METABOLIC DEMAND AND INTERNAL TRAINING LOAD IN ...

Journal of Strength and Conditioning Research Publish Ahead of Print DOI: 10.1519/JSC.0000000000001321

METABOLIC DEMAND AND INTERNAL TRAINING LOAD IN TECHNICAL-

ED

TACTICAL TRAINING SESSIONS OF PROFESSIONAL FUTSAL PLAYERS

Brief running head: Internal training load in professional futsal

Carolina F Wilke1,2, Guilherme P Ramos1,3, Diogo A S Pacheco1, Weslley H M Santos1,

Emerson Silami-Garcia1

1

EP T

Mateus S L Diniz1, Gabriela G P Gonçalves2, João C B Marins3, Samuel P Wanner1,

Exercise Physiology Laboratory, Universidade Federal de Minas Gerais, Belo

Horizonte (MG), Brazil

Nucleus of Sport Sciences Integration, Minas Tênis Club, Belo Horizonte (MG), Brazil

3

Brazilian Football Confederation, Teresópolis (RJ), Brazil

4

Human Performance Laboratory, Universidade Federal de Viçosa, Viçosa (MG), Brazil

A C

C

2

Corresponding author:

Dr. Samuel Penna Wanner

E-mail: [email protected] Telephone: +55 (31) 3409-2328

This investigation was supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), the Fundação de Amparo à Pesquisa do Estado de

Copyright ª 2016 National Strength and Conditioning Association

Minas Gerais (FAPEMIG), the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), the Pró-Reitoria de Pesquisa da Universidade Federal de Minas

A C

C

EP T

ED

Gerais, FUNDEP/Santander and the Brazilian Ministry of Sport.


1

1

ABSTRACT

3

This study aimed to characterize aspects of technical-tactical training sessions of a

4

professional futsal team. We addressed four specific aims: characterize the metabolic

5

demands and intensity of these training sessions, compare the training intensity among

6

players of different positions, compare the intensity of different futsal-specific activities

7

(4 × 4, 6 × 4 and match simulation), and investigate the association between an

8

objective (training impulse; TRIMP) and a subjective method (session rating of

9

perceived exertion; sRPE) of measuring a player’s internal training load. Twelve top-

10

level futsal players performed an incremental exercise to determine their maximal

11

oxygen consumption, maximal heart rate (HRmax), ventilatory threshold (VT) and

12

respiratory compensation point (RCP). Each player’s HR and RPE were measured and

13

used to calculate energy expenditure, TRIMP and sRPE during 37 training sessions over

14

8 weeks. The average intensity was 74 ± 4% of HRmax, which corresponded to 9.3

15

kcal.min-1. The players trained at intensities above the RCP, between the RCP and VT

16

and below the VT for 20 ± 8%, 28 ± 6% and 51 ± 10% of the session duration,

A C

C

EP T

ED

2

17

respectively. Wingers, defenders and pivots exercised at a similar average intensity but

18

with different intensity distributions. No difference in intensity was found between the

19

three typical activities. A strong correlation between the average daily TRIMP and

20

sRPE was observed; however, this relationship was significant for only four of twelve

21

players, indicating that sRPE is a useful tool for monitoring training loads but that it

22

should be interpreted for each player individually rather than collectively.

23 24

Keywords: energy expenditure, situational activity, team sport, training monitoring.


2

1

INTRODUCTION

2

A fundamental pillar for sport performance enhancement is to achieve an adequate

4

balance between training stimuli and athlete recovery. Within this context, monitoring

5

the training load provides coaches and sports scientists with important information that

6

enables adjustments in training sessions and recovery periods, thereby maximizing

7

performance improvement (26). Although many studies have been conducted in sports

8

with extended media coverage, including cycling (17, 28), rugby (22, 43) and soccer

9

(16, 24), fewer studies have focused on less-popular sports, such as futsal (41).

ED

3

EP T

10

Several methods can be used to measure the internal load, which is defined as the

12

individual response to the workload (external load) imposed on athletes during training

13

or competition. The heart rate (HR) is indicative of the overall physiological strain

14

during exercise (3) and can be used to calculate the training impulse (TRIMP), a

15

training load index (35). TRIMP was originally calculated by multiplying the average

16

intensity (expressed as the percentage of HR reserve) by the training duration and by a

17

weighting factor, which was included to account for the non-linear increase in the

18

contribution of the anaerobic system on the energy supply with increasing exercise

19

intensity (35). The concept of TRIMP was later developed by authors who proposed

20

intensity zones based on HR and attributed different weights to the time spent on high-

21

and low-intensity exercises during a training session (18, 28). Lucia et al. (28) used the

22

ventilatory threshold (VT) and the respiratory compensation point (RCP) as criteria to

23

determine three intensity zones, namely, the zones below the VT (considered to be low

24

intensity), between the VT and RCP (moderate intensity) and above the RCP (high

25

intensity). Using Lucia’s method, Milanez et al. (31) described the training load of 8 on-

A C

C

11


3

1

court semiprofessional Brazilian futsal players during 78 training sessions performed

2

before the main competition of the season. The authors observed that the players spent

3

73% of the duration of a technical-tactical training session below the VT, 20% between

4

the VT and RCP and only 7% above the RCP.

5

The session rating of perceived exertion (sRPE) is a simpler method than measuring the

7

HR to quantify the internal training load (21) and consists of assessing athletes’ RPEs

8

using Borg’s 10-point scale (12) 20 to 30 minutes after a training session and then

9

multiplying the score by the session duration. To validate the use of this method in team

10

sports, previous studies tested the correlation between sRPE and TRIMP and found

11

moderate to very strong correlation coefficients (4, 25, 40). However, only one study

12

has investigated the relationship between these variables to quantify the internal load

13

during physical and technical-tactical training sessions in futsal (31). Milanez et al. (31)

14

reported strong to very strong individual correlation coefficients between sRPE and

15

TRIMP and concluded that sRPE is a valid method to measure the internal training load

16

of futsal players.

EP T

C

A C

17

ED

6

18

Futsal is an intermittent, high-intensity sport (14, 11, 15, 38) in which an unlimited

19

number of player substitutions are allowed by official rules (20). To enable the athletes

20

to achieve optimal physical, technical and tactical performance during competitive

21

matches, coaches usually plan training activities with aspects similar to those of

22

different match situations (e.g., small-sided games, specific exercises with the same

23

number of players and court sizes as in official matches but with different rules and

24

instructions, and match simulations (MS)) in an attempt to mimic both the physiological

25

and cognitive demands of matches during the training sessions (42, 44). Although


4

previous studies have shown that some aspects of situational activities, such as the rules,

2

pitch size and number of players, can influence exercise intensity in different sports (1,

3

24), it remains unknown whether the existing information can be used to guide futsal

4

training prescription (11). Thus, based on the importance given to specific training as a

5

determining factor for achieving optimal performance during competitive matches, the

6

intensities of specific activities for futsal with different rules and number of players

7

must be measured and compared.

8

ED

1

Another aspect that deserves further investigation is the stress imposed on players of

10

different tactical positions by training activities. A recent study showed that the

11

concentrations of muscle damage and inflammation markers after matches are similar in

12

defenders, winger and pivots (15), suggesting similar physiological strain among

13

players during the matches. However, to the best of our knowledge, no study has

14

investigated whether physiological strain is also similar among athletes of different

15

playing positions during technical-tactical training sessions.

C

16

EP T

9

Thus, considering the lack of information about training load in futsal, the general aim

18

of this study was to characterize different aspects of training load in professional futsal

19

technical-tactical training sessions. Four specific aims were established: 1) characterize

20

the metabolic demands (estimated VO2) and intensity (percentage of HRmax; %HRmax)

21

of professional futsal training sessions, 2) characterize and compare the training

22

intensity (%HRmax) and intensity distribution (percent of time spent at different intensity

23

zones) among players of different positions (wingers, defenders and pivots), 3)

24

characterize and compare the intensity of typical exercises (4 × 4, 6 × 4 and MS)

25

performed by professional futsal players, and 4) investigate the association between an

A C

17


5

1

objective method (TRIMP) and a subjective method (sRPE) of measuring the internal

2

training load in futsal.

3 4 5

METHOD

ED

6

Experimental approach to the problem

8

This study followed a descriptive longitudinal approach. The internal load of

9

professional futsal players was recorded in every technical-tactical training session over

10

8 weeks. These experiments were conducted between late September and early

11

November, when the team was not participating in any competition and only

12

participated in two friendly matches. The training sessions were planned and executed

13

by the coaches; to ensure ecological validity, the researchers did not interfere with either

14

the planning or execution of the training sessions.

EP T

7

C

15

Prior to the commencement of the study, all of the players performed an incremental

17

maximal test to assess their maximal HR (HRmax) and their HR relative to the VT and to

18

the RCP. To evaluate the internal training load, we measured the HR and RPE of each

19

player in each session over an 8-week period. To assess whether the players’ hydration

20

status could affect the HR measurements during training sessions, the volume of fluid

21

consumed, the body mass loss and the environmental temperature were also measured

22

over the entire study. Estimated oxygen consumption (VO2), energy expenditure, sweat

23

loss, TRIMP and sRPE were calculated for each player in each training session. To

24

guarantee a deeper understanding of the observed data, the training intensity for the

25

different playing positions (winger, defender and pivot) and the intensities of three

A C

16


6

1

futsal-specific training activities were compared. Additionally, the correlation between

2

two internal load indices (i.e., TRIMP and sRPE) and two measures of training intensity

3

(HR and RPE) were tested.

4

Subjects

6

Twelve top-level male futsal players from a first division Brazilian team participated in

7

this study. At the time of the study, the players displayed the following characteristics

8

(mean ± SD): 23.0 ± 4.3 years of age (18 – 32 years range), 70.7 ± 8.5 kg body mass

9

and 54.2 ± 4.2 mlO2.kg-1.min-1 maximal oxygen consumption (VO2max). During the year

10

of this investigation, the team won the second main national competition promoted by

11

the Brazilian Futsal Federation. The players were familiarized with the procedures for 4

12

weeks prior to the commencement of the study and were asked to maintain their

13

habitual diet during the experimental period.

EP T

ED

5

14

All of the procedures described in this study were approved by the Federal University of

16

Minas Gerais Ethics Committee (14376413.0.0000.5149), and all volunteers provided

17

their written consent after being informed of the procedures, potential risks and benefits

18

of their participation in this investigation.

A C

C

15

19 20

Procedures

21 22

Preliminary tests

23

A maximal incremental test (adapted from DITTRICH et al., 16) was performed to

24

assess the players’ VO2max, HRmax, VT and RCP. The test was conducted on a treadmill

25

(Total Health®, HPX 380, Brazil) with an initial speed of 9 km.h-1 and additional


7

increments of 1.2 km.h-1 every 3 minutes until volitional fatigue. If the subject did not

2

interrupt the exercise after finishing the stage corresponding to 16.2 km.h-1, this speed

3

was kept constant, and the inclination was increased by 1% every 3 minutes until

4

fatigue (this was the case for only one player). At least 2 of the following criteria had to

5

be met to determine the VO2max: 1) no further increase in VO2 or HR despite an increase

6

in exercise intensity; 2) RPE higher than 17 on the 6 – 20 scale; and 3) respiratory

7

exchange ratio higher than 1.10. The highest HR and VO2 registered during the test

8

were considered the subject’s HRmax and VO2max.

9

ED

1

VO2 and VCO2 were measured breath-by-breath using an open circuit gas analyzer

11

(GasSys2, BIOPAC System®, USA) calibrated before each test. HR was continuously

12

measured (Team System, Polar®, Finland) and was recorded every minute, and RPE (15

13

point scale – Borg, 12) was assessed immediately after the completion of each stage and

14

at volitional fatigue. Average data from the last minute of each stage were used to

15

determine the individual VO2 values. A linear regression equation was developed using

16

individual HR and VO2 data to establish the association between these two variables.

C

A C

17

EP T

10

18

The VT and RCP were visually identified by two independent investigators, and a

19

consensus-derived value was used. The VT was defined as the HR value corresponding

20

to the moment of increase in VE.VO2−1 without a concomitant increase in VE.VCO2−1,

21

whereas RCP was defined as the moment of increase in both VE.VO2−1 and VE.VCO2−1

22

(6). Whenever there was a disagreement between the two investigators, a third

23

researcher was asked to analyze the data. The point at which two of the three

24

researchers agreed was used for subsequent analyses. Notably, disagreement only

25

occurred for the RCP analysis of one player.


8

Training sessions

2

In total, 37 technical-tactical training sessions were conducted and monitored over 8

3

weeks. The training routine comprised one to two daily technical-tactical training

4

sessions performed on indoor futsal courts six days per week and one to two resistance

5

training sessions per week. During the study, the researchers did not interfere with the

6

training planning or execution, which were coordinated by the coaches of the team.

7

Care was taken to not modify the training session routine to guarantee the ecological

8

validity of the results.

ED

1

9

All of the evaluated sessions were divided in the following two parts: 1) the first 15 to

11

30 minutes consisted of a warm-up that included general exercises such as jogging, low-

12

intensity multidirectional movements and stretching; and 2) the next 60 to 90 minutes

13

consisted of match simulations or activities specific for futsal, with varying instructions

14

and numbers of players, that aimed to improve the technical and tactical skill of the

15

athletes. The second part of each training session was termed the technical-tactical

16

period. The training sessions were performed on courts with the following dimensions:

17

36 × 20 m (n = 23), 31 × 19 m (n = 11) and 25 × 15 m (n = 03). The court dimensions

18

were selected by the coaching staff based on the space available in the multisport

19

training center.

A C

C

EP T

10

20 21

Before each training session, each player’s semi-nude (wearing shorts only) body mass

22

was measured. Then, each player received a HR monitor (Team System, Polar®,

23

Finland) that was used during the entire session. This monitor is reliable for measuring

24

HR during continuous and intermittent exercises (3). Two bottles of water (WAT) and

25

one bottle of a commercial carbohydrate beverage (CHO, 6%), filled with


9

approximately 500 ml of fluid and previously weighed in a 0.02-kg precision balance

2

(MF100, Filizola®, Brazil), were provided for each athlete. They were instructed to

3

consume these fluids ad libitum and not to use the fluids for purposes other than

4

ingestion (such as applying to the skin for refreshment). If necessary, the researchers

5

refilled the bottles with the respective drink during training, thereby guaranteeing ad

6

libitum fluid ingestion. WAT and CHO were offered to maintain the team’s routine

7

because both of these fluids were available during all of the previous training sessions

8

throughout the year.

ED

1

9

HR was continuously measured, and training activities were recorded using a digital

11

video camera (DCR SR87 HD, Sony®, Japan). Wet and dry bulb temperatures were

12

measured using a psychrometer (Alla France, France) every 15 minutes to calculate the

13

average wet bulb globe temperature (WBGT) for indoor environments.

EP T

10

14

At the end of the training sessions, the players dried their sweat with a towel and then

16

underwent a second body mass measurement. Within 15 to 20 minutes after the end of

17

the session, RPE was assessed using a 10-point scale (CR-10; FOSTER et al., 21) by

18

asking each player individually “how was your training?” Finally, all of the bottles were

19

weighed to calculate the volumes of WAT, CHO and total fluid ingested.

A C

C

15

20 21

Total sweat was calculated as the difference between post- and pre-training body mass

22

corrected for the total volume of fluid ingested. Total sweat was divided by the duration

23

of the training session to assess the sweat rate. The difference between post- and pre-

24

training body mass was divided by the pre-training body mass and multiplied by 100 to


10

1

calculate the percent change in body mass (%BM). Body mass losses higher than 3%

2

are associated with exaggerated exercise-induced increases in HR (23, 39).

3

To estimate the VO2 during the training sessions, a linear regression analysis was

5

performed using the individual VO2 and HR values obtained during the incremental test.

6

The average HR measured in each training session for each athlete was then applied to

7

the resultant regression equation to estimate the average VO2 of the training sessions.

8

The energy expenditure was then estimated, considering that the uptake of 1 L of O2

9

corresponds to the expenditure of 4.8 kcal (9).

EP T

10

ED

4

Training intensity was calculated as the average percentage (%) of HRmax and time (in

12

minutes) spent in three HR zones, defined as follows: zone 1) HR values lower than the

13

HR corresponding to the VT; zone 2) HR values between those corresponding to the VT

14

and RCP; and zone 3) HR values higher than the HR corresponding to the RCP. The

15

time spent in each zone was then multiplied by its corresponding score (zone 1 = 1;

16

zone 2 = 2; zone 3 = 3), and the products were summed to calculate the individual

17

TRIMP for each training session (28). The sRPE was also calculated by multiplying

18

each player’s RPE after the training session by the session duration in minutes (21).

A C

C

11

19 20

To compare the training intensity between playing positions [i.e., wingers (n = 5),

21

defenders (n = 2) and pivots (n = 5)], the average values for % HRmax and the time spent

22

in the high, moderate and low intensity zones were calculated for each position in each

23

session (i.e., n = 35 sessions for each of the three playing positions). The head coach

24

was responsible for determining the playing position of each athlete.

25


11

Characterization of different training activities

2

The results from 9 athletes were used to calculate the average %HR and intensity

3

distribution (time spent in zones 1, 2 and 3) for three specific activities frequently

4

performed during the training sessions. All of the selected activities included in the

5

analysis met the following criteria: 1) performed in the morning; 2) held in a 36 × 20-m

6

court; 3) involved the same number of players; and 4) lasted for a minimum of 15

7

minutes. For data analysis, the three specific activities were as follows:

ED

1

8

a) 6 versus 4 (6 × 4): Each team comprised six players on the court, though a

10

numerical advantage was created for the offensive team (attack overload).

11

The opposite team was allowed to maintain only four players on the

12

defensive half of the court, and the other two players stayed in the attack

13

half of the court, without participating in the defense. Furthermore, both

14

teams had a substitute player who could replace one of the team players on

15

the court at any time. The average duration of the two situations selected

16

was 18 ± 4 minutes, and the complete HR recordings from the first 15

17

minutes were analyzed, including the time spent for instructions and

18

substitutions. The main instruction provided by the coach was to maintain

19

ball possession for as long as possible while attacking. Both teams had

20

similar chances to attack during the activity, and the coach instructed the

21

players whenever he felt necessary.

A C

C

EP T

9

22 23

b) 4 versus 4 (4 × 4): Each team comprised four players and a goalkeeper on

24

the court plus one substitute player who could replace one of the players on

25

the court at any time. The average duration of the four situations selected


12

was 21 ± 8 minutes, and the complete HR recordings from the first 15

2

minutes were analyzed, including the time spent for instructions and

3

substitutions. The focus of the instruction provided to the athletes in these

4

activities was the attack movement. Both teams had similar opportunities to

5

attack during the activity, and the coach instructed the players whenever he

6

felt necessary.

ED

1

7

c) Match simulation (MS): This activity involved the same number of players

9

as did the 4 × 4 activity (4 players and a goalkeeper on the court plus one

10

substitute player). The differences between these two activities were mainly

11

in the duration, rules and instruction. The simulations that were analyzed

12

comprised two 10-minute half periods separated by 5 minutes; each athlete

13

was evaluated during the 20 minutes of MS, including the moments when

14

they were playing and when they were substituted. The rules were the same

15

as those of an official match, and there was no interruption for instructions.

16

In addition, no specific instruction about the aim of MS was provided to the

17

athletes by the coach.

A C

C

EP T

8

18 19

Statistical analysis

20

Kolmogorov-Smirnov tests were performed to test data normality. Because all data

21

were normally distributed, they are presented as the mean ± SD. Comparisons of the

22

weekly internal training load (i.e., TRIMP and sRPE) were performed using one-way

23

analysis of variance (ANOVA). Post-hoc differences in TRIMP were assessed using the

24

Hochberg test, and differences in sRPE were assessed using the Games-Howell test;

25

both utilized a Bonferroni correction. To compare the average training intensity (i.e. %


13

HRmax) among the playing positions (i.e. wingers, defenders and pivots), a one-way

2

ANOVA was performed. To compare the training intensity distribution (i.e. below LT,

3

between LT and RCP and above RCP) among the playing positions, a two-way

4

ANOVA ([3] time in each training intensity zone × [3] playing position) was performed

5

followed by Tukey´s post hoc test. Comparisons of the average intensity among the

6

three training activities (i.e. 4 × 4, 6 × 4, MS) were assessed by repeated measures one-

7

way ANOVA, and comparisons of the intensity distribution among these three activities

8

were assessed by repeated measures two-way ANOVA ([3] time in training intensity

9

zone × [3] activities). For these comparisons, data reliability was tested using both

10

relative (intraclass correlation coefficient; ICC (3,3); 45) and absolute (standard error of

11

measurement; SEM) indices, and the effect size was calculated using the omega index

12

(ω). The associations between HR and RPE and between TRIMP and sRPE were

13

assessed using Pearson’s correlation coefficients and classified according to Evans’

14

criteria (0.00 to 0.19 – very weak; 0.20 to 0.39 – weak; 0.40 to 0.59 – moderate; 0.60 to

15

0.79 – strong; above 0.80 – very strong) (19). The difference in the volumes of CHO

16

and WAT consumed was tested using Student’s t-test. A significance level of p < 0.05

17

was adopted. All of the analyses were performed using SPSS 17.0 software for

18

Windows (SPSS Inc.).

A C

C

EP T

ED

1

19 20 21

RESULTS

22 23

The exercise intensities that corresponded to the players’ VT and RCP were 75.8 ±

24

4.8% of HRmax (53.7 ± 8.6% of VO2max), and 89.8 ± 5.1% of HRmax (78.2 ± 9.6% of

25

VO2max), respectively.


14

1

Although the training sessions were not exclusively held on one court size, all of the

3

training data were analyzed and presented together because no differences were found

4

in the average intensities attained by the players on the different court sizes (36 × 20-m

5

court: 74.1 ± 3.6% of HRmax; 31 × 19-m court: 73.7 ± 3.9% of HRmax and 25 × 15-m

6

court: 70.5 ± 1.5% of HRmax; F (2,33) = 1.345, p = 0.274; ω = 0.269).

ED

2

7

The average training intensity was 73.7 ± 3.6% of HRmax, which corresponded to an

9

energy expenditure of 846 ± 129 kcal or 9.3 ± 1.0 kcal.min-1. Analysis of the intensity

10

distribution showed that the athletes exercised at intensities above the RCP for 20.4 ±

11

7.8%, between the RCP and the VT for 28.2 ± 5.6% and below the VT for 51.4 ± 9.7%

12

of the time during the training sessions. The number of monitored sessions per athlete

13

was 23 ± 8, with a minimum of 13 and a maximum of 37 sessions per athlete. The

14

average duration of the sessions was 90.8 ± 11.6 min, and the average environmental

15

conditions were 27.3 ± 2.7°C and 23.0 ± 2.2°C for dry bulb temperature and WBGT,

16

respectively.

C

A C

17

EP T

8

18

The average daily and average weekly TRIMP values over the 8 weeks were 153 ± 21

19

arbitrary units (AU) and 531 ± 148 AU, respectively. As shown in Figure 1, no

20

difference in the average daily TRIMP was found over the 8 weeks (minimum of 140 ±

21

20 AU in week 5 and maximum of 173 ± 21 AU in week 6; F (7,31) = 1.350; p = 0.261; ω

22

= 0.48). However, the weekly TRIMP was lower in the sixth week than in the eighth

23

week of training (266 ± 139 in week 6 vs. 695 ± 105 in week 8; F

24

0.001; ω = 0.96). Additionally, no difference in the daily sRPE was found between the

25

weeks (minimum of 359 ± 58 AU in week 5 and maximum of 514 ± 54 AU in week 6;

(7,63)


= 3.597; p =

15

= 1.477; p = 0.212; ω = 0.52), though the main effect for weekly sRPE was

F

2

significant (F (7,67) = 6.184; p < 0.001; ω = 0.99). sRPE in week 6 (869 ± 416) was lower

3

than in weeks 1, 4, 7 and 8 (2273 ± 423, 2183 ± 804, 2332 ± 845 and 2399 ± 351,

4

respectively; p < 0.001). The lower weekly TRIMP and sRPE observed in week 6 are

5

likely because the team played two friendly matches but performed no other training

6

activities, resulting in fewer sessions compared with the other weeks.

(7,31)

7

Figure 1 about here

8 9

ED

1

During the technical-tactical period, no difference was observed in the average training

11

intensity (i.e. %HRmax) among players of different positions (wingers: 76.3 ± 4.9% of

12

HRmax, defenders: 74.4 ± 4.9% of HRmax and pivots: 76.4 ± 4.0% of HRmax; F

13

2.032; p = 0.136; ω = 0.41). However, these similar average values resulted from

14

different distributions within the intensity zones (Figure 2). The wingers spent more

15

time in the high intensity zone than the pivots (33.6 ± 15.6% vs. 19.1 ± 8.5%; F (2,100) =

16

12.122; p < 0.001; ω = 0.995) and defenders (25.1 ± 12.0%; p = 0.016; ω = 0.995). The

17

pivots spent more time in the moderate intensity zone than the wingers and defenders

18

(pivots 37.4 ± 7.9% vs. wingers 20.9 ± 8.7% vs. defenders 23.3 ± 7.2%; F

19

44.271; p < 0.001; ω = 1.00), and the defenders spent more time in the low intensity

20

zone than did the pivots (wingers 45.7 ± 11.5% vs. defenders 51.6 ± 15.4% vs. pivots

21

43.5 ± 10.4%; F (2,100) = 3.975; p = 0.019; ω = 0.70).

A C

22 23

(2,100)

=

C

EP T

10

Figure 2 about here

24


(2,100)

=

16

1

The three training activities (4 × 4, 6 × 4 and MS) yielded similar average intensity

2

values (F (2,16) = 0.158; p = 0.855; ω = 0.07). The main effect for the two-way ANOVA

3

(activity x intensity zone) was not significant neither for activity (F (2,72) = 0.000047; p =

4

1.000; ω = 0.05) or the time spent in the different intensity zones (F

5

0.066; ω = 0.54). There was also no interaction between the activity and intensity zone

6

(F (4,72) = 0.876; p = 0.483; ω = 0.266) (Table 1).

= 2.824; p =

ED

(2,72)

7

Table 1 about here

8 9

A strong correlation was observed between TRIMP and sRPE (r = 0.70; r2 = 0.49; p