Keywords
neurodegeneration - repetitive head impacts - brain injury - concussion - contact sports
Introduction
Sports participation has been associated with valuable outcomes, including
physiological health promotion and psychosocial benefits [1]. However, there are growing concerns
about the potential adverse effects of sports participation due to repetitive
subconcussive head impacts (RSHIs) observed in contact and collision sports (i. e.
ice hockey, American football, and soccer) [2]
[3]. A subconcussive head
impact is characterized by a cranial impact that does not result in overt symptoms,
such as dizziness, headaches, or short-term memory loss [4], which are present in diagnosed
concussions [5]. Head impacts appear
particularly during body checks in ice hockey [6], tackles/blocks in American football (FB) [7] or headers in soccer [8]. A head impact event leading to
potential meaningful clinical changes has been defined as an event that is caused by
a minimum of 10 g linear acceleration (LA) [9]
[10]. Still, the brain
injury threshold seems to be individualized, as each person’s brain (anatomy /
function) reacts differently to external impact forces [9]
[11]
[12]. Subconcussive head
impact events are believed to have the most negative impacts when they occur in a
cumulative manner [4]. Exposure to such
events may be linked with long-term clinical malfunctions, such as chronic traumatic
encephalopathy (CTE) [13],
neurodegenerative diseases [14], and
acute cognitive deficits [15]
[16]
[17]. However, it is still unclear whether a threshold of 10 g is
plausible and/or if there exists a dose-response relationship between subconcussive
head impacts and brain anatomy and function.
RSHIs may cause structural white matter alterations, such as axonal swelling, axonal
integrity reduction, and neuroinflammation [4]
[18]
[19], mostly in the absence of overt
functional (cognitive) symptoms [20]
[21]. Structural brain
alterations following RSHIs have also been linked with acute and chronic signs of
axonal injury [20] as well as neuronal
loss [21]. Despite the fact that most
researchers did not find overt cognitive symptoms, or report mixed findings [21]
[22], some cognitive changes have been detected after RSHIs. For example,
reduced reaction time [23] and impaired
processing speed abilities [24] have
been reported after recent repetitive heading exposure in soccer. RSHIs, just like
sport-related concussions (SRCs), are characterized by central accelerative
biomechanics (i. e. number of impacts, linear and rotational accelerations, impact
duration, and impulse) acting on the head [25]
[26]. Therefore, they
could potentially share some common symptoms of brain injury. In fact, SRCs may also
lead to alterations on the white matter microstructural level, as indicated by
axonal integrity loss and axonal swelling [27]
[28]. In contrast to
RSHIs, SRCs are often accompanied by observable changes, such as loss of
consciousness, balance deficits, and sleep disturbances [5]
[29]. SRCs are characterized by high acceleration values that surpass
acceleration thresholds of 10 g, as it is assumed for RSHIs. In fact, linear
accelerations of SRCs have been reported as high as 98 g for male collision sports
[30] or 43 g for female ice hockey
[12]. Compared to males, females
seem to sustain SRCs at lower acceleration magnitudes [11]
[12]
[31].
The central biomechanical features in sport-related subconcussive head impacts may be
induced by direct or inertial (i. e. whiplash) loadings on the head and be either
linear or rotational in nature [32].
Despite the acceleration metrics, the impulse and the duration of the impact seem to
critically contribute to the injury mechanism [26]. Linear acceleration of the head has been associated with an increase
in intracranial pressure, while rotational acceleration (RA) is thought to produce
more diffuse injury, due to induced shear forces [32]. Thus far, only head acceleration
events, which are induced by a minimum of 10 g LA, are expected to have negative
impacts on brain health [9]. This might
happen in situations with head-to-head contact in football, during headers in
soccer, or while receiving a punch in boxing [8]
[33]. In fact, male high
school football athletes are exposed to an average magnitude of 26.3±2.8 g LA during
head impacts [34] and the same-aged
female soccer players to 16.1±3.6 g LA [35]. Another investigation of women soccer players reported a median LA
of 12.51 (range 10.0–66.06 g) during games [36]. Despite the magnitude of the impact, the frequency of RSHIs
sustained across a career or season has also been potentially linked with adverse
brain health effects [9]
[37]. Male youth football athletes
experience on average 582.8±444.3 (range of 86–1996) subconcussive head impacts over
a single season [38] while another
investigation of 95 male high school footballers revealed a mean number of 652
impacts (range of 5–2235) across a single season of play [39]. The average number of subconcussive
head impacts across one season of play has been reported in women soccer players
with 142.9±118.8 (range of 86.9–189.3) [35] and 79 in male soccer [40]. Thus, we assume that a combination of high acceleration events in
sports and its repetitive occurrence may have negative consequences on brain
function and anatomy.
However, there is still a lack of understanding of the relationship between specific
head acceleration metrics in RSHIs and their consequences on brain health in
athletes. RSHIs and SRC may share some common structural changes, such as axonal
integrity loss and axonal swelling [4]
[18]
[19]
[27]
[28]. In contrast to SRCs
[5]
[29], functional deficits seem not present
after RSHI exposure [21]
[22]. Whether RSHIs affect specific
cognitive domains, which should be detectable, is yet to be fully clarified.
Additionally, the concrete structural changes induced by RSHIs have not been linked
with specific head acceleration metrics. In fact, the threshold of 10 g is still
debated [9]
[11]
[22]. To our knowledge, no recent systematic review has been conducted
exploring the specific head acceleration metrics (i. e. number of subconcussive
impacts, average linear acceleration, mean rotational acceleration) in RSHIs and
their consequences on brain anatomy and function. Therefore, the aim of this work is
to provide a systematic review of head acceleration metrics in sports and their
functional and anatomical consequences on brain health in athletes exposed to
RSHIs.
Materials and Methods
This systematic review is reported according to the Preferred Reporting Items for
Systematic Reviews and Meta-Analysis (PRISMA) 2020 guidelines [41].
Search strategy
A two-fold literature search was conducted in the following electronic databases
up until July 25, 2023: Web of Science, PubMed, and SPORTDiscus. The search
strings were developed using keywords and by applying the ‘OR’ and ‘AND’ Boolean
operators. The specific search string for the cognitive research domain is
listed in [Table 1] and the one for
the anatomical facet in [Table
2].
Table 1 Search terms cognitive changes and
RSHIs.
|
Boolean operator
|
Search terms
|
|
AND
AND
|
(‘collision sport’ OR ‘contact sport’ OR ‘athlete*’ OR
‘sport’ OR ‘performance’ OR ‘player’)
(‘head impact*’
OR ‘head kinematic*’ OR ‘head acceleration’ OR ‘head
biometric*’ OR ‘sub-concussive’ OR ‘repetitive head
impact’)
(‘neurocogn*’ OR ‘cognit*’ OR ‘neurolog*’ OR
‘function’ or ‘executive function’)
|
Note: Utilized in the databases PubMED, SPORTDiscus, and Web of
Science for cognitive change outcomes; the asterisk is used as a
wildcard symbol that broadens a search.
Table 2 Search terms anatomical changes and
RSHIs.
|
Boolean operator
|
Search terms
|
|
AND
AND
|
(‘collision sport’ OR ‘contact sport’ OR ‘athlete*’ OR
‘sport’ OR ‘performance’ OR ‘player’)
(‘head impact*’
OR ‘head kinematic*’ OR ‘head acceleration’ OR ‘head
biometric*’ OR ‘sub-concussive’ OR ‘repetitive head
impact’)
(‘brain imag*’ OR ‘structural change’ OR
‘MRI’ OR ‘DTI’ OR ‘anatomic* change’ OR ‘white matter’ OR
‘diffuse axonal injury’ OR ‘gray matter’)
|
Note: Utilized in the databases PubMED, SPORTDiscus, and Web of
Science for anatomical change outcomes; the asterisk is used as a
wildcard symbol that broadens a search.
Inclusion criteria
The eligibility criteria are described in line with the PICOS guidelines [42]
[43] and can be found in [Table 3]. As the literature search
was conducted in a two-fold way, we incorporated common inclusion and exclusion
criteria, which are further specified according to the two research questions,
i. e. cognitive and anatomical changes.
Table 3 PICOS criteria for study inclusion.
|
PICOS
|
Inclusion criteria
|
|
Population
|
Active contact sport athletes of all ages and gender
|
|
Intervention
|
Repetitive subconcussive head impact exposure (>10 g)
|
|
Comparison
|
With or without control group
|
|
Outcomes
|
Cognitive: Head acceleration metrics, changes in cognitive
domains (e. g. processing speed, working memory, attention,
reaction time, auditory verbal learning, arithmetic tasks,
task switching)
|
|
Anatomical: Head acceleration metrics, changes in anatomical
metrics (e. g. fractional anisotropy, mean diffusivity,
radial diffusivity, axonal radiation)
|
|
Study type
|
Prospective cohort studies
|
Common inclusion criteria were: (1) written in English language, (2)
peer-reviewed full-text articles, (3) experimental prospective study design, (4)
quantitative in-vivo assessment of head acceleration exposure>10 g, (5)
report of at least one head acceleration metric over the course of the
observation period (e. g. LA, RA, number of subconcussive head impacts).
The subsequent inclusion criteria were chosen to identify articles with a focus
on cognitive changes after RSHIs: (1) utilization of a validated neurocognitive
assessment tool, and (2) report of the repeated measures pre- to post
neurocognitive test performance. Articles targeting anatomical change parameters
were identified according to the following criteria: (1) using a validated
structural brain-imaging method (i. e. DTI or MRI), (2) report of the repeated
measures pre- and post-structural changes of the brain.
Exclusion criteria
In general, studies with one of the following criteria have been excluded: (1)
self-reported head acceleration assessment, (2) retrospective study design, (3)
no in-vivo head acceleration assessment, (4) no respective acceleration data
reported, (5) did not exclude concussed subjects from their analysis, (6) no
report of the repeated measures change parameters (cognitive / anatomical), and
(7) in case of anatomical changes, if they only used functional imaging methods
(i. e. Electroencephalography (EEG), Functional near-infrared spectroscopy
(fNIRS), or Functional magnetic resonance spectroscopy (fMRS)).
Study selection
Record screening was processed using the reference management software Rayyan
[44]. In the beginning, all
duplicates were removed from the literature via automatic detection. The studies
were then reviewed by title and labelled as relevant, irrelevant, or unclear.
Records that were deemed not relevant to one of the research questions have been
removed from further screening. Then, abstracts and titles of articles with
relevance or unclear relevance were screened, and if deemed not relevant, they
got excluded. In the final step, the remaining articles were screened in
full-text version and, if necessary, removed.
Risk of bias assessment
We applied the Newcastle-Ottawa Scale (NOS) for prospective cohort studies [45] to identify the risk of biased
results within the included studies. This checklist includes eight items,
grouped into three categories: (1) group selection, (2) comparability, and (3)
outcome. The assessment was done by two independent raters.
Data extraction
The relevant data from the included studies was extracted and used to populate
summary outcome tables. Extracted information included: the first name of the
author, publication year, sample size, participant information, existence of a
control group, the lengths of exposure, utilized cognitive assessment,
accelerometer device, recording threshold, imaging technique, head acceleration
metrics (either expressed as Median (Mdn) or Mean (M) values),
cognitive/anatomical outcome data, and follow-up measurement timing.
Results
Identification of studies
Regarding the cognitive research domain, we identified 1,470 records in the
initial database search. Three additional records have been identified through
reference searching and author identification. After automatically removing
duplicates and an assessment of eligibility, a total of 14 articles have been
included. The article identification process is presented in [Fig. 1]. The literature search
targeting anatomical changes resulted in a total of 256 records. Additionally,
two records were included from reference searching and author identification.
After screening and combined with the additional records, a total of 11 records
were eligible. [Fig. 2] illustrates
the full literature identification process.
Fig. 1 PRISMA flow diagram of study identification and selection
process for cognitive change parameters [41].
Fig. 2 PRISMA 2020 flow diagram of study identification and
selection process for anatomical change parameters [41].
Risk of bias assessment
The risk of bias assessment of the included studies did not reveal any
inter-rater conflicts. Overall, 20 out of 24 included studies hold at least six
out of nine possible stars. The individual score of each included record is
listed in [Table 4]. Different
scores were mainly identified in the sections of Item 2 and 5, targeting the
inclusion of a control group (Item 2) and the existence of control variables in
the respective data analysis (Item 5).
Table 4 Newcastle-Ottawa Quality Assessment Scale scores
for included cohort studies.
|
First author (year)
|
Item #1
|
Item #2
|
Item #3
|
Item #4
|
Item #5
|
Item #6
|
Item #7
|
Item #8
|
Total (out of 9)
|
|
Asselin (2020)
|
*
|
–
|
*
|
*
|
*
|
*
|
–
|
–
|
5
|
|
Bazarian (2014)
|
*
|
*
|
*
|
*
|
**
|
*
|
*
|
*
|
9
|
|
Breedlove (2014)
|
*
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
Broglio (2018)
|
*
|
*
|
*
|
*
|
**
|
*
|
*
|
–
|
8
|
|
Caccese (2019)
|
*
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
Chrisman (2016)
|
*
|
–
|
*
|
*
|
**
|
*
|
–
|
–
|
6
|
|
Chrisman (2019)
|
*
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
Chun (2015)
|
*
|
*
|
*
|
*
|
*
|
*
|
*
|
*
|
8
|
|
Doan (2022)
|
*
|
–
|
*
|
*
|
*
|
*
|
–
|
–
|
5
|
|
Gong (2018)
|
*
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
Kelley (2021)
|
*
|
–
|
*
|
*
|
–
|
*
|
*
|
*
|
6
|
|
Manning (2020)
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
Marchesseault (2018)
|
*
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
McAllister (2012)
|
*
|
*
|
*
|
*
|
**
|
*
|
*
|
*
|
9
|
|
McAllister (2014)
|
*
|
*
|
*
|
*
|
*
|
*
|
*
|
*
|
8
|
|
Myer (2016a)
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
Myer (2016b)
|
–
|
*
|
*
|
*
|
**
|
*
|
*
|
*
|
8
|
|
Myer (2019)
|
–
|
*
|
*
|
*
|
**
|
*
|
*
|
*
|
8
|
|
Rose (2019)
|
–
|
–
|
*
|
*
|
*
|
*
|
–
|
*
|
5
|
|
Rose (2021)
|
–
|
–
|
*
|
*
|
–
|
*
|
–
|
*
|
4
|
|
Slobounov (2017)
|
*
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
Stojsih (2010)
|
*
|
*
|
*
|
*
|
*
|
*
|
–
|
*
|
7
|
|
Talavage (2014)
|
*
|
–
|
*
|
*
|
*
|
*
|
*
|
*
|
7
|
|
Yuan (2017)
|
–
|
*
|
*
|
*
|
**
|
*
|
*
|
*
|
8
|
Data extraction and synthesis
Data were extracted from the included records and used to provide an overview of
all 24 studies that have been identified as relevant for one of the research
domains. [Table 5] provides an
outcome summary for all 14 articles investigating cognitive changes. The
detailed outcomes of the 11 included articles for the anatomical changes are
listed in [Table 6].
Table 5 Cognitive changes after RSHI exposure in
athletes.
|
First author (Year)
|
Participant information
|
Exposure to RSHIs
|
Specific measurement information
|
Impact sensor (threshold)
|
Mean follow-up time interval
|
Significant cognitive outcomes (pre- to post)
|
|
Sample size (sex)
|
Mean Age (SD)
|
Sport
|
Competitive level
|
Control group
|
|
Asselin (2020)
|
n=28 (M)
|
19.8 a
|
FB
|
NCAA III
|
No
|
One season
|
ImPACT
|
HITS (> 10g)
|
1 week
|
↔
|
|
Bazarian (2014)
|
n=10 (M)
|
20.4 (NR)
|
FB
|
NCAA III
|
Yes
|
One season
|
ImPACT
|
HITS (> 10g)
|
NR
|
↔
|
|
Breedlove (2014)
|
n=13 (M)
|
14–18 b
|
FB
|
High school
|
No
|
One season
|
ImPACT
|
HITS (NR)
|
NR
|
↓ and ↔
|
|
Broglio (2018)
|
n=46 (M)
|
15.9 (0.8)
|
FB
|
High school
|
Yes
|
One season
|
CCAT
|
HITS (NR)
|
1–2 months b
|
↑ and ↔
|
|
Caccese (2019)
|
n=38 (M, F)
|
M: 19.7 (1.2), F: 19.5 (1.0)
|
Soccer, FB
|
NCAA Division I
|
No
|
One season
|
ImPACT, K-D
|
HITS (> 10g)
|
Within 1 week
|
↑ and ↔
|
|
Chrisman (2016)
|
n=17 (M, F)
|
12.6 (1.0)
|
Soccer
|
NR
|
No
|
Weekend tournament
|
ImPact, K-D
|
xPatch (> 10g)
|
2 days
|
↔
|
|
Chrisman (2019)
|
n=46 (M, F)
|
12.27 (NR)
|
Soccer
|
NR
|
No
|
1 month
|
ImPACT, K-D
|
xPatch (> 15g)
|
NR
|
↔
|
|
Doan (2022)
|
n=27 (M)
|
20.7 (1.8)
|
Boxing
|
NCBA
|
No
|
2x 2 min sparring bouts
|
ImPACT, ANAM
|
IBH (> 9.6g)
|
39.50±2.50 minutes
|
↑ and ↓
|
|
Marchesseault (2018)
|
n=15 (M)
|
21.1 (1.5)
|
Lacrosse
|
NCAA III
|
No
|
One season
|
CTMT, SCWT
|
SIM (> 15g)
|
1 week
|
↑ and ↔
|
|
McAllister (2012)
|
n=45 (M, F)
|
19.0 (1.3)
|
FB, Ice Hockey
|
Collegiate
|
Yes
|
One season
|
ImPACT, WRAT-IV, PASAT C, CVLT-II
|
HITS (> 14.4g)
|
25.0±31.0 days
|
↑ and ↔
|
|
Rose (2019a)
|
n=55 (M)
|
P.S.: 10.8 (0.5)
H.S.:
15.9 (0.6)
|
Youth Tackle FB
|
Varsity
|
No
|
Two seasons
|
CCAT, MSVT, WISC-V, ChAMP, TMT, WASI-IV, WASI-II,
T.O.V.A.
|
Ridell InSite (S1:> 10g, S2:> 15g)
|
NR
|
↑, ↓, and ↔
|
|
Rose (2021)
|
n=18 (M)
|
10.6 (0.64)
|
Youth Tackle FB
|
NR
|
No
|
Four seasons
|
CCAT, MSVT, WISC-V, ChAMP, TMT, WASI-IV, WASI-II, T.O.V.A.
|
Ridell Insite (S1:> 10g, S2-S4:> 15g)
|
NR
|
↑ and ↓
|
|
Stojsih (2010)
|
n=55 (M, F)
|
M:22.0
(NR)
F:24.0 (NR)
|
Boxing
|
Amateur
|
No
|
4x 2min sparring bouts
|
ImPACT
|
IBH (> 9.6g)
|
Post: 30 min
Post24: 24h
|
↓ and ↔
|
|
Talavage (2014)
|
n=10 (M)
|
15–19 b
|
Football
|
Varsity, High school
|
No
|
One season
|
ImPACT
|
HITS (> 14.4g)
|
1–3 months b
|
↓ and ↔
|
Note. ↓=sign. decrease; ↑=sign. increase; ↔=no sign. change.
a reported as Mdn,
b reported as range.
ANAM=Automated Neuropsychological Assessment Metrics;
ANT=Attention Network Test; CCAT=CogState Computerized
Cognitive Assessment Tool; ChAMP=Child And Adolescent Memory Profile,
CSx=accelerometer (CSx Systems Ltd; Auckland; New Zealand);
CTMT=Comprehensive Trail Making Test; CVLT-II=Comprehensive
Verbal Learning Test; F=Females; FB=Football; GFT=GForceTracker
accelerometers (GForceTracker, Markham, ON, Canada); HITS=Head Impact
Telemetry System; IBH=The Impact Boxing Headgear;
ImPACT=Immediate Post-Concussion Assessment and Cognitive Tool;
H.S.=High School; K-D=King-Devick Test; M=Male;
MSVT=Medical Symptom Validity Test; NCAA=National Collegiate
Athletic Association; NCBA=National Collegiate Boxing Association;
NR=not reported; PASAT C=Paced Auditory Serial
Addition Test; P.S.=Primary School; RSHIs=repetitive
subconcussive head impacts; S1=Season 1; S2=Season 2; SCWT=Stroop
Color and Word Test; SIM=Smart Impact Monitors; T.O.V.A.=Test of
Variables of Attention; TMT=Delis-Kaplan Executive Function
System Trail Making Test; WASI-II=Wechsler Abbreviated Scale of
Intelligence 2nd Edition/ 4th Edition ;
WISC-V=Wechsler Intelligence Scale for Children 5th Edition;
WRAT-IV=The Wide Range Achievement Test 4th Edition.
Table 6 Anatomical changes after RSHI exposure in
athletes.
|
First author (Year)
|
Participant information
|
Exposure to RSHIs
|
Specific measurement information
|
Impact sensor (threshold)
|
Mean follow-up time interval
|
Significant brain structure alteration (pre- to post)
|
|
Sample size
|
Mean Age (SD)
|
Sport
|
Competitive level
|
Control group
|
|
Asselin (2020)
|
n=28 (M)
|
19.8 a
|
FB
|
NCAA III
|
No
|
One season
|
DTI
|
HITS (> 10g)
|
1 week
|
↓ FA values
|
|
Chun (2015)
|
n=28 (M)
|
16.7 (0.7)
|
FB
|
High school
|
Yes
|
Two seasons
|
DTI
|
HITS (NR)
|
3–5 months b
|
↑ FA values
↓ FA values d
|
|
Gong (2018)
|
n=16 (M)
|
16.0 a
|
FB
|
High school
|
No
|
One season
|
DTI
|
HITS (> 10g)
|
10 days a
|
↑ MD values
|
|
Kelley (2021)
|
n=19 (M)
|
12.1 a
|
FB
|
Youth FB
|
No
|
Two seasons
|
DTI
|
HITS (NR)
|
NR
|
↓ FA values, ↑ MD values
|
|
Manning (2020)
|
n=60 (F)
|
20.13 (1.43)
|
Rugby
|
Varsity Rugby
|
Yes
|
Two seasons
|
DTI
|
GFT (> 15g)
|
2–3 months b
|
↓ FA values, ↑ MD values, ↑ AD values, ↑ RD values
|
|
McAllister (2014)
|
n=80 (M, F)
|
19.5 (1.3)
|
FB, Ice Hockey
|
Collegiate
|
Yes
|
One season
|
DTI
|
HITS (14.4g)
|
NR
|
↔
|
|
Myer (2016a)
|
n=15 (M)
|
16.3 (1.2)
|
Hockey
|
Varsity high school
|
No
|
Half a season
|
DTI
|
GFT (> 20g)
|
2.9±1.8 days
|
↑ MD values, ↑ RD values
|
|
Myer (2016b)
|
n=21 (M)
|
17.13 (0.66)
|
FB
|
Varsity high school
|
No
|
One season
|
DTI
|
GFT (> 20g)
|
7.05±4.61 days
|
↓ MD values, ↓ AD values
|
|
Myer (2019)
|
n=22 (F)
|
15.93 (1.04)
|
Soccer
|
Varsity high school
|
No
|
One season
|
DTI
|
xPatch (> 10g)
|
3.73±4.33 days
|
↓ MD values, ↓ RD values
|
|
Slobounov (2017)
|
n=18 (M)
|
21.6 (1.28)
|
FB
|
Collegiate
|
No
|
One season
|
DTI
|
BodiTrack (> 25)
|
1–7 days b
|
↔ FA, MD, AD, RD
|
|
Yuan (2017)
|
n (S1)=10 (M)
n (S2)=7 (M)
|
(S1): 16.90 a
, (S2): 17.73
a
|
FB
|
High school
|
No
|
Two seasons
|
DTI
|
GFT (> 10g)
|
S1: 12.71±7.87 days
S2: 5.18±2.48 days
|
↓ MD values, ↓ AD values, ↓ RD values
|
Note. ↓=sign. decrease; ↑=sign. increase; ↔=no sign. change.
a reported as Mdn.,b reported as range,
c Team 1, d Team 2. AD=axonal diffusivity;
ANAM=Automated Neuropsychological Assessment Metrics;
ANT=Attention Network Test; CCAT=CogState Computerized
Cognitive Assessment Tool; ChAMP=Child And Adolescent Memory Profile,
CSx=accelerometer (CSx Systems Ltd; Auckland; New Zealand);
CTMT=Comprehensive Trail Making Test; CVLT-II=Comprehensive
Verbal Learning Test; DTI=Diffusion tensor imaging;
F=Females; FA=fractional anisotropy; FB=Football;
GFT=GForceTracker accelerometers (GForceTracker, Markham, ON, Canada);
HITS=Head Impact Telemetry System; IBH=The Impact Boxing
Headgear; ImPACT=Immediate Post-Concussion Assessment and
Cognitive Tool; H.S.=High School; K-D=King-Devick Test;
M=Male; MD=mean diffusivity; MSVT=Medical Symptom
Validity Test; NCAA=National Collegiate Athletic Association;
NCBA=National Collegiate Boxing Association; NR=not reported;
PASAT C=Paced Auditory Serial Addition Test; P.S.=Primary
School; RD=radial diffusivity; RSHIs=repetitive subconcussive head
impacts; S1=Season 1; S2=Season 2; SCWT=Stroop Color and Word
Test; SIM=Smart Impact Monitors; T.O.V.A.=Test of Variables of
Attention; TMT=Delis-Kaplan Executive Function System Trail
Making Test; WASI-II=Wechsler Abbreviated Scale of Intelligence
2nd Edition/ 4th Edition ; WISC-V=Wechsler
Intelligence Scale for Children 5th Edition;
WRAT-IV=The Wide Range Achievement Test 4th Edition.
Functional outcomes and RSHIs
Four out of 14 included articles did not find any significant changes from pre-
to post testing in cognitive performance of the athletes after head impact
exposure [35]
[46]
[47]
[48]
[49]. All of the investigated subjects
competed at an amateur level. A total of ten articles did identify significant
cognitive changes in cognitive performances from the baseline to the
post-testing after head impact exposure. All results from these ten studies
showed mixed outcomes regarding improved, decreased, and no change in cognitive
performance. Four of the studies identified improved outcomes and no change
compared to the baseline [34]
[35]
[50]
[51]. Two of the 14 studies found both
improved and decreased cognitive performances across different domains [52]
[53]. One study reported increased,
decreased, and no change in performance [54], and three found decreased and no change in performance in post
testing [38]
[55]
[56]. Improved performance has been
obtained in the following specific cognitive domains: learning and working
memory (WM) speed, reaction time, arithmetic processing, processing speed,
visual attention, and coding [34]
[35]
[52]
[53]
[54]. The overall cognitive
performance increased in the Comprehensive Trail-Making Test (CTMT), California
Verbal Learning Test (CVLT-II), Paced Auditory Serial Addition Test (PASAT), and
the Child and Adolescent Memory Profile (ChAMP) [50]
[54]
[57]. Decreased cognitive performance
was reported in the following distinct domains: memory functioning, processing
speed, and response time [52]
[53]
[54]
[55]
[56], as well as in the composite
score [38]. No change compared to
the baseline measurement was reported in the overall score of the Immediate
Post-Concussion Assessment and Cognitive Testing (ImPACT), CogState, Stroop
task, Wechsler Intelligence Scale (WISC-V), and Test of Variables of Attention
(TOVA) [35]
[38]
[46]
[47]
[48]
[49]
[50]
[54]
[56]
[57], as well as in these specific
domains: processing speed, attention, WM accuracy [34], verbal and visual memory, visual
motor speed and reaction time [55].
In total, seven of the 14 included articles reported the respective statistical
association between the obtained head impact metrics and cognitive change
parameters from pre to post. Seven out of these 14 investigation did not find
any significant relationship between cumulative head impact metrics (i. e. total
number of impacts sustained) and cognitive change characteristics [38]
[49]
[52]
[53]
[54]
[55]
[56]. Another investigation with
college-aged FB and Ice Hockey athletes identified a significant relationship
between the peak linear acceleration and the composite score reaction time of
the ImPACT, indicating worsened cognitive performance after greater impact
exposure [57]. A groupwise analysis
revealed that players with detected performance reductions sustained more
subconcussive impacts than groups with no changes [56].
Anatomical outcomes and RSHIs
A total of nine out of 11 articles reported significant structural changes after
measuring subconcussive head impact exposure over time. On the other hand, only
one out of 11 identified structural changes after the first season of
subconcussive head impact exposure but not after the second season [58]. Two studies did not find any
significant anatomical changes in after RSHIs [51]
[59]. The included studies only
investigated amateur athletes. The changes were related to decreased fractional
anisotropy (FA) values compared to the baseline measurement in three studies
[46]
[60]
[61]
[62]. Four out of 11 studies found
increased mean diffusivity (MD) values [60]
[61]
[63]
[64]. One study reported increased FA
values (which was only identified in one of the two observed teams) [62]. A total of three out of 11
included articles found reduced MD values over time [58]
[65]
[66]. The change outcomes of axial
diffusivity (AD) showed the following: Three studies reported reduced AD values
over time [58]
[65]
[66], and one study found increased AD
values in their sample [61]. Only
one author reported increased AD values over time [61]. Lastly, radial diffusivity (RD)
decrease was identified in three articles [58]
[65]
[66]. A subset of two articles
highlighted increased RD values over time in the corpus callosum and the
brainstem [61]
[64]. A total of seven out of 11
included studies in this section reported a significant relationship between
cumulative head impact metrics (i. e. number of hits sustained) and anatomical
changes, detected as white matter changes [46]
[51]
[60]
[62]
[63]
[65]
[66]. They investigated FB athletes in
college-age [46], high school-age
[62]
[63]
[65] and youth-age [60], collegiate FB and ice hockey
[51] and high school soccer
athletes [66]. Only one study, which
studied high school FB players, did not identify a significant association
between the obtained head impact metrics and the white matter changes (i. e.
reduced MD values, reduced AD values, reduced RD values) [58].
Discussion
The aim of this systematic review was to answer the question of whether specific head
acceleration metrics in RSHIs affect the brain’s anatomy and function. We therefore
reviewed prospective cohort studies that examined pre- and post-changes in cognitive
functions as well as anatomical changes in the athletes’ brains exposed to
RSHIs.
Brain function and RSHIs
According to this research, the findings on functional (cognitive) changes after
RSHIs are mixed. Additionally, mixed findings are obtained regarding the causal
relationship between the total number of subconcussive hits and functional
changes. In fact, some of the reviewed studies did not find an association
between functional changes and the total number of RSHIs. This has been reported
after a whole season of RSHI exposure [53]
[54], a weekend
tournament [49], and a set of
sparring bouts [52]. In addition, it
seems to be independent of the type of sport [49]
[52]
[53]
[54]. This indicates that immediate
cognitive deficits are not associated with RSHI exposure in active
collision-sport athletes, which has been supported in studies investigating
former [67] and recently active
collision sport athletes [68]
[69]
[70]. We could further highlight this
in studies using a non-contact control group. They found no group [47] or group by time difference [34]
[57] in overall cognitive performance
between contact and non-contact athletes. In contrast to this, a cohort of
female soccer and male football athletes showed a relationship between impaired
visual memory, visual processing speed, and the total number of subconcussive
impacts [35]. But only to the total
number of subconcussive impacts>98 g LA [35]. No such relationship was
observed between the total number of RSHIs above 10 g. Furthermore, the peak
linear acceleration sustained during RSHIs was associated with reduced reaction
time in a group of male ice hockey and football players [57]. Together, this indicates that
RSHIs above 10 g do not induce cognitive deficits, such as impaired memory and
processing speed. Only the maximal peak linear acceleration during RSHIs might
induce cognitive deficits. In contrast, reduced psychomotor speed and lower
verbal learning abilities related to the frequency of subconcussive
impacts>10 g have been obtained in soccer athletes [71]. Monetenigro et al. (2017)
hypothesized that a cumulative number of 7,251 subconcussive impacts needs to be
exceeded to induce cognitive impairments in athletes [72]. In fact, two of the included
studies reported a pattern between reduced cognitive performance and a high
number of subconcussive impacts. Talavage and colleagues (2014) were able to
find players with more subconcussive hits to be more likely to show signs of
cognitive performance reductions after one season [56], which builds on previous work
[73]. Rose et al. (2021)
identified deficits in processing speeds in youth tackle footballers after
receiving a composite mean number of 15,766 subconcussive head impacts over the
period of four consecutive seasons [53]. Thus, longer careers, which are expected to produce higher
numbers of subconcussive impacts [22], might be a risk factor for cognitive dysfunctions [74].
Opposing findings of athletes exposed to RSHIs above 10 g and improvements in
cognitive domains over time such as learning and working memory (WM) speed,
arithmetic processing, and processing speed also exist [34]
[35]
[52]
[53]
[54]. Actively competing athletes may
still benefit from the health effects of sports on cognition [23]
[76], nullifying potential cognitive
declines. Importantly to note, all subjects investigated in the reviewed studies
were non-professional athletes, mostly competing on a collegiate level [35]
[46]
[47]
[50]
[52]
[57]. The level of play seems to be an
important consideration in determining whether athletes are at greater risk of
overt cognitive performance deficits. Former amateur athletes, compared to
professional ones, do not show signs of neurocognitive changes on a
metacognition scale [77], as stated
in other retrospective reports [75]
[78]. A possible
explanation might be the fact that professional athletes are exposed to higher
cumulative numbers of subconcussive head impacts [22]
[79]. This suggests that the most
harmful component for reduced brain function may be the total number of
subconcussive hits experienced over time, particularly prevalent among
professional athletes. Furthermore, one amateur season of exposure might be too
short to induce overt functional changes.
Brain anatomy and RSHIs
The current work shows that exposure to RSHIs exceeding a threshold of 10 g leads
to white matter alterations in the athletes’ brains [46]
[58]
[60]
[61]
[62]
[63]
[64]
[65]
[66]. Specifically, the structural
changes seem to be related to the total number of subconcussive head
impacts>10 g [46]
[51]
[60]
[63]
[65]
[66]. White matter alterations are
mainly characterized by reduced fractional anisotropy (FA) values [46]
[60]
[61]
[62]. Reduced FA, which is regularly
identified in athletes exposed to repetitive subconcussive head impacts [20]
[21]
[22], is interpreted as a signal of
damage to the axons‘ structure and to the myelin sheath surrounding it [20]. The microstructural damage to
the axonal architecture constitutes a marker for neurodegenerative and
neuroinflammatory processes in the brain [19]. Axonal damage may negatively affect the later-life vitality of
contact-sport athletes [28]
[80]. This appears independently of
age, type of sport, and even after one single season of play [46]
[60]
[61]. The reduction of FA values
suggests that even young contact-sport athletes, after one season of play,
already suffer from detectable structural changes. Additionally, a relationship
exists between the total number of RSHIs sustained and the FA value reduction
[46]
[60]
[62]. This dose-response relationship
highlights an accumulative risk of structural axonal damage with longer exposure
periods [81]
[82].
Additionally, microstructural brain alterations after RSHIs have been identified
as an increase in mean diffusivity (MD) [60]
[61]
[63]
[64] values over time. This represents
a chronic irreversible injury state of the brain‘s tissue [19]
[83]. Chronic axonal injury has been
previously observed in youth athletes across different types of contact-sports,
including American football, hockey, and rugby [60]
[61]
[63]
[64]. Compared to a non-contact
control group, increased MD values were only visible in contact-sport athletes
[51]
[61]. Increased MD values remained
after a non-exposure period of three months [61]. This could highlight the danger
of long-lasting irreversible structural changes in contact-sport athletes
exposed to RSHIs [3]. Furthermore,
the number of subconcussive head impacts seems to be related to this MD increase
[60]
[63], suggesting a dose-response
relationship between impact frequency and white matter alteration (i. e. MD
increase) [81]
[82]. In contrast, decreased MD values
are also reported in athletes exposed to RSHIs [58]
[65]
[66], which are related to the number
of subconcussive head impacts [65]
[66]. A decrease in MD
values over time has been associated with an ongoing recovery phase of the
brain’s tissue [84], as well as a
marker of extracellular space compression, axonal swelling, and inflammatory
processes [83]. This injury pattern
has already been reported after one season of RSHI exposure across youth soccer
and football athletes [58]
[65]
[66]. Even at a young age, athletes’
brains seem to show critical signs of structural alterations [20]. Of note, the variability in MD
changes identified in our review might be explained by the variation in
follow-up test timing after the athletes’ last impact exposure [19], as well as the differing
severity and chronicity of brain injury mechanisms potentially accompanying
RSHIs [83]. The white matter
alterations after RSHIs, indicated by reduced FA [46]
[60]
[61] and increased MD [60]
[61]
[63]
[64] values, are similarly reported in
retired concussed athletes [27]
[28]. These signs of brain injury
might represent a common structural injury symptomatology in SRCs and RSHI
exposure in collision sport athletes.
Lastly, athletes did show brain tissue changes in terms of reduced axial
diffusivity (AD) and radial diffusivity (RD) values after RSHIs [58]
[65]
[66]. Reduced AD and RD values have
been generally interpreted as a sign of axonal dysfunction and a loss of axonal
membrane integrity [19]. Thus,
repeated subconcussive head impacts above 10 g induce white matter alterations.
Furthermore, the white matter alterations are related to the total number of
sustained RHSIs.
Linkage between functional and anatomical changes
As functional deficits are hypothesized to be a symptom of structural impairments
[21], the current work aims to
investigate this neurocognitive interplay. However, the results of the
association between structural and functional changes are mixed. As RSHIs above
10 g alter white matter structures, such as the corpus callosum (CC) [58]
[64]
[65], they should induce functional
deficits [85]. In turn, this would
be critical in complex neural execution situations like motor-control and highly
relevant in dynamic sports situations [4]. However, there was no such correlation reported between
structural and functional impairments. We could show that no overt cognitive
deficits can be expected after one single season of RSHI exposure>10 g in
amateur athletes [49]
[52]
[53]
[54]. In contrast, there might be a
dose-response relationship between the total number of hits and the occurrence
of cognitive deficits [53]
[56]
[71]. First, findings from cognitive
neuroscience propose beneficial effects of sport and exercise on cognition [23]
[86], which might positively
contribute to the prevention of functional deficits after RSHIs>10 g. In
addition, the brain’s ability to compensate for structural alterations (i. e.
neural plasticity) [87]
[88] might be another explanation for
the absence of noticeable cognitive impairments in the present study. For this
reason, the brain of active amateur sports athletes might benefit from its
compensatory and repair mechanisms in order to prevent the onset of functional
impairments after structural brain damage. As discussed previously, amateur
collision sport athletes, compared to their professional counterparts, sustain a
lower number of subconcussive head impacts over time [22]
[79]. Only former professional
collision sport athletes show signs of clinical symptoms (incl. overt functional
deficits) after their active careers [75]. The higher number of subconcussive impacts experienced by
professionals suggests that functional symptoms slowly develop over time. In
line with the dose-response hypothesis, it appears that structural damage is
only severe enough to cause noticeable functional changes if a certain number of
impacts are sustained. However, it is yet to be clarified which specific
threshold of structural damage is critical to producing visible functional
impairments. Lastly, if the structural changes to the white matter structure
should, contrary to the results, affect cognition [85], the cognitive tests utilized
(i. e. ImPACT, ANAM, or CogState) may lack sensitivity to detect overt cognitive
changes after one single season of play [5]
[21]
[89]. We also obtained a high level of
methodological heterogeneity in the timing of the follow-up cognitive
assessments across the included studies. This makes it difficult to draw a
causal conclusion between the possible onset and persistence of functional
changes. As a result, we conclude that the brain’s ability to compensate for
structural brain changes, especially those present in active populations, may
protect amateur athletes from suffering from overt functional deficits. However,
to what specific extent this preventive effect may last is yet unknown. Future
research must therefore focus on the neurocognitive interplay between structural
alterations and brain functionality.
Conclusion
It is evident that amateur athletes exposed to RSHIs above 10 g suffer from
structural brain alterations. This is even present after a single season of play.
Structural alterations are mostly found within the white matter by reduced FA [46]
[60]
[61], reduced AD [58]
[65], reduced RD [58]
[66], and increased MD values [60]
[61]
[63]
[64]. These alterations display signs of
neurophysiological brain injury, such as damage to the axons’ structure, the myelin
sheath surrounding them [20],
compromised axonal integrity [19], and
signs of chronic axonal injury [19]
[83]. There exists a dose-response
relationship between such white matter abnormalities and the sustained cumulative
number of RSHIs [46]
[51]
[60]
[63]
[65]
[66]. This implies that the cumulative number, rather than a single impact
event, is causing structural damage. Mixed findings are obtained regarding the
presence of functional (cognitive) changes after one season of RSHIs. A relationship
pattern for cognitive changes over time and the total number of RSHIs>10 g seems
to be present [53]
[56]. Nonetheless, one season of play
might not be severe enough to clearly produce detectable functional changes. Thus,
we conclude that RSHIs above 10 g after one season of play induce neurophysiological
(i. e. white matter) changes that are not displayed in overt functional (cognitive)
symptoms.
