Abstract
Three adolescent football players who had ischemic stroke associated with football practice and play are described. The literature on stroke associated with childhood sports and football in particular is reviewed, and the multiple mechanisms by which football can contribute to ischemic stroke are discussed.
The incidence of childhood stroke in the United States is between 2 and 3 out of 100 000 children 1 month to 18 years old, and most are ischemic. 1 When perinatal stroke is included, the incidence is approximately 2.0 per 100 000 person years. 2 The risk factors and mechanisms of pediatric stroke are not fully understood. American football as played by American children has been linked to multiple types of catastrophic neurologic injury, including ischemic stroke. 3 The authors describe 3 adolescents who had ischemic strokes associated with football play, review the literature on pediatric stroke associated with football and other sports, and discuss possible mechanisms leading to sport-associated infarction.
Case Reports
Case 1
A 17-year-old boy who played the fullback/safety position on his high school football team began to experience confusion, headache, nausea, vomiting, and difficulty understanding speech during football practice. Off the field, he had difficulty understanding text messages. The boy and his parents did not recall any significant head impacts during football play that day. The symptoms lasted approximately 36 hours.
His past medical history was remarkable for a concussion sustained during snowboarding without a helmet 18 months earlier; a baseball injury which shattered 1 eye socket and required facial reconstruction 8 years earlier; and a benign connective tissue tumor in 1 arm which was removed surgically. Family history was remarkable for a paternal grandmother who had what sounded like pulmonary emboli, was found to have the methylene tetrahydrofolate reductase C677T gene variant, and was on warfarin sodium.
He went to see his primary care doctor because the symptoms had persisted more than 24 hours. The physician ordered head computed tomography, which showed an 18-mm hypodensity in the right cerebellar hemisphere. Magnetic resonance imaging was performed several days later to clarify the diagnosis, and showed a 0.5 cm × 1.3 cm lesion in the right cerebellar hemisphere which was bright on fluid attenuated inversion recovery and T2 sequences but not visible on T1 until contrast was given; the outside differential diagnosis was tumor versus vascular malformation versus infectious lesion. He was sent to our emergency room.
Neurological examination at our emergency room (approximately 2 weeks after the event) was unremarkable, including normal finger-to-nose and heel-to-shin testing. Follow-up magnetic resonance imaging and spectroscopy was ordered for 1 month later. Follow-up magnetic resonance imaging showed a nonenhancing 11-mm linear lesion that was bright on fluid attenuated inversion recovery and T2 but did not enhance, and spectroscopy revealed slightly elevated lipids and normal choline and lactate, consistent with old infarct; in retrospect the study 1 month earlier was consistent with subacute infarct. He was placed on daily baby aspirin. No dissection was seen on computed tomography angiography, performed twice over several months. The patient’s prothrombotic workup revealed only heterozygosity for the methylene tetrahydrofolate reductase A1298C mutation and an elevated lipoprotein a that decreased to just above normal when checked fasting after dietary modification. His blood pressure was consistently slightly elevated over several clinic visits with systolic blood pressure in the range of 130–150 mm Hg and was controlled with lisinopril after unrevealing workup by nephrology. He experienced no long-term deficits from the stroke and remained on daily aspirin.
Case 2
This 15-year-old boy developed a headache; right face, arm, and leg weakness; and dizziness, after blowing his nose and feeling a “pop.” The patient was an offensive lineman on his team and had played football 4 days before the incident. The family did not recall any significant head impacts during play. Past medical history was otherwise remarkable only for mild asthma. Head computed tomography at an outside hospital was unremarkable, but magnetic resonance imaging revealed a left pontine stroke. Computed tomography angiography of the head and neck did not demonstrate dissection or other vasculopathy. He was started on intravenous heparin within 8 hours of symptom onset.
Initial workup was remarkable only for a low high density lipoprotein level of 28 (normal >40). The patient by report had 0/5 strength in his right upper and lower extremities for 3 days, then started to recover. He was discharged to rehabilitation on clopidogrel, atorvastatin, and a combination of vitamins B6, B12, and folic acid.
Two months after his stroke, he was ambulating independently, and his neurological examination was remarkable only for mild-moderate right-sided symptoms, with slightly increased muscle tone; very mild weakness in his right arm and leg, with slightly more weakness in the right hand interossea (4/5) and decreased fine motor skill; mild difficulty with reaching for a target and doing rapid alternating movements with his right hand; brisker reflexes in the right arm and leg with positive Babinski; and mild difficulty with tandem gait. Additional prothrombotic workup revealed heterozygosity for both the methylene tetrahydrofolate reductase gene variant C677T and the plasminogen activator inhibitor 1-4G, with a persistently increased lipoprotein a level. Cardiac echocardiogram showed normal structure and function. He was originally maintained on clopidogrel, niacin, and atorvastatin. He was unable to tolerate the niacin and was switched to fish oil and a baby aspirin. He was noted to have elevated blood pressures with systolic pressures often in the 130s–140s. While he initially responded to nonpharmacological management, he eventually required lisinopril. At last follow-up 2 years after his stroke, he had only difficulty hopping on the right foot (he could stand on it), and slightly brisker reflexes on the right.
Case 3
The details of this case have been previously published. 4 This 14-year-old boy was an offensive lineman on his high school football team. He collapsed at the end of a football game and was initially unresponsive. In the emergency room, he had flaccid right hemiparesis and aphasia and was transferred to an intensive care unit in a second hospital, where he improved, with mild right facial weakness, pronator drift, and mixed aphasia. Magnetic resonance imaging and angiography showed scattered left middle cerebral artery-distribution infarcts and what appeared to be thrombus and dissection at M1. Conventional angiography 9 hours after his collapse showed a 95% blockage in the left M1.
A clot retrieval device was used after thrombolytics and anticoagulation failed, but the device tip broke off in the left middle cerebral artery, leading to a 100% occlusion. He was then transferred to our hospital, where, when he awoke from sedation, he had flaccid right hemiparesis, left gaze preference, and expressive aphasia. Prothrombotic workup showed heterozygosity for prothrombin 20210, methylene tetrahydrofolate reductase A1298C gene variant, and plasminogen activator inhibitor 1 4G gene variant. Cardiac echocardiogram was unremarkable. The patient was on a heparin drip until computed tomography showed small hemorrhage, then all blood thinners were stopped. He was started on baby aspirin 2 ½ months after discharge.
At last follow-up a year and a half after his stroke, his hemiplegia had improved significantly. He was able to ambulate independently and had only mild right-sided weakness, with some difficulty opening and closing the right hand and very limited fine finger movements. However, while his language showed some improvement, he continued to have significant aphasia, and was unable to read simple sentences, follow complex commands, or name many common objects.
Discussion
In each of the above cases, ischemic infarction was temporally related to engaging in full-contact football. Multiple mechanisms may have contributed to these infarctions, including several that may have led to delayed presentation after trauma, as in case 2.
Multiple risk factors for pediatric stroke have been identified, with many patients having more than 1 risk factor. 5 Previously identified pediatric stroke risk factors include structural cardiac disease, congenital or acquired intracranial/extracranial arteriopathies (including dissection), head trauma, hematologic disorders, genetic disorders, malignancy, and infection. 1,6 Arteriopathy appears to be a common contributor to childhood stroke. In 1 large retrospective analysis of 185 pediatric stroke patients who underwent cerebral arterial imaging, 79% were found to have vessel abnormalities: proximal large artery narrowing or occlusion in 65%, moyamoya in 18%, dissection in 10%, arteritis in the remainder. 5 The high incidence of cerebrovascular abnormalities in children with pediatric ischemic stroke has been reported several times. 6–9 Early studies of pediatric stroke may have underestimated the importance of vascular abnormalities since vascular imaging was not performed for many cases of “idiopathic” or “spontaneous” pediatric ischemic stroke. Evidence now suggests that, for most cases of pediatric ischemic stroke, imaging of both the brain and cerebral vasculature should be part of the routine workup. 2,6 It is important to identify any underlying arteriopathy: 1 large population-based study found that arteriopathy was associated with stroke recurrence, with recurrence rate as high as 66% within 5 years. 2
Trauma leading to arterial dissection with resulting thromboembolism to the brain is suspected to be an important mechanism leading to stroke triggered by football and by other sports. The degree of force required for dissection is not completely clear. There are many reports of pediatric sports-related strokes—many shown to be secondary to dissection—in both the medical literature and popular press. 10–14 Childhood strokes have been temporally associated with soccer, 15,16 wrestling, 17 lacrosse, 15 basketball, 18,19 bicycling, 19 karate, 15 tobogganing, 17,19 and football. 20–23 However, diagnosing dissection can be very challenging. Presentation with stroke after dissection can be delayed: in 1 study of children with dissection leading to ischemic stroke, the time interval from trauma to symptom onset ranged from 20 minutes to 4 days, with a mean of 1.5 days. 17 Dissection can be a challenging imaging diagnosis to make in some cases, depending on the size and characteristics of the dissection, and the imaging modalities used. 24 Many investigators consider digital subtraction angiography to be the gold standard in diagnosing dissection; however, MRI and magnetic resonance angiography have been increasingly used as a diagnostic method of choice. 8,9,25,26 In the past several years, computed tomography (CT) angiography has emerged as an equivalent method for diagnosing dissection. 24 Some have recommended magnetic resonance angiography over CT angiography in pediatric patients, since it does not use radiation and is less invasive for children able to cooperate with the scan. 6 However, institutional preferences vary. 27 Dissection was suspected in all 3 cases described above, but only demonstrated in 1 case (case 3). Trauma to vessels below the cerebral vasculature can also lead to stroke. One high school football player had a stroke 3 months after a helmet injury to the right shoulder. He had developed a brachiocephalic pseudoaneurysm, which thrombosed, and embolized to his brain as he warmed up for a game. 28
Organized children’s football begins early, with full contact leagues beginning at age 5 years. 29,30 In toddlers and children, trivial head trauma can be associated with infarction even without demonstrated arterial dissection. 31,32 Although it is possible some of these children may have had subtle dissection missed on imaging, there may have been other mechanisms leading to infarction. The cerebral vasculatures of small children may be particularly susceptible to stretching forces leading to vascular injury, since children’s heads are relatively large in proportion to their bodies, and have high moments of inertia. In young children, even mild trauma might lead to shearing of terminal branch cerebral arterial vessels which leave larger vessels (particularly the middle cerebral artery) at near perpendicular angles before brain development is complete. 31,32 These shearing forces may lead to temporary disruption in blood flow secondary to spasm or endothelial damage leading to clot formation, with infarction often affecting the basal ganglia. 31,32
Multiple head impacts with subclinical symptomatology may be a risk factor for brain ischemia in football players, possibly due to long-term vascular and neuronal changes induced by repeated trauma. A recent study involving real-time monitoring of children playing football has found that players sustain mild cognitive changes and have cerebral metabolic changes that can be demonstrated on functional MRI as they accumulate multiple head impacts during routine football play over the course of a season, even without clinical evidence of concussion. 33 Autopsy studies of professional football players and boxers show an accumulation of tau proteins and beta-amyloid deposition. 34,35 Tau proteins can be markers for ischemic injury. Tau proteins increase in cerebral spinal fluid as a result of cerebral ischemia in surgery patients, 36 and are found in pathologic brain specimens after cerebral ischemia. 37 Certain genetic factors, such as having the ApoE-ϵ4 allele can predispose to tau deposition with mild head trauma. 38,39 A recent study showed that multiple blows to the head of less-than-concussive force leads to a release of neurofilament proteins into the cerebral spinal fluid similar to that found in minor brain infarctions. 40,41 Brain autopsy data in young athletes is limited, so it is not possible to know how early permanent changes take place; however, an autopsy of an 18-year-old high school football player who sustained multiple concussions revealed focal deposition of tau proteins around small blood vessels in the frontal and insular cortex. 42 It is not clear to what extent tau and other proteins found in this young football player’s brain were due to ischemic injury, acute cell breakdown at the time of trauma, cytotoxic infarction due to postimpact changes, or other mechanisms. Some authors have described the “second-impact syndrome” where brain edema and infarction occurs when the head is hit a second time in the period shortly after an initial concussion. 43,44 Although the existence of the condition has been disputed, 45 many authors agree the condition does exist. 46 Studies of repeated trauma similar to concussion in in vitro cultured mouse hippocampal cells reveal evidence of microscopic neuronal changes after a second injury. 47 An important area for future work is determining how the repeated mild traumatic brain injuries affect the pediatric vasculature and its ability to perfuse the repeatedly injured young brain.
This study proposes several other possible mechanisms which might contribute to stroke after football injury. Hyperventilation may lead to infarction by vascular constriction in children with pre-existing vasculopathy such as moyamoya disease, 48 and in those with sickle cell disease. 49 If a child playing football has an undiagnosed mild vasculopathy which is exacerbated by hyperventilation due to heavy exertions and/or extreme pain during play, possibly with some dehydration due to heavy sweating, this might lead to stroke associated with football. Paradoxical embolism through a patent foramen ovale is a potential mechanism for stroke in children and young adults. 50 Child and adolescent football players experience repeated injury to the entire body, not just the head, and often valsalva many times during a game because of exertion and pain. It is possible that a young football player with some degree of hypercoagulability, a deep venous thrombosis in the leg secondary to trauma, and a patent foramen ovale could then send an embolus to the brain to cause a stroke. While the authors did not find a child football player with stroke that appeared to be due to this mechanism in the medical literature, 1 was reported in the popular press. 51 In the adult medical literature, multiple adult divers and a weightlifter with stroke due to this mechanism have been reported. 52–54 In addition, the forces experienced by childhood football players continue to increase as the weight of the United States population increases. In 2 recent articles encompassing close to 4000 youth football players in the Midwest, over 45% were categorized as overweight or obese by national criteria. 55,56 Although recent trends have not been examined, from the mid-1970s to 1990, the average body mass index of elite high school football players rose significantly. 57 Another potential risk factor involves the use of anabolic steroids, which 3% to 12% of high school males admit to using. Anabolic steroids can affect the clotting cascade, increase triglycerides, and cause cardiomyopathy, all of which are risk factors for stroke. 58 Case reports of young adult athletes developing strokes while taking anabolic steroids exist, but none specifically related to high school football, although some young athletic stroke patients will not disclose this risk factor. 59 However, the popular press reported a case of a high school football player who suffered a stroke that was attributed to an anabolic steroid-containing supplement. 60 Another possible risk factor is the use of energy drinks by student athletes. These drinks can contain large amounts of caffeine; high doses of caffeine have been associated with stroke, 61 possibly due to vasospasm. In both energy drinks and nutritional supplements, caffeine is sometimes combined with other stimulants which can further increase the risk of vasospasm and stroke. One college varsity football player had a stroke during training, several hours after taking several tablets of a ma huang-containing supplement. Lab analysis revealed that the supplement contained a mixture of ephedrine, pseudoephedrine, methylephedrine, and caffeine. He had moderate vasospasm of the middle cerebral artery on transcranial Doppler ultrasound, and right middle cerebral artery territory infarction on follow-up computed tomography. 62 Two of our subjects had mild hypertension, but were too young to have had the many years of exposure that would lead to chronic vascular injury.
Among sports, American football has a particularly high rate of repeated neurological injury, and severe neurological injury. Approximately 4% of American high school football players receive a mild traumatic brain injury per year; with 1.5 million high school football players this results in 60 000 mild traumatic brain injuries a year. 63 This number is likely a gross underestimate since it is estimated that only approximately 50% of high school football-related mild traumatic brain injuries are reported. 64 The National Center for Catastrophic Sport Injury Research (NCCSI) has collected data on football since 1965 and has expanded coverage to many high school, collegiate, and professional sports. 65 From 1989 to 2009, there have been 440 recorded cases of football play-related cervical spinal cord or cerebral injury with incomplete neurologic recovery. Most of those injured are high school football players since there is such a large number of these players. Many more cases exist with eventual complete neurologic recovery. 22 Most of the injuries documented occurred to defensive backs during tackling. 22 Neurological injuries during football can include impingement of the cervical spinal cord secondary to fracture, intracranial bleeding, cerebral swelling/increased intracranial pressure, or ischemic stroke. In the 2009 national survey of catastrophic football injuries, there is 1 documented case of a high school football player who suffered an ischemic stroke secondary to game play. 22 In 2008, 2 players suffered ischemic strokes secondary to game play according to the national data. 21,22 There may be many more football-associated strokes than those published in the above reports, since symptom presentation can be delayed in cases of dissection or paradoxical embolism. Reporting of football injuries on a national level has led to increased safety precautions, with rule changes, improved protective equipment, and increased monitoring of players for signs of serious injury. 3
There are not clear recommendations as to whether children who have had strokes should play certain sports. In a survey sent to pediatric neurologists, 65% of respondents recommended that a teenager with a provoked dissection and subsequent stroke with resolution of stroke symptoms should never play American football. 66 However, only 40% of respondents recommended that a teenager with idiopathic stroke, normal cerebral vessels, and resolution of stroke symptoms should never play American football. 66 The discrepancy between these two theoretical cases may be due to the fact that at least 1% of patients with prior dissection will have another dissection in the future, 67,68 and the presence of any vessel abnormality on imaging can be a risk factor for a subsequent stroke. 2 Thus football may be a particularly high-risk activity for children with known cerebrovascular abnormalities.
With the number of head contacts each player experiences during a “junior” or high school football season ranging from the hundreds to thousands, 33 and our current incomplete knowledge of the pathogenesis of pediatric ischemic stroke, it is possible that this degree of head trauma is an important risk factor for childhood brain ischemia and/or gross infarction. Organized childhood tackle football in the United States can begin at age 5 years, leading to potentially decades of repeated brain injuries. In addition, the body mass index of the United States pediatric football-playing population continues to increase, so the forces experienced by tackled pediatric players continues to increase. Further work is needed to understand how repeated high-impact large-force trauma from childhood football affects the immature central nervous system.
Footnotes
JRB reviewed the literature and generated the first and subsequent drafts of the paper. MRG provided the cases and mentorship, gave feedback and input on each draft, and participated in the writing.
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The authors disclosed receipt of the following financial support for the research and/or authorship of this article: Dr Golomb was supported by the Department of Neurology at Indiana University School of Medicine and Indiana University Health Physicians.
The study was approved by the Institutional Review Board of the Indiana University School of Medicine (study # 0207-55).