Summary: In eight adults with type 1 diabetes using automated insulin delivery, three smaller isomaltulose servings taken during cycling held plasma glucose steadier (change +1.7 mmol/L) than a single isocaloric serving taken 90 minutes before exercise (change -3.2 mmol/L; p<0.001). The during-exercise approach also blunted incretin, glucagon, and insulin responses and shifted fuel use toward fat. This was a small mechanistic crossover, not a clinical outcomes trial.
PICO Summary
| Element | Detail |
|---|---|
| Population | Eight adults with type 1 diabetes on automated insulin delivery (5 female; mean age 47 years; mean HbA1c 8.3%, 67.5 mmol/mol; diabetes duration 23 years). Randomized two-period crossover; single laboratory centre (Denmark). |
| Intervention | Three isocaloric isomaltulose servings (0.25 g carbohydrate/kg each) taken during 45 minutes of submaximal cycling, with no bolus insulin (DEC). n=8 (within-subject). |
| Comparison | A single isomaltulose serving (0.75 g carbohydrate/kg) taken 90 minutes before exercise with a 25% bolus insulin reduction (PEC). n=8 (within-subject). |
| Outcome | Exercise-induced plasma glucose change diverged between conditions: PEC -3.2 ± 1.2 mmol/L versus DEC +1.7 ± 1.5 mmol/L (p<0.001); rate of change -0.07 versus +0.04 mmol/L/min (p<0.001). During exercise, GLP-1, GIP, glucagon, and total insulin were all lower with DEC (all p≤0.02). Carbohydrate oxidation was lower (p=0.009) and lipid oxidation higher (p=0.014) with DEC. No effect estimates with 95% confidence intervals, ARR, or NNT were reported; outcomes are physiological surrogates, not clinical events. |
Isomaltulose Timing Around Exercise in T1D
RCT crossover · type 1 diabetes · 45-min cycling
Splitting isomaltulose into smaller servings taken during exercise held plasma glucose steady (+1.7 mmol/L), while a single pre-exercise serving let it fall (-3.2 mmol/L). A small mechanistic crossover, not a clinical outcomes trial.
Expert Commentary
This small randomized crossover suggests that, for adults with type 1 diabetes on automated insulin delivery, spreading isomaltulose into smaller servings taken during exercise produces steadier plasma glucose than front-loading a single dose beforehand. The signal is internally consistent: glucose was defended during exercise, and the hormonal and fuel-oxidation shifts point in a biologically coherent direction. The verdict, however, must be read as mechanistic and hypothesis-generating rather than as proof of better clinical control. Only eight participants completed a single bout of cycling, so the findings are fragile, the confidence intervals are necessarily wide, and no patient-centred endpoints such as time in range over days, hypoglycaemia frequency, or symptom burden were measured. The design was unavoidably open-label, since feeding timing cannot be masked, and that introduces behavioural and expectancy effects. The two conditions also differed in more than timing, because the pre-exercise arm included a bolus reduction while the during-exercise arm used none, so timing and insulin dosing are partly confounded. Can I use this with my patients? Cautiously, as a conversation starter for a motivated person on a pump who drops low during steady aerobic activity, but not yet as a firm protocol. Larger studies with free-living glucose monitoring and harder endpoints are needed before this becomes guidance.
References
McCarthy OM, Christensen MB, Tawfik S, Kristensen KB, Hartmann B, Holst JJ, et al. Metabolic and Hormonal Responses to Isomaltulose Ingestion Before or During Sustained Submaximal Exercise in Adults with Type 1 Diabetes Using Automated Insulin Delivery Systems. Nutrients. 2024;16(23):4098. doi:10.3390/nu16234098