Discussion

The results of the present experiment demonstrate that after exercise in a hot environment, resulting in a 1·83 (sd 0·10) % reduction in body mass, the net fluid balance at the end of the study period was greater after the ingestion of a 20 g/l milk protein, 40 g/l carbohydrate solution ( − 181 (sd 280) ml) or a 40 g/l milk protein, 20 g/l carbohydrate solution ( − 107 (sd 126) ml) compared with after ingestion of a 60 g/l carbohydrate solution ( − 470 (sd 154) ml). However, although increasing the milk protein content from 20 to 40 g/l resulted in a slightly greater net fluid balance at the end of the trial period, there was no significant difference between the two protein-containing solutions for any of the measured variables. While there was a statistically significant difference in fluid balance between the C and the CP20 and CP40 trials, this equates to a difference in body mass of approximately 0·3–0·5 %, which is unlikely to confer any enhancement in exercise performance during trial CP20 or CP40 compared with trial C⁽ Reference Watson, Love and Maughan ^{13 )}.

It has previously been shown that after exercise in the heat, resulting in a reduction in body water equivalent to approximately 2 % of subjects' initial body mass, low-fat milk is better retained than a commercially available carbohydrate–electrolyte drink⁽ Reference Shirreffs, Watson and Maughan ^{12 ,} Reference Watson, Love and Maughan ^{13 )}. While there are numerous compositional differences between low-fat milk and a carbohydrate–electrolyte sports drink (energy density, carbohydrate content and type, protein content and electrolyte content), it has recently been demonstrated that the protein contained in low-fat milk is responsible for at least some of the increased fluid retention observed⁽ Reference James, Clayton and Evans ^{8 )}. The study of James et al. ⁽ Reference James, Clayton and Evans ^{8 )} matched solutions in terms of energy density and electrolyte content, making it possible to directly compare the fluid-retaining properties of milk protein (25 g/l) with an isoenergetic amount of carbohydrate. The results of the present study agree with those of James et al. ⁽ Reference James, Clayton and Evans ^{8 )}, and demonstrate that the retention of a rehydration solution consumed after exercise-induced dehydration is increased when some of the carbohydrate in a carbohydrate–electrolyte solution is replaced with an isoenergetic amount of milk protein. The present study demonstrates that the inclusion of some (20–40 g/l) milk protein in a rehydration solution can augment rehydration to a greater extent than an isoenergetic amount of carbohydrate; however, there was no significant difference in any of the measured variables between the rehydration solutions containing either 20 or 40 g/l milk protein. This suggests that in the concentrations of milk protein measured in the present study, there does not appear to be a dose–response relationship with fluid balance after exercise-induced dehydration.

Low-fat milk has a protein content of approximately 36 g/l⁽ Reference Shirreffs, Watson and Maughan ^{12 ,} Reference Watson, Love and Maughan ^{13 )}; however, the present study, as well as previous research⁽ Reference James, Clayton and Evans ^{8 )}, suggests that any increases in drink retention caused by the protein content of low-fat milk can be produced by protein concentrations below that found in low-fat milk. Similarly, Seifert et al. ⁽ Reference Seifert, Harmon and DeClercq ^{14 )} observed an increased retention of a commercially available carbohydrate–protein solution (containing 60 g/l carbohydrate, 15 g/l whey protein) compared with a commercially available carbohydrate solution (60 g/l carbohydrate) and flavoured water. Although, James et al. ⁽ Reference James, Gingell and Evans ^{15 )} demonstrated that this increase in drink retention observed by Seifert et al. ⁽ Reference Seifert, Harmon and DeClercq ^{14 )} might be related to the mismatched energy density of rehydration solutions, as when matched for energy density, no difference in drink retention was observed between a 65 g/l carbohydrate solution and a 50 g/l carbohydrate, 15 g/l whey protein isolate solution⁽ Reference James, Gingell and Evans ^{15 )}. Increasing the energy density of a solution reduces the rate at which it empties from the stomach⁽ Reference Vist and Maughan ^{16 –} Reference Evans, Shirreffs and Maughan ^{18 )}, thus reducing the rate of water uptake into the circulation and offsetting the decline in serum osmolality that occurs with the ingestion of less energy-dense solutions⁽ Reference Evans, Shirreffs and Maughan ^{7 ,} Reference Evans, Shirreffs and Maughan ^{18 )}.

While the results of the present study and those of James et al. ⁽ Reference James, Clayton and Evans ^{8 )} suggest that the inclusion of milk protein in a rehydration solution consumed after exercise-induced dehydration increases the fraction of the ingested solution that is retained, the mechanism by which this effect occurs is currently unknown. The rate at which a solution is emptied from the stomach and the rate at which fluid is absorbed in the small intestine are both affected by the addition of protein to a solution, suggesting that alterations in fluid absorption are likely to be the main reason for the success of these solutions. Careful consideration needs to be given to the interpretation of these data in the hours immediately after drink ingestion, as fluid in the stomach and intestines will not be contributing to fluid balance. Future studies should seek to quantify the volume of fluid contained in the gastrointestinal system to take account of this volume in any fluid balance calculations.

The most important factor affecting rehydration after exercise is the volume of fluid that is consumed, and it has been shown that for effective rehydration, a volume equivalent to approximately 150 % of fluid lost must be ingested⁽ Reference Shirreffs, Taylor and Leiper ^{3 )}. Failure to ingest sufficient fluid will make a return to euhydration impossible. In situations where drinks are consumed ad libitum, palatability and drink perception play an important role in determining the volume of fluid that is ingested. James et al. ⁽ Reference James, Clayton and Evans ^{8 )} observed no difference in drink pleasantness or sweetness between a 65 g/l carbohydrate solution and a 40 g/l carbohydrate, 25 g/l milk protein solution. Interestingly, in the present study, subjects perceived drink C to be more pleasant than drink CP20, and there was a tendency for subjects to perceive drink CP40 as more pleasant than drink CP20 (P= 0·066), with no difference between the C and CP40 drinks. This suggests that if ad libitum fluid intake had been used during rehydration instead of the prescribed drinking that was imposed, then subjects might have ingested a greater amount of drink C or CP40 than drink CP20, which is an important consideration for rehydration in an applied setting.

In the present study, a prescribed volume of fluid of 150 % of body mass loss during exercise was ingested during rehydration. Therefore, although not measured, the observed differences in drink retention must be related to the drinks' effects on gastric emptying, intestinal absorption or fluid retention, all of which might influence serum osmolality and circulating agrinine vasopressin concentrations and thus urine production. Future studies should aim to also quantify these outcomes. Slowing the rate of gastric emptying and/or intestinal absorption might reduce the rate of influx of water into the circulation, offset the decline in serum osmolality that occurs with the ingestion of a large volume of dilute fluid and reduce urine production⁽ Reference Evans, Shirreffs and Maughan ^{18 )}. In contrast, adding ingredients to rehydration drinks that exert an osmotic effect in the extracellular space (e.g. Na) might result in an increase in serum osmolality and a reduction in urine production⁽ Reference Shirreffs and Maughan ^{9 ,} Reference Nose, Mack and Shi ^{19 )}. There is a rationale for milk protein, affecting all three of these components of the rehydration process.

Milk protein is comprised of approximately 80 % casein protein and approximately 20 % whey protein, and in the presence of gastric acid, the casein fraction of milk protein clots⁽ Reference Billeaud, Guillet and Sandler ^{20 )}, which might affect the rate at which a casein or milk protein-containing solution empties from the stomach. It has previously been demonstrated that solutions containing a high fraction of intact casein protein empty from the stomach (measured using a radioactive tracer) slower than solutions containing a low fraction of casein protein or containing casein protein that had been denatured by acidification⁽ Reference Billeaud, Guillet and Sandler ^{20 )}. A reduction in the gastric emptying rate of solutions containing casein or milk protein has also been demonstrated in comparison with glucose and/or lactose⁽ Reference Burn-Murdoch, Fisher and Hunt ^{21 ,} Reference Bowen, Noakes and Trenerry ^{22 )}. In contrast, other studies⁽ Reference Calbert and MacLean ^{17 ,} Reference Calbet and Holst ^{23 )} have demonstrated no effect of milk protein or casein on gastric emptying, although these studies have used a modification of the double-sampling gastric aspiration technique of George⁽ Reference George ^{24 )} as described by Beckers et al. ⁽ Reference Beckers, Rehrer and Brouns ^{25 )}, which involves repeated (at 5–20 min intervals) mixing of the stomach, via an orogastric/nasogastric tube. This repeated mixing is likely to disrupt any clot formation that might affect the rate of gastric emptying.

Calbet & Holst⁽ Reference Calbet and Holst ^{23 )} added ³H to solutions ingested and reported that ³H appearance in the peripheral blood was reduced following the ingestion of a casein protein-containing solution compared with an isoenergetic whey protein-containing solution. As gastric emptying rate of the solutions was not different, this reduced ³H appearance indicates that the rate of intestinal water absorption was reduced in the casein protein-containing solution. As with high-concentration glucose solutions⁽ Reference Evans, Shirreffs and Maughan ^{18 )}, the potential slower delivery of water to the peripheral circulation with a solution containing casein protein might offset the reduction in serum osmolality that occurs with the ingestion of a large volume of a dilute solution⁽ Reference Nose, Mack and Shi ^{19 )}.

Finally, the inclusion of milk protein in a rehydration solution will result in an increase in plasma amino acid concentration⁽ Reference Hall, Millward and Long ^{26 )}, exerting an osmotic effect (oncotic pressure)⁽ Reference Okazaki, Hayase and Ichinose ^{27 )} that would be expected to increase water retention. In line with this, Watson et al. ⁽ Reference Watson, Love and Maughan ^{13 )} reported greater serum osmolality after the ingestion of low-fat milk compared with a 60 g/l carbohydrate solution ingested after exercise-induced dehydration, although this might be accounted for by the greater Na and K contents of low-fat milk. Similarly, Seifert et al. ⁽ Reference Seifert, Harmon and DeClercq ^{14 )} reported greater plasma osmolality after the ingestion of a 60 g/l carbohydrate, 15 g/l whey protein solution compared with a 60 g/l carbohydrate solution after exercise-induced dehydration. Taken together, these results have suggested that the inclusion of milk-derived proteins in a post-exercise rehydration solution might increase plasma osmolality, via an increase in plasma amino acid concentration.

Similar to previous research⁽ Reference James, Clayton and Evans ^{8 ,} Reference Shirreffs, Watson and Maughan ^{12 )}, the osmolality of urine produced after drinking in the present study (2 and 3 h) was greater after the ingestion of milk protein-containing solutions (CP20 and CP40) compared with a carbohydrate–electrolyte solution (C). A change in plasma arginine vasopressin concentration of 1 pg/ml (resulting from a change in plasma osmolality of approximately 3 mosmol/kg) has been shown to lead to a change in urine osmolality of 250 mosmol/kg⁽ Reference Robertson ^{28 )}, and although serum osmolality was not measured in the present study, the increase in urine osmolality observed after drinking during the CP20 and CP40 trials suggests that serum osmolality and circulating arginine vasopressin concentrations would be expected to be greater after ingestion of the protein-containing drinks.

In conclusion, the results of the present study confirm those of James et al. ⁽ Reference James, Clayton and Evans ^{8 )} and suggest that after exercise in a hot environment, the substitution of some of the carbohydrate contained in a carbohydrate–electrolyte rehydration solution with an isoenergetic amount of milk protein can increase the amount of the rehydration solution that is retained. Furthermore, these results demonstrate that there does not appear to be any additional increase in rehydration of increasing the milk protein concentration from 20 to 40 g/l.