*2.7. Statistical Analysis*

A sample size estimation was conducted (G\*power, sample size estimator v.3.1.9.4; Kiel, Germany) for F-test family in a repeated measures design, using the following parameters: average effect size of night time supplementation of 0.39 on CHO oxidation [28], alpha level of 0.05, and minimal power of 0.8, which revealed a minimum sample size of 13 participants. Enrollment was targeted beyond this minimum in the possible event of dropout. All statistical analyses were performed using commercially available software (SPSS v.25, IBM, Armonk, NY, USA). A one-way repeated measures ANOVA was used to determine if differences existed at baseline across conditions (PLA, HGI, LGI). A two-way repeated measures ANOVA was used to analyze the potential impact of condition (PLA, HGI, LGI), exercise intensity (55, 65, 75% VO2peak), and their potential interaction on HR, RPE, GID, RER, % FAT and % CHO utilization. Two-way repeated measures ANOVA models were used to compare condition (PLA, HGI, LGI), distance (each km), and their potential interaction on RPE, HR, and GID during the 5-km TT. Lastly, a two-way ANOVA was used to determine potential differences across condition (PLA, HGI, LGI), time (baseline, one and ten minutes post 5-km TT), and their potential interaction on BG. Tests of normality were performed and Greenhouse–Geisser corrections were utilized if sphericity was violated. As men and women were recruited for this study, exploratory multi-variate ANOVA measures were conducted including sex as an independent covariate in the model for the above analyses. Significant main effects were followed up using Tukey's Honestly Significant Difference, and *p* values were complemented by effect size, which in this model we used partial eta squared (η2). Alpha was set at 0.05. Data are presented as means ± standard deviation.

#### **3. Results**

#### *3.1. Participants*

Fourteen healthy endurance trained males (*n* = 8) and females (*n* = 6) completed all visits for this study. An overview of subject characteristics is presented in Table 1. There were no differences in self-reported sleep duration between visits (*p* = 0.56, η<sup>2</sup> = 0.05, Table 1).


**Table 1.** Subject Characteristics.

**Table 1.** *Cont.*


**Note:** Data expressed as means ± SD. VO2peak, peak oxygen uptake. \* *p* < 0.05 male vs. female.

#### *3.2. E*ff*ects of Supplement on Baseline Measures*

Resting metabolic data are displayed in Table 2. There was no significant effect of supplement on baseline REE (PLA, 1689 <sup>±</sup> 278; HGI, 1701 <sup>±</sup> 308; LGI, 1732 <sup>±</sup> 287 kcal·day−1, *p* = 0.72, <sup>η</sup><sup>2</sup> = 0.03; Table 2). There was a significant interaction of supplement and sex for baseline RER, and thus relative substrate utilization where males displayed a higher %FAT (PLA, 47.6 ± 5.4; HGI, 51.3 ± 11.5; LGI, 48.9 <sup>±</sup> 11.8%, *<sup>p</sup>* <sup>=</sup> 0.02, <sup>η</sup><sup>2</sup> <sup>=</sup> 0.28) utilization compared to females (PLA, 46.9 <sup>±</sup> 13.9; HGI, 28.3 <sup>±</sup> 13.7; LGI, 34.5 <sup>±</sup> 22.2%, *p* = 0.02, <sup>η</sup><sup>2</sup> = 0.28) at rest for HGI and LGI. However, all other baseline measures were unaffected by the supplement at baseline (all *p* > 0.05, Table 2).

**Table 2.** Baseline Measurements.


**Note:** Data are means ± SD. PLA: placebo; HGI: high glycemic index; LGI: low glycemic index; VAS: visual analogue scale; GID: gastrointestinal distress; REE: resting energy expenditure; BG: blood glucose; USG: urine specific gravity. Data expressed as means ± SD.

#### *3.3. E*ff*ects of Supplement on the Response to the Incremental Exercise Test (IET)*

On average, during the IET, the LGI supplement tended to utilize less FAT (PLA, 44.1 ± 10.5; HGI, 39.7 <sup>±</sup> 13.0; LGI, 37.5 <sup>±</sup> 13.7%, *p* = 0.17, <sup>η</sup><sup>2</sup> = 0.14) and more CHO (PLA, 56.4 <sup>±</sup> 10.6; HGI, 60.1 <sup>±</sup> 14.3; LGI, 63.1 <sup>±</sup> 13.9%, *<sup>p</sup>* <sup>=</sup> 0.17, <sup>η</sup><sup>2</sup> = 0.14; Figure 2) than the other two supplements, though this did not reach statistical significance. During the IET, there was no significant effect of supplement on VO2 (*p* = 0.23, η<sup>2</sup> = 0.11, Figure 2C) or RER (*p* = 0.17, η<sup>2</sup> = 0.14, Figure 2D). There was a tendency for an interaction of supplement with intensity for VO2 where values tended to be lower with the PLA during the lower intensity but equalized in the latter stages (*p* = 0.08, η<sup>2</sup> = 0.18, Figure 2D). Expectedly, all metabolic parameters were significantly affected by exercise intensity (all *p* < 0.001, all η<sup>2</sup> > 0.90, Figure 2A–D). The IET elicited an increase in GID (*p* = 0.04, η<sup>2</sup> = 0.23, Figure 3) and RPE (*p* = 0.00, η<sup>2</sup> = 1.00, data not shown). Supplementation had no effect on GID (*p* = 0.28, η<sup>2</sup> = 0.10, Figure 3) or RPE (*p* = 0.55, η<sup>2</sup> = 0.05, data not shown).

**Figure 2.** Metabolic Response to incremental exercise test (IET) at 55, 65, and 75% of VO2peak between placebo (PLA), high glycemic index (HGI), and low glycemic index (LGI) supplements (*n* = 14). (**A**) Relative fat utilization (%FAT), (**B**) relative carbohydrate utilization (%CHO), (**C**) respiratory exchange ratio, and (**D**) VO2. Data expressed as means ± SD. \* effect of intensity, *p* < 0.001.

**Figure 3.** Gastrointestinal distress (GID; categorical scale) during incremental exercise trial (IET) at 55, 65, and 75% of VO2peak (*n* = 14) between placebo (PLA), high glycemic index (HGI), and low glycemic index (LGI) supplements. Data expressed as means ± SD. \* effect of intensity, *p* = 0.04.

#### *3.4. E*ff*ect of Supplement on 5-km TT*

Supplement had no impact on HR during the 5-km TT (*p* = 0.89, η<sup>2</sup> = 0.01, Figure 4A). Although there was a significant effect of running distance during the 5-km TT on GID (categorical scale) (*p* = 0.00, η<sup>2</sup> = 0.58), there was no significant effect of supplement or an interaction of supplement by distance on GID (categorical scale) during the 5-km TT (Figure 4B). RPE was not impacted by supplement (*p* = 0.35, η<sup>2</sup> = 0.01, Figure 4C). Running performance during the 5-km TT was unaffected by supplement (PLA, 21.6 <sup>±</sup> 9.5; HGI, 23.0 <sup>±</sup> 7.8; LGI, 24.1 <sup>±</sup> 4.5 min, *<sup>p</sup>* <sup>=</sup> 0.94, <sup>η</sup><sup>2</sup> <sup>=</sup> 0.01, Figure 4D).

**Figure 4.** (**A**) Heart rate (HR), (**B**) gastrointestinal distress (GID, CS), (**C**) rating of perceived exertion (RPE) and (**D**) time (min) for 5-km time trial performance between placebo (PLA), high glycemic index (HGI), and low glycemic index (LGI) supplements (*n* = 14). Data expressed as means ± SD. \* significant effect for distance *p* < 0.05.

#### *3.5. E*ff*ect of Supplement on Perceptual Responses of GID and Satiety to Exercise*

Supplement had no significant effect on satiety from pre- to post-experimental trial (*p* = 0.39, η<sup>2</sup> = 0.08; Figure 5A). There was no significant effect of supplement or time on pre- to post-experimental trial GID (VAS) (Figure 5B, *p* = 0.56, η<sup>2</sup> = 0.03).

**Figure 5.** *Cont.*

**Figure 5.** (**A**) Satiety (mm; VAS) and (**B**) gastrointestinal distress (GID; mm; VAS) at baseline (pre) and at conclusion (post) of running the 5-km time trial performance between placebo (PLA), high glycemic index (HGI), and low glycemic index (LGI) supplements (*n* = 14). Data expressed as means ± SD.

#### *3.6. Blood Glucose (BG)*

There were significant main effects for time (*p* = 0.00, η<sup>2</sup> = 0.66), where blood glucose at baseline (PLA, 97.7 <sup>±</sup> 8.1; HGI, 99.4 <sup>±</sup> 8.8; LGI, 98.4 <sup>±</sup> 9.3 mg·dL−1) was significantly increased immediately post-exercise (PLA, 127.2 <sup>±</sup> 19.4; HGI, 131.0 <sup>±</sup> 28.8; LGI, 124.4 <sup>±</sup> 27.9 mg·dL<sup>−</sup>1, *<sup>p</sup>* <sup>=</sup> 0.00) and ten minutes post-exercise (PLA, 127.6 <sup>±</sup> 23.9; HGI, 133.3 <sup>±</sup> 24.6; LGI, 126.7 <sup>±</sup> 23.3 mg·dL<sup>−</sup>1, *<sup>p</sup>* <sup>=</sup> 0.00; Figure 6), but no differences were observed between immediate and ten minutes post exercise. There were no significant differences in BG between supplements (*p* = 0.54, η<sup>2</sup> = 0.04) at any time point or an interaction (*p* = 0.87, η<sup>2</sup> = 0.02).

**Figure 6.** Blood glucose levels (mg·dL<sup>−</sup>1) at baseline, 1 min post 5-km TT run, and 10 min post 5-km TT run trial between placebo (PLA), high glycemic index (HGI), and low glycemic index (LGI) supplements (*n* = 14). Data expressed as means ± SD. \* significant effect of time, *p* = 0.00.

#### **4. Discussion**

The present study is the first to assess the effects of pre-sleep supplementation with a novel LGI CHO as compared to HGI CHO or placebo control on next-morning (~8 h later) exercise metabolism, GID, and endurance performance in male and female endurance athletes. It was hypothesized that the nighttime pre-sleep consumption of LGI CHOs would increase fat utilization during morning exercise, decrease GID, and improve 5-km TT performance. The primary findings were as follows: (1) supplementation had no significant effect on REE, CHO, or FAT utilization at rest, though females tended to utilize more CHO in the HGI and LGI supplement at rest; (2) supplementation had no significant effect on substrate utilization during graded submaximal exercise; (3) blood glucose was not different among supplements at any point during the trial; (4) perceptions of GID were not different among supplements; (5) supplementation had no discernable significant effect on 5-km TT performance. Although our data do not support our original hypothesis, the present study suggests that there are no detrimental effects of supplementing with either LGI or HGI CHO pre-sleep in endurance athletes and thus, they may be utilized as a feeding window and fueling strategy to ingest adequate daily energy intake.

The gastrointestinal tract can be very sensitive to the foods and beverages we consume. Unfortunately, nutrient ingestion prior to and during exercise may lead to GID. Baur and colleagues (2016) reported that GID increased after the consumption of the same hydrothermally modified starch (HMS) LGI supplement that our current study used [15]. Baur et al. (2016) compared the HMS to an HGI CHO supplement when ingested prior to, and during, prolonged cycling in ten trained male cyclists and triathletes [15]. It was reported that there were likely large correlations between mean sprint nausea (*r* = −0.51) and total GID (*r* = −0.53) and exercise trial, showing that GID contributed to reduced cycling performance [15]. Further, there was a HMS-associated increase in GID negatively effecting sprint cycling performance [15]. Given that HMS is slow releasing under normal digestion supplements, malabsorption may be the explanation for the primary pathophysiologic mechanism of LGI CHO-induced GID during exercise. Unlike the findings of Baur et al. (2016), the present study found no effect (positive or negative) on GID and performance. Perhaps the pre-sleep ingestion of LGI CHO avoids the LGI CHO-induced increase in GID in morning endurance performance. This is likely because the body can digest the LGI CHO during the overnight period. Participants in the Baur et al. (2016) study consumed LGI CHO during the exercise as well, which likely caused the incidences of GID with HMS ingestion [15].

An LGI CHO may still be an optimal source of CHO for athletes given its previously reported low osmolality, low insulin impact, slow release factor, and maintenance of blood glucose levels [5,24–26]. In general, elevated insulin levels attenuate lipolysis and fat metabolism, thus increasing utilization of CHO. Even though it is well documented that consuming LGI carbohydrates before exercising results in enhanced fat oxidation, or at least maintaining euglycemia during exercise [17,18,21,39–43], and possibly improved performance [44], though not all agree [16,20]. Data from the present study, albeit in a different methodological approach, do not support these findings, as we found no effect of LGI CHO, or HGI CHO for that matter. When comparing LGI to HGI, some studies have reported enhanced exercise performance [19,39,43,45,46] while other studies report no differences [41,42,47–50]. For example, Baur et al. (2016) reported an increase in total FAT oxidation and reduction in CHO oxidation with LGI supplementation 30 min before as well as during exercise [15], which disagrees with the findings of the present study utilizing pre-sleep supplementation of LGI. These inconsistencies may be explained by, principally, time but other methodological differences, such as timing or dose of CHO supplementation, type of exercise protocol (i.e., cycling versus running), or sample size should also be considered. Researchers have reported muscle glycogen sparing with LGI compared to HGI CHO [47], which may be explained by improved fat oxidation. Our findings are in accordance with previous literature that LGI and HGI CHO do not improve running TT performances [21,28].

Glycemic control is extremely important for those training and competing in endurance competitions and increasing fat oxidation could potentially benefit performance by preserving glycogen stores [51]. To maximize glucose fueling, the timing of pre-exercise consumption of CHO is essential, along with the type/amount of exercise being performed. The time of consumption may alter the metabolic effects. Studies have shown that CHO consumed one to four hours prior to exercise resulted in a decline in glucose and insulin basal levels prior to exercise [2,52]. Further research has reported that CHO consumed ≤ 60 min before exercise leads to elevated blood glucose and insulin levels immediately prior to exercise [47,53–55]. These findings emphasize the importance of nutrient timing and the exploring how the body performs from nutrient consumption solely the night before exercise takes place.

The trend towards higher CHO utilization during exercise after pre-sleep consumption of HGI or LGI CHO, perhaps more so in LGI, might suggest that pre-sleep LGI CHO supplementation increases morning CHO availability or more stable bioavailability, though more research is needed as this was not directly investigated in the present study. Due to the exercise paradigm used in the current study, the 5-km TT run lasting ~20–30 min could present itself as a higher intensity glycolytic exercise than longer endurance exercise performance trials. Research on the effects of CHO feeding for endurance exercise indicates that some measures of performance are more sensitive than others, and short duration exercises may not be long enough to cause CHO depletion and reveal potential effects of pre-sleep CHO supplementation [12]. This might explain the insignificant differences in 5-km TT performance in the current study, and perhaps longer bouts, and/or larger sample sizes, are required to reveal an effect. There was, however, a significant effect between supplement and sex for resting CHO and FAT oxidation in this study, where females utilized more CHO with LGI and HGI (PLA was consistent between sexes). This suggests females resting fuel selection may respond differently to pre-sleep LGI or HGI CHO supplementation, but further work is needed.

In the present study, which utilized a graded and shorter duration endurance event, we found no benefit with pre-sleep ingestion on enhancing exercise performance. A contributing factor for the lack of significant positive impact on exercise performance may be attributed to the relatively short duration of the exercise stimulus incorporated in the present study [12], the amount of CHO, and/or sample size. When exercise is prolonged in a moderately intense state, CHO oxidation gradually decreases while fat oxidation increases [51,56]. Muscle glycogen utilization decreases due to reduced muscle glycogen availability [57] hence why CHO supplementation is vital for exercise of longer duration since the body relies on CHO as fuel [13]. The exercise module that was used in the present study was based on previous literature that found an effect of nighttime feeding altering morning metabolism in a 10-km run [57], and was preceded by an incremental exercise trial of three five-minute stages at 55, 65, and 75% VO2peak [57]. That protocol was altered to test a 5-km timed trial with an IET comprised of three three-minute stages at the same intensities. A main reason why those times were chosen include efficiency and time restraints. Additionally, not measuring substrate utilization during the 5-km TT limited the current study's understanding of substrate metabolism to only the initial nine-minute incremental test but this was intentional to allow the athletes to give their best efforts and be minimally distracted. Contrary to our hypothesis, we found that pre-sleep supplementation with LGI CHO tended to the lowest FAT oxidation as compared to HGI and placebo control. In the present study, we cannot ascertain the mechanisms responsible such as altered intramuscular CHO availability, or altered bloods level of glucoregulatory hormones (i.e., insulin and glucagon).

#### *Experimental Considerations*

Future studies should consider measuring exercise performance in live race scenarios, such as overland 5-km running events with performance feedback, for longer duration endurance bouts (e.g., 10 km, half-, or full-marathon), and explore optimal dosing strategies. Additionally, future work should determine if CHO availability is altered with pre-sleep CHO feeding by examining muscle glycogen, and with further consideration for sex differences, as females were shown to have higher CHO utilization than males at rest following both HGI and LGI pre-sleep supplementation. This observation contrasts with relatively established findings, but several factors could have contributed to females utilizing more CHO in the morning; we would like to acknowledge that the study was not designed to test sex differences and there were fewer female participants (*n* = 6) and larger studies may prove otherwise. Females also had, on average, lower VO2peak value (49.9 ± 4.3 mL/kg/min vs. males at 59.5 ± 5.5 mL/kg/min), and lower body weight and thus higher relative CHO loading and thus fitness level and body weight may play a role. Another consideration of this study could be the dose of CHO that was administered; 66 g of CHO may not be enough to last the ~eight hours to the exercise trial. Future studies should investigate different dosages, dosing approaches (e.g., g/kg), and/or timings of nighttime CHO supplementation for next-morning endurance performance in a larger sample, with measurements of circulating glucoregulatory hormones or muscle glycogen which could provide greater mechanistic insight.
