The eGFR dip after starting an SGLT2 inhibitor is the drug working

The commonest reason SGLT2 inhibitors get stopped is a fall in eGFR in the weeks after starting, read as nephrotoxicity. It is the opposite. These drugs reduce intraglomerular pressure by restoring tubuloglomerular feedback, which lowers measured filtration acutely and is the same haemodynamic mechanism that preserves kidney function over years. RAS blockade does the same thing and generates the same misreading.

So set the expectation before you prescribe, not after the result comes back. Tell the patient and whoever will see the repeat blood test that a fall is expected. Check renal function at around four weeks. A dip of up to roughly 30% from baseline is acceptable and does not warrant stopping; it typically stabilises and then the slope of decline is shallower than it would have been. Above that, or if it keeps falling on repeat testing, look for another explanation — volume depletion, an added diuretic or NSAID, obstruction — before blaming the SGLT2 inhibitor.

The other half of this is knowing when to hold. These drugs should be paused during acute illness with volume depletion or reduced intake, and before surgery, because of euglycaemic ketoacidosis risk rather than kidney risk. Confusing those two situations produces both errors at once: stopping the drug permanently for a benign eGFR dip, and continuing it through a vomiting illness where it should have been held.

• Warn the patient and the follow-up clinician that an eGFR fall is expected before you prescribe.
• Check renal function at about 4 weeks; a dip of up to roughly 30% does not warrant stopping.
• If the fall is larger or progressive, look for volume depletion, NSAIDs, diuretics or obstruction first.
• Restart after an interruption; a held drug frequently becomes a stopped drug by default.
• Do pause during acute illness with poor intake and before surgery — for ketoacidosis risk, not kidney risk.

The statistics, in plain English

The acute dip and the long-term benefit are the same physiology measured at two timescales, which is why the trials show curves that cross: the treated group starts lower on eGFR and ends higher, because the slope of decline is shallower. This pattern is consistent across the major SGLT2 inhibitor kidney trials and across RAS blockade trials before them. The roughly 30% threshold comes from trial protocols and long clinical practice rather than from a study designed to find the right cut-off, so treat it as a working rule, not a validated boundary.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.