. 2020 Nov 16;6:54. doi: 10.1186/s40798-020-00283-6

Table 1.

Studies investigating the effects of anticipation on linear (g) and rotational (rad · s⁻²) head accelerations

	Participants, sport and level	Protocol	Anticipation condition	Linear head acceleration (g)	Rotational head acceleration (rad · s⁻²)	Muscle activity (mV)	Significant findings	Limitations
In vivo studies
Kuramochi et al. [47]	9 healthy males 19–30 years	Direct impact to the forehead while sitting in a chair Impact: 4 kg weight released from 25° Uniaxial accelerometer Rectified EMG averaged over onset duration for SCM and TRP	Anticipated (eyes opened) Unanticipated (eyes closed)	3.85 ± 0.33 (37.8 ± 3.3 m · s⁻²) 3.72 ± 0.24 (36.5 ± 2.4 m · s⁻²)	N/A N/A	R/L SCM: 6.3 ± 1.5/ 8.3 ± 2.5 R/L SCM: 17.9 ± 3.7/ 22.4 ± 5.4	Anticipation did not affect the linear head acceleration No difference in the onset latency between conditions Unanticipated condition elicited greater muscle activity than the anticipated condition	A strictly laboratory-based study; findings limited for extrapolating the results to concussive head impact intensities during game play Head accelerations significantly below the proposed 85 g injury threshold Limited to sagittal plane movement
Mihalik et al. [30]	16 male Bantam-level ice hockey players (elite youth) Average age: 14.0 ± 05 years Experience: 7.8 ± 1.7 years	Data collected from single team over 54 games: HIT system (> 10 g) Qualitatively assessed body position from video analysis (CHECC List) Location on the ice (along playing boards, or on the open ice)	Anticipated (‘saw hit coming’), good body position Anticipated, bad body position Unanticipated regardless of body position	20.7 21.4 22.6	1409.4* 1420.4* 1550.0	N/A N/A N/A	Anticipation was not associated with lower linear head accelerations For the medium intensity impacts, anticipation was associated with a reduced severity of rotational head acceleration Anticipation had no effect on head acceleration for the highest intensity collisions (> 75 percentile of collisions according to the HIT severity profile)	Small convenience sample: various player positions, single team, 1 season Investigator’s judgement on the player’s anticipation status from video footage only
Hasegawa et al. [48]	12 male high school rugby players Average age: 16.8 years	5 rugby tackles to the chest Triaxial accelerometer measured head accelerations of the attacker and defender EMG to assess bilateral masseter and SCM¹	No clenching instruction (unanticipated) Tightly clenched (anticipated)	A: 2.64 ± 0.33 D: 2.86 ± 0.23 A: 2.16 ± 0.50* D: 2.30 ± 0.27*	N/A N/A	Masseter A/D: 0.22 ± 0.12/0.29 ± 0.18 SCM A/D: 0.60 ± 0.19/0.70 ± 0.23 Masseter A/D: 0.55 ± 0.25/0.73 ± 0.46 SCM A/D: 0.62 ± 0.18/0.77 ± 0.24	Clenching increased activity of the masseter muscle and decreased linear head acceleration Muscle activity onset occurs prior to body contact	Small sample size Analysed sub-concussive intensity head impacts to minimise injury risk; cannot confidently extrapolate results to concussive head impact intensities Rotational head accelerations not analysed
Narimatsu et al. [49]	11 male high school soccer players Average age: 17.2 years	5 trials of heading a soccer ball under 3 conditions Triaxial accelerometer measured head accelerations EMG ¹ to assess bilateral masseter and SCM Ball projected approx. 9 m from JUGS soccer machine (initial motor velocities set at 28 m · s⁻¹ (RM) and 38 m · s⁻¹ (LM)	No clenching instruction (unanticipated) Clenched w/o mouthguard (anticipated) Clenched with mouthguard (anticipated)	28.4 ± 7.0 23.9 ± 6.2* 21.5 ± 4.6*	N/A N/A N/A	Masseter: 44.0 ± 38.2 SCM: 68.6 ± 47.7 Masseter: 132.7 ± 76.5* SCM: 133.5 ± 74.2* Masseter: 154.0 ± 99.3* SCM: 159.3 ± 76.8*	Clenching in anticipation of heading a soccer ball increases muscle activity of masseter and SCM muscle compared to the no clenching instruction condition only (no mouthguard and clenching interaction), and reduces linear head acceleration Masseter and SCM active before headed ball	Small sample size Not game-like intensity, intentionally eliciting lower intensity head impacts to minimise head injury risk; cannot confidently extrapolate to concussive head impact intensities
Schmidt et al. [50]	32 male high school conference 3A varsity American Football players Average age: 16.7 ± 0.9 years	Data collected during game play HIT system to assess head accelerations Video analysis to subjectively classify impacts as anticipated or unanticipated	Anticipated Unanticipated	25.9 26.5	1605.5 1621.3	N/A N/A	There was a trend (p = 0.07) towards a lower linear head acceleration when the impact was perceived to be anticipated	Small convenience sample: various player positions, single team, over 1 season Limitations of video analysis for determining anticipation level No assessment of concussion risk No assessment of whether perceived anticipation resulted in players moving into a protective position
Simulations and modelling
Jin et al. [51]	Simulated impacts using finite element model Head and neck complex of the Global Human Body Model with validated helmet model 27 pairs of Hill-type muscle elements	Reconstructed concussive and non-concussive American Football-related head impacts Magnitude and timing of impact for simulation and experimental data within 10% of each other; peak impact force approx. 10,000 N 4 conditions simulated	(1) No muscle activity (2) Reactive muscle response (onset at impact, 55 ms to peak activation) (3) Pre-activation response (40 ms before impact) (4) Pre-activation response (40 ms before impact) with 200% strength	113.36 112.18 111.75 110.59	29.25 rad · s⁻¹ 26.90 rad · s⁻¹ 22.72 rad · s⁻¹ 20.16 rad · s⁻¹	N/A N/A N/A N/A	Anticipatory activation of neck musculature reduced injury criteria No change in the injury criteria with double muscle strength, or reactive activation compared to no activation No differences in linear accelerations of the head Pre-activation reduced peak rotational velocity (18.1–31.0%)	Theoretical computational model only Results largely influenced by the model constraints used
Eckersley et al. [52]	Simulated model Duke University Head and Neck model	Simulated 4 head impact conditions with 6 neck muscle activity conditions at 8 impact sites Impacts: (1) Impact of high speed object in flight (2) 80 g helmet-to-helmet impact (short duration) (3) 80 g helmet-to-helmet impact (longer duration) (4) 40 g helmet-to-helmet impact	(1) Relaxed (min. activation to maintain head stability) (2) Maximally activated neck musculature (3) Maximally activated neck flexors (4) Maximally activated neck extensors	N/A N/A N/A N/A	Lowest value ↑ ↑ ↑	N/A N/A N/A N/A	Magnitude of cervical muscle force does not influence short-term (< 50 ms) head kinematics Impacts to the side of the head, higher than ear level consistently produced highest peak resultant angular acceleration Musculature presumably active at time of impact to simulate pre-activation Values not reported	Theoretical computational model only Results largely influenced by the model constraints used Estimated constraints used to simulate a direct impact to the head, rather than using data from real concussive head impacts

Participants, sport and level

Protocol

Anticipation condition

Linear head acceleration (g)

Rotational head acceleration (rad · s⁻²)

Muscle activity (mV)

Significant findings

Limitations

In vivo studies

Kuramochi et al. [47]

9 healthy males

19–30 years

Direct impact to the forehead while sitting in a chair

Impact: 4 kg weight released from 25°

Uniaxial accelerometer

Rectified EMG averaged over onset duration for SCM and TRP

Anticipated

(eyes opened)

Unanticipated

(eyes closed)

3.85 ± 0.33

(37.8 ± 3.3 m · s⁻²)

3.72 ± 0.24

(36.5 ± 2.4 m · s⁻²)

N/A

R/L SCM:

6.3 ± 1.5*/

8.3 ± 2.5*

R/L SCM:

17.9 ± 3.7/

22.4 ± 5.4

Anticipation did not affect the linear head acceleration

No difference in the onset latency between conditions

Unanticipated condition elicited greater muscle activity than the anticipated condition

A strictly laboratory-based study; findings limited for extrapolating the results to concussive head impact intensities during game play

Head accelerations significantly below the proposed 85 g injury threshold

Limited to sagittal plane movement

Mihalik et al. [30]

16 male Bantam-level ice hockey players (elite youth)

Average age:

14.0 ± 05 years

Experience:

7.8 ± 1.7 years

Data collected from single team over 54 games:

HIT system (> 10 g)

Qualitatively assessed body position from video analysis (CHECC List)

Location on the ice (along playing boards, or on the open ice)

Anticipated (‘saw hit coming’), good body position

Anticipated, bad body position

Unanticipated regardless of body position

20.7

21.4

22.6

1409.4*

1420.4*

1550.0

N/A

Anticipation was not associated with lower linear head accelerations

For the medium intensity impacts, anticipation was associated with a reduced severity of rotational head acceleration

Anticipation had no effect on head acceleration for the highest intensity collisions (> 75 percentile of collisions according to the HIT severity profile)

Small convenience sample: various player positions, single team, 1 season

Investigator’s judgement on the player’s anticipation status from video footage only

Hasegawa et al. [48]

12 male high school rugby players

Average age: 16.8 years

5 rugby tackles to the chest

Triaxial accelerometer measured head accelerations of the attacker and defender

EMG to assess bilateral masseter and SCM¹

No clenching instruction (unanticipated)

Tightly clenched (anticipated)

A: 2.64 ± 0.33

D: 2.86 ± 0.23

A: 2.16 ± 0.50*

D: 2.30 ± 0.27*

N/A

Masseter A/D:

0.22 ± 0.12/0.29 ± 0.18

SCM A/D:

0.60 ± 0.19/0.70 ± 0.23

Masseter A*/D*:

0.55 ± 0.25/0.73 ± 0.46

SCM A/D:

0.62 ± 0.18/0.77 ± 0.24

Clenching increased activity of the masseter muscle and decreased linear head acceleration

Muscle activity onset occurs prior to body contact

Small sample size

Analysed sub-concussive intensity head impacts to minimise injury risk; cannot confidently extrapolate results to concussive head impact intensities

Rotational head accelerations not analysed

Narimatsu et al.

[49]

11 male high school soccer players

Average age:

17.2 years

5 trials of heading a soccer ball under 3 conditions

Triaxial accelerometer measured head accelerations

EMG ¹ to assess bilateral masseter and SCM

Ball projected approx. 9 m from JUGS soccer machine (initial motor velocities set at 28 m · s⁻¹ (RM) and 38 m · s⁻¹ (LM)

No clenching instruction

(unanticipated)

Clenched w/o mouthguard

(anticipated)

Clenched with mouthguard

(anticipated)

28.4 ± 7.0

23.9 ± 6.2*

21.5 ± 4.6*

N/A

Masseter: 44.0 ± 38.2

SCM: 68.6 ± 47.7

Masseter: 132.7 ± 76.5*

SCM: 133.5 ± 74.2*

Masseter: 154.0 ± 99.3*

SCM: 159.3 ± 76.8*

Clenching in anticipation of heading a soccer ball increases muscle activity of masseter and SCM muscle compared to the no clenching instruction condition only (no mouthguard and clenching interaction), and reduces linear head acceleration

Masseter and SCM active before headed ball

Small sample size

Not game-like intensity, intentionally eliciting lower intensity head impacts to minimise head injury risk; cannot confidently extrapolate to concussive head impact intensities

Schmidt et al. [50]

32 male high school conference 3A varsity American Football players

Average age:

16.7 ± 0.9 years

Data collected during game play

HIT system to assess head accelerations

Video analysis to subjectively classify impacts as anticipated or unanticipated

Anticipated

Unanticipated

25.9

26.5

1605.5

1621.3

N/A

There was a trend (p = 0.07) towards a lower linear head acceleration when the impact was perceived to be anticipated

Small convenience sample: various player positions, single team, over 1 season

Limitations of video analysis for determining anticipation level

No assessment of concussion risk

No assessment of whether perceived anticipation resulted in players moving into a protective position

Simulations and modelling

Jin et al. [51]

Simulated impacts using finite element model

Head and neck complex of the Global Human Body Model with validated helmet model

27 pairs of Hill-type muscle elements

Reconstructed concussive and non-concussive American Football-related head impacts

Magnitude and timing of impact for simulation and experimental data within 10% of each other; peak impact force approx. 10,000 N

4 conditions simulated

(1) No muscle activity

(2) Reactive muscle response (onset at impact, 55 ms to peak activation)

(3) Pre-activation response (40 ms before impact)

(4) Pre-activation response (40 ms before impact) with 200% strength

113.36

112.18

111.75

110.59

29.25 rad · s⁻¹

26.90 rad · s⁻¹

22.72 rad · s⁻¹

20.16 rad · s⁻¹

N/A

Anticipatory activation of neck musculature reduced injury criteria

No change in the injury criteria with double muscle strength, or reactive activation compared to no activation

No differences in linear accelerations of the head

Pre-activation reduced peak rotational velocity (18.1–31.0%)

Theoretical computational model only

Results largely influenced by the model constraints used

Eckersley et al. [52]

Simulated model

Duke University Head and Neck model

Simulated 4 head impact conditions with 6 neck muscle activity conditions at 8 impact sites

Impacts:

(1) Impact of high speed object in flight

(2) 80 g helmet-to-helmet impact (short duration)

(3) 80 g helmet-to-helmet impact (longer duration)

(4) 40 g helmet-to-helmet impact

(1) Relaxed (min. activation to maintain head stability)

(2) Maximally activated neck musculature

(3) Maximally activated neck flexors

(4) Maximally activated neck extensors

N/A

Lowest value

↑

N/A

Magnitude of cervical muscle force does not influence short-term (< 50 ms) head kinematics

Impacts to the side of the head, higher than ear level consistently produced highest peak resultant angular acceleration

Musculature presumably active at time of impact to simulate pre-activation

Values not reported

Theoretical computational model only

Results largely influenced by the model constraints used

Estimated constraints used to simulate a direct impact to the head, rather than using data from real concussive head impacts

A attacker, D defender, HIT head impact telemetry, LM left motor, RM right motor, SCM sternocleidomastoid muscle activity, TRP trapezius

¹Details of the EMG analysis not provided

*Significant difference compared to the unanticipated condition