The results showed that repeated post exercise administration with a mixture of leucine and glucose caused a marked increase in plasma leucine concentration and decreased levels of the other BCAA and methionine. This agrees with the results observed post-exercise in humans after repeated supplementation with glucose together with a mixture of whey, leucine and phenylalanine . A recent study on horses reported a similar effect on the BCAA profile after a single gastric gavage of leucine and glucose in the early recovery period after glycogen depleting treadmill exercise . In the present study, plasma leucine concentrations increased 4-5 fold after supplementation with leucine, which is less than in the study by Urschel et al.  where plasma leucine concentrations increased 8-10 fold. In that study, the dose was higher (0.3 g/kg BW) and given as a single per oral dose whereas the present study used a lower dose (0.1 g/kg BW) given twice. The administration of leucine at 0 and 4 h post-exercise was chosen since oral administration of leucine stimulates the insulin response for 4 h in horses . Despite this, the decreases in the plasma isoleucine and valine concentrations were in parity to those seen by Urschel et al. .
Although amino acid concentrations tended to be lower during the post-exercise period in the GLU+LEU treatment, significant changes were only found for the other BCAA and methionine. Decreased concentrations of aromatic amino acids and other free amino acids have been seen in humans after leucine infusion or oral intake of leucine or BCAA at rest [8, 10, 11] and before exercise . The changes in plasma amino acids seem to follow a certain pattern, with most marked decreases occurring in isoleucine, -55%, methionine -55%, valine -40%, tyrosine -35%, and phenylalanine, -35% [8, 10]. In the present study, administration of leucine and glucose was given post exercise and the magnitude of change for isoleucine, valine and methionine were more marked than in humans but similar to the values obtained in horses . The differences in results between studies in humans versus horses could be related to species differences, when the mixture was given (at rest or post exercise), the route of administration, the dosage of leucine as well as the sampling time.
Unlike the previous study in horses , the horses in the present study were sampled repeatedly over a period of 360 min post administration, making it possible to study when the changes in amino acid concentrations occurred. The changes in isoleucine occurred earlier (within 60 min after the first leucine supplementation) than the changes in valine and methionine (120 min and 270 min, respectively). This is in agreement with a previous study in humans were methionine levels decreased 45-180 min post ingestion of BCAA . If this is a reflexion in how fast these different amino acids are metabolised in muscle and other tissues (liver, kidney) is not known. Concentrations of plasma free amino acids are often difficult to interpret as they are influenced by many factors such as diet , release from tissues (kidney, liver, muscle) as well as red blood cells.
The lack of correlation between the insulin and isoleucine, valine and methionine responses during the recovery period indicates that the changes in these amino acids are not related to the increased insulin response. This is in contrast to results from a study in human athletes were a negative correlation between the insulin response and the concentrations of plasma amino acids was found . In the latter study the athletes were given a mixture of carbohydrates and amino acids (0.1 g/kg/h leucine and 0.1 g/kg/h phenylalanine) after performing glycogen-depleting exercise. The reason for this discrepancy in results between studies could be related to type and dosage of amino acid supplementation and when it was administered, as well as the degree of insulin response. It is not likely that the marked insulin response was too low to influence the plasma amino acid concentration. Studies in horses have also failed to show an increase in glycogen re-synthesis after supplementation with leucine and glucose during the early post-exercise period after glycogen depleting exercise, despite a marked insulin response [7, 21]. In addition to this, horses do not show increased insulin sensitivity after glycogen depleting exercise in contrast to humans . Taken together, these findings indicate that horses have a different response to insulin (insulin resistant) compared to humans. In humans an increased glycogen re-synthesis [4, 17] is seen after per oral leucine supplementation post exercise as well as an insulin dependent decrease in amino acids [15, 17]. Reduced concentrations of isoleucine, valine and methionine in combination with a low insulin response have been reported in humans after intravenous supplementation with leucine during resting conditions . These results suggest that the leucine induced fall in amino acid concentration is independent of insulin, which is supported by the findings from the present study. Although leucine is known to stimulate insulin secretion in humans and horses [5, 6], the present study could not find a correlation between the leucine and insulin response. One possible explanation for this could be that there is a dose dependant response in insulin until a certain threshold level is reached, and above this the relationship ceases to exist.
The observed changes in amino acid concentrations are likely related to a direct effect of leucine on protein metabolism as leucine has been shown to stimulate protein synthesis and inhibit protein degradation [19, 24, 25]. The reduced amino acid concentrations could also be related to an up regulation of the α-ketoacid dehydrogenase complex  or an increased uptake of amino acids to muscle or liver. The fact that methionine was decreased could also be related to an increase in glucagon post-exercise as previous studies have shown that glucagon stimulates the uptake of methionine from plasma to the liver .
In the present study the horses performed an exercise test in the field in order to deplete muscle glycogen before they were given the two different oral treatments in the recovery period. This exercise test caused alterations in some plasma amino acid concentrations and to a similar degree after both tests. The observed changes in BCAA and aromatic amino acids after exercise and in the recovery period during the CON treatment are in good agreement with a previous study in which horses performed intense standardised exercise on a treadmill to deplete glycogen . Similar increases in the concentrations of leucine and isoleucine, but not in valine, were found in Standardbred trotters after performing intense exercise over a distance of 2000 m . The increase in BCAA and aromatic amino acids after exercise might indicate a certain degree of protein degradation . Amino acid concentrations decreased during the post-exercise period in both treatments but as stated above, the changes were most marked for isoleucine, valine and methionine in the GLU+LEU treatment. This further supports the fact that the decreases seen in these amino acids were related to the treatment and not to an effect of exercise.