Presentation (CTI)

OBJECTIVES: Sodium-glucose co-transporter-2 inhibitors (SGLT-2i) and dipeptidyl peptidase-4 inhibitors (DPP-4i) are widely prescribed glucose-lowering therapies, but their comparative safety regarding acute disturbances in sodium and glucose homeostasis remains uncertain. Because SGLT-2i induce glycosuria, osmotic diuresis, and natriuresis, they may theoretically increase the risk of dysnatraemia, while the two drug classes may differ in their propensity for hypoglycaemia and short-term adverse outcomes. We emulated a target trial comparing SGLT-2i with DPP-4i across various outcomes in patients admitted to University Hospitals Birmingham NHS Foundation Trust: hyponatraemia (serum sodium <130 mmol/L), hypernatraemia (>145 mmol/L), hypoglycaemia (blood glucose <4.0 mmol/L), hyperglycaemia (>11.0 mmol/L), acute kidney injury (AKI), post-prescription length of stay, and 30-day all-cause mortality.
METHODS: For each outcome, the primary analysis excluded patients with the outcome of interest at baseline to assess incident events. Propensity score matching was conducted to balance the exposed and control groups. Cox proportional-hazards models estimated adjusted hazard ratios (HRs) with 95% confidence intervals (CIs), controlling for demographic, socioeconomic, clinical, and medication-related factors. A sensitivity analysis retained patients with the outcome present at baseline.
RESULTS: In the primary analysis, SGLT-2i use was associated with lower risks of hyponatraemia (HR 0.59, 95% CI 0.49-0.72), hypoglycaemia (HR 0.82, 95% CI 0.72-0.94), hyperglycaemia (HR 0.70, 95% CI 0.65-0.77), and 30-day mortality (HR 0.77, 95% CI 0.62-0.94) compared with DPP-4i use. Associations with hypernatraemia (HR 0.80, 95% CI 0.62-1.04) and AKI (HR 0.75, 95% CI 0.32-1.75) favoured SGLT-2i but were not statistically significant. SGLT-2i use was also associated with shorter hospital stay (IRR 0.86, 95% CI 0.83-0.90). The sensitivity analysis including patients with the outcome at baseline left the direction of effect unchanged and preserved the statistical conclusions for every outcome assessed.
CONCLUSIONS: Overall, SGLT-2i demonstrated a favourable safety profile, with lower risks of acute glycaemic and sodium-related events, shorter hospitalisation, and reduced risk of short-term mortality.