Injury trend analysis in the Japan national swim team from 2002 to 2016: effect of the lumbar injury prevention project

Participants

A total of 488 Japan national swimmers who participated in the Olympics (2004, Athens; 2008, Beijing; 2012, London; 2016, Rio de Janeiro), Asian Games (2002, Busan; 2006, Doha; 2010, Guangzhou; 2014, Incheon) or Universiade (2003, Daegu; 2005, Izmir; 2007, Bangkok; 2009, Beograd; 2011, Shenzhen; 2013, Kazan; 2015, Gwangju) from 2002 to 2016 were included in this study.

Injury surveillance

We collected data for the swimmers for the period 2002–2016 from the medical clinic database in the JISS. We compiled the injuries diagnosed by orthopaedic surgeons at the precompetition medical check-up before the international competitions.

Injuries were categorised by the body parts according to the IOC for injury surveillance20 as follows: face, head, neck/cervical spine, thoracic spine/upper back, sternum/ribs, lumbar spine/lower back, abdomen, pelvis/sacrum/buttock, shoulder/clavicle, upper arm, elbow, forearm, wrist, hand, finger, thumb, hip, groin, thigh, knee, lower leg, Achilles tendon, ankle and foot/toe injuries. In this study, injuries were defined as overuse injuries as follows: pain that occurred spontaneously without any obvious traumatic events, which led to difficulties in sports activities, and resulted in restriction of the swimmer’s participation for at least 1 day beyond the day of injury. The orthopaedic surgeons classified the severity in three stages for each diagnosis (Active: treatment and immediate workup; Follow: need re-examination and follow-up; Inactive: history and resolved problems). In this study, we analysed chronic disorders of Active and Follow. We classified swimmers into two groups (injured group: swimmers who experienced more than one injury; non-injured group: those who did not have injuries).

The items extracted were as follows: total number of injuries by body parts and injury incidence rates from 2002 to 2016; the incidence of injury in the top four body parts calculated for every year; and injury incidence rate by body part for each type of swimming stroke.

The lumbar injury prevention project

The Japan Swimming Federation introduced the lumbar injury prevention project to the national team in cooperation with JISS in 2008. The project consisted of two components: deep trunk muscle activation training and evaluation of lumbar disc degeneration using MRI.

Intervention and educational components

For the national team, representative physical therapists provided guidance and conducted workshops on prevention training at the national training camp and at competitions (figure 1). Swimmers were encouraged to perform prevention training during the national training camp and at competitions and in daily exercises in each team. During the training camp and expedition, the physical therapists checked the daily exercises. The roles of the physical therapists were to (1) provide specific instruction to swimmers, and (2) provide verbal feedback to actively encourage players to perform each exercise with good form (eg, trunk positions, hip and lib position). The Japan Swimming Federation published a DVD introducing the preventive training and made it available for purchase and approximately 1000 of them were distributed in 2009. The primary prevention training components are shown in figure 1. The basis of the exercises is trunk stabilisation which consists of draw-in, front bridge, side bridge and hip lift exercises, and may play an important role in lumbar injury prevention and rehabilitation.21–24 These exercises induced contraction of the transverse abdominal muscles and the lumbar multifidus muscle, both of which are important for the prevention of lower back injury.25

Request permission

Details of the exercise are as follows: Front bridge: (A) keep the front bridge for 60 s, (B) front bridge while raising the upper limbs 10×2 repetitions, (C) front bridge with hip extensions on an unstable surface for 10×2 repetitions. Hip lift: (D) normal hip lift 10×2 repetitions, (E) hip lift that raises one leg with a ball in the thigh 10×2 repetitions, (F) move the lifting leg up and down with one single leg hip lift 10×2 repetitions. Side bridge: (G) draw-in sideways 10×2 repetitions, (H) keep the side bridge posture for 3 s and raise and lower the body side 10×2 repetitions, (I) raise the trunk and lower the legs by placing the upper leg on the supporter’s knee in a side bridge position 10×2 repetitions (figure 1).

Lumbar disc degeneration was analysed using T2-weighted MR images before each international competition as a part of the precompetition medical check-up. Two orthopaedic doctors evaluated disc degeneration according to the Pfirrmann classification26 and Grade III and higher were indicative of degeneration. These swimmers received feedback from orthopaedic doctors about the development of intervertebral disc degeneration.

Statistical analysis

We calculated the incident rate of injuries in the top four sites from 2002 to 2016 every year.

The study consisted of two phases: a 7-year (2002–2008) observation period and an 8-year (2009–2016) intervention period. The relative risk of lumbar injuries, incidence rate of total injuries and top four sites were compared between the observation period (2002–2008) and the intervention period (2009–2016) using χ² test. Incidence rates were calculated as the number of injuries per competing athlete and per 100 athletes by body of parts in males, females and total from 2002 to 2016. For incidence rates, 95% CIs were calculated. Incidence rate of injuries in the top three sites was measured and the comparison between the swimming strokes was performed using χ² test. Comparison of the incidence rate by sex was done using χ² test. The average age of each was compared with an unpaired t-test. Statistical significance was set as 0.05. SPSS V.24 (SPSS) was used for all analyses.

Incidence rate of injuries in the top four sites from 2002 to 2016

Figure 2 shows the trends of the injury rates of the lumbar spine, shoulder, knee and ankle joints—the top four injury sites—over 15 years. A downward trend was observed for lumbar injury from 2002 to 2016. The highest incidence, 33.3%, was observed in 2002. However, the incidence in 2009, the year immediately following the implementation of the lumbar injury prevention project, was the lowest at 5.6%.

Request permission

Shoulder joint injury incidence was approximately 10% in 2002–2008, but in 2012 and 2014 it was more than 15%, and the highest, 25.7%, was in 2015. Thus, an increasing trend in shoulder joint injury has been observed in the recent years.

Knee joint injury incidence was around 5%–10% in 2002–2008, but it rapidly increased to 20.7% in 2010. Subsequently, it was more than 20% in 2012, 2014 and 2015, and an increasing trend in knee joint injury was noted in recent years. There was no marked change in the ankle joint injury incidence in 2002–2016 and was around 5%; the highest incidence was 10.5% in 2014.

Effectiveness of the lumbar injury prevention project

The demographic characteristics, including age, height, weight and body mass index of the swimmers, are presented in table 1. During the observation and intervention periods, 51 lumbar injuries (23.5%) and 40 lumbar injuries (14.8%) were recorded, respectively (table 1, figure 3). Lumbar injury incidence decreased significantly during the intervention period (p=0.014, OR 0.564, 95% CI 0.356 to 0.892). The relative risk was 0.11 (95% CI 1.020 to 1.218, p=0.014).

Request permission

In the observation period, 94 out of 217 swimmers (43.3%) experienced some type of injury. In the intervention period, 122 of 271 swimmers (45.0%) experienced some type of injury. There was no significant difference in the type of injury before and after the project (p=0.707, OR 1.071, 95% CI 0.748 to 1.535). In contrast, the knee injury incidence increased significantly during the intervention period (p=0.001, OR 3.171, 95% CI 1.536 to 6.548). There was no significant difference in the shoulder and ankle injury incidence before and after the project (shoulder: p=0.286, OR 1.358, 95% CI 0.773 to 2.385, ankle: p=0.548, OR 1.316, 95% CI 0.536 to 3.236; table 1).

Number of injuries and injury rates in each body part

A total of 283 injuries were recorded for the Japan national swim team from 2002 to 2016. The highest injury rate was observed for the lumbar (total 18.6%, males 15.1%, females 22.9%, table 2), followed by shoulder (11.9%, 10.2%, 13.9%), knee (9.4%, 7.2%, 12.1%) and ankle (4.3%, 2.6%, 6.3%).

Injured parts of the body depending on the type of swimming stroke

The top three injuries for each swimming stroke were as follows: freestyle (n=172), lumbar (16.9%, 95% CI 16.4% to 17.3%), knee (10.5%, 95% CI 10.1% to 10.8%), shoulder (7.6%, 95% CI 7.3% to 7.9%); backstroke (n=78), lumbar (24.4%, 95% CI 23.3% to 25.4%), shoulder (23.1%, 95% CI 22.0% to 24.1%), ankle (6.4%, 95% CI 5.8% to 7.0%); breaststroke (n=86), lumbar (20.9%, 95% CI 23.3% to 25.4%), shoulder (14.0%, 95% CI 13.2% to 14.7%), knee (12.8%, 95% CI 12.0% to 13.6%); butterfly (n=78), lumbar (19.2%, 95% CI 18.2% to 20.2%), shoulder (9.0%, 95% CI 8.3% to 9.7%), knee (6.4%, 95% CI 5.8% to 7.0%); individual medley (n=74), lumbar (13.5%, 95% CI 12.8% to 14.3%), shoulder and knee injury incidence was equal (10.8%, 95% CI 10.1% to 11.5%, each). For all strokes, the lumbar injury was most frequently observed.

Eighteen out of 78 swimmers whose event was backstroke (23.1%) experienced shoulder injuries. The shoulder injury rate was significantly higher among backstroke swimmers than among other swimming strokes (p=0.009; figure 4).

Request permission

Injury rate by sex

Overall, 100 out of 365 male and 116 out of 223 female swimmers experienced an injury in at least one body part. The injury rate was significantly higher in females (52.0%) than in males (37.7%) (p=0.002).

Age of participants in the injured and non-injured groups

Figure 4 shows the average age of the injured and non-injured groups by sex. The average age of males in the injured group was 22±3 years, and the age in the non-injured group was 21±3 years. The average age of females in the injured group was 21±3 years, and that of females in the non-injured group was 20±2 years. The injured group was significantly older than the non-injured group (males, p=0.046; females, p=0.002).

Effectiveness of the lumbar injury prevention project

The most common site of injury was the lower back, followed by the shoulders and knees. While it has been reported that there are many shoulder injuries in swimming,5–12 the possible reasons why there were so many lumbar injuries in this study were: differences in training volume and method, and genetic factors. However, the exact reason was not clear and further study is required.

The introduction of the lumbar injury prevention project in 2008 resulted in a significant decrease in the lumbar injury rate (figure 3). The overall injury rate before (43.3%) and after (45.0%) this project showed no statistically significant difference. However, the lumbar injury rate significantly decreased from 23.5% to 14.8% (p<0.05). Deep trunk muscle training in this project focused on the reinforcement of the deep trunk muscles, such as the transversus abdominis muscle and the multifidus muscles.25 The deep trunk muscles contribute to the prevention of lower back injuries.28–30 The deep trunk muscle training conducted in this study reveals an immediate effect after intervention exercise, which is the lumbar lordosis angle in water motion.31 Therefore, this project was effective in the prevention of lumbar injuries. During the interventional period of the present study, we consider that the strengthened deep trunk muscles contributed to the stabilisation of the lumbar spine during swimming movements and reduced the number of lumbar injuries.

Furthermore, this study provided an intervention protocol and education. The current evidence suggests that exercise in combination with education is effective for preventing low back pain.19 Since we were not able to have a control group for with and without early detection and education in the prevention project after 2008, comparisons were therefore made by using the period before 2008, in which no education of risk management by disc degeneration check using MRI served as the control group. However, we believe that exercise intervention and education of risk management by disc degeneration check using MRI, along with the importance of deep trunk muscle exercises (not implemented before 2008), were also factors contributing to this reduced incidence.

Injury trends over 15 years

Shoulder joint injury incidence increased in swimmers in recent years. The backstroke swimmers have more shoulder injury than other strokes. Impingement of the shoulder joint occurred at shoulder joint elevation and internal rotation.32 Compared with other strokes, the backstroke forces the humeral internal rotation sprain in the upper limb elevation during the catching motion. If swimmers perform a catch motion from the moment they enter the water with insufficient mobility of the thoracic vertebrae, impingement is likely to occur due to excessive inner rotation of the scapulobrachial joint.33 The backstroke swimmers may need to acquire higher thoracic vertebrae and rib cage mobility as opposed to other strokes.

The knee joint incidence also tended to increase over 15 years. Soder et al reported that abnormal knee injury findings were observed in the MRI images in 69.2% of the swimmers, regardless of pain in the knee joints.34 Most reported pain around the patella and the patellar-femoral joint.34 In this study, we did not have MRI on all knee injuries. However, most of orthopaedic surgeons’ diagnosis was patellofemoral pain syndrome, which accounted for most of the knee joint injuries, where the knee joint is affected by the knee joint extension motion, such as the start motion performed during swimming. The increase in the knee joint injury incidence became noticeable in this study for the period 2008–2010, during which the start inclination angle changed from 7.5° to 10° and with starting block. The time when the start block was introduced coincided with the period when the increase in the incidence of knee joint injuries was noted. Studies have reported that the kick-start using this inclined starting block allows a shorter block time, a higher horizontal take-off velocity and better timings for 15 m compared with the track start.35 Therefore, there is a possibility that the load applied to the joint increases as the floor reaction force increases.

Risk factors for injury

In this study, the risk factors for injury were female sex and older age. The higher injury rate in female swimmers is consistent with the results of previous studies.6 7 36 Further studies are required to examine the amount of training, the intensity of training and fatigue level, and clarify why women often have overuse injuries. Regarding older age, findings indicate that there was no relationship between shoulder joint injury and age.9 15 37 38 In contrast, a long history of participation in swimming competitions was a risk factor for injury.20 21 In this study, age was a risk factor because there was a possibility that an older age may correlate with a longer history of training load. Since the incidence of injuries is greater in female swimmers, and in swimmers of an older age, the rate of injuries will be reduced in the future by taking preventive measures especially for young women. Consequently, we recommend that lumbar injury prevention intervention commences at a young age to mitigate the increasing incidence of lumbar injuries in female athletes as they mature.

Limitations

A limitation of this study was that we could not calculate the athlete exposure. It is important to analyse the athlete exposure separately according to the diagnoses, as noted in the study by Zachary6 to determine the cause of the injury. However, athlete exposure was substituted by the injury rate per 100 registered athletes based on the report of Mountjoy in this study.2 In the future, athlete exposure should be calculated and detailed analysis must be conducted. In this study, it was also impossible to establish a control group. The reason for this was that all the subjects in this study were national team swimmers, and could not be distinguished between swimmers. Therefore, in this study, we set up two groups before and after implementation of the prevention project.

Furthermore, we could not calculate the compliance rate of training interventions. During this intervention period, the physiotherapists supervised swimmer’s exercise techniques at major international competitions and national training camps every day, so that swimmers were familiar with the proper exercise techniques and understood the importance of the exercises. In particular, knee, shoulder and lower back injuries increased from 2012 to 2015, which may be related to the change of training such as high-intensity training and weight training, but the causal relationship is unknown.

In the future, we would like to survey the incidence of injuries based on athlete exposure and type of training and investigate the effect of injury prevention intervention for the shoulder and knee joints, because the incidence of shoulder and knee joint injuries was found to be increasing as per an injury surveillance study.