SGLT2 inhibitors in adults aged 75 and over: cohort finds no increased harm versus DPP-4 inhibitors

The study

SGLT2 inhibitors (SGLT2is) have shown cardiorenal benefits in people with type 2 diabetes. Safety data in older adults are limited. This study, published in Diabetes Care, compared SGLT2is with dipeptidyl peptidase 4 inhibitors (DPP-4is) in people aged 75 years or older.

The authors used administrative health databases from British Columbia, Canada. The design was a population-based, active-comparator, new-user cohort study. People started SGLT2is or DPP-4is between 1 January 2021 and 31 December 2023. The primary outcomes were infection, diabetic ketoacidosis, acute kidney injury, fall, fracture and all-cause hospitalisation. The authors used Cox regression with propensity score-based standardised mortality ratio weighting to estimate hazard ratios (HRs).

What they found

The study included 4,997 SGLT2i initiators (median age 79 years; 37% women) and 1,289 DPP-4i initiators (median age 82 years; 50% women).

Compared with DPP-4i initiation, SGLT2i initiation was associated with:

• a lower risk of acute kidney injury (HR 0.76, 95% CI 0.60-0.95)
• a lower risk of all-cause hospitalisation (HR 0.87, 95% CI 0.77-0.98)

No difference was seen for infection. No differences were observed for diabetic ketoacidosis, falls or fractures, although the confidence intervals for these were compatible with both benefit and harm.

The authors concluded that SGLT2i initiation was not associated with increased risk of harm compared with DPP-4i initiation, although estimates were imprecise for some outcomes.

How much weight to give it

This is an observational cohort study. It can show association, not cause. The authors used weighting to adjust for differences between the groups. The abstract does not list the factors used in the weighting, so residual confounding cannot be excluded.

The two groups differed at the start in age (median 82 years for DPP-4i initiators and 79 years for SGLT2i initiators) and in sex (50% and 37% women). The groups also differed in size (1,289 DPP-4i initiators and 4,997 SGLT2i initiators). The authors describe the estimates as imprecise for some outcomes.

For diabetic ketoacidosis, falls and fractures, a result of "no difference" does not rule out a real effect in either direction, because the confidence intervals were compatible with both benefit and harm. The abstract does not report follow-up length, funding or event counts.

The primary outcomes were safety and healthcare use outcomes. The findings are reassuring but not final.

For your clinic

These data may help when you consider an SGLT2 inhibitor for a patient aged 75 or older. In this cohort, starting one was not associated with more infection or hospital admission than starting a DPP-4 inhibitor, and it was associated with less acute kidney injury and hospital admission. The authors say the findings may support considering SGLT2is as a treatment option in this age group.

Do not treat this as proof of safety for every older patient. The cohort was drawn from a Canadian administrative database. The abstract does not describe individual patient factors such as frailty or kidney function. Read the full paper before changing practice, and check current guidelines and product information for your own patients.