For Flux Sake: Isotopic Tracer Methods of Monitoring Human Carbohydrate Metabolism During Exercise

Javier T. Gonzalez; Andy J. King

doi:10.1123/ijsnem.2022-0170

Jump to Content Jump to Main Navigation

For Flux Sake: Isotopic Tracer Methods of Monitoring Human Carbohydrate Metabolism During Exercise

in International Journal of Sport Nutrition and Exercise Metabolism

Click name to view affiliation

Javier T. Gonzalez

Javier T. GonzalezCenter for Nutrition, Exercise and Metabolism, University of Bath, Bath, United Kingdom
Department for Health, University of Bath, Bath, United Kingdom

Search for other papers by Javier T. Gonzalez in
Current site
Google Scholar
PubMed

https://orcid.org/0000-0002-9939-0074^*

and

Andy J. King

Andy J. KingExercise and Nutrition Research Program, The Mary Mackillop Institute for Health Research, Australian Catholic University, Melbourne, VIC, Australia

Search for other papers by Andy J. King in
Current site
Google Scholar
PubMed

DOI:: https://doi.org/10.1123/ijsnem.2022-0170

Keywords:: stable isotopes; glucose; kinetics; physical activity; nutrition

First Published Online:: 29 Nov 2022

In Print:: Volume 33: Issue 1

Page Range:: 60–70

Open access

Get Citation Alerts

Download PDF

Isotopic tracers can reveal insights into the temporal nature of metabolism and track the fate of ingested substrates. A common use of tracers is to assess aspects of human carbohydrate metabolism during exercise under various established models. The dilution model is used alongside intravenous infusion of tracers to assess carbohydrate appearance and disappearance rates in the circulation, which can be further delineated into exogenous and endogenous sources. The incorporation model can be used to estimate exogenous carbohydrate oxidation rates. Combining methods can provide insight into key factors regulating health and performance, such as muscle and liver glycogen utilization, and the underlying regulation of blood glucose homeostasis before, during, and after exercise. Obtaining accurate, quantifiable data from tracers, however, requires careful consideration of key methodological principles. These include appropriate standardization of pretrial diet, specific tracer choice, whether a background trial is necessary to correct expired breath CO₂ enrichments, and if so, what the appropriate background trial should consist of. Researchers must also consider the intensity and pattern of exercise, and the type, amount, and frequency of feeding (if any). The rationale for these considerations is discussed, along with an experimental design checklist and equation list which aims to assist researchers in performing high-quality research on carbohydrate metabolism during exercise using isotopic tracer methods.

In studies of metabolism, tracer methods offer the ability to track the fate of substrates such as carbohydrates, lipids, and proteins. While several tracer methods exist for a variety of applications, for example, dyes for tracking energy into feces (Jumpertz et al., 2011), this review specifically focuses on the theory and practice of employing isotopic tracers to study carbohydrate metabolism during exercise in humans. Tracers can be considered as “labels” which are used to track the trace of interest (e.g., ingested glucose). A common tracer method employs isotopes, which are elements that have the same atomic number (Z) but differ in mass number (A) due to the number of neutrons present. The three key principles that make isotopes useful as tracers are:

(a)rarity; can be distinguished from what is already present in a system
(b)almost identical chemical and functional properties to their tracee; to adequately track the tracee with minimal disturbance to metabolism
(c)different physical properties to their tracee; allowing detection and distinction

Since isotopes have the same number of protons and electrons (which carry charge), their chemical properties are assumed to be identical, yet they can be distinguished from one another based on physical properties (Rennie, 1999). These are two fundamental points of tracer methods, since it supports the principle that the tracer behaves (biochemically) identically to the tracee but can be also distinguished from the tracee. In exercise research, carbon and hydrogen isotopes are commonly used to assess carbohydrate metabolism (Figure 1). While isotopes of hydrogen and oxygen exist and are commonly used in combined form in the exercise sciences for energy expenditure applications (e.g., doubly labeled water), there are some benefits to the use of individual carbon and hydrogen isotopes for measuring the oxidation of carbohydrates. For example, the appearance of carbon isotopes in breath carbon dioxide, and the relatively high natural abundance of the carbon-13 (¹³C) isotope, which can be exploited to obtain measures of carbohydrate metabolism without needing to add a tracer (covered in more detail later). Approximately 98.9% of carbon on earth has an atomic mass of 12, known as carbon-12 (¹²C). This is because ¹²C has six protons and six neutrons (which each have mass), and six electrons (which have negligible mass). Approximately 1.1% of carbon on earth is ¹³C which has an extra neutron, and therefore, greater mass than ¹²C. However, as neutrons are chargeless particles, the chemical properties and interactions are assumed to be identical. ¹³C is a stable isotope since it is not radioactive. Carbon-14 (¹⁴C), however, is much rarer than ¹³C and is a radioisotope (radioactive) due to its nuclear decay (to ¹³C), and thus, can be measured by the radioactivity rather than mass. ¹⁴C isotopes are widely used in metabolic research, however in recent years, human exercise studies tend to favor ¹³C over ¹⁴C isotopes due to the inherently lower risk of stable isotopes. Many of the core principles apply to both stable and radioisotopes, and for further reading on radioisotopes, readers are referred to other sources (Wolfe & Chinkes, 2004).

When using stable isotopes, the proportion of tracer within a tracer–tracee mixture is known as the enrichment, which can be expressed in various ways, including the tracer-to-tracee ratio, atom percent excess or mole percent excess (Kim et al., 2016). All stable isotopes fractions are referenced to international standards, which for ¹³C is the Vienna Pee-Dee Belemnite (¹³C/¹²C = 0.0112372). The unusually high ¹³C content of this reference comes from the original sample of marine fossil (Pee-Dee Belemnite) having extremely high ¹³C content. However, the sample was exhausted, and the International Atomic Energy Agency adopted a reference standard of the same composition. The atomic abundance of a sample is often expressed in ‰ (per mil) and noted as δ¹³C (delta or difference in ¹³C enrichment in parts per thousand from the international standard). When adding tracers to a drink or food for ingestion, the δ¹³C enrichment can increase from approximately −25‰ to at least + 100‰ (Barber et al., 2020; Hearris et al., 2022; King et al., 2018; Rowe et al., 2022). However, even the natural variation in the enrichment/abundance of ¹³C can be used under some experimental conditions to assess aspects of carbohydrate metabolism without the need to specifically add (enrich with) tracers.

Some carbohydrates are naturally higher in ¹³C, which can be high enough to be used in a similar way to enriched tracer methods without the need to add further isotopes as a tracer. The higher ¹³C enrichment in some carbohydrates is due to the ways in which plants photosynthesize. For example, plants known as C4 plants fix carbon from atmospheric CO₂ in such a way that they tend to have a higher ¹³C enrichment than plants known as C₃ plants (Jahren et al., 2006). Many carbohydrates, including those from sugar beet, potato, and wheat, therefore, have a naturally lower ¹³C enrichment than sugarcane or corn, typically δ¹³C −25‰ versus δ¹³C −11 δ‰, respectively (Jahren et al., 2006). These can then be used to assess some aspects of carbohydrate metabolism during exercise without the need to add specific tracers, as described in more detail later.

In addition to different elements and types of isotopes, carbohydrate molecules can be enriched with atomic isotopes at specific positions of the molecule (isotopomers), or in all the possible positions of the molecule. An example is illustrated in Figure 2 for glucose, where the tracee glucose is depicted alongside (a) a glucose molecule where ¹³C is at the C-1 position (known as [1-¹³C]-glucose) and (b) a glucose molecule where all the carbons are ¹³C (known as uniformly labeled; [U-¹³C]-glucose). The choice for the researcher depends on the study design and the outcomes of interest. [U-¹³C]-glucose is often used when enriching an orally ingested carbohydrate, since the chances of all the carbons in endogenous glucose being labeled with ¹³C is extremely unlikely. In contrast, [6,6²H₂]-glucose is often used as an infused tracer since there is almost no chance that both of the ²H atoms will be recycled back into the same C-6 position that they started in after glycolysis and gluconeogenesis (Wajngot et al., 1989). This principle of nonrecycling is important to distinguish the tracer from endogenously produced tracee from metabolic processes. Similarly, depending on the location of the labeling, single carbon atom enriched molecules (e.g., [4-¹³C] glucose) may not represent the full extent of carbohydrate metabolism due to the cleaving of carbons during glycolysis (Metallo et al., 2009).

Strengths and Limitations of Isotope Tracers

Stable isotopes do not emit ionizing radiation, and thus, pose no additional risk to the participant than ingesting the tracee, which is a potential hazard of working with radioisotopes. It should be noted, however, that the dose of ionizing radiation with radioisotope tracers can be very low in many study designs, especially in fully grown adults, and thus, the true risk could be deemed justifiable under some circumstances (Rennie, 1999). The limitations of stable isotopes primarily relate to the cost of purchasing the isotopically enriched substrates and the cost of the equipment required to measure the enrichment of samples. These can limit the questions some studies may be able to answer, and researchers should maximize the specificity of the tracer and tracee of interest. For example, glucose tracers tend to be less expensive than maltodextrin or starch tracers. The cost of starch tracers can be prohibitively expensive if a study requires feeding large numbers of people large amounts of carbohydrate over a prolonged period. While the gold standard would be to use a tracer that is identical to the tracee, there may be some scenarios where a different tracer may be acceptable depending somewhat on the study design, established physiology and outcomes. For example, a glucose tracer may be appropriate for tracing some glucose polymers such as maltodextrin in some conditions as the α1,4-glycosidic bond in maltodextrin is hydrolyzed almost immediately in the gastrointestinal tract, liberating free glucose. Thereafter, responses to glucose polymers such as maltodextrin can sometimes closely agree with glucose ingestion, such as glycemic responses and exogenous oxidation rates (Moodley et al., 1992). However, an identical tracer to tracee should be the priority, as there may be unknown nutrient–nutrient interactions or other effects producing differences in digestion due to a study design, or the research involves the study of a food matrix. Indeed some studies have demonstrated a lack of concordance between metabolic responses of maltodextrin and free glucose during exercise (Saris et al., 1993). Some have achieved this by growing wheat in an environment where the atmospheric carbon dioxide was enriched with ¹³C, which meant that when the wheat photosynthesized, they fixed the ¹³C into the starch within the wheat, producing an intrinsically enriched wheat starch for human exercise metabolism research (Folch et al., 2001).

Using ¹³C carbohydrates at high natural abundance (e.g., in the region of δ¹³C −11‰) has the advantage of being less expensive, but tradeoffs include the relatively small tracer-to-tracee ratio. To help increase the tracer-to-tracee ratio there are at least three aspects of study design that can be considered. First, participants can consume a diet low in ¹³C for several days ahead of a trial to reduce their endogenous carbohydrate ¹³C enrichment (Morrison et al., 2000). Second, the dilution of endogenous ¹³C in tissues can be accelerated by a bout of glycogen depleting exercise prior to the low ¹³C diet (though this will not greatly reduce ¹³C enrichment of adipose tissue). Third, participants can complete an additional background correction trial (discussed in more detail below). These additions do, however, add extra participant burden and need to be balanced with the cost and necessity of using a stable isotope.

The analysis of isotopes also requires specialist equipment and expertise. A common method for assessing the ¹³C (or ²H) enrichment of glucose in plasma is gas chromatography–mass spectrometry (GC–MS), but ¹³C enrichments can also be determined at lower enrichments with gas chromatography coupled with isotope ratio mass spectrometry (GC–C–IRMS; Jeukendrup et al., 1999). Isotope ratio mass spectrometry is also used for assessing enrichment of ¹³C in exhaled CO₂ samples. This equipment can be expensive and resource intensive to maintain, although newer technologies such as isotope ratio infrared spectrometry are becoming available (Sutehall et al., 2021). For ¹⁴C enrichment, the β rays that are emitted are detected in a photomultiplier tube. Moreover, the study design and analysis of the enrichment data require further expertise and understanding to obtain accurate, physiologically meaningful data, without violating key assumptions.

Applications of Different Tracers for Carbohydrate Metabolism During Exercise

Dilution Model—Blood Glucose Appearance and Disappearance Rates

A prominent stable isotope used for the dilution model of blood glucose kinetics is deuterium-labeled glucose (e.g., [6,6²H₂]-glucose), whereas for incorporation models [U¹³C]-glucose is commonly the tracer employed (see later). Circulating glucose concentrations are maintained within a relatively tight range, which is a remarkable feat of physiological regulation during exercise. Insufficient glucose availability can lead to performance impairment (and even coma if not rectified), and excess glucose can induce oxidative damage and modification of proteins. With exercise, the demand of glucose can increase rapidly and drastically as muscle glucose uptake (rate of disappearance) rises from less than 0.5 mmol/min at rest to more than 3 mmol/min within 10 min of cycling-based exercise at 200 W (Wahren et al., 1978). The rate of appearance (R_a) of glucose must, therefore, increase to match the increased rate of disappearance (R_d) if glucose concentrations are to remain stable. In the fasted state, this is almost entirely met by increased hepatic glycogenolysis (Gonzalez & Betts, 2019).

With exercise studies, a common approach to calculating blood glucose R_a and R_d is the continuous infusion method which is based on the dilution model (Edinburgh et al., 2018; Kjaer et al., 1986; van Loon et al., 2001). In this paradigm, a continuous intravenous infusion of a glucose tracer (normally [6,6²H₂]-glucose) is initiated until the enrichment of the blood glucose pool reaches a steady state. This can be accelerated by providing a priming dose when initiating the continuous infusion. The rate of infusion is dictated at the lower end by the sensitivity of enrichment measurement, where sufficient enrichment is required to detect the signal from the noise; and at the upper end by a rate which would perturb metabolism. The equations for calculating R_a and R_d for glucose concentrations mainly rely on three key variables: time, glucose concentration, and glucose enrichment. During steady-state situations where R_a is equal to R_d (e.g., overnight fasted and at rest), only the infusion rate and the isotopic enrichment of blood glucose are needed to calculate R_a (Table 1).

Table 1

Equations for Plasma Glucose Kinetics and Substrate Oxidation Rates

Model	Measure	Equation
Steady-state dilution (e.g., fasted, rest)	${\dot{R}}_{a}$ (μmol·kg⁻¹· min⁻¹)	$= \frac{F}{E}$
Steady-state dilution (e.g., fasted, rest)	${\dot{R}}_{d}$ (μmol·kg⁻¹· min⁻¹)	= R_a
Non-steady-state, single-pool model (e.g., postprandial and exercise)	${\dot{R}}_{a} total$ (μmol·kg⁻¹·min⁻¹)	$= \frac{F - V \cdot (\frac{C_{2} + C_{1}}{2}) \cdot (\frac{E_{2} - E_{1}}{t_{2} - t_{1}})}{(\frac{E_{2} + E_{1}}{2})}$
	${\dot{R}}_{d}$ (μmol·kg⁻¹·min⁻¹)	$= R_{a} total - V \cdot (\frac{C_{2} - C_{1}}{t_{2} - t_{2}})$
	$\dot{{ra}^{o}}$ (μmol·kg⁻¹·min⁻¹)	$= R_{a} total \cdot (\frac{E_{2} + E_{1}}{2} + \frac{C_{2} + C_{1}}{2} \cdot \frac{E_{2} - E_{1}}{t_{2} - t_{1}} \cdot V)$
Non-steady-state, two-pool model (e.g., postprandial and exercise)	${\dot{R}}_{a} total$ (μmol·kg⁻¹· min⁻¹)	$= \frac{F}{E_{1}} - \frac{V_{1} \cdot C_{1}}{E_{1}} \cdot Ė_{1} + k_{12} (\frac{q_{2}^{iv}}{E_{1}} - Q_{2})$
	${\dot{R}}_{d}$ (μmol·kg⁻¹·min⁻¹)	$= R_{a} total - V_{1} \cdot \dot{C} - k_{21} \cdot V_{1} \cdot C_{1} + k_{12} \cdot Q_{2}$
	$\dot{{ra}^{o}}$ (μmol·kg⁻¹·min⁻¹)	$= R_{a} total \frac{F}{r_{1}} - \frac{V_{1} \cdot g}{r_{1}} \cdot {\dot{r}}_{1} + k_{12} [\frac{q_{2}^{iv}}{r_{1}} - q_{2}^{0}]$
Model independent	Metabolic clearance rate (ml·kg⁻¹·min⁻¹)	$= \frac{R_{d}}{C}$
	${\dot{R}}_{a} gut$ (μmol·kg⁻¹·min⁻¹)	${= ra}^{o} [\frac{1}{E_{ing}}]$
	${\dot{R}}_{a} endo$ (μmol·kg⁻¹·min⁻¹)	$= R_{a} total - R_{a} gut$
Natural abundance carbohydrates	Exogenous carbohydrate oxidation (g/min)	$= \dot{V} {CO}_{2} \cdot (\frac{E_{1} - E_{bkg}}{E_{ingA} - E_{ingB}}) \cdot (\frac{1}{k})$
Tracer enriched carbohydrates	Exogenous carbohydrate oxidation (g/min)	$= \dot{V} {CO}_{2} \cdot (\frac{E_{1} - E_{bkg}}{E_{ingA} - E_{bkg}}) \cdot (\frac{1}{k})$
	Plasma glucose oxidation (g/min)	$= \dot{V} {CO}_{2} \cdot (\frac{E_{1} - E_{bkg}}{E_{PG} - E_{PGbkg}}) \cdot (\frac{1}{k})$
	Muscle glycogen oxidation (g/min)	$\begin{matrix} = whole-body carbohydrate oxidation - plasma glucose oxidation \\ or \\ = whole-body carbohydrate oxidation - R_{d} glucose \end{matrix}$
	Liver glucose oxidation (g/min)	$\begin{matrix} = plasma glucose oxidation - exogenous carbohydrate oxidation \\ or \\ = R_{d} glucose - exogenous carbohydrate oxidation \end{matrix}$

Note. R_a = rate of appearance; R_d = rate of disappearance; R_atotal = total rate of appearance; ra^o = rate of appearance of ingested tracer; R_agut = rate of appearance of gut-derived (exogenous) glucose; R_aendo = rate of appearance of endogenous glucose; F = infusion rate; E = enrichment at a given timepoint; E₁ = enrichment at Timepoint 1; E₂ = enrichment at Timepoint 2; E_ing = enrichment of the ingested carbohydrate; E_bkg = enrichment of the background sample (either fasting sample or equivalent timepoint on background trial); E_ingA = enrichment of the high-natural abundance carbohydrate ingested; E_ingB = enrichment of the low-natural abundance carbohydrate ingested; E_PG = enrichment of plasma glucose at the timepoint of interest; E_PGbkg = enrichment of plasma glucose at baseline (fasting, before exercise); V = volume of distribution; C = glucose concentration at a given timepoint; C = change in glucose concentration over time (derivative of C); C₁ = glucose concentration at Timepoint 1; C₂ = glucose concentration at Timepoint 1; t₁ = time at Timepoint 1; t₂ = time at Timepoint 2; K₁₂ = rate constant between the accessible and peripheral compartment (0.05/min); K₂₁ = rate constant between the peripheral and accessible compartment (0.07/min); $q_{2}^{iv}$ = the amount of the infused [²H₂] tracer in the peripheral compartment; Q₂ = amount of tracee in the peripheral compartment; $q_{2}^{o}$ = the amount of the ingested [U-¹³C] tracer in the peripheral compartment; r₁ = ratio of the infusion [²H₂] and oral [U-¹³C] glucose tracer concentrations in circulation; $\dot{r}$ = change in r over time (derivate of r); $\dot{V} {CO}_{2}$ = rate of carbon dioxide production (L/min).

If both R_a and R_d increase (as is seen with exercise), then the blood glucose concentration will remain stable, but the enrichment will decrease by dilution from endogenous glucose production (Figure 3a). If R_a increases but R_d does not change, the enrichment decreases from dilution, but the concentration increases as glucose begins to accumulate in the circulation (Figure 3b). If R_a decreases but R_d does not change, enrichment increases as the concentration decreases (Figure 3c). Finally, if both R_a and R_d decrease, the concentration remains stable, but the enrichment rises.

Using these principles, the equations in Table 1 were developed based on a single-pool model (assuming the glucose is distributed in a single bodily pool) and adapted for use with stable isotopes to calculate rates of appearance of total glucose into the circulation (R_atotal) and R_d (Steele, 1959; Steele et al., 1965). The calculations based on the single-pool model have the advantage of being able to be calculated using widely available software such as Microsoft Excel (Microsoft Corp., Redmond, VA, USA). However, potential disadvantages include some error in the ability of the model to accurately predict R_a and R_d under conditions where these values are changing very rapidly, such as the transition from fasted to fed, or from rest to exercise. Improved accuracy of R_a and R_d calculations can be achieved by:

(a)adjusting the infusion rate to reduce the changes in enrichment (Molina et al., 1990);
(b)smoothing the enrichment and concentration data using techniques such as the optimal segments technique (Taylor et al., 1996); and
(c)using a two-pool model (Basu et al., 2003).

One disadvantage of the two-pool model, however, is that specialist software is required for calculation, such as SAMM II (The Epilson Group). R_atotal and R_d can be calculated using the two-pool model (Radziuk, 1976; Radziuk et al., 1978; Table 1).

Since circulating glucose concentrations can regulate R_d calculating the metabolic clearance rate can give insight into the R_d when taking glucose concentrations into account (Table 1). Furthermore, during moderate-intensity exercise, it has been demonstrated that > 97% of R_d is oxidized (Jeukendrup et al., 1999), and therefore, the R_d of plasma glucose can also be used as an index of blood plasma glucose oxidation rate. As a result, in a fasted state or when glucose-based carbohydrates are ingested, with R_d plasma glucose and whole-body carbohydrate oxidation rates from indirect calorimetry, an estimate of muscle glycogen utilization can also be derived.

It should be noted that the assumptions underlying equations for plasma glucose, muscle glycogen, and liver (endogenous) glucose oxidation can be violated under some conditions. For example, with fructose ingestion during exercise, a substantial fraction is converted into lactate (via gluconeogenesis), which can be oxidized by muscle (Lecoultre et al., 2010). In this scenario, all ingested carbohydrates should be equivalently enriched in ¹³C (e.g., [U-¹³C]glucose and [U-¹³C]fructose) and both glucose and lactate ¹³C enrichments in plasma should be measured. The R_d of both carbohydrates (i.e., [U-¹³C]glucose and [U-¹³C]lactate) should be subtracted from whole-body carbohydrate oxidation to estimate muscle glycogen oxidation, rather than subtracting R_d glucose alone.

Dilution Model—Endogenous and Exogenous Glucose Appearance Rates

Under conditions of carbohydrate ingestion, understanding whether changes in the R_atotal are due to changes in hepatic metabolism or gastrointestinal absorption, requires a distinction between the fraction of R_atotal from endogenous (R_aendo) versus gut (R_agut) sources. The first step is to calculate the R_a of the ingested tracer ra^o which can be calculated using the single-pool model (Table 1).

Under conditions where the R_agut is rapidly changing, such as in the postprandial state following a meal, the dual-tracer approach is sometimes less accurate for estimating R_agut, endogenous glucose production and R_d than one would want. In these scenarios, a third tracer (e.g., [6-³H]-glucose) can be intravenously infused in a pattern which mimics the expected R_agut (Basu et al., 2003). In this scenario, the infused [6-³H]-glucose is the tracer and the ingested [¹³C]-glucose is the tracee. This method can further improve the accuracy of assessing plasma glucose kinetics. However, the additional cost needs to be balanced with the degree of accuracy required and considered alongside the study design. For example, using a single-pool model, the addition of a third tracer meaningfully alters the postprandial time-course (e.g., time to peak: 23 ± 1 min vs. 34 ± 2 min) and peak R_agut (84 ± 4 μmol·kg⁻¹·min⁻¹ vs. 52 ± 5μmol·kg⁻¹·min⁻¹) markedly. However, as the two-pool model already improves the accuracy of postprandial glucose kinetics, the addition of a third tracer still makes a difference to the time-course (time to peak: 26 ± 2 min vs. 37 ± 3 min), but no meaningful difference to the peak R_agut (86 ± 4 μmol·kg⁻¹·min⁻¹ vs. 81 ± 7 μmol·kg⁻¹·min⁻¹; Basu et al., 2003). Since most exercise studies of carbohydrate metabolism involve continuous exercise and frequent carbohydrate ingestion, plasma glucose kinetics are likely to all be in a relative steady state, and therefore, it is probable that a dual-tracer, single-pool model has sufficient accuracy in these scenarios (as opposed to a triple-tracer and/or two-pool model). Under this tracer model, for accurate quantification, it is important that researchers use and report the nature of steady-state exercise under critical power (i.e., when steady-state metabolic indictors such as oxygen consumption, carbon dioxide production and circulating lactate concentrations are not continuously increasing over time; Jansson, 1982).

Tracer Incorporation Model—Gluconeogenesis

Gluconeogenesis can also be determined using stable isotopes by either infusing a labeled glucose precursor such as [3-¹³C]lactate and tracing the labeled carbon into glucose (Emhoff et al., 2013), or ingesting ²H₂O and tracing the deuterium into glucose (Chacko et al., 2008). One benefit over the latter approach is that it is thought this reflects the contribution from all precursors to gluconeogenesis, whereas the use of a labeled precursor requires additional assumptions to extrapolate to other precursors. However, since absolute gluconeogenesis rates are relatively stable in humans with fasting, carbohydrate restriction, and exercise (Gonzalez & Betts, 2019; Nuttall et al., 2008), the measurement of gluconeogenesis will not be prioritized in this review. The reader is directed to other sources for more detail on this measure (Chacko et al., 2008; Coggan, 1999; Wolfe & Chinkes, 2004).

Tracer Incorporation Model—Exogenous Carbohydrate Oxidation

Understanding the rate at which exogenous sources of carbohydrate can be digested, absorbed, and oxidized is useful to establish which products provide the most rapid and efficient fuel to support exercise demands. The measurement of exogenous carbohydrate oxidation rates relies on the principle that the ingested carbohydrate has a higher enrichment than endogenous carbohydrates. Stable isotopes of carbon (¹³C) on the carbohydrate source (see above) are either added as a tracer to a batch of carbohydrates or measured on carbohydrates that are naturally highly enriched in ¹³C. When the ingested carbohydrates are oxidized, the ¹³C appears in exhaled CO₂. Accordingly, by multiplying the total rate of carbon dioxide production by the ¹³C enrichment of exhaled CO₂, the rate of exogenous carbohydrate oxidation can be quantified (Mosora et al., 1976; Table 1).

In conjunction with indirect calorimetry, a further benefit to this method is that endogenous carbohydrate oxidation rates can be noninvasively determined by subtracting exogenous carbohydrate oxidation rates from whole-body carbohydrate oxidation rates. This provides indirect insight into glycogen utilization, which can be specifically apportioned to liver and/or muscle sources if ¹³C enrichment of plasma carbohydrates is measured (Table 1). It should be noted that using baseline (i.e., fasting, preexercise sample) as E_bkg has limitations, notably that the enrichment of endogenously produced CO₂ will likely change due to exercise (Peronnet et al., 1993a). This will differ according to the carbohydrate ingestion regimen and ways to deal with this issue are described below.

Considerations to Accurately Assess Exogenous Carbohydrate Oxidation Rates

Due to the potential for the ¹³C or ¹⁴C tracer to become trapped in either the bicarbonate pool or the acetate pool, and therefore, not appear in expired CO₂, exogenous substrate oxidation may be underestimated. Recovery factors can be applied to correct this but should be based on similar types of exercise and populations. Furthermore, the acetate recovery factor seems be less relevant (or perhaps even inappropriate) for carbohydrate versus fat oxidation, since unphysiological values have been reported when acetate correction factors have been applied to glucose oxidation (Trimmer et al., 2001). To further circumvent these issues, ensuring that the exercise is steady state with respect to VO₂ and VCO₂, and below an intensity where the bicarbonate pool will be changing drastically (e.g., below critical power/speed) can reduce the likelihood of violating these assumptions. Sampling after the point at which steady-state enrichment of the bicarbonate pool is made will minimize disturbances in ¹³CO₂. This has traditionally meant that the samples taken from the first hour of exercise are discounted. However, recent evidence suggests that as little as 25 min may be enough to equilibrate the bicarbonate pool at exercise intensities of at least 50% W_max (Podlogar & Wallis, 2020).

The appropriate choice of sample to represent the background expired CO₂ enrichment is also required. Often, researchers have used a water trial where participants perform identical exercise to the experimental trial except for the ingested drink (i.e., adjustment to carbohydrate as the independent variable), as no carbohydrate is ingested on the background trial. The isotopic enrichment of expired CO₂ changes over the duration of exercise as the endogenous carbohydrate stores are oxidized and can be used as a background correction (Peronnet et al., 1993a). Experiments have also used a preexercise ¹³CO₂ sample as the background correction. However, a limitation of these methods is the assumption that the rate of endogenous substrate oxidation is identical during the experimental trial and the water trial. This is unlikely to be true since carbohydrate ingestion, as well as exercise, can alter the use of muscle and liver glycogen in a dose-dependent manner (Gonzalez et al., 2015; King et al., 2018; Wallis et al., 2007). Peronnet et al. (1990) showed that exogenous carbohydrate oxidation during exercise can be significantly overestimated when the resting background correction is used (Peronnet et al., 1990), and that this is further exacerbated when using high natural abundance carbohydrate sources as the tracer. Therefore, the optimal background trial has been suggested to be identical to the experimental trial in every way, including the carbohydrate ingestion, but using carbohydrates with a lower enrichment (Odell et al., 2020; Peronnet et al., 1993a). For example, beet sugar (approximately −25‰) could be used as the source of sucrose with a low natural enrichment for a background trial, and cane sugar with a high natural enrichment (approximately −12‰) for an experimental trial. One disadvantage to this approach is that since each experimental condition (e.g., dose or type of carbohydrate) could produce different effects on endogenous substrate oxidation—the matched background correction is needed for each experimental trial, and so this doubles the number of trials needed in a study. However, this limitation may be further overcome using added label (tracer) to produce highly enriched carbohydrates. Here, the enrichment is so high that background shifts become negligible in relation to calculating exogenous carbohydrate oxidation rates, and thus, a background carbohydrate trial is not needed. Simply using the baseline expired CO₂ enrichment can suffice as the background correction. Note that the equations differ slightly when using tracers versus natural abundance carbohydrates.

Other considerations come to light when ingesting multiple sources of carbohydrates (e.g., glucose and fructose), especially when the aim is to assess the individual contribution of each substrate to oxidation. One of the approaches to this includes extra experimental trials whereby each carbohydrate source is labeled. For example, Trial 1 could include ingestion of highly enriched glucose alongside lower enriched fructose, and Trial 2 would then include ingestion of lower enriched glucose alongside highly enriched fructose (Adopo et al., 1994; Peronnet et al., 1993b). Alternatively, the glucose and fructose can be labeled with different types of isotopes (e.g., [U-¹³C]glucose and [U-¹⁴C]-fructose), which reduces the need for additional experimental trials but introduces the additional risks of radiation exposure (O’Brien et al., 2013). It should be noted, however, that direct comparison of ¹³C and ¹⁴C methods have revealed discordance, whereby the ¹⁴C method systematically underestimated rates of exogenous glucose oxidation by ∼15% compared with the ¹³C method (Moseley et al., 2005). The source of this discrepancy is unclear, however, it does highlight the tradeoffs with each experimental method.

Future Nutrition Technologies and Applications

The development of food technologies and the drive for advancement in metabolic performance has resulted in new applications of isotopic tracer methods. The development of hydrogel technology for carbohydrate ingestion in sport and exercise, for example, has led to anecdotal and empirical evidence for increased metabolic handling of carbohydrate (King et al., 2020; Sutehall et al., 2018). Similarly, the inclusion of modified starches (Baur & Saunders, 2021; Baur et al., 2016) to adjust osmolality and glycemic index of ingested substrates means consideration of appropriate isotopic tracers is required in future studies. In studies of exogenous substrate oxidation from hydrogel sources, comparable isotopic methods are suitable given the hydrogel formulation does not seem to limit the rate of carbohydrate delivery to the circulation. However, starch formulations which delay intestinal absorption should be labeled with a tracer that closely resembles the biochemical behavior and handling of the ingested substrate.

Summary and Recommendations

Isotopic tracers can be used to assess a variety of aspects of carbohydrate metabolism during exercise, from blood glucose kinetics, including rates of appearance and disappearance of glucose, to exogenous (ingested) and endogenous (muscle glycogen and liver glucose) carbohydrate oxidation. The choice of tracer methods needs to be carefully considered within the context of the study design. The objective of this article was to provide exercise science researchers with some basic background and a framework of quality to help guide experimental design. It was not the intention to provide an exhaustive consideration of the isotopic interactions and measurements that inform the practical execution of experiments. For detailed information on the basic principles of isotopic tracers, the reader is referred to several other comprehensive sources (Peronnet et al., 1990; Rennie, 1999; Wolfe & Chinkes, 2004).

When a researcher is looking to employ stable isotopes for assessing carbohydrate metabolism during exercise, one of the first decisions is whether to employ stable (nonradioactive) or unstable (radioactive) isotopes, each of which have strengths and limitations. Other considerations include the outcomes of interest and the nature of the nutritional and/or exercise interventions. If substrate oxidation rates are of interest, but not blood glucose kinetics, then oral ingestion of carbon-labeled carbohydrates can be used to determine exogenous and endogenous carbohydrate oxidation rates. If blood glucose kinetics are of interest, an infusion of another isotope of glucose is required (normally a deuterium-labeled glucose). For many sport and exercise nutrition studies, which involve exercise and nutritional strategies that achieve steady-state metabolic conditions, this dual-tracer method combined with a single-pool mathematical model can have adequate accuracy to answer many questions. However, if rapidly changing, nonsteady state conditions are expected (e.g., ingestion of a large bolus of carbohydrate, or changes in exercise intensity), then the accuracy of estimating blood glucose kinetics can be more challenging. To improve the accuracy of estimating blood glucose kinetics under these conditions, researchers can consider employing a two-pool model and/or a third tracer (Figure 4).

Further recommendations for using isotopic tracers for carbohydrate metabolism during exercise, particularly when using stable isotopes are found in Table 2 and Figure 5. Researchers are often faced with several threats to experimental rigor when using isotopic tracers. A careful balance of practicality and robust methods is required, and it is suggested that the guidelines outlined here are used to guide these decisions in light of the wider context of the study design, alongside practical challenges and resource availability. To complement this, readers are also referred to broader reporting guidelines for Sport Nutrition and Exercise metabolism research (Betts et al., 2020). Researchers may wish to assess their experimental methods with the checklist provided in Figure 6.

Table 2

Recommendations With Underlying Rationale When Using Isotopic Tracers to Assess Carbohydrate Metabolism During Exercise

Recommendation	Rationale
Pilot test	Ensure adequate enrichment of ingested and/or infused carbohydrates to allow detection in blood and/or breath samples over background
Measure the final enrichment of carbohydrates that were infused and ingested (keep back a sample)	To quantify the enrichment of ingested and infused carbohydrates for accurate calculations
Use the contralateral limb to the infusion for sampling	Prevent contamination of sampling tube with infusate
Accurately time the infusion and weigh infusion syringe before and after infusion on accurate balance scales	To quantify the infusion rate and adjust the equations accordingly
Take at least three blood samples ∼10 min apart (after beginning a tracer infusion and expected equilibration) during the basal state before commencing an intervention (e.g., food intake or exercise)	To check that steady-state enrichment has been achieved, and thus, changes in enrichment are due to the subsequent intervention
If using carbohydrates with natural enrichment, it is preferable to use a background trial which mimics the metabolic status of the experimental trial	To account for background shifts of ¹³C in exhaled breath from the oxidation of endogenous substrates, which can be influenced by the nutritional strategy
Consider 5 days of a diet low in ¹³C foods (e.g., low in corn and sugar cane products) and/or a glycogen depleting bout of exercise to reduce endogenous ¹³C stores	To increase the tracer-to-tracee ratio and decrease the degree of background shift
Use steady-state exercise and confirm steady-state nature with measures such as oxygen consumption, carbon dioxide production, and circulating lactate concentrations	Altering exercise intensity will introduce fluctuations in energy demands, bicarbonate buffering pool size, and substrate source oxidation which take time to stabilize and return to previous levels. Therefore, sampling may not be reflective of sudden changes (e.g., sprints or intervals < critical speed).

Note. These are suggestions and dependent on context and trade-offs in study design. ¹³C = carbon-13.

Acknowledgments

We dedicate this article to the memory of Prof. Kevin D. Tipton. Author Contributions: Gonzalez and King contributed equally to conceptualization, visualization, writing the original draft, and reviewing and editing the final article. Gonzalez is an investigator on research grants funded by BBSRC, MRC, British Heart Foundation, The Rank Prize Funds, The European Society for Clinical Nutrition and Metabolism (ESPEN), Lucozade Ribena Suntory, ARLA Foods Ingredients, Cosun Nutrition Center, and Clasado Biosciences; and has completed paid consultancy for PepsiCo and SVGC.

References

Adopo, E., Peronnet, F., Massicotte, D., Brisson, G.R., & Hillaire-Marcel, C. (1994). Respective oxidation of exogenous glucose and fructose given in the same drink during exercise. Journal of Applied Physiology, 76(3), 1014–1019. https://doi.org/10.1152/jappl.1994.76.3.1014
- Crossref
- PubMed
- Search Google Scholar
- Export Citation
Barber, J.F.P., Thomas, J., Narang, B., Hengist, A., Betts, J.A., Wallis, G.A., & Gonzalez, J.T. (2020). Pectin-alginate does not further enhance exogenous carbohydrate oxidation in running. Medicine & Science in Sports & Exercise, 52(6), 1376–1384. https://doi.org/10.1249/MSS.0000000000002262
- Crossref
- Search Google Scholar
- Export Citation
Basu, R., Di Camillo, B., Toffolo, G., Basu, A., Shah, P., Vella, A., Rizza, R., & Cobelli, C. (2003). Use of a novel triple-tracer approach to assess postprandial glucose metabolism. American Journal of Physiology—Endocrinology and Metabolism, 284(1), E55–E69. https://doi.org/10.1152/ajpendo.00190.2001
- Crossref
- PubMed
- Search Google Scholar
- Export Citation
Baur, D.A., & Saunders, M.J. (2021). Carbohydrate supplementation: A critical review of recent innovations. European Journal of Applied Physiology, 121(1), 23–66. https://doi.org/10.1007/s00421-020-04534-y
- Crossref
- PubMed
- Search Google Scholar
- Export Citation
Baur, D.A., Vargas Fde, C., Bach, C.W., Garvey, J.A., & Ormsbee, M.J. (2016). Slow-absorbing modified starch before and during prolonged cycling increases fat oxidation and gastrointestinal distress without changing performance. Nutrients, 8(7), Article 392. https://doi.org/10.3390/nu8070392
- Crossref
- Search Google Scholar
- Export Citation
Betts, J.A., Gonzalez, J.T., Burke, L.M., Close, G.L., Garthe, I., James, L.J., Jeukendrup, A.E., Morton, J.P., Nieman, D.C., & Peeling, P. (2020). PRESENT 2020: Text expanding on the checklist for proper reporting of evidence in sport and exercise nutrition trials. International Journal of Sport Nutrition and Exercise Metabolism, 30(1), 2–13. https://doi.org/10.1123/ijsnem.2019-0326
- Crossref
- PubMed
- Search Google Scholar
- Export Citation
Chacko, S.K., Sunehag, A.L., Sharma, S., Sauer, P.J., & Haymond, M.W. (2008). Measurement of gluconeogenesis using glucose fragments and mass spectrometry after ingestion of deuterium oxide. Journal of Applied Physiology, 104(4), 944–951. https://doi.org/10.1152/japplphysiol.00752.2007
- Crossref
- Search Google Scholar
- Export Citation
Coggan, A.R. (1999). Use of stable isotopes to study carbohydrate and fat metabolism at the whole-body level. Proceedings of the Nutrition Society, 58(4), 953–961. https://doi.org/10.1017/s0029665199001263
- Crossref
- PubMed
- Search Google Scholar
- Export Citation
Edinburgh, R.M., Hengist, A., Smith, H.A., Travers, R.L., Koumanov, F., Betts, J.A., Thompson, D., Walhin, J.P., Wallis, G.A., Hamilton, D.L., Stevenson, E.J., Tipton, K.D., & Gonzalez, J.T. (2018). Preexercise breakfast ingestion versus extended overnight fasting increases postprandial glucose flux after exercise in healthy men. American Journal of Physiology—Endocrinology and Metabolism, 315(5), E1062–E1074. https://doi.org/10.1152/ajpendo.00163.2018
- Crossref
- PubMed
- Search Google Scholar
- Export Citation
Emhoff, C.A., Messonnier, L.A., Horning, M.A., Fattor, J.A., Carlson, T.J., & Brooks, G.A. (2013). Gluconeogenesis and hepatic glycogenolysis during exercise at the lactate threshold. Journal of Applied Physiology, 114(3), 297–306. https://doi.org/10.1152/japplphysiol.01202.2012
- Crossref
- Search Google Scholar
- Export Citation
Folch, N., Péronnet, F., Massicotte, D., Duclos, M., Lavoie, C., & Hillaire-Marcel, C. (2001). Metabolic response to small and large 13C-labelled pasta meals following rest or exercise in man. British Journal of Nutrition, 85(6), 671–680. https://doi.org/10.1079/bjn2001325
- Crossref
- Search Google Scholar
- Export Citation
Gonzalez, J.T., & Betts, J.A. (2019). Dietary sugars, exercise and hepatic carbohydrate metabolism. Proceedings of the Nutrition Society, 78(2), 246–256. https://doi.org/10.1017/S0029665118002604
- Crossref
- Search Google Scholar
- Export Citation
Gonzalez, J.T., Fuchs, C.J., Smith, F.E., Thelwall, P.E., Taylor, R., Stevenson, E.J., Trenell, M.I., Cermak, N.M., & van Loon, L.J. (2015). Ingestion of glucose or sucrose prevents liver but not muscle glycogen depletion during prolonged endurance-type exercise in trained cyclists. American Journal of Physiology—Endocrinology and Metabolism, 309(12), E1032–E1039. https://doi.org/10.1152/ajpendo.00376.2015
- Crossref
- Search Google Scholar
- Export Citation
Hearris, M.A., Pugh, J.N., Langan-Evans, C., Mann, S.J., Burke, L., Stellingwerff, T., Gonzalez, J.T., & Morton, J.P. (2022). 13C-glucose-fructose labelling reveals comparable exogenous CHO oxidation during exercise when consuming 120 g/h in fluid, gel, jelly chew or co-ingestion. Journal of Applied Physiology, 132(6), 1394–1406. https://doi.org/10.1152/japplphysiol.00091.2022
- Crossref
- Search Google Scholar
- Export Citation
Jahren, A.H., Saudek, C., Yeung, E.H., Kao, W.H., Kraft, R.A., & Caballero, B. (2006). An isotopic method for quantifying sweeteners derived from corn and sugar cane. The American Journal of Clinical Nutrition, 84(6), 1380–1384. https://doi.org/10.1093/ajcn/84.6.1380
- Crossref
- Search Google Scholar
- Export Citation
Jansson, E. (1982). On the significance of the respiratory exchange ratio after different diets during exercise in man. Acta Physiologica Scandinavica, 114(1), 103–110. https://doi.org/10.1111/j.1748-1716.1982.tb06958.x
- Crossref
- Search Google Scholar
- Export Citation
Jeukendrup, A.E., Raben, A., Gijsen, A., Stegen, J.H., Brouns, F., Saris, W.H., & Wagenmakers, A.J. (1999). Glucose kinetics during prolonged exercise in highly trained human subjects: Effect of glucose ingestion. The Journal of Physiology, 515(2), 579–589. https://doi.org/10.1111/j.1469-7793.1999.579ac.x
- Crossref
- Search Google Scholar
- Export Citation
Jumpertz, R., Le, D.S., Turnbaugh, P.J., Trinidad, C., Bogardus, C., Gordon, J.I., & Krakoff, J. (2011). Energy-balance studies reveal associations between gut microbes, caloric load, and nutrient absorption in humans. The American Journal of Clinical Nutrition, 94(1), 58–65. https://doi.org/10.3945/ajcn.110.010132
- Crossref
- Search Google Scholar
- Export Citation
Kim, I.Y., Suh, S.H., Lee, I.K., & Wolfe, R.R. (2016). Applications of stable, nonradioactive isotope tracers in in vivo human metabolic research. Experimental & Molecular Medicine, 48, e203. https://doi.org/10.1038/emm.2015.97
- Crossref
- Search Google Scholar
- Export Citation
King, A.J., O’Hara, J.P., Morrison, D.J., Preston, T., & King, R. (2018). Carbohydrate dose influences liver and muscle glycogen oxidation and performance during prolonged exercise. Physiological Reports, 6(1), Article e13555. https://doi.org/10.14814/phy2.13555
- Crossref
- Search Google Scholar
- Export Citation
King, A.J., Rowe, J.T., & Burke, L.M. (2020). Carbohydrate hydrogel products do not improve performance or gastrointestinal distress during moderate-intensity endurance exercise. International Journal of Sport Nutrition and Exercise Metabolism, 30(5), 305–314. https://doi.org/10.1123/ijsnem.2020-0102
- Crossref
- Search Google Scholar
- Export Citation
Kjaer, M., Farrell, P.A., Christensen, N.J., & Galbo, H. (1986). Increased epinephrine response and inaccurate glucoregulation in exercising athletes. Journal of Applied Physiology, 61(5), 1693–1700. https://doi.org/10.1152/jappl.1986.61.5.1693
- Crossref
- Search Google Scholar
- Export Citation
Lecoultre, V., Benoit, R., Carrel, G., Schutz, Y., Millet, G.P., Tappy, L., & Schneiter, P. (2010). Fructose and glucose co-ingestion during prolonged exercise increases lactate and glucose fluxes and oxidation compared with an equimolar intake of glucose. The American Journal of Clinical Nutrition, 92(5), 1071–1079. https://doi.org/10.3945/ajcn.2010.29566
- Crossref
- Search Google Scholar
- Export Citation
Metallo, C.M., Walther, J.L., & Stephanopoulos, G. (2009). Evaluation of 13C isotopic tracers for metabolic flux analysis in mammalian cells. Journal of Biotechnology, 144(3), 167–174. https://doi.org/10.1016/j.jbiotec.2009.07.010
- Crossref
- Search Google Scholar
- Export Citation
Molina, J.M., Baron, A.D., Edelman, S.V., Brechtel, G., Wallace, P., & Olefsky, J.M. (1990). Use of a variable tracer infusion method to determine glucose turnover in humans. The American Journal of Physiology, 258(1), E16–E23. https://doi.org/10.1152/ajpendo.1990.258.1.E16
- Search Google Scholar
- Export Citation
Moodley, D., Noakes, T.D., Bosch, A.N., Hawley, J.A., Schall, R., & Dennis, S.C. (1992). Oxidation of exogenous carbohydrate during prolonged exercise: The effects of the carbohydrate type and its concentration. European Journal of Applied Physiology and Occupational Physiology, 64(4), 328–334. https://doi.org/10.1007/BF00636220
- Crossref
- Search Google Scholar
- Export Citation
Morrison, D.J., Dodson, B., Slater, C., & Preston, T. (2000). (13)C natural abundance in the British diet: Implications for (13)C breath tests. Rapid Communications in Mass Spectrometry, 14(15), 1321–1324. https://doi.org/10.1002/1097-0231(20000815)14:15< 1321::AID-RCM946> 3.0.CO;2-8
- Crossref
- Search Google Scholar
- Export Citation
Moseley, L., Jentjens, R., Waring, R., Harris, R., Harding, L., & Jeukendrup, A. (2005). Measurement of exogenous carbohydrate oxidation: A comparison of [U-14C] glucose and [U-13C] glucose tracers. American Journal of Physiology—Endocrinology and Metabolism, 289(2), E206–E211. https://doi.org/10.1152/ajpendo.00423.2004
- Crossref
- Search Google Scholar
- Export Citation
Mosora, F., Lefebvre, P., Pirnay, F., Lacroix, M., Luyckx, A., & Duchesne, J. (1976). Quantitative evaluation of the oxidation of an exogenous glucose load using naturally labeled 13C-glucose. Metabolism, 25(12), 1575–1582. https://doi.org/10.1016/0026-0495(76)90110-4
- Crossref
- Search Google Scholar
- Export Citation
Nuttall, F.Q., Ngo, A., & Gannon, M.C. (2008). Regulation of hepatic glucose production and the role of gluconeogenesis in humans: Is the rate of gluconeogenesis constant? Diabetes Metabolism Research and Reviews, 24(6), 438–458. https://doi.org/10.1002/dmrr.863
- Crossref
- Search Google Scholar
- Export Citation
O’Brien, W.J., Stannard, S.R., Clarke, J.A., & Rowlands, D.S. (2013). Fructose-maltodextrin ratio governs exogenous and other CHO oxidation and performance. Medicine & Science in Sports & Exercise, 45(9), 1814–1824. https://doi.org/10.1249/MSS.0b013e31828e12d4
- Crossref
- Search Google Scholar
- Export Citation
Odell, O.J., Podlogar, T., & Wallis, G.A. (2020). Comparable exogenous carbohydrate oxidation from lactose or sucrose during exercise. Medicine & Science in Sports & Exercise, 52(12), 2663–2672. https://doi.org/10.1249/MSS.0000000000002426
- Crossref
- Search Google Scholar
- Export Citation
Peronnet, F., Adopo, E., Massicotte, D., Brisson, G., & Hillaire-Marcel, C. (1993a). Comparison of two methods for computing exogenous substrate oxidation using 13C-labeling. Medicine & Science in Sports & Exercise, 25(2), 297–302. https://www.ncbi.nlm.nih.gov/pubmed/8450736
- Search Google Scholar
- Export Citation
Peronnet, F., Adopo, E., Massicotte, D., Brisson, G.R., & Hillaire-Marcel, C. (1993b). Method for computing the oxidation of two 13C-substrates ingested simultaneously during exercise. Journal of Applied Physiology, 75(3), 1419–1422. https://doi.org/10.1152/jappl.1993.75.3.1419
- Crossref
- Search Google Scholar
- Export Citation
Peronnet, F., Massicotte, D., Brisson, G., & Hillaire-Marcel, C. (1990). Use of 13C substrates for metabolic studies in exercise: Methodological considerations. Journal of Applied Physiology, 69(3), 1047–1052. https://doi.org/10.1152/jappl.1990.69.3.1047
- Crossref
- Search Google Scholar
- Export Citation
Podlogar, T., & Wallis, G.A. (2020). Impact of post-exercise fructose-maltodextrin ingestion on subsequent endurance performance. Frontiers in Nutrition, 7, Article 82. https://doi.org/10.3389/fnut.2020.00082
- Crossref
- Search Google Scholar
- Export Citation
Radziuk, J. (1976). An integral equation approach to measuring turnover in nonsteady compartmental and distributed systems. Bulletin of Mathematical Biology, 38(6), 679–693. https://doi.org/10.1007/BF02458642
- Crossref
- Search Google Scholar
- Export Citation
Radziuk, J., Norwich, K.H., & Vranic, M. (1978). Experimental validation of measurements of glucose turnover in nonsteady state. The American Journal of Physiology, 234(1), E84–E93. https://doi.org/10.1152/ajpendo.1978.234.1.E84
- Search Google Scholar
- Export Citation
Rennie, M.J. (1999). An introduction to the use of tracers in nutrition and metabolism. Proceedings of the Nutrition Society, 58(4), 935–944. https://doi.org/10.1017/s002966519900124x
- Crossref
- Search Google Scholar
- Export Citation
Rowe, J.T., King, R., King, A.J., Morrison, D.J., Preston, T., Wilson, O.J., & O’Hara, J.P. (2022). Glucose and fructose hydrogel enhances running performance, exogenous carbohydrate oxidation, and gastrointestinal tolerance. Medicine & Science in Sports & Exercise, 54(1), 129–140. https://doi.org/10.1249/MSS.0000000000002764
- Crossref
- Search Google Scholar
- Export Citation
Saris, W., Goodpaster, B., Jeukendrup, A., Brouns, F., Halliday, D., & Wagenmakers, A. (1993). Exogenous carbohydrate oxidation from different carbohydrate sources during exercise. Journal of Applied Physiology, 75(5), 2168–2172. https://doi.org/10.1152/jappl.1993.75.5.2168
- Crossref
- Search Google Scholar
- Export Citation
Steele, R. (1959). Influences of glucose loading and of injected insulin on hepatic glucose output. Annals of the New York Academy of Sciences, 82, 420–430. https://doi.org/10.1111/j.1749-6632.1959.tb44923.x
- Crossref
- Search Google Scholar
- Export Citation
Steele, R., Bishop, J.S., Dunn, A., Altszuler, N., Rathbeb, I., & Debodo, R.C. (1965). Inhibition by insulin of hepatic glucose production in the normal dog. The American Journal of Physiology, 208, 301–306. https://doi.org/10.1152/ajplegacy.1965.208.2.301
- Crossref
- Search Google Scholar
- Export Citation
Sutehall, S., Muniz-Pardos, B., Bosch, A.N., Di Gianfrancesco, A., & Pitsiladis, Y.P. (2018). Sports drinks on the edge of a new era. Current Sports Medicine Reports, 17(4), 112–116. https://doi.org/10.1249/JSR.0000000000000475
- Crossref
- Search Google Scholar
- Export Citation
Sutehall, S., Muniz-Pardos, B., Smajgl, D., Mandic, M., Jeglinski, C., Bosch, A., Galloway, S.D., & Pitsiladis, Y. (2021). The validity and reliability of a novel isotope ratio infrared spectrometer to quantify 13C enrichment of expired breath samples in exercise. Journal of Applied Physiology, 130(5), 1421–1426. https://doi.org/10.1152/japplphysiol.00805.2020
- Crossref
- Search Google Scholar
- Export Citation
Taylor, R., Magnusson, I., Rothman, D.L., Cline, G.W., Caumo, A., Cobelli, C., & Shulman, G.I. (1996). Direct assessment of liver glycogen storage by 13C nuclear magnetic resonance spectroscopy and regulation of glucose homeostasis after a mixed meal in normal subjects. The Journal of Clinical Investigation, 97(1), 126–132. https://doi.org/10.1172/JCI118379
- Crossref
- Search Google Scholar
- Export Citation
Trimmer, J.K., Casazza, G.A., Horning, M.A., & Brooks, G.A. (2001). Recovery of (13)CO2 during rest and exercise after [1-(13)C]acetate, [2-(13)C]acetate, and NaH(13)CO3 infusions. American Journal of Physiology—Endocrinology and Metabolism, 281(4), E683–E692. https://doi.org/10.1152/ajpendo.2001.281.4.E683
- Crossref
- Search Google Scholar
- Export Citation
van Loon, L.J., Greenhaff, P.L., Constantin-Teodosiu, D., Saris, W.H., & Wagenmakers, A.J. (2001). The effects of increasing exercise intensity on muscle fuel utilisation in humans. The Journal of Physiology, 536(1), 295–304. https://doi.org/10.1111/j.1469-7793.2001.00295.x
- Crossref
- Search Google Scholar
- Export Citation
Wahren, J., Felig, P., & Hagenfeldt, L. (1978). Physical exercise and fuel homeostasis in diabetes mellitus. Diabetologia, 14(4), 213–222. https://doi.org/10.1007/BF01219419
- Crossref
- Search Google Scholar
- Export Citation
Wajngot, A., Chandramouli, V., Schumann, W.C., Kumaran, K., Efendic, S., & Landau, B.R. (1989). Testing of the assumptions made in estimating the extent of futile cycling. The American Journal of Physiology, 256(5), E668–E675. https://doi.org/10.1152/ajpendo.1989.256.5.E668
- Search Google Scholar
- Export Citation
Wallis, G.A., Yeo, S.E., Blannin, A.K., & Jeukendrup, A.E. (2007). Dose-response effects of ingested carbohydrate on exercise metabolism in women. Medicine & Science in Sports & Exercise, 39(1), 131–138. https://doi.org/10.1249/01.mss.0000241645.28467.d3
- Crossref
- Search Google Scholar
- Export Citation
Wolfe, R.R., & Chinkes, D.L. (2004). Isotopic tracers in metabolic research. Principles and practice of kinetic analysis (2nd ed.). Wiley.
- Search Google Scholar
- Export Citation

^*Gonzalez (J.T.Gonzalez@bath.ac.uk) is corresponding author, https://orcid.org/0000-0002-9939-0074

Print
Save
Cite
Email this content

Share Link

Copy this link, or click below to email it to a friend
Email this content
or copy the link directly:

https://journals.humankinetics.com/view/journals/ijsnem/33/1/article-p60.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

Collapse
Expand

International Journal of Sport Nutrition and Exercise Metabolism

Keywords:
Strengths and Limitations of Isotope Tracers
- Applications of Different Tracers for Carbohydrate Metabolism During Exercise
- Considerations to Accurately Assess Exogenous Carbohydrate Oxidation Rates
Future Nutrition Technologies and Applications
Summary and Recommendations
Acknowledgments
References

View raw image

Figure 1

—Isotopes of carbon, with the most common, carbon-12 (¹²C) on the left, the rare, stable isotope carbon-13 (¹³C) in the center, and unstable carbon-14 (¹⁴C) on the right.

View raw image

Figure 2

—From left to right, glucose with no carbon-13 (¹³C), glucose with one ¹³C at position 1, and glucose with all the carbons as ¹³C.

View raw image

Figure 3

—The principle of the dilution method whereby a stable concentration combined with a decrease in enrichment reflects increased rate of appearance (R_a) alongside increased rate of disappearance (R_d) (a), increased concentration with decreased enrichment reflects increased R_a with stable R_d (b), decreased concentration with increased enrichment reflects decreased R_a alongside stable R_d (c), and stable concentration alongside increase enrichment reflects decreased R_a and decreased R_d (d).

View raw image

Figure 4

—Example of isotope tracer methods of carbohydrate metabolism, where a [6,6-²H₂]-glucose intravenous infusion is combined with [U-¹³C]-carbohydrate (e.g., glucose and/or fructose) ingestion. If a third tracer is infused, this could be [6-³H]-glucose as an additional tracer for highly accurate assessment of rates of glucose appearance from the gut under rapidly changing conditions such as after a meal. The ingested [U-¹³C]-carbohydrate will appear in glucose (and lactate) in the circulation when digested and absorbed, and in CO₂ on the breath when oxidized. The infused [6,6-²H₂]-glucose (and if using a triple-tracer method, [6-³H]-glucose) will appear directly in the circulation when sampling from the contralateral arm.

View raw image

Figure 5

—Considerations for improving the quality of methods in studies using stable isotopes to assess carbohydrate metabolism during exercise. (a) Improve the signal-to-noise ratio by reducing enrichment of endogenous substrates by asking participants to consume a diet low in carbon-13 (¹³C), ensure exercise is of steady state, aim for a sufficiently high tracer enrichment in ingested drinks (ideally uniformly labeled so that all carbons are represented by the tracer), use data after 30 min of commencing exercise or prime the bicarbonate pool, and use contralateral limbs for sampling and infusing. (b) When ingesting carbohydrates of natural enrichment, the preferred method would be to include a true CHO baseline trial, replicating the experimental trial but with ingestion of CHO with low-natural enrichment, if this is not possible, the next best scenario is to include a placebo (water) baseline condition, the final option is to include the preexercise sample as the baseline correction. When ingesting tracer-enriched CHO, the same principles apply, but become obsolete when the enrichment is sufficiently high.

View raw image

Figure 6

—Checklist to assist researchers in designing and evaluating studies using stable isotopes to assess carbohydrate metabolism during exercise in humans. Each experimental design factor (a) is grouped based on potential impact on study quality (b). Note, this assessment is subjective, as the actual impact is dependent on many factors including other aspects of study design and aims. Therefore, this should be treated as a guide to be taken wider context.