1A-02 - SGLT2 Inhibitors, GLP-1 Receptor Agonists, and DPP4 Inhibitors and the Risk of Hypomagnesemia in Type 2 Diabetes: A Multi-Institutional Cohort Study with Target Trial Emulation

Background: Clinical trials have suggested that sodium-glucose co-transporter 2 (SGLT2) inhibitors increase serum magnesium levels. However, whether this translates into a reduced long-term risk of hypomagnesemia in real-world clinical settings remains unclear. Additionally, comparative data on the risk of hypomagnesemia among different classes of antidiabetic agents are limited.

Objectives: To compare the risk of hypomagnesemia among adults with type 2 diabetes treated with SGLT2 inhibitors, glucagon-like peptide-1 receptor agonists (GLP-1 RAs), or dipeptidyl peptidase-4 (DPP4) inhibitors.

Methods: We emulated a target trial using a multi-institutional electronic health record database (TriNetX) to assess the risk of hypomagnesemia among patients initiating SGLT2 inhibitors, GLP-1 RAs, or DPP4 inhibitors between 2016 and 2023. Three pairwise comparisons were conducted: (1) SGLT2 inhibitors vs. DPP4 inhibitors, (2) GLP-1 RAs vs. DPP4 inhibitors, and (3) SGLT2 inhibitors vs. GLP-1 RAs. A 1:1 propensity score matching approach was applied to balance baseline characteristics. Patients were followed from treatment initiation until hypomagnesemia (defined as an ICD-10 diagnosis of E83.42 or serum magnesium levels < 1.80 mg/dL), death, loss to follow-up, or study end (September 14, 2024). Cox proportional hazards models were used to estimate the relative risk of hypomagnesemia.

Results: After propensity score matching, 316,614 matched pairs were included in the SGLT2 inhibitor vs. DPP4 inhibitor cohort, 290,069 in the GLP-1 RA vs. DPP4 inhibitor cohort, and 290,065 in the SGLT2 inhibitor vs. GLP-1 RA cohort. Baseline characteristics were well balanced.

Initiation of SGLT2 inhibitors was associated with a significantly lower risk of hypomagnesemia compared with both DPP4 inhibitors (hazard ratio [HR]: 0.81, 95% confidence interval [CI]: 0.80–0.82) and GLP-1 RAs (HR: 0.87, 95% CI: 0.86–0.89). Similarly, GLP-1 RA use was associated with a lower risk of hypomagnesemia compared with DPP4 inhibitors (HR: 0.92, 95% CI: 0.91–0.93). These associations were consistent across subgroups stratified by age, sex, and hemoglobin A1c levels. A reduced risk of hypomagnesemia was observed across individual agents within the SGLT2 inhibitor (canagliflozin, dapagliflozin, empagliflozin) and GLP-1 RA (dulaglutide, liraglutide, semaglutide) classes compared with DPP4 inhibitors.

Conclusions: Among patients with type 2 diabetes, treatment with SGLT2 inhibitors and GLP-1 RAs was associated with a lower risk of hypomagnesemia compared with DPP4 inhibitors. These findings suggest that SGLT2 inhibitors and GLP-1 RAs may be preferable options for patients at risk of hypomagnesemia.