The 'Northern bias' for injuries in the Australian Football League

The 'Northern bias' for injuries in the Australian Football League
John Orchard, MBBS BA PhD FACSP FACSM FASMF
Visiting Fellow, Sports Medicine Department, University of New South Wales
Australian Turfgrass Management Volume 2.3 (June - July 2000)

Abstract

The Australian Football League surveyed injuries between 1992 and 1998. There were 4681 injuries, that satisfied the definition of missing a regular season game, over 97060 player weeks. Players from teams in Northern states were slightly (14%) more likely to be injured than players from teams in Victoria (RR 1.14, 95% CI 1.07-1.21). There was no significant difference in the risk for any of the categories of upper limb, trunk or head and neck injuries. Many of the lower limb injury categories had greater incidence in players from Northern teams, including ankle injuries (RR 1.71, 95% CI 1.36-2.15), anterior cruciate ligament injuries (RR 1.71, 95% CI 1.13-2.58), groin and hip injuries (RR 1.48, 95% CI 1.23-1.79), calf strains (RR 1.35, 95% CI 1.02-1.77) and quadriceps strains (RR 1.32, 95% CI 1.03-1.71). The majority of these injuries had a non-contact mechanism. Lower limb stress fractures did not show a significant difference between groups, although these injuries were infrequent. This was a correlation study only and further research is required to explain the mechanism of these increased risks. Ground hardness, shoe-surface traction, grass types, shoe selection, training methods and frequent travel are all factors which may have contributed to these findings.

Introduction

The relationship between injuries and ground conditions has become very topical in Australia in recent years due to statistics which have been published as part of the Australia Football League (AFL) injury survey. The Australian Football League Medical Officers Association have followed the paradigm of Van Mechelen [13] which states that injury surveillance is a critical first step towards injury prevention. Injury surveillance has been continuous and consistent since the 1992 AFL season. In addition, the AFL administration has undertaken an injury audit of sorts since 1988, as payments to players who miss games through injury are monitored. The standard AFL player contract allows all players who were playing senior AFL football and then miss games through injury or illness to be granted a payment to compensate them, as if they were playing.

The AFL competition was formed from the Victorian Football League (VFL). Prior to 1982, all VFL teams were based in Victoria, which is Australia's southernmost mainland state. Teams have relocated or been added to the competition so that there are now six teams based outside Victoria ('Northern teams'), with 10 teams based in Victoria ('Victorian teams').

In the early years of the AFL injury survey, it was noticed that the Northern teams were reporting more injuries than the Victorian teams. The definition of an injury in these early years included all injuries that caused players to miss games, but also injuries that prevented players from training fully. The survey coordinators were unsure whether there may have been a bias for Northern teams to be more likely to report injuries because they spent more time with their players, as the teams travelled by plane together for many away matches in Victoria. However, the same discrepancy between Northern and Victorian teams also existed when viewing the AFL injury payments to players. Table 1 shows that every year the Northern teams have paid more injury game payments, on average, than the Victorian teams. There is no reporting bias in these figures, as it is compulsory to pay an injury payment to all players who miss games through injury, and the AFL has a record of which players were missing games through injury. After considering that the AFL's own injury records were free from an important bias that may have existed in the injury surveillance system, the official injury survey definition was changed.

Table - Comparison of senior injury payments (yearly average) between Victorian and Northern teams

	1988	1989	1990	1991	1992	1993	1994	1995	1996	1997
Victorian teams Northern teams	82 106	92 98	88 113	87 91	72 88	58 73	72 79	77 95	88 94	86 110

Methods

The methods for the AFL injury survey are stated in the present tense as they are ongoing. The survey definition of an injury is "an injury or medical condition that causes a player to miss a regular season match". Injuries exist on the database that were reported by teams but did not cause players to miss games and these injuries have not been included in reports since 1996. The change in definition (made in 1996) was retrospective (as detail for injuries from 1992-1995 about missed playing time was available). This study only considers injuries that satisfy the survey definition.

The method of reporting is that at the end of each season, a team medical official sends injury information (in a database or spreadsheet format) to the AFL injury survey coordinator. Games which players have missed through injury are checked with the AFL administration records. Teams are prompted to submit a record for any player who missed a game through injury if there was no injury record originally submitted. Conversely, injuries which were submitted that did not cause missed games are marked on the database as 'unofficial' and not included in survey reports. The major strength of this injury definition is that it is objective, can be verified independently and is consistent between teams and from year to year.

Injury incidence (the rate of new injuries over a given time period) considers only injuries that occur during the regular season. Injury incidence is calculated with number of injuries as the numerator and number of player weeks in the denominator. The unit for injury incidence reported in this paper will be number of injuries per team per season. This unit (injuries/team/season) is used as it is a more meaningful concept than injuries per 1000 player weeks. A regular season is played over 22 weeks (one match per week) and a team is considered to have 40 players on its roster. The unit of injuries/team/season can be converted to a unit of injuries/1000 player weeks by multiplying by 1.136, which is 1000/(40 x 22). Injuries/1000 player weeks is equivalent to injuries/1000 player matches as there is one match per week, although the former is more accurate as training injuries are included.

Once a player returns to play from an injury, the injury is considered to have recovered and that injury record is closed. If the player re-injures the same structure subsequent to returning to play in a match, it is considered a further incidence of injury (a recurrence). If a player re-injures the same body part at training before returning to play in a match, it is considered a continuation of the original injury and it does not contribute further to injury incidence.

Injury prevalence considers injuries that cause regular season matches to be missed, irrespective of when the injury occurred (previous season, off season, pre-season or regular season). Injury prevalence is reported in the annual AFL injury reports but is not relevant to this particular study.

The objective of this study was to compare the relative incidence of injuries between Northern and Victorian teams. Total injury incidence was compared, as were specific injury categories which were formed from pooling similar diagnoses. The injury diagnoses were grouped to allow large numbers in each category. If smaller categories were used, the power of each comparison would have been poor, and confidence intervals would have been large. Lower limb injuries were the most common, so the category diagnoses for these were more specific (e.g. hamstring strains) than injury categories from the other parts of the body (e.g. shoulder joint injuries).

Relative Risk (RR) was calculated by dividing Northern injury incidence by Victorian injury incidence. The 95% confidence intervals for relative risk were calculated using a Taylor series expansion (formulae shown in Figure 1).

Figure 1 – calculation of relative risk and confidence intervals

Results

Between 1992 and 1998 there were 4681 injuries that satisfied the survey definition. During this period there were 1361 AFL players included as part of the survey, who played 4495 player seasons, and a total of 97060 player weeks. The injury incidence was 4681x1000/97060 or 48.2 injuries per 1000 player weeks. This is equivalent to 42.5 injuries per team per season.

Table 2 shows the relative risk (with 95% confidence intervals) for injury incidence of Northern teams compared to Victorian teams. The unit shown for injury incidence is injuries per team per season. Hamstring strains were the most common injury in both groups, followed by groin and hip injuries. The third most common injury category for Victorian teams was back and buttock injuries, whereas for Northern teams it was ankle joint injuries.

Many lower limb injuries showed a significantly increased risk in players from Northern teams, including ankle injuries, knee ACL and cartilage injuries, quadriceps, calf and groin strains. The majority of these injuries had a mechanism not involving player-to-player contact. There was no significant difference between any of the categories of upper limb, trunk or head and neck injuries. No injuries showed a significantly increased risk in Victorian team players, although hamstring strains and back and buttock injuries showed a non-significant trend to being more common in Victorian team players.

Table 2 - Comparison of injury incidence in Victorian and non-Victorian teams (1992-1998) (injuries/team/season)

Victorian incidence

Northern incidence

Relative risk (RR)

95% RR Confidence

  Injury category

N:V

Low

High

Concussion
Other head and neck injuries
Shoulder joint injuries
Acromioclavicular joint injuries
Upper limb fractures
Other upper limb injuries
Trunk injuries
Back and buttock injuries
Groin and hip injuries
Hamstring strain injuries
Quadriceps strain injuries
Anterior cruciate ligament injuries
Knee medial ligament injuries
Posterior cruciate ligament injuries
Knee cartilage/degenerative injuries
Other knee injuries
Leg fractures
Leg stress fractures
Calf muscle strain injuries
Lower limb haematomas ('corks')
Foot fractures (incl. stress fractures)
Ankle joint injuries
Other lower limb injuries
Medical illnesses
TOTAL of all diagnosed injuries

0.9
0.9
1.0
0.9
1.8
0.6
1.2
2.5
3.5
8.0
2.0
0.7
1.2
0.6
1.3
1.2
0.4
0.3
1.8
1.8
0.7
2.2
1.8
1.8
39.1

0.9
1.1
1.1
1.1
1.6
0.6
1.2
2.0
5.2
7.2
2.7
1.1
1.1
0.4
1.9
2.0
0.6
0.4
2.4
2.2
0.5
3.8
1.5
1.8
44.4

1.00
1.29
1.10
1.21
0.88
1.04
0.97
0.81
1.48  *
0.90
1.32  *
1.71 *
0.87
0.72
1.45  *
1.66  *
1.36
1.11
1.35  *
1.19
0.76
1.71  *
0.85
1.03
1.14  *

0.65
0.87
0.75
0.81
0.64
0.62
0.68
0.62
1.23
0.78
1.03
1.13
0.60
0.40
1.06
1.21
0.80
0.54
1.02
0.90
0.45
1.36
0.62
0.76
1.07

1.53
1.93
1.63
1.81
1.20
1.73
1.39
1.07
1.79
1.04
1.71
2.58
1.27
1.30
1.98
2.26
2.29
2.31
1.77
1.58
1.29
2.15
1.17
1.39
1.21

An asterix (‘*’) indicates significantly higher risk of injury for Northern teams (p<0.05 or 95% confidence).

Discussion

These results show an increased incidence in the Northern-based teams of the AFL for lower limb injuries that do not involve player-to-player contact. There are many factors that could be responsible for explaining this increase. The weather is warmer in the Northern states during winter, so grounds are generally harder. Players tend to choose moulded-sole boots when playing in Northern venues, compared to screw-in soles when playing in Victoria in winter. Warm-season grasses such as couch (bermuda) grass (Cynodon dactylon) predominate in Northern venues, whereas Victoria is in a traditional weather zone, and rye grass (Lolium perenne) is the major species. Because of the warmer weather, it is possible that Northern teams train harder and longer in winter because of the more benign conditions. It is also possible that injury management and prevention is better in Victoria than elsewhere in Australia, although it would be unusual if this were the case for lower limb injuries and not for injuries in other parts of the body.

Injuries are traditionally divided into contact and non-contact mechanisms, in which case contact refers to player-to-player contact. Some of the forces involved in a 'non-contact' injury are transmitted from the playing surface to the injured body part, which gives a basis in theory as to why ground hardness may be a risk factor for these injuries.

A study was commenced in 1997 to measure ground conditions at AFL venues, to see whether ground hardness could account for the differences in injury rates between the Northern and Victorian teams. This study received significant publicity in 1998 and there was some pressure on ground managers to intervene to prepare softer grounds in 1999. This in fact occurred, and recent results from the recently-released AFL injury survey [8] showed a reduction in injuries in 1999 compared to 1998 (Table 3). The pattern of injuries which showed a reduction in 1999 was very similar to those injuries that are generally more common in Northern teams. This is an interesting finding, but does not constitute strong scientific evidence that ground hardness is a risk factor for these injuries. Direct measurement of game day conditions constitutes far stronger evidence - to date this has been reported with respect to ACL injuries and weather conditions [9], but not ground hardness readings. Table 4 lists the current state of evidence regarding injuries and ground conditions in the AFL. It is worth noting that grounds may be expected to be harder this year, as the AFL season has been brought forward one month to avoid a clash with the Olympics Games in Sydney.

A recent article was published comparing the Penetrometer and Clegg hammers for measuring ground hardness [4]. A Clegg hammer uses a smaller effective weight than the Penetrometer and hence is influenced more by the grass and thatch, whereas the Penetrometer reading is influenced more by the soil. Further comparison of these devices is advised to determine which, if either of them, is best able to assess injury risk.

Table 3 - Comparison of AFL injury incidence from 1998-1999 (injuries/team/season)

Category

Incidence fell by >0.2 in 1999

TOTAL INJURIES PER TEAM
Concussion
Other head & neck injuries
Shoulder joint injuries
Acromioclavicular joint injuries
Upper limb fractures
Other upper limb injuries
Trunk injuries
Back & buttock injuries
Groin & hip injuries
Hamstring strain injuries
Quadriceps strain injuries
Knee medial ligament injuries
Anterior cruciate ligament injuries
Posterior cruciate ligament injuries
Knee cartilage/degenerative injuries
Other knee injuries
Calf muscle strain injuries
Leg fractures
Leg stress fractures
Ankle joint injuries
Lower limb haematomas ('corks')
Foot fractures (incl. stress fractures)
Other lower limb injuries
Medical illnesses

40.9
0.7
0.9
0.8
0.7
2.2
0.6
0.7
2.0
4.0
7.1
2.8
1.2
0.5
0.3
1.4
1.3
2.4
0.5
0.4
3.4
2.0
0.5
1.5
3.0

38.0
0.5
1.0
0.7
0.6
2.0
0.5
1.6
1.9
3.3
7.9
2.5
1.2
0.4
0.8
1.1
1.4
1.7
0.5
0.1
2.3
1.9
1.1
1.5
1.4

x

x

x

x

x

x
x

x

Table 4 - Current state of evidence regarding ground conditions and AFL injuries

Hypothesis	Evidence	Action
Lower limb injuries are more common in Northern States in the AFL	This is almost certainly scientific fact (evidence presented in this article)	AFL is studying ground conditions to see whether these can explain some of the observed differences
Ground conditions are responsible for the differences in injury risk in various states	This is likely but not certain – other factors may also be responsible	AFL grounds have had Penetrometer readings taken before matches since 1998.
The Penetrometer is the best device for measuring ground hardness	The relevance to injury risk of both the Penetrometer and the Clegg hammer is unknown. Further study is required to assess both these devices.	The Penetrometer was chosen as it is the ‘gold standard’ in the Australian racing industry. There is no ‘gold standard’ for football grounds as we are unaware of any competition in the world that routinely measures all grounds.
Couch grass is a risk factor for ACL injury due to higher shoe-surface traction	This is a popular theory but can’t be tested until there is a reliable device for measuring shoe-surface traction.	The AFL will assess any proposed devices for measuring shear-stress and, if a reliable device is found, will add this to its routine in the future.

Related studies have been performed in other sports. The best comparisons between natural grass and artificial turf (which is generally a harder surface) in the NFL have arisen from the official injury surveillance studies (summarised in Table 5). Bramwell et al. found a higher rate of injury on artificial turf than natural grass over an American football season, with the difference occurring later in the season, when the natural grass surfaces (in Seattle) became softer due to climatic conditions [3]. Andresen et al. found that injuries on muddy or wet surfaces (NG) were less frequent than on good or hard surfaces in Wisconsin [1]. Scranton et al. found that the surface was described as good or dry for the vast majority of non-contact ACL injuries in the NFL, although no direct measurement of ground hardness was made [12].

One study of injuries in Australian horse racing found a much greater incidence of equine injury on fast compared to heavy tracks [2]. Penetrometer readings have been taken for many years in this sport, and have been preferred to Clegg hammer readings [7].

Table 5 - Current state of evidence regarding ground conditions and injury in other sports

Hypothesis	Evidence	Action
Artificial turf is a risk factor for lower limb injury in the NFL	Astroturf moderately increases the risk of ankle injuries [10] , and has a minimal effect on knee injuries [11] .	The NFL continues to monitor injuries each year in a sophisticated injury surveillance system.
NFL players don’t like playing on artificial turf as they perceive it to be too hard	Player surveys consistently show that artificial turf is very unpopular [14] .	Many NFL stadiums are reverting from artificial turf to natural grass [5] .
Long stops are a risk factor for ACL injuries compared to moulded soles	There is one study that suggests that long stops are a risk factor for ACL injuries [6] .	Players are more concerned with performance (not slipping over) than a possible reduction of injury risk. They will continue to use long stops in slippery conditions.
Horses are more likely to be injured when the tracks are hard	There is one Australian study that strongly supports this hypothesis [2] .	It is rare to see racecourses prepare ‘fast’ tracks in recent years. A good to dead track is considered more ideal.

Conclusion

Lower limb non-contact injuries are more common in Northern AFL teams than Victorian teams. This is an important finding and has prompted further research into ground conditions at AFL venues. Ground conditions such as hardness and grass cover may subsequently be shown to be associated with increased injury risk, but other confounding factors may also contribute to the different injury rates.

References

1. Andresen BL, Hoffman MD, Barton LW. High school football injuries: field conditions and other factors. Wis Med J 1989; 88(10): 28-31.

2. Bailey C, Reid S, Hodgson D, Bourke J, Rose R. Flat, hurdle and steeple racing: risk factors for musculoskeletal injury. Equine Veterinary Journal 1998; 30(6): 498-503.

3. Bramwell S, Requa R, Garrick J. High school football injuries: a pilot comparison of playing surfaces. Med Sci Sports 1972; 4(3): 166-169.

4. Ford P. Ground Hardness Measurement. Golf & Sports Turf Australia 1999 (June): 42-48.

5. Garber G. Turf problems keep surfacing. 1999; www.espn.com (Thursday, October 21 1999)

6. Lambson R, Barnhill B, Higgins R. Football Cleat Design and Its Effect on Anterior Cruciate Ligament Injuries: A three year prospective study. American Journal of Sports Medicine 1996; 24(2): 155-159.

7. Neylan J ,Stubbs A. Assessing Racetracks Coditions: A review of available devices. 1998; Rural Industries Research & Development Corporation: Canberra.

8. Orchard J ,Seward H. Injuries on the decline: AFL 1999 Injury Report. AFL Record 2000; 89(5): 29-32.

9. Orchard J, Seward H, McGivern J, Hood S. Rainfall, evaporation and the risk of non-contact anterior cruciate ligament injury in the Australian Football League. Med J Aust 1999; 170(7): 304-6.

10. Powell J ,Schootman M. A Mutivariate Risk Analysis of Natural Grass and Astroturf Playing Surfaces in the National Football League 1980-1989. International Turfgrass Society Research Journal 1993; 7(23): 201-210.

11. Powell JW ,Schootman M. A multivariate risk analysis of selected playing surfaces in the National Football League: 1980 to 1989. An epidemiologic study of knee injuries. Am J Sports Med 1992; 20(6): 686-94.

12. Scranton P, Whitesel J, Powell J, Dormer S, Heidt R, Losse G, Cawley P. A review of selected noncontact anterior cruciate ligament injuries in the National Football League. Foot & Ankle International 1997; 18(12): 772-776.

13. van Mechelen W, Hlobil H, Kemper H. Incidence, Severity, Aetiology and Prevention of Sports Injuries: A Review of Concepts. Sports Medicine 1992; 14(2): 82-99.

14. Wharton D. Turf Wars; The Debate on Merits of Artificial and Natural Grass Resumes as More and More NFL Stars Go Down With Injuries. 1999: LA Times (Los Angeles): Friday, September 24, 1999, page D1.