Effects of leucine or whey protein addition to an oral glucose solution on serum insulin, plasma glucose and plasma amino acid responses in horses at rest and following exercise


  • Funding: This study was supported by grants from Equine Guelph and Land O' Lakes-Purina Mills.

Dr Geor's present address is: Department of Large Animal Clinical Sciences, Michigan State University, D202 Veterinary Medical Center, East Lansing Michigan 48824, USA. Dr McCutcheon's present address is: Department of Pathobiology and Diagnostic Investigation, Michigan State University, F130G Veterinary Medical Center, East Lansing, Michigan 48824, USA. Email: klur222@uky.edu


Reasons for performing study: Providing protein or amino acid mixtures in combination with glucose to post exercise in man has resulted in increases in the post feeding insulin response and in muscle glycogen and protein synthesis rates. However, whether protein and/or amino acids can modify the post exercise insulin responses in horses remains to be fully elucidated.

Objectives: To determine whether whey protein or leucine addition to a glucose solution affects the post gavage plasma insulin, glucose and amino acid responses in horses and whether these responses are different following a period of exercise vs. rest.

Methods: Six mature, conditioned Thoroughbreds received a nasogastric gavage containing either 1 g/kg bwt glucose (G), G + 0.3 g/kg bwt whey protein (GW) or G + 0.3 g/kg bwt leucine (GL), following a period of either rest (R) or an exercise test on a high speed treadmill (EX). Each horse was studied under all 6 treatment conditions, separated by 10 day intervals. Blood samples were collected pre-exercise/rest, pregavage and at regular intervals up to 300 min post gavage. Plasma was analysed for glucose and amino acid concentrations and serum insulin concentrations were determined.

Results: There was a significantly (P<0.05) greater insulin response in GL-R and GL-EX when compared to the other treatments. When compared to rest, post exercise plasma glucose responses were lower in G and GW but unchanged following GL administration. Plasma alanine concentrations were elevated post exercise in all EX treatments. With the exception of markedly elevated plasma leucine concentrations after GL-R and GL-EX, the plasma concentrations of all indispensable amino acids decreased during the post gavage period.

Conclusions: Leucine but not whey protein augmented the serum insulin response to an oral glucose load. Leucine supplementation warrants further investigation as a means to increase the rate of post exercise muscle glycogen synthesis in horses.


Plasma insulin concentrations may be a key determinant of the rate of post exercise glycogen replenishment in horses. Horses have a slow rate of muscle glycogen replenishment following glycogen-depleting exercise, taking up to 72 h to replenish the muscle glycogen pool following high-intensity exercise (Hyyppa et al. 1997; Lacombe et al. 2004) when compared to <24 h in rats and man (Costill et al. 1981; Garetto et al. 1984). Although i.v. glucose administration to horses following exercise has been shown to increase the rate of glycogen replenishment, a similar effect was not seen when the same amount of glucose was administered orally (Geor et al. 2006), suggesting that glucose delivery from the gastrointestinal tract may limit post exercise glycogen synthesis. Furthermore, the insulinaemic response to the oral glucose was ∼50% lower than with i.v. glucose administration (Geor et al. 2006). Although feeding high soluble carbohydrate diets post exercise will increase glycogen synthesis relative to low soluble carbohydrate diets, the post feeding insulinaemic responses reported are ∼3x lower than those reported following i.v. glucose infusion and the rates of glycogen synthesis in the first 6–8 h post exercise were ∼50% lower (Lacombe et al. 2004; Geor et al. 2006; Jose-Cunilleras et al. 2006). Additionally, a recent study found horses do not display an increase in insulin sensitivity following exercise (Pratt et al. 2007). When these observations are considered together it would appear that a pronounced insulinaemic response may be necessary to stimulate muscle glycogen synthesis post exercise. Because of the potential adverse side effects associated with feeding large amount of soluble carbohydrates to horses, including gastrointestinal dysfunction and laminitis (Tinker et al. 1997; Geor and Harris 2009), it is important to investigate non-glucose alternatives to increase insulin secretion.

In man, there is a synergistic effect on plasma insulin concentrations when carbohydrates and protein or amino acids are fed together (Zawadzki et al. 1992; van Loon et al. 2000a,b,c). When a solution providing leucine, phenylalanine and wheat protein hydrolysate was ingested with a moderate amount of carbohydrate (50% glucose, 50% maltodextrin), the insulinaemic response at both rest (van Loon et al. 2000c) and following exercise(van Loon et al. 2000a) was doubled and the post exercise rate of glycogen synthesis was also increased (van Loon et al. 2000b), compared to carbohydrate ingestion alone. Similar effects have been observed when whey protein was added to a carbohydrate (dextrose-maltodextrin) mixture (Zawadzki et al. 1992). In horses, the addition of leucine to a glucose solution following a competition exercise test increased plasma insulin concentrations compared to glucose alone; however, there was no effect on muscle glycogen synthesis (Poso and Hyyppa 1999). On a bodyweight basis, however, the previous equine study only provided 23–45% the amount of leucine and <10% of the carbohydrate (Poso and Hyyppa 1999) administered in the human studies (van Loon et al. 2000a,b,c) and did not study the effects of leucine on the insulinaemic response independent of exercise. In order to gain a greater understanding of the potential role that leucine and other amino acids could play in the stimulation of post-feeding insulin secretion, it is necessary to determine the insulinaemic response of horses, following periods of both rest and exercise, when leucine or protein levels more comparable to those shown to be effective at increasing the insulinaemic response and glycogen synthesis in man (Zawadzki et al. 1992; van Loon et al. 2000a,b,c) are added to a carbohydrate mixture. If there is a significant increase in the insulinaemic response when either leucine or protein is added to the carbohydrate solution, this will provide the basis for future research investigating whether glycogen synthesis increases in response to these same supplements.

The objective of the present study was to determine the effects of leucine or whey protein hydrolysate addition to a glucose solution on serum insulin, plasma glucose and plasma amino acid concentrations following periods of either rest or exercise. Leucine and whey protein were selected for use in this study because they represent a single amino acid and a complete protein, respectively, which have both been shown to increase insulin secretion in man.

Materials and methods


The Animal Care Committee at the University of Guelph approved all experimental procedures. Six mature (age 4–8 years) Thoroughbred horses (2 mares and 4 geldings; mean ± s.d. bwt 449 ± 37 kg) were housed in box stalls (3.5 × 4 m) for the duration of the experimental period. The daily allocation of grass hay (∼6 kg/horse/day; 10.2% crude protein [CP], 46.1% acid detergent fibre [ADF], 54.7% neutral detergent fibre [NDF], 3.1% fat, 11.4% nonstructural carbohydrates [NSC] on a dry matter [DM] basis) and concentrate (∼4 kg/horse/day; 12.9% CP, 23.5% ADF, 26.4% NDF, 10.2% fat, 23.5% NSC on a DM basis) was divided into 2 equal portions and meals were fed at 07.00 and 15.30 h. Water and salt were provided ad libitum. Horses had undergone at least 6 weeks of physical conditioning on a high speed treadmill1 before the experiment and fitness was maintained during the study by running on the treadmill 3 days/week for 30–35 min at a workload equivalent to 45–55% of the peak rate of oxygen consumption (VO2peak).

Before the experiments, horses completed an incremental exercise test on the treadmill for measurement of the VO2peak, as previously described (Pratt et al. 2007). For each horse, linear regression analysis was used to determine the speeds corresponding to 50, 70 and 100% of VO2peak, and these speeds were used for the exercise protocol.

Experimental design

This study used a randomised crossover design with 2 different exercise treatments (either rest [R] or exercise [EX]), followed by one of 3 supplementation treatments: 1 g/kg bwt glucose (G) as a 10% (w/v) solution, G + 0.3 g/kg bwt whey protein2 (GW) or G + 0.3 g/kg bwt leucine2 (GL), for a total of 6 treatments. The whey protein and leucine were mixed with the glucose solution to form a suspension. Taking into account the indispensable amino acid composition of whey protein hydrolysate (van Loon et al. 2000c) (mg/g protein: histidine: 16, isoleucine: 51, leucine: 87, lysine: 84, methionine: 13, phenylalanine: 24, threonine: 66, tryptophan: 12, valine: 45), horses on the GW treatment received 0.03 g/kg bwt leucine. The GL and GW treatments were designed to provide equal amounts of total amino acids (0.3 g/kg bwt), rather than providing equal amounts of leucine, to determine whether any response to the leucine addition was a leucine-specific response or a response that could be elicited by amino acids in general. Each horse was studied on all 6 treatments, with a minimum interval of 10 days between treatments.

Experimental protocol

Exercise protocol: Regardless of whether horses were supplemented following rest or exercise, feed was withheld for 10 h prior to the start of the exercise period to standardise metabolic state prior to treatment. When horses were studied following exercise, the exercise test consisted of a 5 min walk (1.6 m/s), a 5 min warm-up trot (4 m/s), followed by 10 min at a speed that elicited 50% VO2peak (5.0 ± 0.5 m/s, mean ± s.d), 10 min at 70% VO2peak (7.1 ± 0.2 m/s) and three 1 min sprints at 100% VO2peak (10.9 ± 0.3 m/s) with a 5 min walk between and after the sprints. After the initial walk period, the incline on the treadmill was set at 3°. The total distance completed during the exercise test was 9.4 ± 0.2 km. In previous studies in our laboratory, this exercise protocol resulted in a 40–50% decrease in muscle glycogen stores (Geor et al. 2006; Pratt et al. 2007). Horses undergoing the rest treatment stood quietly on the treadmill for an equivalent period of time. All horses were weighed3 before and after the exercise treatment period.

Treatments: Fifteen minutes after completion of the exercise period (EX or R) in each experimental period, a single gastric gavage containing the G, GW or GL supplement was administered to each horse. The leucine and whey doses (0.3 g/kg bwt) were similar to the intakes that resulted in increased plasma insulin concentrations and post exercise glycogen synthesis rates in man (0.2–0.4 g/kg bwt/h) (van Loon et al. 2000a,b) and the glucose intake used was chosen because it was the midway point between the different carbohydrate intakes used in these same human studies (0.8 and 1.2 g/kg bwt/h) (van Loon et al. 2000a,b). Furthermore, when carbohydrate intake was 1.2 g/kg bwt/h, this higher carbohydrate intake appeared to eliminate any benefit of supplemental amino acids or protein on rates of muscle glycogen synthesis (Jentjens et al. 2001) and therefore a lower carbohydrate intake was chosen to provide adequate amounts of glucose for a glycogen substrate without providing too much. Following the gavage, horses were loosely restrained in a box stall. Horses were provided access to fresh water but feed was withheld until the end of the sampling period.

Collection of blood samples: Catheters (Angiocath, 14 gauge, 11.5 cm)4 were inserted percutaneously using aseptic technique into one of the jugular veins. Prior to catheter insertion, the skin overlaying the catheterisation site was desensitised by the subcutaneous injection of 2% mepivacaine (Carbocaine)5. The jugular vein that was catheterised was alternated between experimental study periods. In each of the 6 periods, a blood sample was taken prior to the exercise/rest period (preEX), at the end of the exercise/rest period, prior to the administration of the supplements (postEX) and then following treatment administration: every 10 min for the first hour, every 15 min for the next hour and then every 30 min up to 5 h post gavage. Samples were placed into Vacutainers6 containing sodium heparin or no anticoagulant and were centrifuged for 15 min at 1600 g. Plasma and serum were harvested and stored at −20°C until subsequent analysis.

Sample analysis: Plasma glucose concentration was measured in triplicate using a commercially available spectrophotometric assay kit (Infinity Reagent)7. Serum insulin concentrations were measured in duplicate using a radio-immunoassay kit (Coat-a-Count)8 validated for use in equine samples (Reimers et al. 1982). The intra- and interassay coefficients of variance were 2.2 and 3.1%, and 4.5 and 6.3% for the glucose and insulin assays, respectively. The area under the curve (AUC) for glucose and insulin was determined using the trapezoidal rule using GraphPad Prism 49. Plasma amino acid concentrations were determined in the preEX, postEX, 60 min and 300 min samples using reverse-phase HPLC of phenylisothiocyanate derivatives (Bidlingmeyer et al. 1984; Murch et al. 1996).

Statistical analysis

All data were analysed using the mixed procedure of SAS Version 9.110 and results are expressed as least squares means ± s.e. In all cases, differences were considered significant at P<0.05 and when fixed effects were significant, pre-planned comparisons of the least squares means were made using the pdiff option.

The effects of rest vs. exercise prior to supplement administration on glucose and amino acid concentrations and serum insulin concentrations were assessed using a repeated measures analysis with exercise treatment, time (preEX or postEX) and the interaction of treatment and time as the fixed effects and horse nested in exercise treatment as the subject. The effects of exercise-supplement combination on post-supplement administration serum insulin and plasma glucose were assessed using repeated measures analyses with treatment (exercise-supplement combination), time and the interaction between treatment and time as the fixed effects and horse nested in treatment as the subject. For the amino acid data, values at t = 60 min and 360 min were converted to a percentage of the postEX value and these values were analysed using the repeated measures analysis as described above. For all repeated measures analyses the variance-covariance matrix was selected based on the lowest value for Schwarz's Bayesian criterion. The glucose and insulin AUC data were analysed using a one-way analysis of variance with treatment (exercise-supplement combination) as the fixed effect and horse nested in treatment as the random effect.


All horses completed each of the 6 exercise-supplement combinations. The mean ± s.d. decrease in bodyweight during the exercise protocol were 8.8 ± 3.1, 8.4 ± 3.6 and 9.2 ± 3.8 kg for horses receiving the G-EX, GW-EX and GL-EX treatments, respectively. There was no change in bodyweight when horses remained at rest.

Exercise effects on baseline glucose, insulin and amino acid concentrations

In the horses that remained at rest, there was no effect of time (P>0.05; preEX vs. postEX) on any of the plasma or serum variables studied (Table 1). There was a significant effect of time (P = 0.02) and the exercise by time interaction (P = 0.05) on serum insulin concentrations, with serum insulin concentrations decreasing ∼33% in the EX horses at the postEX time point (Table 1). Exercise resulted in an increase of ∼25% in plasma glucose concentrations (time: P<0.0001; exercise × time: P<0.0001) (Table 1). The only plasma amino acid concentration that changed during the exercise period was alanine, which was approximately doubled postEX vs. preEX in the horses that underwent exercise (exercise: P = 0.002; time: P<0.0001; exercise × time: P<0.0001) (Table 1).

Table 1. Plasma metabolite concentrations in trained, mature Thoroughbreds before (preEX) and after (postEX) a period of rest or treadmill exercise
  1. Values are least squares means ± s.e., n = 4–6/treatment, as determined using the mixed procedure of SAS. abWithin an exercise treatment, values with a different superscript are significantly different. *Indicates that the value is significantly different from the rest treatment at the same time point.

Plasma glucose concentration (mmol/l)    
 Glucose4.9 ± 0.14.9 ± 0.14.6 ± 0.1a5.8 ± 0.1b*
Serum insulin concentration (µiu/ml)    
 Insulin4.7 ± 1.14.4 ± 1.15.4 ± 1.13.6 ± 1.1*
Plasma amino acid concentrations (µmol/l)    
 Alanine179 ± 23182 ± 24208 ± 22a385 ± 22b*
 Arginine62 ± 568 ± 572 ± 574 ± 5
 Glutamine149 ± 21153 ± 22155 ± 20198 ± 20
 Histidine37 ± 243 ± 339 ± 239 ± 2
 Isoleucine42 ± 751 ± 742 ± 648 ± 6
 Leucine83 ± 2294 ± 2481 ± 21148 ± 21
 Lysine67 ± 877 ± 980 ± 886 ± 8
 Methionine7 ± 18 ± 16 ± 17 ± 1
 Phenylalanine53 ± 458 ± 457 ± 467 ± 4
 Proline84 ± 693 ± 794 ± 698 ± 6
 Threonine55 ± 2057 ± 2284 ± 1982 ± 19
 Tyrosine49 ± 454 ± 457 ± 462 ± 4
 Valine147 ± 35165 ± 37147 ± 33218 ± 33

Effect of supplement on plasma glucose and serum insulin concentrations

Time (P<0.0001), exercise-supplement combination (P = 0.003) and the time by exercise-supplement interaction (P = 0.0009) significantly affected serum insulin concentrations (Fig 1a, b). The AUC for insulin (Fig 1a, b), was greater (P<0.05) when horses received the GL supplement (Table 2).

Figure 1.

Least squares means±s.e., n=6/treatment, serum insulin concentrations in mature, Thoroughbred horses after the nasogastric administration of either glucose (G=inline image) or the glucose+leucine (GL=▴) or glucose+whey (GW=○), following a period of either rest (Fig 1a) or exercise (Fig 1b). Overall, there was a significant effect of treatment (P=0.003), time (P<0.0001) and the treatment by time interaction (P=0.0009) on serum insulin concentrations (means separations not shown).

Table 2. The effects of a nasogastric administration of either glucose (1 g/kg bwt), glucose + leucine (1 g/kg bwt + 0.3 g/kg bwt) or glucose + whey (1 g/kg bwt + 0.3 g/kg bwt), following a period of either rest or exercise, on the glycaemic and insulinaemic response in trained, mature Thoroughbreds
GlucoseGlucose + leucineGlucose + wheyGlucoseGlucose + leucineGlucose + wheyPooled s.e.P value
  1. Values are least squares means, n = 6/treatment. abcdWithin a row, values with different superscripts indicate a significant dietary effect (P<0.05).

Area under the plasma glucose response curve (0–300 min)
[(mmol/l)* min]
Area under the plasma insulin response curve (0–300 min)
[(µiu/ml)* min]

The time following supplement administration (P<0.0001) and the interaction of time and the exercise-supplement combination (P = 0.009) significantly affected plasma glucose concentrations (Fig 2a, b). The glycaemic response following supplement administration was significantly impacted by treatment (P<0.0001), with the GL-R, GW-R and GL-EX treatments having a higher AUC than the remaining treatments (Table 2). Also of interest was the very low glycaemic response to glucose only administration following exercise (G-EX). In this treatment, plasma glucose concentrations did not differ from the t = 0 min (postEX) value by 105 min post supplement and then fell below the t = 0 value at 270 min (Fig 2b), resulting in an overall glycaemic response that was ∼10x lower than the response following any of the other exercise-supplement regimens (Table 2).

Figure 2.

Least squares means±s.e., n=6/treatment, plasma glucose concentrations in mature, Thoroughbred horses after the nasogastric administration of either glucose (G=inline image) or the glucose+leucine (GL=▴) or glucose+whey (GW=○), following a period of either rest (Fig 1a) or exercise (Fig 1b). Overall, there was a significant effect of time (P<0.0001) and the treatment by time interaction (P=0.009) on plasma glucose concentrations (means separations not shown).

Post supplement plasma amino acid concentrations

The primary interest was whether the supplements altered plasma amino acid concentrations and whether the plasma amino acid response to the supplements was affected by rest vs. exercise. To facilitate these comparisons, all amino acid concentrations (Table 3) are expressed as a percentage of the t = 0 value (postEX; set at 100). There was a significant effect of time (P<0.05) on the plasma concentrations of alanine, arginine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, tryptophan, tyrosine and valine concentrations (Table 3), a significant effect (P<0.05) of exercise-supplement combination on the plasma concentrations of alanine, isoleucine, leucine, threonine and valine (Table 3) and a significant interaction of time and exercise-supplement combination on the plasma concentrations of alanine, valine, isoleucine and leucine (Table 3). The horses receiving the glucose + leucine supplement (GL-R and GL-EX) had plasma leucine concentrations that were 6.5–9.5x greater than the pre-supplementation level both at t = 60 and t = 300 min and the plasma concentrations of the other branched chain amino acids (isoleucine and valine) decreased more in the horses when they received the GL-R and GL-EX treatments than when they received the other exercise-supplement regimes (Table 3). The plasma alanine concentrations were higher at the time of supplement administration in horses following exercise (Table 1) and decreased more post-supplementation in the horses following the G-EX, GL-EX and GW-EX treatments than following the G-R, GL-R and GW-R treatments (Table 3). However, by t = 300 min, there was no difference in the plasma alanine concentrations between exercise treatment regimes (µmol/l, mean ± s.e.; G-R: 165 ± 31, GL-R: 165 ± 33, GW-R: 162 ± 31, G-EX: 228 ± 31, GL-EX: 196 ± 31, GW-EX: 236 ± 33).

Table 3. Changes in plasma amino acid concentrations in trained, mature Thoroughbreds that received a nasogastric administration of either glucose, glucose + leucine or glucose + whey following a period of either rest or exercise
 Time (min)RestExercise
GlucoseGlucose + leucineGlucose + wheyGlucoseGlucose + leucineGlucose + whey
  1. Values are least squares means ± s.e., n = 4–6/treatment, as determined using the mixed procedure of SAS. abcdWithin a row, values with a different superscript are significantly different. *Indicates that the value is significantly different from the postEX value (t = 0; 100).

Plasma amino acid concentration as a % of postEX (t = 0)
 Alanine6084 ± 6ab71 ± 6a*95 ± 6b77 ± 8ab*75 ± 6a*85 ± 6ab
30083 ± 6b*88 ± 6b93 ± 6b61 ± 6a*54 ± 6a*60 ± 6a*
 Arginine6087 ± 799 ± 692 ± 775 ± 888 ± 775 ± 7*
30059 ± 7*68 ± 7*59 ± 7*56 ± 6*67 ± 7*58 ± 7*
 Glutamine6088 ± 3384 ± 30158 ± 3380 ± 4385 ± 3086 ± 33
30075 ± 3392 ± 3372 ± 33143 ± 3085 ± 3074 ± 33
 Histidine6074 ± 9*83 ± 8104 ± 8106 ± 11102 ± 898 ± 8
30083 ± 985 ± 885 ± 889 ± 879* ± 888 ± 8
 Isoleucine6088 ± 10cd49 ± 9b*100 ± 10d69 ± 13bc18 ± 9a*89 ± 11cd
30056 ± 10c*8 ± 10a*53 ± 10bc*46 ± 9bc*29 ± 9b*51 ± 11bc*
 Leucine6087 ± 92a715 ± 84b*93 ± 92a64 ± 116a657 ± 84b*74 ± 92a
30053 ± 92a780 ± 92b*47 ± 92a50 ± 84a953 ± 84b*40 ± 92a
 Lysine6096 ± 1392 ± 1297 ± 1363 ± 1758 ± 12*106 ± 13
30082 ± 1380 ± 1384 ± 1372 ± 1245 ± 12*68 ± 13*
 Methionine6088 ± 1197 ± 1098 ± 13108 ± 1581 ± 10105 ± 11
30075 ± 1149 ± 11*58 ± 13*79 ± 1055 ± 13*86 ± 11
 Proline6083 ± 7*71 ± 7*99 ± 793 ± 983 ± 792 ± 7
30066 ± 7*51 ± 7*71 ± 7*75 ± 7*62 ± 7*74 ± 7*
 Phenylalanine6092 ± 778 ± 6*87 ± 781 ± 978 ± 6*82 ± 7
30090 ± 751 ± 7*61 ± 7*64 ± 6*54 ± 6*65 ± 7*
 Threonine6085 ± 11ab94 ± 11ab116 ± 11c73 ± 13a75 ± 10a112 ± 12bc
30068 ± 11a*73 ± 11a62 ± 11a*51 ± 10a*68 ± 10a*109 ± 12b
 Tyrosine6093 ± 676 ± 6*95 ± 682 ± 887 ± 6102 ± 6
30061 ± 6*40 ± 6*55 ± 6*60 ± 6*55 ± 6*67 ± 6
 Valine6091 ± 6bc78 ± 6ab*95 ± 6c74 ± 8ab*68 ± 6a*89 ± 6bc
30069 ± 6b*37 ± 6a*66 ± 6b*69 ± 6b*31 ± 6a*70 ± 6b*


The present study showed that following periods of both rest and heavy exercise, addition of leucine to an orally administered glucose solution increased both the insulinaemic and glycaemic response compared to glucose alone (Figs 1 and 2; Table 2) and altered plasma amino acid concentrations.

Leucine supplementation increased the post feeding insulinaemic response

In the present study, the serum insulin response was elevated over glucose administration alone in horses receiving the GL supplement (Table 2, Fig 1), confirming that leucine is an insulin secretagogue in horses (Poso and Hyyppa 1999). The apparent increase in insulin secretion in response to GL administration appears to be related to leucine specifically, as opposed being a general response to dietary amino acids, because the administration of the GW supplement, which provided an equal amount of amino acids to the GL supplement, did not result in any significant increase in insulinaemic response over glucose alone (Table 2, Fig 1). The inability of the whey protein hydrolysate (GW) to elicit a more pronounced increase in serum insulin concentrations is in contrast to research in man, where whey protein administration, following 90 min of exercise at 50–85% of VO2peak, resulted in increased circulating insulin concentrations, relative to carbohydrate administration alone (Zawadzki et al. 1992). In that study (Zawadzki et al. 1992), however, the post exercise intakes (at both 0 and 2 h post exercise) of both whey and milk protein (∼0.55 g/kg bwt) and carbohydrate (∼1.5 g/kg bwt) were greater than the intakes provided to the horses in the present study; therefore, it is possible that the horses in the present study did not receive sufficient whey to elicit a significant increase in the post supplement insulinaemic response.

The peak serum insulin concentrations in horses following administration of GL supplement (∼70 µiu/ml; Fig 1) were similar to the mean serum insulin concentration (75 µiu/ml) during a 0.5 g/kg bwt/h i.v. infusion of glucose (Geor et al. 2006) and were substantially higher than mean serum insulin concentrations following oral glucose (35 µiu/ml) (Geor et al. 2006) or high soluble carbohydrate intake (<∼25 µiu/ml) (Lacombe et al. 2004; Jose-Cunilleras et al. 2006). If the lack of increase in insulin sensitivity in horses following exercise (Pratt et al. 2007) is the key mechanism responsible for the slow rate of muscle glycogen resynthesis in horses (Hyyppa et al. 1997; Lacombe et al. 2004), then it is possible that a threshold insulinaemic response must be exceeded before muscle glycogen synthesis rates can increase. The serum insulin concentrations due to the GL supplement were similar to those resulting from i.v. glucose administration (Geor et al. 2006) and therefore oral leucine plus glucose administration would appear to be an effective alternative to i.v. glucose administration in increasing serum insulin concentrations and potentially rates of glycogen synthesis. In the study by Poso and Hyyppa (1999), leucine addition to a glucose supplement given to horses following exercise was associated with a ∼3–4x increase in plasma insulin concentrations, but there was no difference in muscle glycogen content in 22.5 h after an exercise test compared to glucose supplementation alone. Not only did the present study provide ∼3x more leucine and ∼15x more glucose post exercise than the previous study (Poso and Hyyppa 1999), but there was also a much larger increase in circulating insulin levels (∼13x) compared to pre-supplement levels in response to the GL supplement (Fig 1). Future research must now investigate whether these elevated serum insulin concentrations in response to the GL supplement are in fact associated with comparable rates of glycogen replenishment following exercise as the i.v. glucose infusion (Geor et al. 2006).

Leucine supplementation increased the glycaemic response

All 3 supplements provided the same amount of glucose orally; however, horses receiving the GL supplement had the greatest glycaemic response, with the overall response exceeding that of glucose administration alone by 2 and 31 times, following a period of rest and exercise, respectively. Although the carbon skeleton of leucine cannot be converted to glucose, based on research in other mammals such as rodents and man, it appears that leucine administration can promote an increase in plasma glucose concentrations in 2 different ways. First, an increase in circulating leucine activates the enzyme branched chain ketoacid dehydrogenase in the skeletal muscle of rodents (Aftring et al. 1986), which increases the oxidation of all 3 branched chain amino acids for use as a metabolic fuel (reviewed by Jitomir and Willoughby 2008), sparing the oxidation of glucose. Second, when leucine is oxidised, the resulting amino group is used to form other amino acids such as alanine, which is a key precursor for hepatic gluconeogenesis (Ruderman 1975). If similar metabolic pathways are present and active in the equine muscle, then the elevated glycaemic response following GL administration could indicate a sparing of glucose from being used as a metabolic fuel or an increase in gluconeogenesis. Although both of these possible scenarios for the increased glucose concentrations measured in the GL horses require experimental confirmation, the end result was that there was probably more substrate (glucose) available for glycogen synthesis when horses received the GL supplement.

In both the post rest and post exercise states, the glycaemic response to the GW supplement was intermediate to the G and the GL responses (Table 2; Fig 2), suggesting that there also was glucose sparing and/or increased gluconeogenesis when compared to glucose administration alone, but not to the same extent as when leucine was administered. In contrast to findings in the present report, human studies did not show an increase in glycaemic response, relative to that resulting from glucose ingestion alone, when amino acids or protein hydrolysates were added to a glucose solution given following a period of exercise (van Loon et al. 2000a) or rest (van Loon et al. 2000c), providing further evidence that whole-body glucose metabolism is differently regulated in horses vs. man.

In man, carbohydrate intake plays an important role in the extent to which protein/amino acid supplementation can improve post exercise rates of muscle glycogen repletion. Van Loon et al. (2000b) and Zawadzki et al. (1992) showed that the addition of an amino acid/protein mixture to a solution providing ∼0.8 g/kg bwt/h of carbohydrates resulted in an increase in the rate of muscle glycogen replenishment, compared to carbohydrate ingestion alone. However, 1.2 g/kg bwt/h of carbohydrate ingestion resulted in similar rates of glycogen synthesis compared to the carbohydrate (0.8 g/kg bwt/h)/amino acid/protein hydrolysate mixture (van Loon et al. 2000b) and there was no further increase in rates of muscle glycogen synthesis when the amino acid/protein hydrolysate mixture was provided in combination with 1.2 g/kg bwt/h of carbohydrate (Jentjens et al. 2001). These findings suggest that the amino acids and/or protein mixture may be most effective at increasing rates of muscle glycogen repletion when carbohydrate intake is relatively low. In horses, where we are often concerned about moderating the amount of carbohydrate ingested, these amino acid/protein supplements could provide a means to increase rates of glycogen repletion without providing excessive amounts of carbohydrates. In the present study, we only investigated one level of glucose intake and did not measure rates of glycogen synthesis and therefore additional research is necessary to show: 1) that glycogen synthesis rates can be increased by providing supplemental protein/amino acids in combination with a carbohydrate source and 2) how carbohydrate type and intake level influences the glycogen synthesis response to supplemental amino acids/protein.

Leucine supplementation impacted plasma branched chain amino acid concentrations

Plasma leucine concentrations were 8–10-fold higher than prefeeding values (Table 3) following GL supplement administration. However, plasma concentrations of the other branched chain amino acids, isoleucine and valine, were markedly decreased (to <∼30% of prefeeding values) after GL administration (Table 3). The plasma concentrations of the other indispensable amino acids also declined after GL administration although decreases were smaller in magnitude, raising the possibility that factors other than those related to protein synthesis contributed to the decreases in isoleucine and valine. There is also evidence that a marked increase in circulating leucine concentration is associated with decreases in isoleucine and valine concentrations in rodents (May et al. 1991) and man (Hagenfeldt et al. 1980; Eriksson et al. 1981), probably due in part to the activation of the enzyme branched chain α-keto acid dehydrogenase that results in increased catabolism of all branched-chain amino acids (Aftring et al. 1986).

If the availability of any one of the indispensable amino acids becomes limiting to protein synthesis, then protein synthesis cannot occur despite any leucine-related activation of the translation initiation factors (Escobar et al. 2007). Therefore, it is possible that the GL-supplement had negative effects on protein synthesis, which could have serious consequences for post exercise muscle recovery, although future research is required to evaluate the effect of the GL supplement on protein synthesis. In man, feeding large amounts of leucine without the other branched chain amino acids substantially reduced plasma isoleucine and valine concentrations; however, when the leucine was fed in combination with a protein hydrolysate, there were much smaller changes in isoleucine and valine concentrations, with no reduction in the insulinaemic response (van Loon et al. 2000c). Leucine supplementation in combination with other amino acids as a means to increase the post feeding insulinaemic response and muscle glycogen synthesis requires investigation in the horse.


The addition of leucine to a glucose solution given to horses at rest or following a period of exercise resulted in elevated serum insulin concentrations levels and appeared to increase whole-body glucose availability when compared to glucose alone. The serum insulin concentrations achieved with the leucine supplementation were comparable to those achieved with i.v. glucose infusion (Geor et al. 2006), and higher than those resulting from oral glucose administration (Fig 2b and Geor et al. 2006). Therefore, post exercise leucine administration may have beneficial effects on the post exercise rates of glycogen synthesis without the potential risks associated with oral soluble carbohydrate feeding. However, because leucine addition to the glucose solution resulted in a sharp decrease in plasma concentrations of valine and isoleucine, which could limit protein synthesis, the inclusion of a protein hydrolysate alongside the leucine should be investigated.


The authors gratefully acknowledge technical assistance provided by Leah Larsen and Liana Stewart-Hunt.

Conflicts of interest

There are no conflicts of interest to disclose for any of the contributing authors.

Manufacturers' addresses

1 Sato, BIAB Industrial, Uppsala, Sweden.

2 Sigma Aldrich, Oakville, Ontario, Canada.

3 KSL, Kitchener, Ontario, Canada.

4 Abbott Laboratories, Chicago, Illinois, USA.

5 Deseret, Sandy, Utah, USA.

6 Becton-Dickinson, Parsippany, New Jersey, USA.

7 Thermo Electron Corporation, Waltham, Massachusetts, USA.

8 Diagnostic Products Corporation, Los Angeles, California, USA.

9 GraphPad, La Jolla, California, USA.

10 SAS Institute Inc., Cary, North Carolina, USA.