Nutritional management studies to improve the performance of Mangalarga Marchador (MM) horses during the marcha test are limited. This study was designed to test the hypothesis that chelated trivalent chromium (CR) feed supplementation may reduce the suitability of the length of the interval between concentrate feeding and the marcha test among MM horses. A total of 12 healthy mares (4.25 ^ 0.62 years) were randomly assigned to one of six dietary treatments (0 or 10 mg Cr by concentrate, fed 0.5, 2 or 4 h before exercise), according to a completely randomized design, with a split-plot arrangement. The diet was Cynodon pasture and concentrate (50:50 ratio). The first 29 days of the trial were for diet, Cr and exercise adaptation; during the next 15 days, horses were submitted to three 50-min field marcha tests, once a week. Heart rate (HR) was measured before, during and until 25 min after the exercise. Respiratory rate and rectal temperature were measured; blood samples were collected before, at the end and 25 min after the test. There was no effect of Cr by concentrate feeding strategy on any physiological variables (P . 0.05). Supplementation of Cr increased glycaemia before and soon after the second marcha test (P , 0.01). In addition, Cr reduced HR during the second marcha test and decreased the time to first post-exercise HR recovery (P , 0.05). Insulinaemia was greater when the concentrate was provided 2 h prior to the test (P , 0.05). Concentrate provided 0.5 and 2 h before the test reduced plasma triacylglycerol in the first and second tests, respectively. The interval between concentrate feeding and marcha tests should not be decreased in horses supplemented with Cr. Horses should be fed more than 2 h before that test. Cr supplementation during training may improve the cardiac performance of MM mares during the marcha test.

1. doi:10.1017/S1755254011000043
Comparative Exercise Physiology 7(3); 133–140
q Cambridge University Press 2011
Effect of feed at different times prior
to exercise and chelated chromium
supplementation on the athletic
performance of Mangalarga
Marchador mares
´
´
Lilian de Rezende Jordao1, Jose Aurelio G Bergmann1,
˜
´
Raquel S Moura1, Marılia M Melo2, Maria LL Costa1,
´
´
Patrıcia CB Moss1, Helio M Aquino Neto3
and Adalgiza Souza Carneiro de Rezende1,*
1
Department of Animal Science, School of Veterinary Medicine, Universidade Federal
de Minas Gerais, Belo Horizonte, MG, Brazil
2
Department of Veterinary Surgery, School of Veterinary Medicine, Universidade
Federal de Minas Gerais, Belo Horizonte, MG, Brazil
3
Arapiraca Campus – Vicosa, College of Veterinary Medicine, Universidade Federal
¸
´
de Alagoas, Maceio, AL, Brazil
* Corresponding author: adalgizavetufmg@gmail.com
Submitted 4 July 2010: Accepted 23 January 2011 – First published online 14 March 2011
Research Paper
Abstract
Nutritional management studies to improve the performance of Mangalarga Marchador (MM) horses during the
marcha test are limited. This study was designed to test the hypothesis that chelated trivalent chromium (CR)
feed supplementation may reduce the suitability of the length of the interval between concentrate feeding and
the marcha test among MM horses. A total of 12 healthy mares (4.25 ^ 0.62 years) were randomly assigned
to one of six dietary treatments (0 or 10 mg Cr by concentrate, fed 0.5, 2 or 4 h before exercise), according to
a completely randomized design, with a split-plot arrangement. The diet was Cynodon pasture and concentrate
(50:50 ratio). The first 29 days of the trial were for diet, Cr and exercise adaptation; during the next 15 days,
horses were submitted to three 50-min field marcha tests, once a week. Heart rate (HR) was measured before,
during and until 25 min after the exercise. Respiratory rate and rectal temperature were measured; blood samples
were collected before, at the end and 25 min after the test. There was no effect of Cr by concentrate feeding strategy on any physiological variables (P . 0.05). Supplementation of Cr increased glycaemia before and soon after
the second marcha test (P , 0.01). In addition, Cr reduced HR during the second marcha test and decreased
the time to first post-exercise HR recovery (P , 0.05). Insulinaemia was greater when the concentrate was
provided 2 h prior to the test (P , 0.05). Concentrate provided 0.5 and 2 h before the test reduced plasma triacylglycerol in the first and second tests, respectively. The interval between concentrate feeding and marcha tests
should not be decreased in horses supplemented with Cr. Horses should be fed more than 2 h before that test. Cr
supplementation during training may improve the cardiac performance of MM mares during the marcha test.
Keywords: carbohydrate feeding; horse; insulin; marcha test; micromineral
Introduction
The Mangalarga Marchador (MM) horse is the most
important Brazilian horse breed, and its natural gait is
marcha, rather than trot. The marcha is a comfortable
four-beat lateral and diagonal gait with moments of
triple support and no suspension. The MM horse has
been used for a specific functional trial, called the
marcha test. This test was characterized as an equestrian submaximal intensity competition1 of approximately 50 min with no rest and an average speed of
3.3 m s21. It is a standard evaluation procedure for

2. 134
LR Jordao et al.
˜
Table 1 Nutritional composition of feed ingredients on a DM basis
Feed
DM (%)
Forage
Concentratei
Digestible energy
(Mcal kg21)
CP (%)
Lysine (%)
iii
NDF (%)
34.93
90.49
2.25
3.29ii
8.06
15.32
0.22
0.78iii
68.32
26.06
ADF (%)
Ca (%)
P (%)
Ca:P
Cr (mg kg21)
33.61
6.47
0.21
1.73
0.23
0.63
0.91:1
2.75:1
18.24
9.24
Forage
Concentratei
i
Corn (61%), soyabean meal (19%), wheat bran (15%), limestone (3%), dicalcium phosphate (1%) and mineral salt without Cr
´
´
(1%) (Coequi Plus; Tortuga Cia. Zootecnica Agraria) (basic composition per kg: Na, 120 g; Ca, 185 g; P, 60 g; K, 20 g). ii In addition
to the 0.368 Mcal provided by 50 ml of soyabean oil added to the concentrate daily. iii From Valadares Filho32. DM – dry matter;
CP – crude protein; NDF – neutral detergent fibre; ADF – acid detergent fibre; Ca – calcium; P – phosphorus; Cr – chromium.
most of the official breed championships. The training
and nutritional management carried out in Brazilian
farms are still quite empirical. It is necessary to investigate procedures that have currently been used for
marcha tests to propose protocols for animal training.
Moreover, the proportion of different types of muscle
fibres in the MM horses has not been defined.
Ralston2 contraindicated soluble carbohydrate
supplies to horses with less than 4 h before the
prolonged exercise. The resulting hyperinsulinaemia
may negatively interfere with lipid mobilization and
lead to fatigue3. However, a Cr-enriched diet of MM
horses submitted to the marcha test may facilitate
the use of alternative energy sources and thereby
reduce carbohydrate oxidation. As part of the chromodulin, Cr may potentiate the effects of insulin (INS)4.
Therefore, the purpose of this study was to measure
the effect of three different intervals between concentrate feeding and the beginning of the marcha test and
Cr feed supplementation on biochemical and clinical
responses in MM horses during and after exercise.
Materials and methods
Horses
The experiment was carried out at Haras Catuni
(latitude, 168410 16100 S; longitude, 438310 21000 W; altitude, 784 m), Montes Claros, Minas Gerais, Brazil,
from January to February 2008. A total of 12 healthy
MM mares, offspring of two stallions (mean ^ SD),
4.25 ^ 0.62 years old, 339.40 ^ 23.10 kg body
weight (BWT) and with body condition scoring5
(CS) 2–3 (range 1–5), were used. Prior to the experiment the animals had received no training related
to the marcha test. They were kept in an 8.3 ha
paddock, continuously grazing Coast cross (Cynodon
dactylon £ Cynodon nlemfluensis) with water and
mineral supplements ad libitum. In addition, concentrate (1.25% BWT day21) was offered subsequent to
moderate work6 at 06.00 and 18.00 hours (Table 1).
BWT and CS were measured weekly. All mares were
treated at the beginning of the pre-experimental
phase for internal and external parasites with 2%
´
´
ivermectin (Altec; Tortuga Cia. Zootecnica Agraria,
Sao Paulo, Brazil) and deltamethrin (Butoxw P CE 25;
˜
Intervet/Shering-Plough Brasil, Brazil), respectively.
The Ethics Committee in Animal Experimentation
(CETEA/UFMG) approved the experimental protocol.
Experimental design
Each of the 12 mares was assigned to one of two
groups: Cr (10 mg chelated trivalent Cr capsules;
´
´
Tortuga Cia. Zootecnica Agraria) and CON (no supplemental Cr), in a completely randomized design,
with a split-plot arrangement. The capsules were
orally administered daily, at 18.00 hours just before
concentrate feeding. The distribution of the animals
within the two experimental groups was based on
the following criteria: each group included two
mares, daughters of the same sire, of similar age,
BWT, body CS, mean heart rate for marcha (HRM)
and HR recovery index (D%). Each pair was assigned
for CON or Cr supplementation. Within each group,
two mares were randomly assigned to feed concentrate (0.63 kg 100 kg21 BWT) 0.5, 2 or 4 h before the
marcha test. HRM and D% were observed after a
period of 13 min in a continuous session divided into
the following: 4 min walking, 4 min marching, 2 min
walking and 3 min standing at rest, the latter two
being the recovery period. D% was calculated as
½ðHRM 2 HRR5 Þ=HRM Š £ 100, where HRR5 is the HR at
the end of the recovery period. The first 29 days
(pre-experimental phase) were used for diet, Cr and
exercise adaptation, and the last 15 days were the
experimental phase. During the experimental phase,
the animals were submitted to three 50 min field
marcha tests once a week, in the 1st, 8th and 15th
days of this phase, according to the XXVI Mangalarga
Marchador National Exhibition regulations7.
The adaptation of the animals to simulate the
marcha competition1,8 was carried out on an oval
track (0% slope; approximately 20 m £ 50 m). During
the experimental phase, the ridden animals were submitted to a 5 min walk for warm-up (average speed
1.6 m s21), followed by a 50 min cadenced marcha
(25 min in each direction and average speed

3. Feed prior to exercise and Cr supplementation
135
Table 2 Methods used to determine concentrations of the analytes
investigated in this study
Analyte
Sample
Method
CORT
GLU
GLY
INS
LACT
TG
Serum
Whole blood
Heparinized plasma
Serum
Whole blood
Heparinized plasma
Radioimmunoassay, solid phase
Amperometry, GLU dehydrogenase
Enzymatic colourimetric assay
Radioimmunoassay, solid phase
Amperometry, LACT oxidase
Enzymatic colourimetric assay
3.33 m s21). The marcha tests were performed at
10.00 and 16.00 hours each day, with six animals in
each period. In each period, there was a mare from
each treatment group (Cr by concentrate feeding).
Thus, two mares were used for each concentrate feeding schedule, one supplemented with Cr and the other
with placebo.
The mares remained in the paddock without any
athletic activity 1 day before and 1 day after the
marcha test. In the remaining days of this phase,
they were trained every other day, as in the protocol
adapted in the pre-experimental phase.
´
´mica Basica, Minas Gerais,
(GLY; Bioclin Quibasa Quı
´
´quido
Brazil), and triacylglycerol (TG; triglicerides lı
´
´
´mica Basica)
estavel kit, K055, Bioclin Quibasa Quı
levels were measured BT, AT and 25 min AT (Table 2).
In addition, the INS:GLU ratio was calculated. The
blood samples were collected via the jugular vein in
vacuum tubes using no anticoagulant or heparin.
After the analyses of blood GLU and LACT, samples
were maintained in a cooler with ice and centrifuged
on the same day at 3000 rpm for 20 min. The supernatants were frozen at 2 808C with liquid nitrogen
until processing.
Diet analyses
Feed samples were collected (forage being handplucked9) to determine the chemical composition:
dry matter (DM), crude protein (CP), gross energy
(GE), Ca, P, Cr (measured by atomic absorption spectroscopy10), neutral detergent fibre (NDF) and acid
detergent fibre11 (ADF). Faeces were analysed for DM
and GE concentrations10.
Statistics
The effects of treatments were analysed by ANOVA
and the means compared by F-test using SAEG version
9.1 software (Universidade Federal de Vicosa, Minas
¸
Gerais, Brazil). Data subsets were created to
analyse the effect of the treatments on each marcha
test day. Post hoc comparisons were made, where
appropriate, using the Fisher least significant difference test for RR and the Newman–Keuls test for
the other variables. Significance was accepted when
P , 0.05. A Lilliefors test for normality was conducted.
Logarithmic transformations to approach normality
were applied to RR and INS:GLU variables. A regression adjustment model was used to assess the effect
of time on HR. Correlations between the biochemical
and clinical variables were calculated using Pearson’s
correlation analysis.
Environmental analyses
The room temperature (T), relative air humidity (RH)
(thermo hygrometer MT-242; Minipa Industria e
´
´
Comercio, Sao Paulo, Brazil), temperature–humidity
˜
index (THI) and wet bulb globe temperature (WBGT)
index12 were measured during the days and periods
of the marcha tests. A value of THI or WBGT # 70
indicated a normal condition, not stressful; values
between 71 and 78 indicated a critical condition;
values between 79 and 83 indicated a danger condition; and values . 83 indicated an emergency
condition.
Clinical and biochemical analyses
Polar RS400SD heart rate monitors (Polar Electro Oy,
Kempele, Finland) were used to assess HR immediately
before the test (BT), during the test, soon after the
test (AT), and 5, 10, 15, 20 and 25 min after. Respiratory rate (RR), rectal temperature (RT) and blood
´
INS (BET Laboratories Endocrinologia Veterinaria, Rio
de Janeiro, Brazil), cortisol (CORT; BET Laboratories),
glucose (GLU; Accutrend GCT, Roche Diagnostics,
Mannheim, Germany) and lactate (LACT; Accutrend
Lactate, Roche Diagnostics), and plasma glycerol
Results
Environmental parameters
The WBGT index was (mean ^ SD) 76.62 ^ 3.56
during the marcha tests (critical condition12). T (8C),
RH (%) and THI were high at the time of the tests:
30.32 ^ 2.708C, 47.83 ^ 10.30% and 75.79 ^ 2.74
(critical condition12), respectively.
Biochemical parameters
Considering all three marcha tests, there was no effect
of Cr by the time of concentrate feeding prior to
exercise (FEEDING) and the moment of evaluation
(MOMENT; BT, AT or 25 min AT) on the biochemical
variables analysed (P . 0.05). However, for the three

4. 136
LR Jordao et al.
˜
CORT (ng ml–1)
GLY (mg dl–1)
RR (bpm)
RT (°C)
300
250
200
c
a
b
150
a
100
50
c
b
c
b
a
a
b
b
0
BT
AT
25 min AT
FIG . 1 Blood CORT and plasma GLY levels (P , 0.01), RR and RT (P , 0.0001) of MM mares BT, AT and 25 min AT. Mean ^ SD CV,
34.67% (CORT); 55.56% (GLY); 10.72% (RR); 1.56% (RT). bpm, breaths per min a, b, c: Means followed by different letters vary by the
Newman–Keuls test (CORT, GLY and RT) or Fisher’s least significant difference test (RR).
marcha tests, the effects of MOMENT on CORT, GLY
(Fig. 1) and LACT were significant (P , 0.01). Blood
LACT levels of AT (2.09 mmol l21) and 25 min AT
(2.29 mmol l21) were similar; however, these values
were higher (P , 0.01) than BT (1.84 mmol l21); coefficient of variance (CV) 24.63%. Nevertheless, there
was no effect of either Cr or FEEDING on the variables
GLU and TG (P . 0.05).
Table 3 presents a concentrate feeding strategy by
MOMENT interaction effects on INS (P , 0.05) and
INS:GLU (P , 0.001). The 4 h group showed the least
severe insulinaemia BT, AT and 25 min AT. In this
group, INS was lower AT and 25 min AT than BT. Interestingly, INS was higher when the concentrate was fed
2 h rather than 0.5 h prior to exercise BT, AT and
25 min AT. In both FEEDING groups, INS was higher
BT, followed by AT, reducing further 25 min AT. Also,
concentrate feeding , 2 h before the marcha test
increased INS:GLU AT and 25 min AT. INS:GLU for
the 0.5 h group remained high during the three time
points of analysis.
In the first test, the 0.5 h group showed the lowest
TG (P , 0.001; Table 4). In contrast, the 2 h group
showed the lowest TG on the second test (P , 0.05;
Table 4).
There was a Cr by MOMENT effect interaction on
GLU in the second test (P , 0.01; Table 5). Surprisingly, Cr increased glycaemia BT and AT and maintained almost constant GLU at these moments. In
contrast, the CON group showed a decline in GLU at
the end of the exercise.
In the third test, there was a FEEDING by MOMENT
effect interaction on GLU (P , 0.01; Table 4). The
group that consumed concentrate 4 h prior to the
test showed GLU levels that were almost constant
over the three moments of analysis. The 2 h group
ended the test with lower GLU that returned to
a value close to pre-exercise values by 25 min AT.
The 0.5 h group began the exercise with reduced
GLU in relation to the two moments AT. During the
test, the 0.5 h group showed a higher GLU than the
other groups.
Finally, there was a weak negative correlation between
INS and CORT after exercise (r ¼ 20.42; P , 0.02),
and the level of INS showed a weak positive correlation
with GLU only 25 min AT (r ¼ 0.44; P , 0.01).
Heart rate
The main results of the HR analysis are shown in
Figs 2a and 2b and Tables 3 and 5. The result of the
HRM test was 145 ^ 32 beats min21 (bpm). According
to the regression prediction model (Fig. 2a), the
peak HR (HRpeak) of the marcha test was 153 bpm
at 17 min (P , 0.0001). HR reached stability after
2 min, remaining almost constant until the end of the
activity period.
In accordance with the regression prediction model
of post-exercise HR recovery (Fig. 2b), there was a rapid
reduction in HR over the first 2 min, with 50% reduction
in the HRpeak 25 min after exercise (P , 0.01). However,
recovery values were higher (52 ^ 10 bpm), even at
the end of post-exercise recovery (77 bpm).
Table 3 Blood INS levels, INS:GLU ratio and post-exercise HR
recovery of MM mares fed at different times before the marcha
testsi
Feeding (h:min)
Moment
4:00
2:00
0:30
21
INS (pmol l )*
BT
AT
25 min AT
INS:GLU (pmol mg l21 dl21)***
BT
AT
25 min AT
151.89Ac 222.21Aa
88.32Bc 186.89Ba
42.48Bc 171.70Ca
1.15Ab
0.50Bb
0.29Bb
186.84Ab
159.43Bb
116.67Cb
1.94Aa
1.22Ba
0.79Ba
0.72Ab
1.36Aa
1.22Aa
Post-exercise HR recovery*
Feeding (h:min)
4:00
2:00
0:30
HR (bpm)
87B
85B
97A
* P , (0.05). *** P , (0.001). i Only the results that were statistically different are shown in the table. Feeding, time of concentrate feeding prior to
exercise. Means in rows followed by different lower case letters and in
columns by different upper case letters differ by the Newman–Keuls test.
CV, 64.79; 35.37 and 30.26%, respectively.

5. Feed prior to exercise and Cr supplementation
Table 4 Plasma TG and blood GLU levels of MM
mares from three different marcha testsi
TG (mg dl21)***
Feeding (h:min)
GLU (mg dl21)* and TG (mg dl21)* by test day
First test
4:00
2:00
0:30
41.33A
40.56A
17.48B
4:00
2:00
0:30
40.87A
24.02B
39.24A
Second test
Third test
GLU (mg dl21)**
Feeding (h:min)
4:00
2:00
0:30
BT
131.09Aa
141.63Aa
115.54Ab
AT
130.50Ba
119.50Bb
153.50Aa
25 min AT
126.00Aa
142.25Aa
136.75Aa
* P , (0.05). ** P , (0.01). *** P , (0.001). i Only the results that
were statistically different are shown in the table. Feeding, time
of concentrate feeding prior to exercise. Means in rows followed
by different lower case letters and in columns by different upper
case letters differ by the Newman–Keuls test. CV, 32.71; 49.83
and 17.80%, respectively.
There was no Cr by FEEDING by MOMENT effect
interaction on HR (test and post-exercise recovery;
P . 0.05). Also, there was no effect of the test day
(TEST) on these variables (P . 0.05). However, there
was an interaction of Cr by TEST effect on HR of
test and post-exercise recovery values (P , 0.05;
Table 5). Cr was responsible for the lowest HR in the
second test and the lowest post-exercise HR recovery
in the first test. In addition, the 0.5 h group showed
the highest level of post-exercise HR recovery
(P , 0.05; Table 3).
Respiratory rate and rectal temperature
There was no effect of Cr or FEEDING on RR and RT
(P . 0.05). Also, there was no effect of TEST on
these variables (P . 0.05). However, Fig. 1 presents
higher values of RR and RT AT, followed by 25 min
AT (P , 0.0001). There was a strong positive correlation between RR and RT (r ¼ 0.80; P , 0.0001).
137
approximately 4.86 kcal for each litre of oxygen
utilized6. Additionally, from the HRpeak of marcha and
considering the HRmax as 220 bpm13, the oxygen utilization reached 55.3% of maximum VO2 (VO2max)16.
This value suggests that the marcha test is a moderateintensity exercise (range 35–55% VO2max).
Dietary Cr supplementation has attracted attention
because this micromineral improved both the performance of exercising horses1,17 and GLU tolerance in
humans18,19, horses20 and other farm animals21,22.
It was therefore hypothesized that, in this study, Cr
would reduce the interval between the marcha test
and pre-exercise carbohydrate feeding (according to
Ralston2, 4 h). However, Cr did not interact with
FEEDING and did not prevent the adverse effects of
this nutritional management.
Unexpectedly, Cr was responsible for increased
glycaemia before and after the second test. A possible
explanation could be a blood GLU-sparing effect,
mediated by increased reliance on fat substrate
stores, based on evidence of its lipolytic effect23.
Another possibility is that Cr stimulated hepatic glycogenolysis and gluconeogenesis.
In this study, 10 mg of Cr did not reduce the RR or
RT at the end of the exercise (in contrast to the results
published by Prates24). However, Cr reduced HR
during the test and accelerated the speed of decline
in HR recovery, improving cardiac performance (as
in the study by Prates et al.1). Furthermore, Cr made
the marcha more energy efficient, because there is a
strong relationship between HR and oxygen utilization
at submaximal exercise intensities15.
Horses fed grain and subjected to prolonged
aerobic exercise may have a higher degree of Cr
Table 5 Cr effects on glycaemia and HR
of MM mares during each marcha test and
subsequent post-exercise recoveryi
Blood GLU (mg dl21) on the second day
of testing**
Moment
Cr
CON
Aa
Discussion
The HRpeak for the marcha test was lower than
the maximum HR (HRmax)13 (220–260 bpm). When
associated with the HRpeak value, the stability of the
HR values was used to characterize this exercise as
submaximal intensity (as in the study by Prates
et al.1). In addition, it may be deduced that the
marcha test is essentially aerobic, as the HR remained
close to 150 bpm and blood LACT remained close to
2 mmol l21 during exercise14.
From the HRM, the average VO2 of animals was
estimated at 66.4 ml kg21 min21 15. This value allowed
us to estimate the mean energy expenditure during
the test, assuming that each horse metabolizes
BT
132.12
AT
142.50Aa
25 min AT
125.50Aa
HR during tests (bpm)*
Cr
First
157Aa
Second
139Bb
Third
145ABa
Post-exercise HR recovery (bpm)*
Cr
First
77Bb
Second
98Aa
Third
89Aba
120.14Ab
98.50Bb
134.33Aa
CON
147Aa
159Aa
150Aa
CON
98Aa
91Aa
83Aa
* P , (0.05). ** P , (0.01). i Only the results that
were statistically different are shown in the table.
Means in rows followed by different lower case
letters and in columns by different upper case
letters differ by the Newman–Keuls test. CV, 29.68;
29.75 and 38.36%, respectively.

6. 138
LR Jordao et al.
˜
a) 180
b) 160
160
HR_est
HR_obs
140
140
120
100
HR_est
HR_obs
80
60
HR (bpm)
HR (bpm)
120
100
80
60
40
40
HR = 155.75 – (44.01/T ); R 2 = 91.85%
20
HR = 137.44 – (6.90 T + 0.18 T 2);
R 2 = 87.71%
20
0
0
0
10
20
30
Time (min)
40
50
0
10
20
30
Time (min)
FIG . 2 (a) HR of Cr and CON experimental groups during the 50 min of all three marcha tests (P , 0.0001). (b) HR during the 25 min of
all three post-exercise recoveries (P , 0.01). HR_obs – observed HR; HR_est – estimated HR. T – time (min)
deficiency due to increased excretion in the urine,
especially in situations of high stress25,26. Therefore,
the favourable effect of Cr on the marcha test can
be explained partially by the intense thermal discomfort and high thermoregulatory demand suffered
during the exercise.
Cr did not interfere with INS, CORT, lipolysis (which
would be evidenced by higher GLY and TG) or LACT.
These results are different from those published by
Pagan et al.17, in which exercising horses supplemented with 5 mg Cr yeast had higher plasma TG
levels after a standardized exercise test (SET) and
lower plasma CORT before and at the beginning of
the exercise. There was still a trend towards a decrease
in plasma INS (P , 0.10) 1 h after grain feeding and a
trend towards a decrease in plasma LACT (P ¼ 0.08) in
the final stages of the SET.
In competition, MM horses perform various tests for
about 50 min and are provided with short periods of
rest between, often under intense heat and high RH.
As the animals progress through the various competition stages, they may be allowed to compete for
two to three consecutive days. Thus, a rapid recovery
provides obvious advantages in marcha competitions.
Based on the results of this study, Cr supplementation
in the diet of MM horses submitted to the marcha test
may be considered an effective nutritional ergogenic
aid. However, Cr is neither a panacea nor a drug;
rather, it is a nutrient. The complexity of neuroendocrine regulation during exercise (inherent in the type
of exercise and food status involved) and the different
methodological and environmental conditions could
explain the variation in results recorded in the
literature.
It is possible that the intravenous administration of
GLU before exercise increases the time to fatigue
during moderate-intensity exercise27. However, in
long-duration and low-intensity exercise (. 3 h) such
as endurance racing, the concentrate intake may
increase the level of insulinaemia during exercise.
This would decrease the degree of fat utilization with
a concomitant increase in carbohydrate oxidation
during exercise and acceleration of fatigue3. The
marcha test is a moderate-intensity exercise, and it
has intermediate duration between explosion exercises
(e.g. Thoroughbred racing) and endurance. Thus, its
duration is not extensive enough for hypoglycaemia
to occur, and a marked increase in lactataemia
occurs even if concentrated meals are offered less
than 2 h before exercise. Additionally, regular exercise
may improve glycaemic CON28 and prevent hypoglycaemia caused by concentrate feeding.
Contrary to expectations, animals in the 0.5 h group
did not show an increase in INS as high as that
observed in the 2 h group. When the mares were fed
concentrate 30 min BT, there was probably insufficient
time for digestion of the ingested concentrate, resulting in the recorded blood INS levels. According to
Frape29, ingesta remain in the stomach for 15 min on
average, and their passage through the small intestine
occurs 30–90 min after food intake. Another explanation for this is that there is reduced blood flow to
the splanchnic vascular bed during exercise30, which
may impair the digestion and absorption of nutrients.
The 4 h group experienced minimal fluctuations
in GLU on the third test, which may reflect improved
homeostatic regulation on GLU in this approach to
nutritional management. In addition, the lower INS
and INS:GLU recorded in this group are in agreement
with Ralston2, who maintained that concentrate
should be offered at least 4 h before the long-duration
exercise to avoid hyperinsulinaemia. On the other
hand, the slower return of HR recovery to resting
values in the 0.5 h group is an indication of a clinical

7. Feed prior to exercise and Cr supplementation
abnormality such as severe dehydration, and is associated with fatigue13. Therefore, it is likely that this
group experienced haemodynamic changes related to
the displacement of fluid into the gastrointestinal tract
associated with excessive sweating (subjectively analysed) during the marcha test.
As the concentrate supply 0.5 or 2 h before the
marcha test increased INS and INS:GLU and had a
negative influence on lipid mobilization, we recommend
that researchers should not provide a concentrated
meal within 2 h before the test. The marcha test can
reach 2 h of duration in extraordinary conditions, and
so higher blood INS levels during the exercise may
lead to early fatigue.
Finally, CORT is considered a stress hormone and a
factor for immunosuppression. It has an antagonistic
effect on INS31, and a characteristic response was
noted in this experiment (shown by the negative correlation between INS and CORT at the end of the
test). CORT levels are increased to avoid the deleterious effects of hypoglycaemia, which can occur
during an exercise of prolonged duration.
In conclusion, this study showed that dietary Cr
did not reduce the appropriate interval between concentrate feeding and the marcha test in MM mares.
Cr supplementation during training may improve
cardiac performance during the test, and MM mares
should be fed more than 2 h BT. The HR values
characterized the marcha test as a predominantly
aerobic and moderate-intensity submaximal exercise.
139
4
5
6
7
8
9
10
11
12
13
14
15
Acknowledgements
We thank the FAPEMIG, CAPES, CNPq, Escola de Veter´
´
inaria da UFMG, Pro-Reitoria de Pesquisa da UFMG and
´
´
Tortuga Cia. Zootecnica Agraria for funding this
project. We would also like to thank Mr Joao Carlos
˜
Moreira’s family for providing animals and logistics.
In addition, we are grateful to Neimar Severo (ABS PecPlan), BET Laboratories, ABCCMangalarga Marchador
´mica
and Leonardo de Freitas (Bioclin Quibasa Quı
ˆ
´
Basica) for their support. Dr Angela Lana and Dr
Danusa Soares contributed invaluable advice with
respect to the experimental design.
16
17
18
19
20
References
1
2
3
ˆ
Prates RC, Rezende H, Lana A, Borges I, Moss P, Moura R
and Rezenda ASC (2009). Heart rate of Mangalarga Marchador mares under marcha test and supplemented with
chrome. Revista Brasileira de Zootecnia 38: 916–922.
Ralston SL (1997). Nutritional management of the horse’s
performance on race day. Technical Papers of the Veterinary School of UFMG 19: 59–68.
Jose-Cunilleras E and Hinchcliff KW (2004). Carbohydrate
metabolism in exercising horses. Equine and Comparative
Exercise Physiology 1: 23–32.
21
22
23
Pechova A and Pavlata L (2007). Chromium as an essential
nutrient: a review. Veterinarni Medicina 52: 1–18.
Carrol CL and Huntington PJ (1988). Body condition scoring and weight estimation of horses. Equine Veterinary
Journal 20: 41–45.
National Research Council (NRC), (2007). Nutrient Requirements of Horses. 6th edn. Washington, DC: The National
Academies Press, pp. 341.
ABCCMM, (2007). Regulation of the XXVI National
Exhibition of the Mangalarga Marchador Horse. Belo
Horizonte: Brazilian Association of Creators of the
Mangalarga Marchador horse, pp. 28.
Meirelles JS (1997). Horse endurance. Technical Papers of
the Veterinary School of UFMG 19: 5–10.
Gardner AL (1986). Research Techniques in Pastures and
Applicability of Results in Production Systems. Brasilia:
II AC/Embrapa, pp. 197.
Association of Official Analytical Chemistry (AOAC), (1995).
Official Methods of Analysis. 16th edn. Arlington, VA:
AOAC International, pp. 1025.
Van Soest PJ, Robetson JB and Lewis BA (1991). Methods
for dietary fiber, neutral detergent fiber, and non starch
polysaccharides in relation to animal nutrition. Journal of
Dairy Science 74: 3583–3597.
Hahn GL (1985). Management and housing of farm animals
in hot environments. In: Yousef MK (ed.) Stress Physiology
in Livestock: Ungulates, vol. 2. Boca Raton, FL: CRC Press,
pp. 151–174.
Evans DL (1994). The cardiovascular system: anatomy,
physiology, and adaptations to exercise and training.
In: Hodgson DR and Rose RJ (eds) The Athletic Horse.
Philadelphia, PA: W.B. Saunders Company, pp. 129–144.
´
Castejon F, Riber C, Santisteban R, Tovar P, Trigo P
and Aguera S (2007). Muscular and ergometric evaluation
¨
on the treadmill. In: Opinion Forum of the Spanish Horse
(ed.) Morphofunctional Evaluation in the Selection of
´
Pure Spanish (Andalusian) Horse Breeders. Cordoba:
´
Universidad de Cordoba, pp. 99–112.
Eaton MD, Evans DL, Hodgson DR and Rose RJ (1995).
Effect of treadmill incline and speed on metabolic rate
during exercise in Thoroughbred horses. Journal of
Applied Physiology 79: 951–957.
Eaton MD, Evans DL, Hodgson DR and Rose RJ (1995).
Maximal accumulated oxygen deficit in Thoroughbred
horses. Journal of Applied Physiology 78: 1564–1568.
Pagan JD, Rotmensen T and Jackson SG (1995). The
effect of chromium supplementation on metabolic response to exercise in Thoroughbred horses. In: Equine
Nutrition and Physiology Society (ed.) Proceedings of 14th
Equine Nutrition and Physiology Symposium, Ontario, pp.
96–101.
Anderson RA (1998). Chromium, glucose intolerance and
diabetes. Journal of the American College of Nutrition
17: 548–555.
Cefalu WT and Hu FB (2004). Role of chromium in human
health and in diabetes. Diabetes Care 27: 2741–2751.
Ott EA and Kivipelto J (1999). Influence of chromium tripicolinate on growth and glucose metabolism in yearling
horses. Journal of Animal Science 77: 3022–3030.
Bunting LD, Fernandez JM, Thompson DL Jr and Southern
LL (1994). Influence of chromium picolinate on glucose
usage and metabolic criteria in growing Holstein calves.
Journal of Animal Science 72: 1591–1599.
Amoikon EK, Fernandez JM, Southern LL, Thompson DL,
Ward TL Jr and Olcott BM (1995). Effect of chromium tripicolinate on growth, glucose tolerance, insulin sensitivity,
plasma metabolites, and growth hormone in pigs. Journal
of Animal Science 73: 1123–1130.
Kreider RB (1999). Dietary supplements and the promotion
of muscle growth with resistance exercise. Sports Medicine
27: 97–110.

8. 140
24
25
26
27
LR Jordao et al.
˜
Prates RC (2007). Physiological parameters in the marcha
test of Mangalarga Marchador mares fed a diet supplemented with chromium. Dissertation (Master in Animal
Science), Veterinary School, Minas Gerais Federal University, Belo Horizonte, pp. 47.
Clarkson PM (1997). Effects of exercise on chromium
levels: is supplementation required? Sports Medicine 23:
341–349.
Jackson SG (1997). Trace minerals for the performance
horses: known biochemical roles and estimates of requirements. Irish Veterinary Journal 50: 668–674.
Farris JW, Hinchcliff KW, McKeever KH and Lamb DR
(1995). Glucose infusion increases maximal duration of
prolonged treadmill exercise in Standardbred horses.
Equine Veterinary Journal Supplement 18: 357–361.
28
29
30
31
32
Rose AJ and Richter EA (2005). Skeletal muscle glucose
uptake during exercise: how is it regulated? Physiology
(Bethesda) 20: 260–270.
Frape D (2004). The digestive system. In: Frape D (ed.)
Equine Nutrition and Feeding. 3rd edn. Oxford: Blackwell
Publishing, pp. 1–29.
Manohar M (1986). Effect of furosemide administration on
systemic circulation of ponies during severe exercise.
American Journal of Veterinary Research 47:
1387–1394.
Hyyppa S (2005). Endocrinal responses in exercising
¨
horses. Livestock Production Science 92: 113–121.
Valadares Filho SC (2006). Brazilian tables of cattle feed
composition. 2nd edn. Vicosa: Vicosa Federal University,
¸
¸
pp. 329.