Epidemiology of time-loss injuries within an Australian male professional football club: A 5-year prospective observational study of 21,343 player hours

ABSTRACT

This study aimed to establish injury incidence rates (IIRs) and burden within an Australian male professional football club (n = 73) and to investigate longitudinal trends across five consecutive seasons (2016/17–2020/21). There was an overall IIR of 9.18 injuries per 1000 hours (h) (95% CI [7.89, 10.47]). The IIR was approximately seven times greater (rate ratio (RR): 6.85; 95% CI [5.13, 9.19]; p < 0.01) in matches (31.29 injuries per 1000 h; 95% CI [25.25, 37.33]) compared to training (4.49 injuries per 1000 h; 95% CI [3.51, 5.47]). The overall injury burden was 254.1 days lost per 1000 h (95% CI [220.9, 292.3]). Compared with the reference 2016/17 season, there were significant increases in minimal (RR: 6.94; 95% CI [1.27, 128.73]) and mild injuries (RR: 3.76; 95% CI [1.21, 16.39]) in season 2017/18 and decreases in moderate (RR: 0.40; 95% CI [0.19, 0.80]) and contact injuries (RR: 0.35; 95% CI [0.12, 0.90]) in season 2019/2020. Time-loss injury is common and represents a major burden in Australian male professional football, with injuries more frequently sustained during matches. Injury prevention practices should specifically be directed towards muscle/tendon and ligament injuries of the lower limb, particularly anterior cruciate ligament, ankle sprain and hamstring strain injuries.

KEYWORDS:

• Sports medicine
• athletic injuries
• injury surveillance
• team sport
• soccer

Introduction

Professional football has high injury rates with the average player expected to sustain two time-loss injuries each season (Ekstrand et al., 2013). Injuries, and the consequent unavailability of players for match-play, negatively impacts football team performances within a given season in terms of goals scored, games won, total points accumulated and final league position rank (Eirale et al., 2013; Hägglund et al., 2013). Hence, avoiding time-loss injuries is of utmost importance (Drew et al., 2017). Strategies to prevent time-loss injuries should be based on epidemiological research that has established the aetiology of injury within a specified athlete group (Finch, 2006).

A recent systematic review reports the injury incidence rate (IIR) in male professional football as 8.1 injuries per 1000 hours (h) of exposure (95% CI [7.2, 9.0]), with match injury incidence (36.0 injuries per 1000 h; 95% CI [31.3, 40.8]) almost 10-times higher than that of training (3.7 injuries per 1000 h; 95% CI [3.1, 4.4]) López-Valenciano et al. (2020). A large discrepancy is also reported in match injury burden (504.6 days lost per 1000 h) compared to that of training (60.5 days lost per 1000 h) Ekstrand et al. (2021). However, injury rates are known to vary between regions and leagues (López-Valenciano et al., 2020), and hence league-specific injury epidemiological research is required to best inform sports medicine staff in their respective leagues. Club-specific research may also be warranted to capture the effect of club-specific characteristics on injury rates i.e., staff instability, training surface changes, involvement in international competitions.

Available evidence in Australian male professional football (called the A-League) is limited to two league-wide multi-season epidemiological studies (Gouttebarge et al., 2016; Lu et al., 2020). Match and training exposure durations were not collected in these studies, limiting reporting of injury incidence to number of injuries per round (Lu et al., 2020), or to a clinical incidence measure (number of player injuries relative to the number of players at risk) (Gouttebarge et al., 2016; Knowles et al., 2006). Injury burden statistics have also previously not been reported (Gouttebarge et al., 2016; Lu et al., 2020). This complicates making direct comparisons with international data collected based on the football consensus statement on injury definitions and data collection procedures (Fuller et al., 2006). IIRs consider injury occurrence relative to time-at-risk through comparison to athlete exposure (injuries per 1000 h), which allows comparison of injury rates between different seasons and competitions. Yet they have previously only been reported for sub-elite Australian football (Whalan et al., 2019), limiting the comparison of these findings within the context of the professional A-League.

At present, there remains a paucity of professional football injury epidemiological research within the Australian setting, that aligns to the international consensus statement, particularly data reported as a function of exposure. Given effective injury prevention strategies typically require accurate, standardised and league-specific injury epidemiological data, it is crucial to firstly highlight the extent of the problem prior to any further analyses or interventions. Therefore, this study aims to: i) quantify IIRs and burden per 1000 h exposure for both matches and training within an A-League football club and ii) to investigate longitudinal trends over five consecutive seasons per international standardised methods.

Methods

We conducted a prospective observational study of male professional football players (n = 73 unique players) from a single A-League football club over five consecutive seasons (2016/17–2020/21). Players contracted to the men’s A-League team were eligible for inclusion in the research, with those transferred to and away from the club mid-season also followed for their respective periods. The number of players within the playing squad varied over the five seasons (21 in 2016/17, 25 in 2017/18, 28 in 2018/19, 29 in 2019/20 and 31 in 2020/21, with 45 players’ data included for one season, 11 for two seasons, 7 for three seasons, 6 for four seasons and 4 for five seasons). Similarly, alterations to staff appointments meant there were one physiotherapist (2018/19 season), one strength and conditioning coach (2019/20 season) and three head coach changes (one in the 2017/18 and two in 2019/20 season) across our five-season observational period. Ethical approval for the research was obtained from the University of Newcastle (protocol number H-2018–0229) and all players provided informed consent to participate in the study.

For the first three seasons of this observational study (2016/17–2018/19), pre-season began in July with the in-season commencing in October and ending in May. However, due to the COVID-19 pandemic the 2019/20 and 2020/21 seasons were atypical. While the 2019/20 season began as normal, there was a three-month league-wide postponement of all football-related activities from March to June, with the season then concluding in August. Schedule changes arising from the delayed 2019/20 season then meant a shortened pre-season for the 2020/21 season, with pre-season beginning in October and in-season in December.

A time-loss definition of injury was adopted, with any physical complaint sustained in a football match or football training leading to a player being unable to fully participate in football training or match play recorded. This definition of injury is routinely used internationally in football epidemiology research (Mosler et al., 2018; Werner et al., 2019). A player was considered injured until they were cleared to be available for match selection by club medical staff. Injuries were categorised by severity based on the number of days lost (minimal: time-loss from training and/or match-play between 1 and 3 days; mild: between 4 and 7 days; moderate: between 8 and 28 days and severe: greater than 28 days) (Ekstrand et al., 2011; Mosler et al., 2018). The study design and reporting adhered to the consensus statement on injury definitions and data collection procedures in studies of football injuries (Fuller et al., 2006).

The surveillance period for the study included both the pre-season and in-season phases, inclusive of all matches and training sessions. All injury assessments were performed by the team physiotherapist and/or sports physician, recorded by the team physiotherapist, and were allocated an Orchard Sports Injury Classification System (OSICS) (Version 13.4) diagnostic code (Orchard et al., 2020) by the team physiotherapist. Recorded data included date of injury, injury diagnosis, side of injury occurrence (left/right, dominant/non-dominant), mechanism of injury (contact/non-contact) and date of medical clearance for return to play. Each injury was subsequently classified by body region, as well as by affected tissue type (Bahr et al., 2020). Players transferred to the club with an existing injury or players returning to pre-season training with an injury sustained elsewhere were not excluded from the study, but their existing injury was not recorded as part of the study (Fuller et al., 2006). The team physiotherapist was responsible for treatment, rehabilitation and return to play decisions, guided by the team sports physician. If a player was injured at the end of the season, an estimated return-to-play date was provided based on all available evidence (Fuller et al., 2006; Hägglund et al., 2005). All match and training exposure data was collected according to the football consensus statement (Fuller et al., 2006), with the team strength and conditioning coach and/or sport scientist recording each player’s exposure via their individual global positioning system (GPS) unit. Data was entered for each player into a password protected database by each respective club personnel.

All data was extracted and de-identified from the database and imported into the statistical analysis software R (Version 4.1.0). (R: A Language and Environment for Statistical Computing, 2021) Descriptive statistics were used to determine injury frequency categorised by body region, tissue type, mechanism, injured side, season phase and football activity, and are presented as absolute and relative values. Overall, match and training IIRs are reported as the number of injuries per 1000 hours $((\sum \mathrm{injuries}/\sum \mathrm{exposure}\mathrm{hours})\mathrm{x}1000)$ with corresponding 95% CI. Injury burden is reported as days lost per 1000 hours (IIR x mean severity) (Fuller, 2007) with corresponding 95% CI

$\mathit{lower}95\mathrm{\%}\mathit{CI}=\frac{\mathit{burden}}{\exp \left(\frac{1.96}{\sqrt{\mathit{injury}\mathit{count}}}\right)},\text{breakupper}95\%\mathit{CI}=\mathit{burden}\ast \exp \left(\frac{1.96}{\sqrt{\mathit{injury}\mathit{count}}}\right)$

Poisson regression was used to analyse seasonal trends relating to injury incidence, football activity, mechanism, and severity, with the number of observed injuries as the response variable, season as an independent variable and exposure as an offset. A dispersion test from the “AER” package (Kleiber & Zeileis, 2008) was used to test the null hypothesis of equidispersion of the response variable (Cameron & Trivedi, 1990). A quasi-Poisson or negative binomial regression model was used instead where the assumption of equidispersion was not met. IIRs and rate ratios (RRs) were reported with 95% CI, with intervals that did not include 1.0 indicating statistical significance.

Results

There were 196 time-loss injuries recorded over the five-season study period, with a total exposure of 21,343 player hours (match: 3,292 h, training: 18051 h). There was an overall IIR of 9.18 injuries per 1000 h (95% CI [7.89, 10.47]), with an approximately seven times greater IIR (RR: 6.85; 95% CI [5.13, 9.19]; p < 0.01) in matches (31.29 injuries per 1000 h; 95% CI [25.25, 37.33]) compared to training (4.49 injuries per 1000 h; 95% CI [3.51, 5.47]). An overall injury burden of 254.1 days lost per 1000 h (95% CI [220.9, 292.3]) was observed, with match injury burden (847.2 days lost per 1000 h; 95% CI [698.4, 1027.7]) approximately six times greater than training injury burden (137.4 days lost per 1000 h; 95% CI [110.5, 170.8]).

Lower limb injuries accounted for 85% of all injuries, with a greater proportion found to be moderate (54%) or severe (24%) in severity than mild (13%) or minimal (8%) (Table 1). Muscle/tendon injuries represented 54% of all injuries, yet the most common severe injuries were ligament/joint capsule injuries (37%). Non-contact injuries occurred significantly more often than contact injuries (RR: 1.81; 95% CI [1.34, 2.45]; p < 0.01) yet were relatively evenly distributed for minimal and severe injuries. Injuries to the dominant side of the body (41%) were more frequently reported compared to the non-dominant side (34%) yet there was no significant difference (RR: 1.18; 95% CI [0.85, 1.64]; p = 0.32). Although almost two-thirds (64%) of injuries occurred in-season, there was no significant difference (RR: 0.98; 95% CI [0.73, 1.31], p = 0.87) between the rate of pre-season (9.23 injuries per 1000 h; 95% CI [7.08, 11.38]) and in-season injuries (9.25 injuries per 1000 h; 95% CI [7.63, 10.87]) as a function of exposure.

Table 1. Injury count (n), percentage distribution (%) and severity by body region, tissue type, mechanism, injured side, season phase and football activity.

	Total	Minimal (1–3 days)	Mild (4–7 days)	Moderate (8–28 days)	Severe (>28 days)
Body Region
Head and Neck	6 (3)	1 (5)	2 (6)	3 (3)	0
Upper Limb	7 (3.5)	2 (10.5)	1 (3)	2 (2)	2 (4.5)
Trunk	15 (7.5)	2 (10.5)	6 (19)	4 (4)	3 (6.5)
Lower Limb	166 (85)	14 (74)	22 (69)	90 (91)	40 (87)
Unspecified	2 (1)	0	1 (3)	0	1 (2)
Total	196	19	32	99	46
Tissue Type
Muscle/Tendon	106 (54)	8 (42)	14 (44)	68 (69)	16 (35)
Nervous	11 (5.5)	0	3 (9.5)	6 (6)	2 (4.5)
Bone	13 (6.5)	4 (21)	0	2 (2)	7 (15)
Cartilage/Synovium/Bursa	20 (10)	2 (10.5)	8 (25)	7 (7)	3 (6.5)
Ligament/Joint Capsule	35 (18)	3 (16)	5 (15.5)	10 (10)	17 (37)
Superficial Tissues/Skin	9 (5)	2 (10.5)	1 (3)	6 (6)	0
Vessels	1 (0.5)	0	1 (3)	0	0
Non-Specific	1 (0.5)	0	0	0	1 (2)
Total	196	19	32	99	46
Mechanism
Contact	67 (34)	9 (47)	10 (31)	28 (28)	20 (43.5)
Non-Contact	121 (62)	8 (42)	20 (62.5)	70 (71)	23 (50)
Total	196	19	32	99	46
Injured Side
Dominant	81 (41)	9 (47.5)	13 (41)	40 (40)	19 (41)
Non-Dominant	67 (34)	6 (31.5)	9 (28)	34 (34)	18 (39)
Total	196	19	32	99	46
Season Phase
Pre-Season	71 (36)	8 (42)	14 (44)	30 (30)	19 (41)
In-Season	125 (64)	11 (58)	18 (56)	69 (70)	27 (59)
Total	196	19	32	99	46
Football Activity
Training	81 (41)	9 (47)	14 (44)	36 (36)	22 (48)
Match	103 (53)	8 (42)	13 (41)	59 (60)	23 (50)
Total	196	19	32	99	46

*Percentages (%) represent total column percentages.

*Where column percentages fail to add to 100%, missing data or non-applicable injury categorisations apply.

*Categories with 0 injuries not shown.

Injuries to the thigh were the most common (2.25 injuries per 1000 h), followed by injuries to the lower leg (1.59 injuries per 1000 h) and hip and groin (1.36 injuries per 1000 h) (Table 2). Knee (70.4 days lost per 1000 h) and ankle (40.8 days lost per 1000 h) injuries had the highest injury burden, with anterior cruciate ligament (ACL) injuries (49.4 days lost per 1000 h) and ankle sprains (30.0 days lost per 1000 h) contributing the majority to each category respectively (Figure 1). Hamstring (21.9 days lost per 1000 h) and calf strains (15.9 days lost per 1000 h) resulted in the greatest injury burden for muscular injuries (Figure 1). Meanwhile, injuries to the knee (65 days lost per injury) and foot (60 days lost per injury) had the highest average days lost per injury (Table 2). There were 12 injuries with estimated return to play dates (four knee injuries, two thigh injuries, two lower leg injuries, two ankle injuries, one hip/groin injury and one foot injury).

Figure 1. Evaluation of injury burden over five years within a male professional football club for the 10 injury types with the largest burden using a risk matrix.

Table 2. Injury count (n), percentage distribution (%), incidence, burden and average days lost by football activity and lower limb body region.

	Total	Incidence^a	Injury Burden^b	Average days lost/injury^c
Football Activity
Overall	196	9.18 [7.89, 10.47]	254.1 [220.9, 292.3]	28 ± 46
Match	103 (53)	31.29 [25.25, 37.33]	847.2 [698.4, 1027.7]	27 ± 46
Training	81 (41)	4.49 [3.51, 5.47]	137.4 [110.5, 170.8]	31 ± 49
Lower Limb
Hip/Groin	29 (15)	1.36 [0.87, 1.85]	25.8 [17.9, 37.1]	19 ± 28
Adductor Tendinopathy	5 (3)	0.23 [0.02, 0.44]	12.2 [5.1, 29.3]	52 ± 57
Quadriceps Strain	16 (8)	0.75 [0.38, 1.12]	9.7 [5.9, 15.8]	13 ± 11
Thigh	48 (24)	2.25 [1.61, 2.89]	36.5 [27.5, 48.4]	16 ± 9
Groin Strain	17 (9)	0.80 [0.42, 1.18]	12.8 [8.0, 20.6]	16 ± 8
Hamstring Strain	28 (14)	1.31 [0.82, 1.80]	21.9 [15.1, 31.7]	17 ± 9
Knee	23 (12)	1.08 [0.64, 1.52]	70.4 [46.8, 105.9]	65 ± 95
ACL^d	4 (2)	0.19 [0.01, 0.37]	49.4 [18.5, 131.6]	264 ± 13
Lower Leg	34 (17)	1.59 [1.05, 2.13]	36.2 [25.9, 50.7]	23 ± 45
Calf Strain	21 (11)	0.98 [0.56, 1.40]	15.9 [10.4, 24.4]	16 ± 10
Achilles Rupture	1 (0.5)	0.05 [0.00, 0.14]	12.7 [1.8, 90.2]	270 ± 0
Ankle	25 (13)	1.17 [0.71, 1.63]	40.8 [27.6, 60.4]	35 ± 40
Ankle Sprain	18 (9)	0.84 [0.45, 1.23]	30.0 [18.9, 47.6]	36 ± 38
Foot	7 (4)	0.33 [0.09, 0.57]	19.6 [9.3, 41.1]	60 ± 43
Foot Fracture	2 (1)	0.09 [0.00, 0.22]	10.4 [2.6, 41.6]	111 ± 33
Foot Sprain	5 (3)	0.23 [0.02, 0.44]	9.2 [3.8, 22.1]	39 ± 26

^aInjury incidence reported as the number of injuries per 1000 hours with corresponding 95% CI.

^bInjury burden reported as days lost per 1000 hours (injury incidence x mean severity) with corresponding 95% CI.

^cReported as mean ± standard deviation.

^dAnterior cruciate ligament.

*Lower limb percentage distributions sum to 84.7% as lower limb injury represents 84.7% of total injury count.

*Specific injuries included (in italics) relate to the 10 injury types with the largest injury burden (Figure 1).

In the analysis of time-trends using regression models, the 2016/17 season was used as the reference season (Figure 2). Of the significant trends between seasons, there was a 6.94 multiplicative increase in the number of minimal injuries (RR: 6.94; 95% CI [1.27, 128.73]; p = 0.068) and a 3.76 multiplicative increase in the number of mild injuries (RR: 3.76; 95% CI [1.21, 16.39]; p < 0.05) in season 2017/18. Meanwhile there was a 60% decrease in the number of moderate injuries (RR: 0.40; 95% CI [0.19, 0.80]; p < 0.05) and a 65% decrease in the number of contact injuries (RR: 0.35; 95% CI [0.12, 0.90]; p < 0.05) in the 2019/20 season (COVID-19 affected season).

Figure 2. Injury incidence trends within a male professional football club by a) season only, b) football activity, c) mechanism and d) severity.

Points represent observed incidence rates with 95% CI.*Represents significant change (95% CI ≠ 1.0) in the injury incidence rate compared with the reference season (2016/17).

Injury incidence reported as the number of injuries per 1000 hours (incidence pooled across the five-year study period). Injury severity is defined as the mean number of days lost following injuries. Each point on the graph represents injury burden. Injury burden is reported as days lost per 1000 hours (injury incidence x mean severity). Shaded regions indicate level of injury burden with darker regions representing an increase in the injury burden. Curved lines represent points with equal burden.

Discussion

Our study is the first to describe IIRs and burden as a function of exposure (per 1000 hours) within Australian male professional football. We used internationally standardised injury definitions and data collection procedures outlined in the football consensus statement (Fuller et al., 2006) to allow for direct comparison with previous research.

The IIRs reported in our study are similar to synthesised findings from a systematic review of 44 male professional football studies (overall IIR: 8.1 injuries per 1000 h; 95% CI [7.2, 9.0], match IRR: 36.0 injuries per 1000 h; 95% CI [31.3, 40.8], training IRR: 3.7 injuries per 1000 h; 95% CI [3.1, 4.4]) (López-Valenciano et al., 2020). However, our observed IIRs were greater than those described in a European football injury study of 49 teams over 18 years (match IIR: 23.8 injuries per 1000 h; 95% CI [23.2, 24.4], training IIR: 3.4 injuries per 1000 h; 95% CI [3.3, 3.5]) (Ekstrand et al., 2021). While IIRs in sub-elite football in Australia were as much as double those recorded in our study (overall IIR: 20 injuries per 1000 h; 95% CI [15.9, 23.3], match IIR: 54 injuries per 1000 h; 95% CI [51.2, 57.8], training IIR: 10 injuries per 1000 h; 95% CI [8.2, 11.8]) (Whalan et al., 2019) underlining that rates vary greatly not only internationally, but also within countries by playing level. This highlights the need for sports medicine staff to seek out injury epidemiological data that is contextually specific to their playing group, which will lead to more targeted and impacting prevention strategies.

We identified that incidence rates varied across the years, yet few were statistically significant. Varying methodologies of the two prior injury epidemiological studies within the A-League prevent a direct comparison with findings from our present study, however corresponding data trends may still be highlighted. From 2012–2018 in the A-League, non-contact injuries (3.3–5.1 injuries per round/season) occurred more frequently than contact injuries (1.8–2.3 injuries per round/season) (Lu et al., 2020); a trend we also observed. While we found muscle/tendon and joint/ligament injuries to be the two most common tissue types represented, with percentages similar to those also recorded previously (50% − 60% and 21% − 34% respectively) (Lu et al., 2020). Similarly, the thigh was the injury location with the highest incidence rate, consistent with previous reports (1.9–2.4 injuries per round/season) (Lu et al., 2020). However, our findings contrast with earlier research from the A-League (2008–2013) suggesting knee (3.8–7.7) and ankle (1.8–3.1) injuries contribute the most time-loss injuries per team of 25 players per season (Gouttebarge et al., 2016). Although insufficient information was available to draw conclusions on thigh injuries alone, with hamstring and groin injuries presented independently instead (Gouttebarge et al., 2016). Despite limited previous epidemiological information available within Australian professional football, our study allows mutual trends to be established across seasons and study designs (league-wide and intra-club), allowing interventions to be implemented to reduce the injury incidence impact seen across the league.

We observed relatively high injury burden rates in our study, with an approximately 1.5-fold increase for match burden and a two-fold increase for training burden compared to rates recently reported in European football (match burden: 504.6 days lost per 1000 h; 95% CI [475.6, 533.6], training burden: 60.5 days lost per 1000 h; 95% CI [56.5, 64.5]) (Ekstrand et al., 2021). Burden rates for lower limb injuries in our study (Table 2) were also higher than those previously described, namely for hamstring (13.2 days lost per 1000 h; 95% CI [13.0, 13.4]), quadriceps (7.0 days lost per 1000 h; 95% CI [6.8, 7.1]), adductors (8.0 days lost per 1000 h; 95% CI [7.8, 8.2]) and calf muscles (4.6 days lost per 1000 h; 95% CI [4.5, 4.7]) (Ekstrand et al., 2011). However, up to 20 years’ difference in the respective reporting periods must be noted. The intensity of football matches and the number of high-intensity activities such as running and sprinting are both known to have increased over time (Barnes et al., 2014), which may partially explain the increase in injury burden for specific muscle groups seen in the present study. Alternatively, higher level European football clubs would be expected to have greater medical resources than those in the present study, potentially explaining the differences in injury burden rates. An over-conservative approach by the club’s medical staff to return to play decision making may also be an explanation, given only 19 of the 196 injuries were classified as minimal (1–3 days lost from training and/or match play).

A discernible increase in the overall injury incidence, match injury incidence, and moderate severity injury incidence in the 2020/21 season compared with the two previous seasons is noted (Figure 2). This may be partially explained by a shortened pre-season brought on by schedule changes associated with the COVID-19 pandemic. Only 65 days of pre-season training were available in the 2020/21 season compared with a mean of 109 days from the preceding four seasons. A link has previously been established in football between the number of sessions completed pre-season and injury risk in the forthcoming in-season (Ekstrand et al., 2020). An increase of 10 pre-season sessions is linked with a decrease in injury burden (−21.8 days lost per 1000 h) and severe injury incidence (−0.20 severe injuries per 1000 h) as well as an increase in training attendance (+1.4% points training attendance) and match availability (+1.0% points match availability) (Ekstrand et al., 2020). While this may partially explain the trend seen in this study, we refrain from suggesting causality as injuries occur due to a complex and multivariate system with multiple risk factors and interactions. The increase in overall injury incidence in the 2020/21 season, however, may be alternatively explained by the 2019/20 season having abnormally low injury incidence rates. Significant decreases in contact injury and moderate severity injury in the 2019/20 season compared to the reference season would appear to explain the lower injury incidence rates observed.

When considering the results of this study, several limitations must be acknowledged. In the present study, we utilised a “time-loss” injury definition, whereby injuries were only recorded if resulting in a missed training session or match. While this is an internationally accepted form of injury definition (Fuller et al., 2006), it has various limitations for injury incidence and burden statistics. Firstly, an underestimation of injury incidence is possible due to the time-loss definition not encompassing injuries which do not result in missed football participation, commonly seen in groin injuries (Harøy et al., 2017). The distinction between end of time loss and return to play (injury burden) is also not often objectively clear, with players able to prematurely return to play by adapting technique before injury has fully resolved (Bahr et al., 2020). Conversely, mental or confidence issues when returning from serious injury may extend a players’ time-loss despite the injury having biologically recovered (Bahr et al., 2020). Despite this, the time-loss definition captures injuries with a clear impact for clubs since players are unavailable to train and play matches (Werner et al., 2019).

Lower limb injuries were classified according to the International Olympic Committee (IOC) consensus statements’ “recommended categories of body regions for injuries” (Bahr et al., 2020). However, following the “Doha agreement meeting on terminology and definitions in groin pain in athletes”, groin pain is now subcategorised into three major categories including four clinical entities for groin pain in athletes (Weir et al., 2015), which the IOC categories fail to encompass. This limits comparisons with other large-scale hip/groin studies in which the Doha agreement classification system is now commonplace and considered the gold standard in reporting (Mosler et al., 2018; Werner et al., 2019).

Due to the nature of professional sport and the relatively high turnover of staff, the physiotherapist and strength and conditioning coach at the club at the commencement of the study departed and were replaced through the study (physiotherapist before the 2018/19 season and strength and conditioning coach mid-way through the 2019/20 season). The team treating sports physician however was the same individual throughout the entirety of the study period. Despite changes in staff appointments the agreed methodologies were followed throughout the study period, thus limiting variability within the data collection, injury assessment, recording and classification processes. While there appears to be no increase in IIRs coinciding with the change in physiotherapist, there is a discernible increase in overall, match and moderate IIRs corresponding with the change in strength and conditioning coach (Figure 2). However, this change directly overlapped with the COVID-19 pandemic and as such we refrain from attributing causality to any one variable. The fluctuation of training loads may also have contributed to these findings, with various contextual factors influencing training loads each season, such as varying levels of progression in national and continental cup competitions (Football Australia Cup and Asian Champions League).

As previously mentioned, there were also multiple head coaches within our study period. Coaching leadership style and communication quality between the medical team and the head coach have been shown to have effects on injury burden (Ekstrand et al., 2019) and incidence of severe injuries within professional football (Ekstrand et al., 2018, 2019). While attributing causative factors was beyond the scope of our study, a high number of coaches may have influenced the incidence rates and burden seen in our study. Future research may benefit from the collection of club-specific data to assess the effect of these specific coach and staff changes on IIRs, with larger league-wide data potentially not sensitive enough to detect the effect of these changes.

Finally, a three-month league-wide postponement of all football-related activities was enforced during the COVID-19 pandemic during the 2019/20 season. Spikes have recently been identified in training injury burden upon resumption of team training following COVID-19 associated lockdowns (Waldén et al., 2022), with links to short-term detraining effects due to quarantine/periods of inactivity (Wezenbeek et al., 2022). Despite this link, there appears to have been no noticeable increase in training IIRs in our study (Figure 2).

Conclusion

In our present study we established previously unknown injury incidence rates and burden as a function of exposure within an Australian male professional football club. Time-loss injury was common within this professional football club and contributed to a large burden, particularly sustained during matches compared with training. Most injuries occurred to the lower limb with over half of these injuries affecting muscles and tendons, primarily by non-contact mechanisms. Anterior cruciate ligament, ankle sprain and hamstring strain injuries were the most burdensome observed in the present study, with injury prevention practices recommended to be specifically directed towards these injuries. Injury trends and patterns observed in this study may not be representative of the entire Australian male professional football league.

Acknowledgments

The authors would like to thank all the players of the team considered in the study for their help and cooperation.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This research was supported by an Australian Government Research Training Program (RTP Scholarship).