European Journal of Applied Physiology

, Volume 111, Issue 12, pp 3135–3141

Carbohydrate intake reduces fat oxidation during exercise in obese boys


  • Lisa Chu
    • Children’s Exercise and Nutrition Centre, McMaster University, Children’s Hospital, Chedoke Hospital
  • Michael C. Riddell
    • School of Kinesiology and Health Science, Faculty of HealthYork University
  • Tim Takken
    • Children’s Exercise and Nutrition Centre, McMaster University, Children’s Hospital, Chedoke Hospital
    • Children’s Exercise and Nutrition Centre, McMaster University, Children’s Hospital, Chedoke Hospital
Original Article

DOI: 10.1007/s00421-011-1940-1

Cite this article as:
Chu, L., Riddell, M.C., Takken, T. et al. Eur J Appl Physiol (2011) 111: 3135. doi:10.1007/s00421-011-1940-1


The recent surge in childhood obesity has renewed interest in studying exercise as a therapeutic means of metabolizing fat. However, carbohydrate (CHO) intake attenuates whole body fat oxidation during exercise in healthy children and may suppress fat metabolism in obese youth. To determine the impact of CHO intake on substrate utilization during submaximal exercise in obese boys, seven obese boys (mean age: 11.4 ± 1.0 year; % body fat: 35.8 ± 3.9%) performed 60 min of exercise at an intensity that approximated maximal fat oxidation. A CHO drink (CARB) or a placebo drink (CONT) was consumed in a double-blinded, counterbalanced manner. Rates of total fat, total CHO, and exogenous CHO (CHOexo) oxidation were calculated for the last 20 min of exercise. During CONT, fat oxidation rate was 3.9 ± 2.4 mg × kg fat-free mass (FFM)−1 × min−1, representing 43.1 ± 22.9% of total energy expenditure (EE). During CARB, fat oxidation was lowered (p = 0.02) to 1.7 ± 0.6 mg × kg FFM−1 × min−1, contributing to 19.8 ± 4.9% EE. Total CHO oxidation rate was 17.2 ± 3.1 mg × kg FFM−1 × min−1 and 13.2 ± 6.1 mg × kg FFM−1 × min−1 during CARB and CONT, respectively (p = 0.06). In CARB, CHOexo oxidation contributed to 23.3 ± 4.2% of total EE. CHO intake markedly suppresses fat oxidation during exercise in obese boys.


MetabolismChildhood obesityEnergy expenditureExercise


Exercise is a key component of weight management programs that aim to attenuate the recent surge in childhood obesity. Physical activity is a viable option to increase energy expenditure (EE) and optimize fat oxidation, which are both relevant to the obese child who is at increased risk for metabolic complications (i.e., insulin resistance) (Sinha et al. 2002). Even though understanding of substrate utilization during exercise would seem critical for obese children, only a handful of studies have investigated this issue and very little is known about dietary effects on fat oxidation during exercise.

According to the available evidence, the rate of fat oxidation during exercise is reported to be either similar (Lazzer et al. 2007) or lower (Zunquin et al. 2009a) in obese versus non-obese youth. However, there is consistent evidence that puberty influences fat oxidation during exercise in obese children, with pre-pubertal children demonstrating higher rates of fat oxidation when normalized to fat-free mass (Zunquin et al. 2006, 2009b; Brandou et al. 2006). This loss of capacity to oxidize lipids and a switch to carbohydrate (CHO) oxidation during exercise in obese adolescents is entirely consistent with findings in non-obese youth (Riddell et al. 2008). What the literature is missing, however, is an investigation of how obese children utilize exogenous sources of fuel during exercise, such as CHO. This is particularly relevant given the predisposition of obese children to dysregulation in glucose metabolism (i.e., insulin resistance) and the frequent use of CHO-rich drinks of obese children. To our knowledge, however, no such investigation of exogenous CHO utilization during exercise in obese children has been reported.

In general, non-obese children demonstrate a greater reliance on exogenous CHO (CHOexo) as a source of energy during exercise, compared with adults (Timmons et al. 2003). Puberty also influences CHOexo oxidation during exercise with a loss of the capacity to oxidize CHOexo in more mature individuals (Timmons et al. 2007a, b). In non-obese children, the utilization of CHOexo during exercise also lowers the contribution of fat oxidation to total EE during exercise (Riddell et al. 2000a, b, 2001; Timmons et al. 2003, 2007a, b). However, no study has measured the rate of CHOexo oxidation during exercise in obese children nor has the effect of CHO intake on reducing fat oxidation during exercise in obese youth been studied. A CHO-induced blunting of fat oxidation during exercise may be counterproductive for obese children interested in oxidizing excess adiposity through physical activity. It is possible, therefore, that intake of CHO beverages (e.g., sport drinks and juices) before or during physical activity should not be recommended for obese children, although there is no scientific evidence to support this idea.

Clearly, additional study of endogenous and exogenous substrate utilization during exercise in obese children is warranted and is likely to be valuable for informing exercise programs designed to maximize fat oxidation. Thus, our primary aim was to determine the impact of CHO intake on substrate utilization, including CHOexo oxidation using a non-invasive 13C stable isotope tracer, in obese pre- and early pubertal boys. We hypothesized that CHO intake would reduce fat oxidation during exercise in obese boys, while providing a fuel source that would contribute significantly to total EE.



Seven pre- and early pubertal boys aged 9–12 years with a body mass index (BMI) >95th percentile for their age and sex volunteered to participate in the study. We chose to study pre- and early pubertal boys because studies show that puberty influences CHOexo oxidation in non-obese boys (Timmons et al. 2007b) and fat oxidation in obese boys (Zunquin et al. 2006, 2009b; Brandou et al. 2006). Pubertal status was self-assessed by the child and his parent according to pubic hair development using the criteria of Tanner (Tanner 1962). The boys were either Tanner Stage 1 (n = 3) or Tanner Stage 2 (n = 2). Two of the boys (a 10-year-old and an 11-year-old) did not want to complete the Tanner assessment. To complement the assessment of secondary sex characteristics, we also predicted age of peak height velocity, according to the method of Mirwald et al. (2002). The boys in this study were recruited from the weight management program at the Children’s Exercise and Nutrition Centre (CENC). Interested parents and children signed a consent-to-contact form and were later contacted with more information. This study met the ethical standards of the Hamilton Health Sciences/Faculty of Health Sciences Research Ethics Board.

Study design and procedures

We used a repeated-measures, crossover, counterbalanced, double-blinded, placebo-controlled study design. Participants were required to visit the CENC for three visits, one preliminary visit and two experimental visits.

Preliminary visit

At the preliminary visit, the study was explained in detail, any questions the parent or child had were answered, and written parental consent and child assent were obtained. Anthropometric measurements were then collected. Standing and sitting height (Harpenden Stadiometer, CMS Weighing Equipment LTD.) and body mass (BM; BWB-800, TANITA) of each participant were measured using standard procedures. Measured body mass and height were used to calculate BMI; the Centre for Disease Control growth charts were used to define age-specific BMI percentiles. Body composition (total fat and muscle) was estimated using bioelectric impedance analysis (BIA; bio-electric impedance-101A, RJL Systems). Tanner stage was also acquired at this visit.

An incremental test on a mechanically braked cycle ergometer (Fleisch-Metabo) was then conducted to estimate the exercise intensity that induced the highest fat oxidation rate. We used a modified protocol based on our previous work (Riddell et al. 2008), which required each boy to start cycling at 12.5 W and continue in 3-min stages with 12.5-W increments with each stage. In contrast to our earlier study design (Riddell et al. 2008), participants did not cycle to exhaustion. To satisfy the aims of this study, it was not necessary to have the boys obtain peak oxygen uptake (i.e., \( {\dot{{V}}\text{O}}_{2} \)). Instead, the respiratory exchange ratio (RER) was used to monitor the test, which was ended when the RER reached and was maintained at or greater than 1.00, indicating that the participant was predominantly oxidizing CHO. During the incremental exercise test, breath-by-breath measurements of O2 uptake (\( {\dot{{V}}\text{O}}_{2} \)) and CO2 production (\( {\dot{{V}}\text{O}}_{2} \)) were made using a metabolic cart (SensorMedics Vmax29, Yorba Linda, CA), averaged every 30 s and used to calculate rates of total CHO (CHOtotal) and fat (FATtotal) oxidation, as previously described (Timmons et al. 2003). Fat oxidation rates (g/min) were plotted against the time of exercise. We then visually inspected the graph to identify when during the test the highest fat oxidation occurred. The power output that corresponded to the time of exercise associated with a high fat oxidation rate (highest numerical value for g/min) was then selected for the experimental visits. Consistent with previous reports in obese (Zunquin et al. 2006, 2009a, b; Brandou et al. 2006; Maffeis et al. 2005; Lazzer et al. 2007) and non-obese (Riddell et al. 2008) children, our boys presented with a wide range of exercise intensities during which the rate of fat oxidation was high. In these cases, the lowest power output was chosen from visual inspection. To simplify our approach and ensure that participants could complete the study, we used the lowest intensity associated with the highest fat oxidation rate as the exercise intensity to be used for subsequent experimental visits.

While we acknowledge that more sophisticated statistical procedures (e.g., best-fit polynomial curves) have been used to identify the point of maximal fat oxidation, identifying this precise point was not a critical aspect of our design, since each boy served as his own control and we compared substrate utilization between trials performed at an identical exercise intensity. Heart rate (HR) was also monitored continuously during exercise using an HR monitor (Polar) and recorded at the end of each stage. The preliminary visit was completed at least 1 week before the experimental visits.

Experimental visits

The experimental visits were conducted in a double-blinded and counterbalanced manner, with the only difference between trials being the beverage consumed. Participants were given either a 6% 13C-enriched CHO beverage (CARB) or placebo beverage (CONT) to consume before and during exercise. To standardize habitual routine, each boy (and his parent) was asked to record his physical activity and dietary intake 24 h before each visit and to repeat these activities prior to the final visit. Participants were asked to avoid any corn and corn-derived food products, which contain 13C. The two experimental visits were separated by at least 5 days and were completed as close as possible to the same time of day for each boy, approximately 3 h following their last meal. Figure 1 outlines the protocol for each experimental visit.
Fig. 1

Schematic of protocol during experimental visits

Upon arrival at the laboratory, each boy provided a resting, baseline breath sample for background enrichment of 13C in the expired CO2, prior to the consumption of the experimental beverage. Oxidation rates of CHOtotal and FATtotal were calculated, as described above, from the last 2-min of a 3-min gas collection period (SensorMedics Vmax29, Yorba Linda, CA). A 60-ml syringe was used to draw a sample of the expired gas directly from the tube connecting the participant’s mouthpiece to the metabolic cart and later analyzed for the ratio of 13C/12C. Each boy then started cycling at the intensity estimated to approximate his highest fat oxidation rate. The total exercise duration was 60 min (3 × 20 min bouts, separated by 5-min rest periods). Additional expired breath samples were collected for 3-min periods during exercise (see Fig. 1) to calculate substrate utilization. During each exercise breath collection, an aliquot of breath was collected and later analyzed for the ratio of 13C/12C. Because of the presence of a large bicarbonate pool in the body and because of the delay in measuring 13CO2 production by the tissues at the mouth (Pallikarakis et al. 1991), CHOexo oxidation was calculated for the last 20 min of exercise only, as previously described (Timmons et al. 2003).

HR was monitored continuously during the session and recorded at 10-min intervals. Ratings of perceived exertion (RPE) using Borg’s 6–20 categorical scale were collected at 10-min intervals during the exercise, immediately after the recording of HR, as previously described (Timmons and Bar-Or 2003).

Experimental beverages

The CHO drink contained 60 g of glucose (or 240 kcal) per 1,000 ml of fluid (6% solution), whereas the placebo was artificially sweetened with 3.2 g of Splenda per 1,000 ml of fluid. The beverages were prepared in 13-l containers and coded appropriately by one of the investigators who was not involved in data collection or analysis. The CHO beverage was then artificially enriched with uniformly labeled 13C-glucose to an isotopic composition of +90.0 change per 1000 difference versus the 13C/12C ratio from the international standard 13C Pee Dee Belemnitella-1 (+90.0‰[δ-13C]PDB-1), confirmed by gas chromatography mass spectrometry. Both experimental beverages contained ~9 mM NaCl. The total volume of beverage consumed by each boy was calibrated to his FFM and divided into a bolus (16 mL × kg FFM−1) and four aliquots (4 mL × kg FFM−1). This approach delivered ~1.2 g/min of glucose, which matches our previous studies in non-obese boys while accounting for differences in body composition. The investigator responsible for data collection and the participants were blinded to the contents of the beverages.

Statistical analyses

Data are presented as means ± SD and were analyzed using Sigmastat statistical analysis software. HR, RPE and RER in each experimental session were compared using a two-way (session × time) repeated-measures ANOVA. Where appropriate, a Tukey’s HSD post hoc test was used to determine the location of significance among means. Variables related to substrate utilization, including EE, over the last 20 min in each experimental session were compared using one-tailed, paired t tests. One-tailed t tests were used because we a priori hypothesized that CHO intake would reduce fat oxidation. The threshold for statistical significance was set at p < 0.05 for all tests.


Participant characteristics

On average, the boys were 1.2 years before the age of peak height velocity. Additional characteristics of the boys were as follows: (mean ± SD) age: 11.2 ± 1.0 years; height: 1.5 ± 0.1 m; mass: 67.4 ± 14.8 kg; BMI%ile: 97.8 ± 0.8% body fat: 36.1 ± 4.2%.

Preliminary session

The highest fat oxidation rate during the incremental test averaged 5.8 ± 1.7 mg × kg FFM−1 × min−1. This represented an exercise intensity of 36.0 ± 10.8 W or 0.9 ± 0.3 W × kg FFM−1. The HR at this intensity was, on average, 114 ± 14 bpm.

Experimental sessions

HR and RPE did not show significant differences between CARB and CONT (Table 1). Over the last 20 min of exercise, \( {\dot{{V}}\text{CO}}_{2} \) and \( {\dot{{V}}\text{O}}_{2} \) remained stable, indicating steady state (Table 2). RER was lower during CONT versus CARB (p < 0.001; Table 2).
Table 1

Heart rate and ratings of perceived exertion during exercise in the carbohydrate (CARB) and placebo (CONT) trials


Session time (min)








Heart rate (bpm)


119.0 ± 14.6

121.3 ± 13.6

118.3 ± 12.9

122.7 ± 13.8

122.6 ± 12.6

122.3 ± 13.1


115.7 ± 19.9

120.9 ± 19.4

119.3 ± 17.5

118.1 ± 16.4

118.3 ± 15.5

123.7 ± 15.4

Ratings of perceived exertion


9.1 ± 3.1

9.9 ± 4.2

10.9 ± 3.7

11.7 ± 4.5

10.9 ± 3.8

12.0 ± 4.8


9.4 ± 3.3

10.6 ± 3.8

11.4 ± 5.0

11.6 ± 5.4

12.6 ± 5.6

12.9 ± 5.5

Values are mean ± SD

Table 2

\( {\dot{{V}}\text{O}}_{2} \), \( {\dot{{V}}\text{CO}}_{2} \), and RER measurements during last 20 min of exercise in the carbohydrate (CARB) and placebo (CONT) trials


40 min

50 min

60 min

\( {\dot{{V}}\text{O}}_{2} \)


0.71 ± 0.12

0.70 ± 0.17

0.70 ± 0.21


0.76 ± 0.17

0.74 ± 0.15

0.78 ± 0.21

\( {\dot{{V}}\text{CO}}_{2} \)


0.68 ± 0.11

0.66 ± 0.15

0.65 ± 0.19


0.67 ± 0.18

0.66 ± 0.17

0.68 ± 0.21



0.93 ± 0.02

0.95 ± 0.03

0.93 ± 0.02


0.88 ± 0.06

0.88 ± 0.07

0.86 ± 0.07

Values are mean ± SD. \( {\dot{{V}}\text{O}}_{2} \) and \( {\dot{{V}}\text{CO}}_{2} \) values are in l/min

Total EE, oxidation rates and energy contributions of FATtotal, CHOtotal, CHOendo and CHOexo are reported for the last 20 min of exercise. Total EE was slightly, but significantly (p = 0.04), higher during CONT (73.2 ± 18.4 kcal) than during CARB (69.6 ± 14.1 kcal). To compare with previous literature, oxidation rates are expressed relative to FFM (mg × kg FFM−1 × min−1) and are presented in Table 3. During CONT, FATtotal was 3.9 ± 2.4 mg × kg FFM−1 × min−1 and was significantly reduced (p = 0.02) to 1.7 ± 0.6 mg × kg FFM−1 × min−1 during CARB. CHOendo oxidation was 13.2 ± 6.1 mg × kg FFM−1 × min−1 during CONT and 12.2 ± 2.1 mg × kg FFM−1 × min−1 during CARB, and this did not reach statistical significance (p = 0.33). CHOtotal tended to be higher (p = 0.06) during CARB (17.2 ± 3.1 mg × kg FFM−1 × min−1) than in CONT (13.2 ± 6.1 mg × kg FFM−1 × min−1).
Table 3

Substrate utilization during the last 20 min of exercise in the carbohydrate (CARB) and placebo (CONT) trials








0.07 ± 0.01*

0.72 ± 0.14

0.51 ± 0.07

0.22 ± 0.08


0.16 ± 0.09

0.57 ± 0.32

0.57 ± 0.32


mg × kg FFM−1 × min−1


1.7 ± 0.6*

17.2 ± 3.1

12.2 ± 2.1

5.1 ± 1.5


3.9 ± 2.4

13.2 ± 6.1

13.2 ± 6.1




1.4 ± 0.7*

15.1 ± 3.3

10.9 ± 2.0

4.2 ± 1.6


2.9 ± 1.5

12.3 ± 6.3

12.3 ± 6.3


Values are mean ± SD. FATtotal total fat oxidation rate, CHOtotal total carbohydrate oxidation rate, CHOendo endogenous carbohydrate oxidation rate, CHOexo exogenous carbohydrate oxidation rate, AUC area under the curve in g. * Significantly different from CONT, p < 0.05

During CARB, the rate of CHOexo oxidation was 5.1 ± 1.5 mg × kg FFM−1 × min−1. The contribution of FATtotal, CHOtotal, CHOendo and CHOexo oxidation to total EE was calculated (Fig. 2). The energy contribution from FATtotal was decreased from 43.1 ± 22.9% during CONT to 19.8 ± 4.9% of total EE in CARB (p = 0.02), whereas the % EE from CHOtotal demonstrated the reciprocal response (80.1 ± 5.0% in CARB versus 56.9 ± 22.9% in CONT, p = 0.02). CHOendo did not show statistically significant changes in the energy contribution to total EE during CARB (Fig. 2), compared with CONT. CHOexo contributed to 23.3 ± 4.2% of total EE during CARB. To account for any variations across the last 20 min of exercise, we also calculated area under the curve for this time period (Table 3), with findings similar to oxidation rates.
Fig. 2

Percentage of energy contribution from substrate during the last 20 min of exercise in the carbohydrate (CARB) and placebo (CONT) trials. Values are mean ± SD. Hatched bars total fat oxidation, solid bars endogenous carbohydrate oxidation, open bars exogenous carbohydrate oxidation. *Significantly different between trials, p < 0.02


This study is the first to determine the impact of CHO intake on substrate utilization in obese pre- and early pubertal boys, using 13C stable isotope tracer to quantify exogenous carbohydrate oxidation. The main finding of our study was that CHO intake during exercise performed at an intensity corresponding to high whole body fat oxidation caused more than a 50% reduction in the rate of whole body fat oxidation. A further understanding of energy metabolism during exercise both with and without CHO intake in obese children should help inform exercise therapies to optimize fat oxidation for these children. These findings are consistent with substrate utilization studies conducted with non-obese boys, where the reliance on endogenous fuels during exercise was also reduced with CHO intake (Timmons et al. 2003, 2007b; Riddell et al. 2000b). To our knowledge, the current study is the first to report the contribution of CHOexo to total EE during exercise in obese boys and suggests that during exercise there is considerable reliance on this exogenous fuel source, which in turn blunts endogenous fuel utilization.

During more prolonged exercise in CONT (exercise without CHO intake), the fat oxidation rate, on average, was 4.0 mg × kg FFM−1 × min−1, which was lower than the ~6 mg × kg FFM−1 × min−1 found in non-obese boys (Timmons et al. 2007b). Interestingly, the degree of suppression in whole body fat oxidation observed in the current study (~50%) was similar to the ~55% reduction observed by us in non-obese boys (Timmons et al. 2007b), although boys in the latter study exercised at a higher intensity.

It can be argued that suppression of whole body fat oxidation during exercise is likely to be counterproductive for obese children trying to maximize fat oxidation and weight management through regular physical activity. In our study, CHO intake suppressed fat oxidation by ~50%, while maintaining total EE at the same level. Over the last 20 min of exercise, this would result in ~1.6 fewer grams of fat being oxidized for energy. At this rate of fat sparing, 1 kg of adipose tissue would be “protected” over a 12-month period if ~30 min of exercise with CHO intake was performed every day. The clinical significance of this scenario remains to be determined, but should be balanced against the possible benefits of a sport drink-type beverage, including fluid and electrolytes to promote hydration and assist thermoregulation, which may be less efficient in obese children (Dougherty et al. 2009). However, our study only shows the acute effects of lower intensity exercise and does not take into consideration changes that may occur post-exercise or the benefits of higher intensity exercise, which may result in greater total energy expenditure and greater post-exercise fat oxidation even though fat oxidation during exercise is low.

In light of the significant impact of CHO intake on substrate utilization in obese children, more research is needed to determine how this exogenous source of energy impacts on exercise capacity and other responses relevant to engaging in physical activity (e.g., perceived exertion) in these children. In our study, CHO intake did not influence RPE and the boys perceived the 60-min submaximal exercise task to be, on average, “fairly light” according to Borg’s 6–20 categorical scale. Whether CHO intake would influence perceived exertion during more vigorous exercise remains to be determined, and future studies should investigate children exercising in real-life situations.

One of the limitations of the current study is that greater control of the pre-exercise diet on the day of the experimental sessions could have been made, by providing the participants with pre-packaged food. Although we tried to replicate a real-life situation with our study design (e.g., children being active after school), we acknowledge that small variations in diet could have influenced our results. We did ask each boy to replicate their diet, although we did not measure dietary intake. We think it is unlikely that the dramatic 50% reduction in fat oxidation during exercise could have been solely explained by possible variations in dietary intake more than 3 h before the exercise session. We also tested a relatively small sample size and it is unknown if these results are representative of obese boys in general. Nevertheless, post hoc calculations revealed a statistical power of greater than 90%. Another limitation of our study is that we did not statistically model the point of maximal fat oxidation during the preliminary session. We modified our previously published approach to estimate an exercise intensity that was associated with a high fat oxidation rate. However, because each boy then exercised at the same intensity, the comparison between CARB and CONT trials remains valid.

In conclusion, CHO intake suppresses FATtotal and CHOendo oxidation in obese boys. The suppression of whole body fat oxidation with CHO intake in our study population may not be ideal for optimizing fat reduction. Future work should determine the relationship between substrate utilization and insulin resistance in obese children, as these results should have clinical implications.


We wish to thank the children for their participation and Melanie Wolfe for technical assistance in measuring 13C/12C of expired CO2 gas samples. Funding for this study was made possible by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada to B.W. Timmons. Current address for Lisa Chu is School of Kinesiology and Health Science, Faculty of Health, York University, Toronto, Ontario, Canada. Current address for Tim Takken is Wilhelmina Children’s Hospital, University Medical Center Utrecht, Utrecht, The Netherlands.

Copyright information

© Springer-Verlag 2011