Across 32 studies, hypoglycaemia lengthened the heart rate-corrected QT interval by a pooled 30.2 ms versus euglycaemia (95% CI 17.9-42.4). Clamp-induced episodes produced 44.0 ms of prolongation, spontaneous ones 16.4 ms, and type 1 and type 2 diabetes looked identical (about 27 ms each). Epinephrine surges tracked QTc change across studies (ρ = 0.87), while serum potassium fell 0.44 mmol/L. Mean hypoglycaemic QTc exceeded 450 ms in six studies.
The direction of the effect is consistent, but the size is not: heterogeneity reached I² = 98.4%, and the prediction interval (-40.9 to 101.3 ms) means a future study could plausibly show anything from no effect to marked prolongation. Clamp studies use deliberately engineered hypoglycaemia, so the 44 ms figure overstates what most patients experience day to day. The epinephrine correlation is a between-study association, not proof of mechanism, though it fits the established picture of adrenergic drive and hypokalaemia disturbing cardiac repolarisation. Importantly, this is a surrogate outcome. Nothing here shows that QTc changes during hypoglycaemia cause arrhythmia or sudden death, although they are a plausible link to the 'dead-in-bed' phenomenon and the excess cardiovascular mortality seen with severe hypoglycaemia. The finding is confirmatory rather than paradigm-shifting. Practically, it strengthens the case for avoiding hypoglycaemia, particularly in people taking insulin or sulfonylureas who also have known QT problems or use QT-prolonging drugs such as certain antipsychotics, macrolides or antiarrhythmics. Those patients should discuss glucose targets and medication choices with their clinician.