CHRONIC TRAUMATIC ENCEPHALOPATHY (CTE) is a neurodegenerative disorder characterized by abnormal accumulation of hyperphosphorylated tau protein within the brain.1 It is suspected that CTE is linked to a history of concussive or subconcussive blows that occur with athletes who play sports with a high level of physical contact (eg, American football and ice hockey) and members of the military who experienced multiple blast injuries.2 Omalu et al3–5 are credited with bringing recent attention to this disease by publishing in 2005 the first case report of a 50-year-old former National Football League (NFL) athlete who suffered from deficits in memory and judgment. Subsequently, McKee et al6 in 2009 described CTE in 3 more cases of former athletes. Later, McKee et al7 described postmortem analysis of 85 former NFL and National Hockey League (NHL) athletes with a history of repetitive mild traumatic brain injury, ranging in age from 17 to 98 years, and found 65 (76.5%) had CTE.
Although it is agreed that postmortem analysis is currently the only way to definitely diagnose CTE,8 there are varying opinions regarding the degree to which the presence of tau protein represents trauma-induced CTE versus normal deposits as a result of age and other life factors.9 Gardner et al10 reviewed the 85 cases described earlier by McKee et al7 and found that only 15 cases, instead of 65, met a more stringent criterion of CTE. The National Institutes of Health held a consensus meeting in 2016 with the aim of defining the neuropathological criteria for the diagnosis of CTE.11 Seven neuropathologists blindly evaluated 25 cases of various tau proteinopathies including CTE and a number of dementing brain diseases that are characterized by excessive tau. The results demonstrated reasonably good agreement among the neuropathologists; and, the group was able to add greater specificity to the description and, therefore, the diagnosis of CTE.
The most recent research publication by Mez et al12 on the association of CTE and football reported that, of 111 NFL players, 110 (99.1%) showed evidence of CTE upon postmortem examination. In the total sample of 202 high school, college, and professional football players, various degrees of CTE were diagnosed in 177 of these players (87.6%). All of the athletes represented in the study had their brains donated to the brain bank at Boston University. The authors also provide a discussion of ascertainment bias; simply stated, families of individuals with cognitive deficits are far more likely to donate the brain of their deceased family member for study. Thus, the results may not be representative of the entire population. Still, research to date on CTE based on pathological studies has been interpreted to imply that all athletes who played contact sports professionally have a high probability of eventually experiencing CTE. The subject of CTE has received major public attention,9 including a movie about Dr. Omalu and the resistance of the NFL to accept that concussions could have long-term deleterious effects.
Nevertheless, controversy continues to surround the clinical manifestations of CTE (ie, what are the patterns of behavior and cognitive deficits experienced by the living individual affected by CTE?). The study by Mez et al12 included detailed, blinded retrospective evaluations of the 111 subjects for whom standardized informant reports had been collected. The results indicated that a progressive clinical course was common in 85% of mild cases of CTE and 100% of severe cases. Gavett et al13 conducted interviews of friends and family members of people who had documented CTE. They described a consistent pattern of impairment in cognition, executive function, mood, behavior (impulsivity), and signs of motor neuron disease. Stern et al14 conducted a similar retrospective analysis of 36 deceased athletes—average age of death 56.8 years—with confirmed CTE (mostly former NFL athletes and a few former NHL athletes). Three of these athletes were asymptomatic, 11 had cognitive dysfunction, 13 had behavior alterations that gradually became mood changes, and 10 were diagnosed with dementia. Using next-of-kin interviews, Alosco et al15 studied 25 professional football players (mean age at death = 65 years) with autopsy-confirmed stage III or IV CTE. They found that all 25 of 25 had cognitive symptoms and their age of cognitive decline was inversely related to their cognitive reserve. However, this study did not compare athletes with a control population.
Two studies published in 2017 found very limited evidence of abnormalities consistent with dementia among contact sport athletes. Esopenko et al16 studied 33 former professional ice hockey players, average age 54 years, and compared them with 18 age-matched controls. They found very few differences on a full range of cognitive measures and no differences on critical cognitive factors like memory. Interestingly, they found the athletes performed much better on all of the measures of cognition than the athletes expected they would perform. McMillan et al17 examined a sample of 52 former international rugby athletes, average age 53 years, and 29 age-matched controls. They also found minimal differences, the primary exception being that the athletes did poorly on one test of verbal learning. Importantly, the mean scores for both groups on almost every measure of cognition were in the average range for age and there were no differences in mental health or daily functioning between the retired athletes and the control group.
Since previous postmortem pathological studies have been interpreted by the public as implying that almost all contact sport athletes have, or will have, some form of CTE, the purpose of the current study was to examine a cohort of living former professional athletes from high-contact sports and compare them with a control group of high-level athletes who did not play contact sports. It was important to our goals to provide a comprehensive assessment of cognitive function, executive function, mental health, physical health, and advanced imaging in search of objective biomarkers for the development of CTE. We also studied their diets, lifestyles, drug and alcohol history, and cardiovascular risk factors to assess possible confounders.
The extensive nature of the assessments conducted on our participants dictated the need for at least 3 articles to present the 3 primary facets of the study: (1) cognitive impairment and rates of mild cognitive impairment (MCI) and the risk factors, (2) executive dysfunction and mental health, and (3) advanced brain imaging findings. Based on an initial round of review, a fourth article (the current article) was proposed to provide an overview of the study and address details of the study participants. The reviewers stated that an overview article should describe the participants in greater detail and confront what they outlined as a critical concern, namely the possibility of selection bias in our recruitment of contact sport athletes. The purpose of the current article is to provide a detailed description of the participants and a general summary of findings and conclusions.
We approached the local NHL and NFL alumni associations with the purpose of recruitment. The athletes voiced a number of concerns, not the least of which was the fact that it was researchers who were quoted regularly in the media describing CTE with little apparent concern for the athletes who are still alive. We tried to assure local NHL and NFL alumni association representatives that we would be objective and would have regular contact to allay their fears and benefit from their wisdom. We attended a number of their fund-raising and social events. We had 2 representatives of the alumni associations attend our planning committee meetings, and they regularly offered suggestions for other issues to consider. Alumni members were then contacted by newsletters and e-mails from their leadership and urged to participate in the study.
The alumni associations were interested in our study because they had been reading research and stories in the press regarding the incidence of CTE; and; while they generally wanted us to assess the former players for cognitive and behavioral issues, they were also quite anxious about the impact of the findings on their members. In response to these concerns, once all assessments were completed on an individual, we gave participants the option to receive individual feedback on the state of their physical and mental health along with suggestions from our clinical team regarding treatment. In the end we had 27 former NHL/NFL volunteers agree to participate.
Of the 27 former professional athletes, 22 were included in the studies. There were 20 participants who completed every questionnaire, every neuropsychological examination, the physical examination, and the advanced imaging. There was 1 former NHL athlete who did not complete the magnetic resonance imaging (MRI) (due to claustrophobia) and 1 former NFL athlete who completed the MRI but did not complete all of the questionnaires. Thus, the imaging study had 21 former contact sport athletes (8 NFL and 13 NHL), and the executive function and MCI studies had 21 former contact sport athletes (7 NFL and 14 NHL).
We were concerned that those who participated in the study to be as representative as possible of the alumni members as a whole. We discussed this issue with the alumni representatives and there appeared to be 2 groups of alumni members who did not participate. One was a small group of alumni who were physically unable because of health conditions such as cardiovascular ill health, kidney disease, or stroke. The second group of nonparticipants was those who have significant public presence. This included athletes currently associated with a professional team as a coach, manager, or team representative.
One important issue for our study was determining the characteristics of the most appropriate comparison group. The lack of a control group has been a common criticism of prior research.18 Hence, after considerable discussion, we decided to compare our professional contact sport athletes with athletes who did not play contact sports. Athletes have been described as extroverted, tough minded, and (sometimes overly) aggressive.19 We thought that by having a comparison group of athletes who did not play contact sports it would control for some of the personality traits that may make athletes, as a group, different from the general population.
The noncontact sport athletes were contacted through associations of athletes such as running and swimming clubs that included older athletes on their roster. These master athletes were involved in individual, non-contact sports since their youth and at the time of enrollment. It was not a deliberate choice to recruit control group participants who were currently active in sports, but we did not know how to systematically contact athletes who were no longer participating in sports. Thus, the health status of the control group may have benefitted from their current state of physical fitness.
Inclusion and exclusion criteria
Inclusion criteria for contact sport group (athletes) were players who (1) played professional contact sports (NFL and NHL) for 2 or more seasons; (2) currently retired; and (3) aged 36 to 72 years. Exclusion criteria included (1) medically unsafe to receive an MRI, with implants or medical devices contraindicated for MRI; (2) history of concussion in the past 2 years; (3) history of moderate to severe brain injury from any cause; (4) history of cerebrovascular event that could lead to hypoxia; and (5) history of learning disability. The control group inclusion criteria were (1) master athletes who participated in individual noncontact sports such as running, cycling, or swimming since their youth; (2) aged 36 to 72 years; and (3) no history of self-reported or documented concussions. Exclusion criteria were the same as contact-sport athletes.
A multidisciplinary team of researchers volunteered to participate in the design of the study. The faculty were divided into 6 teams: (1) lifestyle (nutrition, smoking, drug and alcohol history, and activity); (2) physical health (development of a physical examination; measures of important aspects of physical health such as sleep and pain); (3) psychological assessment (depression, anxiety, mental health, executive function, and aggressiveness); (4) cognition (various aspects of memory and attention, visual spatial orientation, intelligence, verbal skills); (5) imaging (determining the neural structures, hemodynamic and neurometabolite concentrations in the brain of prime concern, and the methodology to best assess for dementia); and finally (6) research methods (determining how the assessment procedures will be carried out and how to reduce redundancy). An executive committee was made up of the principal investigators and leaders of each team.
Each team produced a long list of variables of interest and a similarly long list of measures to assess each of the variables. The cognition team was greatly aided by the Boston University team12 who shared their entire list of cognitive measures being employed as part of their ongoing investigation of CTE. Although our study did not include all of the cognitive tests of the Boston University team, we were able to select our measures with confidence that we could ultimately compare our results with theirs. The executive committee had final say as to which questionnaires would be employed based on the amount of time commitment we thought reasonable for each participant. The final protocol consisted of assessments that would require approximately 2 days to complete, not including MR imaging.
We compared participants on norm-referenced neuropsychological test scores in attention, memory, executive function, language, and visuospatial domains. At least 2 neuropsychology tests were included for each domain. For tests with more than one score per test, we included 2 to 5 primary scores for that test. We also used the Behavior Rating Inventory of Executive Function (BRIEF-A), which is a self-report measure of aspects of behavior normally considered part of executive function.
Mild cognitive impairment
According to Jak et al,20 MCI can be diagnosed using several different criteria based on norm-referenced neuropsychological test scores in attention, memory, executive function, language, and visuospatial domains. The most frequently used is the comprehensive criteria (2 test scores below 1 standard deviation [SD] in 1 domain or 1 test score below 1 SD in 3 domains). To validate our findings, we also used the typical criteria (1 test score below 1.5 SD in 1 domain), which is also frequently used.20 , 21 The results for both criteria were similar, so only the results from the comprehensive criteria are presented.
MRI was used to assess the multimodal conventional and nonconventional MRI measures reflective of neural structure, hemodynamic, and neurometabolite concentration impairment in the brain, on a 3T scanner. The following sequences were acquired: proton density/T2-weighted image; fluid-attenuated inversion-recovery; 3D high-resolution T1-weighted imaging using a fast-spoiled gradient echo with magnetization-prepared inversion recovery pulse, diffusion-weighted imaging, susceptibility-weighted imaging, magnetic resonance spectroscopy, and perfusion-weighted imaging.
Image analysts were blinded to the subjects' demographic, clinical characteristics and group status and included detection of white matter signal abnormalities, calculation of global and regional brain atrophy measures, measurement of diffusion tensor imaging tract-based-spatial-statistics measures of mean diffusivity and fractional anisotropy, assessment of cerebral microbleeds and quantitative susceptibility mapping values in various brain structures, measurement of neurometabolites in the corpus callosum, such as N-acetylaspartate, glutamate, and glutamine, relative to the concentration of creatine and phosphor creatine on magnetic resonance spectroscopy, and calculation of perfusion cerebral blood flow, blood volume, and mean transit time within examined brain regions using perfusion-weighted imaging.
General and mental health
To account for possible confounders, we assessed cardiovascular risk factors including history of hypertension, diabetes, and smoking; and drew blood for complete lipid panels. We also assessed their demographics, education, diet, lifestyle, alcohol and drug history, and energy expenditure. Physical activity was assessed using the Yale Physical Activity (Interview) Survey22 that provides an estimate of the kilocalories expended during the average week. Mental health was assessed using the Beck Depression Inventory-II23 and Beck Anxiety Inventory,24 and the Personality Inventory for the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition).25 All are validated self-reported questionnaires.
As part of our data collection, we interviewed participants regarding their history of concussion. We were interested in the relationship between the number of concussions endured during their playing career and their current state of cognition and mental health. This part of the assessment process with the former NFL/NHL athletes became lengthy, and we faced a dilemma. The athletes often used the question to describe how concussion was perceived during their playing careers. Generally, there was no medical assessment for the player who “had my bell rung,” and players often did not report symptoms because it could interfere with playing time. Some athletes then told stories of concussive effects that would last for weeks and how they hid their symptoms from team medical staff and other team members. They also described instances where they were not sure the hit to their head qualified as a concussion or whether the symptoms may have been from something else, like dehydration or a neck injury. While all of the contact sport athletes could be said to have experienced concussive events, we concluded that it would be impossible to obtain an accurate and useful assessment of concussion history that was consistent across players. In the end, we decided that retrospective recall of events involving repeated blows to the head that occurred decades ago was so inaccurate that we had no confidence in determining the number of concussions for one athlete versus another. We decided to exclude these data altogether.
According to the systematic review and meta-analysis by Karr et al,26 an estimated effect size of 0.80 was used for the power analysis for the studies. Based on that analysis and an expectation that athletes will perform worse than controls, in order to achieve a power of 0.80 with a 1-sided 2-sample t test at level .05, a total of 20 participants in each group were required. Ultimately, to ensure that we would discover an unexpected result (ie, contact sport athletes performing better than the noncontact sport athletes), all statistical tests were 2-sided tests and in the case of t tests an assumption of unequal variances was employed.
In this overview article, we compare the demographic and clinical characteristics between the contact sport athletes and noncontact controls using 2 sample 2-sided t tests with unequal variance (level .05) for continuous variables, the Fisher exact test for 2 × 2 tables (level .05), and χ2 tests for larger contingency tables (level .05). When possible, significant confounding variables such as intelligence quotient (IQ) in the MCI article are accounted for in each article. In addition, each of the 3 articles contains its own statistical analysis section, detailing the data analysis for the corresponding variables in that article. In these articles, upon the advice of the statistical reviewers, we did not control for multiple comparisons (MCs) because differences between the groups may be lost due to the conservative nature of MC adjustments. Thus, by not adjusting for MC, we increased the likelihood of finding significant differences between the groups. We also present estimated confidence intervals and effect sizes whenever appropriate so that the readers could evaluate whether a finding or nonfinding was sufficient to draw a conclusion.
Demographics and clinical characteristics
Table 1 presents demographic and clinical characteristics of the study groups. A total of 27 former NFL and NHL players volunteered and provided informed consent. From the pool of 27 volunteers, 5 participants were excluded because of difficulty completing the testing process and 1 participant was excluded because he had a serious brain injury from a motor vehicle crash that was only revealed during testing. Hence, 22 male professional contact sport athletes (8 NFL and 14 NHL) comprised the contact sport athlete group. Twenty-one contact sport athletes participated in the imaging study (8 NFL and 13 NHL) and 21 contact sport athletes participated in the study of executive function and MCI. The average playing career was 8.5 years and all played for at least 2 professional seasons at the highest level (NHL or NFL). Twenty-four noncontact sport athletes volunteered to participate, but 3 did not complete all questionnaires and were therefore considered dropouts. Hence, 21 male noncontact sport athletes comprised the noncontact control group. There were no significant differences between contact athletes and noncontact controls for age and ethnicity. However, contact athletes had significantly higher body mass index compared with noncontact controls (30.1 vs 24.5, P < .0001). There was also higher education in noncontact controls versus contact athletes (P = .024). No significant differences between the study groups were found for current or past smoking, alcohol, or drug abuse, although frequencies were somewhat higher in the contact athletes. There was a trend for more frequent current smoking in contact athletes compared with noncontact controls (4 vs 0, P = .025).
The physical examination results revealed important differences between the 2 groups. Eighteen of the 22 contact athletes had significant sleep disturbance. Sixteen of the 22 contact athletes had surgery due to injuries sustained while playing sports. Four contact athletes had replacement body parts such as knees, hips, or shoulder. Nineteen contact athletes reported chronic pain that was confirmed during the physical examination. In contrast, only one of the noncontact controls complained of chronic pain and one complained of moderate, periodic pain. One of the no-contact control participants described sleep difficulties. The noncontact sport athletes were almost twice as likely to participate in exercise and expended significantly more kilocalories per week compared with the contact sport athletes. It should be noted that all differences found between the groups would tend to bias toward lower cognitive functioning in the group of contact athletes.
All participants were given the opportunity to meet with the clinical team (the physician who completed the physical examination, the neuropsychologist who oversaw the administration of the cognitive testing, and a general psychologist who was familiar with the psychiatric screening results). None of the noncontact control group participants chose to meet with the clinical team. The opposite was true for the contact sport athletes who all chose to receive feedback on their results. Most of the athletes were informed about services available for sleep disturbance and pain management. Occasionally, there was a different specific disorder for which there was referral to a specialist.
The results of the specific substudies are provided in detail in the accompanying articles. The study of executive function and behavior indicated that contact athletes perceived themselves to have issues with working memory and overall rated themselves more poorly on executive function compared with the noncontact controls. However, spousal ratings of both groups were indistinguishable. As well, objective findings of factors associated with executive function revealed no differences between the groups, even for working memory.
The study of cognitive function and MCI also demonstrated relatively few differences between contact athletes and noncontact controls. The contact athletes had less education and lower estimated IQ. Despite a broad range of neurocognitive tests, there were only a few tests that showed that contact athletes were less capable: Letter Fluency and List B Immediate Recall. The study employed a set of comprehensive criteria and neurocognitive tests to classify participants as MCI, generally considered a possible precursor to dementia. The contact athletes were more likely to qualify as MCI (8/21) compared with controls (3/21). These differences only showed a trend toward significance (P = .08) with the differences appearing to be primarily related to education, IQ, and body mass index.
The results of the multimodal imaging study did not find neurometabolic, hemodynamic, functional, or structural differences on brain MRI. An analysis of the contact athletes with MCI also revealed no significant findings. Surprisingly, one of the few differences between the groups was the presence of cerebral microbleeds where more noncontact controls (n = 7) than contact athletes (n = 2) had at least one microbleed. This difference only approached significance.
Perhaps the most important finding in this study is the fact that, after completing extensive evaluation for cognitive function, executive function, and mental health, none of the former NHL or NFL athletes could be diagnosed with early-onset dementia. We began this study assuming some, if not most of the former contact sport athletes, would have dementia and we would use advanced imaging to confirm the MRI abnormalities, suggestive of CTE, based on previous literature.27 Instead, our advanced brain imaging confirmed our findings that these former professional athletes were functioning according to their age in each of the facets being evaluated. Our noncontact sport control group turned out to be better educated and in much better health than our contact sport athletes, but we discovered they were not substantially different in most aspects of functioning except physical activity.
Reviewers for these studies expressed concern about the representativeness of the sample of former NFL and NHL athletes who volunteered to be a part of this study. The authors, Mez et al,12 of the postmortem study of CTE in former contact sport athletes expressed a similar concern about selection bias. It would appear that former athletes who volunteered for our study worried about their mental state and wanted to know, for themselves, whether playing their sports professionally had left them disadvantaged. In contrast, the family members of those who donated brains to the brain bank at Boston University did so because they were convinced that there was something wrong with their family member. In our study, family members (mostly spouses) rated the function of the contact athletes more positively than the athletes themselves. We believe that our study participants represent former contact sport athletes who are alive and do not have dementia, whereas the athletes with CTE appeared to have dementia before they passed away. The participants in the current study were also a convenience sample; but one for which we would have expected a selection bias for worse functioning—if not when compared with the entire population of retired contact sport athletes, at least when compared with noncontact controls still participating in regular exercise.
In an effort to determine whether our sample of participants was comparable to the sample found to have CTE, we reviewed all of the publications on CTE and found they provide surprisingly little demographic information that is comparable. For example, they describe the numbers of years their sample played football but do not separate out the years played in professional football. We know the age at death of those who qualified as severe CTE (median age 71 years), but they do not report the age of death of those who played professional football. We know that our NFL athletes played an average of 4.4 years, and the NFL players association reports that the average playing career today is 3.3 years. But the average career length during the era represented by most of our athletes was apparently more than what it is today, by 1 or 2 years. A survey of NFL alumni conducted in 2001 had a large sample (n = 2552) and the average career length in their sample was 6.7 years.28 It is our assumption that our sample is reasonably representative of those who played in the NFL in the era they represent. In contrast, the average career length for an NHL player is 5 years (according to the NHL Players Association) and the average career for our sample of former NHL athletes was 11.3 years. We are very confident that our sample of NHL athletes had average to above average exposure based on career length.
Our findings support the recent studies by Esopenko et al16 of former professional hockey players, and McMillian et al17 of elite rugby union players. These 2 studies, like our study, did not find evidence of dementia in retired contact sport athletes, suggesting that many of these athletes not only escape the horror of CTE but appear to have relatively normal lives. We believe our study had several important advantages over the Esopenko et al16 and McMillian et al17 studies. We used a clearly defined control group of athletes who did not play contact sports, whereas the other studies simply used age-matched controls. We also used advanced imaging to confirm the brain health of the athletes. Finally, we think it was prudent to examine other factors in the lives of the former professional athletes. For example, physical examinations demonstrated that these athletes had more significant problems with sleep disturbance and pain management than with cognition. The athletes who are experiencing MCI may have more to worry about because of obesity, chronic pain, and sleep disturbance than they do with their history of playing a contact sport.
News coverage has given the public the impression that CTE is inevitable among professional contact sport athletes.9 It is a picture that was reinforced by the postmortem analysis of brains of athletes donated to a brain bank. The results of our comprehensive investigation of a relatively small sample of former athletes do not support this notion. However, this study and similar investigations cited earlier do not provide argument against the existence of CTE but do suggest that the risk is not as great as once believed. The next steps in research should be to try to address the factors, genetic or otherwise, that make the risk of CTE for some athletes much greater than others.
1. McKee AC, Daneshvar DH. The neuropathology of traumatic brain injury
. Handb Clin Neurol. 2015;127:45–66.
2. Kiernan PT, Montenigro PH, Solomon TM, McKee AC. Chronic traumatic encephalopathy
: a neurodegenerative consequence of repetitive traumatic brain injury
. Seminars Neurol. 2015;35(1):20–28.
3. Omalu BI, DeKosky ST, Minster RL, Kamboh MI, Hamilton RL, Wecht CH. Chronic traumatic encephalopathy
in a National Football League player. Neurosurgery. 2005;57(1):128–134.
4. Omalu BI, DeKosky ST, Hamilton RL, et al Chronic traumatic encephalopathy
in a National Football League player: part II. Neurosurgery. 2006;59(5):1086–1092; discussion 1092–1093.
5. Omalu BI, Bailes J, Hammers JL, Fitzsimmons RP. Chronic traumatic encephalopathy
, suicides and parasuicides in professional American athletes: the role of the forensic pathologist. Am J Forensic Med Pathol. 2010;31(2):130–132.
6. McKee A, Stern R, Nowinski C, Gavett B, Cantu R. Chronic traumatic encephalopathy
in professional football players. J Neuropathol Exp Neurol. 2009;68(5):560.
7. McKee AC, Stern RA, Nowinski CJ, et al The spectrum of disease in chronic traumatic encephalopathy
. Brain. 2013;136(1):43–64.
8. Riley DO, Robbins CA, Cantu RC, Stern RA. Chronic traumatic encephalopathy
: contributions from the Boston University Center for the Study of Traumatic Encephalopathy. Brain Inj. 2015;29(2):154–163.
9. Davis GA, Castellani RJ, McCrory P. Neurodegeneration and sport. Neurosurgery. 2015;76(6):643–55; discussion 655–656.
10. Gardner RC, Possin KL, Hess CP, et al Evaluating and treating neurobehavioral symptoms in professional American football players: lessons from a case series. Neurol Clin Pract. 2015;5(4):285–295.
11. McKee AC, Cairns NJ, Dickson DW, et al The first NINDS/NIBIB consensus meeting to define neuropathological criteria for the diagnosis of chronic traumatic encephalopathy
. Acta Neuropathologica. 2016;131(1):75–86.
12. Mez J, Daneshvar DH, Kiernan PT, et al Clinicopathological evaluation of chronic traumatic encephalopathy
in players of American football. JAMA. 2017;318(4):360–370.
13. Gavett BE, Stern RA, McKee AC. Chronic traumatic encephalopathy
: a potential late effect of sport-related concussive and subconcussive head trauma. Clin Sports Med. 2011;30(1):179–188, xi.
14. Stern RA, Daneshvar DH, Baugh CM, et al Clinical presentation of chronic traumatic encephalopathy
. Neurology. 2013;81(13):1122–1129.
15. Alosco ML, Mez J, Kowall NW, et al Cognitive reserve as a modifier of clinical expression in chronic traumatic encephalopathy
: a preliminary examination. J Neuropsychiatry Clin Neurosci, 2017;29(1):6–12.
16. Esopenko C, Chow TW, Tartaglia MC, et al Cognitive and psychosocial function in retired professional hockey players. J Neurol Neurosurg Psychiatry. 2017 88(6):512–519.
17. McMillan TM, McSkimming P, Wainman-Lefley J, et al Long-term health outcomes after exposure to repeated concussion
in elite level: rugby union players. J Neurol Neurosurg Psychiatry. 2017;88(6):505–511.
18. Asken BM, McCrea MA, Clugston JR, Snyder AR, Houck ZM, Bauer RM. “Playing through it”: delayed reporting and removal from athletic activity after concussion
predicts prolonged recovery. J Athl Train. 2016;51(4):329–335.
19. Stoyanova S, Ivantchev N, Petrova K. Connectivity of athletes' personality traits and career period as their predictors. BJCEM. 2016;4(1):41–50.
20. Jak AJ, Bondi MW, Delano-Wood L, et al Quantification of five neuropsychological approaches to defining mild cognitive impairment. Am J Geriatr Psychiatry. 2009;17(5):368–375.
21. Clark LR, Delano-Wood L, Libon DJ, et al Are empirically-derived subtypes of mild cognitive impairment consistent with conventional subtypes? J Int Neuropsychol Soc. 2013;19(6):635–645.
22. De Abajo S, Larriba R, Marquez S. Validity and reliability of the Yale Physical Activity Survey in Spanish elderly. J Sports Med Phys Fitness. 2001;41(4):479–485.
23. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry. 1961;4(6):561–571.
24. Beck AT, Steer RA, Carbin MG. Psychometric properties of the Beck Depression Inventory: twenty-five years of evaluation. Clin Psychol Rev. 1988;8(1):77–100.
25. Krueger RF, Markon KE. The role of the DSM-5 personality trait model in moving toward a quantitative and empirically based approach to classifying personality and psychopathology. Annu Rev Clin Psychol. 2014;10:477–501.
26. Karr JE, Garcia-Barrera MA, Areshenkoff CN. Executive functions and intraindividual variability following concussion
. J Clin Exp Neuropsychol. 2014;36(1):15–31.
27. Koerte IK, Lin AP, Willems A, et al A review of neuroimaging findings in repetitive brain trauma. Brain Pathol. 2015;25(3):318–349.
28. Pietrosimone B, Golightly YM, Mihalik JP, Guskiewicz KM. Concussion
frequency associates with musculoskeletal injury in retired NFL players. Med Sci Sports Exerc. 2015;47(11):2366–2372.