Methods

The principal methods used in this survey have been reported elsewhere.¹⁰ Players were eligible for the survey if they were a member of a Scottish Rugby Union (SRU) affiliated rugby club in the Border Reivers District of the SRU during the 1997–1998 season. In August 1997, research nurses and physiotherapists visited 22 of the 25 clubs who agreed to participate, during the week before the first match, and invited eligible players to complete a questionnaire. Players who did not attend rugby training were contacted by telephone. A total of 803 (84%) players, including all 30 players who were employed by the Border Reivers District or the SRU, supplied information on their physical activity level (none, light, medium, or heavy) and average weekly hours of manual work undertaken during the preseason period (16 week summer period from 26 April to 16 August 1997). Players also provided details of the number of weeks and sessions they had attended rugby training in preseason preparation, and the average number of weekly hours of other fitness and sporting activities undertaken during this period. Players reported if they had been injured or were carrying an injury at the end of the previous season, if an injury had delayed the start of their preseason rugby training, and if they had been injured during preseason training. Players' estimates of their perceived fitness—on a scale of 0 (not fit at all) to 10 (extremely fit)—in terms of speed, strength, and stamina were obtained.

Observers at participating clubs reported injury episodes occurring to any player who was registered with their club during the 1997–1998 season. A rugby injury was defined as an injury sustained on the field during a competitive match or during training that prevented the player from playing or training from the time of injury or from the end of the match or training session in which the injury was sustained. Players completed a diary of matches played during the first and second halves of the season in the middle and at the end of the season respectively.

STATISTICAL ANALYSIS

Fitness and sporting activities were divided into power (weight training and rock climbing) and aerobic activities (all remaining activities). Contractual playing status was defined as a binary variable relating to whether or not the player held a professional contract or was an amateur. Cox's proportional hazards regression was used to assess the effects of previous injuries and preseason physical activity in relation to time to first injury episode (from 10 August 1997, the first day on which matches were played). For the 380 players who were not injured on or before the last match of the season (9 May 1998), time to first injury episode was recorded as at least 274 days (censored times). The following variables were entered into the model in a stepwise manner at the p = 0.05 significance level: contractual playing status; age (<20, 20–29 or ≥30 years on 1 August 1997); level of manual work; manual hours worked; attendance at preseason training (number of weeks and sessions of attendance, and average number of sessions per week); injury status at the end of the previous season, and before and during preseason training; weekly hours of aerobic and power activities; perceived fitness in terms of speed and stamina. Strength was not considered because the question had not been validated. Kaplan-Meier survival estimates were used to calculate the proportion of players who remained free of injury during the season for the six groups of players in fig 1⇓. The six comparison groups were based on the final Cox regression model, with weeks of attendance at preseason rugby training divided into two categories on the basis of the median attendance of 5.5 weeks (all professional players attended for more than five weeks). Stepwise binary logistic regression was also used to assess the effects of these potential risk factors on subsequent injury status, which was defined as a binary variable relating to whether or not a player had been injured during the season. To examine the higher risk of injury in the early part of the season,^3–,⁶ injury over the first three months was also considered. The odds ratios (ORs) and the hazard for the logistic and Cox regressions show which factors increase the risk of injury. A value of 1 implies no increased risk, and a value greater (less) than 1 implies an increased (decreased) risk of injury.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Proportion of rugby players who remain free of injury as the season progresses by contractual status, attendance at preseason rugby training, and injury status at the end of the previous season. (A) Amateur players who attended rugby training for less than six weeks; (B) amateur players who attended rugby training for more than five weeks; (C) professional players (who all attended rugby training for more than five weeks). The number of players still at risk of subsequent injury is given above the curves.

Results

Discussion

Apart from a presentation of a preliminary study to the Fourth World Congress of Science and Football in 1999,¹¹ this is the first report to examine preseason physical activity, fitness, and injury, and relate them to subsequent injury in a large cohort of adult rugby union players. It shows an increased risk of rugby injury for professional players, those who attend preseason rugby training for a longer period, and for players who were injured or were carrying an injury at the end of the previous season.

As information on the level of preseason physical activity and rugby training was collected retrospectively, the accuracy of this information may have been affected by players' memory recall. The different methods of data collection may also have introduced a bias into the survey. However, the survey response rate was high (84%), and it is unlikely that such bias will unduly affect the results of the survey. Although perceived speed and stamina correlated reasonably well with the fitness test in the pilot survey, it was not possible to validate strength, which is an important component for determining fitness in rugby players. In addition, the fitness questions were only validated on a small group of under 19 players, and it is assumed that their fitness perception is reasonably representative of all rugby players. However, the level of fitness did not appear to be an important factor once other factors had been taken into consideration. We do not know whether this was because fitness was not measured scientifically or whether it is not an important factor. However, other measures of fitness (the level and weekly number of hours of manual work and weekly hours of aerobic and power activities undertaken) did not predict subsequent injury status. This suggests that the way in which fitness was measured in this study is less important than may have been expected.

Some of the univariate relations with subsequent injury could have occurred by chance because of the number of statistical tests presented (unadjusted exact p values given). However, the focus is placed on the multivariate analyses, which are completed in two different ways to assess which factors are consistently important. The factors in the multivariate models are adjusted for all other factors already in the model (univariate models adjusted for contractual playing status). Therefore the effects are independently associated with subsequent injury. For example, perceived fitness predicts subsequent injury status univariately, but is not important once contractual playing status, attendance at preseason training, and injury status at the end of the previous season has been taken into consideration.

There are many possible reasons why players who attend more preseason training weeks may have a higher risk of subsequent injury. They will probably have had more practice in rugby related manoeuvres such as tackling. This may boost a player's confidence to make more injury prone manoeuvres which less skilled players would not attempt. These players may also play more intensively. Players who attend training more frequently were more likely to undertake power activities, and players of a larger build have been found to have a higher risk of injury.⁸ However, it is unclear if this is a true injury risk factor or a consequence of the relations among various other risk factors. Personality and the “need to win” could also be influences, although the former was not found to be a significant factor in injuries sustained in the tackle.¹² This could have been because opposing players in the same game are playing, in general, at the same team level, and their “need to win” could be similar. It is possible that personality and the “need to win” differs between the highest and lowest levels of play. Certainly, the injury rate differs; the present survey noted a 180% relative increase in the risk of injury for professional players. This agrees with observations made during the 1995 Rugby World Cup¹³ and other surveys reporting higher injury rates among those who play at a higher team level.^5,6,14 In addition, the duration of the warm up before matches¹⁵ and training sessions and the use of different items of protective equipment¹⁶ could also influence the risk of injury. It was not possible to measure these factors, because of the large number of matches and training sessions covered in this survey. Injury status before and during preseason training affected attendance at preseason training. It might be expected that players injured during their preseason training would attend significantly fewer training sessions during a shorter time span than players who were not injured, but the reverse was true. Players who had an injury that delayed their preseason training did attend rugby training for significantly fewer weeks than non-injured players, but there was no difference in the number of sessions attended. Some of these players may be training more intensively to make up for weeks when they were unable to train because of injury, increasing the likelihood of future injury.¹⁷ There was no evidence that players who attended preseason training more frequently were more likely to obtain subsequent soft tissue injuries, which may indicate overtraining. The number of hours of manual work, aerobic activities, and power activities undertaken during the summer period, level of manual work, and fitness levels before the start of the season had little effect on subsequent injury. This suggests that injury risk is related far more to rugby training (type of activities undertaken in rugby training or personalities and characteristics of players undertaking training more frequently) than to overall player fitness. This agrees with another study that found that strength and flexibility did not exert a significant role in determining injuries in a study of elite young athletes.¹⁸ Although, the preliminary study conducted in New Zealand found that players with ≥40 weekly hours of strenuous activity missed more playing time through injury than those less active,¹¹ no associations between injury risk and the results of the fitness tests were mentioned. Residual fatigue induced by heavy preseason training rather than overtraining may make players more likely to be subsequently injured. Another study, which examined players for three consecutive seasons, has suggested that fatigue may increase injuries in amateur rugby league players.¹⁹ Peripheral vision of rugby players may be affected by fatigue,²⁰ and more than half of injuries that occur in the tackle occur within the tackled player's peripheral vision (19%) or from behind him (33%).¹² However, confounding factors make it difficult to determine exactly why players who attend more weeks of preseason training have an increased risk of rugby injury.

Players who undertook no preseason preparation for the forthcoming season did not have a higher subsequent injury rate. Nevertheless, adequate preseason preparation is important, as lack of preseason training can increase the risk of subsequent injury.⁷ As mentioned above, increased attendance at preseason training may not be a direct cause of increased injury risk; it is possible that other underlying factors, related to increased attendance at preseason training, increase the risk of injury. Therefore a reduction in the number of weeks that players attend preseason training may not necessarily reduce the risk of subsequent injury.

Players who were injured or were carrying an injury at the end of the previous season had a higher risk of subsequent injury. Risk of subsequent injury was also higher for previously injured players in the preliminary study in New Zealand.¹¹ This higher risk of subsequent injury could have occurred for similar reasons as previously mentioned. In addition, their outlook and behaviour towards rugby injury may be a factor. During the 1993–1994 season, one fifth of senior players completed the match in which they were injured.²¹ Players who try to ignore their injuries or who do not seek treatment for them promptly may be more likely to be subsequently injured. These players will become prone to recurrent injuries particularly if they do not rest their injuries sufficiently before returning to play. A casual attitude to injuries has been noted in New Zealand, where 39% of players had played rugby against medical advice at some time in their careers.²² Players could also have a higher risk of subsequent injury if they do not maintain fitness and psychological well being while they are injured.²³

This study has found that players attending preseason training for more weeks and those injured at the end of the previous season had an increased risk of subsequent injury, but there is no clear explanation of why this is the case. Further research is required that can examine preseason fitness in a more detailed and comprehensive manner and relate it to subsequent injury.