*3.1. Blinding, Order E*ff*ects, and Adverse Events*

When queried as to which bar condition participants thought they had consumed, two participants guessed correctly, while all others responded that they were unsure; this indicated the success of the blinding. There were no order effects or order\*time interactions for any of the outcome variables (*p* > 0.05). There were no adverse events associated with the study. None of the participants complained of gastrointestinal discomfort after consumption of the bars.

#### *3.2. Glucose and Insulin Responses*

There was a condition\*time interaction for glucose and insulin. The high-GI condition resulted in higher glucose concentrations than the low-GI condition at 105, 90, and 60 min before the simulated soccer match (*p* < 0.05; Figure 3A). The insulin response was higher in the high-GI condition compared to the low-GI condition at two hours after bar consumption (*p* < 0.05; Figure 3B).

#### *3.3. Serum NEFA and Substrate Oxidation*

NEFA concentration was not different prior to the exercise test in low-GI vs. high-GI conditions (Figure 4). NEFA concentrations significantly increased in the low-GI and high-GI conditions at 45 min and 90 min of exercise (time main effect, *p* < 0.05; Figure 4). No significant difference was seen between the conditions. During the low-GI condition, carbohydrate oxidation was significantly lower compared to the high-GI condition (*p* < 0.05; Figure 5A). There was no difference between conditions for fat oxidation (*p* = 0.14; Figure 5B). There was a time main effect (*p* < 0.05) for fat oxidation with higher rates at 45–51 min versus 63–69 min and 81–87 min (*p* < 0.05). No significant difference was observed for lactate concentrations between the two conditions (Figure 6). There was a time main effect (*p* < 0.01) for lactate, with values increasing at all time-points, except at 54 min (i.e., after half time) compared to baseline (*p* < 0.05). The lactate at 54 min was lower than at 45 min and 90 min (*p* < 0.05).

#### *3.4. Skill Performance and Rating of Perceived Exertion*

There were condition\*time interactions for the agility and heading tests (*p* < 0.05). A significant improvement on the agility test and vertical jump height during simulated heading late in the soccer match (72 min) was observed after consuming the low-GI versus high-GI bar (Table 2; *p* < 0.01). No differences were apparent between bar conditions for skills performance of ball dribbling or kicking accuracy (Table 2). There was a time main effect for rating of perceived exertion (RPE; increasing throughout the simulated soccer match; *p* < 0.05); however, no significant difference was observed between the two conditions (mean RPE throughout the 90 min soccer match: low-GI = 5.2 ± 1.5 versus high-GI = 5.4 ± 1.3; *p* > 0.05).

(B)

**Figure 3.** (**A**) Plasma glucose concentrations before and during the simulated soccer match, (**B**) Insulin concentration before, at half time, and at the end of the simulated soccer match (\* *p* < 0.05 low-GI vs. high-GI sport nutrition bar). "Pre" denotes fingertip blood sample collection prior to consumption of the sport nutrition bars.

**Figure 4.** Non-esterified fatty acid concentrations before, at half time, and at the end of the simulated soccer match for low- and high-GI conditions. Values are means ± standard deviation (SD). Time main effect (*p* < 0.05) with 90 min > 45 min > pre (*p* < 0.05). "Pre" denotes venous blood draw prior to the simulated soccer match.

**Figure 5.** (**A**) Rate of carbohydrate oxidation, (**B**) and fat oxidation during the simulated soccer match in the low glycemic index and high glycemic index conditions. Values are means and SD. \* Main effect for condition for carbohydrate oxidation, with the low-GI condition lower than the high-GI condition; *p* < 0.05. There was a time main effect (*p* < 0.05) for fat oxidation (45–51 min > 63–69 min, 81–87 min; *p* < 0.05).

**Figure 6.** Lactate concentrations before and during the simulated soccer match between low- and high-glycemic index conditions. Values are means ± SD. Time main effect (*p* < 0.05) with all values except 54 min >pre, and 45 min, 90 min >54 min (*p* < 0.05).



aremeansandstandarddeviation(SD).\*Significantly differentinthelow-GIversushigh-GIcondition(*<sup>p</sup>*<0.01).

Values

#### **4. Discussion**

The main finding of this study was that a low-GI sport nutrition bar consumed two hours before and at half time during a simulated soccer match elicited lower carbohydrate oxidation throughout the match and improvements in agility performance and heading (i.e., vertical jump height) late in the match compared to a high-GI sport nutrition bar. In line with this potential for CHO supplementation to improve performance, in a systematic review, Russell and Kingsley stated that six out of eight included studies found that CHO ingestion in the form of 6%–8% solution of glucose, sucrose, or maltodextrin (which would have a high-GI; i.e., GI > 70 [8]) was linked with an improvement of at least one aspect of soccer skill performance [32]. However, to the best of our knowledge this is the first study to address the influence of low- and high-GI sport nutrition bars consumed shortly prior to prolonged, high-intensity, intermittent exercise, which is typical for many team sports. Although we found improvements in some performance measures with the low-GI sport nutrition bar condition at the 72 min time point, this did not persist to the 90 min time point of the simulated soccer match (Table 2). This may be due to lack of adequate statistical power, or perhaps the GI of the bar consumed does not make a difference this late in the match (i.e., glycogen depletion may be at a low enough level in both conditions to impair performance).

Our metabolic findings are in agreement with our previous work with soccer players. In our previous work, we showed that low-GI foods (i.e., lentils; with GI ranging from 29–36; where low-GI is defined as < 55 [8]) consumed before a simulated soccer match on a treadmill reduced carbohydrate oxidation [20], increased fat oxidation [21], and tended to reduce glycogen usage [22] compared to conditions where high-GI foods (i.e., instant mashed potatoes, white bread, and egg whites to match for protein; GI ranging from 75–81; where high-GI is defined as > 70 [8]) were consumed before exercise. In these previous studies we did not see any difference between low-GI and high-GI conditions when performance was evaluated by repeated sprints at the end of the simulated treadmill test. The current study used skill performance that was quite different from these previous studies and more specific to soccer performance. The current study also used a dietary condition (i.e., sports nutrition bars) which is more likely to be used by soccer players before matches when less time is available for food consumption [5].

In line with our previous studies [20,22] glucose concentrations were significantly higher in the high-GI condition in the first 60 min following consumption versus the low-GI condition. Consequently, insulin response in the high-GI condition was higher than the low-GI condition, which might explain the higher carbohydrate oxidation during the test in the high-GI condition. Insulin inhibits fat oxidation, necessitating greater carbohydrate oxidation and potentially greater glycogen usage [11]. Muscle glycogen is a major substrate during prolonged intermittent high-intensity exercise to provide high rate of ATP re-synthesis [3,4,33]. In this study, the carbohydrate oxidation rate was lower in the low-GI condition compared to the high-GI condition averaged across all time points (i.e., there was a "condition" main effect), and the potential for glycogen sparing might have contributed to improved exercise performance (i.e., agility running and heading) late in the simulated soccer match. In line with this, Saltin [6] showed that total walking distance and sprinting speed were reduced in soccer players with lower glycogen content versus those with higher glycogen content late in a soccer match. Bendiksen et al. [34] reported that utilization of muscle glycogen was significantly lower in the last 30 min of a match suggesting an important role of sustained availability of glucose later in the match. It should be noted, however, that we did not directly assess muscle glycogen in the current study; therefore, we cannot make conclusions on whether there was sparing of glycogen during the low-GI condition.

It has been postulated that a reduction in lipolysis rate, and therefore NEFA, will occur following high-GI pre-exercise meals [15,35]; however, we found no significant difference between low-GI and high-GI conditions for appearance of NEFA in the blood. This discrepancy might have been related to the greater intensity of our test protocol versus the other studies. During higher intensity exercise, plasma NEFA concentrations decrease while glucose and glycogen utilization increase in skeletal

muscle [36]. A limited rate of fat oxidation is thought to be connected with a lower flux of long chain NEFA across the mitochondrial membrane [37,38]. An alternative explanation to the lack of difference in NEFA concentration between conditions may be that the high-GI condition resulted in lower intramyocellular lipids utilization rather than lower lipolysis from adipose tissue.

The main limitation of our study was the assumption that reduced carbohydrate oxidation would lead to sparing of muscle glycogen. Direct analysis for glycogen levels by muscle biopsy would strengthen future studies comparing high- versus low-GI foods. We attempted to standardize glycogen levels between trials by having participants match their dietary intake and physical activity levels the day before each trial. For better control, it would be preferable to provide standardized meals to participants the day before trials. We would expect an increase in NEFA release from adipose tissue with the lower insulin concentration in the low-GI condition, but this was not observed. A limitation is that we did not assess glycerol, which may give more precise data on lipolysis. Another limitation is that we tested athletes after an overnight fast. The typical practice of a soccer player would most likely be to have a small breakfast and then consume a small amount of carbohydrate before the soccer match. We supplied enough of the bars to provide 1.5 g/kg available carbohydrate before the soccer match, an amount of carbohydrate that is recommended for improvement in endurance performance [8]. This required consumption of approximately five bars by each participant which totaled approximately 760 kcal (Table 1). We felt the addition of a breakfast before the bar consumption would result in excess fullness in participants. Additional limitations include a relatively small participant number, and the fact that our soccer match was simulated, rather than being an actual soccer match. Future studies could focus on the effects of consuming the bars before actual soccer matches.

#### **5. Conclusions and Practical Application**

Previous studies carried out in our lab using treadmill protocols to measure soccer performance (i.e., 1 min intervals of high-intensity running at the end of a simulated soccer match) were not very specific to soccer performance. This motivated the use of a field simulated soccer test incorporating soccer skills (i.e., agility, dribbling, kicking, and heading) to optimize the specificity of the test to the sport of soccer. Another novel aspect of the study was the assessment of sport nutrition bars given with adequate amount of recommended available carbohydrates (i.e., 1.5 g/kg) before endurance exercise performance. Previous research has generally shown that consumption of sports nutrition bars has no effect on endurance exercise performance [39–41]; however, these studies evaluated the effect of only a single sports nutrition bar before exercise. This would deliver well below the recommended amount of available carbohydrate for improving exercise performance; therefore, consumption of higher number of sports nutrition bars might seem practical. Sport nutrition bars containing high-CHO can act as an immediate snack, in particular, when soccer players are under time constraints before matches.

In conclusion, a low-GI sport nutrition bar consumed before a simulated soccer match elicited a lower carbohydrate oxidation rate and a modest improvement in performance (i.e., better agility and heading performance late in a simulated soccer match) versus a high-GI sport nutrition bar. Further studies are required to investigate how sport nutrition bars varying in GIs could impact soccer skill performance during prolonged match play (i.e., over/extra time, penalty kicks) when carbohydrate stores will be further depleted.

**Author Contributions:** The study was designed by P.D.C. and G.A.Z.; data were collected by M.K., J.J., and S.G.; data interpretation, analysis, and manuscript preparation were undertaken by M.K., P.D.C., and G.A.Z. All authors approved the final version of the paper. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was funded by Agri-food Canada and the Saskatchewan Pulse growers. The sponsors had no role in the design, execution, interpretation, or writing of the study.

**Conflicts of Interest:** The authors declare no conflict of interest.
