Effects of supplementing with an 18% carbohydrate-hydrogel drink versus a placebo during whole-body exercise in −5 °C with elite cross-country ski athletes: a crossover study

Participants attended the laboratory on five separate occasions, firstly completing body composition measurements then a preliminary exercise trial, a familiarization and two experimental trials. They were instructed to abstain from alcohol and to perform only moderate-intensity exercise the day before the preliminary exercise and experimental trials. The preliminary exercise trial was performed in order to determine the submaximal work- \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂ relationship, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂peak and maximal heart rate (HR_max). The familiarization trial was used to identify the individual treadmill speeds required to elicit ~ 70% of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂peak, as well as to familiarize the athletes with the temperature, equipment and procedures used during the two experimental trials. The two experimental trials were conducted using a double-blind, randomized, crossover design and consisted of 120 min of submaximal diagonal-style roller-skiing in −5 °C, followed immediately by a maximal double-poling performance test in ~ 20 °C using a ski ergometer. Immediately prior to and throughout the 120-min submaximal exercise bout participants received either a ¹³C-enriched 18% carbohydrate-hydrogel drink (CHO-HG) or a placebo (PLA), which was designed to mimic the texture and sweetness of the CHO-HG drink.

Following an overnight fast, participants were weighed (Seca 764, Hamburg, Germany) in their underwear and body composition was assessed by dual-energy X-ray absorptiometry (iDXA; GE Medical Systems, Madison, WI, USA). The iDXA was calibrated according to the manufacturer’s guidelines before each measurement. Total lean and relative fat percentages were analyzed using enCore software (version 16.10).

Participants performed an incremental test consisting of four to five, 4-min submaximal stages on a motor-driven treadmill (Rodby Innovation AB, Vänge, Sweden) using the skate roller-skiing technique. The roller skis (Pro-Ski S2, Sterners, Dala-Järna, Sweden) were pre-warmed in order to standardize the rolling resistance and participants wore a safety harness around the waist connected to an automatic emergency brake above the treadmill. The submaximal test was followed by 4 min of active recovery, 5 min of passive recovery and a 5-min active re-warm-up including three, 10–15-s self-paced high-intensity intervals. The maximal test followed, which consisted of 900-m and 1000-m self-paced time-trials for the females and males, respectively. Pulmonary gas exchange was measured throughout both the submaximal and maximal tests using a metabolic cart (AMIS 2001 model C, Innovision A/S, Odense, Denmark) equipped with a flowmeter. The gas analyzers were calibrated with a high-precision two-component gas mixture of 16.0% O₂, and 4.0% CO₂ (Air Liquide, Kungsängen, Sweden). Calibration of the flowmeter was performed with a 3 L air syringe (Hans Rudolph, Kansas City, MO, USA) for low, medium and high flow rates. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙CO₂, and ventilation rate were monitored continuously, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂ values were calculated from 10-s epochs and reported as 30-s averages.

A familiarization trial was used to acquaint the participants with the test procedures and to determine individual treadmill speeds for the subsequent experimental trials. Participants performed a continuous 32-min submaximal effort in an environmental chamber set to −5 °C. To control ambient conditions the chamber utilized a hypoxia controller (Hypoxico, New York, USA), which was set to ‘sea level’ (20.9% O₂), and a customized air-conditioning system controlling room temperature with a stated precision of ±0.5 °C. The exercise was performed using the diagonal-stride technique and classic roller-skis (Pro-ski C2, Sterners, Dala-Järna, Sweden) on a motor-driven treadmill (Rodby Innovation AB, Vänge, Sweden) fixed at a 5° incline. The starting speed was based on the submaximal work- \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙ O₂ relationship derived from the preliminary exercise trial, with continuous adjustments made to the treadmill speed until heart rate (HR) stabilized at an intensity corresponding to ~ 70% of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂peak (equivalent to mean ± standard deviation [SD] 82 ± 3% of HR_max). Treadmill speed was reduced to 4 km·h^{− 1} for 30 s after 10 min and every 20 min thereafter, as well as for 90 s after 20 min and every 20 min thereafter, during the familiarization and experimental trials. The 30-s recovery periods allowed participants to change sub-technique and therefore movement pattern, which is not usually fixed for long durations during XC skiing and was considered a potential injury risk. The 90-s recovery periods again allowed for this precautionary alteration in movement pattern, but also enabled fingertip blood sampling, psychometric data collection and consumption of the drink solution.

Following the 32-min submaximal exercise bout participants exited the environmental chamber, were given the opportunity to empty their bladder, then removed any surplus clothing and changed from ski boots to indoor training shoes. They then completed the TT in ~ 20 °C using a ski ergometer (SkiErg, Concept2, Morrisville, VT, USA). The reasons for moving to room temperature were twofold: 1. The display on the ski ergometer did not function reliably at sub-zero temperatures; 2. The coaches and athletes were uncomfortable performing maximal exercise at sub-zero temperatures at this point in the season, due to the increased risk for harm to the airways associated with high ventilation rates in the cold. The TT was a self-paced, double-poling performance test lasting 2000 m for females and 2400 m for males, with the flywheel resistance set at 6 and 8, respectively. The protocol was designed to simulate the muscular work and physiological responses involved in a biathlon competition, which consists of three or five high-intensity bouts of skiing, each lasting up to 8 min [24]. Participants were instructed to complete the set distance as quickly as possible and no encouragement or feedback was provided, except that distance remaining was visible throughout. Given the elite level of the athletes, and the regularity with which they perform intensive double-poling ergometer exercise as a part of their habitual training, one familiarization trial was considered sufficient to establish reproducibility during the subsequent experimental trials.

Athletes arrived at the laboratory at a standardized time (either 06:00 or 09:00) for their two experimental trials, which were separated by 6 ± 1 (mean ± SD) days. Upon arrival at the laboratory, BM was recorded (Seca 764, Hamburg, Germany) and after resting in a seated position for ~ 5 min, a fingertip blood sample was collected for the subsequent analysis of glucose and lactate concentrations (Biosen C-line, EKF diagnostic GmbH, Magdeburg, Germany). Participants then entered the environmental chamber (−5.0 ± 0.2 °C; Kestrel 5500 Weather Meter, Nielsen-Kellerman Company, Boothwyn, PA, USA), where they received their first drink (220 mL of CHO-HG or PLA) before the onset of exercise (time = 0 min).

The CHO-HG drink provided 2.2 g of CHO·min^{− 1} (132 g·h^{− 1}) in a ratio of 1:0.8 maltodextrin:fructose and had an osmolality of 750 mOsm·kg^{− 1}. Each serving (~ 220 mL) contained 200 mL of water, 44 g of CHO, 0.3 g of NaCl, 0.3 g of sodium alginate and 0.2 g of pectin. In the PLA drink, the CHO was replaced by 0.92 g of erythritol and 20 mg of sweetener (sodium sacharinate, sucralose, L-leucin) per serving, while the amount of fluid, salt and gelling components (i.e., water, NaCl, sodium alginate and pectin) remained the same. Both the CHO-HG and PLA drinks were supplied by Maurten AB (Gothenburg, Sweden), and in vitro tests in simulated gastric acid confirmed gelation of both solutions. The maltodextrin (Cargill Nordic A/S) and fructose (Tate & Lyle Sweden AB) were corn-derived with a ¹³C enrichment of − 11.45‰ and − 11.51‰ vs. Pee Dee Bellemnitella (PDB), respectively. The CHO-HG drink was enriched in ¹³C content by adding U-¹³C glucose and U-¹³C fructose (Cambridge Isotope Laboratories, MA, USA) in proportions 1:0.8 and corresponding to 0.487 per mille of the total CHO content. The ¹³C enrichment of the CHO-HG drink reinforced with U-¹³C glucose and U-¹³C fructose was + 28.00‰ vs. PDB.

The 120-min submaximal exercise bout involved diagonal-style roller-skiing and was performed at a constant incline of 5° and a treadmill speed of 9.7 ± 0.2 km·h^{− 1} for the males and 8.5 ± 0.3 km·h^{− 1} for the females. As described for the familiarization trial, treadmill speed was reduced to 4 km^{− 1} every 10 min to allow for a change of sub-technique and movement pattern. In addition, every 20 min, during the 90-s recovery periods, a fingertip blood sample and overall rating of perceived exertion (RPE; Borg category scale 6–20) was collected. Severity of five GI symptoms (gas, nausea, stomach rumbling, urgency to have a bowel movement and abdominal pain) were also rated on a 0–20 scale (0 = no symptoms, 10 = neutral, 20 = worst conceivable symptoms), and a level of digestive comfort was provided (0 = extremely uncomfortable, 10 = neutral, 20 = extremely comfortable) [26]. Following these measurements participants consumed 220 mL of CHO-HG or PLA before the treadmill speed was increased again at the end of the 90-s period.

Following the 120-min submaximal exercise the participants performed a TT, as described for the familiarization trial. Immediately after completion of the TT, subjective RPE, GI symptoms and level of digestive comfort measures were recorded. A fingertip blood sample was collected 3 min after the TT and subsequently analyzed for glucose and lactate concentrations, as described previously. Post-exercise BM was then measured and total loss in BM, used to represent sweat loss and respiratory water losses, was determined by subtracting post-exercise BM from pre-exercise BM. Heart rate was monitored continuously at 5-s intervals throughout the diagonal-skiing and double-poling trials (M400, Polar Electro Oy, Kempele, Finland) and mean values for each minute were subsequently calculated.

Expired air was collected during the 120-min submaximal exercise bout in 170-L Douglas bags (C Fritze Consulting, Svedala, Sweden) for 35 s per sample after 17.5 min of each 20-min period (i.e., 2–2.5 min prior to reducing the treadmill speed). After collecting each sample the Douglas bags were immediately removed from the environmental chamber and placed on a bag stand in a thermoneutral room and analyzed the same day, following the exercise trials. The fractional concentrations of O₂ were determined with an S-3A oxygen analyzer and CO₂ concentrations were determined with a CD 3-A carbon dioxide analyzer with a P-61B infrared sensor (AEI Technologies Inc., Pittsburgh, PA, USA). Expired gas volume was measured with a 170-L spirometer (Fabri, Spånga, Sweden) with a fast-responding temperature sensor (Greissinger, Würzburg, Germany) attached to the top of the inner cylinder. For the measurement of ¹³C/¹²C in the expired CO₂, two smaller expired gas samples were drawn from each Douglas bag into 65-mL syringes (Kendall, Monoject, UK) connected via a 3-way valve. The samples were then infused into two 12 mL vials (Labco Ltd., Lampeter, UK) for later analysis.

The breath samples were analyzed for ¹³CO₂/¹²CO₂ enrichment (δ¹³C) using a Thermo Scientific Delta Ray isotope ratio infrared spectrometer (IRIS) with a Universal Reference Interface (URI) and a Teledyne CETAC ASX-7100 autosampler. Every two samples were bracketed by calibrating gas (δ ¹³C 27.8 ‰ VPDB). The ¹³C enrichment of beverage content was determined using a Costech Elemental Analyzer (ECS 4010; Costech International, Pioltello, Italy) in continuous flow mode coupled to a Thermo Scientific Delta V plus (ThermoFisher Scientific, Bremen, Germany) isotope ratio mass spectrometer (Friedrich-Alexander-Universität, Erlangen, Germany). All isotope ratio data were normalized to the Vienna Pee Dee Belemnite (VPDB) scale.

Rates of total CHO and fat oxidation (g·min^{− 1}) during the submaximal exercise were calculated from \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂ and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙CO₂ (L·min^{− 1}) using the following stoichiometric equations, [27] with the assumption that protein oxidation during exercise was negligible: 1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \mathrm{CHO}\ \left(\mathrm{g}\cdotp {\min}^{-1}\right)=\left(4.585\times \dot{\mathrm{V}}{\mathrm{CO}}_2\right)-\left(3.226\times \dot{\mathrm{V}}{\mathrm{O}}_2\right) $$\end{document}CHOg·min−1=4.585×V˙CO2-3.226×V˙O22\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \mathrm{Fat}\ \left(\mathrm{g}\cdotp {\min}^{-1}\right)=\left(1.695\times \dot{\mathrm{V}}{\mathrm{CO}}_2\right)-\left(1.701\times \dot{\mathrm{V}}{\mathrm{O}}_2\right) $$\end{document}Fatg·min−1=1.695×V˙CO2−1.701×V˙O2

The isotopic enrichment of the ingested glucose and fructose was expressed as the ‰ difference between the δ¹³C/¹²C ratio of the sample and a known laboratory reference standard [28]: 3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\updelta}^{13}\mathrm{C}=\left(\left(\frac{13C/12C\ sample}{13C/12C\ standard}\right)-1\right)\cdotp {10}^3 $$\end{document}δ13C=13C/12Csample13C/12Cstandard−1·103

The δ¹³C was then related to an international standard (VPDB). In the CHO-HG trial, the rate of exogenous oxidation was calculated using the formula of Mosora et al. [29]: 4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \mathrm{Exogenous}\ \mathrm{CHO}\ \mathrm{oxidation}\ \left(\mathrm{g}\cdotp {\mathit{\min}}^{-1}\right)={VCO}_2\times \left(\frac{\updelta \mathrm{Exp}-{\updelta \mathrm{Exp}}_{bkg}}{\delta Ing-{\updelta \mathrm{Exp}}_{bkg}}\right)\left(\frac{1}{k}\right) $$\end{document}ExogenousCHOoxidationg·min−1=VCO2×δExp−δExpbkgδIng−δExpbkg1kwhere δExp is the ¹³C enrichment of expired CO₂ during exercise, δIng is the ¹³C enrichment of the CHO-HG solution, δExp_bkg is the ¹³C enrichment of expired air in the PLA trial and k (0.7467) is the amount of CO₂ (L·min⁻¹) produced for the complete oxidation of 1 g of glucose. A methodological limitation when calculating exogenous CHO oxidation rates from expired ¹³CO₂ is the retention of ¹³CO₂ in the circulating bicarbonate pool [30]. To take this slow equilibration process into account, and hence the delayed appearance of ¹³C in the breath, the computations were only made during the last 60 min of exercise.

All data were checked for normality using the Shapiro-Wilk test. Independent t-tests were used for between-group comparisons (e.g., sex differences), while paired samples t-tests were used for within-group comparisons (e.g., TT performance and post-TT measurements). A two-way analysis of variance (ANOVA) with repeated measures was performed on all participants (n = 12) to assess differences in breath ¹³C enrichment, RER, substrate oxidation, blood markers and perceptual variables (e.g., RPE and GI symptoms) over time between the two trials (CHO-HG and PLA). A three-way mixed design factorial ANOVA considering time × trial × sex was used to identify differences in metabolic and perceptual variables. Substrate oxidation rates are, unless stated otherwise, expressed as a percent of LBM (g·min^{− 1}·kg LBM^{− 1}·10^{− 2}). Total CHO (CHO_total), exogenous CHO (CHO_exo), endogenous CHO (CHO_endo) and fat (FAT) oxidation rates, as well as RER, were calculated over the last 60 min of exercise. Breath ¹³C enrichment, blood glucose and lactate concentrations, RPE and GI symptoms were calculated over the whole 120-min exercise bout, including pre-exercise (at rest). Bonferroni post-hoc adjustments were used to identify the location of significant differences when the ANOVA yielded a significant F ratio. Analyses were adjusted by use of the Greenhouse-Geisser correction where necessary. Partial Eta-squared (_pη²) was calculated as a measure of effect size for the ANOVA, where values of 0.01, 0.06, and 0.15 were considered as small, medium, and large, respectively [31]. Cohen’s d (d) was calculated as a measure of effect size for pairwise comparisons, where values of 0.2, 0.5, and 0.8 were considered as small, medium, and large, respectively [31]. Results are presented as mean ± SD and statistical significance was set at P < 0.05. All statistical analyses were conducted using SPSS for Windows version 25 (Chicago, Illinois, USA).

Six of the 12 participants correctly guessed the drink solutions (CHO-HG and PLA), while the remaining six guessed incorrectly.

It is well established that compared with a control (i.e., a placebo or water), exogenous CHO provision during prolonged exercise increases total CHO oxidation, decreases fat oxidation and reduces the oxidation of endogenous CHO, and that these alterations in substrate metabolism are entirely attributed to the oxidation of ingested (i.e., exogenous) CHO [21–23]. Further, in contrast to glucose-only feedings during exercise, blood lactate concentration is known to increase in response to fructose ingested within multiple-transportable CHO solutions [3, 32]. The results from the current study support these previous findings, whereby CHO-HG ingestion led to significant increases in blood lactate concentration and total and exogenous CHO oxidation, as well as decreases in endogenous CHO and fat oxidation, in comparison to the PLA trial.

Consuming 2.2 g·min^{− 1} of a 0.8:1 maltodextrin:fructose hydrogel solution in the current study led to a peak exogenous CHO oxidation rate of 1.33 g·min^{− 1} (range 0.89–1.66 g·min^{− 1}) after 120 min. Ingesting 2.4 g·min^{− 1} of a 1:1 glucose:sucrose solution (i.e., 1:0.3 glucose:fructose) has previously been demonstrated to elicit a mean peak oxidation rate of 1.20 g·min^{− 1} at the end of 120 min of exercise at ~ 63% of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂max among cyclists [4]. Two related studies demonstrated peak oxidation rates of 1.70 and 1.75 g·min^{− 1}, respectively, after 150 min of exercise at ~ 60–62% of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂max when ingesting glucose and fructose at 1:0.6 and 1:1 ratios, respectively [3, 32]. Differences in exogenous oxidation rates between the current study compared with those previously reported by Jentjens and colleagues likely resides from differences in the experimental protocols (i.e., amount and type of CHO ingested, and exercise duration). Furthermore, with no plateau observed towards the end of exercise, it may be assumed that the peak exogenous CHO oxidation rate would have exceeded 1.33 g·min^{− 1} in the present study if the submaximal exercise bout had continued beyond 120 min.

Gastrointestinal discomfort is considered to be a limiting factor in moderate- to high-intensity exercise (i.e., ≥ 60% \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂max) lasting ≥ 2 h, and symptoms might be further exacerbated by dehydration and excessive CHO intake [2, 33]. In order to test the potential effects of CHO-HG on GI symptoms, and concomitantly maximize CHO_exo oxidation, a CHO ingestion rate of 2.2 g·min^{− 1} was selected in the present study. This is in excess of current CHO intake guidelines, which recommend up to 1.5 g·min^{− 1} [2]. The CHO solution provided the participants with a similar amount of fluid (i.e., 600 mL·h^{− 1}) previously shown to be ingested during competition by elite XC ski athletes in cold conditions [34]. Despite the high CHO concentration (18%), no differences in GI discomfort or level of digestive comfort were observed in CHO-HG compared to PLA. These findings might be due to the cold ambient conditions during the 120-min submaximal exercise, which has been shown to decrease the incidence and severity of GI symptoms compared to hot conditions [33]. Furthermore, mechanical causes of GI symptoms, such as shaking of the intra-abdominal contents, would likely be reduced during XC skiing compared to running, for example. Thus, future research efforts might address whether the high GI tolerability observed following CHO-HG ingestion can be attributed to the added gelling polysaccharides per se. As well as delineating specific mechanisms on the gastric and intestinal behaviors in response to hydrogel exposure, comparisons with an isocaloric CHO-only control intake under gut-challenging prolonged, high-intensity exercise in different ambient conditions is warranted.

Contrary to the hypothesis, no ergogenic effect was observed for TT performance following CHO ingestion in the current study. This is in contrast to most [35, 36], but not all [37], previous placebo-controlled CHO studies employing similar protocols in terms of the duration and intensity of submaximal exercise (e.g., 105–120 min at ~ 70% \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂max) and subsequent performance tests (e.g., lasting 8–15 min). Plausible explanations for this discrepancy may relate to differences in the exercise modalities used and the training status of the participants. For example, diagonal XC skiing was used in the present study, which, in contrast to the majority of previous studies where cycling protocols have been employed, involves whole-body exercise with the upper body generating ~ 50–75% of the propulsive power output during moderately-intense exercise [38]. Since a substantial portion of the work done to propel the XC skier forwards during the submaximal exercise bout would have been performed by the lower-body, it is possible that endogenous CHO availability was still adequate in the upper-body musculature in PLA to meet the high energy demands of the subsequent ~ 8.4-min double-poling time-trial. That the power output profile patterns did not differ between PLA and CHO-HG, including an increase in power output during the final 20% of the total TT distance (possibly relating to an anaerobic energy reserve), supports this contention that CHO would have still been locally available in the upper body even towards the end of the TT in the PLA trial. However, although muscles of the upper limbs have been shown to be the primary working muscles involved in double poling at lower-exercise intensities, an increasing involvement of the torso, hip and leg muscles is evident at higher exercise intensities [39].

Although muscle glycogen content was not measured in the current study, it is possible that the submaximal exercise was not demanding enough to deplete endogenous glycogen stores in this specific group of elite athletes. A recent meta-analysis [40] of skeletal muscle glycogen utilization concluded that ~ 120 min of exercise at 70% of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂max initiated with normal muscle glycogen content (i.e., 400 mmol∙kg^{− 1} dry weight [d.w.]) would lead to the attainment of critical muscle glycogen levels (i.e., 250–300 mmol∙kg^{− 1} d.w.), which have been associated with reductions in peak power output [9]. However, the majority of participants in the present study were world-class endurance athletes and would likely have a far superior capacity for oxidizing fat and sparing CHO when exercising for a prolonged period of time at this intensity [41]. Future research investigating the impact of CHO supplementation on performance and determinants of fatigue related to substrate utilization during XC skiing with elite skiers and biathletes should aim to increase the duration and/or intensity of the submaximal preload. Moreover, including a TT with the same sub-technique and/or repeated bouts of high-intensity exercise would also allow for a closer simulation of real-world competition demands.

As well as being the first study to investigate exogenous CHO oxidation during XC skiing (i.e., whole-body exercise), the present study is also the first to examine sex differences when ingesting a multiple-transportable CHO solution during exercise. Compared to PLA, CHO-HG ingestion reduced the reliance on endogenous CHO oxidation over the final hour of exercise in both sexes by ~ 18%, which is comparable to reductions previously observed for females and males (~ 15%) in two studies using high ingestion rates (1.5–2 g·min^{− 1}) of glucose only [22, 23]. In the present study, endogenous CHO oxidation contributed ~ 28 and 32% to the total energy yield in the CHO-HG trial for females and males, respectively. In contrast, Riddell et al. [21] showed that the relative endogenous CHO oxidation to the total energy yield was significantly higher in females (~ 14%) than in males (~ 5%). However in that study the ingestion rate was based on BM (1 g glucose·kg BM·h^{− 1}), resulting in ~ 1.0 and 1.3 g CHO·min^{− 1} for the females and males, respectively. On balance, the current and previous studies suggest that when ingesting the same absolute amount of a single- or multiple-transportable CHO, the relative reduction in endogenous CHO oxidation to total energy contribution appears to be similar between the sexes.

Regarding exogenous CHO oxidation, the current and previous studies [20–23] indicate that the relative contribution to total energy expenditure is consistently, although not necessarily significantly, ~ 2–4% higher in females than in males. However, when expressed in absolute terms (g·min^{− 1}), sex differences in exogenous CHO oxidation have showed mixed results. M’Kaouar et al. [20] reported that females oxidized ~ 33% less exogenous CHO compared with males (~ 0.6 versus 0.9 g·min^{− 1}) during 120 min of cycling exercise at ~ 65% of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙O₂max. By contrast, other studies have shown no significant sex differences in absolute exogenous CHO oxidation when cycling for 90–120 min at 57–67% of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \dot{\mathrm{V}} $$\end{document}V˙ O₂max [21–23]. The females in the current study tended (P = 0.064) to oxidize ~ 20% less exogenous CHO than the males during the last hour of exercise (~ 1.2 versus 1.5 g·min^{− 1}). However, the novel data presented in the current study have demonstrated that females have the capacity to substantially increase CHO_exo oxidation when fed a multiple-transportable CHO solution at a high ingestion rate, with observed peak oxidation rates up to 1.61 g CHO·min^{− 1}, well in excess of SGLT1 transporter saturation (i.e., ~ 1 g·min^{− 1}). Due to the small sample sizes used in the current and previous studies (i.e., n = 6–8), further research employing larger samples is necessary to assess whether there is indeed a sex difference in exogenous CHO oxidation following the ingestion of multiple-transportable CHO solutions.

The novel approaches and strengths of this study include the use of an innovative multiple-transportable carbohydrate hydrogel during exercise under conditions where energy requirements (CHO in particular) are expected to be high, and sweat rates low (e.g., whole-body exercise in the cold). Moreover, the involvement of a familiarization trial and the standardized dietary preparation ensured that conditions were controlled between participants and trials. Perhaps most noteworthy, though, is the unusually high level of the participating athletes, most of whom were world-class (with half winning Olympic and World Championship medals in the year of data collection), as well as the mixed-sex nature of the sample.

A number of limitations in the study design should, however, be acknowledged. For example, a CHO control without additional gelling polysaccharides was not administered, and neither was a non-polysaccharide placebo. This was due to the nature of the sample group (i.e., a national team in preparation for an Olympic Games only 6 months away), so prescribing additional long-duration and highly-controlled trials was not possible. Therefore, the experimental solution (CHO-HG) and a placebo with gelling agents but no CHO were prioritized. In addition to this, and a low within-sex sample size, the menstrual phase of the female participants (which may influence substrate oxidation) was not controlled for. However, while ovarian hormones might affect metabolic regulation during exercise [16] results are conflicting [15] and variability in substrate metabolism seems more likely due to between- and within-subject variations than the menstrual-cycle phase. Furthermore, participants were provided with pre-exercise CHO, which has previously been shown to negate the effects of menstrual cycle phase on glucose kinetics by reducing the demand on endogenous glucose production [42]. Three out of six female participants in this study were using hormonal contraceptives, which have also been suggested to alter fat and CHO metabolism during exercise [43]. However, the evidence for this is unclear as no differences in fuel utilization during prolonged exercise were observed between females taking and not taking oral contraceptives [22].

Additional file 1. Exogenous carbohydrate oxidation during exercise in the CHO-HG (carbohydrate-hydrogel) trial in (A) females (n = 6) and (B) males (n = 6). The thick black line represents the group mean and the thin black lines represent individual responses.Additional file 2. Ratings of perceived exertion (RPE) and perceptions of gastrointestinal symptoms after the double-poling time-trial (n = 12).