Twenty-one male recreational endurance runners (age: 26.2 ± 5.8 yrs., height: 179.2 ± 5.0 cm, body mass: 70.3 ± 5.9 kg, VO2peak: 59.5 ± 6.0 ml·min− 1·kg− 1) participated in this study. Prior to all testing, the medical history of all participants was assessed through a standardised questionnaire and a resting ECG was reviewed by a cardiologist to ensure all participants were healthy and physically fit to complete the experimental trial. Participants were informed about possible risks of study procedures and provided their written informed consent prior to inclusion into the study. The study was conducted in accordance with the declaration of Helsinki and approved by the university’s ethical committee (09/2020).
This double-blind randomized-crossover study consisted of four separate testing sessions: A preliminary ramp test to assess peak oxygen uptake (VO2peak) and maximal treadmill running speed (Vmax), as well as three separate experimental trials. The experimental trials were performed in a randomised and counterbalanced order and consisted of a 70-min constant workload test followed by a time to exhaustion test (TTE). The trials were performed after ingesting either isomaltulose, maltodextrin or glucose (Fig. 1). Randomisation was performed by technical staff not involved in data collection and both participants and test personnel were blinded to the experimental conditions. Blinding was only removed after data collection and analysis was completed.
To assess Vmax and VO2peak, a ramp test was performed on a treadmill with a starting inclination of 1%, to reflect the energetic cost of outdoor running . After a 2 min warmup at 2.4 m·s− 1, the test started at 2.4 m·s− 1 and increased by 0.2 m·s− 1 every minute. If participants reached 5.2 m·s− 1, the incline was increased by 1° for each increment. Spirometric data was recorded breath by breath and interpolated for values for each second (Metalyzer® 3B; Cortex Biophysik GmbH, Leipzig, Germany), while heart rate (Polar H7 Sensor; Polar Electro, Kempele, Finland) was recorded every second. The spirometer was calibrated weekly with a reference gas (5% CO2 and 15% O2) and before each test with ambient laboratory air and with a 3-l syringe according to the manufacturer’s specifications. Participants were verbally encouraged to reach voluntary exhaustion. Vmax was defined as the highest increment completed, while for additional degrees of inclination 0.2 m·s− 1 was counted. VO2peak was defined as the highest 30-s moving average oxygen uptake.
Nutritional intake was standardised by meal replacement 24 h prior to each trial, according to the recommendations of the German Society for Nutrition to allow for comparison between conditions. Nutrition was calculated upon a daily requirement of 35 kcal∙kg− 1 (fat: 0.9 g∙kg− 1, carbohydrates: 5.2 g∙kg− 1, protein: 1.5 g∙kg− 1), participants received 60.5 ± 4.7 g fat, 362.7 ± 31.4 g carbohydrates and 106.5 ± 9.7 g protein. Participants were also not allowed to take any over the counter supplements during the study period. Water was allowed ad libitum the day before and after the constant load trial. Testing was carried out in the morning and laboratory visits were separated by at least 72 h. Participants reported to the lab after an overnight fast and were provided with 50 g of either maltodextrin (100% Maltodextrin Carbs, Myprotein, Cheshire, UK), isomaltulose (Risulose, Evonik Creavis GmbH, Marl, Germany) or glucose (100% Glucose Carbs, Myprotein, Cheshire, UK) in 400 ml of water. This quantity was chosen according to recommendations previously outlined, showing that 50 g of isomaltulose was well gastrointestinally tolerated and did not alter gastric emptying rate , while larger amounts may reduce performance due to signs of gastrointestinal discomfort . Similarly, during initial pilot testing in our lab, we found 50 g of isomaltulose to be well tolerated, while larger amounts led to strong symptoms of gastrointestinal discomfort and consequently to failing to completing the exercise session.
After 30 min, a first venous blood sample was drawn (pre) and the constant load trial commenced. The trials consisted of 10 min warm-up at 60% Vmax followed by 60 min at 70% Vmax. Every 10 min the treadmill was stopped for 1 min for capillary blood sampling. Additionally, rate of perceived exertion (RPE) and gastrointestinal discomfort were recorded on a 1-10 scale. After the constant load test, another venous blood sample was taken (post). Following 15 min of passive recovery, participants performed a time to exhaustion test (TTE) at 85% Vmax and a final venous blood sample was taken (pTTE). Spirometric data was recorded breath by breath and interpolated for each second for both tests (Metalyzer 3B, Cortex Biophysik GmbH, Leipzig, Germany). Fat and carbohydrate (CHO) oxidation rates were calculated for the constant load trial for each 10-min block from VO2 and VCO2 data according to the calculations by Peronnet and Massicotte (1991) .
Blood sampling and analysis
Capillary blood samples (20 μl) were drawn from the earlobe into hemolyzing solution cups (EKF Diagnostic Sales, Magdeburg, Germany). Blood lactate and glucose concentrations were measured using the EKF Biosen S-Line Analyser (EKF Diagnostics GmbH, Barleben, Germany). Additionally, venous blood samples were drawn from the antecubital vein into serum separation tubes (BD, Plymouth, UK). After clotting for 10 min at room temperature, serum separation tubes were centrifuged at 1000 g at room temperature (Heraeus® Multifuge® 3 L-R, Kendro Laboratory Products, Newton, USA). Immediately after centrifugation, serum was separated into aliquots and stored at − 80 °C for further analysis. Serum insulin and glucose-dependent insulinotropic polypeptide (GIP) were assessed using the Insulin ELISA Kit (EIA-2935; DRG Instruments GmbH, Marburg, Germany) and Human GIP (Total) ELISA Kit (EZHGIP-54 K; Merck KGaA, Darmstadt, Germany). Samples were analysed in duplicate using a microplate reader (Multiscan™ FC; Thermo Scientific™, Waltham, USA) and the mean was used for statistical analysis.
Calculations and statistical analysis
All data are presented as mean ± standard deviation (SD), apart from percent change, where mean and [95% confidence intervals] are reported. Statistical analysis was performed using SPSS 27.0 (SPSS, IBM Statistics, New York, US). Residual histograms, residual plots and Q-Q-plots were visually checked for homoscedasticity and normality prior to statistical analysis. Incremental area under the curve (AUC) was calculated for VO2, fat oxidation, CHO oxidation and RER using the trapezoid rule. Glucose fluctuation was calculated as the maximal difference in absolute glucose values assessed during the constant load trial (i.e. maximum glucose concentration - minimum glucose concentration). For better visualization, we additionally expressed the glucose fluctuation as percentage. Baseline differences (glucose, insulin, GIP) and differences between conditions (AUC VO2, AUC fat oxidation, AUC CHO oxidation, AUC RER, glucose fluctuation and gastrointestinal discomfort) were tested using a one-way analysis of variances (ANOVA). For time and interaction effects, a mixed factorial analysis of covariance (ANCOVA) was performed with Bonferroni correction for post-hoc tests. For this purpose, measurement times (i.e. minutes 0-70 during the constant load trial for blood glucose, VO2, fat and CHO oxidation rates and RER or pre, post and pTTE for insulin and GIP) were defined as within-group variables and isomaltulose, maltodextrin and glucose ingestion as between-condition variable. Effect sizes for main effects of the ANOVA and ANCOVA were reported as partial η2. To assess associations between changes in glucose, GIP and insulin concentrations across all conditions, Pearson product-moment correlation coefficients r were calculated. For all tests, statistical significance was accepted at p < 0.05.