.2169/internalmedicine.49.2597 [DOI] [PubMed] [Google Scholar]
- 18.Pearle MS Goldfarb DS Assimos DG, et al.; American Urological Assocation. Medical management of kidney stones: AUA guideline. J Urol. 2014;192(2):316–324. doi: 10.1016/j.juro.2014.05.006 [DOI] [PubMed] [Google Scholar]
- 19.Centers for Medicare & Medicaid Services. Risk Adjustment; 2022. Accessed April 9, 2024. https://www.cms.gov/Medicare/Health-Plans/MedicareAdvtgSpecRateStats/Risk-Adjustors.html [Google Scholar]
- 20.Diez Roux AV Merkin SS Arnett D, et al. Neighborhood of residence and incidence of coronary heart disease. N Engl J Med. 2001;345(2):99–106. doi: 10.1056/nejm200107123450205 [DOI] [PubMed] [Google Scholar]
- 21.Strope SA, Sarma A, Ye Z, Wei JT, Hollenbeck BK. Disparities in the use of ambulatory surgical centers: a cross sectional study. BMC Health Serv Res. 2009;9:121. doi: 10.1186/1472-6963-9-121 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Oh MS. A new method for estimating G-I absorption of alkali. Kidney Int. 1989;36(5):915–917. doi: 10.1038/ki.1989.280 [DOI] [PubMed] [Google Scholar]
- 23.Ferraro PM Ticinesi A Meschi T, et al. Short-Term changes in urinary relative supersaturation predict recurrence of kidney stones: a tool to guide preventive measures in urolithiasis. J Urol. 2018;200(5):1082–1087. doi: 10.1016/j.juro.2018.06.029 [DOI] [PubMed] [Google Scholar]
- 24.Prochaska M, Taylor E, Ferraro PM, Curhan G. Relative supersaturation of 24-hour urine and likelihood of kidney stones. J Urol. 2018;199(5):1262–1266. doi: 10.1016/j.juro.2017.10.046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Borghi L, Meschi T, Amato F, Briganti A, Novarini A, Giannini A. Urinary volume, water and recurrences in idiopathic calcium nephrolithiasis: a 5-year randomized prospective study. J Urol. 1996;155(3):839–843. doi: 10.1016/s0022-5347(01)66321-3 [DOI] [PubMed] [Google Scholar]
- 26.Borghi L Meschi T Schianchi T, et al. Urine volume: stone risk factor and preventive measure. Nephron. 1999;81(suppl 1):31–37. doi: 10.1159/000046296 [DOI] [PubMed] [Google Scholar]
- 27.Scholtes RA Muskiet MHA van Baar MJB, et al. Natriuretic effect of two weeks of dapagliflozin treatment in patients with type 2 diabetes and preserved kidney function during standardized sodium intake: results of the DAPASALT trial. Diabetes Care. 2021;44(2):440–447. doi: 10.2337/dc20-2604 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Fegan J, Khan R, Poindexter J, Pak CY. Gastrointestinal citrate absorption in nephrolithiasis. J Urol. 1992;147(5):1212–1214. doi: 10.1016/s0022-5347(17)37520-1 [DOI] [PubMed] [Google Scholar]
- 29.Sas KM Kayampilly P Byun J, et al. Tissue-specific metabolic reprogramming drives nutrient flux in diabetic complications. JCI Insight. 2016;1(15):e86976. doi: 10.1172/jci.insight.86976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Schaub JA AlAkwaa FM McCown PJ, et al. SGLT2 inhibitors mitigate kidney tubular metabolic and mTORC1 perturbations in youth-onset type 2 diabetes. J Clin Invest. 2023;133(5):e164486. doi: 10.1172/jci164486 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Hamm LL. Renal handling of citrate. Kidney Int. 1990;38(4):728–735. doi: 10.1038/ki.1990.265 [DOI] [PubMed] [Google Scholar]
- 32.Cameron MA, Maalouf NM, Adams-Huet B, Moe OW, Sakhaee K. Urine composition in type 2 diabetes: predisposition to uric acid nephrolithiasis. J Am Soc Nephrol. 2006;17(5):1422–1428. doi: 10.1681/ASN.2005121246 [DOI] [PubMed] [Google Scholar]
- 33.Pak CY Sakhaee K Moe O, et al. Biochemical profile of stone-forming patients with diabetes mellitus. Urology. 2003;61(3):523–527. doi: 10.1016/s0090-4295(02)02421-4 [DOI] [PubMed] [Google Scholar]
- 34.Novikov A Fu Y Huang W, et al. SGLT2 inhibition and renal urate excretion: role of luminal glucose, GLUT9, and URAT1. Am J Physiol Renal Physiol. 2019;316(1):F173–F185. doi: 10.1152/ajprenal.00462.2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Suijk DLS van Baar MJB van Bommel EJM, et al. SGLT2 inhibition and uric acid excretion in patients with type 2 diabetes and normal kidney function. Clin J Am Soc Nephrol. 2022;17(5):663–671. doi: 10.2215/CJN.11480821 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Park CW Kim HW Ko SH, et al. Long-term treatment of glucagon-like peptide-1 analog exendin-4 ameliorates diabetic nephropathy through improving metabolic anomalies in db/db mice. J Am Soc Nephrol. 2007;18(4):1227–1238. doi: 10.1681/ASN.2006070778 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data cannot be shared. The individual level data are proprietary and cannot be shared publicly. The statistical code and readme files are available at https://doi.org/10.7302/55an-x018 to reconstruct the analysis should others wish to obtain access to the individual level data.