Volume repletion after exercise-induced volume depletion in humans: replacement of water and sodium losses

Susan M. Shirreffs, Ronald J. Maughan

Abstract

Sodium and water loss during, and replacement after, exercise-induced volume depletion was investigated in six volunteers volume depleted by 1.89 ± 0.17% (SD) of body mass by intermittent exercise in a warm, humid environment. Subjects exercised in a large, open plastic bag, allowing collection of all sweat secreted during exercise. For over 60 min beginning 40 min after the end of exercise, subjects ingested drinks containing 0, 25, 50, or 100 mmol/l sodium (trials 0, 25, 50, and 100) in a volume (ml) equivalent to 150% of the mass lost (g) by volume depletion. Body mass loss and sweat electrolyte (Na+, K+, and Cl) loss were the same on each trial. The measured sweat sodium concentration was 49.2 ± 18.5 mmol/l, and the total loss (63.9 ± 38.7 mmol) was greater than that ingested on trials 0 and25. Urine production over the 6-h recovery period was inversely related to the amount of sodium ingested. Subjects were in whole body negative sodium balance ontrials 0 (−104 ± 48 mmol) and 25 (−65 ± 30 mmol) and essentially in balance on trial 50(−13 ± 29 mmol) but were in positive sodium balance ontrial 100 (75 ± 40 mmol). Only ontrial 100 were subjects in positive fluid balance at the end of the study. There was a large urinary loss of potassium over the recovery period on trial 100, despite a negligible intake during volume repletion. These results confirm the importance of replacement of sodium as well as water for volume repletion after sweat loss. The sodium intake on trial 100 was appropriate for acute fluid balance restoration, but its consequences for potassium levels must be considered to be undesirable in terms of whole body electrolyte homeostasis for anything other than the short term.

  • fluid balance
  • electrolyte balance
  • sweat electrolyte loss
  • dehydration
  • rehydration

the need to replace electrolytes after volume depletion is linked to the loss of electrolytes in sweat; ingestion of large volumes of plain water after exercise-induced volume depletion results in a rapid fall in plasma osmolality and in the plasma sodium concentration (12-14), and both of these effects stimulate urine output. Addition of sodium to ingested fluids maintains the circulating vasopressin levels, and an increase in urine production is avoided. There is, however, no clear indication of the amount of sodium that is necessary to optimize volume repletion after sweat loss: excessive sodium intake will stimulate a natriuresis, which will be accompanied by an increased urine flow. It might seem reasonable to replace the electrolytes in amounts approximately equal to those lost in sweat. The composition of human sweat appears, however, to be highly variable (8), and the electrolyte content as well as the total volume of the sweat loss determines the extent of electrolyte loss.

The importance of the addition of sodium to volume repletion fluids has been systematically evaluated by Maughan and Leiper (9), who volume depleted subjects by the equivalent of 2% of body mass by intermittent exercise in the heat; subjects then ingested one of the test drinks in a volume equal to 150% of their mass loss. The test drinks contained ∼0, 25, 50, or 100 mmol/l sodium. Urine output over the subsequent few hours was inversely proportional to the sodium content of the ingested fluid, and only when the sodium content exceeded 50 mmol/l did subjects remain in positive fluid balance throughout the recovery period. At the end of the study period the mean difference in net fluid balance between the trials with 0 and 100 mmol/l sodium was substantial, amounting to 787 ml. The plasma volume was calculated (2) to have decreased with volume depletion on all trials and increased after volume repletion on all trials; the mean decrease over all trials amounted to ∼4%. The increase in plasma volume after drinking occurred less rapidly with the 0 mmol/l beverage, so that 1.5 h after the end of the fluid ingestion period, the increase in plasma volume was 6.8% compared with increases of 12.4 and 12.0% with the 50 and 100 mmol/l drinks, respectively. There was no difference between trials in the extent of the change in plasma volume 5.5 h after the end of the period of fluid consumption, but there was a tendency for it to be related to the quantity of sodium consumed, with the smallest increase with the 0 mmol/l beverage and the greatest increase with the 100 mmol/l beverage. These observations were confirmed in a further study in which the volume of fluid ingested and the sodium content of drinks administered after exercise-induced hypohydration were varied systematically (17). This study showed that even when a volume equal to twice the sweat loss was ingested, subjects did not remain in a positive fluid balance when a low-sodium (23 mmol/l) drink was consumed; when a drink containing 61 mmol/l sodium was consumed, subjects maintained a positive fluid balance when the volume of fluid ingested was ≥1.5 times the loss.

In none of these previous studies was the sweat composition measured, with the result that electrolyte balance could not be calculated. The present study aimed to investigate the effect on fluid balance of variations in the quantities of sodium consumed with beverages and to relate the effectiveness of volume repletion to the sweat electrolyte loss.

METHODS

Six subjects (4 men and 2 women) volunteered to take part in this investigation. All were more-or-less physically active on a recreational basis, but none were training systematically at the time of the study. All gave their written informed consent to participate, and the investigation was approved by the local Ethics Committee. Their physical characteristics (means ± SD) were as follows: 28 ± 8 yr of age, 168 ± 11 cm height, 69.0 ± 13.8 kg body mass, 53.8 ± 5.9 ml ⋅ kg−1 ⋅ min−1maximum O2 uptake (V˙o 2 max).

Four experimental trials were undertaken by each subject, and a different beverage was ingested on each. Each trial had three distinct phases. The first, the volume depletion phase, consisted of intermittent exercise in the heat, such that a moderate level of hypohydration (∼2% of body mass) was induced. The second, the volume repletion phase, consisted of fluid ingestion over a 1-h period commencing 40 min after the end of exercise. In the third and final phase, the recovery phase, the subject rested in the laboratory for 6 h. Blood and urine samples were collected at intervals throughout the study.

Before starting the experimental trials, each subject underwent three preliminary trials. The first involved measurement ofV˙o 2 max, which served to determine the exercise workload to be used during the volume depletion exercise; this was carried out by means of a discontinuous, incremental cycle ergometer test. On the second preliminary trial, subjects underwent only the volume depletion procedures, as described below, and were then free to leave the laboratory. On the third preliminary trial, subjects underwent the full experimental procedure, but the recovery phase was reduced to only 2 h; blood samples were not collected from those individuals who were accustomed to the procedure and had previously undertaken studies of a similar nature. The second and third preliminary trials were undertaken to familiarize the subjects with all the procedures involved in the study and to determine the duration of exercise that was required to volume deplete subjects by 2% of body mass. In these trials, subjects were weighed every 10 min to monitor their body mass loss; no sweat samples were collected during the preliminary trials.

All experimental trials commenced in the morning after an overnight fast, except for the ingestion of ∼500 ml of plain water 2 h before the scheduled start of the experiment. Trials took place at 7-day intervals, so that, for each subject, trials were conducted on the same day of the week. For the 2 days preceding each trial, subjects followed similar diets and activities to standardize the starting conditions for each trial; they kept a diary on the 2 days before the first trial and reproduced the same pattern of eating and physical activity on the 2 days preceding each of the subsequent three trials.

On arrival at the laboratory, subjects rested in a seated position in a room maintained at ∼24°C. After 15 min of seated rest, a venous blood sample was collected without stasis from an antecubital vein. After the blood sample was collected, subjects were asked to give a urine sample by emptying their bladders as completely as possible and collecting the entire volume.

Subjects showered and washed themselves with soap and water and then rinsed themselves with 4 liters of distilled, deionized water with a specific resistance >18 MΩ ⋅ cm. Subjects were instructed to wash themselves thoroughly with copious amounts of soap in the shower. On leaving the shower, they stood in a stainless steel tray and washed with the deionized water. Subjects then dried themselves carefully with a prepared washed towel and put on a pair of plastic operating theater overshoes to avoid contact with the floor. Nude body mass was measured to the nearest 10 g. Subjects then entered the exercise room.

Volume depletion was induced by intermittent cycle ergometer exercise in the heat. Subjects performed a series of 5-min bouts of exercise at an intensity corresponding to ∼60% ofV˙o 2 max; each 5-min exercise period was followed by a 5-min rest period. Exercise was performed in a climatic chamber maintained at 34°C and 60–70% relative humidity. The cycle ergometer was contained within a large polythene bag that was stretched over a plastic frame. The procedure has been fully described previously (16). The bags used in this study (Anaplast, Ayrshire, UK) were obtained from an agricultural merchant and were intended to contain silage. Preliminary trials showed that the bags were free from contamination with electrolytes. When the subject entered the laboratory, the frame and ergometer were in place, but the plastic bag was pulled up to a height of only ∼600–800 mm. The subject removed the plastic overshoes and stepped into the bag; the bag was then pulled up fully and stretched over the top of the plastic frame. The subject then put on the prepared washed shorts and was ready to begin exercise. Subjects were dressed only in shorts: shoes were not worn. The intention was to volume deplete each subject by 2% of body mass; the duration of exercise required by each subject to achieve this was determined in the preliminary trials and ranged from 30 to 70 min (median 43 min). The total sweat loss in all trials was therefore approximately the same when expressed relative to each subject’s body mass, although sweating rates were very different.

On completion of the prescribed exercise, the subjects removed their shorts and washed themselves with 4 liters of deionized water to which had been added ammonium sulfate (20 mmol/l). This water was presented in four 1-liter sports drink bottles, and no other washing implements were provided. Subjects were instructed to wash themselves as completely as possible and to remove as much surface water as they could before leaving the bag. An additional 1 liter of deionized water containing ammonium sulfate was then used to wash down the inside of the bag and the bicycle, with care being taken to ensure that as much liquid as possible was flushed toward the bottom of the bag.

The liquid in the bag at this stage contained the sweat that had not evaporated plus the water used to wash the subject at the end of the trial. The volume of unevaporated sweat was calculated from the dilution of the added ammonium and sulfate; preliminary studies had failed to detect the presence of ammonium or sulfate in sweat. Duplicate samples of the water in the bag were collected using a syringe and needle and stored for later analysis.

At the end of the volume depletion phase, subjects dried themselves completely before nude body mass was again measured. The difference in mass was used to estimate total sweat loss, with the assumption that the specific gravity of sweat is 1.0 (7). The respiratory water loss (48 ± 10 g) and mass loss due to substrate oxidation (43 ± 10 g) were estimated for each subject in the preliminary trials (11); these were assumed to be the same on each trial and were taken into account in the estimation of the sweat loss.

Subjects showered and dressed before returning to sit in a comfortable environment within 25 min of the end of exercise. After ∼10 min of seated rest, a 21-gauge venous cannula was inserted into a superficial forearm vein and was flushed with isotonic saline. A further 5 min elapsed before the first postexercise blood sample was withdrawn; this was collected, therefore, 40 min after the end of exercise and after at least 15 min in a sitting position. A urine sample was then obtained; for this, as for all samples, subjects were instructed to empty their bladders as completely as possible and to collect the entire volume. The volume repletion period then commenced, and over the following 60 min subjects consumed a drink containing 0, 25, 50, or 100 mmol/l sodium (Table 1) plus 125 ml/l low-energy lemon flavoring; for the three sodium-containing drinks, 25 mmol/l sodium was from sodium chloride and the remainder, where applicable, was from sodium acetate. A different beverage was consumed on each trial, but in each case the volume consumed (in ml) was equivalent to 150% of the mass loss (in g), with the assumption that the entire mass loss was due to water loss. Further blood samples were collected at the end of the volume repletion period and at 1, 2, 4, and 6 h thereafter, and further urine samples were collected at the end of the volume repletion period and at 1, 2, 3, 4, 5, and 6 h thereafter. Subjects remained within the laboratory environment at all times and were seated for at least 15 min before collection of each blood sample to minimize the effects of postural changes on the redistribution of water between the major body water compartments. When blood and urine sample times coincided, the urine sample was collected immediately after blood sample collection.

View this table:
Table 1.

Drink composition on each trial

Analytic methods.

For each blood sample, 8 ml were collected. Part of this (2.5 ml) was added to a K2EDTA tube and was used for measurement of hemoglobin concentration (by the cyanmethemoglobin method) and of packed cell volume (by microcentrifugation). The values so obtained were used to calculate changes in blood, red cell, and plasma volumes using the equations described by Dill and Costill (2); all calculations were made relative to the blood sample collected 40 min after exercise. Part (2.5 ml) of each sample was added to a chilled K2EDTA tube and centrifuged at 1,500 g for 5 min at 4°C before the plasma was removed. This was then used for measurement of plasma vasopressin and aldosterone concentrations by RIA (Euro-Diagnostica, Cornwall, UK; Serono Diagnostics, Rome, Italy). The remainder of the sample was allowed to clot before being centrifuged, and the serum was then removed. The serum was analyzed for sodium and potassium concentration by flame photometry (model 410C clinical flame photometer, Corning, Halstead, Essex, UK) and for chloride concentration by coulometric titration (Cl meter, Jenway, Dunmow, Essex, UK). Serum osmolality was assessed by freezing-point depression (Osmomat 030 cryoscopic osmometer, Gonotec, Berlin, Germany).

For each urine sample the volume was measured and a 5-ml aliquot was retained. Urine sodium, potassium, and chloride concentrations and osmolality were measured by the same methods used for serum analysis.

Sweat sample analysis was carried out by ion chromatography using an ion chromatograph system (model DX-100, Dionex, Sunnyvale, CA). Samples were first diluted 1:100 in filtered, distilled, deionized water, and sodium, potassium, chloride, ammonium, and sulfate were analyzed.

All analyses were carried out in duplicate, and the mean value of the duplicates was taken, except for packed cell volume, which was measured in triplicate, and the serum vasopressin and aldosterone concentrations, which were analyzed singly.

Statistical analysis.

Values are means ± SD or medians and ranges where they were found not to be normally distributed. Statistical comparisons were made after the normality or otherwise of the distribution of the data points was established. Analysis was by repeated-measures ANOVA followed by one-way ANOVA and Tukey’s multiple range test or the Kruskal-Wallis and Mann-Whitney tests where appropriate. Differences between and within trials were considered significant whereP < 0.05.

RESULTS

The subjects’ preexercise body mass was the same on all four trials: 69.05 ± 13.99, 69.00 ± 13.82, 69.20 ± 13.64, and 69.17 ± 13.97 kg for trials 0, 25, 50, and 100, respectively (P = 1.000). Subjects exercised for the predetermined duration, established from the preliminary procedures, on each of their trials, and they achieved the same degree of body mass loss on each trial (P = 0.984; Table 2); over all trials the volume depletion was 1.89 ± 0.17% of body mass. Over all trials the mean sweat sodium concentration was 49.2 ± 18.5 mmol/l and the mean sweat potassium concentration was 5.5 ± 1.4 mmol/l. The quantities of sodium (P = 0.999), potassium (P = 0.979), and chloride (P = 0.939) lost in the sweat did not differ from week to week (Table 2). Because the total volume of sweat loss was the same on each trial, the volume of drink consumed also did not differ between trials. The volume of drink consumed and the electrolyte intake on each trial are shown in Table 2.

View this table:
Table 2.

Sweat electrolyte loss during volume depletion and the volume of drink consumed on each trial together with the electrolyte intake

Urine output and fluid balance.

The volume of urine produced at each collection point over the duration of the study is shown in Fig. 1. On each trial the greatest urine output occurred over the 1-h period ending 1 h after the end of the volume repletion period. There was no statistical difference in the volume produced at this time between the trials (P = 0.457), but there was a tendency for the greatest volume on trial 0 and the least on trial 100. Over the 1-h period ending 2 h after the end of the volume repletion period, the volume of urine produced on trial 100was significantly less than that produced on trial 0 or 25(P = 0.006). One hour later, 3 h after the end of the volume repletion period, the same volume of urine was produced on trials 50 and100; on both trials the volume was less than the volume produced on trials 0 and 25(P = 0.016).

Fig. 1.

Urine output over time. Preexercise sample has not been shown. Points are median values.

Over the entire time after volume depletion (i.e., the volume repletion and recovery periods), the total volume of urine produced was inversely proportional to the quantity of sodium ingested and amounted to 1,182 ± 438, 970 ± 340, 800 ± 353, and 578 ± 280 ml on trials 0, 25, 50, and100, respectively (P = 0.033).

From the volume of drink consumed and the volume of urine produced together with the volume of sweat lost, whole body net fluid balance relative to the starting euhydrated position can be calculated (Fig.2). On losing water, in the form of sweat or urine, subjects move in the direction of negative fluid balance, and on gaining water by drinking, subjects move in the direction of positive fluid balance. When the sodium-free drink was consumed, subjects were significantly volume depleted at the end of the study period (−642 ± 342 ml) and, indeed, had been in a state of negative fluid balance from 2 h after the end of the volume repletion period. The same was true on trials 25and 50, although the extent of whole body negative fluid balance was less on trial 25 (−403 ± 394 ml) than ontrial 0 and less ontrial 50 (−266 ± 366 ml) than on trial 25. After consuming the 100 mmol/l drink the subjects remained in positive fluid balance for the entire recovery period, and at the end of the study period, 6 h after the end of the volume repletion period, subjects were essentially euhydrated (16 ± 322 ml).

Fig. 2.

Whole body net fluid balance calculated from estimated sweat loss, volume of fluid ingested, and urine output over course of experiment. Preexercise urine sample is not included in these calculations. Points are mean values.

Urine composition and electrolyte balance.

Despite large differences in the quantities of sodium ingested on each trial, the quantities excreted in the urine after volume repletion ontrials 0, 25, and50 did not differ from each other (P = 0.272), but there was a tendency (P = 0.093) for a larger excretion of sodium on trial 100 (Table3).

View this table:
Table 3.

Cumulative urinary excretion of Na+, K+, and Cl after volume repletion

On trial 25, only 9% (−60 to 60%) of the ingested sodium had been retained by the end of the study period, 6 h after the end of the volume repletion period; this represents a retention of ∼4 mmol. On trial 50, 53% (2–83%) or ∼53 mmol were retained, and on trial 100, 72% (56–80%) or ∼148 mmol were retained. The fraction retained was significantly less on trial 25 than ontrial 100(P = 0.013).

Whole body net sodium balance was calculated as for water balance, with the sweat sodium losses taken into account (Fig.3). On trials 0 and 25, insufficient sodium was ingested with the drink to replace the losses incurred with volume depletion, and subjects were therefore in sodium deficit throughout the recovery period. On trial 50, sufficient sodium was consumed with the beverage to replace the sodium losses, and with the urinary excretion over the recovery period, subjects were in slight net negative sodium balance (−13 ± 29 mmol) at the end of the study period. Ontrial 100 the amount of sodium consumed with the beverage (206 ± 55 mmol) was far in excess of the quantities lost in the sweat during volume depletion (65 ± 46 mmol), and subjects were in a state of positive sodium balance at the end of the volume repletion period. There was some urinary sodium loss during the recovery period (Table 3), but at the end of the study, 6 h after the end of the volume repletion period, subjects were still in a significant positive sodium balance (75 ± 40 mmol).

Fig. 3.

Whole body net sodium balance calculated from sweat sodium loss, amount of sodium ingested, and urine excretion of sodium over course of experiment. Preexercise urine sodium loss is not included in these calculations. Points are mean values.

The potassium content of the drinks was low (0.6 ± 0.1 mmol/l) relative to that of sweat (5.5 ± 1.4 mmol/l) and was not sufficient to replace the sweat losses (Table 2). However, although this was small and did not differ between trials, the amount of potassium excreted in the urine after volume repletion was greater (P = 0.006) after consumption of the 100 mmol/l drink than on any of the other trials (Table 3); the greatest excretion occurred over the 1-h periods ending (0 h) and 1 and 2 h after the end of the volume repletion period. The intake of chloride in the beverages was the same on trials 25, 50, and 100, but chloride was virtually absent from the sodium-free drink (Table 2). However, even on trials 25, 50, and100, insufficient chloride was consumed to replace the sweat losses. A smaller quantity of chloride tended to be excreted over the 1-h periods ending 3 and 4 h after the end of the volume repletion period (bothP = 0.058), and as a result the cumulative quantity excreted over the entire recovery period tended to be less on trial 100 than on the other trials (Table 3).

Urine osmolality values generally mirrored the urine volume. The osmolality was higher on trial 100than on trial 0 at 2, 3, 4, and 5 h after the end of the volume repletion period, was greater than that ontrial 25 at 2, 3, and 4 h after the end of the volume repletion period, and was also greater than that ontrial 50 at 2 h after the end of the volume repletion period. On each trial the lowest osmolality was recorded 1 h after the end of the volume repletion period, which corresponded with the time of greatest urine volume production (Fig.1).

Blood and serum measurements.

Blood volume was calculated to have decreased with volume depletion on all trials, and over all trials the decline amounted to 3.6 ± 2.5% (Table 4;P = 0.782). With beverage consumption there was a restoration to preexercise levels on all trials, and the increase was generally influenced by the amount of sodium ingested. Plasma volume generally changed in the same manner as the blood volume, but the magnitude of the changes was larger (Fig.4); the decrease with volume depletion amounted to 5.5 ± 4.7% over all trials (P = 0.556). Red cell volume decreased with volume depletion on all trials, and over all the trials the decline amounted to 1.3 ± 1.6% (Table 4). With volume repletion the red cell volume increased on trial 0 (P = 0.000), whereas on trial 100 there was no change from the volume-depleted volume (P = 0.637); these volume changes were significantly different between the trials at the end of the volume repletion period (0 h) and 1 h later.

View this table:
Table 4.

Calculated change in blood and red cell volume

Fig. 4.

Calculated change in plasma volume over duration of study. Calculations were made relative to values obtained 40 min after end of exercise (postexercise). Points are mean values.

On all trials there was a tendency for the serum sodium concentration to increase with volume depletion (Table5). On trial 0 the serum sodium concentration at the end of the volume repletion period (0 h) and at 1 and 2 h later was lower than the preexercise level, but on none of the other trials did the serum sodium concentration fall below the preexercise level. The serum potassium concentration was little affected by any of the treatments (Table 5); on trial 100 the serum potassium concentration 1, 2, 4, and 6 h after the end of the volume repletion period was lower than the preexercise level. The serum chloride concentration tended to increase with volume depletion and was restored to preexercise levels with volume repletion (Table 5). There were no significant differences between trials at any time, nor did the serum chloride concentration change significantly over the course of any trial.

View this table:
Table 5.

Serum Na+, K+, and Cl concentrations over duration of each trial

Serum osmolality was little affected by any of the treatments over the duration of the study on any trial (Table6). Although not significantly different between trials, serum osmolality tended to increase with volume depletion (over all trials the increase amounted to ∼4 mosmol/kg) and to decline to values slightly below the preexercise value on all trials over the recovery period.

View this table:
Table 6.

Serum osmolality over duration of each trial

The plasma aldosterone and vasopressin concentrations did not differ between trials at any time, although both tended to fluctuate over the course of each trial. With volume depletion there tended to be an increase in concentrations from 34.1 (range 24.8–77.8) to 43.4 (range 19.6–187.3) pg/ml for aldosterone and from 9.4 (range 4.0–21.9) to 10.8 (range 4.7–18.9) pg/ml for vasopressin (over all trials), followed by a return to the preexercise levels over the 6-h recovery period. On trials 25, 50, and 100, when sodium was consumed in the beverages, the peak aldosterone concentration was recorded at the end of the volume repletion period, but on trial 0, when no sodium was consumed, the peak plasma aldosterone concentration was recorded 1 h after the end of the volume repletion period.

DISCUSSION

The results of this study confirm that the addition of sodium to fluids ingested after exercise-induced volume depletion has a significant effect on the restoration and maintenance of whole body fluid balance. After volume depletion by 1.89% of body mass and fluid ingestion equivalent to 150% of this volume, subjects retained a mean of 71% of the ingested volume when the beverage had a high (102 mmol/l) sodium concentration but a mean of only 37% when virtually no sodium was present (1 mmol/l).

In the present study the volume of drink consumed was 150% of the water loss with volume depletion; this ensured a positive fluid balance at the end of the volume repletion period (0 h). However, on only two of the trials in this study was sufficient sodium ingested to replace the losses in the sweat, despite the relatively large volume of fluid consumed. In previous studies of a similar nature (9, 10), water loss could be reasonably reliably quantified, but electrolyte losses could only be estimated. The development of a simple and reliable method for whole body sweat collection (16) made possible the measurements in the present study and allowed calculation of electrolyte as well as water balance. The sweat had a mean sodium concentration of 49.2 ± 18.5 mmol/l, and the sweat sodium loss amounted to 63.9 ± 38.7 mmol over all trials (Table 1).

On trial 50 the quantity of sodium ingested was >50% as much as the sweat loss (99 mmol intake vs. 64 mmol loss; Table 2), and despite the urinary losses of sodium over the 6 h recovery period, subjects remained in the region of whole body net sodium balance for the 6 h of the recovery period (Fig. 3). However, subjects were in a state of whole body fluid deficit for the majority of this time (Fig. 2). On trial 100the sodium intake was ∼300% of the sweat loss (206 mmol intake vs. 65 mmol loss; Table 2), and even with the tendency for a greater urinary excretion of sodium during the 6-h recovery period on this trial relative to the other trials (Table 3), subjects were in a state of positive sodium balance for the entire 6-h recovery period (Fig. 3). Over this time, subjects also remained in a state of positive fluid balance, and at the end of the 6-h recovery period they were essentially euhydrated relative to their preexercise condition (Fig.2). The relationship between whole body net fluid balance and whole body net sodium balance for the last 2 h of the recovery period is shown in Fig. 5; there is a close relationship between the whole body sodium balance and the whole body fluid balance (correlation coefficient = 0.521).

Fig. 5.

Whole body net sodium balance (□, ▪) and whole body net cation balance (⋄, ◆) vs. whole body net fluid balance 5 (□, ⋄) and 6 (▪, ◆) h after end of volume repletion period. For net sodium balance, y = 59.75 + 0.28 x;R 2 = 0.521. For net cation balance, y= −1.68 + 0.23x; R 2 = 0.577. Points are mean values. Fitted lines are calculated from 6-h values only.

It is generally agreed that, provided a sufficient volume of water is consumed, sodium replacement is the most important factor in achieving effective restoration of fluid balance (4, 9, 14). The requirement for sodium replacement stems from its role as the major ion in the extracellular fluid and the major electrolyte lost in sweat. If sufficient sodium and water are ingested, some of the sodium remains in the vascular space, with the result that plasma osmolality and sodium concentration do not markedly decline as may occur if plain water is ingested (13). As a result, the plasma levels of vasopressin and aldosterone are maintained, and a diuresis that would result in creating a whole body net negative fluid balance is prevented. There was little difference between trials in the plasma concentrations of vasopressin and aldosterone at any time in the present study; the same was true of serum osmolality (Table 6), although serum sodium concentration (Table 5) did tend to decline on trial 0, when the sodium-free drink was consumed. The effects on urinary output of any changes in plasma sodium concentration and osmolality will, however, be delayed slightly because of the time to inactivate circulating vasopressin and aldosterone or stimulate their release; vasopressin has a half-life in plasma of ∼10–20 min (1,3), and aldosterone acts via the synthesis of new proteins, including Na+-K+-ATPase, so a time lag of ∼30–90 min occurs before its effects take place (6). The set point of the osmoregulatory system can be altered by a change in blood volume or pressure, and this will therefore influence vasopressin secretion (15). Differences in blood volume were observed between treatments in the present study, such that the general pattern was for the increases in plasma volume to be a direct function of the quantity of sodium ingested (Fig. 4); this would, however, cause changes in the opposite direction to the effect seen. Therefore, it seems most likely that the renal response that resulted in each trial was appropriate, in terms of water and sodium excretion, thus maintaining the relatively constant vasopressin, aldosterone, and sodium concentrations.

The measurement of water, sodium, potassium, and chloride losses in sweat and urine and intake via the drinks allowed calculations to be made of the whole body water and electrolyte (Na+, K+, and Cl) balance over the duration of the study. In the same way as an excess of water must be ingested to have a chance of restoring fluid balance, an electrolyte intake in excess of the losses must also occur to compensate for ongoing urinary losses. In terms of the cations investigated (Na+ and K+), on trial 100, when the large amount of sodium was ingested, a significant fraction of it was retained [72% (range 56–80%)], although it was not readily apparent that this excess was being retained in the extracellular space; there was no increase in serum sodium concentration (Table 5), but the greater expansion of plasma volume on this trial would accommodate some of this sodium. Because of the rapid exchange of water between the plasma and the other fractions of the extracellular fluid, it is reasonable to assume that the changes in the plasma volume will reflect those of the extracellular fluid as a whole. From the height and body mass of the subjects in the present study, their plasma volume was estimated to be 2,309 ml (7) and their extracellular fluid volume as a whole to be 9,236 ml (5). Relative to the preexercise situation, the plasma volume expansion was by ∼8.2% on trial 100. This represents an increase in plasma volume of 189 ml and in extracellular fluid volume of 757 ml; the plasma volume is approximately one-fourth of the extracellular volume (5). Given the minimal difference in plasma sodium concentration ontrial 100 between the start and end of the study (Table 5), the increased volume could accommodate an extra 117 mmol of sodium. At the end of the study, subjects were in positive sodium balance by 75 ± 40 mmol (Fig. 3), and this all could be accommodated in the increased fluid volume.

However, even though the losses of potassium in sweat are small relative to those of sodium (7.1 ± 3.0 vs. 63.9 ± 38.7 mmol over all trials; Table 2) and virtually no potassium was consumed with the beverage on any trial (1.1 ± 0.4 mmol over all trials; Table2), on trial 100 a large amount of potassium was excreted over the 6-h recovery period relative to the other trials. This electrolyte excretion (high K+, low Cl) is most likely to be due to a metabolic alkalosis as a result of the metabolism of the acetate. Bicarbonate is likely to have been secreted in greater amounts on this trial than the others, but this was not measured in our study. This has the effect of giving the whole body net cation balance status shown in Fig. 5. However, in terms of potassium balance, subjects were in deficit throughout the recovery period due mainly not to the loss of potassium in sweat during exercise but, rather, to the loss in urine during the recovery period. Calculations suggest that only a small fraction of this (∼0.5 mmol) came from the extracellular space and that the majority, as expected, came from the intracellular fluid.

The quantification of sweat electrolyte losses during exercise and replacement in the postexercise recovery period has allowed calculation of whole body water and electrolyte balance. In the present study this has enabled the monitoring of the recovery of water and sodium balance. The 102 mmol/l sodium drink in this study appears to maximize the acute restoration of fluid balance, but its consequences on potassium levels must be considered undesirable in terms of whole body electrolyte balance for anything other than the short-term situation.

Footnotes

  • Address reprint requests to R. J. Maughan.

REFERENCES

View Abstract