Article Text

The psychological factor ‘self-blame’ predicts overuse injury among top-level Swedish track and field athletes: a 12-month cohort study
  1. Toomas Timpka1,2,3,
  2. Jenny Jacobsson1,2,
  3. Örjan Dahlström1,4,
  4. Jan Kowalski1,5,
  5. Victor Bargoria1,2,6,
  6. Joakim Ekberg1,2,3,
  7. Sverker Nilsson1,2,
  8. Per Renström1,7
  1. 1Athletics Research Center, Linköping University, Linköping, Sweden
  2. 2Department of Medical and Health Sciences, Linköping University, Linköping, Sweden
  3. 3Unit for Health Analysis, Centre for Healthcare Development, County Council of Östergötland, Linköping, Sweden
  4. 4Department of Behavioural Sciences and Learning, Linköping University, Linköping, Sweden
  5. 5Department of Clinical Science, Intervention and Technology, Pediatric Unit, Karolinska Institutet, Stockholm, Sweden
  6. 6Department of Orthopaedics and Rehabilitation, Moi University, Eldoret, Kenya
  7. 7Department of Molecular Medicine and Surgery, Center for Sports Trauma Research and Education, Karolinska Institutet, Stockholm, Sweden
  1. Correspondence to Professor Toomas Timpka, Department of Medical and Health Sciences, Linköping University, SE-581 83 Linköping, Sweden; toomas.timpka{at}


Background Athletes’ psychological characteristics are important for understanding sports injury mechanisms. We examined the relevance of psychological factors in an integrated model of overuse injury risk in athletics/track and field.

Methods Swedish track and field athletes (n=278) entering a 12-month injury surveillance in March 2009 were also invited to complete a psychological survey. Simple Cox proportional hazards models were compiled for single explanatory variables. We also tested multiple models for 3 explanatory variable groupings: an epidemiological model without psychological variables, a psychological model excluding epidemiological variables and an integrated (combined) model.

Results The integrated multiple model included the maladaptive coping behaviour self-blame (p=0.007; HR 1.32; 95% CI 1.08 to 1.61), and an interaction between athlete category and injury history (p<0.001). Youth female (p=0.034; HR 0.51; 95% CI 0.27 to 0.95) and youth male (p=0.047; HR 0.49; 95% CI 0.24 to 0.99) athletes with no severe injury the previous year were at half the risk of sustaining a new injury compared with the reference group. A training load index entered the epidemiological multiple model, but not the integrated model.

Conclusions The coping behaviour self-blame replaced training load in an integrated explanatory model of overuse injury risk in athletes. What seemed to be more strongly related to the likelihood of overuse injury was not the athletics load per se, but, rather, the load applied in situations when the athlete's body was in need of rest.

  • Athletics
  • Psychology
  • 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.


In the past few years, epidemiological evidence on injury rates, patterns and potential risk factors, have been established for athletics contestants, and track and field athletes. At international athletics championships, about 10 injuries occur per 100 participating athletes.1 The few studies available that cover an entire athletics season have reported that about two out of three athletes sustain at least one injury each year,2 ,3 and most of these injuries have been classified as from ‘overuse’.

The cause of overuse injuries is different from the mechanisms that lead to acute injuries. Overuse injuries result from the interplay between functional over-reaching, repeated exposure to microtrauma and the periodicity of recovery.4 ,5 There are several theories on how athletes experience and psychologically manage acute injuries including, for example, the stress and injury model,6 and various cognitive appraisal models.7 ,8 Psychological risk factors have been investigated for acute injuries.9–11 Regarding factors associated with successful return to sport after an acute injury, fear of relapse is a prominent emotional response at the time of transition back to sport.12 This research suggested that clinicians should address fear, autonomy and competence-related psychological factors in rehabilitation. The authors also recommended that routine screening of injured athletes during the rehabilitation was to identify those at risk of developing maladaptive psychological responses.12

There is increasing evidence that psychological factors may contribute to the processes that lead to overuse injuries.13 ,14 The affective adaptation framework,15 a relatively recent account of cognitive appraisal theory, is based on the premise that a person becomes cognitively aware of sensory information, such as perception of the body, predominantly if it is not expected. After a bodily sensation has been noticed, an explanatory process takes place in which the individual tries to make sense of the world. Here, explanation serves to suppress fear and anxiety.

If the sensation cannot be explained, an affective reaction is initiated, activating additional cognitive functions. The consequential thought patterns and behaviours may range from escape and avoidance, to increase and overuse.16 For example, an athlete can be frightened of experiencing a new type of bodily sensation not previously recognised. If the sensation is interpreted as non-threatening (eg, considered a temporary nuisance), the athlete typically resumes training, often after a period of diminished activity. The athlete will then test and correct explanations of the sensation, keeping them in line with actual experiences.17 ,18

On the other hand, denial can be used as a coping behaviour when perceiving pain or a new bodily sensation. An athlete's anxiety related to the new sensation can be relieved even by an incorrect explanation (‘This is probably a temporary nuisance’) of the pathological process that has begun. Habitually using denial or other maladaptive explanations of pain and other sensations increases the risk of long-term overuse conditions.19 This implies that psychological interventions that correct maladaptive thought processes may be useful in primary and secondary prevention programmes. For example, the psychological flexibility model is used for understanding and clinically adjusting maladaptive thought patterns in patients suffering from chronic pain.19 Psychological flexibility denotes a person's capacity to persist or change behaviour in a way that includes conscious contact with thoughts and feelings, in addition to having an appreciation of one's most important goals and values.20 ,21

Therefore, we examined the relevance of enduring psychological factors (defined below) associated with the affective adaptation framework in an integrated model of overuse injury risk in athletics. Enduring psychological factors are psychological characteristics that are stable over time, such as traits, as opposed to versatile psychological states, for example, mood. In short, we tested whether certain psychological elements appear to predict overuse injuries; if this were the case, it would open up another domain (psychological) to consider in sports injury prevention.


A cohort design of 52 weeks’ duration22 was used for the study involving Swedish male and female youth and adult elite track and field athletes. Individuals ranked in the national top 10 in each athletics event were invited to participate, regardless of their injury status. An overuse injury was defined as a condition to which no identifiable single external transfer of energy could be associated but that led to the athlete being unable to take full part in athletics training.23 A psychological model based on the affective adaptation framework was constructed to identify factors that interfere with day-to-day explanations and behaviours when athletes experience pain or other activity-limiting sensations. For the study, a gainful explanation of such sensations was defined according to the affective adaptation framework as “an explanation that includes taking body signals and affective responses earnestly, and optimally enlisting the help of others’ expertise in reaching explanations and regulatory behaviours as accurately as possible”. The risk indicators assessed were thus grouped into two categories (figure 1): enduring psychological/behavioural factors associated with the affective adaptation framework and psychological flexibility (body consciousness and hyperactivity, coping behaviours, perceived motivational climate and commitment to exercise), and standard epidemiological items (sex, age, athletics event, recent serious injury and training load).

Figure 1

Overview of the three models of overuse injury risk (epidemiological, psychological and integrated) used to structure the data collection and analyses in the study.


Ethical approval for the study was obtained from the Ethical Committee in Linköping, in November 2008 (dnr M-201–08). Informed written consent was obtained from all participants in the study before the baseline surveys were distributed. For those under 18 years of age, approval was also obtained from their parents.

Data collection

At the study baseline, in March 2009, the psychological data were collected by a postal survey. A web questionnaire was distributed in parallel to ask for personal sports-specific and sociodemographic data. The following four psychological/behavioural areas were covered by the survey.

Body consciousness and hyperactivity

Measurement of body consciousness was based on the Body Consciousness Scale (BCS).24 In order to also take tendency to hyperactivity into account, items from the hyperactivity definition in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) were added to form the BCS-HA questionnaire (21 items; the last six items addressed hyperactivity). Scores are recorded and computed separately for private body consciousness, public body consciousness, body competence and hyperactivity.

Coping behaviours

The practices that athletes used to understand and manage perceptions of pain were measured by the Brief Cope instrument.25 Adaptive behaviours were covered by items for active coping, emotional support, instrumental support, positive reframing, planning, mood, acceptance and religion; the maladaptive behaviours were covered by the items self-distraction, denial, substance, behavioural disengagement, venting and self-blame.

Perceived motivational climate

Perceived demands of the social sporting environment were measured using the Perceived Motivational Climate in Sport Questionnaire (PMCSQ). The PMCSQ records two separate scores for the mastery/task accomplishment and performance orientations.26

Commitment to exercise

Tendencies to rigidity in implementing training was measured by the Commitment to Exercise Scale (CtES).27 This is an eight-item questionnaire designed to assess an individual’s psychological commitment to the activity of exercising.

Injury and training data were collected using web-based questionnaires (SiteVision V.2.5, Senselogic AB, Örebro, Sweden).22 Injuries were defined as recurrent if they occurred in the same location, were of the same type and took place within 2 months of a previous injury. Further injuries occurring after a first injury were defined as subsequent injuries. In this study, no attempt was made to classify recurrent and subsequent injuries in detail. The data were self-reported by the athletes; assistance from parents was promoted for athletes under 18 years of age. An injury was defined as any new musculoskeletal pain, soreness or injury that resulted from athletics training or competition and caused changes in the mode, duration, intensity or frequency of normal training/competition from the current or subsequent training and/or competition sessions. Only injuries occurring while participating in athletics, training or competition were included.24 Emails were sent weekly to participants with questions about the preceding week; one reminder was sent to those who did not respond to the first email.

The weekly training load was quantified by combining training hours and intensity on a relative basis for a 6-week period in association with the start of the study. A training load rank index (TLRI) defining the relative training load was constructed by first multiplying the reported training intensity (light=2, moderate=3, hard=5) with minutes of training performed during the week.3 The athletes were grouped by athlete category and event group, and then ranked and separated into quartiles by their training load score into TLRI categories Q1–Q4.

Statistical analysis

All data were first presented using descriptive statistics, namely, mean, median, minimum and maximum for continuous data, and frequencies and proportions (%) for categorical data. The Mann-Whitney U test was used to analyse differences between youth and adult athletes with regard to baseline scoring of the psychological instruments. The ensuing analyses were performed at the level of athletes suffering and not suffering an overuse injury during the surveillance period.

Data from the participating athletes were entered into the analyses from the first day that they were free of injury and in normal training. At baseline, athletes identified as being injured were left censored until the week after they reported being back in normal training after injury. Participants discontinuing their study participation or suffering traumatic injuries were right censored at the time of that event. Cox proportional hazards analyses using single explanatory variables were first performed to study differences in overuse injury risk with regard to athlete category (combining gender and age group), event group, injury history, number of training hours per week, weekly training intensity, the TLRI, perceived motivational climate (PMCSQ), coping behaviours (Brief Cope), body consciousness and hyperactivity (BCS-HA) and commitment to exercise (CtES).

As previous studies found that injury risk was associated with athlete category and previous injury,2 ,3 we also tested for interactions with combinations of these factors. Thereafter, Cox proportional hazards models using several explanatory variables (multiple models) were compiled using time to overuse injury as the endpoint and the risk indicators that showed statistically significant associations with injury risk in the previous analyses as eligible explanatory variables.

The analyses were run using a forward stepwise procedure where, at each step, the variable with the highest explanatory power (ie, lowest p value based on the Wald statistic) was added to the model until no more variables could be added with p<0.05 (final models were also compared with results using backward stepwise procedures to assure identical models).

Three types of multiple models (including several explanatory variables) were compiled: (1) an epidemiological model without the psychological variables, (2) a psychological model excluding the epidemiological variables and (3) an integrated model including all study variables.

All tests were two-sided and p<0.05 was regarded as statistically significant. The calculations were carried out using SPSS V.18 or higher (IBM Inc).


Among the 278 athletes entering the surveillance study with complete baseline data, 96% (n=266) submitted complete 12-month data sets and 78% (n=216) also completed the psychological survey (figure 2). There were significantly more males (p=0.014) and adults (p=0.006) among those athletes not completing the psychological survey compared with the survey responders. There was no meaningful difference with regard to having sustained a severe injury the previous year between survey completers and non-completers (p=0.34).

Figure 2

Overview of study process and populations (F, female; M, male).

The mean age of the adult athletes completing the 12-month surveillance was 24 years (range 18–37) and the mean age of the youth athletes was 17 years. There were no differences between youth and adult athletes with regard to perceived motivational climate or commitment to exercise. Regarding body consciousness, youth athletes scored lower than adult athletes on the public body consciousness subscale of the BCS-HA (p<0.001), suggesting a lower tendency among the youth athletes to be concerned with the public appearance of their bodies. Adult athletes were more likely than youth athletes to use planning (p<0.001) and active coping (p=0.011) as coping behaviours, while the youth athletes were more prone to maladaptive strategy behavioural disengagement (p=0.042). No other differences regarding body consciousness or coping behaviours were observed between youth and adult athletes at baseline (see online supplementary table S1).

Seventy-one per cent of the athletes reported an injury event during the study period. The median time to the first reported injury event among those injured was 41 days (mean 72 days; 95% CI 60 to 83). Ninety-seven per cent of reported injuries were classified as associated with overuse. The interactions between athlete category and previous injury, the TLRI, and the psychological variables self-blame and religious beliefs were all found to be individually associated with injury risk (see online supplementary table S2).

Epidemiological multiple model

The epidemiological multiple model comprised the TLRI (p=0.033), and an interaction between athlete category and injury history (p<0.001; table 1). Athletes in the third (p=0.013; HR 1.76; 95% CI 1.13 to 2.76) and fourth (p=0.007; HR 1.81; 95% CI 1.18 to 2.80) TLRI quartiles had almost twice the risk of overuse injury compared with their peers in the first quartile, and youth male athletes who had sustained a severe injury the previous year presented a more than twofold increased risk (p=0.005; HR 2.57; 95% CI 1.34 to 4.92) of sustaining a new injury compared with adult males with no previous severe injury.

Table 1

Risk indicators included in the epidemiological multiple model compiled by the Cox proportional hazards method (n=266)

Psychological multiple model

The multiple model restricted to psychological risk indicators included variables accounting for coping behaviours (table 2). The maladaptive coping practice of self-blame (p=0.018; HR 1.27; 95% CI 1.04 to 1.55) was found to be associated with increased risk.

Table 2

Risk indicator included in the psychological multiple model compiled by the Cox proportional hazards method (n=216)

Integrated multiple model

The integrated multiple model included an interaction between athlete category and history of serious injury (p<0.001; table 3). The risk of sustaining a new overuse injury during the prospective year was reduced to less than half compared with the reference category (adult male athletes with no previous injury) for youth female (p=0.034; HR 0.51; 95% CI 0.27 to 0.95) and youth male (p=0.047; HR 0.49; 95% CI 0.24 to 0.99) athletes with no severe injury the previous year. The maladaptive coping behaviour self-blame (p=0.007; HR 1.32; 95% CI 1.08 to 1.61) remained associated with increased risk for overuse injury. The TLRI was not found to be statistically associated with injury risk in the integrated model.

Table 3

Risk indicators included in the integrated multiple model compiled by the Cox proportional hazards method (n=216)


We examined the relevance of enduring psychological factors associated with the affective adaptation framework in an integrated model of overuse injury risk in athletics. A previous analysis that did not include psychological factors identified two risk indicators: training hours and intensity, jointly measured by the TLRI, and an interaction between athlete category (defined by age and gender) and recent injury history.3 In this study, we found that the TLRI was replaced in an integrated explanatory model by a maladaptive coping behaviour: self-blame reflecting negative thinking. The second previously identified risk indicator (interaction between athlete category and injury history) remained in the integrated model; youth athletes who had not suffered a severe injury the previous season were at a decreased risk of sustaining a new injury episode.

Maladaptive coping behaviour, self-blame, predicted overuse injury

In the integrated explanatory model developed in this study, the maladaptive coping behaviour self-blame was a stronger predictor of overuse injury than self-reported training behaviours represented by the TLRI. In other words, what seemed to matter in overuse injury causation was not the athletics load per se, but the load applied in situations when the athlete's body was in need of rest and restoration. This observation suggests that enduring psychological factors that disturb adequate interpretations of body perceptions or balanced behavioural responses to these perceptions should be included when musculoskeletal overuse injury risk in track and field athletes is regarded in clinical and research settings. More specifically, our study indicates that overuse injuries in athletics are partly the consequences of inadequate self-perception among athletes.

Self-perception involves detecting and accurately utilising relevant cues about how one typically feels, thinks and behaves.28 Self-perception can be distorted by two fundamental motives: a self-enhancement motive (ie, the desire to perceive the self positively) and a self-verification motive (ie, the desire to confirm one's identity).29 Applying this reasoning to the present results, top-level track and field athletes regard themselves as being able to challenge the national record in their event. Athletes who use self-blame to cope with failure to reach such performance goals may be more prone to suffer frustration and this may lead to a vicious cycle characterised by negative thinking, undue acceptance of pain and irrational task persistence in training.30 ,31

We noted that a recent review12 did not include maladaptive coping behaviours among the psychological factors associated with return to sport following injury. Instead, positive psychological responses in terms of motivation, self-esteem and autonomy-need satisfaction promoted a greater likelihood of return to the preinjury level of participation in that systematic review. These observations suggest that the associations between perceptions of pain and other bodily sensations and the psychological/behavioural responses may differ between the postacute injury and preoveruse injury contexts.

Nonetheless, in both settings, to reinitiate or change a training activity in the presence of pain or other activity-limiting sensations is something that track and field athletes need to manage on a regular basis. The affective adaptation framework draws attention to what information is utilised in this decision-making process and the conditions under which decisions are made. The framework can also be used to identify factors that have negative influence on the decision-making and for selection of psychological interventions to prevent such influence. This study showed that self-blame may be one potential influence of this kind.

Relevance in the sport of track and field

The last decade has ushered in rapid development and dissemination of evidence-based therapies for mental illness and psychological problems.32 For prevention of overuse injuries, similar interventions can also be provided to track and field athletes. In the psychological flexibility model,19 ,20 for instance, such interventions are divided into six areas:

  1. ‘Acceptance’, where the athlete is supported to open up to unwanted experiences when to do this will serve his or her goals. An example in the self-blame context is when a long-distance runner attributing a recent championship failure to not having pushed himself hard enough during preseason training, accepts that he perceives disquieting Achilles tendon pain while preparing for the next major competition, abstains from pushing through further high-intensity workouts, and consults his physiotherapist.

  2. ‘Cognitive defusion’ denotes the ability to make a distinction between thoughts and the things they describe. For instance, a 400 m sprinter can learn to isolate maladaptive thoughts (such as “I do not deserve to be selected for the relay team because I do not push myself through pain as the others do”) as ‘just thoughts’ and thereby avoid submitting to inflexible and potentially harmful behaviours.

  3. ‘Flexible present-focused attention’ is also called mindfulness.

  4. ‘Self-as-observer’ addresses the ability to attain a standpoint where the athlete is not harmed by own thoughts and feelings.

  5. ‘Goals and values’ targets desires or qualities that the athlete defines as the most important.

  6. ‘Committed action’ is where the athlete is promoted to persist with a course of action guided by important personal goals and values.

Limitations and strengths

The present study has several limitations. We had a 74% response rate to the baseline psychological survey and fewer than expected males and adults responded. Higher research participation rates among females has also been reported from other settings,33 while the higher participation rate among the youth athletes may be explained by parental influence. Nonetheless, there was no difference between responders and non-responders with regard to having sustained a severe injury the previous year. Except for the hyperactivity subscale included in BCS-HA, which reflects each item in the hyperactivity section in DSM-IV, all psychological instruments used in the study had previously been validated and were used according to their respective protocol. The hyperactivity subscale was developed for this study in cooperation with clinical psychologists working with client populations in late adolescence and young adulthood. Before use in the survey, the subscale was tested for face validity and informally for content validity. No formal tests of content or construct validity of the subscale were performed.

Also, it should be remembered that all psychological instruments used in the study had been developed and validated in adult populations, and their specific validity among adolescents 16–17 years of age had not been established. When comparing response patterns for youth and adult athletes, we observed no or small differences with regard to perceived motivational climate, commitment to exercise and coping behaviours. However, youth athletes scored lower than adult athletes on public body consciousness, possibly indicating a progress of this body awareness aspect. The lack of psychological instruments validated for populations in late adolescence and young adulthood is a general problem, and this issue is also a weakness of the present study.

Moreover, being one of the first epidemiological investigations of its kind, the choice of theoretical model, research instruments and psychological constructs may not have been optimal for studies of overuse injury in athletics. Here, besides the more enduring psychological factors included in this study, versatile mood states may also play a role. The Mood-as-Input (MAI) model34 has been suggested to provide an account of both persistence and avoidance behaviour in a pain setting,35 particularly taking into consideration the influence of mood status.36 Task persistence is seen as the result of current mood interacting with stop rules adopted when performing open-ended tasks.34 Accordingly, when adopting an achievement-oriented stop rule, elite athletes may ask themselves whether they have performed well enough, and they will persist in a given task as long as the answer is negative. A negative mood may then facilitate task persistence because it signals to the athlete that the performance goals have not been reached yet. In further research on overuse injury predictions using psychological factors, inclusion of versatile states, such as mood, in the models is thus warranted.


We examined the relevance of enduring psychological factors derived from the affective adaptation framework in an integrated explanatory model of overuse injury risk in athletics. The maladaptive coping behaviour self-blame replaced a training load index in the predictive model where both psychological and standard epidemiological factors were represented. The observed associations between athlete categories, coping behaviours and overuse injury risk, can be used to inform further epidemiological studies and rigorously designed trials of psychological interventions endorsing primary and secondary prevention of overuse injuries among track and field athletes.

What are the findings?

  • The coping behaviour ‘self-blame’ recorded by the Brief Cope instrument replaced a training load rank index in an integrated explanatory model of overuse injury risk in a cohort of top-level Swedish athletics contestants, and top-level Swedish track and field athletes.

  • The replacement of a training load index by a habitual maladaptive coping behaviour in the explanatory model suggests that what predicts overuse injuries may not be the athletics load per se, but, rather, the load applied in situations when the athlete's body is in need of rest.

  • A psychological characteristic added explanatory value to integrated prediction model for overuse injury risk in athletics.

How might it impact on clinical practice in the future?

  • Track and field athletes who display self-blame as coping behaviour may be at increased risk of injury.

  • Interventions that promote psychological flexibility in response to self-blame and injury should be considered when planning programmes for rational clinical management of overuse injury among track and field athletes.

  • We encourage researchers who assess overuse injury risk in track and field athletes to include measures of baseline psychological characteristics as well as monitoring training load, injury symptoms and mood state changes.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors TT and JJ made substantial contributions to conception and design of the project, and the data collection, provided directions for the analysis of data, drafted and revised the article and provided final approval of the version to be published. ÖD and JK analysed the data. ÖD, JK, VB, JE, SN and PR provided contributions to the study design, data collection and data analysis and they critically revised the article for important intellectual content and provided final approval of the version to be published.

  • Funding This study was supported by a research grant (P2014-0167) from the Swedish Center for Sports Research (CIF).

  • Competing interests None declared.

  • Ethics approval Regional Research Ethics Board in Linköping, Sweden.

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

Linked Articles

  • Warm up
    Mats Börjesson Magnus Forssblad Jón Karlsson