Article Text

Download PDFPDF

Throwing workload and injury risk in elite cricketers
  1. Richard Saw1,
  2. Rebecca J Dennis2,3,
  3. David Bentley1,
  4. Patrick Farhart4
  1. 1Health and Exercise Science, School of Medical Science, University of New South Wales, New South Wales, Australia
  2. 2NSW Injury Risk Management Research Centre, University of New South Wales, New South Wales, Australia
  3. 3University of Ballarat, Victoria, Australia
  4. 4Cricket NSW, Sydney Cricket Ground, New South Wales, Australia
  1. Correspondence to Mr Richard Saw, Health and Exercise Science, School of Medical Science, University of New South Wales, High Street, Kensington, New South Wales 2052, Australia; richard.saw{at}


Objective To investigate the risk between throwing workload and upper limb injury in elite cricketers.

Design Prospective cohort study.

Setting Elite Australian cricket.

Participants 28 adult male cricketers aged 18–32 years.

Assessment of risk factors Daily throwing workload and injury were prospectively monitored over the 2007–2008 cricket season. Risk ratios (RRs) were calculated to describe the association between throwing workload and injury.

Main outcome measurement Upper limb injury associated with throwing.

Results Seven (25%) players sustained an injury during the season. Injured players threw approximately 40 more throws/week (p=0.004) and 12.5 more throws per throwing day (p=0.061) than uninjured players. Players were at a significantly increased risk of injury if they completed more than 75 throws/week (RR=1.73, 95% CI=1.03 to 2.92), and there was a trend towards an increased risk if they completed more than 40 throws per throwing day (RR=1.41, 95% CI=0.88 to 2.26). Injured players also completed more throws and had more throwing days (and consequently less rest days) in the week before injury, as compared with the rest of their season preceding that point.

Conclusion An increased throwing workload is a risk factor for the development of upper limb injury in elite cricketers. Investigation of the kinematics of throwing in elite cricketers would complement this study, and further research is required to develop detailed throwing workload guidelines for cricketers across a range of ages.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Cricket is a popular sport largely played in British Commonwealth countries. There has been relatively limited research into cricket injuries. While some broad injury surveillance has been conducted,1 2 the majority of research to date has focused on the overuse injuries sustained by fast bowlers.3 While the relationship between bowling workload and injury has been explored, no published research has investigated the association between throwing workload and upper limb injury in cricket.

A study of elite Australian cricketers4 found that the average incidence of shoulder injuries over a 10-year period was only 1.0 injury per 10 000 playing hours, while shoulder injury prevalence was only 0.75% (percentage of matches missed). However, the burden of upper limb injury is often not elicited by formal injury surveillance, as the current international consensus definition only includes injuries that prevent the player from being selected or from batting, bowling, or wicketkeeping during a match.5 While many players with upper limb injury can continue to bat and bowl, there may be some impact on functional ability to throw. In a recent English study, 23% of professional cricketers self-reported a shoulder injury during the 2005 season, although only 1.7% were unavailable for a match during the season because of a shoulder injury.6 Sports medicine staff attached to elite cricket teams suggest that upper limb injury is a substantial problem in cricket, and the need for research into overuse injuries to the shoulder has been identified by injury researchers.6,,8

Research regarding throwing injuries has been conducted among other sports, in particular, baseball. A prospective study of junior/adult players found that the majority of baseball injuries occurred in the upper extremity (58%), in particular, the shoulder.9 The general opinion among baseball coaching staff supports the notion that throwing workload is a risk factor for injury. However, there is surprisingly little published data on workload and injury in adult elite baseballers,10,,12 with most research to date focusing on junior baseballers. This research has reported that pitches thrown per game and per season have been associated with increased shoulder and elbow pain.10 11 While throwing workloads have been investigated for baseball pitchers, the differences in the context and the nature of baseball pitching and cricket throwing may limit the transferability of the guidelines. Therefore, we aimed to explore the impact of throwing workload among cricketers, specifically with the hypothesis that high throwing workload will be associated with an increased risk of upper limb injury.



Thirty-three elite male cricketers from the New South Wales (Australia) State Cricket Squad were recruited to participate in this study. The players were prospectively monitored throughout the 2007–2008 season, from the beginning of their preseason training in July 2007 until the season end in April 2008. This project was approved by the Human Research Ethics Advisory Panels of the University of New South Wales (NSW).

Of these 33 players, five were excluded because of incomplete workload data (resulting from ongoing commitments to the Australian National team and from absence because of a non-throwing related injury). The remaining 28 players had a mean age of 24.4 (3.9) years (range, 18–32 years).

Evaluating throwing workload

Throwing workload was monitored by either video recordings or direct observation. Only overarm/overhead throws were considered:

  1. throw-downs (TDs), where a player throws the ball to a batsman or a keeper, in an attempt to simulate the batsman/keeper receiving a ball from a bowler in a match;

  2. fielding drill throws (drill throws): throws that occur as part of fielding drills at a training session;

  3. warm-up throws: throws that occur on the day of a match outside of match play and, when interstate, the day before a match;

  4. match throws: throws that occur during match play.

TDs and drill throws that occurred at NSW State Squad training sessions were directly observed by the researchers. For practical reasons, TDs that occurred outside of training sessions were player recorded.

Player self-recordings were validated by an observer for six training sessions. The players' recordings produced a correlation of 0.99, with a mean error of 1 (2) throws.

Throws that occurred during the 1st and 2nd XI matches were recorded by review of video footage of the matches. Match warm-up throws were recorded by the NSW State Squad physiotherapist. Throws that occurred during Grade club cricket matches and other practice matches were not counted because the participants in this study played for 14 different clubs and there were no video recordings available. However, the dates of these matches were retrieved from scorecards, which facilitated the compilation of a complete data set of the dates of throwing sessions. Participation in Grade clubs' training sessions was not recorded; however, attendance at these sessions by the participants was minimal, as they usually coincided with NSW State Squad sessions.

Injury surveillance

Surveillance of injuries was conducted by the NSW State Squad physiotherapist, who records, diagnoses, and treats all injuries for the entire squad. An injury that was eligible for inclusion in this study was a condition that caused a player to present to the state team physiotherapist, complaining of shoulder or elbow pain that was associated with throwing.

While some injury surveillance authors recommend the use of “match time loss”13 as a standardised injury definition, they acknowledge that broader definitions can be more suitable for analysing specific injury types. Also, it has been reported that 70%–92% of all injuries are “transient” and consequently normally overlooked by “missed match” injury definitions.14 Throwing can be a non-essential component of cricket match play, so a player will not always miss game time from a throwing-related injury. A study of professional British cricketers found that 58% of players with a shoulder injury reported that they fielded in a specific position to avoid throwing.6 These injuries can influence a player's performance and impact on the healing process, thus are relevant for inclusion.14

Statistical analysis

Analysis was conducted using the same approach as that of published fast-bowling workload studies.3 15 16 For players who were uninjured, total throwing workload for the duration of the season was calculated. However, for injured players, only throws completed before their first injury were calculated, as their workload postinjury may be reduced subsequent to the injury itself.

The average weekly workload for each player was calculated in 7-day periods from the date of the first squad training session. Weeks in which the player did not throw were not included when calculating the average weekly workload so that the mean was not influenced by non-scheduling of sessions (eg, during Christmas break). Additionally, when calculating the average number of throws completed per day per week, only the days per week in which actual throw counts were recorded were included.

Independent t tests were used for all parametric data to compare the mean of injured and uninjured groups and were processed using SPSS v14.0. Tests returning a p value of less than 0.05 were considered significant. Paired t tests were used to compare the workload of injured players in the weeks before injury (1, 2, 3 and 4 weeks before) and in the rest of their season preceding that point.

Risk of injury was determined by comparing players with a high workload with players with a low workload. The group with a low throwing workload was chosen as the reference group when calculating risk ratios (RRs) so that high workload could be investigated as a risk factor for injury. For reasons of completeness, the risk factor variables were categorised into these high/low groups in two ways using the values for the 50th percentile (above or below the median) and the values for the 75th percentile (above or below the upper quartile) as the cut-points. For practical interpretation, these cut-points were rounded to the nearest appropriate number.


Descriptive results

The 28 players completed a total of 2102 throwing sessions and more than 42 000 throws during the course of the season. Match sessions (mean throws for the squad, 10.48 (10.39) throws) involved substantially less throws than match warm-up (27.07 (19.24)), drill (42.48 (26.28)) and TD (66.52 (40.00)) sessions. Seven players sustained injuries that were eligible for inclusion in the analysis.

Relationship between workload and injury

Throws per week

Injured players had a higher mean number of throws per week than uninjured players, as seen in table 1.

Table 1

Difference between injured and uninjured players in the mean (SD) number of throws per week and throws per day

When comparing above and below the median, players who threw above 75 throws/week were at an increased risk of injury (RR=1.73, 95% CI=1.03 to 2.92). There was a trend towards an increased risk of injury for workloads above 100 throws/week (upper quartile), but it did not reach significance (RR=2.59, 95% CI=0.83 to 8.13).

Throws per day

Injured players completed more throws per throwing day than uninjured players as seen in table 1, with this difference approaching significance (p=0.061). There was also a trend towards an increased injury risk for players completing more than 40 throws per throwing day (RR=1.41, 95% CI=0.88 to 2.26).

Rest days per throwing day

Subsequent to a player's first throwing day, every day without a throwing session was considered a rest day. There was no difference (p=0.555) in the average number of rest days per throwing day between the injured (mean=2.61 (0.91)) and uninjured (mean=2.88 (1.07)) groups.

Workload in the week before injury

In the week before the injury onset, the injured group had significantly increased their workload, with fewer rest days per throwing day and significant rises in the average number of throwing days per week and throws per week (table 2).

Table 2

Difference in workload in the week before injury onset compared with all other data in preceding weeks

In 2, 3 and 4 weeks before the injury onset, the injured group again had fewer rest days per throwing day. There was no difference in throws per week or throwing days per week.


Previous research has been conducted in baseball to describe the association between throwing workload and injury,9,,12 Yet, this is the first published study to investigate this in cricketers. This study has demonstrated a consistent association between high throwing workload and an increased risk of upper limb injury. Injured players completed more throws per day and per week and had significantly heavier workloads in the week before injury than in preceding weeks, completing more throws with fewer rest days.

It is possible that the development of full-time professional cricketers at the domestic level has resulted in players, particularly batsmen, throwing more TDs at each other to practice. In the current study, TDs throwing session had the largest average workload. This greater volume and the anecdotally perceived strain of TDs have implications towards overuse injuries, and, therefore, workload research such as this is essential to provide evidence-based injury prevention guidelines.

A key finding was that injured players threw 40 extra throws/week and 12.5 extra throws/day than their uninjured teammates. Throwing more than 75 throws/week puts players at 1.73 times the risk of injury. This is quite intriguing considering that baseball studies have only found throwing above 75–80 pitches per match to be a risk factor.10,,12 One study11 found that throwing more than 75 pitches puts junior players at a 50% increased risk of elbow pain and a 1.5% increased risk of shoulder pain for every throw beyond the 75th, while another study found that players throwing more than 80 pitches per match were at a fourfold increased risk of injury.12

Throwing injuries may develop gradually over time before manifesting when stress loads breach threshold levels. This concept is supported by the results of this study; injured players had heavier workload overall compared with non-injured players and had substantially greater loads in the week before injury compared with their already above-average workload.

In the week before injury, the injured group threw 38.91 more throws in 1.77 more days, with 2.19 less rest days per throwing day than their average before that time. This has strong implications about the risks associated with an acute increase in throwing workload and is an area for players to be mindful of in the future. It is interesting to note that these parameters (except rest days per throwing day) were not significantly increased in the 2, 3 or 4 weeks before injury onset, suggesting that a workload spike in only 1 week is sufficient to trigger an injury.

A limitation of this study and previous baseball studies is that it does not measure the intensity of throws, which is significant because the kinematics of maximal and submaximal effort throws have been shown to vary significantly.17 Potential differences between baseballers' and cricketers' throwing average intensity also limit transferability of baseball throwing volume guidelines.

Cricket fielders have to throw a variety of distances, at a wide range of angles and occasionally off balance. In this regard, cricket fielders are best compared with baseball position players. While there is limited research focusing on position players, an American study of college baseballers found that 32% of shoulder injuries occurred in position players, which is comparable with the percentage of throws they undertake in an average baseball match.9

There is also no available data comparing possible differences in the throwing techniques of baseballers and cricketers. Recent research has found that the sidearm and submarine throwing styles are associated with greater peak forces at the elbow and the shoulder, respectively, while the biomechanics of the overhead style generally facilitated a minimum injury risk.17 Also, pitchers usually start from a “windup” position, which involves different kinematics from those of throwing from a “set” position.17

These factors highlight the practical limitations of applying baseball literature to develop recommendations in cricket and the importance of throwing kinematic and workload research that is specific to cricket.

Increased humeral external rotation and decreased internal rotation have been well documented in baseball pitchers18,,22 and position players.21 These changes are thought to be due to adaptations in both soft tissue and bony architecture in response to high throwing workloads during skeletal development.19 20 22 By facilitating greater external rotation, a certain degree of retroversion may allow greater throwing performance while placing less strain on the anterior capsulolabral complex.20 23 Hence, adaptive remodelling during skeletal development in overhead athletes may lead to increased humeral retroversion, facilitating decreased injury risk and improved throwing performance.20

The non-essential nature of throwing in cricket in comparison with bowling makes it possible that junior cricketers may miss out on this adaptive remodelling, although there is no evidence to support this at present. This theory, along with other differences in kinematics already discussed, could explain why cricketers in this study obtained injuries at lower workloads than baseball pitchers. However, it is also important to consider differences in study designs and possible missed throws from our workload recordings.

A recent study of professional baseballers found that pitchers had decreased glenohumeral internal rotation and total range of motion immediately after throwing and to a lesser extent 24 h after throwing.24 The authors hypothesised that acute decreases in glenohumeral range of motion may place throwers at greater risk of shoulder injury if they continue to throw while their range of motion is acutely reduced. This theory could explain some of the results of the current study, whereby a reduced number of rest days in the week before injury placed players at an increased risk of injury, but overall rest days for the season did not. The results of our study coupled with proposed acute and long-term biomechanical adaptations highlight the need for a project on cricketers that integrates musculoskeletal screening measurements, such as humeral torsion, with workload recordings.

Bowling workloads were not considered in this study; however, all seven injuries have been previously identified as being common to overhead-throwing athletes,25,,27 and previous research has acknowledged that the majority of glenohumeral injuries in cricketers occur as a direct result of throwing.28 29 Of the injured players, only one was a bowler and one was an all-rounder, which further downplays the possible influence of bowling loads. Future studies should analyse the throwing and injury patterns of specific player roles (batsmen, bowlers and all-rounders). The small sample size of the current study prevented such analysis.

A limitation of the study is the missing throw counts from Grade matches as explained in the methods section. However, it should be noted that there is no systematic bias in the missing data between injured and uninjured players, and even though throws in these sessions could not be counted, the dates of all of these sessions were recorded in the data set.

It is also possible that players did not self-record all TD sessions or other throws completed outside of match and training sessions. However, it should again be noted that there is no foreseeable systematic bias, and it would be impractical to physically monitor unscheduled training TDs for each squad member. Nonetheless, it is important to acknowledge that throwing volumes may be underestimated in this study; hence, further research is required to establish comprehensive workload guidelines.

In conclusion, this study has shown that an increased throwing workload is associated with an increased injury risk. The implications of these results suggest that players and coaches should be mindful of excessive throwing workloads and should particularly avoid acute increases in workload because these place players at a greater risk of sustaining shoulder pain. Strategies for reducing training throwing workload (particularly TDs) that also accommodate the practice needs of batsmen need to be developed—for example, bowling machines that are more realistic and accepted by batsmen or hiring baseball pitchers to throw TDs to batsmen. There is a need for cricket-specific research into throwing kinematics and musculoskeletal markers such as humeral retroversion, along with workload studies with a larger cohort of cricketers, to allow the development of more detailed throwing workload injury prevention guidelines.

What is already known on this topic

No published research has investigated the association between throwing workload and injury in cricket. Baseball studies have reported that pitches thrown per game and per season have been associated with increased shoulder and elbow pain.

What this study adds

The key finding of this study was the significantly increased injury risk associated with throwing more than 75 throws/week. There was also a trend towards an increased risk for more than 40 throws/day. Injured players also threw more throws with less rest days in the week before their injury, suggesting that acute increases in workload and overall workloads are risk factors for throwing-related upper limb injuries.


The authors would like to thank Cricket NSW, the players who participated, David Gilbert, and the coaching and support staff of the SpeedBlitz Blues including Matthew Mott, Andrew Ware and Bob Barter. The authors also wish to acknowledge Jodie Duggan, Tenzing Tsewang, Anna O’Brien and Thomas Davies for their assistance with data collection.



  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval This study was conducted with the approval of the Human Research Ethics Advisory Panels of the University of NSW.

  • Provenance and peer review Not commissioned; externally peer reviewed.