Key points

Introduction

In recent years, concussion has been brought to the forefront of both public and scientific literature. This attention includes increased concern about removing athletes from the field of play following a concussion, along with establishment of return-to-play protocols for subsequent practices and competition. In addition to these safeguards, which are thus far clinically rather than empirically driven, more players and physicians are being faced with the difficult question of how many concussions are too many to continue participation in a given sport. Now, in the wake of increased attention on the subject, players across several professional sports are electing to retire or terminate contracts because of concerns over the long-term consequences of concussion. As many athletes attempt to balance their passion for sport and financial security with their long-term health and safety, physicians are in the unique position to offer education and guidance around how many concussions are too many.

Introduction

In recent years, concussion has been brought to the forefront of both public and scientific literature. This attention includes increased concern about removing athletes from the field of play following a concussion, along with establishment of return-to-play protocols for subsequent practices and competition. In addition to these safeguards, which are thus far clinically rather than empirically driven, more players and physicians are being faced with the difficult question of how many concussions are too many to continue participation in a given sport. Now, in the wake of increased attention on the subject, players across several professional sports are electing to retire or terminate contracts because of concerns over the long-term consequences of concussion. As many athletes attempt to balance their passion for sport and financial security with their long-term health and safety, physicians are in the unique position to offer education and guidance around how many concussions are too many.

Epidemiology

The Centers for Disease Control and Prevention estimates that there are between 1.6 and 3.8 million recreation and sport-related concussions per year. Marar and colleagues studied high school athletes over a 2-year period and found concussions occurred at a rate of 2.5 per 10,000 athlete exposures. Football had both the highest number of concussions (47%) and the highest rate of concussion (6.4 per 10,000 exposures). The sport with the greatest proportion of concussions compared with other injuries was boys’ ice hockey (22%). In sex-comparable sports, girls had a higher concussion rate (1.7) than boys (1.0). Overall, 11.5% of athletes sustaining a concussion had previously sustained a sports-related concussion in either that season or a previous season. Most athletes (55.3%) returned to play in 1 to 3 weeks, with 22.8% returning in less than 1 week and 2.0% returning in less than 1 day. In a 2014 survey of Division I National Collegiate Athletic Association athletes, 11.7% reported one concussion and 4.5% reported multiple concussions during their collegiate career. The highest rates were in women’s and men’s ice hockey, wrestling, and football. The highest reported rate of multiple concussions was in football.

In professional sports, the stakes and financial incentives are higher. There were 228 reported concussions in the 2013 National Football League season, including preseason and regular season games and practices. The National Hockey League (NHL) saw 559 concussions from the 1997 through 2004 seasons totaling 1.8 concussions per 1000 player hours. There were 53 concussions in the 2014 NHL season, down from 78 in the previous season.

Past and current recommendations

The question of when to retire following sport-related concussion is not a new one. Historically, recommendations were based on a proposal by Quigley in 1945, which was later published by Thorndike in 1952. This article suggested that play should be suspended for at least the remainder of the season if an athlete sustained 3 concussions. At the time, concussion required a loss of consciousness. Although these recommendations were based on anecdotal evidence and using a concussion definition that we no longer consider valid, they have been used as the basis for current return-to-play guidelines, league protocols, medical decision making, as well as the hypothetical threshold for research into the cumulative effects of repeated injury. In 2002, McCrory published a review questioning the scientific validity of the 3-strike rule and suggesting that there is no set number of concussions after which a player should retire. A general set of guidelines was proposed based on common clinical scenarios, including the recommendations that players with persistent cognitive or neurologic symptoms should be withheld from collision sports until symptoms fully resolve and also highlighting the importance of differentiating postconcussive headaches from headaches with an alternative cause.

McCrory proposed that moderate to severe traumatic brain injury (TBI) resulting in subarachnoid hemorrhage (SAH), persistent neurologic deficit, or TBI requiring craniotomy should be a contraindication to further participation in collision sport. Recommendations for moderate TBI resulting in epidural hematoma include reevaluation for participation in collision sport after a minimum of 12 months. If symptoms have resolved and neurologic sequelae have normalized, return to play may be considered.

Cantu and Register-Mihalik published a statement in 2011 addressing the issue of retirement from sport following concussion. Contraindications to returning to sports were persistent postconcussive symptoms, increasing symptoms in the setting of decreasing impacts, symptomatic neurologic or pain-producing abnormalities about the foramen magnum, permanent central neurologic sequelae, hydrocephalus, spontaneous SAH from any cause, and second impact syndrome (SIS). Of note, the topic of SIS remains both controversial and questionable, as evidence for such a phenomenon is minimal based on critical review of these cases. These observations raise critical questions about the validity of SIS as a criterion for retirement from sport.

In 2011, Sedney and colleagues published a review of the current literature and offered a list of specific season-ending and career-ending features of concussion. Season-ending features included prolonged postconcussion syndrome (PCS) (defined as symptoms >6 months, lowered concussion threshold, diminished athletic or academic performance), 3 or more concussions or 2 or more major concussions (symptomatic for longer than a week) in a single season, or diminished academic or athletic performance. Career-ending features included persistent PCS (without a clear definition of persistence), Chiari malformation, intracranial hemorrhage, diminished academic performance or cognitive abilities, lowered threshold for concussion, 3 or more major concussions, computed tomography or MRI evidence of structural abnormality, nonresolving MRI deficits, or symptoms of chronic traumatic encephalopathy (CTE). The investigators emphasized the potential for cumulative and, rarely, progressive symptoms after concussion.

Concannon and colleagues reviewed the subject in 2014, creating a similar list of absolute contraindications to return to play, including persistent PCS (with PCS defined as symptoms lasting longer than 3 months), SAH, permanent neurologic injury, SIS, hydrocephalus, imaging that increases risk for future brain injury (edema, hemorrhage, hydrocephalus, arachnoid cysts), and permanent deficit on neuropsychological testing. On the topic of in vivo diagnosis of CTE as a criterion for return to play, this article called into question the actual incidence and prevalence of CTE, emphasizing the clinical variability in symptoms and weak correlation between symptoms and histopathology of tissue samples. The investigators recommended further study into risk factors for developing CTE as well as the actual incidence and prevalence. These same questions and other critical issues have been raised with regard to CTE, which thus far remain unanswered. Given the preliminary nature of the evidence, it remains premature to use CTE (or symptoms presumed to be related to this condition) as a criterion for clinical decision making, including retirement from a given sport. The investigators do recommend discussing the current state of this diagnosis with families as the literature base evolves.

Although some of the recommendations for retirement from sport have a strong evidence base and others are consistent among experts, such as SAH, some of the cited recommendations need clarification. Persistent neurologic deficits and persistent symptoms provide a vague timeline, which cannot easily be enforced when a player is exhibiting a prolonged recovery. Although it is clear that a player should not return to play while symptomatic, a prolonged recovery in itself does not necessitate retirement. It is not clear at what point in time symptoms or deficits become persistent and, thus, become a contraindication to return to play. Additionally, although in the past Chiari malformations were thought to preclude athletes from participating in collision sport, recent studies have shown that small asymptomatic Chiari malformations provide no additional neurologic risk to collision sport athletes. There is some question in regard to the return to play in the setting of craniotomy. Multiple reports of safe return to play in multiple sports exist.

Prolonged symptoms

PCS is a feared yet poorly understood consequence of concussion in sport. Of those who sustain concussions, 95% or more resolve within a few days to around a 3-month time period, leaving a small minority (5% or less) of concussed athletes with prolonged symptoms. Who of these will be diagnosed with PCS is uncertain because of inconsistencies in definition and the contribution of confounding diagnoses. A recent survey of American College of Sports Medicine physicians highlighted the lack of consensus over the duration and quantity of symptoms required to make the diagnosis of PCS. The greatest proportion of physicians surveyed (33%) replied that they make the diagnosis after 1 to 3 months of symptoms; however, 27% selected “less than 2 weeks” and 20% selected the “2 weeks to 1 month” option. This lack of consensus carries over into retirement recommendations that use prolonged symptoms as a criterion for retirement. Several sources cite persistent symptoms as a condition for retirement; however, few quantify how many weeks or months of symptoms qualify as persistent. Care must also be taken in making this diagnosis, as there are many conditions that could be falsely designated as sequelae of concussion. Preexisting headache and mood disorders can be misattributed to the concussive injury and obscure the trajectory of the recovery, leading to incorrect diagnoses, failure to treat these conditions, and misguided recommendations to athletes regarding return to play and retirement.

Iverson calculated the effect size of a wide variety of nonconcussive-related factors across several meta-analyses (eg, chronic benzodiazepine use/withdrawal; psychiatric conditions, such as bipolar disorder, attention-deficit/hyperactivity disorder [ADHD], and depression; and litigation). These factors had substantially greater impact on neuropsychological functioning than concussion following the 3-month recovery window. Nonathlete studies have demonstrated that preexisting psychiatric illness or mood disorder as well as high anxiety levels 1 week after injury both correlated with unresolved symptoms at 3 months. There are few factors that may be predictive of prolonged recovery including symptom score at presentation, age, amnesia, photophobia, phonophobia, and history of migraines. Although it would be clinically useful to be able to predict the onset of a prolonged recovery, there are more questions to be answered in order to make evidence-based retirement recommendations based on prolonged recovery in athletes.

The questions that continue to elude us are as follows: Is there increased risk associated with having symptoms that persist outside of the normal recovery period and do these risks resolve once the symptoms have resolved? These answers likely lie in the physiologic mechanism of recovery. We know there is a window of metabolic vulnerability following concussion whereby the brain experiences a high cerebral energy state with altered glucose metabolism, cerebral perfusion, and cellular hypofunction. We do not yet know whether these metabolic changes are altered in prolonged recovery and, if so, what vulnerability is imparted and for what duration or if persistent symptoms are entirely unrelated to such metabolic changes much as the neuropsychological literature suggests the importance of non-neurological factors.

There is no evidence to suggest that a player who experiences prolonged symptoms after a concussion will have an abnormal recovery following subsequent concussions. So what is the real risk to returning an athlete to play after their symptoms have resolved? Many successful athletes have missed months of play and returned to their sport without further apparent consequence. Based on the current evidence, the authors recommend that players with prolonged recovery (persistence of symptoms outside the typical 10-day recovery) are worked up for other causes of their symptoms and not returned to play until concussion-related symptoms have resolved.

Cumulative effects of concussion

Although there is significant evidence that athletes who have sustained concussions are at increased risk for future concussions and mildly slower recovery of symptoms, the evidence for cumulative effects of multiple concussions is less conclusive. The possibility exists that there are no cumulative effects of repeat concussion and that, once a concussion has healed, the metabolic activity in the brain is reset to baseline. However, there are also differing yet not mutually exclusive theories that describe static and progressive neurologic changes following repeat concussion. The breadth of research on the subject is growing, but we have yet to reach a consensus on the effect of multiple injuries to an athlete’s brain.

Several studies have examined the effects of multiple concussions on symptom presentation at subsequent concussion. Collins and colleagues reported in 2002 that athletes with 3 or more prior concussions (self-reported via standardized concussion history form) were more likely to experience on-field loss of consciousness (LOC), anterograde amnesia, and confusion following a concussion and were 9.3 times more likely to have 3 to 4 symptoms following subsequent concussion than those athletes with fewer concussions. Brooks and colleagues studied the effects of multiple concussions on adolescent athletes and found that, among athletes 13 to 17 years old, there were significantly more reported baseline symptoms, however, no significant difference on neurocognitive testing in athletes with 2 or more reported prior concussions. Guskiewicz and colleagues performed a large-scale study of 2905 college football players over 3 seasons and identified 196 concussions in 184 athletes. This study showed evidence for susceptibility to repeat concussion within the typical concussion recovery period. Eleven of 12 (92%) repeat concussions occurred within 10 days of the initial injury. Statistical analysis revealed that athletes with 3 or more prior concussions (by self-report) were 3 times more likely to sustain a concussion than those with no concussion history. This study also measured duration of symptoms and found that athletes with prior concussions experienced slightly longer recovery than those with no concussions (greater than 1 week more).

In a study of high school athletes, grouped by number of prior concussions, it was shown that those with a history of 2 or more concussions had higher rates of headaches, balance problems, and dizziness than players with 1 or no concussions. They also reported higher rates of physical symptoms and symptoms associated with sleep. Those with a history of 2 or more concussions reported higher ratings on nausea and fatigue compared with those with no history of concussion. However, the studies mentioned earlier used self-report as a means to determine incidence of prior concussion, which inherently diminishes the validity of their outcomes. Furthermore, several studies conducted to date do not reflect a relationship between previous mild TBIs (MTBIs) and persistent PCS symptoms. It is also well understood that the symptoms of PCS are nonspecific and, therefore, highly problematic for the purpose of clinical diagnosis as well as in conducting research on outcomes from concussion. A strong body of literature from sport and nonsport literature has developed that shows high rates of postconcussive symptoms in association with non-neurological factors, such as orthopedic injuries, psychological problems (eg, posttraumatic stress disorder and depression), involvement in personal injury claims, and even healthy controls. Given these complexities, consultation with a neuropsychologist is oftentimes quite helpful to assist with the differential diagnosis of concussion based on acute symptoms at the time of injury and to objectively assess cognitive functions rather than rely on self-reported symptoms to determine the effects of injury and best course of treatment moving forward. Additional benefits of such consultations are described further later.

Although no animal brain can serve as a perfect model for humans, animal models allow us to examine pathophysiologic correlations under conditions we are unable to reproduce in human models. One animal model attempted to characterize the effects of cumulative head trauma in mice and demonstrated that repeated mild closed head injury (mCHI) in mice worsened vestibulomotor, motor, short-term memory, and conflict learning impairments as compared with a single mCHI. These learning and memory impairments were sustained 3 months after injury. In this study, repeated mCHIs also reduced cerebral perfusion, prolonged the inflammatory response, and, in some animals, caused hippocampal neuronal loss.

Luo and colleagues demonstrated that animals that received repetitive MTBI showed a significant impairment in spatial learning and memory when tested at 2 and 6 months after injury. Astrogliosis and increased p-Tau immunoreactivity were observed on postmortem pathologic examinations in mice that sustained 3 impacts versus sham mice. Astrogliosis increased with increasing number of MTBIs.

In a series of animal and human studies, Vagnozzi and colleagues have shown the value of magnetic resonance spectroscopy in evaluating the metabolic changes within the brain following multiple concussive injuries with variable spacing. These studies demonstrated that the metabolic effect of recurrent concussion was maximized with a 3-day separation in rats. Although animal models continue to provide insight into the pathologic effects of cumulative brain injury, there has yet to be a study to evaluate the influence of time between concussions on recovery time to account for the effect of healing following the primary concussion.