Aims: There has been uncertainty whether SGLT2 inhibition predisposes to hyperkalaemia or is protective from it. We therefore performed a meta-analysis to assess effects of SGLT2 inhibition on serum-potassium and hyperkalaemia-events in T2DM.

Methods: MEDLINE and PubMed databases were searched for 'hyperkalaemia' or 'potassium', with SGLT2 inhibitors in T2DM, to 31st December 2020. Randomised controlled trials, with potassium or hyperkalaemia as primary or secondary outcomes, were included. Cochran's Q test and I2 statistic assessed statistical heterogeneity. Meta-analyses were performed using Cochrane-RevMan with two outcomes: i) Odds ratio (OR) of hyperkalaemia-events between SGLT2 inhibitor and placebo (fixed-effects), ii) Mean difference (MD) in change from baseline potassium between SGLT2 inhibitor and placebo (random-effects).

Results: Of 1724 identified publications, nine were included in the meta-analysis (n=3 hyperkalaemia event; n=5 serum-potassium; n=1 reported both outcomes). Pooled OR for hyperkalaemia-events for SGLT2 inhibitor vs placebo was 0.72 [95% confidence interval (CI) 0.61 to 0.85, P<0.001], I2 of 9%. The pooled MD in serum-potassium concentration with SGLT2 inhibitor vs placebo was -0.04 mmol/L [95% CI -0.08 to 0.00 mmol/L; P=0.04], I2 of 89%.

Conclusions: Use of SGLT2 inhibitors in T2DM reduced odds of inducing hyperkalaemia but had a minimal effect of lowering serum potassium.

Keywords: SGLT2; diabetes mellitus; hyperkalaemia; nephrology; potassium.