Article Text

Download PDFPDF

Lifetime history of sexual and physical abuse among competitive athletics (track and field) athletes: cross sectional study of associations with sports and non-sports injury
  1. Toomas Timpka1,2,3,
  2. Staffan Janson4,
  3. Jenny Jacobsson1,2,
  4. Örjan Dahlström1,5,
  5. Armin Spreco1,2,3,
  6. Jan Kowalski1,2,
  7. Victor Bargoria1,2,6,
  8. Margo Mountjoy7,8,
  9. Carl Göran Svedin1,9
  1. 1 Athletics Research Centre, Linköping University, Linköping, Sweden
  2. 2 Department of Medical and Health Sciences, Linköping University, Linköping, Sweden
  3. 3 Unit for Health Analysis, Centre for Healthcare Development, Region Östergötland, Linköping, Sweden
  4. 4 Department of Women ́s and Children ́s Health, Uppsala University, Uppsala, Sweden
  5. 5 Department of Behavioural Sciences and Learning, Linköping University, Linköping, Sweden
  6. 6 Department of Orthopaedics and Rehabilitation, Moi University, Eldoret, Kenya
  7. 7 Department of Family Medicine, McMaster University, Hamilton, Ontario, Canada
  8. 8 International Olympic Committee Medical Commission, Games Group, Lausanne, Switzerland
  9. 9 Barnafrid, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden
  1. Correspondence to Prof. Toomas Timpka, Department of Medical and Health Sciences, Linköping University, Linköping SE-581 83, Sweden; toomas.timpka{at}


Objective To examine associations between lifetime sexual and physical abuse, and the likelihood of injury within and outside sport in athletes involved in competitive athletics.

Methods A cross sectional study was performed among the top 10 Swedish athletics athletes using 1 year prevalence of sports and non-sports injuries as the primary outcome measure. Associations with sociodemographic characteristics, lifetime abuse history and training load were investigated. Data were analysed using simple and multiple logistic regression models.

Results 11% of 197 participating athletes reported lifetime sexual abuse; there was a higher proportion of women (16.2%) than men (4.3%) (P=0.005). 18% reported lifetime physical abuse; there was a higher proportion of men (22.8%) than women (14.3%) (P=0.050). For women, lifetime sexual abuse was associated with an increased likelihood of a non-sports injury (OR 8.78, CI 2.76 to 27.93; P<0.001). Among men, increased likelihood of a non-sports injury was associated with more frequent use of alcoholic beverages (OR 6.47, CI 1.49 to 28.07; P=0.013), while commencing athletics training at >13 years of age was associated with a lower likelihood of non-sports injury (OR 0.09, CI 0.01 to 0.81; P=0.032). Lifetime physical abuse was associated with a higher likelihood of sports injury in women (OR 12.37, CI 1.52 to 100.37; P=0.019). Among men, athletes with each parents with ≤12 years formal education had a lower likelihood of sustaining an injury during their sports practice (OR 0.37, CI 0.14 to 0.96; P=0.040).

Conclusions Lifetime sexual and physical abuse were associated with an increased likelihood of injury among female athletes. Emotional factors should be included in the comprehension of injuries sustained by athletes.

  • athletics
  • sexual harassment
  • sporting injuries
  • epidemiology

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.

What are the findings?

  • A lifetime history of sexual abuse was associated with an increased likelihood of sustaining a non-sports injury among female athletics athletes, while a history of physical abuse increased the likelihood of sports injury.

  • The characteristics of the associations between sexual and physical abuse history, sociodemographic factors and different adverse health outcomes among competitive athletics athletes differed between women and men.

How might it impact on clinical practice in the future?

  • Attention to lifetime abuse and other emotional factors helps in understanding a competitive athletics athlete’s likelihood of sustaining injury within and outside of sport.

  • Physical and sexual abuse have different correlates among female and male athletes, implying that clinical investigations need to be adjusted for gender.


Prospective studies in athletics (track and field) have reported that overuse injuries constitute a major problem among competitive athletes.1 2 Overuse injuries in sport represent bodily harm caused by inadequate balancing of load and recovery, thus harming body tissue through repeated micro trauma.3 In a consensus statement on sport training load as a cause of injury, the International Olympic Committee summarises the current evidence by stating that high loads can have either positive or negative influences on injury risk; the rate of load application and the athlete’s internal risk factors are the important determinants.4 5 If load is applied in a progressive manner, high training loads may even offer a protective effect. In other words, what is damaging is not the exercise load per se but rather the application of load in situations when the athlete’s body is in need of rest and restoration. Behaviours that are performed with the knowledge that they may or will result in some degree of harm to oneself are denoted as self-injurious.6 Therefore, when an adult athletics athlete continues physical exercise despite physical warnings of pain and distress indicative of overuse injury, this behaviour can be described as including self-injurious aspects.

To accommodate the rapidly growing evidence on the topic, self-injury has been sub-categorised using intention and injury severity as the main criteria. For example, self-defeating behaviour occurs when individuals repeatedly engage in behaviours that result in unintended and unforeseen harm,7 while non-suicidal self-injury is defined in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) as intentional self-inflicted damage to the surface of an individual’s body without conscious suicidal intent,8 colloquially known as ‘cutting’. Self-injury is more often observed among women, and the underlying causes also differ between the genders.9 However, lifetime sexual and physical abuse have been shown to be consistent correlates across settings.10 A retrospective community based study involving 11 000 adults reported that childhood physical, but not sexual, abuse was associated with self-injury in adulthood. For women, the self-blame item in the Brief Cope instrument measuring coping behaviours largely accounted for the association of physical abuse and self-injury.11 Of note, in a prospective study of sports injury risk in competitive athletics athletes, this self-blame measure was found to be a significant injury risk indicator.12

Several studies have described victimisation prevalence and risk factors for sexual or physical abuse in sport.13 However, post-exposure correlates (reported to have occurred after the abuse) have been investigated to a lesser extent.14 15 The aim of this study was to examine associations between lifetime sexual and physical abuse history, sociodemographic characteristics, training load and 1 year injury prevalence within and outside sport among competitive athletics athletes. Prevalence is the preferred measure of injury burden in sports where overuse injuries dominate.16 While not suggesting equivalence, it is assumed that different types of injuries sustained by athletes and bodily harm instigated by self-injurious acts share causal pathways. Consequently, a history of physical or sexual abuse is hypothesised to be associated with increased 1 year injury prevalence. For cross validation purposes, both sports and non-sports injuries, which differ with regard to aetiology, are analysed. The former injuries are among athletics athletes characteristically sustained through repeated micro trauma during pre-scheduled training and competition,17 and the latter by accidents in everyday community settings.18


This study used a cross sectional design. One year injury prevalence was used as the primary outcome measure. Data on sociodemographics, injuries and lifetime abuse experiences were collected from athletics athletes competing at the national elite level using a web questionnaire (SiteVision v2.5; Senselogic AB, Örebro, Sweden). Ethics approval for the study was obtained from the research ethics committee in Linköping, Sweden (Dnr 2012/140-31).


The study population consisted of the top 10 Swedish adult or junior (under 20 years of age) athletes in an athletics event (representing the event categories sprints, hurdles, throws, jumps, middle and long distance running, combined events and race walks) at the time of recruitment (August 2013). Four top 10 lists (men/women, adults/youths) for each event were compiled by the Swedish Athletic Association. If an athlete was ranked among the top 10 in more than one event, the athlete was only included in his/her self-reported main event.

Definitions and data collection

A web questionnaire was developed based on a working model for analysis of sports and non-sports injury likelihood among athletics athletes (see online supplementary figure 1). The hypothetical working model was developed by the research team using its overall clinical and research expertise in the field of study. A sports injury was defined as a substantial musculoskeletal condition that made the athlete partially or completely abstain from training or competition for a 3 week minimum injury period during the past year.2 A non-sports injury was defined as an unintentional injury sustained outside the athletics context requiring treatment at an emergency medical facility. Sexual abuse was defined as any sexual interaction involving physical contact with a person(s) of any age that is perpetrated against the victim’s will, without consent or in an aggressive, exploitative, manipulative or threatening manner. More precisely, the dependent variable for the study was defined from the statement and questions originally developed by Mossige19: “Sometimes people are persuaded, pressed or forced to do sexual activities they cannot protect themselves from. The following questions are about such situations. Have you been exposed to any of the following against your will? (it is possible to choose several alternatives): (a) Somebody exposed himself or herself indecently towards you, (b) Somebody has pawed (touched your body in an indecent way) you, (c) You masturbated somebody else, (d) You have had sexual intercourse, (e) You have had oral sex and (f) You have had anal sex". Physical abuse was defined as being deliberately hurt by an adult person causing injuries such as bruises, broken bones, burns or cuts.20 More precisely, the questionnaire asked whether it had happened that an adult had done any of the following to the athlete: “(a) Pushed, shoved or shook you up, (b) Thrown something at you, (c) Hurt you with her/his hands, (d) Kicked, bit or hit you with her/his fists, (e) Hurt you with a weapon, (f) Burned or scalded you, (g) Tried to smother you (took stranglehold) and (h) Physically attacked you otherwise."

Supplementary file 1

The questionnaire first asked for data on demographics and athletics background. Secondly, data on sexuality, lifetime sexual and physical abuse within or outside sport were gathered, including perpetrators, and indicators for a risky lifestyle, such as hazardous behaviours, were collected. The point in time of the abuse was specifically determined. The respondents were asked whether they had suffered an injury within and outside of athletics during the past year (1 year injury prevalence). Regarding injuries suffered in athletics, structured follow-up questions were asked about body location and injury type, and the athlete was asked to describe probable cause and/or provide a diagnosis made by a physician or physiotherapist.

Data analysis

The sports injury data were broken down by body location and injury type,17 while the data on 1 year prevalence of sexual and physical abuse were described by age, sex, sexual orientation and event group. For the analyses of injury likelihood, the data collected for the explanatory variables were recoded into a binary format: lifetime sexual abuse involving physical contact (yes/no), lifetime physical abuse (yes/no), both parents had ≤12 years formal education (yes/no), at least one parent immigrant (yes/no), family perceived by athlete as financially disadvantaged (yes/no), age >13 years when commencing training athletics (yes/no), training load >20 hours/week (yes/no), use of alcoholic beverages at least twice a week (yes/no) and binge use of alcoholic beverages (equivalent to one bottle of wine) at least once a week (yes/no). Athletics events were recoded as speed/power (sprints, hurdles, throws, jumps, combined events) and endurance events (middle and long distance running, race walking).

We first performed analyses with simple models (logistic regression analyses with one explanatory variable) and thereafter analyses with multiple models (logistic regression analyses with several explanatory variables) using sports injury during the study year (yes/no) and healthcare treated non-sports injury during the study year (yes/no) as outcomes. An odds ratio (OR) <1 indicates an enhanced likelihood for injury. The explanatory variables were abuse types (sexual/physical), event category and the binary risk indicator variables. The multiple models were fitted using backward elimination of non-significant variables (ie, variables with P≥0.05 were stepwise eliminated). The analyses of injury likelihood were performed on the total respondent population and separately for male and female athletes. Nagelkerke R2 was computed for all multiple models to estimate their determination levels.21 The Statistical Package for the Social Sciences (SPSS) for Windows V.23.0 was used for the analyses. All statistical tests were two sided and outcomes with P<0.05 were considered to be statistically significant.


Of 507 eligible athletes, 197 (38.9%) returned complete data sets; 105 (53.3%) of the respondents were women. The mean age of the respondents was 23.0 years (SD 4.7 years; range 17–37 years): juniors 19.2 years (SD 0.8 years; range 17–20 years), seniors 26.1 years (SD 4.2 years; range 21–37 years). The event groups were equally distributed among responders and non-responders: 31% of respondents were middle or long distance runners, 25% throwers, 20% jumpers, 19% sprinters and 5% were combined event athletes. Ninety-two per cent of the athletes described their sexual orientation as heterosexual. The majority stated that during their childhood and adolescence, they had been living in families with a stable economy (83.8%) and at least one parent/custodian having more than 12 years of formal education (72.6%) (table 1). Twelve per cent of the athletes reported the use of alcoholic beverages twice a week or more, and 9.6% reported binge drinking (equivalent to one bottle of wine) at least once a week.

Table 1

Study population characteristics (n=197)

Lifetime sexual and physical abuse

Eleven per cent of the participating athletes reported having sustained lifetime sexual abuse (table 2); a higher proportion of women (16.2%) than men (4.3%) (P=0.005). Six athletes (3.0%) reported sexual abuse in the athletics setting. The perpetrators of sexual abuse in athletics included men and women, coaches and peers in late adolescence. Two athletes reported suicidal ideation because of the abuse experienced in the athletics setting. Regarding lifetime physical abuse, 18.3% of the participating athletes reported victimisation; a higher proportion of men (22.8%) than women (14.3%) (P=0.050). Non-heterosexual athletes tended to more often report lifetime physical abuse than their heterosexual peers (P=0.058). The perpetrators of physical abuse were predominantly parents/custodians; only three athletes reported having been physically abused by a coach. All abuse episodes were reported to have occurred more than 1 year before the data collection.

Table 2

Proportions of lifetime sexual abuse victims (in the athletics setting) and lifetime physical abuse victims in the study sample, by sex, athletics event and sexual orientation (n=197)

One year injury prevalence

The 1 year prevalence of sports injury among the participating athletes was 55.3% (men 57.6%; women 53.3%). The most common injury cause was overuse (89.0%), while the most frequent location was the foot/ankle (24.8%), followed by the thigh (14.7%), back/trunk (14.7%) and lower leg (13.8%) (table 3). Regarding non-sports injury, 16.8% of the athletes (men 15.2%; women 18.1%) reported having sustained an injury outside of training or competition that required emergency medical care.

Table 3

Proportions of reported sports injury (n=109) types and body locations, by gender

Associations between lifetime sexual and physical abuse and 1 year injury prevalence

The simple model analyses showed that lifetime physical abuse in women was associated with an increased likelihood of sports injury (OR 4.18, CI 1.10 to 15.83; P=0.035) (table 4). Among men, athletes with both parents having ≤12 years of formal education showed a lower likelihood of sustaining an injury during training and competition (OR 0.37, CI 0.14 to 0.96; P=0.040). In the analyses of multiple models, lifetime physical abuse was associated with a higher likelihood of sports injury for female athletes (OR 12.37, CI 1.52 to 100.37; P=0.019; R2 =0.14), while among males athletes with both parents having ≤12 years of formal education, a lower likelihood was found (OR 0.37, CI 0.14 to 0.96; P=0.040; R2 =0.03). With regards to non-sports injury (table 5), the simple model analyses displayed that lifetime sexual abuse among women was associated with a higher likelihood of injury (OR 8.78, CI 2.76 to 27.93; P<0.001), while among men an increased likelihood of non-sports injury was associated with a more frequent use of alcoholic beverages (O.R, 4.26, C.I, 1.17 to 15.55; P=0.028) and binge drinking (OR 4.80, CI 1.15 to 20.01; P=0.031). In the multiple models for women, lifetime sexual abuse was associated with an increased likelihood of non-sports injury (OR 8.78, CI 2.76 to 27.93; P<0.001; R2 =0.20). Similarly, in the multiple models for male athletes (R2 =0.14), frequent use of alcoholic beverages was associated with an increased likelihood of non-sports injury (OR 6.47, CI 1.49 to 28.07; P=0.013) and commencement of athletics training at >13 years of age with a lower likelihood (OR 0.09, CI 0.01 to 0.81; P=0.032).

Table 4

Associations between lifetime sexual and physical abuse, personal background factors and 1 year prevalence of sports injury (>3 weeks' time loss) among athletics athletes (n=197)

Table 5

Associations between lifetime sexual and physical abuse, personal background factors and non-sports injury among athletics athletes (n=197)


This study was designed to examine associations between lifetime history of sexual and physical abuse, sociodemographic factors, training load and injuries sustained within and outside of sport among competitive athletics athletes. Lifetime physical abuse among female athletes was found to be associated with a 12-fold increased likelihood of sports injury and lifetime sexual abuse with an 8-fold increased likelihood of non-sports injury. The lifetime prevalence rates of physical (18.3%) and sexual (10.7%) abuse were similar to those reported in recent community studies in Sweden,20 22 whereas the prevalence of sexual abuse sustained in athletics (3.0%) was in the lower range compared with reports from other sports.13

Physical abuse and sports injury

It is increasingly recognised that the emotional backdrop is an important regulator of adaptation to sport load in high level sport, and that the athlete’s emotional state has to be considered in the timing of training and performance.23 Among female competitive athletes, we found that lifetime physical abuse was associated with an increased likelihood of sports injury, even though this type of abuse was more prevalent among male athletes. Sex differences in correlates to abuse have previously been reported from studies in general populations.9 24 Considering that our data almost uniquely identified parents as the physical abuse perpetrators, the associations with sports injury may be explained by disciplinary interactions with the young athlete. Developmental research has shown that the inability to describe emotions can lead to abused young people expressing affect through their body.25 Hypothetically, repeated physical disciplinary corrections may thus have taught the young athlete that coping with unsatisfactory performance is linked with tolerance of physical pain. In other words, the athlete may have been habituated to associate correction and improvement with suffering pain.26

Sexual abuse and risky behaviours

We found lifetime sexual abuse to be unrelated to sports injury, but among female athletes it was associated with the likelihood of sustaining non-sports injury in community settings. Of note, slightly more women than men reported having sustained a non-sports injury, while the opposite is known from studies in the general population.27 28 Injuries sustained outside of sport18 differ from injuries sustained in athletics17 with regards to typical causal mechanisms. Our results suggest that these differences may also comprise influences from self-injurious elements. Based on their context and causal influences, self-injury has been suggested to be divided into several discrete diagnostic entities with different aetiologies.6 29 30 Risky lifestyles theory31 states that hazardous behaviours, such as alcohol or drug use, are associated with exposure to certain contexts, while social control theory32 contends that adult supervision prevents youth from engaging in these contexts. In this study, male athletes reporting frequent use of alcoholic beverages and female athletes with a sexual abuse history showed a higher likelihood of non-sports injury, while male athletes commencing athletics during adolescence displayed a lower likelihood of such injuries. These findings regarding lifestyle and non-sports injury have in common that they reflect aspects of social control; alcohol consumption at a young age represents disapproval of contexts distinguished by adult supervision, while choosing an individual sport in adolescence mirrors the opposite position. From this perspective, having parents with a low formal educational background can also play a role in explaining the increased risk of sports injuries among boys and non-sports injuries among both sexes, potentially due to different behaviour practices. Likewise, sexual abuse victimisation may indirectly have increased non-sports injury risk by predisposing the victim to distrust and avoid community settings where older people establish norms and regulations. In other words, the association between lifetime sexual abuse and non-sports injury among female athletes may have been mediated by other self-injurious mechanisms than those mediating sport injury (i.e. by risky lifestyle rather than through habituated pain tolerance). There are few studies of injuries sustained by competitive athletes outside sport,33 implying that the present observations warrant thorough confirmation in prospective studies using data collected at multiple levels.

Primary and secondary prevention

The results of this study indicate that primary prevention of sexual and physical abuse should be included in routine sports practices, and also secondary prevention by offering therapeutic measures.34 Just like physical capacities,35 the competitive athlete’s stress resilience needs to be understood and addressed in educational programming and individual training schedules. Athletes should be supported to disclose abuse experiences and be informed about existing reporting and counselling structures. Active screening of a history of abuse can be conducted by trained individuals, as it has been proposed to hold value for intervention planning in non-clinical settings.23 For primary prevention, awareness raising initiatives are needed to inform athletes, coaches, parents, sport managers and federation boards. Institution of sport governing policies and procedures are also important tools of primary prevention. This study and other recent results36 have revealed a more nuanced abuse perpetrator profile than that of the older male coach sexually assaulting a younger athlete, often brought forward in media coverage.37 This is also underlined by a recent Swedish national study including 5873 students in grade 3 at the Upper Secondary School where only a small minority of the sexually abused respondents stated that the perpetrator was a sports coach or supervisor.38 Primary prevention initiatives should therefore also include measures aimed at contexts outside the coach–athlete relationship (eg, at sports parenting, athlete peers and abuse by minors). The design and evaluation of measures for primary and secondary prevention of sexual and physical abuse in sport is an important area for future research.


Research on correlates to psychological, physical and sexual abuse is methodologically challenging, and there is a general paucity of longitudinal analyses.23 The present study was designed as a retrospective analysis conducted using best practice procedures for the research topic.39 Retrospective analysis of abuse experiences has previously been stated to display reasonable accuracy,40 and the determination levels observed in this study for the multiple models on associations between abuse experiences and injury were acceptable. Adolescence and emerging adulthood are the high risk periods for sexual and physical abuse.21 41 In this study, all abuse episodes had occurred before the recorded injury episodes. Also, the response rate (38.9%) may have contributed to selection bias, but the non-response analysis did not display important differences between responders and non-responders.38 Although not influencing the main inferences drawn from the study, it is possible that some athletes avoided participation  given that providing information on physical and sexual abuse is sensitive to elite athletes exposed to extensive media and public attention, rendering a likely underestimation of abuse experiences. This type of information is sensitive for elite athletes exposed to extensive media and public attention, and hence there may have been an underestimation of abuse experiences. Moreover, due to the cross sectional design of the study, we restricted the analyses to examinations of single risk factors, and did not consider more complex mechanisms, such as mediation and interaction.42 43 Accordingly, explanatory power may have been attributed to factors that are better understood as links or mediators rather than primary risk factors. Finally, the study was planned and executed in competitive athletics and the findings may not be applicable to recreational sport and non-athletics competitive sport.


In this study of associations between lifetime physical and sexual abuse and injuries sustained within and outside sport among competitive athletics athletes, we found stronger associations among women and that different abuse experiences were associated with different categories of adverse health correlates. In community settings, physical and sexual abuse have been correlated with lower health status long after the abuse has stopped.44–46 Our results underscore the importance of lifetime abuse and other emotional indicators in the comprehension of injuries sustained by competitive athletes. Further sex specific prospective research on the influences from athletes’ emotional backgrounds on their performance and general health is needed to clarify causal pathways and support intervention designs.



  • Contributors TT is the lead investigator. TT, CGS and SJ designed the study. All authors were involved in the study design and reviewed the draft of the report. TT, CGS and SJ coordinated the data management, and TT drafted the report. AS carried out the statistical analysis, reviewed by ÖD and JK. All authors approved the final version of the report. TT is the guarantor.

  • Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Ethics approval The study was approved by the research ethics committee in Linköping, Sweden.

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