Sports M e d 2CO7; 37 (4^5): 344-347

0112-1642/O7/CXXM-O344/S44.W/0

CONFERENCE PAPER

O 2007 Adls Data Infarmatton BV. All rights reserved.

Nutrition Strategies for the Marathon

Fuel for Training and Racing
Louise M. Burke
Department of Sports Nutrition, Australian Institute of Sport, Belconnen, Australian Capital
Territory, Australia

Abstract

Muscie glycogen provides a key fuel for training and racing a marathon.
Carbohydrate 'loading' can enhance marathon perfonnance by allowing the
competitor to run at their optimal pace for a longer period before fatiguing. For the
well trained runner, this may be achieved by tapering exercise over the final days
before the marathon and ensuring carbohydrate intakes of 10-12 g/kg/day over the
36-48 hours prior to the race. Sports nutrition guidelines recommend that the
runner consumes sufficient carbohydrate to promote restoration of muscle glycogen between training sessions. This strategy should allow the runner to 'train
harder' and recover optimally between workouts. A recent hypothesis suggests
that runners might 'train smarter' by training with low glycogen stores, since this
might promote greater stimulation of the training response. However, there is no
evidence that a low carbohydrate diet enhances the outcomes of training or
provides benefits as a depletion phase prior to carbohydrate loading. In fact, a low
carbohydrate diet may even impair performance if carried out for extended
periods. If there are benefits to manipulating glycogen stores for some workouts,
this is likely to happen as the natural outcome of the periodisation of the highvolume programme of an elite runner.

To run 42.195km is one of the ultimate challenges in sport. Therefore, it is not surprising that as
the science of sports nutrition has evolved over tlie
past 30 years, marathon runners have been quick to
put new knowledge into practice in the field. This
review will focus on muscle glycogen as a fuel for
training and racing, noting how the marathon helped
to popularise the discovery of this important exercise fuel.
1. Carbohydrate Loading
1.1 History of Protocols
Because depletion of muscle glycogen stores occurs so predictably during the latter stages of a

marathon, the term 'hitting the wall', which describes this overwhelming fatigue, has become part
of everyday jargon. Carbohydrate loading, or supercompensating the muscle glycogen stores in preparation for prolonged exercise, resulted from studies
undertaken in the late 1960s."' Using percutaneous
biopsy techniques to examine fuel utilisation and
enzyme activities in the muscle, Scandinavian sports
scientists found that several days of low-carbohydrate eating depleted muscle glycogen and reduced
cycling endurance compared with the results associated with a moderate carbohydrate intake. However,
the subsequent intake of a high-carbohydrate diet
over several days super-compensated glycogen
stores and further increased cycling time to exhaus-

Nutrition for the Marathon

tion. These pioneering studies on healthy but untrained men produced the classic 7-day model of
carbohydrate loading; a 3-4 day depletion phase of
hard training and low carbohydrate intake followed
by a 3-4-day loading phase of high carbohydrate
intake and exercise taper. Carbohydrate-loading
practices were taken up first by competitive marathon runners, including 1969 European marathon
champion Ron Hill, and later by recreational runners
who joined the running boom of the 1970s and
1980s.
A modified version of carbohydrate loading was
developed when well trained mnners were shown to
super-compensate their glycogen stores without the
necessity of a severe depletion or glycogen stripping
phase.'^' The modified protocol, consisting simply
of 3 days of high carbohydrate intake and taper, was
offered as a more practical competition preparation
which avoided the fatigue and complexity of the
extreme diet and training requirements of the previous depletion phase. More recently, muscle glycogen concentrations were measured after 1 and 3 days
of rest and a high carbohydrate intake (10 g/kg body
mass per day) in well trained male athletes.'^' After 1
day, muscle glycogen content increased significantly from pre-loading levels of =90 mmol/kg wet
weight to values of =180 mmol/kg wet weight.
Thereafter, glycogen levels remained stable despite
another 2 days of rest and carbohydrate intake. This
provides a time course of glycogen storage, showing
that super-compensation may not require a full
72-hour period in rested athletes. Rather than promoting a unique strategy for carbohydrate loading as
suggested by the authors, this study shows that
optimal refuelling is probably achieved within
36-48 hours following the last exercise session, at
least when the athlete rests and consumes adequate
carbohydrate.

1.2 Gender and Corbohydrate Loading

Most studies of glycogen storage have involved
male participants, with the assumption that the results also apply to females. There have been claims
2007 Adls Data Information BV. All rights reserved.

345

that female athletes fail to super-compensate muscle

glycogen stores compared with males and fail to
show a performance benefit following carbohydrate
loading.'"! However, this is at least partly explained
by the relatively smedler amounts of dietary carbohydrate and restricted energy intakes of many female athletes. Indeed, when female athletes are provided with sufficient energy and carbohydrate intake, they are able to achieve a significant increase
in glycogen storage, similar to that seen in a male
population.'^' Menstrual status of female athletes
may affect glycogen storage, with greater storage
occurring during the luteal rather than the follicular
phase. Further studies of gender differences in the
achievement of, and response to, carbohydrate loading are warranted. In the meantime, it is likely that
the main challenge related to carbohydrate-loading
practices of females is the issue of adequate die
carbohydrate and energy.
1,3 Carbohydrate Loading and
iVlarathon Performanoe
Theoretically, carbohydrate loading can enhance
performance in sporting events that would otherwise
be limited by glycogen depletion. Although a half
marathon event is too short to benefit from supercompensated carbohydrate stores, carbohydrate
loading has been shown to enhance overall performance of a 30km cross-country run, a 30km treadmill
run in trained men and a 25km treadmill run in
moderately trained men.f*' Where perfonnance enhancements have been found, carbohydrate loading
was associated not with an increase in overall running speed but with maintenance of race pace during
the last part of the run compared with the control
trial or control group.'*' Even when carbohydrate
loading did not cause a statistically significant enhancement of total running time, participants were
found to run faster over the last 5km than in a
control trial. Therefore, even though field studies
have not been conducted on the marathon, it appears
highly likely that it will benefit from a carbohydrate
loading preparation.

Sports Med 2007: 37 (4-5)

Burke

346

2. Fat Adaptation: a Twist on

Pre-Loading Depletion
Marathon runners should have a high capacity for
fat oxidation during exercise as a legacy of their
training. However, this capacity can be further upregulated by as little as 5 days of training while
following a low-carbohydrate (<2.5 g/kg/day) and
high-fat (=65-70% of energy) diet. In trained individuals, such 'fat adaptation' achieves a markedly
increased fat oxidation and reduced utilisation of
muscle glycogen ('glycogen sparing') during subsequent sub-maximal exercise. However, there are no
clear benefits to performance and there is the possibility that long periods of exposure (>4 weeks) may
impair training outcome and cause some health risks
(for a review, see Burke and Kiens'^').
Research during the 1990s showed that the metabolic changes created by fat adaptation are robust. In
fact, they persist for at least 24 hours despite aggressive tactics to increase carbohydrate availability
(e.g. restoration of muscle glycogen and the consumption of carbohydrate before and during exercise). In effect, fat adaptation could be viewed as a
slightly extended 'depletion phase' prior to carbohydrate loading. Such a combination of dietary strategies would seem the perfect preparation for a mai athon, simultaneously optimising carbohydrate stores
while maximising the capacity for fat oxidation.
Curiously, the effect on performance is unclear.'^'
The apparent failure of translation of metabolic
changes has been variously explained as a failure of
scientists to detect small changes in perfonnance
that might be worthwhile in real-life sport, or the
existence of 'responders' and 'non-responders' to
fat adaptation strategies.
There is now evidence that what was initially
viewed as glycogen sparing may be, in fact, a downregulation of carbohydrate metabolism or 'glycogen
impairment'. Fat adaptation/ carbohydrate restoration strategies are associated with a reduction in the
activity of a key enzyme regulating carbohydrate
metabolism, pymvate dehydrogenase (see Burke
and Kiens'^l)- Such a change would impair rates of
glycogenolysis at a time when muscle carbohydrate
requirements are high. This explains the observation
2007 Adls Data Information BV. All rights reserved.

that when fat adaptation/carbohydrate restoration

are applied to exercise protocols that mimic a reallife race (i.e. self-pacing, and the interspersing of
high-intensity and moderate-intensity exercise),
there is a compromised ability to perfonnance highintensity sprints. Although a marathon is viewed as
an endurance event, the critical activities in a race
(i.e. the breakaway, the surge up a hill or the sprint
to the fmish line) are all dependent on the runner's
ability to work at high intensities. With growing
evidence that this critical ability may be impaired, it
now seems clear that fat adaptation or pre-loading
depletion strategies should not be undertaken by
marathon runners.
3. Training and iVIuscie Glycogen
According to the current guidelines for sports
nutrition,'^! the everyday diet of a marathon runner
should provide enough carbohydrate to cover the
fuel costs of their training programme and restore
glycogen between workouts. This is likely to be in
the range of 7-12 g/kg/day according to volume and
intensity of training sessions. However, data pertaining to nutrient-gene interactions and the cellular
signalling pathways that promote muscle adaptations to training have recently suggested that training with low or moderate glycogen levels may accelerate the transcription of several important genes.
In support of this, one study has shown that training
with low glycogen levels might accelerate training
outcomes. In this study, untrained subjects achieved
greater increases in muscle enzyme content and
exercise endurance in the leg trained with a protocol
promoting depleted glycogen stores, than the contralateral leg, which undertook the same volume of
training in a glycogen-recovered state.''' However,
other chronic studies of diet and training interventions in well trained athletes, including runners^'"'
have shown that higher carbohydrate intakes that
allow greater glycogen recovery are associated with
fewer symptoms of overtraining during high-volume periods and greater training adaptations.'"*' In
real life, it is likely that high level runners optimise
training outcomes by periodising their training and
diet so that some sessions are undertaken with relaSports M e d 2007; 37 (4-5)

Nutrition for the Marathon

tive glycogen depletion while important sessions are

targeted for better refuelling.
4. Summary
Carbohydrate loading techniques are recommended as a race preparation to help the marathon
runner cross the finish line according to their potential. For the well trained runner, this may be as
simple as tapering exercise over the final days
before the marathon and ensuring carbohydrate intakes of 10-12 g/kg/day over the 36-48 hours prior
to a race. It is no longer considered beneficial to
undertake a depletion phase prior to carbohydrate
loading and this may even impair performance if
carried out for an extended period. The restoration
of glycogen between workouts remains a principle
of the training diet and may require daily carbohydrate intakes in the range of 7-12 g/kg body mass.
There may be some benefits to manipulating glycogen stores for some workouts, but this is likely to
happen as the natural outcome of the periodisation
of the high-volume programme of an ehte runner.
Acknowledgements
Louise M. Burke is a contracted researcher for Nestec Ltd.

2007 Adis Data Information BV. All rights reserved.

347

References
1. Ahlborg G, Bergstrom J, Brohult J. Human muscle glycogen
content and capacity for prolonged exercise after difference
diets. Forsvarsmedicin 1967; 3: 85-99
2. Sherman WM, Costill DL, Fink WJ, et al. Effect of exercise-diet
manipulation on muscle glycogen and its subsequent utilisation during performance. lnt J Sports Med 1981; 2: 114-8
3. Bussau VA, Fairchild TJ, Rao A, et al. Carbohydrate loading in
human muscle: an improved 1 day protocol. Eur J Appl Physiol 2002; 87: 290-5
4. Burke L. Middle and long-distance running. In: Practical sports
nutrition. Champaign (IL): Human Kinetics, 2007: 109-139
5. Tamopolsky MA, Zawada C, Richmond LB, et al. Gender
differences in carbohydrate loading are related to energy intake. J Appl Physiol 2001; 91: 225-30
6. Karlsson J, Saltin B. Diet, muscle glycogen, and endurance
perfonnance. J Appl Physiol 1971; 31: 203-6
7. Burke LM, Kiens B. "Fat adaptation" for athletic performance:
the nail in the coffin? J Appl Physiol 2006; 100 (1): 7-8
8. Burke LM, Kiens B, Ivy JL. Carbohydrates and fat for training
and recovery. J Sports Sci 2004; 22: 15-30
9. Hansen AK, Fischer CP, Plomgaard P, et al. Skeletal muscle
adaptation: training twice every second day vs training once
daily. J Appl Physiol 2005; 98: 93-9
10. Achten J, Halson SH, Moseley L, et al. Higher dietary carbohydrate content during intensified running training results in
better maintenance of performance and mood state. J Appl
Physiol 2004; 96: 1331-40

Correspondence: Louise M. Burke, Department of Sports

Nutrition, Australian Institute of Sport, Belconnen, ACT
2616, Australia.
E-mail: Louise.burkeausport.gov.au

Sports Med 2007; 37 (4-5)