Impact of Sodium-Glucose Cotransporter-2 Inhibitors on Urolithiasis

Introduction

Urolithiasis is a common disease with an estimated global prevalence of up to 15% and a recurrence rate of up to 50%. Diabetes mellitus has been implicated as a risk factor for urolithiasis.¹ The mechanism underlying the increased risk of urolithiasis in patients with diabetes mellitus may involve high uric acid excretion, impaired ammoniagenesis, and a subsequent decrease in urinary pH levels resulting from insulin resistance. Sodium-glucose cotransporter-2 (SGLT2) inhibitors have diuretic and anti-inflammatory effects.² We previously conducted in vivo and in vitro studies showing that SGLT2 inhibitors may suppress renal stone formation via anti-inflammatory effects.³ We also reported that the prevalence of urolithiasis in males with diabetes mellitus was significantly lower in the SGLT2 inhibitor prescription group than in the non-SGLT2 inhibitor prescription group.³ A nationwide cohort study employing Danish health registries also revealed a lower risk of nephrolithiasis in SGLT2 inhibitor users than that in dipeptidyl peptidase-4 inhibitor users.⁴ Moreover, empagliflozin, an SGLT2 inhibitor, reduced urolithiasis risk by approximately 40% in a pooled analysis of randomized clinical trials.⁵ In contrast, a previous report showed that the adverse effects of SGLT2 inhibitors might include increased nephrolithiasis risk.⁶ A clinical trial is underway to assess the impact of empagliflozin on urolithiasis.⁷

Therefore, we investigated the relationship between diabetes mellitus and urolithiasis in Japanese patients. First, we explored the relationship between several antidiabetic medications and the likelihood of developing urolithiasis in patients with diabetes mellitus using large-scale epidemiologic data. Thereafter, we examined the relationship between SGLT2 inhibitors and urolithiasis in patients without diabetes mellitus.

Results

The current cross-sectional study used the Japanese administrative database maintained by MDV analyzer (Medical Data Vision Co., Ltd., Tokyo, Japan), a web tool containing the medical information of patients from Japanese acute Diagnosis Procedure Combination hospitals.

The study population comprised patients with diabetes mellitus, and data collection was conducted from January 1, 2021, to December 31, 2021. Patients with the International Classification of Diseases, 10th revision codes of E10–14 and N20 were defined as having diabetes mellitus and renal or ureteral stones, respectively (Supplementary Table S1). First, we investigated the relationship between diabetes mellitus and the likelihood of developing urolithiasis. Then, we calculated the presence or absence of diabetes mellitus and urolithiasis, and the crude or unadjusted odds ratio (OR). Here, we found that the International Classification of Diseases, 10th revision–assigned urolithiasis was common in patients with diabetes mellitus (OR, 1.12; 95% confidence interval [CI], 1.10–1.13 for male and OR, 1.70; 95% CI, 1.67–1.73 for female) (Table 1). The multivariate relative risk of prevalent urolithiasis in males with diabetes mellitus compared with that in males without diabetes mellitus has been reported to be 1.31 (95% CI, 1.11–1.54),¹ which is consistent with the results of this study. Next, we examined the associations between medication use and urolithiasis in patients with diabetes mellitus. We compared patients who used α-glucosidase inhibitors, biguanides, dipeptidyl peptidase-4 inhibitors, glinides, SGLT2 inhibitors, sulfonylureas, and thiazolidinediones with patients who did not use these medications. A list of each antidiabetic medication constituting the major types of oral antidiabetic medications available on the Japanese market as of December 2021 is shown in Supplementary Table S2. For combined medications, each one was counted as a single medication.

Table 1.

The relationship between diabetes and urolithiasis

Male (n = 5,704,430)		urolithiasis (+)	urolithiasis (−)	OR	95% CI	P value
Diabetes (+)	1,026,574	25,707 (2.50%)	1,000,867 (97.50%)	1.12	1.10–1.13	<0.0001
Diabetes (−)	4,677,856	104,979 (2.24%)	4,572,877 (97.76%)

Female (n = 6,227,050)		urolithiasis (+)	urolithiasis (–)	OR	95% CI	P value
Diabetes (+)	720,418	11,912 (1.65%)	708,506 (98.35%)	1.70	1.67–1.73	<0.0001
Diabetes (−)	5,506,632	53,922 (0.98%)	5,452,710 (99.02%)

CI, confidence interval; OR, crude/unadjusted odds ratio.

Only the use of SGLT2 inhibitor was significantly negatively associated with urolithiasis (OR, 0.95; 95% CI, 0.91–0.98: male and OR, 0.91; 95% CI, 0.86–0.97: female); however, α-glucosidase inhibitors (OR, 1.51; 95% CI, 1.45–1.58: male and OR, 1.70; 95% CI, 1.60–1.81: female), biguanides (OR, 1.20; 95% CI, 1.16–1.23: male and OR, 1.23; 95% CI, 1.18–1.29: female), dipeptidyl peptidase-4 inhibitors (OR, 1.13; 95% CI, 1.11–1.16: male and OR, 1.24; 95% CI, 1.19–1.29: female), glinides (OR, 1.16; 95% CI, 1.10–1.23: male and OR, 1.30; 95% CI, 1.19–1.42: female), sulfonylureas (OR, 1.16; 95% CI, 1.11–1.21: male and OR, 1.26; 95% CI, 1.18–1.35: female), and thiazolidinediones (OR, 1.29; 95% CI, 1.21–1.38: male and OR, 1.33; 95% CI, 1.19–1.49: female) were significantly positively associated with the likelihood of developing urolithiasis (Table 2).

Table 2.

The relationship between antidiabetic medications and urolithiasis in patients with diabetes mellitus

Group	α-glucosidase inhibitor (+)	α-glucosidase inhibitor (−)	OR	95% CI	P value
Male (n = 1,064,443)	2469 (3.67%)	24,474 (2.45%)	1.51	1.45–1.58	<0.0001
Female (n = 742,911)	1107 (2.72%)	11,355 (1.62%)	1.70	1.60–1.81	<0.0001

	Biguanide (+)	Biguanide (−)	OR	95% CI	P value
Male (n = 1,117,154)	5304 (2.90%)	22,765 (2.51%)	1.20	1.16–1.23	<0.0001
Female (n = 775,247)	2162 (1.98%)	10,745 (1.61%)	1.23	1.18–1.29	<0.0001

	DDP4 inhibitor (+)	DDP4 inhibitor (−)	OR	95% CI	P value
Male (n = 1,183,957)	8358 (2.74%)	21,275 (2.42%)	1.13	1.11–1.16	<0.0001
Female (n = 812,738)	3541 (1.96%)	10,027 (1.59%)	1.24	1.19–1.29	<0.0001

	Glinide (+)	Glinide (−)	OR	95% CI	P value
Male (n = 1,056,018)	1312 (2.88%)	25,132 (2.49%)	1.16	1.10–1.23	<0.0001
Female (n = 737,502)	572 (2.12%)	11,649 (1.64%)	1.30	1.19–1.42	<0.0001

	SGLT2 inhibitor (+)	SGLT2 inhibitor (−)	OR	95% CI	P value
Male (n = 1,109,212)	3731 (2.37%)	23,845 (2.50%)	0.95	0.91–0.98	0.0018
Female (n = 762,217)	1163 (1.51%)	11,339 (1.65%)	0.91	0.86–0.97	0.0039

	Sulfonylurea (+)	Sulfonylurea (−)	OR	95% CI	P value
Male (n = 1,068,899)	2182 (2.91%)	24,985 (2.51%)	1.16	1.11–1.21	<0.0001
Female (n = 744,102)	881 (2.06%)	11,477 (1.64%)	1.26	1.18–1.35	<0.0001

	Thiazolidinedione (+)	Thiazolidinedione (−)	OR	95% CI	P value
Male (n = 1,041,666)	925 (3.19%)	25,221 (2.49%)	1.29	1.21–1.38	<0.0001
Female (n = 727,946)	308 (2.18%)	11,764 (1.65%)	1.33	1.19–1.49	<0.0001

CI, confidence interval; DDP4, dipeptidyl peptidase-4; OR, crude/unadjusted odds ratio; SGLT2, sodium-glucose cotransporter-2.

Because several SGLT2 inhibitors are indicated for heart failure and chronic kidney disease in Japan, the study was also validated in patients without diabetes mellitus. In males, the prescription of SGLT2 inhibitors in patients without diabetes mellitus was significantly associated with a low likelihood of developing urolithiasis (OR, 0.42; 95% CI, 0.35–0.51). In contrast, in females, the prescription of SGLT2 inhibitors in patients without diabetes mellitus was not significantly associated with a low likelihood of developing urolithiasis (OR, 0.90; 95% CI, 0.68–1.19) (Supplementary Table S3).

Discussion

The relationship between SGLT2 inhibitor use and the likelihood of developing urolithiasis is topical and forms the research basis for an ongoing prospective observational study.⁷ One systematic review found no significant relationship between SGLT2 inhibitor use and the likelihood of developing urolithiasis.⁸ However, the aforementioned study had a limited sample size because nephrolithiasis was not a prespecified endpoint in the included trials, and the number of nephrolithiasis new onset cases was low. Previous reports have used populations of approximately 10,000 patients; however, the current study used a large database of approximately 1,000,000 patients.^4,5 A novel finding in this study was that SGLT2 inhibitor use predicted a low likelihood of developing urolithiasis, whereas use of other antidiabetic medications was associated with an increased likelihood of developing urolithiasis. Furthermore, SGLT2 inhibitors were significantly associated with a low likelihood of developing urolithiasis in males without diabetes mellitus.

Only SGLT2 inhibitors were associated with a low prevalence of urolithiasis among the oral antidiabetic medications of interest. Renal stones, primarily calcium oxalate stones, are associated with renal inflammation and fibrosis and induce reactive oxygen species–mediated inflammation in renal cells.⁹ Therefore, targeting inflammation might be effective in preventing renal stone formation.³ Diuresis is a possible mechanism by which SGLT2 inhibitors prevent renal stone formation. However, in vivo and in vitro studies suggest that SGLT2 inhibitors prevent renal stone formation via their anti-inflammatory and anti-fibrotic effects.³ A decrease in urinary pH is a risk factor for urolithiasis. SGLT2 inhibitors increase urinary pH by inhibiting Na⁺-H⁺ exchanger activity and bicarbonate reabsorption in the proximal tubules. This is one possible mechanism by which SGLT2 inhibitors prevent urolithiasis.

This study has some limitations. First, we would have liked to examine not only diabetes mellitus but also hypertension and obesity as risk factors for urolithiasis; however, we could not examine comorbidities in detail using our database. Second, considering that this was a cross-sectional study, the causal relationships between the medications and urolithiasis risk could not be established. However, the database used herein has a large number of patients registered in Japan, and only SGLT2 inhibitors are negatively associated with urolithiasis, in contrast to other antidiabetic medications. Therefore, further prospective observational studies are needed to show a causal relationship between SGLT2 inhibitors and urolithiasis.

In conclusion, in patients with diabetes mellitus, SGLT2 inhibitors are associated with a low likelihood of developing urolithiasis, whereas other antidiabetic medications are associated with a high likelihood of developing urolithiasis. In addition, in male patients without diabetes mellitus, SGLT2 inhibitors are associated with a decreased likelihood of developing urolithiasis.

Disclosure

All the authors declared no competing interests.

Supplementary Material

Supplementary Methods.

Supplementary Results.

Supplementary References.

Table S1. International Classification of Diseases, 10th revision codes for diabetes mellitus and urolithiasis.

Table S2. The list of oral α-glucosidase inhibitors, biguanides, DDP4 inhibitors, glinides, SGLT2 inhibitors, sulfonylureas, and thiazolidinediones.

Table S3. The relationship between SGLT2 inhibitor and urolithiasis in nondiabetic patients.