Design: Systematic review and meta-analysis of 85 double-blind randomised trials
Population: 70,543 adults with diabetes, heart failure, CKD or cardiovascular disease
Primary outcome: Change in haemoglobin and haematocrit
Effect: Haemoglobin +0.60 g/dL (95% CI 0.55–0.66); haematocrit +2.36% (2.19–2.52) vs control

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

Haemoglobin rose by 0.55 g/dL on SGLT2 inhibitors and fell slightly on control, a difference of 0.60 g/dL (95% CI 0.55–0.66). Haematocrit rose by 2.36 percentage points more than control. Erythropoietin rose and transferrin saturation fell slightly (−0.84 percentage points).

The authors say the clinical significance is unclear. The practical point is interpretive: a modest rise in haemoglobin after starting an SGLT2 inhibitor is an expected drug effect, not polycythaemia to investigate or haemoconcentration from dehydration by default. The small fall in transferrin saturation is a reminder to check iron when anaemia coexists.

Expect haemoglobin to rise by about 0.5–0.7 g/dL after starting an SGLT2 inhibitor
Do not read a modest haematocrit rise on these drugs as dehydration or polycythaemia without other evidence
Check iron studies in patients with anaemia, as transferrin saturation fell slightly
In CKD patients on erythropoiesis-stimulating agents, factor the drug into dose reviews

Why it matters

It explains a routine lab change that otherwise prompts unnecessary work-up or fluid advice.

The statistics, in plain English

The 95% confidence interval of 0.55 to 0.66 g/dL is narrow, so the size of the rise is well estimated. Statistically certain is not the same as clinically important: 0.6 g/dL is small for any one patient.