In the United States, over 300,000 sports-related concussions occur annually, and the likelihood of suffering a concussion while playing a contact sport is estimated to be as high as 19% per year of play. More than 62,000 concussions are sustained each year in high-school contact sports, and among college football players, 34% have had one concussion and 20%, multiple concussions. Concussions often cause significant and sustained neuropsychological impairments in information-processing speed, problem solving, planning, and memory, and these impairments are worse with multiple concussions.

Suffering a second concussion while still having symptoms from a previous concussion can be lethal. In 1973, Schneider was the first to describe the deaths of two athletes who died after suffering a relatively minor head injury during recovery from a previous concussion. In 1984, Saunders and Harbaugh reported the same scenario in a 19-year-old college football player and coined the term "second-impact syndrome" (SIS). Since then, at least 26 deaths have been attributed to SIS, 20 of them occurring in the past 10 years.

Physicians, trainers, and others who care for injured athletes disagree about the amount of time that must pass before an athlete with a concussion can safely return to play or the potential danger of returning too soon. Many athletes resume play prior to resolution of cognitive deficits because, until recently, there were no valid tests available to assess cognitive function during recovery. Thus, there are no widely accepted evidence-based guidelines that coaches or trainers can use to be sure the athlete has recovered. At least 14 different concussion scales have been published since 1973, with widely disparate criteria for grading concussion severity and resuming athletic activity. Moreover, because post-concussive symptoms can be quite subtle, coaches and athletic trainers often miss the diagnosis. In fact, physicians and sports medicine researchers do not even agree on the definition of "concussion." Previous attempts to objectify the diagnosis of concussion or post-concussive syndrome using multiple concussion scales, computed tomography (CT), magnetic resonance imaging (MRI), and EEG have been unsuccessful.

Most experts believe that the signs and symptoms of concussion are related to a metabolic dysfunction in the inferior parietal, prefrontal, and cingulate cortex. Decreased cerebral blood flow, hyperglycolysis, glutamate-induced excitotoxicity, and abnormal cellular ionic fluxes occurring after TBI have been implicated as the cause for this dysfunction. Recent studies have suggested that delayed metabolic dysautoregulation is caused by excitatory amino acid-induced ionic shifts with increased Na/K ATPase activation and resultant hyperglycolysis. The decreased cerebral blow flow that occurs with post-traumatic dysautoregulation is not well understood. Such dysautoregulation may not be seen until 2-3 days after injury and often persists for more than a week. It has also been postulated that metabolic dysregulation, until fully resolved, may make the brain more vulnerable to a second injury, thus explaining the severe neurological dysfunction or death when a second impact occurs before these abnormalities resolve. These hypotheses, however, are based primarily on metabolic and physiologic information obtained from animal experiments or studies of severe TBI in humans. Quantitative, and possibly also qualitative, differences likely exist ior victims of concussion.

The neuropsychological deficits that result from sports-related concussion have been extensively reported. Barth et al. gave baseline neuropsychological tests to 2300 football players from 10 college teams. Approximately 200 players suffered concussions during the period of the study. Re-administration of the tests to each injured player at 24 hours, 5 days, and 10 days showed subtle post-concussive impairments in their ability to process new information, which typically resolved by the tenth day. Macciocchi et al. compared cognitive function and post-concussive symptoms between 183 college athletes with concussions and age-matched control subjects and found impaired performance and increased headaches, concentration difficulties, and behavioral problems in the injured group. Among 33 amateur soccer players, Matser et al. found moderate to severe impairments in memory and planning in 27%, compared with only 7% of swimming or track athletes. They attributed the high rate of deficits in soccer players to concussions sustained during "heading" of the ball. Jordan et al. found impaired concentration, attention, and memory in all 42 boxers they studied. The degree of cognitive dysfunction was proportional to the boxer's sparring exposure, a finding that supports the concept that multiple concussions have a cumulative adverse effect on cognitive function. Collins et al. studied post-concussion recovery in almost 400 athletes. They reported that athletes who suffered multiple concussions performed more poorly on neuropsychological tests and were more likely to have prolonged learning difficulties than those with a single or no concussion.

All of these studies have been of college or professional athletes, and we have not found any studies that examined the effects of concussion in high-school athletes. And yet, most at-risk athletes are at the high-school level or younger. Moreover, the majority of the 17 athletes who died of SIS between 1992-1997 were high-school students. Although current return-to-play guidelines are applied identically to all age groups, no study has examined whether vulnerability to concussive injury differs with age. Such age-related differences are suggested, however, by our previous studies, which showed significant post-concussion differences on standard neuropsychological tests between high-school and college athletes

These age-related neuropsychological differences may well be related to previously described differences in post-traumatic pathophysiology in the adolescent versus the adult train. Children and teenagers are more likely than adults to have diffuse injury and prolonged brain swelling. This may in part be related to the fact that the immature brain is approximately 60 times more sensitive to glutamate-mediated N-methyl-D-aspartate excitotoxic brain injury. Therefore, high-school athletes might be expected to have a slower recovery than college-aged or older athletes and to be more susceptible to severe neurological deficits should they be re-injured during recovery. On the other hand, some argue that younger athletes should have a greater potential for recovery after concussion because of their greater potential for cortical reorganization compared with adults. Studies comparing functional outcome after hemispherectomy found that younger animals had a more complete functional recovery than older ones. This finding supports previous clinical evidence of marked synaptic excess in children, relative to adults, allowing for neural pathway rerouting during recovery and functional plasticity in the developing brain.

In addition to age, gender appears to affect outcome after a concussion Females with mild TBI are more likely than males to report sleep disturbances and headaches, and are less likely to be employed or in school 1 year after injury. In their study of individuals who presented to an emergency departrnent after a concussion, Bazarian, et al. also found that females were more likely than males to have post-concussive syndrome at 1, 3, and 6 months after injury, though these differences reached statistical significance only at months. The authors attributed these differences to significant differences in the mechanism of injury: more females were involved in motor-vehicle collisions (57% vs. 22% of males:), while more males had sports injuries (33.3% vs. 8.6% of females). Gerberich evaluated the effects of TBI on college academic performance. Individuals with TBI were compared to uninjured controls and to controls hospitalized for other types of injuries. Concussion led to a significant decrease in grade point average (GPA) for the females compared to their controls; this was not observed in the males who had a concussion. Another study looking only at those who were working before injury showed no gender differences in return-to-work rates, but significant differences between males and females in reported capacity for unrestricted employment' as measured through the DRS employability subscale. Most studies to date have found that even after controlling for other demographic, premorbid, and event-related factors, women have worse outcomes than men after a concussion, and a recent meta-analysis involving 8 studies and 20 outcome variables showed outcome was worse in women than men for 85 % of the variables studied.

In contrast to the literature citing more symptoms and poorer outcomes for females after TBI, animal studies suggest that female hormones may play a neuroprotective role after TBI. Roof found that endogenous or synthetic progesterone caused a greater reduction of cerebral edema in female rats compared to males. Progesterone also has been shown to limit lipid peroxidation, facilitate cognitive recovery, and reduce secondary neuronal loss after cortical contusion in male rats. Estrogen appears to have a protective effect in males' but it increases mortality and worsens motor function in females. Other studies have found that estrogen maintains normal cerebral blood flow and, when given acutely, reduces mortality after experimental TBI. These findings are thought to be due to estrogen-mediated neuroprotective mechanisms such as antioxidant effects and reduction of levels of excitotoxic mediators of secondary brain injury.

In clinical drug trials, gender differences have been reported for aspirin efficacy in stroke prevention and for tirilizad efficacy in the treatment of subarachnoid hemorrhage. These drugs were found to be much more beneficial for males than females. However, studies evaluating gender-dependent differences in the treatment of TBI are largely non-existent.

Thus, many questions remain regarding the pathobiology of concussion, the effects of age and gender, and the time course for recovery. One approach would be to examine brain regional activity during cognition as a function of concussion. Functional MRI (fMRI) is a noninvasive means of imaging changes in local cerebral blood flow and in brain activation that occur with cognitive or physical activity. fMRI monitors brain function with a relatively high degree of spatial and temporal resolution. The technology is based on the sensitivity of MRI to magnetic effects induced by the modulation of the oxygenation status of hemoglobin (oxy/deoxyhemoglobin) that results from local variations in blood flow. It has been shown that such variations can be induced by task activation or cognitive processes such as language or mental imagery. The MRI signal is increased by a few percent when physical or mental activity activates a region of the brain, causing a sharp increase in local blood flow and oxygen utilization. The functional activation of brain regions as recorded by fMRI has been shown to correlate quite well with regional brain electrical activity as recorded using event-related evoked potentials, and with regional glucose metabolism as detected with positron-emission tomography (PET).

To date, only one study has been published describing the use of fMRI in patients with TBI. McAllister et al. compared brain-activation during a working-memory task (n-back) between 11 healthy subjects and 12 patients who had a concussion within the past month. They found activation by the two groups to be significantly different for working-memory circuitry in response to different processing loads. In the control subjects, stimulation with a low processing load led to maximum activation in the frontal and parietal lobes, and more difficult tasks produced very little increase in activation in these areas. For the 12 concussion patients, the level of activation stimulated by the low processing load was slightly less than in controls, but the level of activation in the frontal and parietal lobes was significantly higher than for control subjects with the high processing Loads. The investigators interpreted their results as suggesting that patients with a concussion perceive a change in their ability to engage working memory, and they experience this change as having to work harder to maintain accurate test performance. These findings also suggest that injury-related changes in the ability to modulate memory processing may underlie the typical findings of worsened memory after a concussion.

Other diagnostic modalities have shown that regional metabolic abnormalities correlate with cognitive dysfunction. Fontaine et al. used PET to measure regional glucose metabolism in 13 TBI patients with deficits in memory, attention, executive function' or behavior. They found cognitive and behavioral disorders to be closely linked to decreased metabolism in the prefrontal and cingulate cortex. Results of memory and executive function tests correlated with regional metabolism in the mesial and lateral prefrontal cortex and the cingulate gyrus. Behavioral disorders correlated with mesial prefrontal and cingulate metabolism.

In summary, mild TBI (concussion) causes significant cognitive deficits that appear to be worse and more prolonged with repeated injury. The few imaging studies done to date implicate abnormal function or metabolism in the prefrontal, parietal, and cingulate cortex as likely causes for these cognitive deficits. However, there have been no large prospective studies that combine cognitive testing with a functional imaging study to elucidate the relationship between specific post-concussive cognitive deficits and abnormalities in various brain regions. In this regard, several studies suggest that there may be significant differences related to age and gender, but it is not clear if the prognosis is better or worse for younger as compared to older individuals or males versus females. As a result, current management of mild TBI is relatively arbitrary and extremely variable. With no standard criteria to assess cogrutive damage or recovery, high-school and college athletes who suffer a concussion may return to play too soon and are at risk for SIS. Students who return to school too soon after a concussion may be unable to work to their full potential, become discouraged because of a decline in academic performance, and unnecessarily limit their long-term goals.

There is a need for a large prospective study that will more clearly define the patholaiology of concussion and identify the differences in neuroanatomic and neurophysiologic recovery from concussion that are related to time after injury, age, and gender. The results of such a study would provide the scientific basis for the formulation of guidelines concerning who can safely return to work, school, or athletic competition after one or repeated concussions. Such guidelines may well be found to be different for individuals of different ages, or for males and females. Information derived from such a study also could be expected to guide the development of therapies aimed at preventing or limiting the symptoms of post-concussive syndrome.