Fluid and fuel intake during exercise
EDWARD F. COYLE
The Human Performance Laboratory, Department of Kinesiology and Health Education, The University of Texas at
Austin, Austin, TX 78712, USA
Accepted 7 August 2003
The amounts of water, carbohydrate and salt that athletes are advised to ingest during exercise are based upon
their effectiveness in attenuating both fatigue as well as illness due to hyperthermia, dehydration or
hyperhydration. When possible, fluid should be ingested at rates that most closely match sweating rate. When
that is not possible or practical or sufficiently ergogenic, some athletes might tolerate body water losses
amounting to 2% of body weight without significant risk to physical well-being or performance when the
environment is cold (e.g. 5–108C) or temperate (e.g. 21–228C). However, when exercising in a hot environment
(4308C), dehydration by 2% of body weight impairs absolute power production and predisposes individuals to
heat injury. Fluid should not be ingested at rates in excess of sweating rate and thus body water and weight
should not increase during exercise. Fatigue can be reduced by adding carbohydrate to the fluids consumed so
that 30–60 g of rapidly absorbed carbohydrate are ingested throughout each hour of an athletic event.
Furthermore, sodium should be included in fluids consumed during exercise lasting longer than 2 h or by
individuals during any event that stimulates heavy sodium loss (more than 3–4 g of sodium). Athletes do not
benefit by ingesting glycerol, amino acids or alleged precursors of neurotransmitter. Ingestion of other
substances during exercise, with the possible exception of caffeine, is discouraged. Athletes will benefit the most
by tailoring their individual needs for water, carbohydrate and salt to the specific challenges of their sport,
especially considering the environment’s impact on sweating and heat stress.
Keywords: carbohydrate, dehydration, fatigue, gastrointestinal function, hyperthermia, sodium.
Introduction
When athletes exercise during training or while
competing, it is clear that they sometimes benefit by
ingesting various mixtures of water, carbohydrate and
electrolytes (Convertino et al., 1996; Casa, 2000). The
benefits can be expressed through improved perfor-
mance and/or reduced physiological stress, on an
athlete’s cardiovascular, central nervous and muscular
systems. Although ample scientific evidence exists to
support the general theory for encouraging athletes to
consume water, carbohydrate and electrolytes during
exercise, the practical recommendations for optimally
applying these general theories is not simple. This is
due to the quite varied nature of the physical stresses
encountered during training and competition for a wide
range of sports, as well as the unique rules of each sport
regarding the allowance for fluid and fuel intake during
competition. Furthermore, variations in the physical
intensity, duration and environment, as well as indivi-
dual characteristics of a given athlete, might alter their
optimal rate of water, carbohydrate and salt intake as
well as the rate of gastrointestinal absorption and
feelings of fullness. This task of developing consensus
for general recommendations becomes less daunting
when approached systematically through logical inter-
pretation of the vast literature that has grown exponen-
tially in the past two decades. It is also important to
incorporate the experience of athletes recorded in
carefully controlled ‘field’ studies or case reports, as
this serves as a reality check regarding the conditions in
which theories and laboratory findings can be applied
directly to the overall welfare of athletes.
The aim of this article is to develop practical
recommendations for fluid and fuel intake during
exercise based upon interpretation of the scientific
literature, with heavy reliance upon controlled labora-
tory studies as well as careful study of athletes in the
field during training and competition. The focus of this
review will be on identifying the conditions during
which fluid and fuel intake positively or adversely affects
either athletic performance or the physical well-being
(i.e. acute and chronic health) of the individual athlete.* e-mail: coyle@mail.utexas.edu
Journal of Sports Sciences, 2004, 22, 39–55
Journal of Sports Sciences ISSN 0264-0414 print/ISSN 1466-447X online # 2004 Taylor & Francis Ltd
DOI: 10.1080/0264041031000140545

Discussion of the physiological mechanisms mediating
the performance effects will be limited to that which
helps elucidate the practical application or refine the
recommendation for fluid and fuel intake during
exercise. Several excellent scientific review articles have
already detailed the physiological mechanisms mediat-
ing reduced physiological strain during exercise by
ingesting fluids, carbohydrate and electrolytes (Gallo-
way and Maughan, 1998; Sawka and Coyle, 1999;
Sawka and Montain, 2000; Cheuvront and Haymes,
2001; Cheuvront, 2001; Sawka et al., 2001). The last 40
years have seen dramatic shifts in both scientific
thinking and popular practice regarding water, carbo-
hydrate and salt intake during exercise. Here, I develop
practical guidelines for the exercising athlete, based on
modern science, recognizing lessons from past applica-
tion and the limitations of our present understanding.
Water or fluid intake
A person’s physiological drive for fluid intake during
exercise is perceived through ‘thirst mechanisms’ and it
has long been known that when given ad libitum access
to fluid, and thus drink voluntarily, that these mechan-
isms compel people to drink at a rate that replaces
approximately one-half of their fluid losses and at best
two-thirds (Pitts and Consolazio, 1944; Hubbard et al.,
1984). The concept that thirst during exercise does not
drive people to take in fluid at the rate of fluid loss is
termed ‘voluntary dehydration’. In the 1960s, athletes
were generally advised ‘to drink only a little water
during exercise’ and to ‘ignore their thirst’ and to thus
replace a small percentage of lost fluid. Furthermore,
the scientific literature in the 1960s was interpreted to
suggest that dehydration by less than 3–4% of body
weight caused insignificant hyperthermia or impairment
of physiological function and performance, although it
was recognized that dehydration by more than 4% is
dangerous to health (Wyndham and Strydom, 1969).
This belief that dehydration by 3–4% was tolerable
prevailed despite evidence to the contrary from carefully
conducted studies around the time of the Second
World War that had clearly shown that dehydration by
less than 3–4% during prolonged marching in the heat
impaired performance and caused exhaustion and
collapse (Pitts and Consolazio, 1944; Adolph, 1947;
Ladell, 1955; Coyle and Montain, 1992a,b). The
athletic community of this era appears to have remained
unaware or unconvinced that the demands of marching
were not unlike athletic endurance events, except, of
course, for the speed and practical aspects of drinking
while running compared with walking.
It was not until the 1970s that athletes were generally
advised to ingest something more than just ‘a little
water’ during exercise. In 1975, the American College
of Sports Medicine (ACSM) published its first position
stand entitled ‘Prevention of heat injury during distance
running’. In 1985, the ACSM published another
position stand that advised runners to drink 100–
200 ml of fluid after every 2–3 km. This recommenda-
tion acknowledged the large variation in ideal rate of
fluid replacement among runners, with faster runners
needing more and slower runners less fluid intake per
hour. However, these broad recommendations did not
provide sufficiently practical guidelines for drinking
relative to thirst or sweating rate. At the extremes, this
recommendation could be interpreted to suggest that it
is permissible for slow runners (10 km
h
71
) to drink
only 330 ml
h
71
, whereas the fastest runners should
drink as much 2000 ml
h
71
. The latter is a rate of fluid
intake that is unrealistically high for most fast runners.
However, this 1985 recommendation has theoretical
merit in that it accurately set the limits for rates of
sweating in slow and fast runners (330–2000 ml
h
71
).
However, the ideal application of these broad guidelines
would require the athlete to devise the schedule that
meets not only their assumed need for fluid based on
sweating rate, but also the rate of fluid replacement that
is practical for them individually. It is unrealistic to
expect that brief guidelines, which are naturally general,
can be practised by all athletes in all sports under all
conditions.
In its most recent position stand (Convertino et al.,
1996), the ACSM recommended that: ‘During ex-
ercise, athletes should start drinking early and at regular
intervals in an attempt to consume fluids at a rate
sufficient to replace all the water lost through sweating
(body weight loss), or consume the maximal amount
that can be tolerated’. This document further stated
that ‘individuals should be encouraged to consume the
maximal amount of fluids during exercise that can be
tolerated without gastrointestinal discomfort up to a
rate equal to that lost from sweating’.
In an attempt to bring attention to the potential for
developing hyponatraemia from drinking excessively
large volumes of fluid during marathon running, Tim
Noakes has been critical of the 1996 guidelines from
ACSM (Noakes, 2003). Noakes quotes the 1996
guidelines for his statement that ‘athletes are now
advised to replace all the water lost through sweating
(that is, loss of body weight), or consume the maximal
amount that can be tolerated, or drink 600–1200 ml per
hour’. It is perplexing to the author of the current
review how Dr. Noakes could interpret the 1996
guidelines from ACSM (Convertino et al., 1996) to
suggest multiple recommendations, including the no-
tion that athletes drink more fluid than is lost during
exercise and therefore gain body weight. In an attempt
to minimize current misinterpretation, misrepresenta-
40 Coyle