Acute Effects of Acarbose and Canagliflozin, Alone and in Combination, on Postprandial GLP ‐1 and Glycaemic Responses in Post‐Bariatric Hypoglycaemia: A Randomised Crossover Study

ABSTRACT Aim To evaluate the acute effects of acarbose and canagliflozin, alone and in combination, on postprandial total GLP‐1 and glycaemic responses during a mixed meal test in adults with post‐bariatric hypoglycaemia (PBH). Materials and Methods This single‐centre, randomised, four‐period crossover physiological study included 12 adults with PBH after Roux‐en‐Y gastric bypass. Participants underwent meal tests after placebo, acarbose 100 mg, canagliflozin 300 mg or combination treatment. Co‐primary outcomes were total GLP‐1 area under the concentration–time curve (AUC) from 0 to 240 min and peak total GLP‐1 concentration. Secondary outcomes included glucose, insulin, C‐peptide, GIP and glucagon. Results Eleven participants completed all four visits and all 12 were analysed. During the placebo mixed meal test, 3/12 participants reached glucose < 3.0 mmol/L, while 10/12 reached < 3.5 mmol/L and 10/12 reported postprandial symptoms suggestive of hypoglycaemia. All active treatments reduced peak glucose and glucose excursion versus placebo, without changing glucose nadir or glucose AUC 0–240 min. Peak insulin, insulin AUC 0–240 min, peak C‐peptide and C‐peptide AUC 0–240 min were lower with both acarbose‐containing regimens than with placebo. Total GLP‐1 AUC 0–240 min did not differ between treatments, while peak total GLP‐1 was higher with canagliflozin than with the other treatments. Peak GIP was lower with acarbose‐containing regimens than with placebo, and GIP AUC 0–240 min was lower with combination treatment than with placebo. Peak glucagon was higher with combination treatment than with placebo. Conclusions Acarbose‐containing regimens attenuated postprandial beta‐cell responses, but adding canagliflozin did not provide clear additive benefit over acarbose alone on the prespecified GLP‐1 outcomes or glucose nadir. Changes in glycaemic and beta‐cell responses without a parallel reduction in total GLP‐1 suggest that mechanisms beyond GLP‐1 secretion contribute to the acute postprandial phenotype in PBH.