Design: Systematic review and meta-analysis of 85 randomised, double-blind trials
Population: 70,543 adults with CVD, diabetes, heart failure or CKD
Primary outcome: Change in haemoglobin and haematocrit
Effect: Haemoglobin +0.60 g/dL (95% CI 0.55–0.66); haematocrit +2.36 points vs control

This JAMA meta-analysis pooled 85 double-blind randomised trials, 70,543 participants with cardiovascular disease, diabetes, heart failure or kidney disease, followed a median 24 weeks.

Against control, SGLT2 inhibitors raised haemoglobin by 0.60 g/dL (95% CI 0.55 to 0.66) and haematocrit by 2.36 percentage points. Erythropoietin rose slightly and transferrin saturation fell by 0.84 points, consistent with stimulated red cell production drawing on iron.

The authors call the clinical significance unclear. In practice the main value is interpretive: a rising haemoglobin or haematocrit after starting these drugs is expected, and a mild fall in iron indices may follow. This matters in heart failure and kidney disease, where anaemia and iron deficiency are common and monitored.

Expect haemoglobin to rise by about 0.5–0.6 g/dL after starting an SGLT2 inhibitor.
A rising haematocrit on these drugs is usually a drug effect, not haemoconcentration alone.
Check iron indices periodically in heart failure and kidney disease, as transferrin saturation fell slightly.
Consider the SGLT2 effect when interpreting anaemia work-up or erythropoietin dosing in CKD.

Why it matters

It explains a laboratory change clinicians see often and may otherwise misread.

The statistics, in plain English

The confidence interval is narrow because the trials are many and large, so the 0.6 g/dL estimate is precise. Statistically significant does not mean clinically important; the authors themselves say the clinical meaning is unclear.