Contact or Collision Sport History, Repetitive Neurotrauma, and Patient-Reported Outcomes in Early to Midadulthood

Participants were recruited via word of mouth, local sport and recreation groups, email, and printed flyers. We recruited a total of 113 adults (53 [46.9%] men, 60 [53.1%] women; age = 34.88 ± 11.80 years), representing 4 distinct groups with various RHI histories and physical activity status and described elsewhere (Table 1).²²  The 4 groups represented physically inactive (NON) and active (NCA) individuals with no RHI exposure through contact or collision sports and physically active individuals with either previous RHI exposure through contact or collision sports (HRS) or prolonged RHI exposure through rugby participation beyond college (RUG). In addition, the groups self-reported as physically active (yes or no) based on the definition in the American College of Sports Medicine²³  physical activity guidelines (150 min/wk of moderate or 60 min/wk of vigorous physical activity). Briefly, the NON (non-RHI exposed, inactive) group represented physically inactive individuals without a history of playing contact or collision sports. The NCA (non-RHI exposed, physically active) group represented physically active individuals without a history of contact or collision sport participation. The HRS (RHI exposed through high-risk sports, physically active) group represented physically active individuals who previously participated in contact or collision sports with RHI exposure, but ceased participation at age 22 years. Lastly, the RUG (prolonged RHI exposure through rugby, physically active) group represented physically active individuals who reported participating in rugby beyond college but had since ceased participation. Volunteers were excluded if they self-reported current pregnancy; a concussion within 6 months of enrollment; a history of stroke or neurodegenerative disease; or any preexisting neurologic, balance, hearing, vestibular, or ocular disorder. All participants provided oral and written informed consent in accordance with the Institutional Review Board of the University of Delaware, which approved the study.

Participants completed online questionnaires (Qualtrics) consisting of relevant descriptive variables, physical activity status (yes or no according to the American College of Sports Medicine²³  guidelines), sport history, and contact or collision sport career duration (sum of all years played of each contact or collision sport). They also completed various measures of mental health and patient-reported outcomes via Qualtrics. We chose these patient-reported outcomes because they have been well used in the concussion and RHI literature, and some have shown group differences in later life among rugby and football players.^5,18,21,24 

The 12-Item Short-Form Health Survey (SF-12) consists of 12 questions related to one’s functional health and well-being. Scores are grouped into SF-12 Physical Component Summary (PCS) and Mental Component Summary (MCS) subscales reflecting perceived physical and mental health, respectively.²⁵  Scores range from 0 to 100, with higher scores indicating better quality of life and a score of 50 ± 10 representing the average American.²⁵ 

The Apathy Evaluation Scale–self-rated version (AES-S) consists of 18 questions on self-reported apathy in the last 4 weeks.²⁶  Scores range from 18 to 72, with higher scores indicating worse apathy and scores ≥34 signifying clinically meaningful apathy.²⁶ 

The Satisfaction With Life Scale (SWLS) is a 5-statement questionnaire in which individuals rate life satisfaction from 1 (strongly disagree) to 7 (strongly agree).²⁷  Scores range from 5 (lowest) to 35 (highest satisfaction with life), with a score <20 considered dissatisfaction with one’s life.²⁷ 

The Sport Concussion Assessment Tool–5th Edition (SCAT5) Symptom and Symptom Severity Checklist is a 22-item checklist in which individuals rate their symptoms based on how they typically feel on a scale from 0 (none) to 6 (severe).²⁸  Symptoms are reported as the total number of symptoms, with a rating ≥1 indicating the presence of that symptom, for a total symptom count ranging from 0 (none) to 22 (every symptom present). Symptom severity is the summed ratings of each symptom, with a possible symptom severity burden ranging from 0 to 132; a larger number indicates a greater symptom burden and greater severity.²⁸ 

We used either a 1-way analysis of variance to compare the 4 groups or a Mann-Whitney U test to compare the contact or collision groups on all descriptive data (Table 1). Data were tested for normality using the Shapiro-Wilk statistic and violated the assumption of normality (Shapiro-Wilk value < .05). To test our hypotheses, we first compared groups in an unadjusted model (model 1) across all patient-reported outcomes (ie, SF-12 PCS and MCS, AES-S, SWLS, SCAT5 symptom count and symptom severity) using a 1-way analysis of variance. For outcomes that were different, we performed a pairwise comparison procedure and Bonferroni correction for multiple comparisons. We used χ² analysis to compare the proportions of participants in each group meeting the clinical cutoff scores for the SF-12 PCS and MCS, AES-S, and SWLS. Outcomes were followed up with a post hoc z test of the adjusted residuals with the Bonferroni correction to determine which groups differed from each other.

To control for potential contributors to model 1, we compared the groups in model 2 using an analysis of covariance, adjusting for the known covariates of age, sex, and concussion history that may affect patient-reported outcomes.^8,10,29  For outcomes that were different, we performed a post hoc analysis using the Bonferroni test for multiple comparisons. Effect sizes for each test are reported as η² for unadjusted models or partial η² for adjusted models and interpreted as small (0.01), medium (0.06), or large (0.14).³⁰ 

To investigate the relationship between career duration and patient-reported outcomes, we conducted a linear regression using the enter method to determine the association between all patient-reported outcomes in collision sport athletes (ie, HRS and RUG groups) and contact or collision sport career duration (years). For follow-up analyses, we used a multiple linear regression to evaluate the association between patient-reported outcomes and career duration and known covariates: age (years), sex, and concussion history. The test assumptions of independence of observations, linearity, homoscedasticity, multicollinearity, and normal distribution of the residuals were all confirmed. The α level was set at .05, and all analyses were performed using SPSS (version 28; IBM Corp).

Mean, SD, range, and 95% CI for each physical and mental health patient-reported outcome (ie, SF-12 PCS and MCS, AES-S, SWLS, SCAT5 symptom count and symptom severity) are presented in Table 2.

In the unadjusted model, we found a difference between groups for the SF-12 PCS (F_3,106 = 4.143, P = .008) with a post hoc difference identified between NCA and RUG (adjusted P = .01) and NCA and NON (adjusted P = .04), whereby the SF-12 PCS was higher (ie, better) in the NCA group than in both the NON and RUG groups (Figure 1A). We observed no group differences for the SF-12 MCS (F_3,106 = 0.702, P = .55; Figure 1B), total symptom count (F_3,106 = 0.824, P = .48), or symptom severity (F_3,106 = 0.974, P = .41; Table 3).

A difference existed in the unadjusted model between groups for the AES-S (F_3,106 = 3.601, P = .02), but no post hoc differences were evident (P > .05; Figure 1C;Table 3). Lastly, we noted a difference between groups for the SWLS (F_3,106 = 4.161, P = .008), with post hoc differences between HRS and NON (adjusted P = .03) and NCA and NON (adjusted P = .03), such that the NON group had a lower (ie, worse) SWLS outcome than the NCA and HRS groups (Figure 1D). These results are presented with the abnormal scores shaded in grey in Figure 1A–D.

After adjustments for age, sex, and concussion history, a slightly attenuated main effect of group for the SF-12 PCS with a medium effect size (F_3,106 = 3.530, P = .02; η² = 0.091) was found with post hoc differences between the NCA and NON (P = .03), whereby the NON group had lower (ie, worse) self-rated physical function (Table 3). Only group was a contributor to the model (P = .02); age (P = .09), sex (P = .47), and concussion history (P = .13) were not. We demonstrated no group differences for the SF-12 MCS (F_3,106 = 1.349, P = .26).

A group difference (F_3,106 = 4.230, P = .007) with medium effect size (η² = 0.107) was seen for the AES-S with post hoc differences characterized, such that the NON group had a higher (ie, worse) score than the NCA (P = .03) and HRS (P = .040) groups. Both sex (P = .01) and group (P = .007) were contributors to the AES-S adjusted model. Further group differences were observed for the SWLS (F_3,106 = 4.354, P = .006) with a medium effect size (η² = 0.110), whereby the NON group had a lower (ie, worse) SWLS score than the NCA (P = .03) and HRS (P = .03) groups; only group was a contributor to the adjusted model (P = .006). Lastly, no group differences occurred for symptom count (F_3,106 = 0.953, P = .42) and symptom severity (F_3,106 = 0.851, P = .47).

None of the models were associated with any of the patient-reported outcomes (P > .05; Table 4; Figure 2A–F).

The multiple regression model was different for the SF-12 PCS (F_4,51 = 3.899, P = .008, adjusted R² = 0.174), indicating a medium effect size.³⁰  The predictor variables (age, sex, career duration, and concussion history) explained 20.9% of the variability in the SF-12 PCS. However, only age added to the model (P = .004): every 1-year increase in age was associated with a 0.138 decrease in the SF-12 PCS score (Table 4).

None of the other models were different for any of the patient-reported outcomes (P > .05). Nonetheless, age added to the SF-12 MCS model (P = .04), whereby every 1-year increase in age was associated with a 0.242 increase in the SF-12 MCS score.

The percentages of abnormal responses (ie, meeting the standard of a clinical cut point cited in the earlier test descriptions) are also highlighted (Table 2). The χ² findings are reported in Table 5. A post hoc z test of the adjusted residuals revealed that the proportion of individuals in the NON group with abnormal (ie, low) SF-12 PCS scores was greater than that in the NCA group. In addition, the proportion of participants with abnormal SWLS scores (ie, dissatisfaction with life) in the RUG group was different from that in the other 3 groups, such that a greater proportion of participants in the RUG group self-reported dissatisfaction with life.

Contrary to our hypothesis, RHI exposure history was unrelated to patient-reported outcomes in early to midadulthood among individuals who were physically active. However, supporting our hypothesis, individuals who were not physically active had worse patient-reported outcomes (ie, worse SF-12 PCS, AES-S, and SWLS scores). Second, counter to our hypothesis, the duration of the contact or collision sport career was unrelated to adverse patient-reported outcomes in early to midadulthood among current and former contact or collision sport athletes, as evidenced by the lack of contribution to the models. Overall, these findings suggest that RHI exposure, whether previous or prolonged, may not negatively affect patient-reported outcomes among physically active individuals in early to midadulthood; still, the effects in later life remain to be elucidated.

Previous studies^14,18  on postcollege-aged individuals have been limited to primarily young or older male cohorts and have not accounted for current physical activity, a modifiable risk factor for numerous health conditions. Our results suggested that patient-reported outcomes were not worse among early to middle-aged adults with a history of sport-related RHI exposure who reported being physically active than among both physically active and inactive individuals without a history of sport-related RHI exposure. This finding is consistent with research on former National Collegiate Athletic Association Division III athletes in midlife, which indicated that collision sport athletes did not have worse neurobehavioral quality-of-life outcomes but that a concussion history was associated with worse quality of life.³  Indeed, the benefits of exercise and physical activity on mood, anxiety and depression, dementia, cardiovascular disease, metabolic syndrome, and other noncommunicable diseases are well established.⁶  Therefore, the reason that neither RHI-exposed group (HRS and RUG) displayed worse scores on any outcome than the NCA group may be due to the neuroprotective and beneficial effects of exercise and regular physical activity and to the fact that these participants were in their mid-30s. In addition, given that the NON group’s self-reported scores were worse than those of the NCA group for 3 patient-reported outcomes (ie, SF-12 PCS, AES-S, and SWLS) despite never having experienced RHI through contact or collision sport further highlights the detrimental effects of a lack of physical activity and presumably low levels of cardiorespiratory fitness, which have been major risk factors linked with many diseases and health conditions (eg, cardiometabolic disease, all-cause deaths, systemic inflammation, stress, anxiety, Alzheimer disease).⁶ 

Despite many group differences in the adjusted models, a considerable proportion of participants in each group had scores outside the clinical cut points (Tables 2 and 3; Figure 1). For instance, the NON group self-reported lower SF-12 PCS scores in the adjusted models than the NCA group. As evidenced in Tables 2 and 3, a greater proportion of individuals in the NON group (17.9%) had SF-12 PCS scores below the population average of 50 compared with the NCA group (0%). The χ² analysis, however, only revealed a difference in the proportion of individuals with low SF-12 PCS scores in the NON group compared with the NCA group. Moreover, a greater proportion of participants in the RUG group (18.5%) expressed dissatisfaction with life as shown in SWLS scores <20 of 35 than the NON (14.3%), NCA (3.4%) and HRS (0%) groups, which may have been related to longer career durations in RHI sports.²⁷  Herein lies a unique trend: although most of our group differences reflected worse performance in the NON group than the NCA group in the adjusted models, signifying that former contact or collision sport athletes (ie, HRS and RUG) were not performing worse than physically active individuals who never experienced RHI, we observed a substantial proportion of individuals in the NON, HRS, and RUG groups who displayed clinically meaningful dysfunction (up to 37%). Indeed, RHI exposure confers risk, but for most of these athletes, the benefits appeared to outweigh the negative risks related to patient-reported outcomes. Yet a subpopulation of athletes in whom the risk-to-benefit ratio inclines toward a greater risk of dysfunction may exist. Although the progression of this dysfunction cannot be determined from these data, it highlights a clear trend that some individuals in these cohorts performed worse than their peers, which warrants investigation to measure their later-life patient-reported outcomes.

Counter to our secondary hypothesis, the duration of the contact or collision sport career was unrelated to patient-reported outcomes among former contact or collision sport athletes (HRS and RUG). Our study expanded on previous research in which authors assessed career duration and patient-reported outcomes in rugby players¹³  by including physical activity status and other collision sport athletes. Indeed, a variety of contact or collision sports have been linked to adverse mental health, a heightened prevalence of anxiety and depression compared with the general population, and worse quality of life than in low- or noncontact athletes.^5,8,9  Moreover, contact or collision sport participation has been associated with a variety of neurologic and neurodegenerative diseases.^4,20  However, conflicting information has suggested that contact or collision sport participation is unrelated to worse mental and psychological health outcomes in middle life^13,18,31  and unrelated to neurologic disease diagnoses in American football players.¹²  Consistent with a study of community rugby players (approximately 32 years old) who had an average contact or collision sport career length of approximately 13 years, we found that career duration was unrelated to patient-reported outcomes.¹³  Herein, our sample of contact or collision sport athletes was relatively the same age (approximately 35 years old) with a similar contact or collision sport career duration (15 years), but importantly, both samples consisted of approximately 41% women, as female athletes have been historically underrepresented in traumatic brain injury–related research.³²  As such, career duration may be unrelated to patient-reported outcomes across sexes in early to midadulthood but should remain a variable of interest due to its relationship with later-life neurodegenerative disease.²⁰ 

The primary limitation of our work was that the groups were not matched by age and sex. Nonetheless, the groups did not differ in the age or sex breakdown. Moreover, the inclusion of a large proportion of women was a relative strength because female athletes have been underrepresented in the previous literature.³²  Another limitation was the use of career duration as a metric for RHI exposure. Despite its identification by the National Institute of Neurological Disorders and Stroke consensus statement on traumatic encephalopathy syndrome¹⁴  as the only modifiable risk factor, RHI exposure does not account for the sport, position, level of play, or volume of play over a lifetime. Our study was limited in its design, which prevented us from delineating the effects of physical activity and RHI on patient-reported outcomes. Future investigators should use more sensitive measures of physical activity to determine the chronicity of lifetime physical activity to further elucidate the potential protective effects of exercise on the relationship between repetitive neurotrauma and patient-reported outcomes in early to midadulthood. Lastly, our sample may have been susceptible to respondent and survivor bias, whereby collision sport athletes who had physical and mental dysfunction may have ceased participation in sport, not volunteered for our study, or both, whereas healthy, more resilient athletes continued to participate and subsequently volunteered.

Neither a history of contact or collision sport participation nor the duration of the contact or collision sport career negatively affected patient-reported outcomes among physically active individuals in early to midadulthood. However, a sedentary lifestyle was negatively associated with patient-reported outcomes in the absence of RHI exposure. Importantly, a nontrivial proportion of participants across groups had abnormal patient-reported outcomes, suggesting a potential threshold between RHI exposure risk and the benefits of sport and exercise. These findings fill critical gaps in the literature and add to the growing body of research,^12,13,18  suggesting that career duration is unrelated to psychological health dysfunction in early to midadulthood. Future authors should expand on these results by evaluating older cohorts and using more precise measures of physical activity status and history as well as RHI exposure history to better understand the risk-to-benefit ratio between RHI exposure and physical activity throughout the lifespan.

This study was made possible and funded in part by the University of Delaware Unidel Distinguished Graduate Scholars Fellowship and the Department of Kinesiology and Applied Physiology Doctoral Research Fund (Dr Hunzinger). Dr Hunzinger acknowledges funding support in part by the Department of Defense grant W81XWH-21-1-0590 (principal investigator: Dr Andrea Schneider), the Penn Injury Science Center, and National Institutes of Health/National Institute of Neurological Disorders and Stroke brain injury training grant T32 NS043126.