Sport-related concussion (SRC ) is a common occurrence in young athletes, accounting for 8.9 % of all high school injuries and 5.9 % of all collegiate injuries [1]. Acute symptoms in children are similar to those described in adult populations, including somatic, cognitive, emotional, and sleep-related difficulties [2–6]. However, conclusions based on adults do not necessarily generalize to younger children. The biomechanics and pathophysiological response to trauma, as well as the risk of repeat injury may be age related.

Anatomical and mechanical properties of the body and brain differ between developing and mature individuals. With regard to mechanics, it is accepted that concussion occurs when linear and/or rotational forces are applied to the head, neck, face, or body such that an impulsive force is imparted to the brain [7]. While rotational forces are thought to be more instrumental in SRC than linear [8], both are subject to neck strength. This was examined in Collins and colleague’s 2014 [9] study of 6704 high school athletes, which found that smaller neck circumference, smaller mean neck to head circumference, and weaker mean overall neck strength were associated with incidence of concussion. By extension, children or adolescents may be at greater risk of injury by nature of their still developing musculature.

Ongoing brain development also distinguishes the risk and response to injury in young and mature athletes. Brain water content, cerebral blood volume, myelination, skull geometry, and suture elasticity are related to maturation. A combination of these developmental factors may place children at increased risk for more widespread and prolonged cerebral swelling following head injury. Youths are also vulnerable to greater metabolic sensitivities compared to similarly injured adults. These differences are seen in experimental models of concussion that report a longer temporal window of vulnerability following trauma in younger animals and suggest that the brain may be particularly sensitive to additional injury during the acute recovery phase [10–14].

Although research examining the longer-term impact of pediatric concussion is limited, longitudinal studies of more severe traumatic brain injury (TBI ) in younger children show that insult occurring during infancy and preschool is associated with worse outcomes compared to injury sustained in later childhood or adolescence. Undeveloped or developing skills, particularly executive control processes, are particularly vulnerable [15–18]. Moreover, contextual factors including access to care and academic demands may impact recovery in younger athletes [3, 19–21].

In sum, ongoing development that occurs throughout childhood and adolescence may play a role in recovery from concussion. This has implications for both the risk and response to injury in the younger athlete. The clinical and neuropathological effects of sustaining repeat concussion are reviewed within this context.

Research indicates that recovery from a single concussion typically occurs within 2–14 days, although younger athletes may take longer [22–27]. However, there is evidence of acute differences between individuals with a prior history of concussion and those without. First, individuals with a prior history of concussion are likely to experience more severe markers of injury with subsequent trauma, including loss of consciousness, amnesia, and confusion [28]. Similarly, repeat concussion is associated with a higher initial symptom burden following injury such that individuals with a prior history of concussion endorse more symptoms and rate them as being more severe than individuals with no history of concussion [29, 30].

Studies also demonstrate different trajectories of recovery following repeat concussion. For example, a recent investigation by Eisenberg and colleagues [30] demonstrated that children presenting to the emergency department with a prior history of concussion are more likely to report longer symptom duration than those with no prior history. Similar findings have been shown in athletic populations [31]. For example, Guskiewicz and colleagues found that college football players with a self-reported history of three or more concussions experienced longer recovery times than players with no history of concussion [32]. In addition, prolonged impairment has been found to extend beyond self-report measures of function, with some studies demonstrating more chronic deficits on objective measures of postural stability following repeat concussion [33, 34].

Neurocognitive findings in athletes with a history of concussion are variable. The effects of a single concussion typically result in short-lived deficits in executive function, speed of information processing, attention, and memory [22, 35]. However, there is evidence indicating that concussed athletes who deny subjective symptoms demonstrate worse performance on neurocognitive measures compared to uninjured controls, but better performance than their symptomatic counterparts [26]. This finding suggests that subtle deficits signifying incomplete recovery may linger beyond reported symptoms, which raises concern that the risk of re-injury is greater than previously thought. It also highlights the utility of neurocognitive testing in determining full recovery from acute symptomatology following concussion.

While several studies have shown that repeat concussion is associated with more pronounced neurocognitive deficits[24, 36–39], others do not show any difference between those with a history of concussion and uninjured controls [40–44]. These inconsistencies highlight the heterogeneity of concussion, as well as some of the methodological limitations inherent to cross-sectional analyses. In the extreme (e.g., boxing), however, repetitive trauma is associated with more consistent findings of neurocognitive impairment, with data showing a dose–response relationship [45]. Readers are referred to the section entitled “Long-Term Effects of Chronic Repetitive Concussion” for a more detailed review of prolonged exposure to repeat injury.

The most consistently reported finding in studies examining athletes with a history of concussion is increased susceptibility to additional injury [32, 36, 46–48]. In high school and college athletes, the risk of re-injury for those who report a prior history of concussion is 2–5.8 times greater than in non-injured athletes [49]. Several factors have been suggested to explain this risk, including persistent neuropathological changes, personality/behavioral factors (e.g., aggressive playing style), position, or size and strength. Of these, the number of previous concussions and time elapsed since the most recent prior concussion may be the best predictor of repeat injury [30, 32]. Guskiewicz et al. [32], for example, found a dose–response relationship between history of concussion and new injury in their sample of 2905 college football players when controlling for division (i.e., NCAA I, II, III), position, years of participation, academic level, and body mass index. This study also indicated that the risk of repeat concussion was greatest within the first 7–10 days after an initial injury [32].

Eisenberg and colleagues’ [30] recent study of patients presenting to the emergency department with concussion confirms previous findings demonstrating a dose–response effect of previous injuries. Results also suggest a temporal vulnerability to prolonged recovery with repeat injury. Specifically, subjects who had suffered a concussion within the previous year experienced prolonged symptom duration following repeat injury, whereas subjects with a history of a single concussion occurring more than a year prior to a second injury were indistinguishable from those with no history of concussion. Similar findings are well documented in animal models of concussion [50, 51]. For example, Meehan et al. [51] showed that shorter intervals (1 day) between concussive injury in mice were associated with worse performance on measures of learning and memory compared to those with longer, 1 month intervals between injury or uninjured controls. Moreover, the mice that received daily concussions demonstrated reduced functioning 1 year post injury.

At its extreme, repetitive head injury within a short period of time can have devastating effects. Although presumed to be related to an extraordinarily rare cascade of events, second impact syndrome (SIS) is characterized by catastrophic brain injury occurring when an athlete sustains a second, typically benign head trauma while still symptomatic from a previous concussion [52]. Physiologically, SIS is thought to reflect cerebrovascular congestion or a loss of cerebrovascular autoregulation, which leads to increased intracranial pressure and, ultimately, brain herniation through the foramen magnum [53, 54]. To date, most cases of SIS have been reported in children, adolescents, and young adults, suggesting that the developing brain may be more vulnerable to the effects of this condition, although this may simply reflect the larger numbers of athletes in this age group participating in collision sports. Other investigators have questioned the existence of SIS [55–57] and suggested that the syndrome reflects malignant brain edema, a rare but equally devastating effect of mild brain trauma in youths that has been documented in neurosurgical literature [58]. Despite the debate about SIS and its extremely low incidence, experts generally agree that athletes should be removed from play until all symptoms have resolved to prevent its occurrence [52, 55].

Although useful in ruling out more serious pathology, traditional neuroimaging techniques [e.g., computed tomography (CT) or magnetic resonance imaging (MRI)] are unable to detect concussion [7, 59]. However, functional imaging studies reveal differing patterns of activation in patients diagnosed with concussion compared to uninjured controls [60–63]. Of these, investigations utilizing functional MRI (fMRI) have consistently shown physiological differences following concussion that tend to correlate with self-reported symptoms. Similar findings are documented in studies involving positron emission tomography (PET) or single-photon emission computed tomography (SPECT). None of these studies has looked specifically at the effect of repeat concussion. However, in one investigation examining the acute effects of a single concussion on metabolic function with magnetic resonance spectroscopy (MRS), participants who sustained a second injury during the study period showed more prolonged metabolic dysfunction compared to those who sustained only one concussion [64]. Although speculative, a similar trajectory of prolonged hypo- or hyper-activation may exist in patients who continue to experience symptoms following repeat concussion.

Emerging brain imaging techniques capable of detecting white matter integrity have demonstrated structural injury in the absence of findings on CT or MRI studies. The most consistent findings are reported in studies utilizing diffusion tensor imaging (DTI). This technique measures directionality and regularity of white matter tracts, and it is particularly sensitive to axonal injury. Initial findings indicate that both neurocognitive test performance and self-reported symptoms are associated with DTI findings [65, 66]. Moreover, as with other studies examining repetitive concussion, current research suggests a dose–response relationship such that the extent of white matter or axonal damage may be associated with exposure to concussion [67, 68].

There is also evidence of neurophysiological dysfunction following concussion. While standard electroencephalogram (EEG) readings have shown some utility in detecting change immediately following head injury [69], long-term effects are shown less consistently. However, variants of EEG that are sensitive to cognitive and sensory processing, including evoked potentials (EP) and event-related potentials (ERP), have demonstrated a cumulative effect of multiple concussions [70–72]. Consistent with studies utilizing neurocognitive function as an outcome, there is evidence that altered brain function seen in ERP studies may extend beyond reported symptoms [71–73].

Investigations involving athletes with prolonged exposure to repetitive head injury indicate a risk of significant longer-term neurobehavioral and neuropathic sequelae [74–76]. Martland [77] was the first to describe what was initially referred to as punch-drunk syndrome in his review of boxers who presented with confusion, loss of coordination, problems with speech, and upper body tremors [77]. More recently, neuroanatomical investigations of similarly injured boxers have revealed cerebral atrophy, cortical and subcortical neurofibrillary tangles (NFTs), and cellular loss in the cerebellum with some consistency [74, 78]. Not unexpectedly, these neuropathic changes are often accompanied by neurocognitive impairment, with most studies suggesting a dose–response relationship wherein the degree of impairment is associated with the number of bouts fought [45]. Extrapolating from data collected in boxers, the genetic protein apolipoprotein E (apoE) є4 allele may be a risk factor for development of neurobehavioral and neuropathic impairment in those exposed to chronic repetitive trauma [79].

Within the past decade, scientific inquiry has extended beyond boxers in its effort to gain a better understanding of the cumulative effects of concussion. Much of this followed literature suggesting a greater risk of Alzheimer’s dementia [76], suicidality [80], and depression [81] in retired professional football players. Methodological limitations in studies to date have prompted investigators to push for more well-controlled, prospective epidemiological studies to determine the true risk of chronic repetitive concussion [80, 82, 83]. Of particular interest to researchers and athletes is the constellation of symptoms first described in 1928 by Martland and later detected neuropathically by Corsellis in 1973, chronic traumatic encephalopathy (CTE) .

As currently understood, CTE represents a progressive neurodegenerative disease associated with repetitive symptomatic and asymptomatic brain trauma. Diagnosis is based upon the presence of uniquely distributed tauopathy at postmortem neuropathological examination [75, 84–87]. As noted, while early descriptions of the disease focused on boxing, recent investigations extend to a more diverse group of individuals exposed to repetitive head impacts, most notably American football players.

Clinical Presentation: Studies documenting the clinical presentation of CTE are somewhat variable, as they rely on proxy reports of the patient’s premortem behaviors. However, a pattern tends to emerge of progressive cognitive deficits, irritability, aggression, chronic headaches, slurred speech, and parkinsonism [88–90]. Within this broad cluster of neurobehavioral/cognitive symptoms, a recent investigation describes two distinct clinical presentations of CTE [91]. One subset is characterized by more prominent behavioral or mood changes whereas the other is more likely to exhibit cognitive symptoms. Age of onset between these subtypes is reported to differ such that the type involving behavior or mood changes emerges earlier than the type involving cognitive symptoms, although all forms of CTE are thought to eventually result in cognitive impairment.

It is important to note that CTE is distinct from acute or prolonged symptoms of concussion. The clinical symptoms of CTE typically do not manifest until one to two decades after retirement from contact or collision sports [87], although there are documented cases in younger athletes showing symptoms consistent with the disease [75]. Furthermore, there has not been any clear evidence showing a relationship between prolonged acute concussion symptoms and CTE [83, 92].

Following the suicide of several high-profile professional athletes, some have posited a higher risk of depression and suicidality in those with exposure to repeat concussion. There is specific concern that there may be an association with CTE; the media has reported suicide as the cause of death in six retired NFL players between 2011 and 2013 [80]. While clearly worrisome, there is little empirical support showing a direct relationship between CTE and suicidality [80, 92]. To the contrary, there is evidence that the rate of suicide is higher in the general public than in retired professional football players, where it is well established that there is a high risk of repeat head injury. In a large cohort of retired NFL players, completed suicide deaths due to “intentional self-harm” were observed in 9 per 100,000 compared to 21.8 per 100,000 in the general population [93]. In nearly all cases, the risk factors associated with suicide, including depression, substance abuse, violence, and disinhibition, make it difficult to establish a causal relationship, and this remains true for CTE [80, 82, 83].

Neuropathology: CTE is characterized by gross neuropathological findings. The hallmark features include cerebral atrophy, enlargement of the lateral and third ventricles, thinning of the corpus callosum, large cavum septum pellucidum with fenestrations, and scarring and neuronal loss of the cerebellar tonsils [74]. At a microscopic level, CTE is characterized by uniquely distributed tauopathies, including NFTs, neuropil threads (NTs), and glial tangles (GTs) [90].

Tau i s an intracellular protein found primarily in neurons of the central nervous system. It functions to support and strengthen the architecture of microtubules. When defective, tau collects abnormally and causes neuronal dysfunction. Tau depositions themselves are not unique to CTE, with 1 review citing 20 different neuropathic conditions involving tauopathies [83]. However, there is a unique pattern of tau distribution in CTE that includes the entorhinal cortex, hippocampus, and neighboring cortical areas [87, 94]. Although beta-amyloid (Aβ) deposits have also been identified in cases of CTE, they are significantly less than the Aβ deposits that characterize nearly all cases of Alzheimer’s disease [83]. In 2013, McKee and colleagues published a large case report on individuals with neuropathologically confirmed CTE, presenting proposed criteria for four stages of CTE pathology based on the severity of the findings [88, 95]. Validation of the reliability of this staging system for diagnosis of CTE is currently being performed.

Risk Factors: At this time, the most consistently reported risk for developing CTE is exposure. In boxing, where the goal is to induce a concussion in one’s opponent, the risk of exposure to repetitive trauma (i.e., head impact) is obvious. However, literature also shows an impressive risk of head impact in other sports, as well. For example, in American football, reports suggest as many as 1400 head impacts per season for offensive lineman and 2000 for bidirectional players [96–101]. In addition to type of sport played, position and duration of involvement are closely tied to exposure [98, 102, 103]. While it is generally accepted that CTE results from repetitive injury, this is not without some debate. Some authors have suggested that the disease can result from a single impact [104]. In sports with purposeful contact, this has implications for position played and the risk of CTE. For example, data on direct measurements of head impact exposure in college football indicated that compared to other players, running backs and quarterbacks suffer hardest and most severe blows to the head, while linemen and linebackers suffer a greater number of head impacts during a game [98].

As noted, previous investigations have shown that genetic factors may be an important risk factor in the development of CTE. Studies have linked the apolipoprotein E (APOE) gene to worse cognitive functioning in athletes and prolonged recovery following a single head injury [79, 104, 105]. The APOE є4 allele is the largest known genetic risk factor for sporadic AD and has been associated with Aβ, but not tau deposition in cognitively normal aging. There is suggestion that it could also be related to onset of CTE in retired athletes. Other health-related variables may also contribute to the neurodegeneration and clinical symptom spectrum associated with CTE including chronic inflammation associated with obesity, hypertension, diabetes, and heart disease [106].

Age has also been noted as a possible risk factor [87]. Previous literature has indicated that the increased plasticity of a younger brain may allow an individual to better compensate and recover after head injury [107]. However, more recent research indicates that a younger brain may be more susceptible to diffuse brain injury, which could lead to more pronounced and prolonged symptoms, particularly cognitive deficits, over time [27]. A recent study examined the relationship between exposure to repeated head impacts through tackle football prior to age 12 in former NFL players. Results indicated that those whose first impact was before age 12 performed significantly worse on measures of executive functioning, verbal memory, and reading, after controlling for total number of years of football played and age at the time of evaluation [108]. It is possible that early, repeated exposure to head injury could increase an individual’s risk for neurocognitive difficulties later in life and potentially lead to CTE. Well-controlled, longitudinal research is needed in this area.