Is there a strong association between substantial injuries and previous injuries in adolescent elite athletes? A 1-year prospective cohort study

ABSTRACT Introduction Adolescent elite athletes have a high injury risk and many risk factors for injury have been suggested, where the most conclusive risk factor is a previous injury. However, there is a lack of longitudinal data on a complete season in adolescent elite athletes. Objectives The aim of the study was to explore the relationship between substantial injuries and previous injuries in adolescent elite athletes. A secondary aim was to explore sex differences in terms of this relationship. Methods Injury problems and substantial injury was monitored in adolescent elite athletes (n = 320) using the validated the Oslo Sports Trauma Research Center Questionnaire over 52 weeks. Results In total, 74% (n = 237) athletes reported at least one substantial injury during the study period. Previous injury problems were reported by 82% (n = 195), where 48% (n = 183) of all substantial injuries occurred within the same body location as a previous injury. Forty-four percent (n = 83) of the substantial injuries occurred within 2 weeks after occurrence of an injury problem. There was no association between sex and number of substantial injuries (p = .956, χ2 = 1.7). Poisson regression analysis demonstrated that 16-year-old athletes had a significantly (p = .034) increased risk of reporting substantial injuries (IRR 1.19, 95% CI: 1.01–1.39), compared to 18-year-old athletes. A high prevalence of substantial injuries occurred closely followed a previous injury in the same body location. Conclusion Exploring rehabilitation following injuries in this age, addressing injury risk behavior when an injury problem has occurred and increasing awareness of the relationship between injury problems and substantial injuries are suggested to be important strategies to reduce substantial injuries in adolescent elite athletes.


Introduction
Increased sport participation, the competitive nature of youth sports, and early sport specialization (intensive, year-round training in a single sport) before puberty in recent decades have been suggested to contribute to a high number of sport injuries (Brenner, 2007;DiFiori et al., 2014;Goldberg, Moroz, Smith, and Ganley, 2007;Post et al., 2020). Sport is the main cause of injury in adolescents, and adolescent elite athletes have a high injury risk partly related to high levels of exposure to sports at a period characterized by rapid physical growth (Abernethy and Bleakley, 2007;DiFiori et al., 2014;Maffulli, Baxter-Jones, and Grieve, 2005). Sport-related injuries can either consist of acute injuries that occur in a single traumatic event or overuse injuries that occur over a longer period of time due to repetitive stress of the musculoskeletal system without sufficient recovery to enable structural adaption (Abernethy and Bleakley, 2007;DiFiori et al., 2014;Yang et al., 2012). The reported injury incidence in adolescent elite athletes varies from 2 to 18/1000 hours of training and up to 22/ 1000 hours of competition versus 6/1000 hours for college-aged athletes (von Rosen et al., 2018a;Yang et al., 2012). The majority of injuries in adolescent elite athletes normally affect the lower extremities, such as the foot and knee regions, often involving the apophyses (Adirim and Cheng, 2003;Brenner, 2007;von Rosen et al., 2018a;Yang et al., 2012). Even if several longitudinal injury surveillances on adolescent athletes are available, the relationship between occurrence of multiple injuries has not been explored fully.
A previous injury is suggested to be the strongest indicator of future injury (DiFiori et al., 2014). This may be explained by inadequate rehabilitation, persistent symptoms, and physiological deficiencies, such as decreased muscle strength, endurance, and proprioception (Emery, 2003). It may also be related to a specific injury risk behavior or trait associated with the previously injured athlete. Findings derived from football studies, have showed a significant risk of developing future injury if a previous injury was reported during the earlier season (Emery, Meeuwisse, and Hartmann, 2005;Turbeville et al., 2003). In an American study where 60% of the included football-playing adolescents reported injury, a doubled risk for a future injury occurred after a previous injury and 3-times the risk of injury if two or more previous injuries were reported (Kucera et al., 2005). A previous injury is classified as a recurrent injury, if the injury is of the same type and located in the same body region as a previous injury (DiFiori et al., 2014;Fuller et al., 2006). However, it is possible that other body regions are also affected even if the injury is not located in that specific region. For instance, an athlete with a foot injury may not be able to load their knee or hip to a similar extent as before the injury, making the athlete more susceptible to injuries in these regions as well. If an injury occurs in these regions, followed a foot injury, it would not be classified as a recurrent injury.
Adolescent female athletes may be more likely to suffer from overuse injuries compared to adolescent male athletes who rather incur more acute injuries (Cuff, Loud, and O'Riordan, 2010;Stracciolini et al., 2014;von Rosen, Heijne, and Frohm, 2016;Yang et al., 2012). Two previous studies on adolescent elite athletes at National Sports High Schools in Sweden and Colleges in the USA showed an increased average weekly injury prevalence and a higher severity of injuries in adolescent females compared to adolescent male athletes (von Rosen et al., 2018a;Yang et al., 2012). The location of injury differs also by sex, for instance, female athletes have an increased risk for knee injuries compared to male athletes (Cuff, Loud, and O'Riordan, 2010;McGuine, 2006;Messina, Farney, and DeLee, 1999;Powell and Barber-Foss, 2000;Rommers et al., 2020).
As previous literature has documented an increase in sports-related injuries (Abernethy and Bleakley, 2007;Brenner, 2007;Goldberg, Moroz, Smith, and Ganley, 2007), there is a great interest in further investigation of the time before an adolescent elite athlete gets an injury. By exploring the time before the injury onset, some risk factors can possibly be identified. These findings can lead to improved injury prevention methods. The aim of the study was therefore to explore the relationship between substantial injuries, defined as any physical complaint leading to a moderate and severe reduction in training volume, reduction in performance or complete inability to participate, and previous injuries in adolescent elite athletes. A secondary aim was to explore sex differences in terms of this relationship.

Design
This study is part of the larger KASIP-study (Karolinska Athlete Screening Injury Prevention) project, which aims to understand injury occurrence and associated risk factors in Swedish adolescent elite athletes. In this study, data was collected using an injury questionnaire over 52-weeks, between September 2013 and December 2015. Ethics approval was granted by the Regional Ethical Committee in in Sweden (2011/749-31/3).

Data collection
Adolescent elite athletes from 24 National Sports High Schools in Sweden were invited to participate in an injury surveillance study. All athletes were informed of the purpose of the study and the voluntary nature of participation. Written consent was obtained from all athletes.
An injury questionnaire was e-mailed to the athletes over a period of 52 weeks. During the first year of the study the adolescent elite athletes received the questionnaire every week and every other week for year two. If no response were registered, the adolescent elite athlete received a reminder by e-mail 4 days later. If there was still no response, the athlete was contacted again as usual and also asked to respond the previous missed week.
During the first week of the study, all adolescent elite athletes were asked to fill in an online background questionnaire regarding personal and sports-specific information. The Questback online survey software (version 9.9; Questback AS, Oslo, Norway) was used for data collection.

Participants
A total of 517 adolescent elite athletes from National High Schools in Sweden, aged between 15 and 19 years, were included to participate in the study. The athletes were considered to be elite athletes since the athletes participated in national teams or at the highest national level of their age.
All of the athletes were connected to one of the included 24 National Sports High Schools and 17 difference sports were included. The included sports were American football, athletics, bowling, canoeing, cycling, cross-county skiing, downhill skiing, freestyle skiing, golf, handball, orienteering, rowing, ski orienteering, triathlon, running, water skiing, and wrestling.
Due to insufficient data such as low response rate or lack of background information a total of 197 adolescent elite athletes were excluded. The final cohort consisted of 320 adolescent elite athletes (girls = 165 and boys = 155). The excluded athletes did not differ from the main cohort under investigation regarding sex, age, and injury at study start (p > 0.05).

Questionnaire
The injury questionnaire contained the validated and translated version of the Oslo Sports Trauma Research Center (OSTRC) Overuse Injury Questionnaire and questions used by Jacobsson et al. (2010); Jacobsson et al. (2013)) in the cohort study performed on adolescent and senior elite athletics in Sweden (Ekman et al., 2015). The questionnaire consisted of three parts: questions about training variables and prevalence of injury problems, questions about any new injury occurrence, and questions about the return to sport after an injury.

Data preparation
In this study, we decided to include the adolescent elite athletes who responded for a continuous period of at least 20 weeks. Missing reporting for a maximum of 2 weeks during the continuous period was considered acceptable to be included. If the athlete had responded for at least 20 weeks and a break in reporting was noted, the athletes could be included again if reported for at least 4 continuous weeks.

Injury definition and classification
All injury data were self-reported, and the definitions of an injury problem or substantial injury is based on the definitions in the OSTRC Overuse Injury Questionnaire. An injury problem was defined as any physical complaints that affected participation in normal training or competition or led to reduced training volume or experience of pain. A substantial injury was defined as any physical complaint leading to a moderate or severe reduction in training volume, a moderate or severe reduction in performance, or the complete inability to participate in sports related to an injury (Clarsen, Myklebust, and Bahr, 2013;Ekman et al., 2015). A new substantial injury in the same region as a previous substantial injury was only recorded if the athlete had been injury free for at least 4 weeks in between. Substantial injuries from the foot, lower leg, knee, and hip were included in the definition of lower extremity. Lowerand upper back complaints and neck injuries were included in the definition of torso, and substantial injuries in the shoulder, elbow, and hand were included in the definition of upper extremity. Time to onset of substantial injury and the total time with substantial injury were noted. Time to onset of substantial injury was calculated based on the inclusion date or time from reported to be injury free for at least four weeks, in case of multiple substantial injuries.
A substantial injury was classified as a global or local injury based on the location of previous injuries. If a substantial injury was reported in the same lower/ upper extremity or torso as a previous injury problem, it was classified as a global injury. For example, if an athlete reported substantial injury in the knee and previously had reported injury problems in the hip, the substantial injury was classified as a global injury. Substantial injury reported in the same specific region (foot, lower leg, knee, hip, lower-and upper back, neck, shoulder, elbow, or hand) as a previous injury problem was classified as a local injury. For example, if an athlete reported substantial injury in the shoulder and previously had reported injury problems in the shoulder, the substantial injury was classified as local injury. The reason for classifying substantial injury as local and global injury was to more explore the relationship between previous injuries and substantial injury more in-depth. Since injuries may affect multiple body regions, even if the injury is not located in that region, we wanted to explore the proportion of such injuries (i.e. global injury) and contrast these to injuries in the same body region (i.e. local injury).

Data analysis
Chi-square tests were applied to explore the relationship between sex and number of substantial injuries, and sex and type of previous injury. A Poisson regression was used to estimate the incidence rate ratios (IRRs) for substantial injuries. Possible independent variables were age, sex, and sport types. Sport types was divided into: endurance sports (i.e. cross-country skiing, cycling, orienteering, running, ski-orienteering, and triathlon); contact sports (American football, handball, and wrestling); and other sports (athletics, bowling, canoe, downhill skiing, freestyle skiing, golf, rowing, and water skiing). Robust standard errors were calculated to account for any overdispersion. We evaluated how well the data were approximated by the Poisson model by determining the amount of overdispersion. If data follow a Poisson distribution, then the variance will be approximately equal to the mean.
To explore the proportion of all substantial injuries that was classified as a global or local injury, we excluded substantial injuries up to 26 weeks. As we have limited data on previous injuries when a substantial injury occurs during the first weeks, we wanted to explore the proportion of global and local injuries occurring later in the study. A p-value ≤ 0.05 was considered as statistically significant. All analyses were conducted using the R statistical system version 3.5.2 (R Foundation for Statistical Computing, Vienna, Austria, 2021).

Substantial injuries
A total of 381 unique events of substantial injury were identified, and over 1-year 74% (n = 237) of the adolescent elite athletes reported at least one substantial injury. These injuries were mainly located in the lower extremity (65%, n = 248) ( Table 1). Of the adolescent elite athletes that reported substantial injury, the median time for sustaining these injuries was 5 weeks (25 th -75 th percentile 1-15). There was no association between sex and number of substantial injuries (p = .956, χ2 = 1.7) (Figure 1). The median duration of substantial injury was 4 weeks (IQR 2-9). At the time the study was finished, 29% (n = 94) reported still injury problems and 16% (n = 51) substantial injury.
The Poisson regression analysis demonstrated that 16-year old athletes had a significantly (p = .034) increased risk of reporting substantial injuries (IRR 1.19, 95% CI: 1.01-1.39), compared to 18-year old athletes (Table 2). Sport types or sex was not associated with number of substantial injuries.

Injury classification
Of the adolescent elite athletes with substantial injuries (n = 237), the majority (82%, n = 195) had previously reported injury problems (Figure 2). Sixty-five percent (n = 248) and 48% (n = 183) of all substantial injuries (n = 381) were classified as a global and local injury, respectively. There was no association between sex and number of previous injury problems, global injury, and local injury (p = .672). There was an increased probability that the injury was a global or local injury the later a substantial injury was reported in the study period ( Figure 3). For instance, excluding all substantial  injuries occurring before week 26 (n = 299, 78%), 70% (n = 82) of the remaining substantial injuries was classified as a local injury. The average time between an injury problem and a local injury was 9 weeks (SD 9.8), where 58% (n = 108) of all local injuries occurred within 6 weeks after occurrence of an injury problem.

Discussion
The study aimed to explore the relationship between substantial injuries and previous injury problems.
The main finding was that adolescent elite athletes that sustained substantial injuries had a high prevalence of previous injury problems. In this study, 82% of the adolescent elite athletes with substantial injuries reported previous injury problems and 48% of all substantial injuries were classified as local injuries.
There was no association between sex and number of previous injury problems, global injury and local injury.
In this study, 74% of the adolescent elite athletes reported at least one substantial injury and 32% reported more than one substantial injury. Our results emphasize that a high part of substantial injuries (48-70%) occur in the same body location as a previous injury. In addition, as many as 44% (n = 83) of all substantial injuries occurred within 2 weeks after occurrence of an injury problem. Several studies have highlighted that adolescent elite athletes have a high injury risk (Jacobsson et al., 2013;Moseid et al., 2018;von Rosen et al., 2018a), however the causal relationship is not fully clear. Even if the majority of the included athletes specialize in a single sport, there are contraindicatory findings regarding if sport specialization may increase the risk of injury Table 2. Showing incidence rate ratios (95% confidence interval (CI)), standard errors, and P-values, based on the Poisson regression model for number of substantial injuries.  Figure 2. Number of substantial injuries, injuries with history of injury problems, number of global and local injury. (Garcia et al., 2021;Jayanthi et al., 2020;Moseid, Myklebust, Fagerland, and Bahr, 2019). However, we do know that a previous injury is one of the most important risk factors for developing a future injury (Emery, Meeuwisse, and Hartmann, 2005;Kucera et al., 2005;Turbeville et al., 2003). This is suggested to be explained by inadequate rehabilitation, persistent symptoms, and physiological deficiencies such as decreased muscle strength, endurance, and proprioception or by specific injury risk behavior (DiFiori et al., 2014;Emery, 2003;Emery, Meeuwisse, and Hartmann, 2005;Jacobsson et al., 2013). Of these factors, which is the most relevant for development of injuries in adolescent elite athletes is not clear, although challenges with rehabilitation in this age group has been highlighted (von Rosen et al., 2018b). Our study emphasizes that even in young athletes, it is common to have had minor injuries in the same body location closely before occurrence of a substantial injury. This suggests that young athletes have a history of injury problems before many substantial injuries occur, which could be relevant in injury prevention measures. There were no significant differences between sex and number of previous injury problems, global and local injuries, suggesting that both adolescent male and females have a similar distribution of previous injuries. Our findings also showed that 16-year-old athletes had a significantly increased risk of reporting substantial injuries compared to 18-year-old athletes. The results are in line with previous studies that showed an increased risk for injury in 14-year-old football players compared to 18-year-old players (Emery, Meeuwisse, and Hartmann, 2005) and the general perception that the risk of injury gradually decreases with age (Adirim and Cheng, 2003). However, in a recent study that examined young and older elite level soccer players, relative age, body weight, leg length, as well as total growth and leg growth were associated with an increased injury risk in both younger and older soccer players (Rommers et al., 2020). We believe there is no consensus which age group has the greatest injury risk and individuals go through a period of varying growth velocity that also is associated with injury risk.
At the baseline, 18% of the adolescent elite athletes reported a substantial injury and at the time the study was completed, 16% reported still substantial injury. The results therefore indicate that almost a fifth of the adolescent elite athletes at National Sports High Schools in Sweden constantly have a substantial injury. That 32% of the adolescent elite athletes reported more than one substantial injury and 48% of the substantial injuries were classified as local injuries suggest that the proportion of overuse injuries may be underestimated. Since symptoms of overuse injuries appear gradually and may be transient in nature, it is hard to fully capture the extent of overuse injuries in this age group, even though recent studies have suggested it to be high (Jacobsson et al., 2013;von Rosen, Heijne, and Frohm, 2016). Future injury surveillance studies with longer follow-up are warranted. The strengths of this study are the large number of adolescent elite athletes, the prospective design and that the athletes were followed over 1 year. Regular self-reporting has been performed every week or every second week based on a validated questionnaire which is sensitive to capturing all kinds of physical complaints (Clarsen, Myklebust, and Bahr, 2013;Ekman et al., 2015). Another strength of this study is that a continuous period of self-reporting was chosen to exclude sporadic report occasions and thus enable to explore the time before substantial injury. On an average, the response rate of the included athletes was 82%. Missing data was not explored.
The study also has limitations. Adolescent elite athletes entered the study at different times of seasons within their own sports (training or competition season), which may result in different injury risks between different sports. However, the adolescent elite athletes were followed over 1 year, which should limit and minimize these seasonal bias. Another limitation of this study is that substantial injury was classified as a local injury based on the same body location. We do not have medical data to confirm that the new injury is of the same type as a previous injury. Due to few participants in team sports, our results can mainly be extended to individual sports as they accounted for almost 80% of our sample. A few individual sports (i.e. bowling, canoeing, cycling, and freestyle skiing) also contributed with few participants and thus our results may not accurately reflect these sports. Nevertheless, we managed to include almost a third of all athletes studying on National Sports High Schools in Sweden.

Clinical implications
In summary, the results of the study demonstrated that a high prevalence of substantial injuries is related to previous injury problems in adolescent elite athletes. Specifically, between 48-70% of all substantial injuries occurred in the same body location as a previous injury within the same season. The results of the study therefore confirm that a previous injury is a strong predictor of future substantial injury in adolescent elite athletes and that almost half of the local injuries occurred within 2 weeks from a previous injury problem. Exploring rehabilitation following injuries in this age, addressing injury risk behavior when an injury problem has occurred, and increasing awareness of the relationship between injury problems and substantial injuries are suggested to be important strategies to reduce substantial injuries in adolescent elite athletes.

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