Comparison of Various Metrics of Repetitive Head Impact Exposure And Their Associations With Neurocognition in Collegiate-Aged Athletes

Abstract

Objective

Characterize the levels of various metrics of repetitive head impacts (RHI) in contact (CS) and non-contact (NCS) sport athletes and determine the extent to which they are associated with fluid cognition.

Methods

Collegiate-aged athletes (n = 176) completed semi-structured interviews about participation in contact sport. RHI was operationalized based on current sport (CS/NCS), the cumulative number of years of participation, age at first exposure (AFE), and based on recently proposed traumatic encephalopathy syndrome (TES) categories. The NIH Toolbox Cognition Battery assessed fluid cognition. General linear models compared RHI metrics between CS and NCS athletes and tested associations of RHI measures with fluid cognition.

Results

CS athletes had more years of RHI exposure, higher rates of “extensive” exposure based on TES criteria, and were more likely to have AFE before age 12 relative to NCS (ps < .001). A subset of NCS athletes, however, reported prior RHI at levels categorized as being “extensive” based on TES criteria (5%), while a larger minority had AFE before 12 (34%). No adverse associations of RHI and fluid cognition were observed (ps > .05). Across all RHI metrics, more or earlier RHI was associated with better episodic memory (ps ≤ .05). Secondary analyses showed this effect was driven by women.

Conclusions

Current results find no evidence that RHI in collegiate-aged athletes is associated with worse neurocognition. Although there was extensive overlap among RHI measures, results demonstrate that categorizing athletes based on their current sport undercounts the lifetime RHI exposure in many NCS athletes.

Introduction

There is growing concern that cumulative, lifetime exposure to repetitive head impacts (RHI) that occur through routine participation in contact sports may have negative effects on brain function and health. This concern has largely been driven by findings in former professional athletes (e.g., American football players) that suggest RHI, rather than concussion or mild traumatic brain injury, is associated with adverse neuropathological (i.e., chronic traumatic encephalopathy) and neurobehavioral outcomes, including deficits in fluid cognitive abilities such as executive functioning (Alosco et al., 2017; Alosco & Stern, 2019; McKee et al., 2013; Montenigro et al., 2017).

More recently, there has been an effort to determine if similar associations of RHI and cognitive function are present in younger, active athletes. However, one critical issue hindering research to date is that no gold-standard for estimating RHI exists. While it is possible to quantify impacts with head impact sensors across a season or even multiple seasons, in most cases this is not a feasible approach to capture all RHI that occurs throughout an individual’s lifespan. Instead, RHI exposure is typically estimated based on self-report or interview-based assessments of an athlete’s current and/or prior participation in sports where RHI are known to occur. Several studies have compared groups of athletes categorized based on their current primary sport (e.g., contact sport vs. non-contact sport, football vs. non-football). Investigations of neurocognitive outcomes using this approach to characterize exposure have not yielded a consistent pattern of differences across contact and non-contact sport athletes. Specifically, high school contact sport athletes had worse visual motor speed and reaction time scores on the Immediate Post-Concussion Assessment Testing (ImPACT) battery relative to non-contact sport athletes (Tsushima et al., 2019; Tsushima, Siu, Yamashita, Oshiro, & Murata, 2018). Conversely, a large study from the Concussion Assessment, Research and Education (CARE) Consortium (n > 15,000) found that, relative to non-contact sport athletes, contact sport athletes had worse performance on the Standardized Assessment of Concussion (SAC) and slower reaction times on ImPACT, but better visual and verbal memory (Katz et al., 2018). One concern about this approach is that it does not capture all RHI that an athlete might experience across a multi-sport career (e.g., a current collegiate track athlete that played multiple years of football in high school), nor does it allow for stratification of the extent of RHI within the contact sport group.

Another approach that has been widely used in the literature involves estimating RHI exposure across the lifespan in a single relevant sport or across all contact sports. For example, several studies in retired, as well as active athletes, have examined associations of cognition with continuous measures of RHI, such as years of participation in football or other similar measures that account for expected head impact exposure (e.g., based on position played), with mixed results (Brett et al., 2020; Brett et al., 2022; Gallo et al., 2022; Meier et al., 2021; Montenigro et al., 2017; Singh et al., 2014). This approach is generally consistent with the conceptualization of RHI exposure levels (i.e., “extensive”, “substantial”, and “none/less than substantial”) proposed by a recent expert consensus statement on the potential diagnostic criteria for the clinical disorder thought to be associated with CTE, traumatic encephalopathy syndrome (TES) (Katz et al., 2021).

Yet another alternative metric that has garnered recent interest is the age of first exposure (AFE) to contact sport based on several studies demonstrating that a younger AFE (often binarized as occurring before the age of 12 or later) was associated with neurocognitive testing performance, magnetic resonance imaging abnormalities, and self-reported symptoms in former American football players (Alosco et al., 2017; Schultz et al., 2018; Stamm et al., 2015a, 2015b). There has been limited evidence of a similar association in younger athletes, to date. A recent study in over 1800 high school and collegiate football players showed that AFE to football was not associated with neurocognitive performance (Brett, Huber, Wild, Nelson, & McCrea, 2019), while no association between AFE to contact sport and neurocognitive performance was observed in a cohort of National Collegiate Athletic Association (NCAA) athletes from US military service academy members (Caccese et al., 2020b). Similarly, a separate CARE Consortium study (n > 6000) showed that early AFE to contact sport was associated with better ImPACT verbal and visual memory scores in women only (n = 1891) (Caccese et al., 2020a).

While there is certainly overlap between these various proxies of RHI, it is unknown whether these markers are differentially related to cognitive outcome measures. Accordingly, the current study in active collegiate-aged athletes with varying levels of exposure had two goals: To characterize the levels and inter-relationship of various metrics of RHI in contact sport (CS) and non-contact sport (NCS) athletes, and To determine the extent to which different metrics of RHI exposure are associated with measures of fluid cognition (e.g., measures of attention, executive function, and episodic memory as opposed to “crystallized” measures such as word recognition). We hypothesized that while current NCS athletes would have lower exposure to RHI than CS, a minority of NCS athletes would still report extensive RHI exposure based on more refined metrics. Furthermore, we tested the hypothesis that greater RHI exposure across all athletes is associated with worse performance on fluid neurocognitive tasks.

Methods

Participants were recruited from the community and enrolled in an ongoing study on the chronic effects of concussion and sport participation. Current analyses focused on data collected from study onset December 2018 to April 2022. Inclusion criteria included a current age of 18–26 years and collegiate sport participation within the last 12 months. Exclusion criteria included self-reported psychiatric, neurodevelopmental (including attention-deficit/hyperactivity disorder), neuroendocrine, severe neurological, or severe autoimmune diseases; use of neuroactive medication; history of moderate or severe TBI; concussion within the past 6 months or incomplete recovery from previous concussion; excessive alcohol or substance use (defined as scores ≥2 on the CAGE Questions Adapted to Include Drug Use, CAGE-AID); contra-indication to study procedures; and injury/illness precluding participation in study protocol. For contact-sport athletes, study procedures were conducted outside of their respective sport’s competitive season when possible; one participated in mixed martial arts throughout the year and one reported being in season for rugby. Visits were completed a median of 3.79 [1.53, 6.12] months since their last in-season RHI exposure. The local Institutional Review Board approved this study and all participants gave written informed consent.

Cognitive Battery

A clinical assessment battery was administered to each participant. Cognitive functioning was evaluated using the National Institutes of Health (NIH) Toolbox Cognition Battery [iPad version] (NIHTB-CB) (Gershon et al., 2013). The NIHTB-CB is a well-validated and reliable computerized-based testing instrument that assesses multiple domains of cognitive function, including executive function, episodic memory, language, processing speed, attention, and working memory (Carlozzi et al., 2014; Dikmen et al., 2014; Gershon et al., 2014; Tulsky et al., 2014; Zelazo et al., 2014). In addition to individual domain measures, composite scores of crystallized and fluid cognition can be derived based on performance across individual subtests (Heaton et al., 2014). The current study focused on fluid cognition using uncorrected standard scores from the following tests: the Flanker Inhibitory Control and Attention Test (assessing attention and executive function), the Picture Sequence Memory Test (assessing episodic memory), the List Sorting Working Memory Test (assessing working memory), the Dimensional Change Card Sort Test (assessing executive function), and the Pattern Comparison Processing Speed Test (assessing processing speed). Performance validity was evaluated using the first 10 items of the Test of Memory Malingering (TOMMe10), with two or more errors used to flag participants for closer inspection of neurocognitive testing performance.

Semi-Structured Interviews for Concussion and Sport History

A semi-structured interview was conducted to quantify cumulative repetitive head impacts from contact sport exposure. Current and previous contact sport participation were systematically collected to allow estimation of various proxies of RHI exposure. Specifically, detailed information was collected for the following sports: football, rugby, wrestling, soccer, men’s lacrosse, men’s ice hockey, and combat sports with routine head impacts. While an athletes’ current sport was recorded, contact sport exposure was collected across all relevant sports at the youth, high school and collegiate level. From this interview, several estimates of RHI exposure history were derived for the current study including the cumulative number of years of exposure across all contact sport (years of exposure), non-cumulative years of contact sport exposure (i.e., participation in multiple sports in a single year is counted as a single year), and age at first exposure (AFE) to any contact sport. Given the strong correlation between cumulative years of exposure with non-cumulative years of exposure (see Results), only the cumulative years of exposure was used in subsequent analyses. AFE was categorized as occurring before 12 years of age (<12) or either occurring after 12 or having no contact sport experience (AFE 12+/none) based on prior work suggesting exposure before 12 years of age as a risk factor for negative outcomes (Alosco et al., 2017; Stamm et al., 2015a, 2015b). Traumatic encephalopathy syndrome (TES) criteria delineate three step-wise levels of exposure, including “extensive” exposure (i.e., >11 years of contact sport exposure including ≥2 years at high school or later levels), “substantial” exposure (i.e., ≥5 years of contact sport exposure including ≥2 years at high school or later levels), and “non-substantial” exposure (i.e., non-substantial or extensive exposure, or no exposure) (Katz et al., 2021). For the purposes of this study, the TES variable was binarized as being “extensive” exposure or “non-extensive” exposure based on the cumulative years of exposure for each participant.

An additional semi-structured interview was conducted to collect retrospective diagnoses of concussion. For each participant, the total number of prior concussions were retrospectively diagnosed based on criteria from the American Congress of Rehabilitation Medicine: loss of consciousness (<30 min), post-traumatic amnesia (<24 h), retrograde amnesia, or alteration in mental status (i.e., being dazed or confused). Experienced research staff administered both semi-structured interviews (i.e., prior concussion and contact RHI exposure) based on a standardized script with question prompts developed by the study team to systematically collect relevant information.

Statistical Analysis

Statistical analyses were conducted in IBM SPSS software (version 27; IBM Corp., Armonk, NY). Independent samples t-tests, chi-square tests, Fisher’s exact tests, and Mann–Whitney U tests were used to compare demographic variables and RHI measures in CS versus NCS athletes. Pearson correlation coefficients were used to determine the relationship between cumulative years of exposure, non-cumulative years of exposure, and AFE as a continuous factor for descriptive purposes. Separate general linear models (GLMs) were used to assess associations of RHI measures (CS vs. NCS, years of exposure, AFE: <12 vs. 12+/none, TES: extensive vs. non-extensive) with NIH Toolbox Cognitive Battery fluid measures (Flanker Inhibitory Control and Attention Test, the Picture Sequence Memory Test, the List Sorting Working Memory Test, the Dimensional Change Card Sort Test, and the Pattern Comparison Processing Speed Test). For all GLM, sex, age, and the number of prior concussions (recoded into 0, 1, 2, or 3+ intervals) were included as covariates. Finally, based on recent evidence of sex-specific associations between RHI and neurocognitive performance in collegiate athletes (Caccese et al., 2020a), secondary GLM were run for each metric in men and women separately, with age and the number of prior concussions again included as covariates. An alpha of 0.05 was used for all analyses. Data are available from the corresponding author upon reasonable request with the execution of necessary data-use agreements.

Results

Demographics

Demographic data and associated statistics are shown in Table 1; a flow diagram of screening and enrollment is provided in Fig. 1. A total of 176 athletes met study criteria and were categorized as current contact sport (CS; n = 115) or current non-contact sport (NCS; n = 61) athletes based on information collected in a semi-structured interview. CS athletes were older, more likely to be men and also reported more prior concussions than NCS athletes (ps < .005), but did not differ in race, ethnicity, socioeconomic status based on the Hollingshead Four Factor Index (Hollingshead, 1975), or performance on the crystallized composite score of the NIH toolbox. A total of 8 participants (5 CS, 3 NCS) had 2 errors on the TOMMe, but inspection of performance on overall clinical battery did not suggest abnormal performance, thus their data are retained.

RHI Metrics in Contact and Non-contact Sport Athletes

As expected, CS athletes had greater cumulative years of exposure to contact sport than NCS (Mean Diff [MD] (SE) = 12.96(0.87), 95% Confidence Interval [11.24, 14.68], p < .001). However, as seen in Table 1, NCS on average did have multiple years of prior exposure to contact sport (Mean (SD) = 2.69(4.71). Similarly, a greater proportion of CS had “extensive” exposure based on TES criteria than NCS (86% vs. 5%, p < .001) and were more likely to have AFE before age 12 (83% vs. 34%, p < .001), though again a subset of NCS athletes had “extensive” exposure and AFE before age 12. In athletes with prior exposure, age of first exposure to any contact sport was inversely associated with years of contact sport exposure (r = −.54, n = 139, p < .001) and the non-cumulative years of contact sport exposure (r = −.58, n = 139, p < .001). The primary years of exposure measure was strongly correlated with the non-cumulative years of exposure (r = .88, n = 176, p < .001).

Association of RHI Metrics with Neurocognitive Performance

Statistics and associated effect sizes for the association of RHI metrics with NIH toolbox performance are shown in Table 2. Episodic memory as assessed on the picture sequence task was better in CS athletes relative to NCS, positively associated with years of exposure, and better in athletes with extensive relative to non-extensive RHI exposure based on TES criteria (ps < .05). A similar effect was observed for AFE, with athletes with AFE < 12 having better episodic memory than those with no exposure or exposure after 12 (AFE 12+/none), though this was not statistically significant (p = .05). There were no other significant associations between RHI metrics and performance on other NIH toolbox tasks. Sensitivity analyses showed that the inclusion of NIH crystallized cognition composite scores as an additional covariate had no impact on results. An additional set of sensitivity analyses showed that the inclusion of a variable accounting for both race and ethnicity also had no impact on results.

Secondary Analyses in Men and Women Separately

Secondary analyses were conducted in men and women separately to determine if similar patterns were present in both subgroups. In women, more RHI exposure was significantly associated with better episodic memory for all RHI metrics. Specifically, better episodic memory performance was observed in CS relative to NCS, in women with extensive relative to non-extensive exposure based on TES criteria, in women with AFE < 12 relative to AFE 12+/none, and was also positively associated with years of exposure (ps < .05; Table 3). In contrast, although the direction of the relationships was consistent with those seen in women, none of the RHI measures were significantly associated with episodic memory performance in men (ps > .05; Table 4). Finally, there were no associations between any measure of RHI and any other cognition measure in either men or women.

Discussion

The current study investigated various metrics of RHI in collegiate-age CS and NCS athletes and their associations with measures of fluid cognition. Contrary to findings in older cohorts that have shown associations between greater RHI exposure and lower neurocognitive performance (Alosco et al., 2017; Montenigro et al., 2017), we found that RHI exposure is associated with better episodic memory regardless of the metric used to approximate RHI across the whole cohort. Secondary analyses showed that this effect was primarily driven by women, demonstrating the importance of considering men and women separately when investigating the potential cumulative effects of RHI exposure. Finally, as expected, current CS athletes had more and earlier RHI exposure compared to current NCS athletes on a variety of RHI metrics investigated. However, a notable subset of NCS did report prior RHI exposure when using more refined metrics that account for lifetime exposure, suggesting that classifying athletes based solely on their current sport as a proxy for RHI may be an oversimplification that does not account for prior exposure in some NCS athletes.

Several studies in older, retired athletes have shown associations between RHI history and subjective neurocognitive difficulties, including performance based-cognition, subjective cognitive difficulties, and psychological symptoms (Alosco et al., 2017; Montenigro et al., 2017; Schultz et al., 2018; Stamm et al., 2015a, 2015b). It has been suggested that RHI history is a risk factor for, if not causally related to, the development of the neuropathological diagnosis of CTE (McKee et al., 2013; Nowinski et al., 2022). Given the large number of athletes that participate in contact sports at various, non-professional levels of play, it is of critical public health importance to determine whether similar effects are present in athletes at younger ages. While some studies in younger, active athletes have reported similar associations of RHI with adverse cognitive outcomes (Singh et al., 2014; Tsushima et al., 2019; Tsushima, Siu, Yamashita, Oshiro, & Murata, 2018), other large scale studies have failed to find similar associations or have observed both subtle positive and negative associations of RHI with cognition (Brett, Huber, Wild, Nelson, & McCrea, 2019; Caccese et al., 2020a, 2020b; Katz et al., 2018).

The current findings expand upon the existing literature by demonstrating similar effects across a variety of potential RHI metrics obtained from a thorough semi-structured interview of sport history, using a well-validated, reliable neurocognitive testing battery with strong psychometric properties, and focusing on athletes across multiple sports without comorbidities that could potentially influence neurocognitive performance. It is important to note, however, that this does not rule out the possibility that individuals with extensive RHI exposure at younger ages may be at increased risk to develop cognitive deficits at later ages. Longitudinal studies are needed to determine what risk factors might be associated with these adverse outcomes observed in some former, aged athletes.

Rather than having adverse effects, we observed that more RHI exposure, regardless of metric, was associated with better episodic memory, an effect that was largely driven by women. Similar positive associations between RHI and better cognitive performance have been reported in large-scale studies, specifically visual and verbal memory on ImPACT (Caccese et al., 2020a; Katz et al., 2018). Interestingly, Caccese and colleagues reported positive associations between younger AFE and improved visual and verbal memory scores only in women, supporting our findings that the association between episodic memory and RHI in the current study were most evident in women (Caccese et al., 2020a). The mechanism underlying the association of better episodic memory with more extensive RHI is unclear, and future research is required to determine if other unmeasured factors associated with contact sport participation could be driving this relationship. For example, there may be aspects of contact sports besides the accumulation of RHI that promote development of episodic memory, such as the type and extent of physical activity or the requirement to learn complex game strategies and plays. While this is only conjecture at this point, these results nevertheless demonstrate the importance of investigating the potential effects of RHI on neurobehavioral outcomes of interest for men and women separately, when possible. Future studies should investigate the extent to which other sex-specific factors such as menstrual cycle or oral contraceptive use may influence current and long-term health of women with cumulative RHI exposure.

Finally, it is unsurprising that CS athletes had more exposure than NCS on all metrics of RHI assessed. It is important to note, however, that a subset of NCS athletes reported prior contact sport exposure (e.g., 5% had extensive RHI exposure based on recent TES criteria) and 34% reported having their first contact sport exposure before the age of 12. Continuous measures that account for all contact sport history offer a more complete assessment of RHI exposure across the lifespan in contrast to comparing groups of athletes solely based on their current sport alone. This finding has important implications for future studies investigating the potential associations of RHI with other clinical and physiological outcomes. Similarly, sub-analyses in athletes with prior contact sport exposure showed that AFE is moderately associated with total years of exposure, though both showed similar relationships with the studied cognitive outcomes. Thus, investigators should carefully consider what RHI related construct is relevant to their hypotheses (e.g., developmental disruptions with younger AFE) and outcomes of interest. Future studies can compare these simple measures of cumulative RHI exposure with other developed measures incorporating more detailed information, such as sports and positions played, estimates of sport-specific, quantifiable head impact exposure based on published sensor data, and other relevant variables (Kerr et al., 2015; Montenigro et al., 2017).

Limitations

The current work has several strengths, including the inclusion of men and women across a variety of sports, strict exclusion criteria limiting analyses to a cohort without comorbidities that can potentially confound neurocognitive test performance, the use of a well-validated neurocognitive assessment battery, and semi-structured interviews of sport history that allow for assessment of lifetime exposure to RHI using a variety of metrics. There are, however, limitations that should be considered. First, the study is cross-sectional, and we are unable to prospectively assess RHI and their associations with changes in cognition over time. Second, this study focused on collegiate-aged athletes and it is uncertain the extent to which these findings generalize to younger or older athletes. Third, the study sample is relatively demographically homogeneous and it is unclear if findings generalize to historically marginalized groups. Additional studies in larger, more diverse populations of athletes at various ages are required.

Conclusion

Current results show no evidence that cumulative exposure to contact sports assessed using a variety of RHI metrics is associated with worse neurocognitive test performance in active, collegiate-aged athletes. Finally, although all assessed metrics of RHI showed overlap, these results demonstrate that more refined measures of RHI are better able to capture RHI history in all athletes, including those currently competing in NCS, compared to simple characterizations based on the current sport type.

Funding

This work was supported by the National Institute of Neurological Disorders and Stroke of the National Institutes of Health under award number R01NS102225. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The REDCap electronic database and the Adult Translational Research Unit used for this project were supported by the National Center for Advancing Translational Sciences, National Institutes of Health, award number UL1TR001436. BLB acknowledges support from the National Institute of Neurological Disorders and Stroke (award number L30NS113158) and the National Institute on Aging (award number K23AG073528) of the National Institutes of Health. This research was completed, in part, with computational resources and technical support provided by the Research Computing Center at the Medical College of Wisconsin.

Conflict of Interest

Dr. Meier receives compensation as a member of the Clinical and Scientific Advisory Board for Quadrant Biosciences Inc. All other authors report no conflicts.

Acknowledgements

The authors thank Amy Nader for study coordination, Daniel Huber and Lezlie España for database management, and all Brain Injury Research staff at the Medical College of Wisconsin for overall project support.

References