Developing a simple risk metric for the effect of sport-related concussion and physical pain on mental health.

Depression has become an increasing concern across sport. This may be due to prevalence rates ranging from 21% in the USA [ 1 ] and 43% in the UK [ 2 ] in young adults that visibly exceed the 3.8% of the general population that suffer from the disorder [ 3 ]. Several risk factors are commonly reported, including biological sex [ 2 , 4 ], physical pain [ 5 – 7 ], and sport-related concussion [SRC; 8 – 10 ]. Despite these links, it is continuously reported that athletes will accept the risk of experiencing physical pain [ 11 – 13 ] and sustaining SRC [ 14 ] when competing. This may be due to the complexity of these issues, in that athletes may report that they understand the risks, but in fact do not. Becker [ 15 ] reported that athletes respond better to simple stimuli from their coaches rather than complex. Therefore, a similar approach such as presenting a simple risk metric should be adopted when attempting to inform athletes of the dangers of SRC and physical pain. The influence of biological sex on depressive symptoms is difficult to pin-point. Wolanin et al. [ 4 ] found that females are twice as likely to experience depression at some point in their lifetimes than males. Reasons for this include biological differences [ 16 ], environmental factors [ 17 , 18 ], stress response [ 19 , 20 ] and self-esteem [ 21 ]. Despite this males are more likely to die by suicide [ 22 – 25 ] which questions our understanding of male mental health and coping strategies. This comes in a period where there are many large initiatives promoting male mental health support [ 26 – 28 ] such as Movember by the Movember Foundation and the Best Man Project by the Campaign Against Living Miserably charity. Therefore, the role of biological sex is one that ought to be continuously investigated to further uncover mental health in men and women. Many athletes and non-athletes [ 29 ] live daily with acute physical pain. From a sporting perspective, physical pain is often reported as a risk factor for depressive symptoms in sportspeople [ 2 , 6 , 7 ] due to the prevalence of increased pain within contact sports [ 30 , 31 ]. Athletes from contact sports regularly continue through the pain barrier [ 11 ] perhaps due to a culture of being perceived as weak or in fear of the prospect of losing their position in the team during their absence [ 32 ]. Even in non-contact sports physical pain is likely as athletes often compete with overuse injuries [ 7 ] such as repetitive strain [ 33 ]. However, physical pain also impacts non-athletes. There are conditions that people live with such as fibromyalgia [ 34 , 35 ] and endometriosis [ 36 , 37 ] that leave the sufferer in high levels of physical pain exposing them to depression, despite not being athletes. It is also normal for people to experience everyday pain such as headache [ 38 ] or musculoskeletal pain [ 39 , 40 ]. With the knowledge that athletes and non-athletes experience physical pain, and that this pain is associated with poorer mental health [ 6 ], we must protect them from physical pain where possible. Non-athletes that suffer from musculoskeletal pain or headaches may not have engaged in behaviours that encouraged this. However, athletes do engage in activity that promotes and exacerbates physical pain. Therefore, as they are engaging in physical activity that they believe to be positive for their physical [ 41 , 42 ] and mental health [ 43 – 45 ], they ought to be aware of the potential consequences to their mental health, by exposing themselves to the risk of physical pain so an informed decision can be made regarding their participation. Another reason for increased physical pain in athletes could be sustaining SRC. Research into this area was primarily interested in cognitive impairments [ 46 – 48 ], before a focus on mental health derived [ 49 – 51 ]. Didehbani et al. [ 49 ] found that affective scores obtained using BDI-II were significantly higher for athletes with a history of concussion compared with matched athletes with no history of concussion ( M = 1.59 vs. 0.38). Research like this has promoted examining mental health in those that are likely to sustain SRC, with depression continuously linked [ 8 – 10 ]. As with physical pain, the attitudes of many sportspeople towards sustaining SRC are harmful, and many perceive head injury as a risk that they are willing to take to succeed in their sport [ 14 ]. However, it is unlikely that these athletes are truly aware of the danger of SRC and the likelihood of developing depression following this type of impact. Therefore, as with physical pain, it is necessary to identify the likelihood of developing depression following SRC and informing sportspeople of the danger that SRC has to their mental health. Given what we know about the influence of physical pain and SRC on depression, it is plausible to suggest that the same risk is not applicable to all sports. It is reasonable to predict that we may expect a greater chance of sustaining SRC and experiencing elevated levels of physical pain in contact sports such as rugby than in non-contact sports such as tennis. However, it is equally important to understand the likelihood of developing depression in non-contact sports and contact sports and in non-athletes, given that physical pain is still present in all groups and engaging in physical activity has been found to alleviate depressive symptoms [ 43 – 45 ]. It is important to understand the influence of sport-type on depression, as this will corroborate or undermine the influence of SRC and physical pain on this outcome. Though there are numerous studies indicating that females are more likely to be depressed than males, the disparity in suicide cases suggests continued research into the role of biological sex on depressive symptoms is required. Physical pain and SRC are continuously found to be linked with depression, however, a simple risk metric is required to illustrate how much risk athletes are at if experiencing physical pain and/or have sustained SRC. Communicating this will allow sportspeople the opportunity to make an informed decision as to whether they would like to take the risk of competing. The present study therefore aimed to provide such a metric by exploring the likelihood of developing depression based on biological sex, physical pain scores, SRC history, with sport-type included to corroborate or contradict SRC and pain findings.

Seventy-six (52.8%) participants scored ≥16 and were therefore included in the depressed group for logistic regression whereas the remaining 68 (47.2%) participants that scored < 16 on the CESD were included as the non-depressed group. It is also important to note that 47 (92%) of those that took part in contact sports had previously sustained concussion compared with 7 (36%) of those that took part in non-contact sports and 8 (11%) of participants that took part in no sport at all. Scores of physical pain and depression by sport-type are presented in Table 2 . Binary logistic regression was performed to assess the impact of a set of predictor variables on the odds that respondents would report depressive symptoms. Meaningful depressive symptoms were operationalised as scoring ≥16 on the CESD [ 54 ]. The model contained 4 independent variables (sex, physical pain, SRC history and sport-type). Sport-type was either non-sport for non-athletes or contact or non-contact sports for athletes. The full model containing all predictors was statistically significant χ 2 (5, N = 144) = 41.61, p = .000, indicating that the model was able to distinguish between respondents whose CESD scores were below or above the cut-off score of 16. The model correctly classified 70.8% of cases. As shown in Table 3 , all four predictors made a unique statistically significant contribution to the model. The strongest predictor of depressed categorisation was having sustained SRC with an odds ratio of 56.98, indicating that the likelihood of being depressed is almost 57 times more likely than those that have not sustained SRC. Respondents were also 1.4 times more likely to be depressed for every score of physical pain they reported, and females were 2.9 times more likely to be depressed than males. However, those that competed in sport had a lesser likelihood of developing depression than those that did not take part in sport, regardless of sport-type (Contact sports = 71 times less likely, Non-contact sports = 4.4 times less likely). *- Significant at p < .05 **- Significant at p < .01 ()–Calculated to indicate lesser likelihood

The main purpose of this study was to provide simple risk metrics for sportspeople, to aid in their decision making process of competing in their sports. From the present study we can aid athletes with this decision by informing them that they are 1.38 times more likely to suffer from depression for every score of physical pain they experience (0–10), and that they are 57 times more likely to be depressed if they sustain SRC. This simple risk metric will help sportspeople make an informed decision on whether they want to take part in their sport, with them aware of which type of sports are prone to these factors. We also provide evidence that females are nearly three times more likely to be depressed than males, and therefore the impact of SRC and physical pain may be more pronounced in females. This has been previously reported with physical pain [ 2 ] but not with SRC.

The present study provides simple risk metrics for the likelihood of developing depression following SRC or when in physical pain. To our knowledge, we are the first to calculate these and therefore the present findings add value to the literature. It is vital to continue researching the role of biological sex on mental health disorders, such as depression, as females are often found to be at greater risk [ 2 ], while it appears that males may conceal this personal information. Continuing to examine this area allows us to support and protect vulnerable sportspeople. This study, however, still suggests that engaging in sport and physical activity is beneficial for mental health. Therefore, sportspeople should not be deterred from engaging in sport and physical activity but should be aware of the simple risk metrics developed in this study regarding SRC and physical pain to make an informed decision on their participation.

There is a risk of self-selection bias due to the voluntary nature of our recruitment and the way the study was advertised, though this is common in the literature [ 2 ]. Another limitation of the present study is regarding sport-type SRCs. What constitutes contact sports and non-contact sports is debatable but there are not many that would dispute what category a certain sport belongs to. However, there are instances whereby even when the sport is defined as non-contact due to the laws of the game, SRC can still occur. For example, cycling would be considered a non-contact sport for many but if falling during a sprint the chance of sustaining SRC increases. In the present study, touch rugby is defined as a non-contact sport but there are a lot of SRCs that have occurred. This is likely a reflection of participants only reporting their predominant sport at the time of participating, and therefore it is not possible to truly determine where SRC was sustained. In future, studies should ask participants to not only report their predominant sport at time of taking part, but also what sports they were taking part in when SRC occurred, as this could be different, or there could be SRCs across multiple sports. Including this would amend for potential misclassifications that may be present in this study. Additionally, regardless of which sport participants predominantly took part in, the present study utilised a self-report technique regarding SRC history. Without diagnoses, it is possible that the total number of SRCs are inflated in this study, as can be viewed in Appendix B in S1 Appendix , and therefore the reader should take this into consideration when inferring the results. However, concussion is a lived experience and therefore we would argue that if an athlete is reporting having sustained concussion then it is likely because they have experienced concussion symptoms. As we know that not all concussions present symptoms, it could be that there is in-fact an underreporting of total number of SRCs. Therefore, although there is less control here than if we had sought clinical diagnoses of concussion, there is still value in trusting the athletes lived experience of what they deem concussion to be.

A convenience sample of 144 participants (Age, M = 22.79, SD = 5.61) was obtained consisting of 68 males (Age, M = 24.41, SD = 5.50) and 76 females (Age, M = 21.34, SD = 5.33). Participants were recruited via the online departmental recruitment system within the university, advertisements on social media platforms LinkedIn and Twitter as well as word of mouth. Men and women over 18 years of age were welcomed to take part in the study. All participants self-reported whether they had sustained SRC (yes/no) and if so, how many SRCs have they sustained in total and how many months have passed since the latest one (free-text responses). A question asking whether they had sustained concussion away from sport was also asked, with all participants responding that they had not. No formal diagnoses were collected in this study. For those that had sustained concussion, a minimum of 28 days must have passed before taking part in this study to avoid exacerbating post-concussion symptoms due to the surveys being displayed on a computer screen. Those that had sustained concussion < 28 days were ineligible to take part until this period had elapsed. There were 74 participants that did not take part in any sport at all, 19 that competed in different non-contact sports (athletics, netball, squash, touch rugby, cricket, baseball, weightlifting, handball, cycling, equestrian, badminton, and dance), and 51 that took part in contact sports (rugby union/league, football, skiing, boxing, and taekwondo). A more in-depth break down is presented in Appendix A in S1 Appendix . Sixty-two participants reported having sustained SRC in the past four years (months since last concussion– Range , 1–48, M = 18.87, SD = 14.01), totalling 223 SRCs between them ( Range , 1–12, M = 3.60, SD = 2.80). Appendix B in S1 Appendix presents the number of SRCs by sport-type and non-sport. These three factors were used for data analysis alongside scores of physical pain. Sample size calculations were calculated post-hoc using G*Power 3.1.9.7 software. Logistic regression using 144 participants provided excellent power (β = 0.97) to detect an OR of 3 and effect size of f 2 = 0.5 (large effect size) at α = 0.05. A large effect size was used in calculations due to the time difference between certain dependent variables (SRC history being 28+ days ago compared with physical pain in the past week). General Information Questionnaire (GIQ) : included data on biological sex, age, sport-type (non-sport, non-contact sport, contact sport), SRC history, and physical pain experienced in the past week. Pain was measured using the Numeric Rating Scale-11 [NRS-11; 52 ] providing a score ranging 0–10. Centre for Epidemiological Studies Depression Scale [CESD; 53 ] : This tool contains 20-items that measure depressive symptoms over the past week using a four-point Likert scale (0–3). 0 = ‘rarely or none of the time’ (less than once a week) 1 = ‘some or a little of the time’ (1–2 days a week) 2 = ‘occasionally or a moderate amount of time’ (3–4 days a week) 3 = ‘most or all of the time’ (5–7 days a week) Four items were reverse coded due to the nature of the question. A total score ≥16 of 60 represents the respondent may be experiencing some form of depression [ 54 ] and a higher total score reflected more severe depressive symptoms. Cronbach’s alpha analysis revealed an internal consistency score of α = .94 which is considered excellent [ 55 ]. Participants completed the questionnaires on the online survey platform Qualtrics (Qualtrics, Provo, UT) and were fully informed about the study with their consent obtained prior to participation. This was followed by the participants completing the study questionnaires (GIQ and CESD). British Psychological Society (BPS) ethical guidelines were adhered to with data collection commencing after ethical approval was obtained from the University’s Departmental Research Ethics Committee (DREC). A participant information sheet informed participants of the nature of the study and their rights as a participant including details on the withdrawal of data if they wished to do so. All participants were 18 years or older at the time of completing the study and were able to provide informed consent. Consent was obtained via tick box options on the consent form that had validation options set disallowing participants to continue if they did not provide consent. Participants developed a unique ID following this page which was only used if participants wished to withdraw their data, making their data identifiable for the researcher to do so. A debrief form was displayed reiterating the aims of the study. Binary logistic regression investigated the odds ratios of the four independent variables (sex, sport-type, physical pain and SRC history) on the likelihood of depressed categorisation (scoring ≥16 on CESD). It was predicted that being male, engaging in contact sport, in physical pain and having sustained SRC would increase the chances of depressed categorisation. There are some assumptions that are required for logistic regression to provide a valid result [ 56 ]. Firstly, the dependent variable ought to be binary. In this study, depressed categorisation (depressed/non-depressed) was the dependent variable, and therefore this serves as a binary dependent variable and satisfies this assumption. We have justified our sample in sample size calculations and therefore this assumption is also satisfied. Finally, logistic regression requires little to no multicollinearity among independent variables. Correlation analysis displayed in Table 1 shows independent variables are not highly correlated with one another. This is due to the general rule of thumb that correlation coefficients between two variables are less than 0.9 [ 57 ], of which all are in our analysis. *- p < .01

(DOCX) Click here for additional data file. (DOCX) Click here for additional data file.