1. The study examined the physiological demands of a simulated Muay Thai boxing match through measuring oxygen uptake, heart rate, carbon dioxide production, and other variables in 10 trained male athletes. 2. Results showed that energy expenditure during the match was high, at an average of 10.75 kcal/min, and relied on both aerobic and anaerobic energy systems. 3. There was an initial high recruitment of anaerobic glycolysis in the first round as shown by a spike in excess carbon dioxide production, which then gradually decreased over the rounds as aerobic metabolism increased.

1. 143
Physiological responses and energy cost during a
simulation of a Muay Thai boxing match
Antonio Crisafulli, Stefano Vitelli, Ivo Cappai, Raffaele Milia, Filippo Tocco,
Franco Melis, and Alberto Concu
Abstract: Muay Thai is a martial art that requires complex skills and tactical excellence for success. However, the energy
demand during a Muay Thai competition has never been studied. This study was devised to obtain an understanding of the
physiological capacities underlying Muay Thai performance. To that end, the aerobic energy expenditure and the recruit-
ment of anaerobic metabolism were assessed in 10 male athletes during a simulation match of Muay Thai. Subjects were
studied while wearing a portable gas analyzer, which was able to provide data on oxygen uptake, carbon dioxide produc-
tion, and heart rate (HR). The excess of CO2 production (CO2 excess) was also measured to obtain an index of anaerobic
glycolysis. During the match, group energy expenditure was, on average (mean ± standard error of the mean), 10.75 ±
1.58 kcalÁmin–1, corresponding to 9.39 ± 1.38 metabolic equivalents. Oxygen uptake and HRs were always above the level
of the anaerobic threshold assessed in a preliminary incremental test. CO2 excess showed an abrupt increase in the first
round, and reached a value of 636 ± 66.5 mLÁmin–1. This parameter then gradually decreased throughout the simulation
match. These data suggest that Muay Thai is a physically demanding activity with great involvement of both the aerobic
metabolism and anaerobic glycolysis. In particular, it appears that, after an initial burst of anaerobic glycolysis, there was
a progressive increase in the aerobic energy supply. Thus, training protocols should include exercises that train both aero-
bic and anaerobic energetic pathways.
Key words: martial arts, exercise, energy expenditure, oxygen uptake, anaerobic glycolysis, carbon dioxide excess.
´ ´ ¨ ¨ ´
Resume : Le muay thaı ou boxe thaılandaise est un art martial exigeant des habiletes complexes et une tactique de haut
´ ´ ´ ´
niveau pour reussir. Il n’y a pas encore d’etude sur les sources d’energie de cet art martial lors d’une competition. Le but
´ ´ ´ ´ ` ´ ´
de cette etude est d’evaluer les capacites physiologiques des individus en competition. A cette fin, on evalue la depense
´ ´ ´ ´ ´
d’energie aerobie (EE) et l’implication du metabolisme anaerobie chez dix sujets masculins au cours d’une competition si-
´ ¨ ´ ´
mulee de muay thaı. L’evaluation des sujets en competition se fait au moyen d’un analyseur de gaz portable fournissant
´ ` ´
les donnees de consommation d’oxygene, de production de gaz carbonique et de frequence cardiaque (HR). On evalue ´
´
aussi le surplus de production de gaz carbonique afin d’obtenir un indice de la sollicitation de la glycolyse anaerobie. Au
cours du match, la EE est en moyenne (± erreur type sur la moyenne) de 10,75 ± 1,58 kcalÁmin–1, ce qui equivaut a
´ `
´
9,39 ± 1,38 METs. Les valeurs de consommation d’oxygene et de HR sont toujours au-dessus du seuil anaerobie preala- ´
´ ´
blement determine au cours d’un test d’effort progressif. Au cours du premier round, le surplus de dioxyde de carbone pro-
duit presente une augmentation marquee et affiche la valeur de 636 ± 66,5 mLÁmin–1. Cette valeur s’abaisse graduellement
´ ´
` ¨
au cours du match. D’apres ces observations, le muay thaı est une discipline sportive exigeante sur les plans de la sollicita-
´ ´ ´ ` ´ ´ ´
tion des deux metabolismes, aerobie et anaerobie. Apres une sollicitation marquee du metabolisme anaerobie au debut du ´
´ ´ `
match, on observe un engagement graduel du metabolisme aerobie. Il faudrait donc veiller a solliciter les deux modalites ´
´ ´ ˆ
de fourniture d’energie dans l’elaboration d’un programme d’entraınement.
´ ´ ´ ` ´
Mots-cles : arts martiaux, exercice physique, depense d’energie, consommation d’oxygene, glycolyse anaerobie, surplus de
production de gaz carbonique.
´
[Traduit par la Redaction]
Introduction worldwide, with 5 continental federations, under a sole and
unified regulatory body.
Muay Thai, often translated into English as Thai boxing, Muay Thai requires complex skills and tactical excellence
is the national sport of Thailand and is a martial art with ori- for success. Matches are characterized by dynamic phases of
gins in the ancient battlefield tactics of the Siamese army. short duration, during which athletes try to strike their oppo-
During the latter half of the 20th century, Muay Thai was nent or defend themselves from the attacks of their oppo-
exported to many countries, and now the International Fed- nent. Fighters wear boxing gloves and use several parts of
eration of Muaythai Amateur claims 110 member countries the body for offensive and defensive purposes, including
Received 27 May 2008. Accepted 09 January 2009. Published on the NRC Research Press Web site at apnm.nrc.ca on 28 March 2009.
A. Crisafulli,1 S. Vitelli, I. Cappai, R. Milia, F. Tocco, F. Melis, and A. Concu. Department of Science Applied to Biological
Systems, Section of Human Physiology, University of Cagliari, Cagliari, Italy.
1Corresponding author (e-mail: crisafulli@tiscali.it).
Appl. Physiol. Nutr. Metab. 34: 143–150 (2009) doi:10.1139/H09-002 Published by NRC Research Press

2. 144 Appl. Physiol. Nutr. Metab. Vol. 34, 2009
fists, elbows, knees, and feet, but headbutting an opponent is Experimental protocol
not allowed. Therefore, Muay Thai shares many similarities
with several forms of martial arts and with boxing. A typical Preliminary test
match usually consists of 3 to 5 rounds (depending on the Each subject underwent a preliminary incremental exer-
category of fighters), 3 min per round, with a 1-min break cise test on a motorised treadmill (Runrace, Technogym,
between each round. `
Forlı, Italy) to assess their anaerobic threshold (AT) and
From a physiological point of view, Muay Thai appears _
maximal oxygen uptake (V O2 max). The test consisted of a
to be an intermittent physically demanding sport, with linear increase of running velocity of 2 kmÁh–1 every 3 min,
short phases of maximal or supramaximal intensity spaced starting at 6 kmÁh–1, up to exhaustion, which was considered
by brief recoveries. It is, thus, likely that both aerobic the exercise level at which the subject was unable to main-
and anaerobic energy systems are recruited during a tain the running speed (i.e., muscular fatigue).
match.
Fighting simulation test
To obtain an understanding of the physiological capacities
underlying Muay Thai performance, it would be useful to On a separate day from this preliminary test (the interval
know the energy demands and whether the anaerobic metab- was at least 3 days), each subject underwent a simulation of
olism is recruited during a match. However, while there are a Muay Thai match. To construct a fighting simulation as
studies dealing with the energy demands of some martial real as possible, the assistance of a skilled trainer, who had
arts, such as Judo, Karate, and Taekwondo (Beneke et al. been involved in national and international competitions
2004; Degoutte et al. 2003; Francescato et al. 1995; Ima- with excellent results, was enlisted. This simulation was
mura et al. 1999), to the best of our knowledge, the energy conducted in our laboratory, where a space with the same
requirement during a Muay Thai competition has never been dimensions as a Muay Thai ring was prepared. The subject
studied. This information would provide benchmarks for im- under study performed a 15-min warm-up and then rested
proving and monitoring athletes’ training. on a bench until his cardiorespiratory variables returned to
the pre-exercise level. Recovery was considered complete
This study was devised to study energy demand during a when HR was not more than 10 beatsÁmin–1 higher than
competition of Muay Thai and to test the hypothesis that pre-exercise level, and when the respiratory ratio, calculated
Muay Thai is a physically demanding activity that recruits as the carbon dioxide : oxygen uptake ratio, was less than
both aerobic and anaerobic energy systems. In particular, 0.9. The last 3 min of sitting were used to gather the resting
we were interested in measuring aerobic energy expenditure values of the variables and, after this period, the simulation
during a competition and in discovering whether and to match began.
what extent anaerobic glycolysis was recruited. This
The simulation consisted of 3 rounds, each followed by
information would be useful for coaches to design specific
1 min of recovery, during which the subject sat on a bench.
training programmes capable of inducing the specific adap-
The rounds consisted of a series of 6 attacks and 6 defensive
tations required by Muay Thai. To this end, some physio-
_ actions, each lasting 15 s, for a total duration of 180 s
logical variables, such as oxygen uptake (V O2), carbon (3 min). During the attack phases, the subject fought against
dioxide production (V _
_ CO2), pulmonary ventilation (V E),
a sparring partner, who was the aforementioned skilled
and heart rate (HR), were assessed during a simulation trainer equipped with padded arm-shields (Fig. 1). The se-
match, during which athletes wore a portable gas analyzer quence of strikes was planned ahead, and included strikes
able to measure these variables. with knees, elbows, fists, and kicks. The fighter was ver-
bally encouraged to perform maximally throughout the test.
After the recovery following the last round, 3 min of further
Materials and methods recovery was allowed (final recovery). Hence, the whole
Subjects simulation test lasted a total of 18 min: 3 min of resting be-
Ten male Muay Thai athletes (mean ± standard error of fore the beginning of the match; 3 rounds, each lasting
the mean (SEM) of age, height, and body mass were 3 min, spaced with 3 min of recovery (for a total of
23.7 ± 1.5 years, 174.3 ± 0.9 cm, and 65.1 ± 1.2 kg, re- 12 min); and 3 min of final recovery. At the end of the test,
spectively), who regularly took part in competitions in the athletes were asked to compare the effort expended during
previous 2 years, were enrolled in the study. None had any the simulation with that expended during a real Muay Thai
history of cardiac or respiratory disease or was taking any match. They gave a score ranging from 1 to 5, with 1 indi-
medication at the time of the study, and none showed any cating not similar and 5 indicating very similar.
abnormalities on physical examination or on resting elec- All experiments were conducted between 0900 and 1400
trocardiogram. Subjects were skilled athletes who trained hours in a temperature-controlled room (room temperature
for 8 to 10 h a week and had been involved in regular set at 22 8C, relative humidity at 50%). Subjects had a light
training program for at least 3 years. In the previous year, meal at least 2 h before exercising. Subjects were also asked
6 of them had participated in international competitions, to avoid caffeine and alcohol ingestion the day before tests
while the other 4 participated in matches at the national were scheduled.
level. Thus, our group represented the Muay Thai fighter
at the middle–upper level. The study was performed ac- Variables
cording to the Declaration of Helsinki and was approved
by a local ethics committee. All subjects gave written in- Assessment of respiratory variables and heart rate
formed consent. _ _ _
Values of V O2, V CO2, V E, and HR were obtained
Published by NRC Research Press

3. Crisafulli et al. 145
Fig. 1. (A) One of the subjects in the study wearing the portable _ _
metabolic system, which provided average of V O2, V CO2,
metabolic system (MedGraphics VO2000) while sitting on a bench _ E, and HR values throughout the test.
V
before starting the simulation. The face mask, breathing valve, har-
ness, and battery pack can be seen. The metabolic unit, which is Measurement of aerobic energy expenditure and anaerobic
placed on the back, can be seen in (B), which shows the subject glycolysis
engaged in simulated fighting. During the simulation match, the aerobic energy expendi-
ture (EE, expressed as kcalÁmin–1) was calculated with the
Weir equation (Weir 1949; Mansell and Macdonald 1990):
_ _
EE ¼ 3:941 Â V O2 þ 1:106 Â V CO2
This equation was used when the respiratory exchange ratio
(RER) was <1, while an oxygen caloric equivalent of 5.04
was used when EE became >1. In this case, it was assumed
that all aerobic energy was derived from carbohydrate oxi-
dation. To obtain an index of anaerobic glycolysis, excess
CO2 production (CO2 excess) was assessed, as follows (An-
derson and Rhodes 1989):
_ _
CO2 excess ¼ V CO2 À ðRERrest  V O2 Þ
where RERrest is the respiratory exchange ratio at rest, and
CO2 excess represents an index of lactic acid and H+ accumu-
lation, since, at tissue pH, lactic acid dissociates and pro-
duces H+, which is buffered by –HCO3 and other cell
buffers. The amount being buffered by –HCO3 leads to
H2CO3 production, then to H2O and CO2 (Beaver et al.
1986b; Hirakoba et al. 1993). In this way, CO2 excess is pro-
duced and is superimposed on the CO2 normally derived
from aerobic metabolism. Actually, CO2 excess was found to
correlate well with the rate of lactate accumulation in the
blood during exercise and with the anaerobic capacity (Hir-
akoba et al. 1993, 1996; Yano et al. 2002). The start of the
lactate increase and CO2 excess were found to have good in-
tercorrelation, even though the interindividual prediction of
lactate concentrations from CO2 excess is not straightforward
(Roeker et al. 2000). This parameter has been recently uti-
lized to assess the rate of anaerobic glycolysis during var-
ious kinds of exercise, including field testing and training
sessions involving dynamic phases and recoveries (Crisafulli
et al. 2002, 2006a, 2006b).
Statistical analysis
throughout the preliminary and the simulation tests, by Data were averaged for 3 min during the rest period be-
means of a portable metabolic system (MedGraphics fore the simulation match, during rounds, and during recov-
VO2000, St. Paul, Minn.), which provides a 3-breath aver- ery after the simulation, while a 1-min average was
age of variables through telemetric transmission. This sys- employed for the recovery periods between rounds. In this
tem has been shown to be reliable and to have good way, information about the time course of studied variables
agreement with a standard metabolic cart for laboratory use was gathered, and differences among the various periods of
(Byard and Dengel 2002; Olson et al. 2003). The device the protocol were detected. Responses are reported as
weighs about 1.2 kg and includes the metabolic unit, battery means ± SEM. Comparisons between periods were per-
pack, harness, chest belt for HR monitoring, face mask, and formed using the repeated measures analysis of variance
breathing valve. It is worn on the subject’s chest with a har- (ANOVA), followed by Neuman–Keuls post hoc, when ap-
ness, and does not limit the athlete’s movements. Prior to propriate. Significance was set at a p value of < 0.05. De-
testing, the VO2000 was calibrated according to manufac- scriptive statistics were performed on each variable before
turer’s instructions. During the incremental test, AT was de- the ANOVA to confirm the assumptions of normality by
termined using the V-slope method, which detects AT using means of the Kolmogorov–Smirnov test. The a level was
_
computerized regression analysis of the slopes the of V CO2 set at p < 0.05. Statistics were calculated with a commer-
_
vs. V O2 plot during exercise (Beaver et al. 1986a), while cially available software (Graph-Pad Prism).
_ _
V O2 max was calculated as the average V O2 during the last
30 s of the exercise test. Results
During the simulation test, subjects wore the portable All subjects completed the study protocol. Table 1 shows
Published by NRC Research Press

4. 146 Appl. Physiol. Nutr. Metab. Vol. 34, 2009
Table 1. Mean group values ± standard error of _
Fig. 4. Group values of HR (A) and V O2 (B) during the various
the mean (SEM) of maximum oxygen uptake (ex- periods of the simulation. A horizontal dotted line identifies the le-
pressed as absolute and indexed by body mass va- vel of anaerobic threshold. Values are means ± SEM (n = 10).
lues), maximum heart rate, oxygen uptake at *, p < 0.05 vs. rest; {, p < 0.05 vs. final recovery.
anaerobic threshold, and HR at anaerobic threshold
reached by subjects during the preliminary incre-
mental test.
Parameter Mean SEM
_
V O2 max (mLÁminÁkg–1) 48.52 1.7
_
V O2 max (mLÁmin) 3158.6 102.4
HRmax (beatsÁmin–1) 182.9 1.6
_
V O2 AT (mLÁminÁkg–1) 30.8 1.6
_
V O2 AT (mLÁmin) 2024.6 101.6
HRAT (beatsÁmin–1) 137.5 4.5
_
Note: V O2 max, oxygen uptake; HRmax, maximum heart
_
rate; V O2 AT, oxygen uptake at anaerobic threshold;
HRAT, heart rate at anaerobic threshold.
Fig. 2. Example of time course of heart rate (HR) and pulmonary
_
ventilation (V E) of 1 subject during the simulated match.
col, while Table 2 shows mean values of variables during
the active phases of the match (excluding recoveries).
HR (Fig. 4A) increased during the simulated fighting, in
comparison with rest. This HR elevation was present during
the whole test, including the recovery phases between
_
Fig. 3. Example of time course of oxygen uptake (V O2) and carbon rounds and the 3 min of final recovery after the test. This
dioxide production (V_ CO2) of 1 subject during the simulated occurrence means that the resting periods between rounds
match. did not allow complete recovery. Moreover, it is noteworthy
that HR was above the value of AT assessed during the pre-
liminary incremental test for the entire period of the simula-
tion.
_
Similarly, V O2 (Fig. 4B) rose during the match, compared
with rest, reaching values well above the AT, with no signif-
icant difference between rounds and recoveries. However,
contrary to what was described for HR, during the period of
_
final recovery, V O2 returned to values no different from
_
rest. A very similar behaviour was shown by V CO2 and V E _
(Fig. 5A and 5B, respectively), which increased throughout
the test but returned to baseline during the period of final
recovery.
Figure 6A shows the EE time course, which was obvi-
_
ously very similar to that of V O2. It should be noted that
the EE during the 9 min of the simulation match (i.e., ex-
the results of the incremental preliminary test. Figures 2 and cluding recoveries) was, on average, 13.94 ± 0.7 kcalÁmin–1
_ _ _
3 exhibit an example of HR, V E, V O2, and V CO2 time (or 0.21 ± 0.01 kcalÁmin–1Ákg–1), which corresponds to
course in 1 subject during the simulation match. Figures 4–6 12.15 ± 0.64 metabolic equivalents (METs; Fig. 6B), while
depict results of the ANOVA test applied to the mean during the whole test (i.e., including recoveries), this param-
value of variables during the various periods of the proto- eter was, on average, 10.75 ± 1.58 kcalÁmin–1 (or 9.39 ±
Published by NRC Research Press

5. Crisafulli et al. 147
_ _
Fig. 5. Group values of V CO2 (A) and V E (B) during the various Fig. 6. Group values of aerobic energy expenditure (EE), expressed
periods of the simulation. Values are means ± SEM (n = 10). *, p < as kcalÁmin–1 (A) and as metabolic equivalents (METs) (B), during
0.05 vs. rest; {, p < 0.05 vs. final recovery. the various periods of the simulation. (C) Time course of excess of
carbon dioxide production (CO2 excess). Values are means ± SEM
(n = 10). *, p < 0.05 vs. rest; {, p < 0.05 vs. final recovery; {, p <
0.05 vs. recovery 1.
1.38 METs). Figure 6C depicts the behaviour of CO2 excess.
This variable showed an abrupt increase in the first round,
and reached its maximum during the first recovery min be-
tween rounds, when it reached a value of 636 ±
66.5 mLÁmin–1. CO2 excess then gradually decreased, even
though it never returned to baseline.
Finally, as far as the similarity of the simulation to a real
match was concerned, the mean score given by athletes was
4.1 ± 0.3 (with 5 being very similar).
Discussion
This study aimed at characterizing the energetic require-
ments during a typical Muay Thai match. According to the
initial hypothesis, from our data, it appears that Muay Thai
is a physically demanding activity that recruits both aerobic 1 min was probably not sufficient to recover from the effort
and anaerobic energy systems. This finding is in accordance made during the previous round.
with what has been found in studies dealing with the energy The suggestion that Muay Thai is physically demanding
demands of other martial arts (Beneke et al. 2004; Frances- also emerges from the analysis of the EE; during the whole
cato et al. 1995). On average, during the whole simulation, simulation, which lasted 18 min, EE was, on average,
_
including both active phases and recoveries, V O2 and HR 10.75 ± 1.58 kcalÁmin–1, which corresponds to 9.39 ±
were above the values of AT previously assessed, and ap- 1.38 METs, while during the 9 min of the 3 rounds, it was,
_ _
proached the level of V O2 max. Also, V E greatly increased, on average, 13.94 ± 0.7 kcalÁmin–1, (or 12.15 ± 0.64 METs).
reaching, on average, the maximum value of 117.5 ± The sixth edition of the American College of Sports Medi-
12.7 LÁmin–1 during the second round, with peaks in some cine (2000) guidelines for exercise testing and prescription
subjects that reached 200 LÁmin–1. It is noteworthy that reports that the aerobic requirements for ring boxing and
even during the recovery periods between rounds, the phys- Judo correspond to 13.3 and 13.5 METs, respectively. These
iological variables did not decrease to resting values. Thus, values are very similar to what we found for Muay Thai,
Published by NRC Research Press

6. 148 Appl. Physiol. Nutr. Metab. Vol. 34, 2009
Table 2. Mean group values ± SEM of heart rate, oxygen uptake, carbon di-
oxide production, pulmonary ventilation, energy expenditure (expressed as
kcalÁmin–1 and as METs), and carbon dioxide excess during the 3 rounds (i.e.,
excluding recovery phases) of the simulation match.
Round 1 Round 2 Round 3
HR (beatsÁmin–1) 159.7±13.7 165.2±16.4 174±10.9
_
V O2 (mLÁmin–1) 2526.5±112.5 2927.5±185.2 2912.7±125.4
_
V CO2 (mLÁmin–1) 2685±122.9 3166.4±178.6 2939.1±79.1
_
V E (LÁmin–1) 90.5±8.3 117.5±12.6 110±8.4
EE (kcalÁmin–1) 12.6±0.5 14.6±0.9 14.5±0.6
EE (METs) 10.9±0.4 12.7±0.8 12.7±0.6
CO2 excess (mLÁmin–1) 307.3±77.5 405.8±95.5 195.7±93.6
_
Note: For statistical results, see figures. HR, heart rate; V O2, oxygen uptake;
_ _
V CO2, carbon dioxide production; V E, pulmonary ventilation; EE, energy expendi-
ture; CO2 excess, carbon dioxide excess.
suggesting that these fighting activities have similar meta- and the following recovery, whereas, during the remaining
bolic requirements. time, there was a progressive reduction in its utilization.
It is to be noted that this EE very likely underestimated This is in accordance with previous findings showing that,
the real energy requirement, since it did not take into ac- during intermittent maximal bouts of exercise, the EE of the
count the energy derived from the anaerobic metabolism. first bout is derived mainly from phosphocreatine degrada-
Actually, even the anaerobic lactacid metabolism seems to tion and anaerobic glycolysis, while, during the latter stages
have been widely recruited, as can be seen by the high level of exercise, there is a significant shift to aerobic metabolism
of CO2 excess reached in athletes, especially during the first and a reduced anaerobic energy yield (Bogdanis et al. 1996;
recovery between rounds (Fig. 6C). This respiratory index Gaitanos et al. 1993). Thus, it appears that in our simulation
has been found to correlate well with the rate of lactate ac- test, after an initial burst of anaerobic metabolism, there was
cumulation in the blood and the anaerobic capacity during a progressive increase in the aerobic energy supply. This
exercise (Hirakoba et al. 1993, 1996; Volkov et al. 1975; suggestion can also be seen in Table 2, which shows that
Yano et al. 2002). Thus, its assessment allows continuous _
V O2 was higher during rounds 2 and 3 than during round 1,
measuring of the recruitment of anaerobic lactacid metabo- even though statistics applied to the overall protocol phases
lism during exercise, without requiring the athlete to stop so (i.e., including rest and recoveries) did not find any differ-
that blood can be drawn. This parameter has been recently ence among these conditions, probably because the number
used in various kinds of efforts to detect whether the lacta- of subjects enrolled was not sufficient to reach significance.
cid metabolism is involved in the exercise being performed Another finding deserving attention is that, among meas-
(Crisafulli et al. 2002, 2006a, 2006b). In our investigation, ured variables, HR was the only one that did not return to
the mean group value of CO2 excess during the active phases rest level during the 3 min of final recovery, with the note-
of the test was about 341.9 ± 51.7 mLÁmin–1, with a peak of worthy exception of CO2 excess. This means that this variable
636.2 ± 66.5 mLÁmin–1 during the first recovery between _ _
had a slower recovery time course than V O2, V CO2, and
rounds. This value is similar to what was reported in pre- _
V E. In particular, this occurrence suggests that there was a
vious investigations, where athletes performed maximal or _
sort of dissociation between HR and V O2, which caused an
even supramaximal exercise tests requiring massive recruit- increase in HR over the real metabolic engagement. A very
ment of anaerobic glycolysis (Crisafulli et al. 2002, 2006a). similar HR behaviour was described in a recent paper (Cri-
Moreover, it should be noted that during the dynamic phases safulli et al. 2006a), which reported that when a substantial
_
of the simulation fight, mean values of V O2 were above the amount of CO2 is produced, such as when the exercise is
_
level of AT, and close to 90% of V O2 max, especially during characterized by alternate phases of maximal exercise and
rounds 2 and 3 (Table 2). This high metabolic requirement recovery, HR provides overestimated values of oxygen up-
likely led to lactate generation, as suggested by previous take. Magosso and Ursino 2001 explained this phenomenon
findings examining the production of lactate in humans over by considering that carbon dioxide has a significant impact
a range of power outputs, from 25% to 250% of V O2 max _ on the systems controlling the cardiovascular apparatus, and
(Spriet et al. 2000). Taken together, these findings (i.e., ele- that hypercapnia may induce tachycardia. In our study, a
_
vated CO2 excess and V O2 constantly above the level of AT substantial elevation of CO2 excess was present throughout
_
throughout the simulation and close to 90% V O2 max during the test, explaining the HR behaviour. Furthermore, maximal
the dynamic phases) strengthen the concept that during a and supramaximal bouts of exercise have a profound impact
Muay Thai match there is the recruitment of lactacid ca- on cardiovascular homeostasis, since they modify cardiac
pacity (i.e., the capacity of anaerobic glycolysis to resynthe- preload, afterload, and contractility, which stress the cardio-
sise ATP). vascular regulatory systems and induce compensatory tachy-
From the results of this work, it appears that anaerobic cardia (Crisafulli et al. 2004, 2006c). Hence, our study
glycolysis was recruited especially during the first round supports the concept that the use of HR monitoring to assess
Published by NRC Research Press

7. Crisafulli et al. 149
the intensity of exercise may be unreliable in activities that in- method is based on 2 assumptions: that the proteins oxida-
volve repeated bouts of maximal and supramaximal exercise tion during exercise is negligible; and that when RER be-
and lead to a massive recruitment of anaerobic glycolysis. comes >1, only carbohydrates are being oxidized. Both
Yet, other factors, such as heat stress and dehydration, assumptions are clearly wrong, since a slight quantity of
may have caused disproportionate HR elevation in relation proteins is oxidized during exercise and a RER > 1 does
to metabolic stress (Gilman 1996). All these factors (i.e., not necessarily mean that fat oxidation is not occurring, as
CO2 excess, cardiovascular stress caused by modifications in lactate accumulation and the consequent CO2 excess genera-
preload and afterload, heat stress, and dehydration) may ex- tion lead to an overestimation of the actual RER. Thus, there
_
plain the noticed dissociation between HR and V O2 during is considerable uncertainty when assessing substrate oxida-
the final recovery period, and suggest that caution should tion rates in vivo from gas exchange (Frayn 1983). How-
be used in drawing conclusions about the intensity of an ef- ever, the potential error in assessing EE with this method is
fort from HR. not wide, and was calculated within 2.5% (Mansell and
MacDonald 1990). Furthermore, it should be considered
Limitations of the study that there is not a reliable alternative method for estimating
One possible limitation of our study is that it did not ana- EE during exercise, especially during field tests. A final
lyze a real Muay Thai match but a simulation. However, in- consideration is the fact that the exercise protocol did not
asmuch as portable gas exchange analyzers are not allowed control for the effect of fatigue on power output being per-
in official competitions, it is impossible to measure variables formed. For instance, whether or not fatigue was impairing
during a real fight. Therefore, a simulation test was set with performance during the final round, compared with the first
the assistance of a skilled trainer. As testified by the mean round, was not controlled for. This is a clear limitation of all
score given by the fighters enrolled in the study, the effort field studies not conducted in the laboratory setting, where it
made during this simulation was similar to that experienced is possible to obtain physiological-biomechanical indexes of
during a real match. Hence, it is conceivable that the simu- fatigue. Nevertheless, it is likely that athletes performed
lation resembled a typical Muay Thai competition. The maximally during the third round, as can be argued by
number of subjects enrolled was only 10 because of the dif- _
Fig. 4 and Table 2; neither HR nor V O2 decreased during
ficulty of recruiting Muay Thai fighters who met the inclu- round 3, compared with the other 2 rounds. However, it can
sion criteria. However, subjects appeared to be very not be excluded that power output was lower (i.e., that effi-
homogenous in terms of age, height, body mass, and train- ciency decreased) in the last round.
ing level. Therefore, it is conceivable that the number of In conclusion, these data suggest that Muay Thai is a
subjects enrolled represent the typical Muay Thai fighter at physically demanding sport with great involvement of both
the middle–upper level. Moreover, other studies dealing aerobic metabolism and anaerobic glycolysis. This leads us
with martial arts employed the same or fewer subjects to speculate that training protocols should include exercise
(Beneke et al. 2004; Francescato et al. 1995). Another po- that train this metabolic pathway. Moreover, interval periods
tential limitation is the use of CO2 excess as a measure of between rounds do not allow a complete recovery. Coaches
blood lactate accumulation. The relationship between this should consider these suggestions when preparing the train-
parameter and blood lactate has been investigated several ing program of athletes.
times, and some studies found a good correlation (Volkov
et al. 1975; Hirakoba et al. 1993, 1996; Yano et al. 2002), Acknowledgements
while others did not (Roeker et al. 2000). In particular, This study was supported by the University of Cagliari,
Roeker et al. (2000) concluded that ‘‘the start of the lactate the Italian Ministry of Scientific Research, and PRISMA
increase and excess-CO2 showed good intercorrelation,’’ Onlus.
even though ‘‘an inter-individual prediction of lactate con-
centrations from the excess-CO2 would be difficult.’’ They References
also thought that ‘‘this parameter could be more a measure
for the formation rate of new lactate than the blood lactate American College of Sports Medicine (ACSM). 2000. General
concentration alone.’’ Our study was not intended to indi- principles of exercise prescription. In ACSM’s guidelines for ex-
rectly assess blood lactate level; rather, it was devised to ercise testing and prescription. 6th ed. Lippincott Williams &
gather qualitative information (i.e., whether the anaerobic Wilkins, Philidelphia, Penn. pp. 152–153.
lactacid metabolism was recruited or not). To the best of Anderson, G.S., and Rhodes, E.C. 1989. A review of blood lactate
our knowledge, no studies have questioned the concept that and ventilatory methods of detecting transition thresholds.
Sports Med. 8(1): 43–55. doi:10.2165/00007256-198908010-
CO2 excess during exercise qualitatively reflects the recruit-
00005. PMID:2675254.
ment of anaerobic glicolysys and capacity. The only concern
Beaver, W.L., Wasserman, K., and Whipp, B.J. 1986a. A new
is whether it is possible to extrapolate a blood lactate value method for detecting anaerobic threshold by gas exchange. J.
from CO2 excess. Therefore, it is likely that the use of this in- Appl. Physiol. 60(6): 2020–2027. PMID:3087938.
direct index, on which the conclusions of our study depend, Beaver, W.L., Wasserman, K., and Whipp, B.J. 1986b. Bicarbonate
did not influence the outcome, since the measurement of ab- buffering of lactic acid generated during exercise. J. Appl. Phy-
solute values of blood lactate was not essential to detect the siol. 60(2): 472–478. PMID:3949651.
involvement of anaerobic glycolysis in the exercise per- ¨
Beneke, R., Beyer, T., Jachner, C., Erasmus, J., and Hutler, M.
formed. 2004. Energetics of karate kumite. Eur. J. Appl. Physiol. 92(4–
A further limitation of this study may be the use of indi- 5): 518–523. doi:10.1007/s00421-004-1073-x. PMID:15138826.
rect calorimetry and the Weir equation to assess EE. This Bogdanis, G.C., Nevill, M.E., Boobis, L.H., and Lakomy, H.K.A.
Published by NRC Research Press

8. 150 Appl. Physiol. Nutr. Metab. Vol. 34, 2009
1996. Contribution of phosphocreatine and aerobic metabolism Hirakoba, K., Maruyama, A., and Misaka, K. 1993. Effect of acute
to energy supply during repeated sprint exercise. J. Appl. Phy- sodium bicarbonate ingestion on excess CO2 output during in-
siol. 80(3): 876–884. PMID:8964751. cremental exercise. Eur. J. Appl. Physiol. Occup. Physiol.
Byard, A.D., and Dengel, D.R. 2002. Validity of a portable meta- 66(6): 536–541. doi:10.1007/BF00634306. PMID:8394808.
bolic measurement system. Med. Sci. Sports Exerc. 34(5): Hirakoba, K., Maruyama, A., and Misaka, K. 1996. Prediction of
S149. doi:10.1097/00005768-200205001-00829. blood lactate accumulation from excess CO2 output during con-
Crisafulli, A., Melis, F., Tocco, F., Laconi, P., Lai, C., and Concu, stant exercise. Appl. Human Sci. 15(5): 205–210. doi:10.2114/
A. 2002. External mechanical work versus oxidative energy con- jpa.15.205. PMID:8979401.
sumption ratio during a basketball field test. J. Sports Med. Imamura, H., Yoshimura, Y., Nishimura, S., Nakazawa, A.T.,
Phys. Fitness, 42(4): 409–417. PMID:12391434. Nishimura, C., and Shirota, T. 1999. Oxygen uptake, heart rate,
Crisafulli A., Carta, C., Melis, F., Tocco, F., Frongia, F., Santoboni, and blood lactate responses during and following karate training.
P., et al. 2004. Hemodynamic responses following intermittent Med. Sci. Sports Exerc. 31(2): 342–347. doi:10.1097/00005768-
supramaximal exercise in athletes. Exp. Physiol. 89(6): 665– 199902000-00019. PMID:10063825.
674. doi:10.1113/expphysiol.2004.027946. PMID:15328308. Magosso, E., and Ursino, M. 2001. A mathematical model of CO2
Crisafulli, A., Pittau, G.L., Lorrai, L., Cominu, M., Tocco, F., Melis, effect on cardiovascular regulation. Am. J. Physiol. Heart Circ.
F., et al. 2006a. Poor reliability of heart rate monitoring to assess Physiol. 281(5): H2036–H2052. PMID:11668065.
oxygen consumption during field training. Int. J. Sports Med. Mansell, P.I., and Macdonald, I.A. 1990. Reappraisal of the Weir
27(1): 55–59. doi:10.1055/s-2005-837504. PMID:16388443. equation for calculation of metabolic rate. Am. J. Physiol.
Crisafulli, A., Salis, E., Pittau, G., Lorrai, L., Tocco, F., Melis, F., 258(6 Pt 2): R1347–R1354. PMID:2360685.
et al. 2006b. Modulation of cardiac contractility by muscle me- Olson, T.P., Tracy, J.E., and Dengel, D.R. 2003. Validity of a low-
taboreflex following efforts of different intensities in humans. flow pneumotach and portable metabolic system for measure-
Am. J. Physiol. Heart Circ. Physiol. 291(6): H3035–H3042. ment of basal metabolic rate. Med. Sci. Sports Exerc. 35(5):
doi:10.1152/ajpheart.00221.2006. PMID:16782848. S143.
Crisafulli, A., Tocco, F., Pittau, G.L., Lorrai, L., Porru, C., Salis, Roeker, K., Mayer, F., Striegel, H., and Dixkhuth, H.H. 2000. In-
E., et al. 2006c. Effect of differences in post-exercise lactate ac- crease characteristics of the cumulated excess-CO2 and the lac-
cumulation in athletes’ hemodynamics. Appl. Physiol. Nutr. Me- tate concentration during exercise. Int. J. Sports Med. 21(6):
tab. 31(4): 423–431. doi:10.1139/H06-017. PMID:16900232. 419–423. doi:10.1055/s-2000-3836. PMID:10961517.
Degoutte, F., Jouanel, P., and Filaire, E. 2003. Energy demands Spriet, L.L., Howlett, R.A., and Heigenhauser, G.J.F. 2000. An en-
during a judo match and recovery. Br. J. Sports Med. 37(3): zymatic approach to lactate production in human skeletal muscle
245–249. doi:10.1136/bjsm.37.3.245. PMID:12782550. during exercise. Med. Sci. Sports Exerc. 32(4): 756–763. doi:10.
Francescato, M.P., Talon, T., and di Prampero, P.E. 1995. Energy cost 1097/00005768-200004000-00007. PMID:10776894.
and energy sources in karate. Eur. J. Appl. Physiol. Occup. Physiol. Volkov, N.I., Shirkovets, E.A., and Borilkevich, V.E. 1975. Assess-
71(4): 355–361. doi:10.1007/BF00240417. PMID:8549580. ment of aerobic and anaerobic capacity of athletes in treadmill
Frayn, K.N. 1983. Calculation of substrate oxidation rates in vivo running tests. Eur. J. Appl. Physiol. Occup. Physiol. 34: 121–
from gaseous exchange. J. Appl. Physiol. 55(2): 628–634. 130. doi:10.1007/BF00999924. PMID:1193088.
PMID:6618956. Weir, J.B. 1949. New methods for calculating metabolic rate with
Gaitanos, G.C., Williams, C., Boobis, L.H., and Brooks, S. 1993. special reference to protein metabolism. J. Physiol. 109: 1–9.
Human muscle metabolism during intermittent maximal exer- PMID:15394301.
cise. J. Appl. Physiol. 75(2): 712–719. PMID:8226473. Yano, T., Horiuchi, M., Yunoki, T., and Ogata, H. 2002. Kinetics
Gilman, M.B. 1996. The use of heart rate to monitor the intensity of CO2 excessive expiration in constant-load exercise. J. Sports
of endurance training. Sports Med. 21(2): 73–79. Med. Phys. Fitness, 42(2): 152–157. PMID:12032409.
PMID:8775514.
Published by NRC Research Press