The Effect of Ingesting Carbohydrate and Proteins on Athletic Performance: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Endurance athletes participating in sporting events may be required to complete multiple training sessions a day or on successive days with a limited recovery time. Nutritional interventions that enhance the restoration of endogenous fuel stores (e.g., liver and muscle glycogen) and improve muscle damage repair have received a lot of attention. The purpose of this review is to investigate the effect of ingesting carbohydrate (CHO) and protein (PRO) on athletic performance. Studies were identified by searching the electronic databases PubMed and EMBASE. Random-effects meta-analyses were conducted to examine the intervention efficacy. A total of 30 randomized controlled trials (RCT), comprising 43 trials and 326 participants in total, were included in this review. The meta-analysis showed an overall significant effect in Time-To-Exhaustion (TTE) and Time-Trial (TT) performance, when ingesting carbohydrates and proteins (CHO-PRO) compared to CHO-only (p = 0.03 and p = 0.0007, respectively). A subgroup analysis demonstrated a significant effect in TTE by ingesting CHO-PRO compared to CHO, when supplements were provided during and/or following an exercise bout. CHO-PRO significantly improved TTE compared to CHO-only, when a long-term recovery (i.e., ≥8 h) was implemented (p = 0.001). However, no effect was found when the recovery time was short-term (i.e., ≤8 h). No significant effect was observed when CHO-PRO and CHO-only supplements were isocaloric. However, a significant improved TTE was evident with CHO-PRO compared to CHO-only, when the supplements were matched for carbohydrate content (p < 0.00001). In conclusion, co-ingesting carbohydrates and proteins appears to enhance TTE and TT performance compared to CHO-only and presents a compelling alternate feeding strategy for athletes.


Introduction
Endogenous carbohydrate (CHO) is stored as liver and muscle glycogen [1][2][3]. Glycogen is a branched polymer of glucose where cells store and utilize glucose to meet their energetic demands. Muscle glycogen is a major energy source during prolonged moderate-to-high intensity exercise [4,5]. The development of fatigue during exhaustive exercise is often associated with low muscle glycogen concentrations [6,7]. Moreover, recent studies [6,7] have reported a relationship between low glycogen content and reduced calcium (Ca 2+ ) release from sarcoplasmic reticulum (SR), an association ultimately leading to fatigue. Muscle glycogen levels are usually restored to pre-exercise levels within 24 h, if a sufficient amount of carbohydrate (CHO) is provided [1,2]. The restoration of glycogen requires the translocation of glucose transporter carrier protein-4 isoform (GLUT-4) from the intracellular domain      (Table 1). They all included healthy participants ≥18 years. The majority of the investigations included male participants. In total approximately 79.8% of the participants were males and~20.2% were females (Table 1).

Study Protocol
All studies were of cross-over design, except one which was a parallel study [39]. Included investigations varied in protocol designs (e.g., duration of recovery, timing of supplementation and mode of exercise) ( Table 1). Approximately 80% of the studies included a prior depletion exercise of which~67% included a recovery period in their protocol ( Table 1).
The majority of the investigations conducted exercise protocols on a bike, including cyclists and/or triathletes, whereas four studies [13,30,38,39] assessed a running performance including, e.g., recreationally active males.

Intervention and Control Products
The combined ingestion of CHO-PRO was used as the intervention for all studies. This was compared to a control product, which contained carbohydrates only.

Effect of Carbohydrate and Protein (CHO-PRO) vs. Carbohydrate (CHO) Supplementation on Time-To-Exhaustion (TTE)
A total of 24 trials (n = 256 participants) derived from 16 publications investigated the effect of CHO-PRO on time-to-exhaustion performance. A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 3.62, CI: 0.44, 6.79, p = 0.03), favouring the intake of CHO-PRO with an overall effect size of Z = 2.23 min (Figure 3). A moderate heterogeneity was present across trials I 2 = 33% (p = 0.06).

Exhaustion (TTE)
A total of 24 trials (n = 256 participants) derived from 16 publications investigated the effect of CHO-PRO on time-to-exhaustion performance. A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 3.62, CI: 0.44, 6.79, p = 0.03), favouring the intake of CHO-PRO with an overall effect size of Z = 2.23 min (Figure 3). A moderate heterogeneity was present across trials I 2 = 33% (p = 0.06).

Effect of CHO-PRO vs. CHO Intake during TTE on Performance
Nine trials (n = 121 participants) examined the effect of CHO-PRO supplementation on TTE performance, when provided during exercise. This was compared to CHO.
A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 8.53, CI: 3.52, 13.53, p = 0.0008), favouring the intake of CHO-PRO with an overall effect size of Z = 3.34 min (Figure 4). A low heterogeneity was present across trials I 2 = 0% (p = 0.79).

Effect of CHO-PRO vs. CHO Intake during TTE on Performance
Nine trials (n = 121 participants) examined the effect of CHO-PRO supplementation on TTE performance, when provided during exercise. This was compared to CHO.
A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 8.53, CI: 3.52, 13.53, p = 0.0008), favouring the intake of CHO-PRO with an overall effect size of Z = 3.34 min (Figure 4). A low heterogeneity was present across trials I 2 = 0% (p = 0.79).
CHO-PRO on time-to-exhaustion performance. A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 3.62, CI: 0.44, 6.79, p = 0.03), favouring the intake of CHO-PRO with an overall effect size of Z = 2.23 min (Figure 3). A moderate heterogeneity was present across trials I 2 = 33% (p = 0.06).

Effect of CHO-PRO vs. CHO Intake during TTE on Performance
Nine trials (n = 121 participants) examined the effect of CHO-PRO supplementation on TTE performance, when provided during exercise. This was compared to CHO.
A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 8.53, CI: 3.52, 13.53, p = 0.0008), favouring the intake of CHO-PRO with an overall effect size of Z = 3.34 min (Figure 4). A low heterogeneity was present across trials I 2 = 0% (p = 0.79).

Effect of CHO-PRO vs. CHO Intake during Recovery on Subsequent TTE
Fourteen trials (n = 130 participants) examined the effect of CHO-PRO supplementation on TTE performance, when provided during a recovery period. This was compared to CHO. No significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 1.55, CI: −3.06, 6.16, p = 0.51) ( Figure 5). A significant heterogeneity was present across trials I 2 = 54% (p = 0.009).
Fourteen trials (n = 130 participants) examined the effect of CHO-PRO supplementation on TTE performance, when provided during a recovery period. This was compared to CHO. No significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 1.55, CI: −3.06, 6.16, p = 0.51) ( Figure 5). A significant heterogeneity was present across trials I 2 = 54% (p = 0.009).

Effect of CHO-PRO vs. CHO Intake during Long-Term Recovery on Subsequent TTE
Three trials (n = 37 participants) derived from three publications examined the effect of CHO-PRO supplementation on TTE performance, when conducting a long-term (i.e., >8 h) recovery. This was compared to CHO. A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 10.59, CI: 4.18, 17.01, p = 0.001), favouring the ingestion of CHO-PRO during recovery periods >8 h, with an overall effect size of Z = 3.24 min (Figure 7). A low heterogeneity was present across trials I 2 = 5% (p = 0.35). It should be noted, that Romano-Ely et al. Fourteen trials (n = 130 participants) examined the effect of CHO-PRO supplementation on TTE performance, when provided during a recovery period. This was compared to CHO. No significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 1.55, CI: −3.06, 6.16, p = 0.51) ( Figure 5). A significant heterogeneity was present across trials I 2 = 54% (p = 0.009).

Effect of CHO-PRO vs. CHO Intake during Long-Term Recovery on Subsequent TTE
Three trials (n = 37 participants) derived from three publications examined the effect of CHO-PRO supplementation on TTE performance, when conducting a long-term (i.e., >8 h) recovery. This was compared to CHO. A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 10.59, CI: 4.18, 17.01, p = 0.001), favouring the ingestion of CHO-PRO during recovery periods >8 h, with an overall effect size of Z = 3.24 min (Figure 7). A low heterogeneity was present across trials I 2 = 5% (p = 0.35). It should be noted, that Romano-Ely et al.

Effect of CHO-PRO vs. CHO Intake during Long-Term Recovery on Subsequent TTE
Three trials (n = 37 participants) derived from three publications examined the effect of CHO-PRO supplementation on TTE performance, when conducting a long-term (i.e., >8 h) recovery. This was compared to CHO. A significant overall effect was found between the ingestion of CHO-PRO and CHO on TTE performance (MD: 10.59, CI: 4.18, 17.01, p = 0.001), favouring the ingestion of CHO-PRO during recovery periods >8 h, with an overall effect size of Z = 3.24 min (Figure 7). A low heterogeneity was present across trials I 2 = 5% (p = 0.35). It should be noted, that Romano-Ely et al. 2006b [17] and Saunders et al. 2004b [16] provided supplementations both during exercise as well as during the recovery period between two exercise bouts.  [16] provided supplementations both during exercise as well as during the recovery period between two exercise bouts.

Effect of Isocaloric Supplementation (i.e., CHO-PRO vs. CHO) on TTE Performance
Six trials (n = 59 participants) derived from five publications examined the effect of isocaloric supplementation of intervention and control product on TTE performance. No significant effect was found between the ingestion of isocaloric CHO-PRO and CHO on TTE performance (MD: 1.27, CI:

Effect of Non-Isocaloric CHO-PRO vs. CHO Supplementation on TTE Performance
Seventeen trials (n = 186 participants) derived from 12 publications examined the effect of a nonisocaloric supplementation of intervention and control product on TTE performance. A significant effect was found between the ingestion of non-isocaloric CHO-PRO and CHO on TTE performance (MD: 3.99, CI: 0.12, 7.87, p = 0.04), favouring the intake of CHO-PRO with an overall effect size of Z = 2.02 min (Figure 9a). A moderate heterogeneity was present across trials I 2 = 44% (p = 0.03).
When doing a subgroup-analysis looking at the effect of CHO-PRO matched for volumetric content with the control product no significant effect was found (p > 0.5). However, when supplements were matched for CHO content (i.e., iso-CHO) a significant effect was evident favouring the ingestion of CHO-PRO compared to CHO-only on TTE performance (p < 0.00001) (Figure 9b).   [16] provided supplementations both during exercise as well as during the recovery period between two exercise bouts.

Effect of Isocaloric Supplementation (i.e., CHO-PRO vs. CHO) on TTE Performance
Six trials (n = 59 participants) derived from five publications examined the effect of isocaloric supplementation of intervention and control product on TTE performance. No significant effect was found between the ingestion of isocaloric CHO-PRO and CHO on TTE performance (MD: 1.27, CI: −4.73, 7.26, p = 0.68) (Figure 8). A low heterogeneity was present across trials I 2 = 0% (p = 0.45).

Effect of Non-Isocaloric CHO-PRO vs. CHO Supplementation on TTE Performance
Seventeen trials (n = 186 participants) derived from 12 publications examined the effect of a nonisocaloric supplementation of intervention and control product on TTE performance. A significant effect was found between the ingestion of non-isocaloric CHO-PRO and CHO on TTE performance (MD: 3.99, CI: 0.12, 7.87, p = 0.04), favouring the intake of CHO-PRO with an overall effect size of Z = 2.02 min (Figure 9a). A moderate heterogeneity was present across trials I 2 = 44% (p = 0.03).
When doing a subgroup-analysis looking at the effect of CHO-PRO matched for volumetric content with the control product no significant effect was found (p > 0.5). However, when supplements were matched for CHO content (i.e., iso-CHO) a significant effect was evident favouring the ingestion of CHO-PRO compared to CHO-only on TTE performance (p < 0.00001) (Figure 9b).

Effect of Non-Isocaloric CHO-PRO vs. CHO Supplementation on TTE Performance
Seventeen trials (n = 186 participants) derived from 12 publications examined the effect of a non-isocaloric supplementation of intervention and control product on TTE performance. A significant effect was found between the ingestion of non-isocaloric CHO-PRO and CHO on TTE performance (MD: 3.99, CI: 0.12, 7.87, p = 0.04), favouring the intake of CHO-PRO with an overall effect size of Z = 2.02 min (Figure 9a). A moderate heterogeneity was present across trials I 2 = 44% (p = 0.03).
When doing a subgroup-analysis looking at the effect of CHO-PRO matched for volumetric content with the control product no significant effect was found (p > 0.5). However, when supplements were matched for CHO content (i.e., iso-CHO) a significant effect was evident favouring the ingestion of CHO-PRO compared to CHO-only on TTE performance (p < 0.00001) (Figure 9b).

Effect of CHO-PRO vs. CHO Supplementation on Time-Trial (TT) Performance
Nineteen trials (n = 190 participants) derived from 14 publications examined the effect of a CHO-PRO on TT performance compared to CHO. A significant effect was found between the ingestion of CHO-PRO and CHO on TT performance (MD: −1.50, CI: −2.37, −0.63, p = 0.0007), favouring the intake of CHO-PRO with an overall effect size of Z = 3.39 min (Figure 10). A moderate heterogeneity was present across trials I 2 = 29% (p = 0.11).

Effect of CHO-PRO vs. CHO Supplementation on Time-Trial (TT) Performance
Nineteen trials (n = 190 participants) derived from 14 publications examined the effect of a CHO-PRO on TT performance compared to CHO. A significant effect was found between the ingestion of CHO-PRO and CHO on TT performance (MD: −1.50, CI: −2.37, −0.63, p = 0.0007), favouring the intake of CHO-PRO with an overall effect size of Z = 3.39 min (Figure 10). A moderate heterogeneity was present across trials I 2 = 29% (p = 0.11).

Discussion
The primary objective of the present systematic review and meta-analysis was to present the effect of co-ingesting carbohydrates and proteins on TTE performance when comparing to carbohydrate only. The key finding was that the ingestion of CHO-PRO significantly improves the

Discussion
The primary objective of the present systematic review and meta-analysis was to present the effect of co-ingesting carbohydrates and proteins on TTE performance when comparing to carbohydrate only. The key finding was that the ingestion of CHO-PRO significantly improves the overall effect in TTE performance compared to CHO-only by 2.23 min (Figure 3).
Co-ingestion of carbohydrates and proteins has gained attention as an alternate feeding strategy during limited recovery time (i.e., ≤8 h), as specific amino acids (AA) and proteins may have a potential effect in restoring the glycogen stores via insulin-mediated-pathways [10,14,40] and may promote muscle damage repair [41]. Many investigations have reported an effect with CHO-PRO on TTE performance, when comparing to CHO-only. However, not all investigations [11,12,17,[20][21][22]26,28,32,33,36,38,39] report additional improvements in athletic performance with CHO-PRO compared to CHO. Differences in protocol design across studies likely explain inconsistencies in findings of CHO-PRO supplementation. Differences in the duration of recovery across studies, could impact the performance outcome, as the glycogen stores are typically restored to pre-exercise levels in 20-24 h, given that sufficient CHO is provided [1,2]. A subgroup analysis on TTE performance showed, when ingesting supplements during exercise and/or during recovery it significantly prolongs TTE performance. However, in the present meta-analysis, a significant effect on performance was only evident during long-term recovery periods (i.e., ≥8 h), no significant effect was found when the recovery period was short-term (i.e., ≤8 h). Three trials [10,15,17] assessed the effects of a long-term (i.e., ≥8 h) recovery between two exercise bouts. The result from these trials revealed a significant longer subsequent TTE (p = 0.001) after ingesting CHO-PRO compared to CHO ( Figure 6). However, two of these trials [15,17] also provided supplementation during exercise. This could have had an additional effect in prolonging the performance duration to fatigue. In addition, when interpreting these results, one should also consider that only three trials examined the effect with long-term recovery. More studies implementing long-term recovery, when investigating the effect of CHO-PRO supplementation, are needed. Furthermore, differences in a preceding exercise to deplete the muscle glycogen levels exists across trials. First, some studies (n = 8) [15][16][17][18]27,30,32,39] did not include a preceding exercise bout to deplete the muscle glycogen levels, which could impact the potential effect of adding protein to a CHO beverage. Second, in studies including a prior depletion protocol the exercise protocol varied across investigations (e.g., mode, intensity, and duration) ( Table 1). In some studies, the preceding exercise bout was ≤70 min, which may not provide enough time to deplete the muscle glycogen content sufficiently. A~50% decline in muscle glycogen content has been reported, when cycling at >70% VO2max for minimum 50 min [42], but very low levels of glycogen have mostly been shown, when cycling for >1.5 h [1,2,8].
In addition, differences in the exercise type employed to investigate athletic performance exists across the trials. Many studies have used TTE as the performance test. However, in some studies a TT performance has been employed. A TT performance test is considered a more relevant physiological test, as it is a highly reliable and repeatable test and an effective predictor of cycling performance [43,44]. To our knowledge this is one of the first meta-analysis to comprehensively assess the effect of CHO-PRO vs. CHO on both TTE and TT, respectively, including both long-term and short-term recovery periods. In the present review, we found a significant effect in TT performance, when ingesting CHO-PRO compared to CHO-only ( Figure 10), with an improvement in performance by 3.39 min. The protocol design and duration of recovery varied across the pooled investigations using TT performance tests, these have likely contributed to the differences in performance outcome across the trials. The majority of studies conducted a preceding exercise bout [12,28,29,31,[33][34][35][36][37][38], while some studies did not [27,30,32,39]. Moreover, differences in duration and intensity of the preceding exercise bout was evident, which could result in different muscle glycogen levels prior to and during the TT performance test between trials. In addition, the duration of recovery also varied. Some studies did not include a recovery period [12,27,30,32,33,[37][38][39], whereas others implemented a short-term recovery [28,29,[34][35][36] and in one trial the recovery time was long-term [31]. Furthermore, there were inter-study differences in the TT performance tests with regards to distance and modality (e.g., cycling, running). Variables such as gender and age of subjects could also affect TTE and TT performance outcome. Physiological and morphological characteristics may account for both gender differences (e.g., differences in muscle mass, body fat, aerobic capacity due to genetic and hormonal responses) and age-related differences seen in athletic performance. Only two of the investigations included in this review compared the performance differences in regard to gender. Interestingly, both found no significant differences in TTE performance between men and women [15,21]. In addition, muscle damage repair and endogenous glycogen restoration are possible mechanisms which have been linked to improved recovery and performance. Prior trials have shown that CHO-PRO administration can increase the rate of muscle glycogen synthesis beyond that of CHO alone [8,40] and hence may contribute to improved exercise performance in a subsequent bout. However, a study by Ferguson-Stegall [34] demonstrated an improved TT but no significant differences in glycogen re-synthesis when comparing CHO-PRO with CHO-only supplement. Furthermore, some investigators have proposed that reducing muscle damage that occurs during endurance exercise is a possible mechanism for improved performance with CHO-PRO ingestion. Saunders et al. [15,16] has reported significant improvements in TTE concomitant with significant reductions in markers of muscle damage when comparing CHO-PRO to CHO-only [15,16]. While others have reported significant reduction in muscle damage without an improvement in performance [17,18]. Hence, the association between muscle damage, glycogen restoration and improved cycling performance is not fully elucidated. A study by Ferguson-Stegall et al. [34] demonstrated an enhanced protein synthesis (i.e., upregulated mTOR activation) with CHO-PRO supplementation compared to CHO-only.
Other methodological differences across investigations include differences in the supplements provided. In some studies, beverages were matched for caloric content, while in others they were matched for CHO content. A subgroup analysis in this review demonstrated that providing a higher caloric content by adding protein to a carbohydrate-matched supplementation significantly improved performance time (Figure 9b), whereas an isocaloric content did not (Figure 8). In addition, when supplements were matched for volume only, no significant effect was found with CHO-PRO compared to CHO. Indicating, that a sufficient CHO content is important in improving performance.
Investigating the effect of the different types of protein and CHO on athletic performance was beyond the scope of the present review. However, the majority (n = 18 studies) of the investigations included in this review used whey as the protein source [10][11][12][13][16][17][18][20][21][22]24,26,[30][31][32][35][36][37]. Proteins and specific amino acids (AA) have been shown to stimulate pancreatic secretion of insulin [45][46][47]. Especially, a synergistic effect has been observed by certain AAs to result in a potent stimulation of the β-cells [45,47]. Leucine, in particular, seems to be the potent stimulator of insulin secretion [8,45,48]. A study by Upshaw et al. [29] reported no effect in same-day-TT performance, when investigating the effect of different protein types, including animal-and plant-based proteins. In addition, a former study [49] observed a similar improved insulin sensitivity in type 2 diabetics, when comparing the effect of diets high in animal-based protein to diets high in plant-based protein. This suggests that the macronutrient and caloric content may be more important for athletic performance than the protein type. However, to our knowledge only a few studies have investigated the effect of plant-based protein on performance. Hence, more studies investigating the effect of different protein sources are therefore needed.

Future Research
The intent of this manuscript was to provide a systematic review of clinical trials investigating the effect of CHO-PRO on athletic performance compared to CHO. It is evident that methodological differences exist across trials to a great extent, which could explain the differences in performance outcome in the literature. Considering that the digestive properties and AA profile of proteins may differentially affect muscle protein synthesis and glycogen synthesis (by mediating insulin-secretion) [8,45,46,48], future research should assess whether manipulating the source of protein (e.g., animal-or plant-based) exerts enhanced effect on performance. Furthermore, more studies investigating long-term recovery, assessing muscle glycogen levels through biopsy techniques, as well as the timing of CHO-PRO intake (e.g., during exercise and/or recovery) are warranted.
Due to the need to consume beverages or gels containing high amount of sugars during sporting events, athletes have been considered as being at high risk for oral disease (e.g., carious lesions) [50,51]. Therefore, future research investigating the potential risk effects of energy supplements on dental health and risk of e.g., carious lesions in endurance athletes are needed, especially during sporting events lasting over several days, when the intake of high sugary supplements may be high.

Limitations
This review contains some limitations. First, only full-text articles written in English were included. Second, some of the trials tested the effect of treatment when subjects were fed while others when they were fasting, this could moderate the effect of CHO-PRO. Third, we did not take into account differences in the amount or type of protein provided between the trials. Fourth, when interpreting the results, one should consider the differences in the protocol design across studies. This includes the preceding exercise bout (e.g., duration and intensity) and timing of supplement intake, which could affect the level of glycogen depletion. Nevertheless, it was not possible to estimate the extent of glycogen depletion, based on the description in the included trials. In addition, not all studies measured the level of muscle glycogen via biopsies.

Conclusions
In the present Systematic Review and Meta-analysis, a total of 30 RCT studies were included, comprising 43 trials and 326 participants.
In conclusion, we found a significant overall effect in both TTE and TT performance, when CHO-PRO was ingested compared to CHO. A subgroup analysis demonstrated that this was significant when CHO-PRO was provided during and/or following an exercise bout. We found no significant effect in TTE, when the recovery time was short-term (i.e., ≤8 h), but TTE was significant, when a long-term recovery was implemented. Although the meta-analysis found no differences between CHO-PRO and CHO, individuals with a limited time to recover should ensure they have an adequate CHO intake to efficiently replenish glycogen deposits. Moreover, results demonstrate an ergogenic effect of CHO-PRO, offering performance benefit when protein is added to an optimal amount of CHO supplement (e.g., matched for CHO content) as opposed to supplements matched for caloric content.