Review

Role of Sodium in Fluid Homeostasis with Exercise

Rick L. Sharp, PhD
Exercise Physiology Laboratory, Department of Health & Human Performance, Iowa State University, Ames, Iowa
Key words: sodium, fluid, hyponatremia, exercise, physical activity, heat exhaustion
This paper provides a review of recent literature concerning the interactive effects of sodium and fluid
ingestion in maintaining fluid homeostasis during and following exposure to heat and exercise. Heavy sweating
during exercise combined with heat exposure commonly produces fluid deficits corresponding to 1 8% loss in
body mass. Thus, a great deal of attention has been focused on developing fluid replacement guidelines and
products for active people. Recently, there have been reports of more frequent cases of hyponatremia among
individuals who tend to over-ingest water during exercise lasting more than four hours, and inclusion of sodium
chloride in the fluid replacement beverage is often suggested as a potential means of reducing risk of
hyponatremia. Although hyponatremia is not likely to be a major risk factor for the general population,
ultra-endurance athletes and people with occupational physical activity and heat exposure may benefit from these
recommendations. Replacement of fluid deficits after exercise and heat exposure is another area that has received
considerable attention. Studies in this area suggest that if water is consumed, the volume ingested needs to
exceed the fluid deficit by approximately 150% to compensate for the urinary losses that will occur with water
ingestion. Inclusion of sodium chloride and other solutes in the rehydration beverage reduces urinary water loss,
leading to more rapid recovery of the fluid balance. Data are presented in this paper that suggest a quantifiable
interactive relationship between sodium content and fluid volume in promoting rapid recovery of fluid balance
after exercise and thermal-induced dehydration.

INTRODUCTION

replacement were developed and shared with the medical community, race organizers, and to the general public. Specialty
beverages were developed by food companies to provide fluid,
carbohydrate, and electrolyte replacement and were designed to
be used before, during and after exercise to help meet the
elevated demands for these nutrients in the exercising public.
The composition of sports beverages was adjusted over the next
30 years in response to both research findings and taste preferences. It is the purpose of this paper to review the recent
scientific literature concerning sodium balance and its relationship to hydration both during and following exercise, particularly when performed under environmental heat stress.

In the 1960s it was not uncommon to find salt tablet

dispensers in locker rooms at various sports venues. This was
because of the widespread belief that excessive losses of sodium in sweat during physical activity could lead to a depletion
of sodium and result in heat-cramps. Subsequent research,
however, showed that sweat is hypotonic and that the sodium
concentration is lower than plasma. This finding led to the
realization that the nutrient lost in greatest abundance during
exercise in the heat is water rather than sodium. Further research confirmed this finding by showing that during exercise
in hot and humid conditions causes an increase in plasma
sodium concentration [1], implying that water replacement may
be more important than sodium replacement during exertional
heat stress.
With the popularity of running in the 1970s, it became
apparent that heat illness was a major risk for those individuals
running in hot and humid environments. Guidelines for fluid

WATER AND SODIUM LOSSES

DURING EXERCISE
Sweat production during exercise in the heat depends on
exercise intensity, duration, clothing, hydration status of the

Address reprint requests to: Rick L. Sharp, Ph.D., 250 Forker Building, Department of Health & Human Performance, Iowa State University, Ames, IA 50011. E-mail:
rlsharp@iastate.edu

Journal of the American College of Nutrition, Vol. 25, No. 3, 231S239S (2006)
Published by the American College of Nutrition
231S

Sodium in Fluid Homeostasis with Exercise

individual, heat-acclimation of the individual, and environmental conditions [25]. When performing physical activity in high
environmental temperature, evaporation of sweat from exposed
skin is the predominant mechanism for heat loss. If heat loss is
not matched to the rate of metabolic heat production (intensity
of exercise), body heat storage rises and core temperature can
quickly reach dangerous levels. Maintaining a high capacity for
sweat production is therefore critical in thermoregulation and
prevention of heat illness. During high intensity athletic events,
sweat rates up to 3 L/hr are possible under hot and humid
conditions [6,7]. This leads to a loss of body water or dehydration equivalent to 1 8% of body mass. Coupled with sweat
sodium concentrations ranging on average between 40 60
mEq/L [6 9], such sweat rates can lead to sodium depletion
rates of about 150 mmol/hr with additional sodium losses in
urine production.
A recent study by Mao et al. measured sweat electrolyte and
urinary electrolyte concentration and excretion in 13 adolescent
(16 18 yr) soccer players during 1-hour soccer practices conducted in the heat (3237C, 30 50% relative humidity) on eight
days [10]. Mean sodium concentration in sweat was 55 mmol/L.
Average sweat loss during the 1-hour practice sessions was 1.54 L
(SD ! 2.06 L). Calculated sweat loss of sodium averaged
82 mmol (SD ! 62 mmol). Urinary loss of sodium averaged 110
mmol (SD ! 36 mmol). Thus average sodium excretion accounted for by sweat and urinary excretion was 192 mmol (Table
1). Because no dietary intake data were reported for these subjects,
sodium and fluid balance could not be calculated. Likewise, no
data were obtained to assess either performance or physiological
consequences of these fluid and electrolyte losses. Nonetheless,
these observations suggest large losses of both sodium and water
during exercise in the heat.
It is possible that the sweat collection method used by Mao
et al. overestimated the whole body sodium losses in sweat due
to regional variations in sodium concentration of sweat [11,12].
In the Mao et al. study, sweat was collected from the backs and
chests of the subjects for 5 min during the exercise sessions.
The measured sodium concentration of 55 mmol/L is similar to
the Na concentration of sweat collected by Shirreffs using a
whole-body washdown method [12]. Shirreffs et al. measured
sweat sodium concentrations of 51.6 mmol/L during exercise
producing a 2% dehydration of subjects. It is therefore unlikely
that the data obtained by Mao et al. are grossly overestimated.
In a study by Sanders et al. [13], water and sodium losses
were measured during 4 hr cycling exercise at 20 C at exercise
intensity equivalent to 55% of peak VO2. During the exercise

subjects ingested 3.85 L of an 8% carbohydrate-electrolyte

drink containing 5, 50, or 100 mmol/L of sodium. Sweat losses
averaged between 3.7 and 3.9 L for each of the trials. Sodium
concentration of sweat ranged from 43 48 mmol/L, producing
a sweat sodium loss between 150 and 190 mmol over the 4 hr
of exercise. Combined with the urinary sodium losses, subjects
experienced a negative sodium balance of 198 mmol when
ingesting the 5 mmol/L Na beverage, 36 mmol when ingesting
50 mmol/L Na beverage, and experienced a positive sodium
balance of 159 mmol when ingesting the beverage containing
100 mmol/L sodium (Fig. 1). In addition to assuring a positive
sodium balance throughout exercise, ingestion of the beverage
containing 100 mmol/L sodium reduced total fluid lost during
exercise in comparison to the other beverages. Calculation of
water compartment changes revealed a significant loss of fluid
from ECF ("1.1 L) in the 5 mmol/L sodium trial, no change in
ECF in the 50 mmol/L sodium trial, and expansion of ECF
volume (#0.5 L) in the 100 mmol/L sodium trial. Despite the
better maintenance of hydration status in the 50 and 100
mmol/L sodium trials, cardiovascular responses (e.g. heart rate
response) was similar among the three trials.

HYPONATREMIA
During the last 20 years, persons engaged in long duration
endurance exercise in the heat have been advised to drink as
much fluid as possible during the exercise to prevent dehydration, preserve the sweating response and thereby maintain
thermoregulatory capacity [14]. Unfortunately, this advice has
led to an increase or at least a recognition of hyponatremia in
many athletes competing in these events [1519]. Hyponatremia may result because of excessive loss of sodium due to a
heavy sweating response, or alternatively, due to a dilution of
plasma sodium as a consequence of overzealous hydration [16].
Various recommendations for preventing hyponatremia are
made in the literature and include reducing the emphasis on
fluid ingestion [20] and/or increasing sodium content of beverages ingested during exercise [2124].

Prevalence of Hyponatremia
Several authors have described cases of hyponatremia during endurance exercise in the heat. Speedy et al. have published
the largest field-based study of the occurrence of hyponatremia

Table 1. Body Fluid and Sodium Losses during 1-Hour Soccer Practices among Adolescent Boys

Mean
SE

Body Mass
(kg)

Fluid Loss
(L)

Sweat [Na#]
(mmol/L)

Sweat Na# Loss

(mmol)

Urinary Na# Loss

(mmol)

Total Na# Loss

(mmol)

62.5
6.8

1.54
0.57

55
27

82
62

110
36

192

Data were derived from Mao et al. [10].

232S

VOL. 25, NO. 3

Sodium in Fluid Homeostasis with Exercise

Prevention of Hyponatremia

Fig. 1. Sodium balance at the end of 4-hr cycling exercise in 20C

(dry-bulb) environment. Trials were repeated with ingestion of 3.85 L
of an 8% carbohydrate-electrolyte beverage with either 5, 50, or 100
mmol/L sodium concentration. Adapted from Sanders et al. [13].

[18]. In this study, 330 finishers of a triathlon competition (6 9

hr) were studied. Based on plasma sodium concentration less
than 135 mmol/L, 58 (18%) of the finishers were hyponatremic. Eleven of these subjects were described as severely hyponatremic ($ 130 mmol/L) and seven of these were symptomatic. The authors also noted that those subjects with the
most severe cases of hyponatremia had less change in body
weight during the race, implying that fluid overload was the
cause of the hyponatremia in most of the cases.
Other authors suggest that hyponatremia may only be a
significant risk factor in extraordinarily long duration physical
activity such as marathon running and triathlon lasting 4 hours
or more. Noakes et al. [20] point out that most cases of
hyponatremia are observed in the less well trained participants
who take considerably longer to finish the race than do the top
finishers. The longer duration of exercise coupled with greater
total fluid intake as a result of the longer duration, therefore
puts these persons at greater risk of developing hyponatremia.
Because the cases of exercise-induced hyponatremia are
mostly confined to extraordinary physical efforts lasting longer
than 4 hr, hyponatremia is not likely to be particularly widespread among the general population who engage in exercise
lasting less than 2 hrs per day. Various mechanisms have been
proposed to explain the development of hyponatremia in some
individuals. These causes include fluid overload or dilution
effect [17], excessive sodium loss during the exercise [21], and
inappropriate response of arginine-vasopressin leading to excessive retention of ingested fluids [25]. Findings of greater
prevalence of hyponatremia among women suggests either a
biological sex effect on fluid homeostasis or behavioral differences between men and women that may lead women to be
more compliant with advice to drink as much fluid as possible
during endurance exercise [27].

JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION

If fluid overload is an important contributor to the development of hyponatremia, one would expect plasma sodium concentration to fall during exercise in proportion to the volume of
low- or no-sodium fluid ingested. Vrijens and Rehrer [24] have
examined this question by recruiting 10 male subjects to exercise for 3 hr in an environmental chamber kept at 34C. The
subjects performed this exercise on two separate days; once
while ingesting sodium-free water every 15 minutes to match
fluid loss, and once while ingesting a commercial sodiumcontaining (18 mmol/L Na#, 63 g/L carbohydrate, 3 mmol/L
potassium) beverage to match fluid loss. During the water
ingestion trial, average plasma sodium concentration declined
from 140 mmol/L before exercise to 134 mmol/L by the end of
exercise (Fig. 2). In the carbohydrate-electrolyte trial, plasma
sodium concentration did not decrease significantly (140
mmol/L before exercise, 138 mmol/L at end of exercise). The
authors conclude that hyponatremia is possible even when fluid
intake matches fluid loss during long duration exercise when
sodium is not included in the fluid replacement beverage.
Other authors have also recommended inclusion of sodium
in beverages consumed during exercise [7,22,23,26]. Gisolfi
[26] recommended that persons exercising for 13 hr should
consume between 800 1600 ml/hr of fluid containing 10 20
mmol/L sodium and that persons exercising for more than 3 hr
should consume 500 1000 ml/hr of fluid containing 20 30
mmol/L sodium. Lutkemeier et al. [22] suggested that saline
ingestion before exercise can help preserve the plasma volume
and may lead to beneficial changes in endurance exercise
performance. In a review article published by Rehrer [7] inclusion of sodium in a fluid replacement beverage at concentration
ranging between 30 and 50 mmol/L was suggested as possibly
beneficial to those engaged in long duration exercise (3 hr or
more) in the heat.
Consistent with the hypothesis that excessive sodium loss is

Fig. 2. Plasma sodium concentration before and after 3-hr exercise in

34C (dry bulb) environment with ingestion of either plain water to
match fluid loss or a commercial carbohydrate-electrolyte beverage to
match fluid loss. Adapted from Vrijens and Rehrer [24].

233S

Sodium in Fluid Homeostasis with Exercise

the primary cause of exercise-induced hyponatremia, Hiller et
al. [21] suggested 12 g sodium ingestion per hour of exercise
to prevent hyponatremia. Assuming fluid ingestion of 1 liter per
hour to match fluid lost through sweating, this amount of
sodium requires a beverage containing 43 87 mmol/L sodium.
This recommendation is slightly higher than that recommended
by Rehrer and represents a sodium concentration roughly 2 4
times as high as that found currently in most commercial fluid
replacement beverages. Barr et al. argues that the reduced
palatability of such beverages would likely lead to less fluid
consumption among the general population and result in a
greater risk of dehydration [28].
There are also several studies that provide evidence that
sodium supplementation during exercise along with fluid replacement is not necessary [28 32]. Barr et al. had 8 subjects
perform 6 hr exercise at 55% VO2max in a heat chamber held
at 30C [28]. Each subject completed this exercise on separate
occasions to evaluate the possible effects of water ingestion,
water plus sodium (25 mmol/L), or no fluid. When the subjects
were not provided with fluid during the exercise, core temperature and heart rate rose rapidly while plasma volume declined
throughout exercise. Under this condition, only one subject was
able to complete the full 6 hr exercise and the mean time of
exercise was 4.5 hr. The subjects who failed to complete the
exercise did so because heart rate exceeded 95% maximum
heart rate (n ! 1), core temperature exceeded 40C (n ! 1), or
volitional exhaustion (n ! 5). In the water and saline trials,
seven of the eight subjects completed the 6 hr of exercise.
There were no differences in either heart rate or core temperature response between water and saline ingestion and both
trials resulted in smaller rise in these variables than was observed when no fluid was ingested. Plasma volume dropped
less when ingesting the saline beverage than when ingesting
water. Plasma sodium concentration decreased by small
amount in both the saline (change ! "3.0 mmol/L) and water
(change ! "3.9 mmol/L) trials but there were no significant
differences in plasma sodium concentration between these trials. Calculation of overall sodium balance revealed a sodium
deficit in the water trial ("207 mmol) that was significantly
larger than observed in the saline trial ("91.3 mmol). Based on
these results, the authors concluded that sodium concentration
equivalent to that found in commercial sports drinks do not
prevent the fall in plasma sodium during exercise when fluid
intake matches fluid lost through sweating. They further suggest that sodium replacement is not necessary in exercise
lasting less than 6 hr.
Based on these reviewed studies, it is apparent that inclusion
of sodium in fluid replacement beverages can offset some of
the losses of sodium that occur during prolonged and heavy
sweating. It is less clear that doing so will prevent hyponatremia or that this improves either exercise performance or thermoregulation. As suggested by Sanders et al., however, sodium
ingestion likely preserves the plasma volume during exercise at
the expense of the intracellular fluid volume. What effect this

234S

relative dehydration has on muscle metabolism and function

has not yet been studied. An additional finding common to
most of these studies is that even if sodium ingestion does not
affect plasma sodium concentration, it does reduce the sodium
deficit that occurs during prolonged exercise in the heat. This
may be significant for people who are involved in daily exercise or occupations that involve prolonged physical activity in
hot, humid environments.

ROLE OF SODIUM IN
REHYDRATION AFTER EXERCISE
Despite efforts to replace fluid losses during exercise, mild
dehydration after exercise remains a common finding. Dehydration equivalent to less than 2% loss of body mass is associated with reduced performance and impaired thermoregulation during subsequent exercise if the fluid deficit is not
corrected. Thus, considerable research has been devoted to
understanding the rehydration process and the role played by
sodium in restoring body fluids lost during prior exercise.
In studying rehydration after exercise-induced body water
loss, investigators have employed three models for rehydration:
allow subjects to drink fluids ad lib during the rehydration
period [3335], prescribe fluid intake during the rehydration
period to match the fluid lost during the prior exercise [36 38],
and prescribe fluid intake in excess of the fluid lost in the prior
exercise [39 43]. The advantage of allowing ad lib rehydration
is that factors regulating thirst can be studied while the advantage of prescribing fluid intake equal to fluid lost restores
plasma volume while total body water remains somewhat contracted. The rationale for the approach that involves prescribing
fluid intake in excess of that lost in the prior exercise is that
both plasma volume and total body water are restored by the
end of the rehydration period. Finally, there are also hybrid
models in which varied amounts of fluid and sodium content
are studied to allow for evaluation of independent effects of
sodium and fluid volume on the rehydration process.

Ad Libitum Rehydration
Nose et al. dehydrated six subjects by 2.3% using thermal
and exercise induced dehydration [34]. Over the next 3 hr,
subjects were seated in a thermoneutral environment and allowed to rehydrate ad libitum using tap water (15C), placebo or
capsules containing NaCl to produce sodium concentration of
75 mmol/L. The purpose of this approach was to examine the
effect of sodium on drinking behavior and restoration of body
fluid compartments. Average fluid loss in the dehydration
period was 1550 ml and was followed by ingestion of 1100 ml
in the water trial and 1216 ml in the water plus sodium trial,
leaving the subjects in a fluid deficit after 3 hr of rehydration.
When urine production is subtracted from fluid ingestion, net
fluid gain during rehydration was 826 ml in the water trial and

VOL. 25, NO. 3

Sodium in Fluid Homeostasis with Exercise

1045 ml in the water plus sodium trial. Despite the persistent
negative fluid balance even after 180 min, plasma volume had
returned to pre-dehydration by 90 min of recovery in the water
plus sodium trial while plasma volume remained slightly below
the pre-dehydration level even at 180 min of recovery. Calculation of fluid compartment recovery based on chloride space
showed that by the end of the rehydration period, total body
water had recovered by 52% in the water trial and by 76% in
the water plus sodium trial. Recovery of intracellular fluid was
not different between water and water plus sodium trials. Both
ECF and PV were more completely restored in recovery in the
water plus sodium trial (84% and 100%, respectively) compared with water only (44% and 77%, respectively). These
findings illustrate the following points: 1) thirst is inadequate to
assure complete recovery of total body water deficits likely due
to early restoration of plasma volume, thereby removing the
volume dependent dipsogenic drive, 2) the presence of sodium
in the rehydration beverage stimulates greater drinking likely
due to greater osmotic dipsogenic drive, 3) the presence of
sodium in the rehydration beverage accelerates the recovery of
extracellular fluid and plasma volume in particular, and 4)
sodium in the rehydration beverage reduces urinary losses of
water, allowing a greater fraction of the ingested fluid to be
retained. These findings were later confirmed by Wemple et al.
using a similar dehydration and rehydration protocol [35].

Rehydration with Fluid Intake ! Sweat Loss

Several studies have examined recovery of body water
losses after exercise by providing an amount of fluid to subjects
that is equal to the amount of water lost during the exercise as
a consequence of sweating. Most of these studies attempted to
achieve complete rehydration within a relatively short period
lasting between 2 and 4 hours. The early study by Costill and
Sparks [36] dehydrated eight male subjects using intermittent
exposure to dry heat (70C) until 4% of body mass was lost.
Once the prescribed dehydration was reached, the men returned
to a thermoneutral environment to begin the rehydration period.
At the beginning of rehydration and at 15-min intervals the
subjects drank a volume of fluid equal to 7.7% of the volume
lost during the dehydration. This was continued for 3 hr so that,
by the end of the 3 hr rehydration period, the subjects had
ingested the same total volume of fluid as lost in dehydration.
The procedure was repeated once when ingesting plain water as
the rehydration fluid and once using a carbohydrate-electrolyte
(CE) drink for rehydration. The CE drink contained 22 mmol/L
sodium, 17 mmol/L chloride, 2.6 mmol/L potassium, 3.9
mmol/L phosphate, and 10.6 g/100ml glucose with osmolality
of 444 mOsm/L.
Urine production was significantly higher when subjects
rehydrated with water (602 ml) than when using the CE beverage (367 ml). Despite drinking a volume of fluid equal to that
which was lost in dehydration these subjects were only able to
recover 62% of their body mass loss during the rehydration.

JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION

This was mostly due to urinary and insensible loss of water

during the rehydration period. Plasma volume had dropped by
an average of 12% with dehydration and 38% of this loss was
recovered during rehydration with water while 67% of the loss
in plasma volume was recovered when drinking the CE beverage. The authors concluded that the presence of electrolytes and
carbohydrate in the rehydration favored a more complete refilling of plasma volume, but that neither beverage was adequate for completely restoring either plasma volume or total
body water when 100% of the dehydration volume is consumed
over a 3 hr period.

Rehydration With Fluid Intake > Fluid Loss

Based on the earlier observations of incomplete body water
restoration when either thirst regulates fluid intake or fluid
intake matches the fluid lost in the prior dehydration, most
recent studies have provided fluid in excess of that which was
lost in dehydration [39 43]. Authors recognized that additional
fluid was needed to offset the obligatory urinary losses, continued sweat water loss, and water loss through respiration.
These studies fail to demonstrate complete body water restoration during rehydration lasting up to 6 hours unless the
ingested fluid is coupled with sodium ingestion. A convenient
method of providing both fluid and sodium during rehydration is
to select a rehydration beverage or food providing both fluid and
sodium with other nutrients (carbohydrate and potassium, e.g.)
that may be vital in restoring normal function after dehydration.
Maughan and Leiper [39] examined the role of varied
concentrations of sodium in the rehydration beverage in achieving euhydration after mild dehydration of approximately 2%.
Their approach involved ingestion of 150% of the fluid lost
during a 30 minute period after a dehydration protocol consisting of intermittent cycling exercise in a 32C environment.
Recovery of physiological markers of dehydration was followed for 5.5 hr after ingesting the rehydration beverages. The
four beverages compared included sodium concentrations of 2,
26, 52, and 100 mmol/L. Although the fluid intake was considerably larger than used in the prior research, neither the 2
mmol/L nor 26 mmol/L beverages resulted in complete recovery of body water (66% and 82% recovery of body mass loss,
respectively) (Fig. 3). Both of the higher sodium beverages
resulted in complete (100%) rehydration by the end of the 5.5
hr monitoring period.
In an ambitious study designed to assess the interactive
effects of both sodium content and volume of fluid ingested in
rehydration, Shirreffs et al. [41] rehydrated subjects using
either 50%, 100%, 150%, or 200% of the volume lost and each
of these volumes contained either low sodium (23 mmol/L) or
higher sodium (61 mmol/L) concentration. Based on the net
fluid balance presented, body water recovery was nearly complete (91% for both) with the lower sodium fluid when consumed in both 150% and 200% excess but was incomplete with
either 50% volume (39% recovery) or 100% volume (60%

235S

Sodium in Fluid Homeostasis with Exercise

Fig. 3. Percent recovery of fluid balance during a 5.5-hr rehydration

period in which fluid was ingested at a volume equal to 150% of the
fluid deficit that was incurred. Rehydration was compared between
beverages containing 2100 mmol/L sodium. Adapted from Maughan
and Leiper [39].

recovery) (Fig. 4). With the higher sodium content in ingested

fluid, recovery of the fluid deficit was complete with ingestion
of 150% of volume lost (107% recovery) while ingestion of
200% of volume lost resulted in a surplus of fluid (127%
recovery). Neither the 50% volume nor 100% volume fully
restored whole body fluid balance (38% recovery and 81%
recovery, respectively). Urine volume was positively related to
the volume of fluid ingested and inversely related to the content
of sodium in the rehydration beverage.

Multiple Regression of Sodium Concentration and

Fluid Volume
That recovery of total body water would depend on both the
sodium intake and the volume of fluid ingested may seem
intuitively obvious. The above reviewed studies provide an

Fig. 4. Percent recovery of fluid balance during 6-hr rehydration period

in which both volume and sodium concentration of beverage were
varied. Adapted from Shirreffs et al. [41].

236S

evidentiary framework for quantifying this interactive effect.

Although each of these studies has compared rehydration between different volumes and between different intakes of sodium, there have been no attempts to use the combined data
from several studies in estimating the independent and interactive contributions of fluid volume and sodium concentration to
the rehydration process. The data displayed in Table 2 summarizing the findings of several rehydration studies were therefore
used in a multiple regression analysis to assess the relative
contributions of sodium concentration and fluid ingestion. In
each study, the data that were presented in the published paper
were either used directly (when provided by the authors) or the
relevant data were calculated from other results reported by the
authors. For the purpose of this analysis, whole-body rehydration (dependent variable) was expressed as the percentage
recovery of the fluid loss that had occurred during the dehydration protocol. The reported sodium concentration of the
rehydration solution and the volume of this solution were used
as independent variables. Initially, additional variables were
entered into the regression model but none of the other variables achieved statistical significance (p $ 0.05). The variables
which did not significantly contribute to the prediction of fluid
recovery included urine volume during dehydration (likely due
to colinearity with sodium concentration), body mass (due to
low range of body mass in the reported studies), and duration
of rehydration period (which ranged from 2 6 hr).
The final regression model included both sodium concentration (mmol/L) of the rehydration fluid and volume of this
solution consumed during the rehydration period (ml) as significant predictors of percent recovery of fluid balance (Table
3). The resulting regression equation was

% rehydration ! 22.7 " %0.406 * &Na'( " %0.021 * volume(

In the example of a 75 kg person who dehydrates by 2.5% and
ingests 100% of the volume lost during rehydration, a sodium
concentration of approximately 93 mmol/L would be required
to achieve fluid balance within 6 hr. On the other hand, if fluid
intake is increased to 150% of that lost in prior dehydration, the
regression model predicts that full rehydration could be
achieved with a sodium concentration of approximately 50
mmol/L. However, it must be noted that the regression model
accounts for only 66% of the variance in body water recovery.
It is likely that additional variables including temperature of the
ingested fluid, presence of other electrolytes (potassium, calcium, magnesium) and nutrients (carbohydrate, amino acids),
arginine vasopressin and aldosterone, and osmolality of the
rehydration fluid also play important roles but are not included
in this regression model. Thus the present analysis is incomplete but does support the contention that both fluid volume and
sodium concentration are important considerations in the selection and/or design of optimal rehydration solutions.

VOL. 25, NO. 3

Sodium in Fluid Homeostasis with Exercise

Table 2. Summary of Papers Used in Multiple Regression to Describe Relationship between Fluid Volume and Sodium
Concentration of the Rehydration (RH) Solution
Reference
Costill & Sparks 1973 [36]
Maughan & Leiper 1995 [39]

Maughan et al. 1996 [40]

Shirreffs et al. 1996 [41]

Shirreffs & Maughan 1998 [42]

Ray et al. 1998* [38]

Mitchell et al. 2000 [43]

Body Mass
(kg)

Change in Body
Mass (kg)

Volume Ingested
During RH (ml)

Sodium
Concentration
(mmol/L)

Urine Volume
(ml)

% Recovery of
Fluid Balance

71.7
71.7
71.8
71.8
71.8
71.8
66.1
66.2
71.5
71.5
71.5
71.5
73.2
73.2
73.2
73.2
69
69
69
69
72.0
72.3
72.0
72.2
79.6
79.6
79.6
79.6

2.74
2.74
1.36
1.36
1.36
1.36
1.36
1.36
1.49
1.45
1.50
1.46
1.52
1.52
1.50
1.59
1.27
1.29
1.31
1.36
1.80
2.00
1.80
1.70
2.26
2.26
2.28
2.28

2740
2740
2045
2045
2045
2045
2042
2042
746
1448
2255
2927
758
1522
2243
3180
1912
1938
1968
2035
1800
2000
1800
1700
2280
2280
3390
3390

0
60
2
26
52
100
21
21
23
23
23
23
61
61
61
61
0
25
50
100
0
21
18
35
25
50
25
50

602
367
1350
940
610
580
940
935
135
493
867
1361
194
260
602
1001
1182
970
800
578
232
310
188
231
300
180
600
540

73
73
66
82
100
100
75
73
41
69
101
103
40
83
106
136
50
69
80
101
76
76
75
78
71
104
76
101

* Sodium concentration calculated based on amount of sodium provided by ingestion of soup broth and soup diluted by additional water ingested during rehydration period.
Change in body mass from pre-dehydration to pre-rehydration.
Calculated as percentage recovery in body mass lost or net fluid balance depending on how the data were expressed in referenced paper.

Table 3. Multiple Regression of Percent Recovery of Fluid Balance as a Function of Both Volume and Sodium Concentration of
Fluid Ingested during Rehydration

Constant
Na conc
Volume

Regression
Residual
Total

Coefficient

Std Error

22.70
0.406
0.021

9.17
0.093
0.004

2.48
4.38
5.46

0.020
$0.001
$0.001

DF

SS

MS

23.9

$0.0001

2
7764
3882
25
4055
162
27
11819
438
Y ! 22.7 " %0.406Na conc( " %0.021vol intake)
R ! 0.81
R2 ! 0.66

Data were extracted from references shown in Table 2.

Rehydration with Food

One study from our laboratory [38] examined the question
of whether ingestion of food containing fluid and sodium is
effective in restoring fluid and sodium balance after a dehydrating bout of exercise and heat. Subjects were dehydrated by

JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION

2.5% using intermittent exposure to heat and exercise. Once the

prescribed fluid loss was achieved, subjects ingested 355 ml of
either chicken broth, chicken soup with noodles, a carbohydrate-electrolyte beverage, or tap water. Thereafter, the subjects
ingested an average of 290 ml water every 20 min so that total
fluid intake by 2 hr matched fluid loss. The decision to choose

237S

Sodium in Fluid Homeostasis with Exercise

these products was based on commercial availability to
consumers as well as their varied amounts of electrolytes and
osmolality. With regard to sodium intake, chicken noodle soup
and chicken broth treatments provided a total sodium ingestion
of 50 mmol and 39 mmol, respectively. This is considerably
less than the sodium intake associated with the prior studies in
which subjects ingested 150% of the fluid loss with a sodium
concentration of 50 100 mmol/L. Using the regression model
from above, it is expected that the chicken broth and the
chicken noodle soup treatments would not fully restore the
fluid deficit in 3 hr (estimated % rehydration ! 73% for both).
Measured fluid recovery was 76% and 78% for the chicken
broth and chicken noodle soup, respectively. Although total
body fluid balance was not fully recovered in rehydration,
plasma volume was fully restored with the chicken broth and
the chicken noodle soup trials, but not with either a commercial
carbohydrate-electrolyte beverage or with water.
These findings illustrate the importance of ingestion of
sodium during the rehydration period not only for encouraging
increased retention of ingested fluids but also for restoration of
the plasma volume, which can be re-filled ahead of total fluid
balance when sufficient sodium is provided either in the rehydration drink or in food consumed during rehydration. In addition, these findings show that it may not be necessary to
include sodium in every aliquot of fluid ingested during rehydration if sufficient sodium is provided early in the rehydration
period either as a constituent of fluid or food.

SUMMARY AND CONCLUSION

Both sodium and fluid ingestion play important roles in
maintaining health and physiological function during physical
activity in hot environments. Whether people engage in prolonged endurance exercise such as marathons and triathlons or
if they are involved in occupational heat exposure during
physical activity, it is important that both fluid and sodium are
provided to offset the losses in both nutrients that occur as a
consequence of heavy sweating. People involved in vigorous
exercise in hot environments lose up to 3 liters of water and 3.5
grams of sodium per hour through sweating. Preventing these
fluid and sodium deficits helps to maintain both performance
and thermoregulation in such environments. The evidence from
published literature shows that fluid intake during exercise in a
warm environment is absolutely essential to attenuate the rise
in core temperature. These studies also demonstrate that unless
sodium is provided in the fluid replacement beverage, fluid
intake that matches or exceeds fluid loss may cause hyponatremia in some individuals participating in at least 4 hr of
exercise. Thus, many authors now recommend sodium concentration of 20 50 mmol/L in beverages consumed during the
physical activity.
In designing a nutritional strategy for recovery from exercise and heat exposure that results in mild dehydration, the dual

238S

and interactive roles of fluid and sodium intake should be

considered. This synergistic association between fluid volume
and sodium intake is reflected in recommendations to consume
fluid in excess of that lost during the prior exercise and to
include sodium to increase the retention of the ingested liquids
by minimizing urine production. The papers reviewed here
suggest that plasma volume can be fully restored before total
body water deficits are fully corrected when sodium intake is
consumed either as a component of the rehydration beverage
with sodium concentration of approximately 20 mmol/L or
with food consumed in the early part of a rehydration period.
Using the meta-analysis presented in this paper, full recovery of
the fluid deficit within 6 hrs requires ingestion of a rehydration
solution containing 100 mmol/L sodium if consuming the same
volume of fluid that was lost in the prior dehydration. Alternatively, correction of the fluid deficit can also be achieved by
ingesting 150% of the volume lost if the rehydration solution
contains 50 mmol/L sodium.

REFERENCES
1. Senay LC Jr: Relationship of evaporative rates to serum [Na#],
[K#], and osmolarity in acute heat stress. J Appl Physiol 25:149
152, 1968.
2. Shapiro Y, Pandolf KB, Goldman RF: Predicting sweat loss response to exercise, environment and clothing. Eur J Appl Physiol
48:8396, 1982.
3. Allan JR, Wilson GG: Influence of acclimatization on sweat sodium concentration. J Appl Physiol 30:708712, 1971.
4. Sawka MN, Toner MM, Francesconi RP, Pandolf KB: Hypohydration and exercise: effects of heat acclimation, gender, and
environment. J Appl Physiol 55:11471153, 1983.
5. Sawka MN, Montain SJ: Fluid and electrolyte supplementation for
exercise heat stress. Am J Clin Nutr 72(Suppl):564S72S, 2000.
6. Costill DL: Sweating: its composition and effects on body fluids.
Ann NY Acad Sci 301:160174, 1977.
7. Rehrer NJ: Fluid and electrolyte balance in ultra-endurance sport.
Sports Med 31:701715, 2001.
8. Costill DL, Cote R, Fink W: Muscle water and electrolytes following varied levels of dehydration in man. J Appl Physiol 40:6
11, 1976.
9. Consolazio CF, Matoush LO, Nelson RA, Harding RS, Canham
JE: Excretion of sodium, potassium, magnesium and iron in human
sweat and the relation of each to balance and requirements. J Nutr
79:407415, 1963.
10. Mao IF, Chen ML, Ko YC: Electrolyte loss in sweat and iodine
deficiency in a hot environment. Arch Environ Health 56:271277,
2001.
11. Costa F, Calloway DH, Margen S: Regional and total body sweat
composition of men fed controlled diets. Am J Clin Nutr 22:5258,
1969.
12. Shirreffs SM, Maughan RJ: Whole body sweat collection in humans: an improved method with preliminary data on electrolyte
content. J Appl Physiol 82:336341, 1997.
13. Sanders B, Noakes TD, Dennis SC: Sodium replacement and fluid

VOL. 25, NO. 3

Sodium in Fluid Homeostasis with Exercise

14.

15.

16.
17.

18.
19.

20.
21.

22.

23.

24.

25.

26.

27.
28.

29.

30.

shifts during prolonged exercise in humans. Eur J Appl Physiol

84:419425, 2001.
Convertino VA, Armstrong LE, Coyle EF, et al.: American College of Sports Medicine position stand: exercise and fluid replacement. Med Sci Sports Exerc 28:ivii, 1996.
Speedy DB, Campbell R, Mulligan G, et al: Weight changes and
serum sodium concentrations after an ultraendurance multisport
triathlon. Clin J Sport Med 7:100103, 1997.
Noakes TD: The hyponatremia of exercise. Int J Sport Nutr 2:205
228, 1992.
Irving RA, Noakes TD, Buck R et al.: Evaluation of renal function
and fluid homeostasis during recovery from exercise-induced hyponatremia. J Appl Physiol 70:342348, 1991.
Speedy DB, Noakes TD, Rogers IR, et al.: Hyponatremia in
ultradistance athletes. Med Sci Sports Exerc 31:809815, 1999.
Hew TD, Chorley JN, Cianca JC, Divine JG: The incidence, risk
factors, and clinical manifestations of hyponatremia in marathon
runners. Clin J Sport Med 13:4147, 2003.
Noakes T: Hyponatremia in distance runners: fluid and sodium
balance during exercise. Curr Sports Med Rep 1:197207, 2002.
Hiller WDB, OToole ML, Fortress EE, Laird RH, Imbert PI, Sisk
TD: Medical and physiological considerations in triathlons. Am J
Sport Med 15:164167, 1987.
Luetkemeier MJ, Coles MG, Askew EW: Dietary sodium and
plasma volume levels with exercise. Sports Med 23:279286,
1997.
Twerenbold R, Knechtle B, Kakebeeke TH, et al: Effects of
different sodium concentrations in replacement fluids during prolonged exercise in women. Br J Sports Med 37:300303, 2003.
Vrijens DMJ, Rehrer NJ: Sodium-free fluid ingestion decreases
plasma sodium during exercise in the heat. J Appl Physiol 86:
18471851, 1999.
Montain SJ, Sawka MN, Wenger CB: Hyponatremia associated
with exercise: risk factors and pathogenesis. Exerc Sport Sci Rev
29:113117, 2001.
Gisolfi CV, Duchman SM: Guidelines for optimal replacement
beverages for different athletic events. Med Sci Sports Exerc
24:679687, 1992.
Eichner ER: Exertional hyponatremia: why so many women.
Sports Med 24:5456, 2002.
Barr SI, Costill DL, Fink WJ: Fluid replacement during prolonged
exercise: effects of water, saline, or no fluid. Med Sci Sport Exerc
23:811817, 1991.
Brandenberger G, Candas V, Follenius M, Kahn JM: The influence
of the initial state of hydration on endocrine responses to exercise
in the heat. Eur J Appl Physiol 58:674679, 1989.
Cade R, Spooner G, Schlein E, Pickering M, Dean R: Effect of

JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION

31.

32.

33.

34.

35.

36.
37.

38.

39.
40.

41.

42.

43.

fluid, electrolyte, and glucose replacement during exercise on

performance, body temperature, rate of sweat loss, and compositional changes of extracellular fluid. J Sports Med Phys Fitn
12:150156, 1972.
Greenleaf JE, Brock PJ: Na# and Ca2# ingestion: plasma volumeelectrolyte distribution at rest and exercise. J Appl Physiol 48:838
847, 1980.
Powers SL, Lawler J, Dodd S, Tulley R, Landry G, Wheeler K:
Fluid replacement drinks during exercise: effects on minimizing
exercise-induced disturbances in homeostasis. Eur J Appl Physiol
60:5460, 1990.
Nose H, Mack GW, Shi X, Nadel ER: Involvement of sodium
retention hormones during rehydration in humans. J Appl Physiol
65:332336, 1988.
Nose H, Mack GW, Shi X, Nadel ER: Role of osmolality and
plasma volume during rehydration in humans. J Appl Physiol
65:325331, 1988.
Wemple RD, Morocco TS, Mack GW: Influence of sodium replacement on fluid ingestion following exercise-induced dehydration. Int J Sport Nutr 7:104116, 1997.
Costill DL, Sparks KE: Rapid fluid replacement following thermal
dehydration. J Appl Physiol 34:299303, 1973.
Gonzalez-Alonzo J, Heaps CL, Coyle EF: Rehydration after exercise with common beverages and water. Int J Sports Med 13:399
406, 1992.
Ray ML, Bryan MW, Ruden TM, Baier SM, Sharp RL, King DS:
Effect of sodium in a rehydration beverage when consumed as a
fluid or meal. J Appl Physiol 85:13291336, 1998.
Maughan RJ, Leiper JB: Sodium intake and post-exercise rehydration in man. Eur J Appl Physiol 71:311319, 1995.
Maughan RJ, Leiper JB, Shirreffs SM: Restoration of fluid balance
after exercise-induced dehydration: effects of food and fluid intake. Eur J Appl Physiol 73:317325, 1996.
Shirreffs SM, Taylor AJ, Leiper JB, Maughan RJ: Post-exercise
rehydration in man: effects of volume consumed and drink sodium
content. Med Sci Sports Exerc 28:12601271, 1996.
Shirreffs SM, Maughan RJ: Volume repletion after exerciseinduced volume depletion in humans: replacement of water and
sodium losses. Am J Physiol 274(Renal Physiol 43):F868F875,
1998.
Mitchell JB, Phillips MD, Mercer SP, Baylies HL, Pizza FX:
Postexercise rehydration: effect of Na# and volume on restoration
of fluid spaces and cardiovascular function. J Appl Physiol 89:
13021309, 2000.

Received January 9, 2006.

239S