Original Article

Journal of Concussion
Volume 3: 1–7
The changing nature of concussion ! The Author(s) 2019
Article reuse guidelines:
in rugby union: Looking back sagepub.com/journals-permissions
DOI: 10.1177/2059700219860641
to look forward journals.sagepub.com/home/ccn

Thomas S Owens , George Rose , Christopher J Marley,

Thomas A Calverley, Benjamin S Stacey, Priscilla Williams,
John PR Williams and Damian M Bailey

Abstract
Introduction: Concussion is regularly observed in rugby union and has generated a growing public health concern, yet
remains one of the least understood injuries facing the sports medicine community. Evidence suggests that multiple
concussions may increase susceptibility to long-term neurological complications that present decades after the initial
injury for reasons that remain unclear. We aimed to determine the incidence rate and risk factors for concussion
amongst community-level rugby union-15s players active during the 1980s given that it may help to better understand
the risks and mechanisms of injury.
Methods: Injury data were collected from clubs by the coach at the time of injury in players using a 15-item ques-
tionnaire (1982–1984).
Results: Seventy games were recorded throughout 1982–1983 and 1983–1984 rugby union seasons. Forty-two docu-
mented concussions accounted for
6% of injuries corresponding to an incidence rate of 0.64 per 1000 playing hours,
more than a third lower than the ‘modern-day’ equivalent. Tackling (relative risk 1.60, p < 0.05), collisions (relative risk
0.95, p < 0.05) and gum shield use (relative risk 1.69, p < 0.05) were independently associated with concussion whereas
no associations were observed for ground condition, quarter of play or players playing out of position (p > 0.05).
Conclusion: Despite limitations due to the retrospective focus and reliance on questionnaire data notwithstanding
raised awareness of concussion, the incidence rate of concussion during the 1980s appears to be appreciably
lower compared to the present-day game. This is the likely outcome of improvements in the clinical understanding
of concussion, data collection tools, reporting methods and clinical management of concussive injuries, including changes
to both player and game. However, the findings of this study help better understand the risks and mechanisms of injury
once encountered by rugby union players active during the 1980s, of which some of those risks are still apparent.

Keywords
Concussion, traumatic brain injury, incidence, risk factors
Date received: 9 July 2018; accepted: 4 June 2019

Introduction Alzheimer’s disease.2,3 Subsequently in recent years,

Concussive injury has generated a growing public health focus has centred towards retired contact sport athletes
concern yet remains one of the least understood injuries who may be more susceptible to neurological sequela as
facing the sports medicine community today. Emerging a result of recurrent concussion.1
evidence suggests that multiple concussions sustained
during young adulthood may increase susceptibility to Neurovascular Research Laboratory, Faculty of Life Sciences and
long-term neurological complications that typically pre- Education, University of South Wales, Pontypridd, UK
sent decades after the initial injury for reasons that
Corresponding author:
remain unclear.1 These complications include accelerat- Thomas S Owens, Neurovascular Research Laboratory, University of
ed cognitive decline, chronic traumatic encephalopathy South Wales, Alfred Russell Wallace Building, Pontypridd CF37 4AT, UK.
(CTE), depression, Parkinson’s syndrome and Email: tom.owens@southwales.ac.uk
Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons Attribution-
NonCommercial 4.0 License (http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and
distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.
sagepub.com/en-us/nam/open-access-at-sage).
2 Journal of Concussion

Since the advent of professionalism was introduced consensus definitions and methodologies to standardise
to rugby union in 1995, participation throughout the the recording of injuries and reporting of studies which
United Kingdom (UK) has risen to 2.5 million play- was introduced by the Rugby Injury Consensus Group
ers.4 A recent meta-analysis of community sub-elite (RICG) in 2007.20 This is accompanied by a research
rugby union players identified a concussion incidence determined definition of concussion that is a ‘traumatic
rate of 2.08/1000 player match hours and speculated brain injury induced by biomechanical forces’ which is
this trend would continue to rise with the current accompanied by a number of symptoms including
reporting methods.5 Furthermore, professional rugby headache, dizziness, balance/gait abnormalities, confu-
union players in the UK are more likely than not to sion, amnesia and various others which can occur with-
sustain a concussion within 25 games,6 thus generating out loss of consciousness.21 Clinical questionnaires
future health concerns among the medical community specific to concussion including the Sports
and players alike.
Concussion Assessment Tool 5th edition and the
A number of intrinsic and extrinsic factors contrib-
Head Injury Assessment further allow certified athletic
ute towards concussion risk including: playing posi-
trainers and medical professionals alike to recognise
tion, tackling technique, use of protective head/mouth
concussion and remove athletes from play, while gov-
apparatus, neck strength, warm-up strategy, foul play,
quarter of play, ground condition and weather, includ- erning bodies have pre-defined return to play protocols
ing various others.6–10 Given that the majority of the to ensure athletes have recovered adequately before
available literature relating to concussion and risk of returning to competition.22 Despite these comparative
neurological impairment centres towards retired con- difficulties, the importance of a detailed injury history
tact sports athletes aged between 50 and 80,1 it is has time again been emphasised to be of relevance
seldom supported by injury data from those cohorts, when diagnosing neurological disorders.11
other than individual recall of past events11 and ques- Given the evidence presented, retrospective injury
tions whether today’s athletes are exposed to the same data from the period may be of relevance to retired
risks as once encountered by their senior counterparts. contact sports athletes and clinicians, to better under-
This uncertainty is likely due to the primitive recorded stand the risks and mechanisms of injuries once
injury data throughout the ‘amateur years’ of rugby encountered. We sought to determine the incidence
union. However, some rugby union injury data and corresponding in-game variables and risk factors
throughout this period in both the adult and schoolboy for concussion among Welsh rugby union players who
levels exists. Durkin12 observed injures in British adult were active during the 1980s.
rugby union players over the course of the 1972–1976
seasons and observed that 5.6% of all injures were
concussions. Sparks13 recorded over half a million
Methods
hours of schoolboy rugby between 1950 and 1980 and Participants
observed 9885 injuries, of which 513 (5.2%) were con-
cussion. South African school boy rugby injuries were Information was obtained from a total of 708 college
reported by Nathan et al.14 and Roux et al.15 who and senior level rugby union-15s players from clubs
observed concussion in 22% and 12% of all injuries across Wales using a 15-item questionnaire (Figure 1)
respectively. at the time of injury by the team coach between the
In addition, the understanding of concussion has 1982–1983 and 1983–1984 rugby union seasons. All
improved, thus improving the standard of clinical players and coaches from the selected clubs were invit-
care provided to those with suspected injury and ed to participate. All players and coaches who partici-
making comparisons between the ‘amateur’ and pre- pated provided written and verbal informed consent
sent game difficult. For instance, throughout the begin- with data collection overseen by a general practitioner
ning of the 1980s, a concussion was clinically defined as and consultant orthopaedic surgeon-player.
a loss of consciousness or a loss of awareness following
a blow to the head,16 which was later evolved by Procedures
Cantu17 into three categories (mild, moderate or
severe). The long-term neurological consequences of Questionnaires included intrinsic and extrinsic factors
concussion were poorly understood and while literature associated with injuries and each player was assigned
had documented dementia pugilistica18 among boxing an identification code with anonymised datasets
cohorts and later described as CTE,19 no such evidence subsequently uploaded to a computer database for
existed in rugby union. analysis. Concussion was defined by loss of conscious-
Data collection for all injuries in rugby union have ness or a loss of awareness following a blow to the
improved drastically in the modern day due to the head,16 including symptoms of amnesia (personal
Owens et al. 3

Figure 1. Fifteen-item questionnaire for injury reporting.

communication with general practitioner). The inci- square (v) tests were used however in the event that
dence rate was subsequently calculated as more than 20% of variables had expected cell counts
below five, likelihood ratios (LH) were calculated as a
Total number of players with concussion
  1000ðhÞ surrogate measure.23 Throughout association tests, rela-
Exposure ðnumber of matches tive risk (RR) of injury were computed simultaneously
match duration  number of playersÞ and incidence rates were calculated for all variables
manually. Players with cases of missing data for the
Statistical analysis observed factor were excluded from the overall analyses.

Statistical analyses were completed using the Statistics

Package for Social Scientists (Version 23.0). Data were Results
first categorised into concussion or other injuries for Seventy games were observed throughout 1982–1983
each category observed via the 15-item questionnaire. and 1983–1984 rugby union seasons among 708 play-
To determine association of injury, 2  2 Pearson chi ers. A total of 178 injuries were classified as head
4 Journal of Concussion

Table 1. Factors associated with concussion incidence.a

Frequency Percent IR RR 95% CI p value

Playing position
Front row 4 11 0.06 0.32 0.12–0.84 0.001
Lock 6 17 0.09 1.40 0.58–3.41 0.47
Loose forward 4 11 0.06 0.50 0.19–1.36 0.15
Inside back 9 26 0.14 1.73 0.84–3.58 0.14
Midfield back 3 9 0.05 1.09 0.31–3.84 0.89
Outside back 9 26 0.13 1.28 0.65–2.54 0.48
Cause of injury
Tackleb 21 50 0.32 1.60 1.08–2.36 0.02
Foul 2 5 0.03 0.18 0.05–0.73 0.001
Collisionc 2 5 0.03 0.95 0.89–1.02 0.02
Ruck 11 26 0.16 0.98 0.55–1.74 0.93
Head clashd 4 10 0.06 0.91 0.32–2.62 0.86
Accidente 1 2 0.02 1.06 0.11–9.96 0.96
Scrum 1 2 0.02 0.80 0.09–6.94 0.83
Gum shield
Yes 18 69 0.27 1.69 1.12–2.51 0.02
No 8 31 0.12 0.52 0.28–0.96
Ground condition
Hard 6 16 0.09 0.85 0.37–1.93 0.69
Firm 11 29 0.16 0.70 0.41–1.12 0.17
Soft 15 39 0.23 1.41 0.87–2.30 0.18
Muddy 5 13 0.08 1.09 0.43–2.80 0.86
Icy 1 3 0.02 0.77 0.09–6.68 0.81
Quarter of play
1 5 16 0.08 0.95 0.37–2.48 0.92
2 10 32 0.15 1.20 0.66–2.18 0.55
3 7 23 0.11 0.70 0.34–1.45 0.32
4 9 29 0.13 1.37 0.68–2.80 0.39
Usual playing position
Yes 35 92 0.53 1.03 0.93–1.16 0.58
No 3 8 0.05 0.72 0.22–2.38
IR: incidence rate (per 1000 playing hours); RR: relative risk; CI: confidence interval.
a
Percentages may not total to 100% due to rounding;
b
Tackle defined as a collision where opposing player uses arms to ground player in possession of the ball.
c
Collision, collision where opposing player does not use arms to ground player in possession of the ball.
d
Head clash, contact of heads when a player was in possession/not in possession of the ball.
e
Accident, an unintended collision while a player was in possession/not in possession of the ball. The values given in bold highlight the significant
findings (p < 0.05).

injuries (26% of all injuries). We observed 42 concus- were observed between concussion and ground condi-
sions (
6%) corresponding to an incidence rate of tion (LH ¼ 2.27, p > 0.05), quarter of play (v ¼ 1.34,
0.64 per 1000 playing hours (
1 concussion every p > 0.05) and players playing out of position
1.7 games). (LH ¼ 0.31, p > 0.05).
Injury data are outlined in Table 1. Tackling
(v ¼ 4.84, p < 0.05, RR, 1.60, 95% CI, 1.08–2.36), col-
Discussion
lisions (LH ¼ 5.81, p < 0.05, RR 0.95, 95% CI, 0.89–
1.02) and gum shield use (v ¼ 5.82, p < 0.05, RR 1.69, Our descriptive findings have provided a unique insight
95% CI, 1.12–2.51) were independently associated with into the changing nature of concussion and associated
concussion. In contrast, front row players were at risk factors from rugby union during the 1980s against
lowest risk of injury compared to the backs (v ¼ 7.12, the modern day game. Notwithstanding the limitations
p < 0.05, RR 0.32, 95% CI, 0.12–0.84) and fouling of the current investigation, the incidence rate of con-
posed the lowest risk of concussion (v ¼ 8.78, cussion nearly four decades ago aligned closely to other
p < 0.05, RR 0.18, 95% CI, 0.05–0.73). No associations injury data available from rugby union players during
Owens et al. 5

the 1980s. Furthermore, this retrospective data have changes to knowledge, identification, reporting and
identified risk factors once encountered by past ath- management of concussion within modern day
letes, of which some of those risks are still apparent rugby union.20,21
in the modern era. Comparatively, concussive inci- Principally, the introduction of the consensus defi-
dence was seen to be appreciably lower and some risk nitions and methodologies to standardise the recording
factors were not entirely consistent with what has been of injuries and reporting of studies20 has altered injury
reported in the published literature during the modern reporting within rugby union to great effect. Injuries
era. This is a result of greater clinical management of are defined and data are now collected in accordance to
concussion in modern rugby union, assisted by meth- whether an injury is: recurrent, non-fatal or catastroph-
ods that better recognise and remove an athlete from ic, and classified by severity, location, type, diagnosis
play safely following injury. and cause. All injuries are further recorded in relation
to training and match exposures, providing detailed
medical records for all athletes, thereby allowing qual-
Historical comparisons ified health professionals and coaches to better recog-
In the present study, head injuries were shown to nise concussive injuries and typical severity
account for approximately one quarter of all injuries characteristics. Moreover, as reporting and recognition
corresponding to an incidence rate of 0.64 concussions of concussion has developed among health professio-
per 1000 playing hours. We further calculated that nals and coaching staff, athlete under-reporting of con-
concussive injury accounted for 6% of all injures cussive injury has been identified as a key area of
which replicates the earlier findings of Durkin.12 improvement.25 In turn, concussion awareness and
Additionally, Sparks13 documented that 16.9% of all education programmes are now utilised to varying
injuries recorded were to the head and neck which is degrees from school level onwards, in a bid to enable
appreciably lower than our observations, however the athletes to better recognise and self-report concus-
overall percentage of concussive injuries were similar sive symptoms.26
(5%). Our observations of concussion were substantial- During the 1980s, no such tools or consensus agree-
ly lower than the 22% and 12% documented by ments had been formed, thus highlighting the limita-
Nathan et al.14 and Roux et al.15 in South African tions of this investigation. However (to the best of our
school-level rugby union. However, our results align knowledge), the 15-item questionnaire utilised through-
with data from other southern hemisphere regions out the 1982–1984 rugby union seasons in the current
during that period as Davidson24 observed 24.5% of investigation was the first of its kind within the United
injuries to the head and neck among Australian rugby Kingdom and indeed may be of relevance to better
union players. understand the mechanisms of previous injuries that
While data for factors associated with concussion may apply to a number of retired athletes. For exam-
were primitive, previous literature acknowledged that ple, the 15-item questionnaire (Figure 1) shares eight
tackling was the primary mechanism for injury13,15 and similarities between the Injury Report Form for Rugby
front row players were at lower risk of concussion rel- Union as constructed by the RICG,20 which was intro-
ative to the backs,15 which corresponds with our find- duced some 25 years later and still utilised today.
ings. However, our observations revealed that hard Indeed with the advent of professionalism, training
ground did not increase the risk of concussion, con- methods have changed such that the ‘modern game’
trary to the findings of others.13,15 The discrepancies now sees players who are more skilful, powerful, con-
between these studies are likely due to the variation ditioned and heavier27 with increased speed and force
of data collection tools, study sizes, definition of con- of contact events, duration of time the ball is in play
cussion and subsequent clinical management provided and number of tackles/rucks per game28 that collective-
following injury, including international differences in ly increase concussion risk. In support, tackling was
health care procedures. identified as one of the primary risk factors for concus-
sion and continues to prevail in the modern game espe-
cially, with the number of tackles seen to quadruple
Modern comparisons following the advent of professionalism.28 Front row
Given the inevitable discrepancies in injury definition players were at a lower risk compared to the backs,
across studies, our calculated incidence rate is appre- again consistent with the published literature,8 likely
ciably lower than the ‘modern-day’ equivalent of 2.08 due to limited opportunity to run with the ball and
(range of 1.2–6.9) cited in a recent meta-analysis of fewer tacking incidents.29
players at a similar standard (community, sub-elite However, some of the risk factors identified in the
15s) who are at greatest risk of injury.5 This more 1980s were not entirely consistent with what has been
than tripling in incidence is the likely consequence of reported in the modern game.5 Fouling has previously
6 Journal of Concussion

been associated with an increased risk of concussion30 some of those risks are still apparent in the modern
whereas we observed the contrary. Likewise, we failed era and may be priming athletes for future neurological
to confirm previous reports of an increased incidence of symptoms.
concussion during the third quarter of play (40–60 min)
subsequent to insufficient warm up following the half- Acknowledgements
time break31 and play on hard ground.30 Finally, gum We acknowledge the cheerful cooperation of all players and
shield use that was beginning to be actively encouraged the coaching staff. Raw data collected for this research can be
during the 1980s (personal communication personal accessed by direct contact with the lead author. Thomas
communication JPR Williams) increased concussion Owens, Neurovascular Research Laboratory, University of
risk in contrast to recent findings.30 With consideration South Wales, Alfred Russell Wallace Building, Faculty of
towards the biomechanics and attendant forces during Life Sciences and Education, Pontypridd CF37 4AT, UK.
rugby events, the extent that gum shields could reduce
the incidence of brain injury and concussion Declaration of conflicting interests
remains unclear. The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of
Limitations this article.

There are inevitable limitations to this study given its

Funding
retrospective focus and reliance on questionnaire data.
The understanding of concussion from the period of The author(s) disclosed receipt of the following financial sup-
data collection to the modern day has evolved to the port for the research, authorship, and/or publication of this
extent that the definitions used to diagnosis are differ- article: Damian Bailey is a Royal Society Wolfson Research
ent and may highlight the underreporting of such inju- Fellow (#WM170007). Tom Owens is a PhD student funded
ries throughout the ‘amateur years’. Although medical by the Higher Education Council of Wales. The research was
assessments were carried out by qualified clinicians fol- funded by the JPR Williams Research Fellowships.
lowing injury, data collection forms were populated via
the team coach and may subsequently overlook rele- ORCID iD
vant medical information relating to an athlete’s injury. Thomas S Owens https://orcid.org/0000-0003-0097-0616
Furthermore, we were not in a position to record George Rose https://orcid.org/0000-0002-9598-6372
player demographics including concussion history
thus information on concussion severity, residual References
symptoms from any prior concussions reported and 1. Guskiewicz KM, Marshall SW, Bailes J, et al.
medical clearance to return-to-play were not captured. Association between recurrent concussion and late-life
Finally, we did not assess the long-term functional cognitive impairment in retired professional football
alterations in these players that would have allowed players. Neurosurgery 2005; 57: 719–724.
us to determine to what extent, if indeed any, enduring 2. Blennow K, Brody DL, Kochanek PM, et al. Traumatic
cumulative cognitive, cerebrovascular and motor func- brain injuries. Nat Rev Dis Primers 2016; 2: 16084.
tion impairments were incurred as a result of concus- 3. Mez J, Daneshvar DH, Kiernan PT, et al.
sions sustained decades earlier. Clinicopathological evaluation of chronic traumatic
encephalopathy in players of American football. JAMA
2017; 318: 360–370.
Conclusion 4. International Rugby Board. Global Rugby Particpation,
The present findings highlight the changing nature of https://pulse-static-files.s3.amazonaws.com/worldrugby/
concussion incidence rates in rugby union since the document/2017/03/09/766b1947-4543-4bc5-9240-8d1390
1980s. The incidence rate of concussion during the 46b653/2016-PARTICIPATION-MAP-FINAL.pdf
1980s appears to be appreciably lower compared to (2016, accessed 4 June 2018).
5. Gardner AJ, Iverson GL, Williams WH, et al. A system-
that reported in the modern (present-day) game,5 the
atic review and meta-analysis of concussion in rugby
likely outcome of improvements in the clinical under-
union. Sports Med 2014; 44: 1717–1731.
standing of concussion, data collection tools, reporting 6. Rafferty J, Ranson C, Oatley G, et al. On average, a
methods and clinical management of concussion,20 professional rugby union player is more likely than not
including changes to both player and game.27 From a to sustain a concussion after 25 matches. Br J Sports Med
clinical perspective, this report allows us to better 2018; 52(6): 1–5.
understand the risks and mechanisms of injury once 7. Benson BW, McIntosh AS, Maddocks D, et al. What are
encountered by rugby union players that were active the most effective risk-reduction strategies in sport con-
during the ‘amateur period’ of the sport, of which cussion? Br J Sports Med 2013; 47: 321–326.
Owens et al. 7

8. Brooks JHM, Fuller CW, Kemp SPT, et al. procedures for studies of injuries in rugby union. Br J
Epidemiology of injuries in English professional rugby Sports Med 2007; 41: 328–331.
union: part 1 match injuries. Br J Sports Med 2005; 21. McCrory P, Meeuwisse W, Dvorak J, et al. Consensus
39: 757–766. statement on concussion in sport-the 5(th) international
9. Collins CL, Fletcher EN, Fields SK, et al. Neck strength: conference on concussion in sport held in Berlin, October
a protective factor reducing risk for concussion in high 2016. Br J Sports Med 2017; 51: 838–847.
school sports. J Prim Prev 2014; 35: 309–319. 22. World Rugby. Concussion Guidance, http://playerwel
10. McCrory P. Do mouthguards prevent concussion? Br J fare.worldrugby.org/concussion (2018, accessed 4
Sports Med 2001; 35: 81–82. April 2018).
11. Stewart W, McNamara PH, Lawlor B, et al. Chronic 23. €
Ozdemir T and Eyduran E. Comparison of chi-square
traumatic encephalopathy: a potential late and under rec- and likelihood ratio chi-square tests: power of test.
ognized consequence of rugby union? QJM 2016; J Appl Sci Res 2005; 1: 242–244.
109: 11–15. 24. Davidson R. Factors associated with Rugby injuries in
12. Durkin TE. A survey of injuries in a 1st class Rugby schoolboys. In: A paper delivered at the Australian Sports
Union Football Club from 1972–1976. Br J Sports Med Medicine Federation conference, Adelaide, Australia, 18
1977; 11: 7–11. August 1978.
13. Sparks JP. Half a million hours of rugby football. The 25. Williamson I and Goodman D. Converging evidence for
injuries. Br J Sports Med 1981; 15: 30–32. the under-reporting of concussions in youth ice hockey.
14. Nathan M, Goedeke R and Noakes TD. The incidence Br J Sports Med 2006; 40: 128–132.
and nature of rugby injuries experienced at one school 26. Bagley AF, Daneshvar DH, Schanker BD, et al.
during the 1982 rugby season. S Afr Med J 1983; Effectiveness of the SLICE program for youth concus-
64: 132–137. sion education. Clin J Sport Med 2012; 22: 385–389.
15. Roux CE, Goedeke R, Visser GR, et al. The epidemiol- 27. Sedeaud A, Marc A, Schipman J, et al. How they won
ogy of schoolboy rugby injuries. S Afr Med J 1987; Rugby World Cup through height, mass and collective
71: 307–313. experience. Br J Sports Med 2012; 46: 580–584.
16. Cook A, Hunt W and Mosberg W. Glossary of head 28. Eaves S and Hughes M. Patterns of international rugby
injury, including some definitions of injury to the cervical union teams before and after the introduction of profes-
spine. Clin Neurosurg 1966; 12: 386–394. sional status. Int J Perform Anal Sport 2003; 3: 103–111.
17. Cantu RC. Guidelines for return to contact sports after a 29. Duthie G, Pyne D and Hooper S. Applied physiology
cerebral concussion. Phys Sports Med 1986; 14: 75–83. and game analysis of rugby union. Sports Med 2003;
18. Denny-Brown D and Russell WR. Experimental cerebral 33: 973–991.
concussion. J Physiol 1940; 99: 153. 30. Chalmers DJ, Samaranayaka A, Gulliver P, et al. Risk
19. Roberts AH. Brain damage in boxers: A study of the prev- factors for injury in rugby union football in New
alence of traumatic encephalopathy among ex-professional Zealand: a cohort study. Br J Sports Med 2012;
boxers. London: Pitman Medical & Scientific Publishing 46: 95–102.
Co., Ltd., 1969. 31. Williams S, Trewartha G, Kemp S, et al. Meta-analysis of
20. Fuller CW, Molloy MG, Bagate C, et al. Consensus injuries in senior men’s professional Rugby Union. Sports
statement on injury definitions and data collection Med 2013; 43: 1043–1055.