SGLT2 Inhibitors Cut Kidney Stone Risk by 28% in Type 2 Diabetes Patients
A meta-analysis of 10 studies finds SGLT2 inhibitors significantly lower kidney stone risk in type 2 diabetes — adding another metabolic benefit to this drug class.
Summary
A new systematic review and meta-analysis pooling data from 10 studies found that SGLT2 inhibitors — popular diabetes drugs like empagliflozin, dapagliflozin, and canagliflozin — were associated with 28% lower odds of developing kidney stones in adults with type 2 diabetes. The protective effect was strongest compared to non-users (45% lower odds) and also held against placebo and other diabetes drug classes including GLP-1 receptor agonists and DPP-4 inhibitors. While the findings are consistent and statistically robust, extremely high heterogeneity across studies cautions against a purely causal interpretation. Most evidence comes from observational data, and researchers call for future trials with rigorous stone outcome tracking and urinary biochemical profiling to confirm the mechanism.
Detailed Summary
Kidney stones affect roughly 10% of adults and are more common in people with type 2 diabetes — a population already burdened by cardiovascular, renal, and metabolic complications. Understanding whether widely used diabetes drugs influence stone risk carries both clinical and quality-of-life significance, particularly as kidney stone episodes are painful, costly, and can accelerate renal decline with age.
This systematic review and meta-analysis searched PubMed, Embase, and CENTRAL through April 2026, identifying 10 comparative studies contributing 14 unique drug comparisons. The primary outcome was nephrolithiasis incidence. Pooled effect estimates were calculated using restricted maximum likelihood random-effects models, and subgroup analyses stratified results by comparator drug class.
Across all comparisons, SGLT2 inhibitors were associated with a 28% reduction in kidney stone odds (OR 0.72, 95% CI 0.65–0.80). The strongest benefit appeared versus non-user controls (OR 0.55), followed by placebo (OR 0.65), DPP-4 inhibitors (OR 0.74), and GLP-1 receptor agonists (OR 0.79). Results were consistent regardless of study design or how stone outcomes were identified. Comparator class explained roughly 30% of the between-study variability, and follow-up duration did not significantly modify the association.
The plausible mechanism involves SGLT2 inhibitor-induced changes in urinary chemistry: reduced urine glucose lowers the risk of calcium oxalate and uric acid stone formation, while osmotic diuresis may dilute lithogenic solutes. These drugs also lower uric acid levels systemically — a known stone-forming substrate.
For clinicians managing type 2 diabetes patients with a history of nephrolithiasis, these findings add a meaningful secondary benefit to an already compelling drug class. However, the I² of 97.5% signals substantial heterogeneity, and the predominantly observational evidence base limits causal claims. Adjudicated stone outcomes and urinary biochemistry data in future randomized trials are needed to solidify recommendations.
Key Findings
- SGLT2 inhibitors associated with 28% lower kidney stone odds across all comparators (OR 0.72) in type 2 diabetes.
- Strongest protection vs. non-users (OR 0.55, 45% lower odds); benefit also held vs. placebo and GLP-1 agonists.
- Results were consistent across both randomized and observational study designs.
- Comparator drug class explained ~30% of between-study variability, highlighting active comparator bias.
- High heterogeneity (I² = 97.5%) and observational evidence base limit causal conclusions.
Methodology
The meta-analysis included 10 studies (14 comparisons) identified from PubMed, Embase, and CENTRAL through April 2026. Restricted maximum likelihood random-effects models were used to pool odds ratios, with subgroup analyses by comparator class, study design, and outcome ascertainment method. Heterogeneity was assessed using I² statistics and 95% prediction intervals.
Study Limitations
Substantial heterogeneity (I² = 97.5%) means pooled estimates must be interpreted cautiously, and the 95% prediction interval crosses 1.0, indicating the effect may not hold in all clinical contexts. The evidence base is largely observational, limiting causal inference. Summary is based on the abstract only, as the full text is not open access; urinary biochemical mechanisms and individual drug-level data could not be evaluated.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: