Lithium Concentrations After Initiation or Discontinuation of SGLT2 Inhibitors in Psychiatric Inpatients
Pierre Duranleau-Gagnon, Laurent Béchard, Katryn Langlois-Parisé, Marie-France Demers, Olivier CorbeilBackground:
Sodium-glucose cotransporter 2 inhibitors (SGLT2i) reduce cardiovascular and renal risk, but coprescription with lithium raises concern because lithium elimination is linked to renal sodium handling. Prescribing information recommends lithium monitoring after SGLT2i initiation or dose changes, although clinical evidence remains limited, particularly after discontinuation. We describe lithium concentrations and clinical evolution after SGLT2i initiation or discontinuation in lithium-treated psychiatric inpatients.
Methods:
We conducted a retrospective case series at a psychiatric hospital. Pharmacy records identified patients who received lithium and an SGLT2i between January 1, 2014 and March 31, 2026. Patients were included if the SGLT2i was initiated or discontinued while lithium was continued and ≥1 lithium concentration was available before and after modification. Clinical pharmacists reviewed records for lithium dose and concentrations, SGLT2i exposure, renal function, interacting medications, acute illness, hydration status, and adverse events.
Results:
Six cases were identified: 3 initiations and 3 discontinuations. Lithium concentrations showed no consistent directional change. After initiation, concentrations remained stable or changed minimally in 2 cases and increased slightly in one. After discontinuation, lithium remained stable in an older patient with chronic kidney disease, increased slightly in 1 case confounded by lithium dose escalation, and increased from 0.6 to 0.8 mmol/l in 1 case complicated by pneumonia, diarrhea, and dehydration. No case required urgent lithium discontinuation, toxicity treatment, or prolonged hospitalization attributed to lithium.
Conclusions:
SGLT2i initiation or discontinuation was not associated with consistent or clinically concerning lithium changes. These observations suggest coprescription may be manageable with individualized monitoring when clinically indicated.