• Nenhum resultado encontrado

Lower limb muscle activation patterns in ice-hockey skating and associations with skating speed

N/A
N/A
Protected

Academic year: 2023

Share "Lower limb muscle activation patterns in ice-hockey skating and associations with skating speed"

Copied!
28
0
0

Texto

(1)

1

Lower Limb Muscle Activation Patterns in Ice-Hockey Skating

1

and Associations with Skating Speed

2

Sami Kaartinen1, Mika Venojärvi1, Kim Lesch1, Heikki Tikkanen1, Paavo Vartiainen2, 3

Lauri Stenroth2 4

1 Institute of Biomedicine, Sports and Exercise Medicine, School of Medicine, University 5

of Eastern Finland, Kuopio, Finland; 2 Department of Applied Physics, University of 6

Eastern Finland, Kuopio, Finland 7

8

ORCiDs:

9

Sami Kaartinen: https://orcid.org/0000-0002-6962-1054 10

Mika Venojärvi: https://orcid.org/0000-0003-1327-9760 11

Paavo Vartiainen: https://orcid.org/0000-0003-0974-0913 12

Lauri Stenroth: https://orcid.org/0000-0002-7705-9188 13

Heikki Tikkanen: https://orcid.org/ 0000-0001-6067-4101 14

15

Corresponding author:

16

Sami Kaartinen 17

University of Eastern Finland 18

School of Medicine 19

PO Box 1627 20

70211 Kuopio, Finland 21

tel. +358456935233, email: [email protected] 22

23

(2)

2 Abstract

24

In this study, we aimed to describe lower limb kinematic and muscle activation patterns 25

and then to examine the potential associations between those variables and skating speed 26

in highly trained ice-hockey players. Twelve players (age 18.4-22.0 years) performed five 27

maximal 30-meter forward skating sprints. Skating speeds, muscle activities from eight 28

lower limb muscles (gluteus maximus, gluteus medius, adductor magnus, rectus femoris, 29

vastus lateralis, biceps femoris, tibialis anterior, and soleus), and sagittal plane joint 30

angles from the hip and knee joint were measured. A lower activity of the gluteus 31

maximus (r=-0.651, p=0.022, β=-0.08) and a reduced gluteus maximus to rectus femoris 32

coactivity (r=-0.786, p=0.002, β=-3.26) during the recovery phase were found to be 33

associated with faster skating speed. No significant associations were observed between 34

sagittal plane hip and knee kinematics and skating speed. This study provides evidence 35

that muscle activities during the recovery phase of skating may have an important role in 36

skating performance.

37 38

Key Words: ice-hockey, recovery phase, skating cycle, biomechanics, EMG, 39

electromyography 40

(3)

3 Introduction

41

Ice-hockey is a fast team sport in which good skating skills and fast skating speed are 42

important characteristics for the players. To enhance training in ice-hockey, it would be 43

important to understand the technical determinants of fast skating performance. Lower 44

limb kinematics and muscle activity patterns can be considered as key technical 45

determinants of a motor task where lower limb kinematics represent the movement 46

pattern and muscle activity the cause of the movement pattern.

47

During the skating cycle, there is a movement of the hip joint in sagittal, frontal and 48

horizontal planes. In addition, the knee joint extends and flexes whereas the ankle joint 49

dorsiflexes and plantarflexes during different phases of the skating cycle (Haché, 2002;

50

Goudreault 2002). When the skating speed increases, greater hip abduction occurs during 51

the propulsion phase (skate is in contact with the ice, more details in Supplementary 52

material, Pearsall et al., 2000; Haché, 2002). It could be summarized that skating consists 53

of complex movement patterns, which pose special demands on motor skills (Haché, 54

2002).

55

Previous studies of muscle activity during skating in ice-hockey have indicated that the 56

gluteus maximus (Pearsall et al., 2000) and the vasti muscles are active during the 57

propulsion phase to extend, abduct and externally rotate the hip and extend the knee, 58

respectively (Chang et al., 2009; Buckeridge et al., 2015). Hamstring muscles are reported 59

to be most active during the gliding phase (isometric phase) of skating to increase the 60

stiffness of the knee joint with co-activity of knee extensors (de Boer et al., 1987).

61

Additionally, the biceps femoris shows high activity also during the propulsion phase in 62

ice-hockey skating when extension of the hip occurs (Chang et al., 2009). The activity of 63

(4)

4

the tibialis anterior is at its highest during the gliding phase to stabilize the ankle and 64

during the recovery phase to dorsiflex the ankle (Goudreault 2002).

65

Rather few studies have examined how the previously described kinematic characteristics 66

of ice-hockey skating are related to the skating speed. During the propulsion phase, 67

greater range of motion (ROM) of the hip joint (Upjohn et al., 2008), and more extended 68

knee seem to be important determinants of higher skating speed in ice-hockey players 69

(Buckeridge et al., 2015). During the recovery phase (skate is not in contact with the ice, 70

more details in Supplementary material) greater hip adduction was found to be associated 71

with a higher skating velocity (Lafontaine, 2007). Since the most above-mentioned 72

studies included players from two contrasting groups (different levels of players) one 73

needs to be careful when interpreting causality from the observed associations. However, 74

one could summarize the findings to indicate that large hip ROM during the propulsion 75

phase, may be an important determinant of fast skating speed.

76

While there is little information available on the association between the kinematic 77

characteristics of skating and skating speed in ice-hockey, it is not known if differences 78

in muscle activities between the players explain differences in skating speed between the 79

players. Overall, there has been rather little research published on skating biomechanics 80

in ice-hockey and the available data on ice-hockey originates from players with variable 81

skill and performance levels (Robbins et al., 2018; Buckeridge et al., 2015; Upjohn et al., 82

2008). Hereby it is difficult for coaches to apply evidence-based protocols in the training 83

of elite ice-hockey players. Therefore, this study aimed first to describe the lower limb 84

muscle activation and joint kinematics patterns during the maximal forward skating phase 85

in highly trained ice-hockey players. Secondly, this study aimed to examine the 86

association between the skating speed and lower limb kinematics and muscle activities 87

during the different phases of the skating cycle.

88

(5)

5

We hypothesized that muscle activities of the lower limb muscles and ROM of the hip 89

and knee joint would be associated with a faster forward skating speed. Based on previous 90

studies suggesting greater ROM at the hip and knee during the propulsion phase to be 91

associated with greater skating speed (Buckeridge et al., 2015) we hypothesized that 92

greater hip and knee extension and high muscle activity of the respective agonist muscles 93

(vastus lateralis and gluteus maximus) during the propulsion phase would enable 94

powerful propulsion and thus be associated with a higher maximal skating speed. Fast 95

lower limb flexion in addition to the high activity of the respective agonist muscles (rectus 96

femoris, biceps femoris, and tibialis anterior) during the recovery phase was hypothesized 97

to also be associated with higher skating speed as it could shorten the duration of the 98

recovery phase enabling the skater to achieve higher skating cycle frequency (Lafontaine, 99

2007; Buckeridge et al., 2015) and increase propulsion power.

100

Methods 101

Participants 102

Participants were recruited from the elite Finnish hockey league or the elite Finnish junior 103

hockey league. Inclusion criteria were male gender, age over 18 years. Players with acute 104

injury and players recovering from a previous injury were excluded. This study was aimed 105

to be a preliminary investigation and therefore we set out to detect strong correlations 106

(r<0.65) between muscle activities and lower limb kinematics with skating speed. Based 107

on a priori sample size calculation, a sample size of N=16 is needed for achieving 80%

108

statistical power to detect strong correlations with a two-tailed test (Faul et al., 2007). We 109

recruited 17 participants. All recruited participants signed an informed consent to 110

demonstrate that they were aware of the physical demands of the testing protocols and 111

were willing to participate in the study as volunteers. The ethics committee of the Hospital 112

(6)

6

District of Northern Savo approved the study protocol. Four participants were not able to 113

participate due to acute injure after the recruitment, and data from one participant needed 114

to be excluded during the data analysis due to invalid data leaving us with a sample size 115

of N=12 (18.4 - 22.0 years, 81.0 ± 6.0 kg, 1.82 ± 0.03 m, and BMI 25.2 ± 1.6 kg/m2).

116

Protocol 117

The participants performed a warm-up session according to their own protocol with which 118

they had been familiarized in their training groups. After the warm-up, each participant 119

performed a 30-meter maximal sprint from a standing start at least five times. Additional 120

trials were performed if technical problems were noticed during the data collection if the 121

participant fell or stumbled during the sprint or if there was a technical error in the data 122

collection. The number of trials performed varied between five and ten. In the skating 123

test, participants wore a helmet, their own skates (which were sharped as they were used 124

to), gloves, and tracksuit. Recovery time between the skating sprints was 90 s.

125

Electromyography, acceleration and joint kinematics 126

Electromyography (EMG), acceleration, and electrogoniometer signals were collected 127

using a data logger Biomonitor ME6000 (Bittium Biosignals Oy, Finland) and sent 128

wirelessly to a laptop computer. The sampling frequency of the signals was 1000 Hz.

129

Surface EMG was used to measure muscle activity during the skating test. The electrodes 130

(measuring area 95 mm2, Ambu®, Denmark) were attached according to the guidelines 131

of the Seniam project (Hermens et al., 2000) on pre-determined muscles of the right leg 132

with a 22 mm distance between the electrodes. The selected muscles were adductor 133

magnus, rectus femoris, biceps femoris, vastus lateralis, tibialis anterior, soleus, gluteus 134

maximus, and gluteus medius. Since there are no guidelines for electrode placement on 135

the adductor magnus muscle in the Seniam guidelines, ultrasound imaging was used to 136

(7)

7

identify the most suitable electrode placement during the pilot measurements. The 137

placement was in the upper 1/3 of the thigh. An experienced researcher attached the 138

electrodes. The skin in the region of the electrodes was shaved, lightly rubbed with 139

sandpaper, and cleaned with alcohol before attachment of the electrodes, which were 140

secured using adhesive tape. After the electrodes were placed on the participant´s skin, 141

they performed maximal isometric contractions in order to collect EMG normalization 142

data. The following tasks were performed for EMG normalization: push against a wall 143

while standing on the ball of the foot, foot dorsiflexion in a seated posture with manual 144

resistance, a seated knee extension with the shank fixed, knee flexion in a prone posture 145

with manual resistance, hip flexion against manual resistance in a supine posture, hip 146

extension in a prone posture, hip abduction against manual resistance lying on the side, 147

and hip adduction in a semi-seated posture against a foam roll held between the thighs.

148

For each maximal isometric muscle testing, the participant held the maximal contraction 149

for three seconds with strong verbal encouragement by the researcher. The participants 150

executed maximum voluntary isometric contraction (MVIC) for EMG normalization and 151

each MVIC task was performed twice with a 90 s rest between the contractions. EMG 152

activities of each muscle recorded during the skating were normalized to the maximum 153

values obtained during these contractions irrespective of in which task the maximum 154

values had been obtained.

155

A triaxial accelerometer (Bittium Biosignals Oy, Finland) was attached on the lateral side 156

of the plastic part of the right foot skate using adhesive tape, to measure accelerations of 157

the skate for detecting the phases of skating cycles (Buckeridge et al., 2015).

158

During skating knee and hip joint angles were measured using electrogoniometers 159

(Bittium Biosignals Oy, Finland) which were attached to the right leg using tape and a 160

bandage. The arms of the electrogoniometers were placed over the distal part of the femur 161

(8)

8

and the proximal part of the tibia for the knee joint, the arms crossed the joint in the 162

sagittal plane. Care was taken to position the electrogoniometers on the mid-lateral aspect 163

of the leg in order to measure the sagittal plane angles. Goniometer signals were zeroed 164

to the values obtained while the participants were standing with hips and knees straight 165

(i.e. anatomical position).

166

Data analysis and reductions 167

The first five successful sprints were selected for the analysis from each participant. The 168

last two full skating cycles (occurring within the final 15 meters) of the 30-meter skating 169

were extracted from the data based on detection of the cycles from the acceleration signal 170

and voltage pulse sent to the data logger Biomonitor ME6000 (Bittium Biosignals Oy, 171

Finland) when the skater passed the photocells at the end of the 30-meter sprint. Hence, 172

a total of 10 skating cycles were extracted for analysis from each participant.

173

The 30-meter skating time, characterizing the performance of the 30-meter sprint was 174

measured using photocells (Chronojump Boscosystem®, Spain) which were placed 25 175

cm from the starting line in the direction of skating and 25 cm behind the finish line. The 176

photocells were mounted 100 cm above the ice surface to avoid interference by the ice- 177

hockey stick. The average skating speed during the final 15 meters was measured by 178

analyzing the time taken to skate the final 15 meters from a high-speed (frame rate 120 179

Hz) video (GoPro 3, GoPro Inc., USA). The calculated skating speed reflects near- 180

maximal skating speed and is later referred to as “skating speed15-30m“. The camera was 181

located in the middle of the 30-meter sprint, 16 meters from the line of skating with its 182

field of view perpendicular to the skating direction. Vertical poles were positioned 183

between the camera and the skating line to mark specific distances over the 30-meter 184

sprint. The camera parallax was considered in the location of the marks and consequently, 185

(9)

9

the 30-meter mark was positioned at a distance of 29.06 meters from the starting line 186

(Figure 1).

187

Accelerometer, goniometer, and EMG data processing were performed in Matlab (v.

188

R2014a, MathWorks Inc., USA) using custom-made scripts. Skating cycle detection was 189

based on acceleration signals, which allowed the identification of time instants for the 190

start of the cycles (skate comes into contact with the ice) and the start of the recovery 191

phase (skate leaves the ice). The acceleration signals were first high-pass filtered using a 192

300 Hz bidirectional 4th order Butterworth filter to accentuate high-frequency 193

acceleration peaks in the data. Then, the time instants of the beginning of the skate cycles 194

and the beginning of the recovery phases were manually detected from the signal. EMG 195

data were first band-pass filtered between 20 and 450 Hz using bidirectional 4th order 196

Butterworth filter after which the data were full-wave rectified and low pass filtered using 197

12 Hz bidirectional 4th order Butterworth filter to create EMG envelopes (Thelen et al., 198

2005). The EMG values obtained during the skating were normalized by dividing them 199

with the maximal values obtained during the maximal isometric contractions (processed 200

similarly as data from skating). Electrogoniometer signals were low pass filtered using a 201

5 Hz bidirectional 4th order Butterworth filter.

202

The EMG envelopes and electrogoniometer data were time normalized and averaged to 203

create ensemble averages of the ten skating cycles for each participant. The ensemble 204

average cycle was divided into propulsion (i.e. ice contact) and recovery phases (i.e.

205

swing) based on the average duration of the propulsion phase across the ten cycles. The 206

propulsion phase was further divided into three equally spaced phases (beginning, middle, 207

and end) and the recovery phase was divided into two equally spaced phases (beginning 208

and end). From the ensemble averages, the mean EMG values were calculated from the 209

propulsion and recovery phases and their sub-phases. In addition, the coactivity of the 210

(10)

10

agonist-antagonist pairs (gluteus maximus/rectus femoris, tibialis anterior/soleus, 211

adductor magnus/gluteus medius, and vastus lateralis/biceps femoris) was calculated as 212

the ratio of the normalised activities between the two muscles in each phase (Ervilha, 213

et.al., 2012). From the electrogoniometer data, hip and knee joint angles were extracted 214

from the beginning and the end of the propulsion and recovery phases. In addition, the 215

range of motion during both phases was calculated. Finally, the duration of each phase 216

was extracted, and the average distance traveled during the whole skating cycle and both 217

skating cycle phases were calculated based on the measured skating speed15-30m as 218

distance = velocity * duration.

219

Statistical analyses 220

Statistical analysis was performed with SPSS statistical analysis software (v. 23.0.0.2 for 221

Windows, SPSS Inc., Chicago, IL, USA). Linear regression analysis was performed to 222

examine the association between muscle activities and skating speed15-30m and between 223

joint kinematics and skating speed15-30m. Correlation coefficients were calculated using 224

Pearson product-moment correlation. The normality of the data was checked using the 225

Shapiro-Wilk test. Skating speed15-30m (dependent variable of the regression analyses) 226

was normally distributed (p=0.407) while some of the independent variables were not 227

(p<0.05). We confirmed using Spearman´s rank correlation that the conclusion of the 228

correlation analyses was not altered when the normality assumption was violated.

229

Homoscedasticity was visually confirmed and none of the associations showed clear 230

deviation from the linearity assumption. The level of statistical significance was set at 231

p<0.05.

232

(11)

11 Results

233

Mean values for skating speeds and spatio-temporal parameters of all participants are 234

presented in Table 1. At the beginning of the propulsion phase, the hip joint flexion angle 235

was on average 46.8 ± 10.0º (Figure 2). Extension of the hip continued during the 236

propulsion phase reaching 15.3 ± 6.2º hip flexion at the beginning of the recovery phase 237

with continuing extension in the early recovery phase. At the beginning of the propulsion 238

phase, the average knee joint flexion angle was 67.3 ± 21.1º. The knee joint extended 239

during the propulsion phase, reaching the minimum flexion angle shortly before the skate 240

left the ice. At the beginning of the recovery phase, the average knee joint flexion angle 241

was 18.5 ± 17.7º. The knee joint reached maximal flexion immediately after the midpoint 242

of the recovery phase, followed by a minor knee extension before the skate hit the ice 243

again (beginning of the propulsion phase).

244

Mean normalized muscle activity patterns are presented in Figure 3 and mean normalized 245

muscle activities during the propulsion and recovery phases are displayed in Table 2. The 246

soleus muscle demonstrated the highest neuromuscular activity during the propulsion 247

phase, immediately after the skate hit the ice, and at the mid-portion of the propulsion 248

phase. In contrast, the tibialis anterior was neuromuscularly active at the beginning of the 249

propulsion phase, but the highest activity occured during the recovery phase. Biceps 250

femoris showed the highest neuromuscular activity at the beginning of the propulsion 251

phase reaching the peak activity at the beginning of that phase. Vastus lateralis illustrated 252

high neuromuscular activity from the very end of the recovery phase until the second half 253

of the propulsion phase and at the very end of the recovery phase, its activity increased 254

steeply. Rectus femoris exhibited two activity peaks, the first in the mid-portion of the 255

propulsion and the second in the mid-portion of the recovery phase. Adductor magnus 256

demonstrated highest neuromuscular activity mostly at the beginning of the propulsion 257

(12)

12

phase and at the time when the propulsion phase switches to the recovery phase after 258

which its activity decreased evenly until the mid-portion of the recovery phase. Gluteus 259

medius showed the highest neuromuscular activity during the first half of the propulsion 260

phase and during the mid-portion of the recovery phase.Gluteus maximus neuromuscular 261

activity was the highest at the beginning of the propulsion phase until the final parts of 262

the phase and again during the latter half of the recovery phase.

263

The results of the correlation and regression coefficient analyses are presented in Tables 264

3-5. The activity of the gluteus maximus during the whole recovery phase was negatively 265

correlated with the skating speed (r=-0.651, p=0.022, β=-0.08). Additionally, a similar 266

correlation was found during the second half of the recovery phase (r=-0.701, p=0.011, 267

β=-0.05). Coactivity between gluteus maximus and rectus femoris during the recovery 268

phase correlated negatively with skating speed (r=-0.786, p=0.002, β=-3.26).

269

Discussion and Implications 270

In this study, we characterised lower limb muscle activity patterns in near maximal speed 271

ice-hockey skating in highly trained players and examined the associations between 272

muscle activities and joint kinematics with skating speed. In general, uniarticular hip, 273

knee and ankle extensors showed high neuromuscular activity during the propulsion 274

phase whereas ankle dorsiflexor showed high activity during the recovery phase. In 275

contrast to our hypothesis, no significant associations were observed between the knee or 276

hip joint sagittal plane kinematics and skating speed. Analysis of the association between 277

muscle activities and skating speed showed a negative correlation between gluteus 278

maximus activity during the recovery phase and skating speed. In support of this finding, 279

a low co-activity of gluteus maximus and rectus femoris during the recovery phase was 280

associated with faster skating speed. These findings were in line with our hypothesis of 281

(13)

13

fast hip flexion during the recovery phase being beneficial for achieving fast skating speed 282

as low activity of hip flexion antagonist may facilitate fast hip flexion. However, we did 283

not observe further significant correlations between muscle activities and skating speed 284

which contradicted our hypothesis. Overall, these findings suggest that muscle activities 285

during the recovery phase may be important technical factors related to motor control 286

helping a player to achieve fast skating speed providing a potential target for training 287

interventions.

288

Muscle activity patterns during the maximal skating phase 289

In this study, muscle activity patterns were evaluated for eight major lower limb muscles 290

during the maximal skating phase in high-level players. Soleus, vastus lateralis, gluteus 291

maximus, gluteus medius, and rectus femoris were highly active during the propulsion 292

phase, a phase during which the extension of the lower limb occurs. In contrast, adductor 293

magnus and tibialis anterior were noted to be involved during the recovery phase.

294

Additionally, rectus femoris showed high levels of activity during both the propulsion 295

phase when the knee and hip were extending and during the recovery phase when the hip 296

and knee were flexing. A similar activity profile with two distinct activity periods was 297

found for the gluteus medius. The first activity period occurred during the propulsion 298

phase and the second peak occurred during the recovery phase. These two periods of 299

gluteus medius activation are probably related to the stabilization of the pelvis by 300

increasing hip joint stiffness in the frontal plane (propulsion) and positioning the leg in 301

preparation for the propulsion phase (recovery). In addition, the muscle may assist 302

propulsion by abducting the hip in the propulsion phase.

303

The timing of the adductor magnus activity suggests that adduction of the hip initiates the 304

recovery phase. Adductor magnus is known to play an important role at the beginning of 305

(14)

14

the recovery phase during hip adduction. (Kiel & Kaiser, 2021) Additionally, the adductor 306

magnus seems to support the main hip flexor muscles during the hip flexion. It has been 307

reported that the activation level of the adductor magnus increases at higher skating speed 308

(Chang et al., 2009). When skating speed increases, propulsion needs to be directed more 309

laterally as compared to the acceleration phase (Pearsall et al., 2000; Haché 2002), which 310

sets higher demands on the adductor muscles to work eccentrically during the end of the 311

propulsion phase and concentrically at the beginning of the recovery phase. Proper hip 312

adduction during the recovery phase ensures the optimal length of the stride whereas 313

insufficient adduction may shorten the propulsion and limit the skating speed (Chang et 314

al., 2009). Rectus femoris may possess a double role during the skating cycle. Firstly, it 315

may take part in the propulsion phase by extending the knee. Secondly, during the 316

recovery phase, the rectus femoris can flex the hip along with other hip flexor muscles.

317

Rectus femoris is a knee extensor and a hip flexor muscle. We observed similar activation 318

of the muscle in both propulsion (knee and hip are extending) and in recovery (knee and 319

hip are flexing) phases with a decrease in activation between the phases. The finding 320

shows that rectus femoris is used for both knee extension and hip flexion in ice-hockey 321

skating. However, due to concomitant knee and hip extension and flexion, the length 322

changes in rectus femoris muscle may be limited during these activation peaks allowing 323

the muscle to operate close to isometric conditions. Similarly, to the situation with rectus 324

femoris, gluteus medius shows high activity during both the propulsion and recovery 325

phases. Gluteus medius can offer support and balance for the hip joint during the 326

propulsion phase. Furthermore, gluteus medius acts as an antagonist for the adductor 327

magnus during the hip adduction; thus eccentric activation occurs in the middle of the 328

recovery phase while controlling the hip adduction. Tibialis anterior is an ankle 329

dorsiflexor and it needs to be activated during the recovery phase. If dorsiflexion of the 330

(15)

15

ankle is insufficient, the front part of the skate blade hits the ice first, which may increase 331

the friction between the skate and the ice thus limiting the skating capability and even 332

causing the skater to fall. The activity of the major hip and knee extensor muscles (vastus 333

lateralis and gluteus maximus) begins at the end of the recovery phase by pre-activity.

334

When the skate hits the ice, it can be assumed that muscles need initially to work 335

isometrically to acquire the proper gliding position while in the gliding phase (beginning 336

of the propulsion phase) there is a rapid extension of the joint.

337

Individual neuromuscular activities are presented in Figure 4. The timing of muscle 338

activity of vastus lateralis, gluteus maximus, and soleus showed small between- 339

participant variability (i.e. consistent timing of the maximum and minimum muscle 340

activation). In contrast, there was greater between-participant variability in biceps 341

femoris, rectus femoris, and gluteus medius. For instance, biceps femoris of participant 342

12 showed relatively small alterations in the activity during the cycle compared to other 343

participants. Rectus femoris of participant 1 did not show a similar increase in the activity 344

during the recovery phase than other participants. In contrast, the extensive activity of the 345

rectus femoris compared to gluteus maximus was noticed in participants 3 and 4.

346

Interestingly participant 3 had the second-worst mean time of the five 30 m skating 347

sprints. Participant 12 had the fastest average 30-meter skating time and when observing 348

Figure 4, both rectus femoris and gluteus maximus showed high activity during the 349

propulsion phase. These observations need to be put into the context of qualitative 350

observations from single participants but may suggest that motor skill and muscle 351

coordination are essential factors for ice-hockey skating performance.

352

Associations between skating speed, muscle activity, and kinematics 353

(16)

16

Lower activity of the gluteus maximus and lower coactivity of the gluteus maximus and 354

rectus femoris during the recovery phase were found to be associated with faster skating 355

speed. During the recovery phase, the hip flexes rapidly, and relaxation of the gluteus 356

maximus may enable the skater to execute greater and more rapid hip flexion, which may 357

help the skater to reach a higher skating speed. Gluteus maximus muscle is mainly a hip 358

extensor muscle and an antagonist muscle for hip flexion (Krause et al., 2020). In fact, it 359

has been found that high-level players flex the hip more quickly than their lower-level 360

counterparts (Robbins et al., 2021). Furthermore, earlier studies have shown that the 361

optimal coactivation between agonist and antagonist muscles enables a better function of 362

the muscles i.e. a more favourable muscle activation pattern and effective movement 363

(Latash, 2018).

364

Limitations and suggestions for future studies 365

As in any correlation-based study, the number of observations is critical in identifying 366

associations between variables and for accurate estimates of the strength of the 367

associations. Due to acute injuries and technical problems we were not able to include the 368

sample size that was calculated to be sufficient based on a priory sample size calculations.

369

Because of the limited number of participants and their rather homogeneous nature (all 370

were high-level ice-hockey players), we may not have observed all potentially important 371

associations between skating speed and muscle activities. On the other hand, some 372

potentially coincidental correlations were not detected. The identified statistically 373

significant correlations, therefore, reflect the strongest associations observed in this 374

dataset and serve as a starting point for future studies. In this study, kinematics were 375

available only in the sagittal plane. We do not consider this as a limitation regarding the 376

knee joint as it is common to assume the knee to function as a 1 degree of freedom hinge 377

joint as most of the movement is in the sagittal plane. However, measuring kinematics 378

(17)

17

only in the sagittal plane prevented us from examining potentially important factors 379

associated with the hip joint motion in which frontal plane movement is significant. In 380

this study, maximum voluntary isometric contractions (MVICs) were used to normalize 381

the EMG data. It is possible that some participants were not able to maximally activate 382

their muscles during these tasks and thus the results (Fig. 3) should not be viewed to 383

directly indicate the percentage of maximal muscle activation but rather the percentage 384

of activity compared to the one measured during maximal effort isometric task.

385

In the future, the results of the current study should be replicated with larger sample size 386

and with 3D kinematics measurements. In addition, the potential causality of the observed 387

associations should be investigated in an intervention study. Moreover, in the future, the 388

studies of the forward skating kinematics during fatigue as well as the muscle activations 389

and kinematics involved in backward skating, crossover skating, and different turns and 390

pivots would be beneficial to develop skating skills more comprehensively in ice-hockey 391

players. These kinds of studies would offer useful knowledge to develop ice-hockey as a 392

sport and help ice-hockey players improve their skills.

393

Practical implications 394

In this study, it was found that improper co-activity of agonist and antagonist muscles 395

may limit skating performance. Therefore, coaches need to understand the role of each 396

muscle during the different phases of the skating cycle and their function as agonists or 397

antagonists while the skater is in motion. These concepts should be also conveyed to the 398

players through specific motor skill training exercises performed either on or off the ice 399

to teach the players to activate and deactivate the muscles in the optimal order and time.

400

Here we also described muscle activity patterns of eight lower limb muscles relevant for 401

skating. The data can be useful for improving the basic understanding of the neuro- 402

(18)

18

muscular function during ice-hockey skating with potential implications on the design of 403

strength training routines or drills. For example knowing that vastus lateralis, gluteus 404

maximus, and soleus are highly active during the propulsion phase can help design a 405

functional exercise aiming to improve propulsion force. Based on the results the exercise 406

should highly activate at least vastus lateralis, gluteus maximus, and soleus in addition to 407

replicating the kinematic pattern observed during the propulsion.

408

Conclusion 409

It is concluded that muscle activity patterns (timing or magnitude) in the recovery phase 410

of the skating cycle seem to play an important role in achieving fast skating speed. The 411

novel finding in this study is that not only the powerful knee and hip extension during the 412

propulsion phase but also the ability to deactivate the gluteus maximus muscle during the 413

recovery phase appears to be important for achieving the highest possible skating speed.

414

This study provides practitioners with an overview of lower limb muscle activity patterns 415

during near-maximal speed skating and highlights the importance of considering also the 416

recovery phase muscle function, in addition to the propulsion phase, when developing 417

training methods used to develop maximal skating speed.

418

Acknowledgments 419

This work was supported by the European Regional Developments Fund and the 420

University of Eastern Finland under the project: Human measurement and analysis - 421

research and innovation laboratories (HUMEA, project identifiers: A73200 and A73241).

422

The authors thank Markku Keinänen (Tuplajäät Oy, Kuopio Finland) to offer facilities 423

for on-ice data collection and all participants for their excellent co-operation.

424

Disclosure statement 425

(19)

19

Sami Kaartinen is the co-founder and head of R&D of the company Pro Prospect that 426

provides ice-hockey coaching solutions. He is also a strength and conditioning coach for 427

the ice-hockey team KalPa Hockey (Kuopio, Finland).

428 429

References 430

de Boer, R.W., Cabri, J., Vaes, W., Clarijs, J.P., Hollander, A.P., de Groot, G., van 431

Ingen Schenau, G.J. (1987). Moments of force, power, and muscle coordination 432

in speed-skating. International Journal of Sport Medicine, 8, 371–378.

433

Buckeridge, E., LeVangie, M.C., Stetter, B., Nigg, S.R., Nigg, B.M. (2015). An on-ice 434

measurement approach to analyse the biomechanics of ice-hockey skating. PLoS 435

One, 10.

436

Chang, R., Turcotte, R., Pearsall, D. (2009). Hip adductor muscle function in forward 437

skating. Sport Biomechanics, 8, 212–222.

438

Ervilha, U.F., Graven-Nielsen, T., Duarte, M. (2012). A simple test of muscle 439

coactivation estimation using electromyography. Brazilian Journal of Medical 440

and Biological Research, 45, 977-981.

441

Faul, F., Erdfelder, E., Lang, A.G., and Buchner, A. (2007). G*Power 3: A flexible 442

statistical power analysis program for the social, behavioral, and biomedical 443

sciences. In Behavior Research Methods, (Psychonomic Society Inc.), 175–191.

444

Goudreault, R. (2002). Forward skating in ice-hockey: comparison of EMG activation 445

patterns at three velocities using a skate treadmill. National Library of Canada.

446

Haché, A. (2002). The Physics of Ice-hockey. Johns Hopkins University Press, 69-84.

447

Hermens, H.J., Freriks, B., Merletti, R., Stegeman, D., Blok, J., Disselhorst-Klug, C., 448

Rau, G., Hägg, G. (2000). Development of recommendations for SEMG sensors 449

(20)

20

and sensor placement procedures. Journal of Electromyography and Kinesiology, 450

10, 361–374.

451

Kiel, J., Kaiser, K. (Updated 2021). Adductor Strain. StatPearls [Internet].

452

Krause Neto, W., Soares, E.G., Vieira, T.L., Aguiar, R., Chola, T.A., Sampaio, V. de L., 453

and Gama, E.F. (2020). Gluteus maximus activation during common strength and 454

hypertrophy exercises: A systematic review. Journal of Sports Science and 455

Medicine, 19, 195–203.

456

Lafontaine, D. (2007). Three-dimensional kinematics of the knee and ankle joints for 457

three consecutive push-offs during ice-hockey skating starts. Sports 458

Biomechanics, 6, 391–406.

459

Latash, M.L. (2018). Muscle coactivation: Definitions, mechanisms, and functions.

460

Journal of Neurophysiology, 120, 88–104.

461

Pearsall, DJ., Turcotte, R., Murphy, S.D. (2000). Biomechanics of ice-hockey. Exercise 462

and Sport Science, 675–692.

463

Robbins, S.M., Renaud, P.J., and Pearsall, D.J. (2021). Principal component analysis 464

identifies differences in ice-hockey skating stride between high- and low-calibre 465

players. Sports Biomechanics, 20, 131–149.

466

Thelen, D.G., Chumanov, E.S., Best, T.M., Swanson, S.C., Heiderscheit, B.C. (2005).

467

Simulation of biceps femoris musculotendon mechanics during the swing phase 468

of sprinting. Medicine & Science in Sports & Exercise, 37, 1931–1938.

469

Upjohn, T., Turcotte, R., Pearsall, D.J., Loh, J. (2008). Three-dimensional kinematics of 470

the lower limbs during forward ice-hockey skating. Sports Biomechanics, 7, 206–

471

221 472

473

(21)

18

Table 1. Descriptive spatio-temporal parameters of the maximal 30-meter forward skating sprints (N=12).

Parameter (unit) Mean ± SD

30 m time (s) 4.14 ± 0.08

Skating speed15-30 m (m/s) 8.39 ± 0.12 Skating cycle frequency (cycles/s) 1.64 ± 0.14 Recovery phase proportion (% of cycle) 42 ± 4 Distance travelled during the recovery phase (m) 2.15 ± 0.27

Skating speed was measured as the average speed of the last 15 meters of the sprint. Skating cycle frequency, recovery phase proportion and distance travelled during the recovery phase were determined from skating cycles taken during the last 15 meters of the sprint.

(22)

22 Table 2. Normalized neuromuscular activity during the propulsion and recovery phases (N=12).

Muscle Propulsion phase Recovery phase

(Mean ± SD) (Mean ± SD)

Soleus 68.5 ± 26.2 18.4 ± 19.6

Tibialis anterior 25.0 ± 8.6 39.2 ± 12.9

Bicep femoris 31.9 ± 18.2 14.6 ± 6.8

Vastus lateralis 82.5 ± 28.7 15.4 ± 8.0

Rectus femoris 46.9 ± 13.9 41.9 ± 15.1

Adductor magnus 19.5 ± 27.2 17.9 ± 7.2

Gluteus medius 54.7 ± 23.6 35.3 ± 29.0

Gluteus maximus 52.3 ± 24.1 24.7 ± 13.3

The values represent mean muscle activities normalized to activities measured during maximal isometric voluntary contractions.

(23)

23

Table 3. Correlation (r) and regression (β) coefficients (and their p-value) between the dependent variable (skating speed15-30 m) and the independent variables (neuromuscular activity) during the entire propulsion phase and separately for the early (0-33%), mid (33-66%) and late (66-100%) subphases of the propulsion.

Propulsion

Muscle Entire phase 0-33% 33-67% 67-100%

r p β (95% CI) r p β (95% CI) r p β (95% CI) r p β (95% CI)

SOL -0.284 0.371 -0.02 (-0.60;0.20) -0.452 0.140 -0.03 (-0.70; 0.10) -0.046 0.886 0.00 (-0.50; 0.50) -0.223 0.486 -0.01 (-0.40; 0.20) TA 0.157 0.626 0.03 (-1.00; 1.60) -0.160 0.620 0.02 (-0.70; 1.10) -0.227 0.479 0.03 (-0.60; 1.20) -0.058 0.857 -0.01 (-1.10; 0.90) RF -0.132 0.684 -0.02 (-1.00; 0.70) -0.099 0.760 -0.01 (-0.60; 0.40) -0.347 0.269 -0.03 (-0.90; 0.30) -0.153 0.635 0.01 (-0.50; 0.70) VL 0.080 0.805 0.00 (-0.40; 0.50) -0.009 0.978 0.01 (-0.30; 0.30) -0.159 0.622 0.01 (-0.30; 0.40) 0.092 0.776 0.01 (-0.40; 0.60) BF 0.199 0.536 0.02 (-0.50; 0.80) 0.165 0.609 0.01 (-0.30; 0.50) 0.189 0.556 0.01 (-0.40; 0.70) 0.218 0.496 0.04 (-0.80; 1.50) AM 0.153 0.636 0.01 (-0.03; 0.05) 0.157 0.625 0.01 (-0.03; 0.04) 0.125 0.700 0.01 (-0.04; 0.06) 0.169 0.600 0.01 (-0.03; 0.06) GlutMax 0.269 0.398 -0.02 (-0.07; 0.03) -0.378 0.226 -0.02 (-0.05; 0.01) -0.202 0.529 -0.01 (-0.05; 0.03) 0.013 0.967 0.00 (-0.09; 0.09) GlutMed 0.053 0.869 0.00 (-0.05; 0.05) -0.172 0.592 0.01 (-0.02; 0.04) -0.414 0.181 -0.03 (-0.09; 0.02) 0.230 0.471 0.02 (-0.03; 0.06) GlutMax/RF -0.065 0.840 -0.17 (-1.99; 1.65) -0,289 0.362 -0.56 (-1.88; 7.50) 0.051 0.874 0.13 (-1.67; 1.93) 0.147 0.648 0.36 (-1.34; 2.06) VL/BF -0.221 0.490 -0.21 (-0.85; 0.44) -0.200 0.532 -0.19 (-0.84; 0.46) -0.096 0.767 -0.05 (-0.40; 0.31) -0.193 0.547 -0.19 (-0.89; 0.50) TA/SOL 0.136 0.673 0.97 (-4.00; 5.94) 0.360 0.250 1.26 (-1.04; 3.55) 0.133 0.681 0.45 (-1.93; 2.84) -0.073 0.823 -0.20 (-2.15; 1.75) AM/GlutMed 0.119 0.712 0.47 (-2.31; 3.25) 0.134 0.677 0.50 (-2.09; 3.09) 0.119 0.712 0.49 (-2.41; 3.40) -0.017 0.958 -0.04 (-1.56; 1.49) SOL, soleus; TA, tibialis anterior; RF, rectus femoris; VL, vastus lateralis; BF, biceps femoris; AM, adductor magnus; GlutMax, gluteus maximus; GlutMed, gluteus medius; GlutMax/RF, activity ratio between gluteus maximus and rectus femoris; VL/BF, activity ratio between vastus lateralis and biceps femoris; TA/SOL, activity ratio between tibialis anterior and soleus;

AM/GlutMed, activity ratio between adductor magnus and gluteus medius.

(24)

24

Table 4. Correlation (r) and regression (β) coefficients (and their p-value) between the dependent variable (skating speed15-30 m) and the independent variables (neuromuscular activity) during the entire recovery phase and separately for the early (0-50%), late (51-100%) subphases of the recovery phase.

Recovery

Muscle Entire phase 0-50% 50-100%

r p β (95% CI) r p β (95% CI) r p β (95% CI)

SOL 0.105 0.745 0.02 (-1.00; 1.40) -0.091 0.779 0.01 (-0.70; 0.90) 0.105 0.745 0.03 (-1.40; 2.00) TA 0.130 0.688 0.02 (-0.70; 1.10) 0.132 0.682 0.01 (-0.60; 0.80) 0.096 0.766 0.01 (-0.90; 1.10) RF 0.320 0.310 0.04 (-0.40; 1.10) 0.270 0.396 0.02 (-0.30; 0.80) 0.210 0.512 0.02 (-0.40; 0.80) VL -0.143 0.658 -0.03 (-1.80; 1.20) -0.066 0.838 -0.03 (-3.90; 3.30) -0.144 0.655 -0.02 (-1.00; 0.60) BF 0.408 0.499 0.10 (-0.60; 2.60) 0.265 0.406 0.07 (-1.00; 2.30) 0.491 0.105 0.10 (-0.30; 2.30) AM 0.191 0.551 0.04 (-0.12; 0.21) 0.142 0.659 0.02 (-0.09; 0.13) 0.162 0.614 0.04 (-0.12; 0.20) GlutMax -0.651 0.022 -0.08 (-0.15; -0.01) -0.178 0.580 -0.03 (-0.13; 0.08) -0.701 0.011 -0.05 (-0.09; -0.01) GlutMed 0.243 0.446 0.01 (-0.03; 0.05) 0.147 0.684 0.01 (-0.03; 0.05) 0.330 0.295 0.02 (-0.02; 0.06) GlutMax/RF -0.786 0.002 -3.26 (-5.07; -1.45) -0.428 0.165 -4.63 (-11.51; 2.25) -0.620 0.032 -1.36 (-2.58; -0.15) VL/BF -0.349 0.266 -0.83 (-2.41; 0.74) -0.284 0.371 -0.98 (-3.32; 1.36) -0.268 0.400 -0.39 (-1.37; 0.59) TA/SOL 0.130 0.688 0.02 (-0.70; 1.10) -0.122 0.705 -0.13 (-0.87; 0.61) -0.217 0.498 -0.25 (-1.04; 0.54) AM/GlutMed 0.020 0.951 0.07 (-2.40; 2.54) 0.199 0.534 0.32 (-0.78; 1.41) 0.006 0.986 0.04 (-4.49; 4.56)

SOL, soleus; TA, tibialis anterior; RF, rectus femoris; VL, vastus lateralis; BF, biceps femoris; AM, adductor magnus; GlutMax, gluteus maximus; GlutMed, gluteus medius; GlutMax/RF, activity ratio between gluteus maximus and rectus femoris; VL/BF, activity ratio between vastus lateralis and biceps femoris; TA/SOL, activity ratio between tibialis anterior and soleus;

AM/GlutMed, activity ratio between adductor magnus and gluteus medius.

Table 5. Correlation (r) and regression (β) coefficients (and their p-value) between the dependent variable (skating speed15-30 m) and the independent variables (knee and hip kinematics) during different phases of the skating cycle.

Kinematic variable r p β (95% CI)

Knee flexion at start of the propulsion phase 0.067 0.845 0.00 (-0.04; 0.05) Knee flexion at start of the recovery phase -0.197 0.561 -0.01 (-0.07; 0.04) Knee ROM during the recovery phase 0.263 0.435 0.01 (-0.03; 0.05) Knee ROM during the propulsion phase 0.239 0.480 0.02 (-0.04; 0.07) Hip flexion at start of the propulsion phase -0.271 0.420 -0.05 (-0.17; 0.08) Hip flexion at start of the recovery phase 0.517 0.104 0.15 (-0.04; 0.33) Hip ROM during the recovery phase -0.409 0.211 -0.07 (-0.18; 0.05) Hip ROM during the propulsion phase -0.502 0.116 -0.08 (-0.20; 0.03)

(25)

25 Figure captions

Figure 1. Camera and split mark locations for measuring skating speed during the final 15 meters. Above, a screenshot from the video recording with 15- and 30-meter marks and photocells highlighted. Below, a schematic of the measurement setup showing the camera parallax correction considered in the location of the split time marks, The camera was located in the middle of the 30-meter sprint, 16 meters from the line of skating with its field of view perpendicular to the skating direction. The 30-meter mark was positioned at a distance of 29.06 meters from the starting line to account for cameral parallax.

(26)

26

Figure 2. Average sagittal hip and knee joint angular changes during the whole skating cycle. The solid line represents the mean value of all skaters and the shaded area represents standard deviation. The area on the left side of the dotted vertical line represents the propulsion phase and on the right side the recovery phase.

Figure 3. Muscle activities during the whole skating cycle. Activities are normalized to maximal activity obtained in maximal isometric contractions. The solid line represents the mean value of all participants and the shaded area represents the standard deviation. The area which is located on the left side of the dotted vertical line represents the propulsion phase and the right side the recovery phase.

(27)

27

Figure 4. Heat map of muscle activities for each participant and muscle during maximal skating phase (15-30 meters of maximal sprint). Activities were scaled individually between 0 and 100% to highlight parts of the skating cycle with high and low muscle activities relative to the range of muscle activities observed for the individual.

(28)

28

Referências

Documentos relacionados