Content uploaded by Peter Edholm
Author content
All content in this area was uploaded by Peter Edholm on Oct 13, 2017
Content may be subject to copyright.
Half-time re-warm up increases performance capacity in male
elite soccer players
P. Edholm1, P. Krustrup2,3, M. B. Randers2
1School of Health and Medical Sciences, Örebro University, Örebro, Sweden, 2Department of Nutrition, Exercise and Sports,
Copenhagen Centre for Team Sport and Health, University of Copenhagen, Copenhagen, Denmark, 3Sport and Health Sciences,
College of Life and Environmental Sciences, St. Luke’s Campus, University of Exeter, Exeter, UK
Corresponding author: Peter Edholm, School of Health and Medical Sciences, Örebro University, Fakultetsgatan 1, 70282 Örebro,
Sweden. Tel: +019 303969, Fax: +019 303486, E-mail: peter.edholm@oru.se
Accepted for publication 19 March 2014
This study investigated the acute effects of a half-time
re-warm up on performance and movement patterns in
soccer match play. Using a crossover design, 22 profes-
sional male players performed traditional passive rest
(CON) or a low-intensity re-warm up (RW) during the
half-time period of two soccer matches. Before and after
the first half and before the second half, maximal sprint
and jump performance were evaluated. Time–motion
analysis of the first 15 min of each half was conducted.
Sprint and jump performance were reduced (P<0.05) by
2.6% and 7.6%, respectively, during the half-time period
in CON, whereas sprint performance was maintained
and the decrement in jump performance (3.1%; P<0.05)
was lower after RW. No significant interaction for high-
intensity running was observed, but less defensive high-
intensity running was observed after RW than CON
(0.14 ±0.06 vs 0.22 ±0.07 km; P<0.01). Moreover, RW
had more possession of the ball in the beginning of the
second half. In conclusion, traditional passive half-time
rest leads to impaired sprint and jump performance
during the initial phase of the second half in professional
soccer players whereas a re-warm up effectively attenu-
ates such deteriorations. Less defensive high-intensity
running and more ball possession were observed after
RW, indicating a game advantage at the onset of the
second half.
The ability to perform large amounts of high-intensity
running is generally considered as an important perfor-
mance marker in soccer, and most (Reilly & Thomas,
1976; Ekblom, 1986; Bangsbo et al., 1991; Carling
et al., 2008; Ingebrigtsen et al., 2012) but not all (Di
Salvo et al., 2009; Rampinini et al., 2009) studies have
shown that professional soccer players perform more
high-intensity running during all stages of a match com-
pared with lower level soccer players. Interestingly,
several studies have reported that soccer players at all
levels including professional players perform less high-
intensity running during the initial phase of the second
half compared with the first half of a match (Mohr et al.,
2003, 2005; Bradley et al., 2009; Weston et al., 2011). In
fact, Mohr et al. (2005) showed that as much as 20% of
elite soccer players have their least intense 15-min
period in a match during the initial part of the second
half. This is remarkable, especially as the players just
before this period have had a 15-min half-time period,
aiming to give the players time to rest and recover so that
they can resume the second half of the game with
renewed vigor. The reason for this significant reduction
in high-intensity running is unknown. However, several
hypotheses have been suggested, including tactical
aspects, accumulated fatigue, half-time length, ambient
temperature, and lack of physical preparation prior to the
second half (Mohr et al., 2004; Mugglestone et al., 2013;
Zois et al., 2013). In addition to high-intensity running,
soccer players’ capacity to produce explosive actions
such as maximal sprint and jump performance is also an
important performance marker (Stolen et al., 2005).
Even small differences might be crucial in the critical
duels influencing the results of the game and a difference
of only 0.05 s in a 10-m sprint equals 0.25–0.5 m
depending on the running speed. In line with this, pro-
fessional soccer players have repeatedly been reported
to have higher sprint and jump capacity than non-
professionals (Stolen et al., 2005).
Before a match professional soccer players usually
perform a warm up routine typically lasting ∼30 min and
consisting of low- to high-intensity exercises (Towlson
et al., 2013). This type of physical preparation has been
shown to elevate muscle temperature, which in turn is
associated with enhanced sprint and jump performance
(Sargeant, 1987; Stewart & Sleivert, 1998; Bishop, 2003;
Mohr et al., 2004). Thus, the warm up regime performed
by professional soccer players before matches induces
physiological changes assuring that players are well pre-
pared to perform maximal efforts directly from the start
of match. Given the above discussion, it is interesting to
Scand J Med Sci Sports 2014: ••: ••–••
doi: 10.1111/sms.12236
© 2014 John Wiley & Sons A/S.
Published by John Wiley & Sons Ltd
1
note that re-warm up regime specifically set up to
prepare the players for the second half of the game are
rarely used (Towlson et al., 2013). Indeed, the dominat-
ing routine during the half-time period in professional
soccer is traditionally restricted to passive seated recov-
ery, rehydration, and brief tactical advice from the coach
(Towlson et al., 2013). Such half-time regimes are prob-
ably not optimal as it has been shown that both core and
muscle temperature decreases during a passive half-time
period (Mohr et al., 2004; Lovell et al., 2007, 2013).
Moreover, those studies also reported that as a conse-
quence of the decrements in core and muscle tempera-
ture, the soccer players’ exercise capacity was impaired
including sprint, jump, and dynamic strength perfor-
mances. Thus, this sequence of event with passive seated
half-time recovery, decreased muscle temperature, and
impaired exercise capacity might at least, in part, be the
cause of the reduced amount of high-intensity running
seen during the initial phase of the second half in soccer
matches.
The issue of re-warm up before the second half during
90-min soccer matches is poorly studied and has been
addressed in only two reports (Mohr et al., 2004; Lovell
et al., 2013). Nevertheless, both of these two novel
papers suggest that a short-active re-warm up regime
(7 min of low-to-moderate aerobic running exercises at
approximately 70% of HRmax as a component of the latter
part of the 15-min half-time period) is an efficient
method to maintain elevated muscle temperature and
preserve sprint and jump performance over the half-time
period. However, both of these studies have some limi-
tations that need to be considered. First, both studies
were conducted on amateur players. Second, the study
design included either a friendly match where the
players repeatedly were replaced by substitute players
during muscle temperature sampling or a soccer game
simulation agility course aiming to replicate a competi-
tive match.
Thus, it is still unknown whether the proposed impair-
ments in sprint and jump performance after the half-time
period also are present in well-trained professional
players during real soccer matches and subsequently if a
half-time re-warm up regime outlined as above is an
effective method to attenuate such potential decreases in
performance capacity. Finally, it is also unknown
whether there are any connections between the reduced
performance capacity (i.e., sprint and jump perfor-
mance) and the reduced amount of high-intensity
running that are often observed during the initial part of
the second half, and whether a re-warm up during the
half-time period also results in an increased amount of
high-intensity running performed in the initial phase of
the second half.
Therefore, the aims of this investigation were to study
the effect of an active re-warm up program as a part of
the half-time period during match play in professional
soccer players. More specifically we investigated (a)
changes in maximal sprint performance and counter-
movement jump (CMJ) performance and (b) movement
pattern and technical skill during the initial 15 min of
each half.
Methods
Participants
Twenty-two male professional soccer players with a mean age of
25 (range: 18–33) years, an average height of 182 (175–195) cm,
and a mean body mass of 78.6 (69.3–93.6) kg playing in the same
team in the top league in Sweden (Allsvenskan) participated in the
study. The players participated in two 90-min matches separated
by 6 days of regular training. Only field players with complete test
data from both matches were included in the statistical analysis.
Three players were not able to complete both matches due to
injuries (one during the first match and two during training
between the matches).
Thus, a total of 17 players were included in the analysis. All
players were informed of the experimental procedures and pos-
sible discomforts associated with the study before giving their
written consent to participate. The study was conducted in accor-
dance with the guidelines set by the Swedish Central Ethical
Review Board.
Experimental design
The players were divided into two teams according to playing
position and coach-evaluated skill level. The selection of the two
teams was made in close collaboration with the head coach to
make sure that the teams were of equal skill level. Thereafter, two
90-min matches separated by 6 days of training were completed
with the same subjects playing the same field position in the same
team during both matches. During the 3 days preceding each
match, training was standardized and consisted of low-intensity
soccer training. Before the start of each match, players performed
a 30-min standardized warm up, consisting of low-to-moderate
speed running, light individual calisthenics, and stretching exer-
cises followed by short bursts of high-speed running and different
playing exercises with the ball at a moderate intensity. During the
first match, the control group (CON) had a traditional 15-min
half-time period in the break between the first and the second half
of the game (including seated passive recovery, rehydration, and
coaching instructions), whereas the experimental group performed
a re-warm up regime (RW) essentially as described by Mohr et al.
(2004).
Thus, the RW protocol consisted of 7 min of traditional passive
rest (as described above), followed by 7 min of low-/moderate-
intensity jogging and light calisthenics at ∼70% of maximal heart
rate. The RW regime ended 1 min before the start of the second
half. The study was conducting using a crossover design, so that
the team who had the traditional half-time rest in the first match
performed the RW regime in the second game and vice versa.
To mirror a real competitive match as much as possible, the two
teams were separated during the matches, using separate dressing
room, test stations, and performed the half-time activities sepa-
rately. In addition, the two matches were supervised by official
referees with extensive experience from the top Swedish soccer
league. To minimize the influence of match tactics on individual
movement patterns, the playing system and tactics were standard-
ized in close collaboration with the head coach, as defensive/
offensive tactics might result in large inter-individual variation in
the amount of high-intensity running among players (Bradley
et al., 2011). During the matches the teams only had support of an
assistant coach, whose role was to assist and remind the players of
the preordained tactics.
Edholm et al.
2
The player’s net fluid loss during the matches was determined
by weighing the players without clothes immediately before and
after the matches, using a digital scale (Seca 708, Seca Ltd.,
Birmingham, UK). The players were allowed to drink water ad
libitum during the matches and their fluid intake was measured.
The fluid loss was calculated as fluid loss =Δbody weight +fluid
intake. The players wore soccer boots and clothing during all
trials, and were instructed to refrain from consumption of caffeine
(24 h) and alcohol (48 h) before each trial. The players maintained
their usual diet throughout the study period. We noted similar
weather conditions (5 vs 6 °C, 87% vs 84% humidity and 3 vs
2 m/s wind) when the two matches were played.
Performance tests
The players completed performance tests on three occasions: (a)
after the pre-match warm up, before start of the first half; (b) after
the first half, before start of half-time period; and (c) after the
half-time period just before the start of the second half (Fig. 1). All
the players were familiarized to all of the testing procedures 1
week before the first experimental day.
Each test occasion included two maximal 10-m sprints and two
maximal CMJs. The tests were always performed in the same
order during all test occasion. The 10-m sprint tests were per-
formed on grass at the side of the field with players wearing soccer
boots. Each player performed two trials separated by a ∼90-s walk
back recovery and the best result was used for data analysis. The
players were instructed to start with their feet in a parallel position
standing behind a mark, which were placed 0.5 m behind the first
set of photocells. The players were also instructed not to do any
kind of backward movement with the upper body or the feet, as
such procedure allowed the player to break the first photocell with
a forward acceleration movement and thereby reducing the sprint
time. Sprint time was recorded by infrared light sensor, with a
precision of 0.01 s and placed at a height of 1.0 m (Musclelab
4000, Ergotest, Oslo, Norway).
Immediately after the sprint test, the players performed two
maximal CMJs separated by ∼90 s of rest. The jumps were per-
formed indoors on a 25-mm hard tartan rubber carpet, which
allowed the player to perform the jumps while wearing soccer
boots. The best result was used for data analysis. The CMJ started
from a standing position with hands fixed to the hips. To make sure
that the player landed on their toes (with ankles in a dorsal flexion
position), the players were instructed to perform a small post-jump
direct after landing (not measured). This procedure ensured that
the players were jumping and landing on the ground with their toes
first and with straight legs, consequently giving a correct time
in the air. Jumping height was measured using portable
MicroMusclelab (Ergotest) units connected to infrared light rails.
Jumping height was calculated as h=tf2·g·8
−1where h=jump
height, t=time in air, and g=gravity. Each team had their own
test station (sprint and jump) to make sure that the performance
test took a minimum amount of time.
Movement patterns and heart rate
During both matches all field players were filmed close up
throughout the first 15 min of each half (Fig. 1) using 20 digital
video cameras (Canon DM-MV 600, Canon Inc., Tokyo, Japan).
An additional camera was used to follow the ball. The cameras
were positioned at the side of the pitch, at the level of the midway
line, at a height of about 15 m, and at a distance of 30–40 m from
the touchline. The recordings were later replayed on a monitor for
computerized coding of activity pattern as described in detailed by
Bangsbo et al. (1991). The following locomotion categories were
used: standing (0 km/h), walking (6 km/h), jogging (8 km/h), low-
speed running (12 km/h), moderate-speed running (15 km/h),
high-speed running (18 km/h), sprinting (30 km/h), and backward
running (10 km/h). The locomotion categories were chosen in
accordance with previous studies (Bangsbo et al., 1991; Mohr
et al., 2003). High-intensity running was defined as all runs with
speeds above 15 km/h (encompassing moderate-speed running,
high-speed running, and sprinting). The number of occurrences of
each activity was recorded in each of the first 15 min of each half
in both matches. The distance covered in each activity for each
interval was determined as the product of the total time spent in
that activity category and mean speed for that activity. The total
distance covered during the first 15 min of each half was calcu-
lated as the sum of the distance covered during each type of
activity. The number of shoots, dribbles, and passing was counted
and possession of the ball was analyzed.
Offensive high-intensity runs were defined as high-intensity
runs while the team had possession of the ball. The pitch was
divided into three zones (defending zone (team 1)/attacking zone
(team 2), neutral zone, and attacking zone (team 1)/defending zone
(team 2). The neutral zone was defined as 19 m on each half of the
pitch from the halfway line. Effective playing time was defined as
the total playing time minus time used for stoppages. Data from
each 15-min period were subsequently divided into 5-min periods.
All recordings were analyzed by the same experienced observer.
Prior to the present analysis the experienced observed analyzed the
same matches not related to the present study with 6 months
between the analysis and a coefficient of variation for test–retest
analysis of <5% for each locomotion activity was found. The
observer was blinded for re-warm up regimes in both matches. The
recordings from the first and the second half of each match as well
as players from both teams were analyzed in a random order. Heart
rate was monitored in 5-s intervals throughout the games including
the half-time periods using Polar System heart rate monitors (Polar
Electro OY, Kempele, Finland).
Statistical analysis
Statistical analysis was performed using SigmaStat software
(version 12.0, Systat Software Inc., San Jose, California, USA).
Changes in sprint and jump performance, movement pattern, and
ball skill and technical standards were evaluated by a two-way
analysis of variance with repeated measures. When a significant
Test 1 Test 2 Test 3
Warm up 30 min 1st half 45 min
Half-time rest
15 min
(CON/RW) 2nd half 45 min
Heart rate monitoring
Video filming
15 min
Video filming
15 min
Fig. 1. Schematic view of the experimental design during the two matches. Sideward arrows denote end of filming; upward arrows
denote test of sprint and jump performance.
Half-time re-warm up and soccer performance
3
interaction was detected, data were subsequently analyzed using a
Newman–Keuls post-hoc test. Effect sizes (ESs; η2) were calcu-
lated and values of 0.01, 0.06, and >0.15 were considered small,
medium, and large, respectively (Cohen, 1988). Test–retest analy-
sis revealed high-reliability coefficient for both CMJ and 10-m
sprint (α=0.97 and 0.93), respectively. The differences in heart
rate between the matches and the first and second half were tested
with Student’s paired t-test. Correlation between two variables
was assessed using Pearson’s correlation coefficient. P-val-
ues <0.05 were considered statistically significant. Data are pre-
sented as mean ±standard deviation unless otherwise stated.
Results
Sprint performance
A significant time ×condition interaction was found for
10-m sprint performance (P=0.014; η2=0.05, i.e.,
small ES). Post-hoc test revealed that no differences in
the player’s 10-m sprint performance were found before
the first half between re-warm up (RW) and control
condition (CON) (1.89 ±0.04 vs 1.88 ±0.06 s). After
the first half sprint performance was reduced by ∼3%
during both RW and CON to 1.95 ±0.06 and
1.93 ±0.05 s, respectively (P<0.05). Immediately after
the half-time period, sprint performance was further
reduced by another 2.6% during CON (1.98 ±0.06 s),
whereas no further change was observed after RW
(1.94 ±0.05 s, P<0.05 vs CON; Fig. 2(a)).
Jump performance
Also for CMJ a significant time ×condition interaction
was found (P=0.043; η2=0.05, ES =small). Post-hoc
test revealed that before the start of matches, there were
no differences in the players’ CMJ performance between
RW and CON (38.7 ±3.7 vs 39.0 ±2.9 cm) and after the
first half CMJ performance was still unchanged during
both RW and CON (38.7 ±3.8 and 39.4 ±3.8 cm,
respectively). However, directly after the half-time
period CMJ performance was significantly reduced by
3.1% during RW (37.5 ±3.7 cm) and 7.6% during CON
(36.4 ±3.9 cm), with a larger (P<0.05) reduction in
CON than in RW (Fig. 2(b)). Before the start of match,
there was a significant correlation between sprint and
CMJ performance during both RW and CON (r=−0.52,
P<0.05 and r=−0.58, P<0.05 for RW and CON,
respectively), whereas no such correlations were
observed after the first half or after the half-time period.
Heart rate
The mean heart rate for the players during the two
matches were 162 ±10 and 164 ±10 beats/min for RW
and CON, respectively, which corresponded to
84% ±2% (RW) and 85% ±2% (CON) of maximal
heart rate. The average heart rate was higher (P<0.05)
during the first half compared with the second half of the
matches during both RW (167 ±7vs157±12 beats/
min) and CON (167 ±8vs161±11 beats/min). More-
over, the average heart rate was also higher (P<0.05)
during the first 5 and 15 min of the first half compared
with the second half both in RW (163 ±7vs159±10
and 168 ±8vs161±12 beats/min) and CON matches
(164 ±9 vs 158 ±14 and 168 ±9vs163±12 beats/
min). Immediately before start of the second half, the
heart rate was higher (P<0.05) after RW compared with
CON (117 ±10 vs 109 ±12 beats/min), and the time for
the heart rate to increase to values taxing average heart
rate of the second half was faster (P<0.05) after RW
(71 ±34 s) compared with CON (129 ±44 s; Fig. 3).
Weight loss
The average body weight loss for the players was
0.9 ±0.3 and 1.0 ±0.3 kg during RW and CON
#
*
#
*
#
(a) (b)
Pre 1st half
Time (min) Time (min)
Pre 2nd halfPost 1st half
0
Pre 1st half Pre 2nd halfPost 1st half
101
100
99
98
97
96
95
94
93
0
104
102
100
98
96
94
92
90
88
RW
CON
RW
CON
% of pre-match sprint performance
% of pre-match jump performance
Fig. 2. Ten-meter sprint performance (a) and countermovement jump performance (b) during soccer matches. Closed bars (■)
represent re-warm up (RW) condition and open bars (□) represent control (CON) condition during half-time rest. N=17. Data are
presented as means ±standard error of the mean. #P<0.05 post first half vs pre second half. *P<0.05 RW vs CON.
Edholm et al.
4
condition matches, respectively. The fluid intake was
0.3 ±0.2 kg (RW) and 0.3 ±0.2 kg (CON); thus, the cal-
culated total fluid loss was 1.2 ±0.4 kg (1.5% ±0.3% of
body weight) and 1.3 ±0.3 kg (1.7% ±0.3% body
weight) during RW and CON, respectively. There were
no significant differences between RW and CON condi-
tion matches regarding fluid loss.
Match activities, distance covered, and technical skill
Total distance covered during the initial part of the
second half was significantly shorter than during the first
half for both RW (9%, P<0.001) and CON (4%,
P<0.05), with a larger drop in RW compared with CON
(P<0.05, η2=0.02, ES =small). No significant interac-
tion was observed for distance covered with high-
intensity running (P=0.294) or sprinting (P=0.804).
Re-warm up during half time had no effect on the
number of high-intensity runs or sprints as no
time ×condition interaction was found. When focusing
on the initial 5 min of each half the statistical
analysis showed no significant time ×condition
interactions (Table 1).
Defensive and offensive high-intensity runs
For the distance covered with defensive high-intensity
running, a significant time ×group interaction was
observed (P<0.05, η2=0.06, ES =medium) and post-
hoc tests showed that this distance was lower in the
initial 15 min of the second compared with the first half
(0.14 ±0.06 vs 0.21 ±0.07 km, P<0.01) in RW but
not in CON (0.22 ±0.07 vs 0.21 ±0.11 km). The
defensive high-intensity running distance was different
between the two conditions in the second (P<0.01)
but not in the first half (P=0.937). These differences
between the two conditions were due to the number
of defensive high-intensity runs, which decreased
(P<0.05) in the second half for RW (13.9 ±5.7 vs
22.3 ±7.8) but not for CON (21.1 ±4.8 vs
22.9 ±10.6), whereas no difference was observed in the
mean length of the defensive high-intensity runs for
either RW (10 ±2vs10±2 m) or CON (10 ±2vs
9±2 m). No difference was observed in offensive
high-intensity running in the initial 15 min of each
half. In the initial 5 min of each half, a significant
interaction was found for the distance covered with
defensive high-intensity running (P<0.05, η2=0.06,
ES =medium). Post-hoc tests showed that the distance
tended (P=0.095) to be lower in the second half
compared with the first half in RW (0.05 ±0.04 vs
0.08 ±0.03 km) but not in CON (0.08 ±0.04 vs
0.08 ±0.04 km) with a significant difference between
RW and CON in the second (P<0.01) but not in the
first half (P=0.683). No difference was observed in
the number of defensive high-intensity runs or the
mean distance of those runs. Moreover, no significant
interactions were found for the distance covered with
or the number of offensive high-intensity running
during the initial 5 min of each half.
Effective playing time
The effective playing time was 69% ±1% for the initial
15 min of the second half and 80% ±0% for the similar
period in the first half. Also for the initial 5 min of each
half less effective playing time was observed for the
second half compared with the first half (67% ±10% vs
75% ±6%).
(a)
Heart rate (beats/min)
(b)
*
0
180
170
160
150
140
130
120
110
30 60 90
Time (s)
120 150 1800
Mean RW
Mean CON
Heart rate (beats/min)
0
180
170
160
150
140
130
120
110
30 60 90
Time (s)
120 150 1800
Mean RW
Mean CON
Fig. 3. Mean heart rate kinetics response during the initial 3 min of (a) first half and (b) second half. Dashed line ( ) represent
re-warm up (RW) condition and dotted line ( ) represent control (CON) condition during half-time rest. N=17. Data are presented
as means ±standard error of the mean. *P<0.05 RW vs CON.
Half-time re-warm up and soccer performance
5
Match activities adjusted for effective playing time
When the match activities were related to effective
playing time and analyzed as distance covered per min
effective playing time (MEPT), a significant time ×con-
dition interaction was observed for total distance covered
(P<0.05, η2=0.02, ES =small) and post-hoc test
showed that a significant longer total distance was
covered in the first 15 min of the second half compared
with the same period in the first half for both RW
(0.17 ±0.02 vs 0.16 ±0.01 m per MEPT, P<0.05) and
CON (0.19 ±0.02 vs 0.17 ±0.02 m per MEPT,
P<0.001), with a shorter distance covered in RW than
CON in the first 15 min of the second half (P<0.001)
but not in the first half (P=0.120). In accordance
with the main analysis, high-intensity runs (P<0.05,
η2=0.05, ES =small) and sprinting (P<0.05,
η2=0.02, ES =small) were performed less frequent in
RW than in CON (3.3 ±0.7 vs 3.8 ±1.3 and 0.5 ±0.2 vs
0.6 ±0.2 times per MEPT).
Ball possession and specific actions
Ball possession increased in RW in both the initial 5 and
15 min in the second half compared with the correspond-
ing period in the first half. Moreover, the ball possession
increased in both the attacking and defending zone,
whereas ball possession in the neutral zone decreased in
both periods (Table 2).
The number of ball entries into the attacking zone was
12 and 11 in RW and 8 and 11 in CON in the first 15 min
of the second half compared with the similar period of
first half. In the initial 5 min of the second half and the first
half, the number of ball entries into the attacking zone was
4 and 3 in RW and 2 and 3 in CON, respectively.
The number of passes and dribbles was lower in the
second half compared with the first half for RW (100 ±4
vs 113 ±13 and 20 ±2vs26±11, respectively) and
CON (89 ±7vs108±9 and 17 ±4vs22±11), whereas
2±1 shots on goal in the second and 1 ±2 shots on goal
in the first half were observed in both conditions.
Table 1. Distance covered with and number of occurrences of sprinting, high-intensity running, and in total during the initial 5- and 15-min periods in the
first and second half after active half-time re-warm up (RW) or a passive half-time period (CON)
Distance (km) Sprinting High-intensity running Total
RW CON RW CON RW CON
First half 0–5 min 0.03 ±0.02 0.03 ±0.02 0.13 ±0.05 0.15 ±0.07 0.64 ±0.07 0.68 ±0.07
5–10 min 0.03 ±0.02 0.04 ±0.04 0.14 ±0.06 0.16 ±0.08 0.68 ±0.07 0.70 ±0.09
10–15 min 0.03 ±0.02 0.03 ±0.02 0.11 ±0.04 0.10 ±0.06 0.63 ±0.05 0.63 ±0.06
0–15 min 0.08 ±0.03 0.10 ±0.04 0.37 ±0.11 0.41 ±0.19 1.95 ±0.15 2.01 ±0.18
Second half 45–50 min 0.03 ±0.02 0.05 ±0.03 0.13 ±0.05 0.15 ±0.06 0.60 ±0.07 0.67 ±0.07
50–55 min 0.02 ±0.02 0.02 ±0.02 0.09 ±0.04 0.12 ±0.05 0.58 ±0.07 0.63 ±0.06
55–60 min 0.03 ±0.02 0.04 ±0.02 0.12 ±0.05 0.15 ±0.06 0.59 ±0.07 0.63 ±0.09
45–60 min 0.08 ±0.03 0.10 ±0.04 0.34 ±0.03 0.42 ±0.14 1.77 ±0.17‡1.92 ±0.18*†
No. of occurrences (n)
First half 0–5 min 1.8 ±1.4 2.4 ±1.5 14.1 ±6.4 15.7 ±6.7 102 ±14 102 ±9
5–10 min 1.8 ±1.4 2.8 ±1.9 14.3 ±3.6 15.8 ±6.6 104 ±9 102 ±13
10–15 min 1.9 ±1.1 1.7 ±1.2 11.3 ±4.1 11.5 ±5.8 95 ±893±11
0–15 min 5.4 ±2.8 6.9 ±2.4 39.6 ±10.4 43.0 ±17.4 301 ±19 296 ±30
Second half 45–50 min 2.0 ±1.2 2.4 ±1.5 12.6 ±4.7 14.7 ±6.7 97 ±12 101 ±13
50–55 min 1.1 ±1.1 1.4 ±1.9 8.3 ±3.2 12.3 ±6.6 82 ±12 88 ±12
55–60 min 1.9 ±1.1 2.1 ±1.2 12.4 ±3.7 13.7 ±5.8 88 ±10 90 ±13
45–60 min 5.1 ±1.9 6.0 ±1.7 33.2 ±6.9 40.7 ±11.7 267 ±27 278 ±34
N= 17. Distance covered (km) and number of occurrences (n) are expressed as average per player. Data are presented as means ±standard deviation.
*Denotes significantly different from RW (P<0.001).
†Denotes significantly different from first half (P<0.05).
‡Denotes significantly different from first half (P<0.001).
Table 2. Possession of the ball during the initial 15 min of both halves in game with active half-time re-warm up (RW) or a passive half-time period (CON)
in total and possession of the ball in defending, neutral, and attacking zone
RW CON
Total
possession (%)
Defending
(%)
Neutral
(%)
Attacking
(%)
Total
possession (%)
Defending
(%)
Neutral
(%)
Attacking
(%)
0–15 min first half 53.7 20.1 24.2 9.5 46.3 18.2 20.6 7.5
0–15 min second half 57.2 23.5 22.3 11.2 43.0 14.2 19.8 8.9
0–5 min first half 46.7 19.0 21.9 5.8 53.3 22.0 22.2 8.8
0–5 min second half 55.2 27.8 19.9 7.5 44.8 18.9 17.6 8.3
Edholm et al.
6
Discussion
The main findings of the current study were that an
active half-time re-warm up (RW) regime attenuated the
decreases in sprint and jump performances observed
after traditional passive half-time period (CON) in pro-
fessional soccer match play. These findings are in line
with previous results obtained for sub-elite players
(Mohr et al., 2004; Lovell et al., 2013) and emphasized
that these types of performance decrements also occur in
well-trained professional soccer players during actual
match play.
The magnitude of the impaired sprint performance
after passive half-time period in this study (CON, 2.6%)
was similar to that reported by Mohr et al. (2004) (2.6%)
but lower than Lovell et al. (2013) (6.2%). These differ-
ences might be attributed to several different factors,
including the fact that Lovell et al. (2013) used a soccer
match simulation agility course to mimic a soccer game
instead of investigating changes in the player’s physical
performance capacity in relation to a real soccer match
play. Thus, the players movement pattern and exercise-
induced fatigue might have differed from actual soccer
match play. However, other factors such as starting pro-
cedure during the sprint test (3.0 vs 0.5 m rolling start)
and differences in the player’s physical capacity (sub-
elite vs professional soccer players) might also contrib-
ute to differences in sprint performance deteriorations.
In the study by Mohr et al. (2004), sprint performance
was measured through a repeated sprint test (3 ×30 m,
rest 25 s). However, movement analysis has revealed that
the mean distance of a single sprint during a soccer game
is 12–15 m and it is very seldom that players performed
high-intensity running over 30 m or longer (Reilly &
Thomas, 1976; Bangsbo et al., 1991; Valquer et al.,
1998; Mohr et al., 2003; Di Salvo et al., 2009).
Previous studies have also shown that while profes-
sional soccer players have a significantly better sprint
capacity over short distances (e.g., 10 m) than sub-elite
soccer players, such patterns do not seem to exist at
longer distance (e.g., 30 m) (Cometti et al., 2001; Stolen
et al., 2005). This suggests that long sprint distances
such as a 30-m dash do not mirror actual game situa-
tions. Thus, from a physiological point of view it might
be argued that data on performance capacity over shorter
distance such as 10 m that were used in the current study
are more relevant if the aim is to examine professional
soccer players’ maximal sprint performance capacity.
Nevertheless, result from the current study supports the
previous findings by Mohr et al. (2004) and Lovell et al.
(2013) in that a short, low-intensity re-warm up program
is an efficient method to limit the decrements in sprint
performances observed during traditional passive half-
time period in soccer players.
In the present study, jump performance was signifi-
cantly reduced after the half-time period during both
RW and CON condition (pre half-time period vs post
half-time period). However, the RW regime was associ-
ated with a reduction in the CMJ performance deterio-
ration as the decrease jump height was significantly
lower after RW compared with CON condition (3.1% vs
7.6%). This indicates that the RW regime at least to some
extent also had a positive effect on preserving jump
capacity over the half-time period in professional soccer
players. Studies investigating the effect of different
warm up protocols on jump performance in well-trained
athletes have shown that warm up protocols, which
include some brief but high-loading exercises such as
jumping or squats with heavy load, are more efficient in
improving the jumping ability than a traditional warm up
(5-min jogging and some half-squats with low load)
(Saez Saez de Villarreal et al., 2007). In line with this,
Gullich and Schmidtbleicher (1996) suggest that in
addition to elevated muscle temperature, the warm
up-induced increases in jump performance might also be
ascribed to an increased recruitment of higher order
motor units. Thus, it might be that the RW regime, which
consisted of low-intensity jogging and light calisthenics,
should be modified to contain some brief but more
neuromuscularly challenging stimuli such as squat
jumping in order to increase its efficiency regarding
preservation of the players jump performance.
After RW players restarted the game with higher heart
rate levels than after CON condition (+7.3%) and were
able to reach mean second half heart rate values faster
than after CON (Fig. 3(b)). Previous studies have
reported that a close relationship exists between heart
rate and oxygen uptake responses during varying non-
steady states activities (Bot & Hollander, 2000) and
faster oxygen uptake kinetics and the ability to maintain
sprint performance during repeated sprints (Dupont
et al., 2010). Therefore, the association between RW and
both faster recruitment of the aerobic system and
reduced contribution of anaerobic metabolism to meet
the metabolic demand appears to be a reasonable
hypothesis.
In this study, the muscle temperature of the players
was not measured and this could be regarded as a limi-
tation. The reason for not including such measurement
was that this procedure would have either required an
extended half-time period or that the players were tem-
porarily taken out of the game for sampling, as per-
formed in the study by Mohr et al. (2004). As one of the
aims of the present study was to investigate the re-warm
up concept during real match play and to evaluate its
potential effect on movement pattern both of those
options were excluded. Moreover, previous studies have
repeatedly shown that muscle temperature decreases
during passive half-time rest and that a RW regime such
as that utilized in this study inhibits such decreases
(Mohr et al., 2004; Lovell et al., 2013). Mohr et al.
(2004) also reported that there was a correlation between
changes in muscle temperature and repeated sprint per-
formance over the half-time period, suggesting that
Half-time re-warm up and soccer performance
7
muscle temperature has a key role in the reduced physi-
cal performance capacity seen in soccer players after the
half-time period. However, other mechanisms such as
increased neuromuscular function, elevated blood flow
in working muscle, and oxygen uptake kinetics could
also be involved (Hodgson et al., 2005). Nevertheless,
the exact mechanism through which soccer players’ per-
formance capacity is compromised during passive half-
time rest, and consequently that a RW regime effectively
attenuates remains unclear, thus warrants further
research.
The second key finding of this study was that the total
distance covered was lowered from the initial 15 min of
first half to the initial 15 min of the second half in both
RW and CON, but to a higher extent in RW (9% vs 4%,
respectively). Interestingly, no significant interaction
was found for distance covered with high-intensity
running and sprinting; thus, high-intensity running was
not lowered in the initial 15 min of the second half as
often observed (Mohr et al., 2003, 2005; Bradley et al.,
2009; Weston et al., 2011) in neither RW nor CON.
The lower physical match performance observed in
RW may be related to lower physical demands as it was
observed that ball possession was higher in RW. A pre-
vious study has shown that a team’s physical perfor-
mance is related to the level and performance of the
opposing team (Rampinini et al., 2007). Moreover,
another study by the same research group showed that
successful teams covered less total distance, high-
intensity running, and very high-intensity running than
less successful teams, but covered significantly more
distance with possession of the ball in the same catego-
ries (Rampinini et al., 2009). Similar findings have been
shown in another study (Di Salvo et al., 2009), indicating
that high-intensity running distance covered with or
without the ball may be a better indicator of performance
although it should be noticed that ES were small to
moderate. Thus, the shorter distance observed in RW in
this present study may simply be a consequence of their
higher percentage of ball possession and therefore higher
physical demands put on CON chasing and trying to
regain possession of the ball. It should however be noted
that the ES on almost all workload parameters were only
small. Although the used design tries to limit confound-
ing variables, match-to-match variation in workload,
especially high-intensity running and sprinting, between
consecutive soccer matches has been shown to be high,
which complicates the use of real matches when evalu-
ating interventions and therefore possibly resulting in
small ESs (Gregson et al., 2010).
Few studies have highlighted the importance of relat-
ing workload to the effective playing time instead of only
total playing time as these parameters may vary from
game to game and throughout the periods of the game
(Castellano et al., 2011; Dellal et al., 2013). In this study,
the effective playing time was 11% and 8% lower in the
initial 15- and 5-min period of the second half compared
with the corresponding periods in first half. However, as
most high-intensity running is performed during effec-
tive playing time (Castellano et al., 2011), we divided
distances covered with the effective playing time in the
respective periods as applied by others (Lago-Peñas
et al., 2012; Dellal et al., 2013). When related to the
effective playing time, more total distance was covered
in the first 15 min of the second half compared with the
similar period in the first half, which is in contrast to
when analyzed as total time. Moreover, when related to
effective playing time more high-intensity running and
sprinting was performed in the initial phase of the
second half than in the first half. Although relating work-
load to effective playing time changed the results regard-
ing differences between the first and second half, no shift
was observed in the relation between RW and CON.
Thus, CON had a higher workload than RW during the
initial part of the second half. This may also be explained
by the higher ball possession in RW, and thereby more
ball chasing by CON. Moreover, these results highlight
the importance of taking effective playing time into
account when comparing workload between different
periods of a match or between matches.
It might be argued that the type of RW regime used in
this study could result in accumulated fatigue and thus
compromise the player’s ability to perform maximally
during the later stage of the game (e.g., through acceler-
ated glycogen depletion (Gollnick et al., 1973), dehydra-
tion (Edwards et al., 2007), and/or increased thermal
heat strain (Nadel et al., 1980). However, results from
both Mohr et al. (2004) and Lovell et al. (2013) contra-
dict this speculation as no differences in muscle tem-
perature or sprint performance were observed during the
later stage of the game when comparing traditional
passive half-time period and an active re-warm up
regime. Moreover, although HR was elevated after per-
forming RW, no differences in HR were observed during
the later part of the second half and no differences in
fluid loss (1.5% vs 1.7% of body weight) were seen
between RW and CON matches. Accordingly, this type
of short-to-moderate intensity RW regime does not
induce additional fatigue and therefore does not compro-
mise the player’s performance capacity during later
stage of match.
Although outside the scope of this study and accord-
ingly not investigated here, it is very interesting to note
that several researchers have reported an increased risk
of muscle injury (i.e., non-contact injuries) during the
initial stage of the second half of match play in soccer
(Hawkins & Fuller, 1996; Rahnama et al., 2002). As it is
well known that a decrement in muscle temperatures
such as those typically reported after passive half-time
period in soccer players (1.5–2.0 °C) (Mohr et al., 2004;
Lovell et al., 2013) is associated with an increase risk of
muscle injuries (Safran et al., 1989), and that a re-warm
up regime such as that utilized in this study attenuates
such decrements (Mohr et al., 2004; Lovell et al., 2013),
Edholm et al.
8
it might be speculated that half-time re-warm up also is
an effective method to prevent muscle injuries during the
initial phase of the second half. Thus, the potential of
half-time re-warm up regimes as an injury prevention
tool in professional soccer warrants further research.
In summary, the present study demonstrates that a
traditional passive half-time period during soccer match
play results in impaired sprint and jump performance
during the initial phase of the second half in professional
soccer players and that a re-warm up protocol effectively
attenuates such deteriorations. Despite an increased
physical work capacity, RW did not increase the amount
of high-intensity running performed by the players
during the initial phase of the second half. Nevertheless,
time–motion analysis did reveal a higher ball possession
and less defensive high-intensity sprints after RW, which
indicates a game advantage at the onset of the second
half.
Perspectives
This study adds to the growing body of evidence
showing that a traditional passive half-time period
during soccer match causes temporary impairment in the
players’ physical performance capacity and, accord-
ingly, is not an optimal procedure to prepare the players
for the second half of the game. The study also shows
that if a short-duration low-intensity re-warm up regime
is performed during the second part of the half-time
period, such deteriorations in performance capacity are
diminished or neutralized. Moreover, time–motion
analysis revealed that inclusion of such a re-warm up
regime might result in game advantages as players
increased their ball possession and reduced their amount
of defensive high-intensity running during the initial
phase of the second half after re-warm up. Taken
together, the enhanced performance capacity in profes-
sional soccer players after an active re-warm up regime
might contribute to decisive actions during the initial
phase of the second half of the game. As the initial phase
of the second half during soccer match also has been
identified as a period of increased “vulnerability” to inju-
ries (Reilly & Hardiker, 1981), the results from this
study provide a strong rationale for professional soccer
team to evaluate and adapt their half-time period regime
so that the players are optimally physically prepared to
perform during the first part of the second half in com-
petitive matches.
Key words: Football, high-intensity exercise, sprint per-
formance, countermovement jump, time–motion analysis.
Acknowledgements
The authors would like to gratefully acknowledge the assistance
from Jimmy Högberg, Johan Svensson, and Mattias Olsson; the
cooperation with the coaches and soccer players participating in
this study; and the financial support from the Swedish Football
Association.
References
Bangsbo J, Norregaard L, Thorso F.
Activity profile of competition soccer.
Can J Sport Sci 1991: 16: 110–116.
Bishop D. Warm up II: performance
changes following active warm up and
how to structure the warm up. Sports
Med 2003: 33: 483–498.
Bot SD, Hollander AP. The relationship
between heart rate and oxygen uptake
during non-steady state exercise.
Ergonomics 2000: 43: 1578–1592.
Bradley PS, Carling C, Archer D, Roberts
J, Dodds A, Di Mascio M, Paul D,
Diaz AG, Peart D, Krustrup P. The
effect of playing formation on
high-intensity running and technical
profiles in English FA Premier League
soccer matches. J Sports Sci 2011: 29:
821–830.
Bradley PS, Sheldon W, Wooster B,
Olsen P, Boanas P, Krustrup P.
High-intensity running in English FA
Premier League soccer matches. J
Sports Sci 2009: 27: 159–168.
Carling C, Bloomfield J, Nelsen L, Reilly
T. The role of motion analysis in elite
soccer: contemporary performance
measurement techniques and work
rate data. Sports Med 2008: 38:
839–862.
Castellano J, Blanco-Villasenor A,
Alvarez D. Contextual variables and
time–motion analysis in soccer. Int J
Sports Med 2011: 32: 415–421.
Cohen J. Statistical power analysis for the
behavioral sciences (2nd ed.). Hillsdale,
NJ: Lawrence Erlbaum, 1988: 273–288.
Cometti G, Maffiuletti NA, Pousson M,
Chatard JC, Maffulli N. Isokinetic
strength and anaerobic power of elite,
subelite and amateur French soccer
players. Int J Sports Med 2001: 22:
45–51.
Dellal A, Lago-Penas C, Rey E, Chamari
K, Orhant E. The effects of a congested
fixture period on physical performance,
technical activity and injury rate during
matches in a professional soccer team.
Br J Sports Med 2013: doi:
10.1136/bjsports-2012-091290
Di Salvo V, Gregson W, Atkinson G,
Tordoff P, Drust B. Analysis of high
intensity activity in Premier League
soccer. Int J Sports Med 2009: 30:
205–212.
Dupont G, McCall A, Prieur F, Millet GP,
Berthoin S. Faster oxygen uptake
kinetics during recovery is related to
better repeated sprinting ability. Eur J
Appl Physiol 2010: 110: 627–634.
Edwards AM, Mann ME, Marfell-Jones
MJ, Rankin DM, Noakes TD,
Shillington DP. Influence of moderate
dehydration on soccer performance:
physiological responses to 45 min
of outdoor match-play and the
immediate subsequent performance of
sport-specific and mental concentration
tests. Br J Sports Med 2007: 41:
385–391.
Ekblom B. Applied physiology of soccer.
Sports Med 1986: 3: 50–60.
Gollnick PD, Armstrong RB, Sembrowich
WL, Shepherd RE, Saltin B.
Glycogen depletion pattern in human
skeletal muscle fibers after heavy
exercise. J Appl Physiol 1973: 34:
615–618.
Gregson W, Drust B, Atkinson G, Salvo
VD. Match-to-match variability of
high-speed activities in premier league
soccer. Int J Sports Med 2010: 31:
237–242.
Gullich A, Schmidtbleicher D. MVC-
induced short-term potentiation of
explosive force. New Stud Athletics
1996: 11: 67–81.
Hawkins RD, Fuller CW. Risk assessment
in professional football: an examination
of accidents and incidents in the 1994
Half-time re-warm up and soccer performance
9
World Cup finals. Br J Sports Med
1996: 30: 165–170.
Hodgson M, Docherty D, Robbins D.
Post-activation potentiation: underlying
physiology and implications for motor
performance. Sports Med 2005: 35:
585–595.
Ingebrigtsen J, Bendiksen M, Randers
MB, Castagna C, Krustrup P,
Holtermann A. Yo-Yo IR2 testing of
elite and sub-elite soccer players:
performance, heart rate response and
correlations to other interval tests. J
Sports Sci 2012: 30: 1337–1345.
Lago-Peñas C, Rey E, Lago-Ballesteros J.
The influence of effective playing time
on physical demands of elite soccer
players. Open Sports Sci J 2012:
188–192.
Lovell R, Midgley A, Barrett S, Carter D,
Small K. Effects of different half-time
strategies on second half soccer-specific
speed, power and dynamic strength.
Scand J Med Sci Sports 2013: 23:
105–113.
Lovell RJ, Kirke I, Siegler J, McNaughton
LR, Greig MP. Soccer half-time
strategy influences thermoregulation
and endurance performance. J Sports
Med Phys Fitness 2007: 47: 263–269.
Mohr M, Krustrup P, Bangsbo J. Match
performance of high-standard soccer
players with special reference to
development of fatigue. J Sports Sci
2003: 21: 519–528.
Mohr M, Krustrup P, Bangsbo J. Fatigue
in soccer: a brief review. J Sports Sci
2005: 23: 593–599.
Mohr M, Krustrup P, Nybo L, Nielsen JJ,
Bangsbo J. Muscle temperature and
sprint performance during soccer
matches – beneficial effect of
re-warm-up at half-time. Scand J Med
Sci Sports 2004: 14: 156–162.
Mugglestone C, Morris JG, Saunders B,
Sunderland C. Half-time and
high-speed running in the second half
of soccer. Int J Sports Med 2013: 34:
514–519.
Nadel ER, Fortney SM, Wenger CB.
Effect of hydration state of circulatory
and thermal regulations. J Appl Physiol
1980: 49: 715–721.
Rahnama N, Reilly T, Lees A. Injury risk
associated with playing actions during
competitive soccer. Br J Sports Med
2002: 36: 354–359.
Rampinini E, Coutts AJ, Castagna C,
Sassi R, Impellizzeri FM. Variation in
top level soccer match performance. Int
J Sports Med 2007: 28: 1018–1024.
Rampinini E, Impellizzeri FM, Castagna
C, Coutts AJ, Wisloff U. Technical
performance during soccer matches of
the Italian Serie A league: effect of
fatigue and competitive level. J Sci
Med Sport 2009: 12: 227–233.
Reilly T, Hardiker R. Somatotype and
injuries in adult student rugby footfall.
J Sports Med Phys Fitness 1981: 21:
186–191.
Reilly T, Thomas V. A motion analysis of
work-rate in different positional roles in
professional football match-play. J
Hum Mov Stud 1976: 87–97.
Saez Saez de Villarreal E,
Gonzalez-Badillo JJ, Izquierdo M.
Optimal warm-up stimuli of muscle
activation to enhance short and
long-term acute jumping performance.
Eur J Appl Physiol 2007: 100:
393–401.
Safran MR, Seaber AV, Garrett WE Jr.
Warm-up and muscular injury
prevention. An update. Sports Med
1989: 8: 239–249.
Sargeant AJ. Effect of muscle temperature
on leg extension force and short-term
power output in humans. Eur J Appl
Physiol Occup Physiol 1987: 56:
693–698.
Stewart IB, Sleivert GG. The effect of
warm-up intensity on range of
motion and anaerobic performance.
J Orthop Sports Phys Ther 1998: 27:
154–161.
Stolen T, Chamari K, Castagna C,
Wisloff U. Physiology of soccer:
an update. Sports Med 2005: 35:
501–536.
Towlson C, Midgley AW, Lovell R.
Warm-up strategies of professional
soccer players: practitioners’
perspectives. J Sports Sci 2013: 31:
1393–1401.
Valquer W, Barros T, Sant’anna M. High
intensity motion pattern analyses of
Brazilian elite soccer players. In:
Tavares F, ed. IV world congress of
notational analysis in sports. Porto,
Brazil: HM Tavares Fernando, 1998:
80: 23–27.
Weston M, Batterham AM, Castagna C,
Portas MD, Barnes C, Harley J, Lovell
RJ. Reduction in physical match
performance at the start of the second
half in elite soccer. Int J Sports Physiol
Perform 2011: 6: 174–182.
Zois J, Bishop D, Fairweather I, Ball K,
Aughey RJ. High-intensity re-warm-ups
enhance soccer performance. Int J
Sports Med 2013: 34: 800–805.
Edholm et al.
10