Key points

Introduction

Female athletes are participating in athletics at all levels (scholastic, collegiate, and professional) at higher rates than previously. Although male athletes still comprise the majority of participants in sports like football, many collision and contact sports, such as soccer, basketball, lacrosse, hockey, and wrestling, have seen increased female participation in recent years. Female athletes also compete in sports, such as gymnastics, cheerleading, and competitive dance, in greater numbers than their male counterparts. Although these sports often are not considered collision or contact sports in the traditional sense, several athletes experience sports-related concussions (SRCs) while participating, because they are using their bodies as projectiles. Postconcussion, women report greater symptomatology, score more poorly on neurocognitive testing, and are noted to have higher mortality rates compared with their male equivalents. This article discusses the epidemiology of concussions in girls and women; what is known about female risk factors for concussions; suggestions for concussion treatment and prevention; and future directions to improve knowledge and care of female athletes susceptible to concussion.

Epidemiology

Female athletes experience concussions in similar sports as their male counterparts (eg, soccer, basketball, lacrosse, and softball) but also have unique activities in which they participate in greater numbers than male athletes (eg, cheerleading, gymnastics, and field hockey), which are associated with high concussion rates. Although male athletes experience a greater number of overall concussions, female athletes have a greater incidence of SRCs compared with male athletes. Of the youth sports with the highest rates of concussion, women’s soccer is second only to football with concussion injury burden ^, and more recent data have highlighted that high school female soccer players actually may have a greater incidence of concussions compared with high school football players. In sex-comparable sports (basketball, ice hockey, soccer, and baseball/softball), high school female athletes had 1.4-times greater concussion incidence than male athletes. In soccer and basketball, collegiate female athletes had a higher duration of time loss from sports after concussion compared with their male counterparts.

Ice hockey and lacrosse are some of the fastest growing sports in the United States. Play is different in the respective genders, because the rules are designed to encourage less contact in the female versions of these sports (eg, no body checking). ^, Researchers tracking concussions during international hockey tournaments found that adult female hockey players had lower rates of concussion than adult male hockey players (1.0/1000 player-games for women and 1.4/1000 player-games for men). At the under-18 level of participation, however, female players and male players had similar concussion rates (1.4/1000 player-games for both girl and boy hockey players). Youth-level coaches (28% of whom were women) of female hockey players had similar levels of concussion knowledge as coaches of male youth hockey players.

Men’s lacrosse is a collision sport, whereas women’s lacrosse has been considered a contact sport. Injuries have been decreasing in men’s lacrosse but have been noted to be increasing in women’s lacrosse. Recent rule changes may be leading to more contact in women’s lacrosse (eg, subbing on the fly and establishment of the shot clock). Over the past 2 decades, women’s lacrosse has added protective eyewear and in some regions also has added protective helmets. As women’s sports continue to evolve to involve more contact, injury rates will likely go up, as they have in women’s lacrosse, despite equipment modifications. Early research examining the effect of soft headgear on injury rates in women’s lacrosse found no significant reduction in linear or rotational impacts from high-velocity ball impacts or lateral impact response for stick impact. The soft headgear, however, did lead to a small decrease in frontal impacts of stick impact to the head. More research is needed to better ascertain risk reduction from head equipment.

Baseball and softball are similar sports but with important differences. The distance from the pitcher’s mound to the batter is shorter in softball compared with baseball, as is the difference between the bases. A systematic review examined traumatic brain injuries (TBIs) in baseball and softball. Female softball players were more than twice as likely to experience a concussion compared with male baseball players. In high school players, a majority of concussions were experienced by the catcher in softball (29.7%), but by the batter in baseball (50.6%). At the collegiate level, a majority of TBIs occurred to the batter in softball and to the middle infielder in baseball.

Female athletes participate in gymnastics, competitive dance, and cheerleading at higher rates than athletes. ^{, , ,} There is a lack of understanding of the full epidemiologic concussion burden in these activities, but the few published studies demonstrate that there is concussion injury burden in these sports. ^{, ,} Uncertainty of SRC rates in competitive dance and cheer has been compounded by evolving definitions of what is considered a sport. Competitive cheerleading only recently has been classified a sport, which contributes to the paucity of injury data. Additionally, there is a deficit of specific tracking information for competitive dance in sports injury epidemiology.

Anatomy/physiology/chemical differences

Concussion symptoms are thought to result from an alteration and dysregulation of cerebral blood flow. Thus, research has focused on monitoring brain perfusion during the recovery process. Studies have shown a baseline sex difference in healthy, noninjured young adults, with female athletes having a higher global cerebral blood flow compared with male athletes on PET scan. In a study of 9 concussed female athletes and 13 nonconcussed, healthy athlete controls, cerebral blood flow was monitored via transcranial Doppler (TCD). Concussed athletes demonstrated significantly lower minute ventilation and CO ₂ sensitivity but greater cerebral blood flow at similar exercise workloads in association with symptom onset and premature exercise cessation. These findings were thought to be due to increased perfusion needs from injury. The minute ventilation, CO ₂ sensitivity, and cerebral blood flow changes of the concussed athletes improved after subsymptom threshold aerobic exercise treadmill therapy. In contrast, a TCD study of 58 postconcussive adolescent athletes ages 14 years old to 19 years old demonstrated that female athletes had no changes on TCD even when still reporting symptoms subjectively; this was not observed in the male participants. This suggests that there are subclinical changes that differ by sex that need to be considered both in clinical evaluation and within research studies examining cerebral perfusion changes.

In both in vitro rat and human stem cell axonal cultures, female axons were consistently smaller than male axons and displayed greater damage with dynamic stretch-injury mechanics compared with male axons. This led to greater swelling and loss of normal calcium signaling in female tissues compared with male tissues. These axonal level findings could help explain symptomology differences between the sexes but need to be evaluated with greater depth in further human studies.

Chronic traumatic encephalopathy (CTE) is caused by repetitive TBI and defined by neuropathology showing degenerative changes in brain tissue on autopsy: hyperphosphorylated tau in gray matter, myelin loss, axonal injury, and white matter microstructural changes. Much of the anatomic research involving long-term effects of concussion, in particular those of CTE, have focused on the brains of male athletes, whereas data are lacking in female athletes. In 1 small study of female athletes, 8 young adult, nonconcussed female soccer and hockey players were compared with 10 soccer, hockey, and water polo female players who were 6 months postconcussion. There were no differences on diffusion tensor imaging in the corticospinal tract between the 2 groups, but, within the corpus callosum, the concussed group had lower diffusivity in the area that leads to the prefrontal cortex (whose responsibilities include planning and high-level thinking). Because the function of the corpus collosum is to synthesize motor, sensory, and cognitive functions of the 2 hemispheres of the brain, such a finding suggests considerable neurocognitive deficiencies that should be explored further in athletes of both sexes.

Autonomic responses to postural changes were examined in a study that compared 65 concussed adolescent athletes (60% female) to a control group of 54 nonconcussed athletes (46% female). Heart rate variability, heart rate, and blood pressure were monitored during 6 weekly clinic visits. Less heart rate variability was found in concussed versus nonconcussed female athletes, with no significant difference noted between the male athlete groups. In the overall cohort, the postconcussive participants had greater baseline seated blood pressures and heart rates compared with the controls. These parameters resolved by the end of the study, but the heart rate variability decrease did not resolve in the female participants, suggesting a possible autonomic dysregulation sex disparity, potentially contributing to longer duration of symptomatology. Because vestibular coordination is known to be a factor in symptomatology and is related to postural motions, understanding this difference in the body’s autonomic response to injury may help explain the prolonged recovery repeatedly observed in female athletes postinjury.

An area of concussion research that is expanding rapidly and leading to new areas of focus in diagnosis, management, and time to recovery is the field of concussive biomarkers. In a cohort of 415 healthy, nonconcussed collegiate athletes (39% female), β-amyloid 42, total tau, S100 calcium binding protein B (S100B), ubiquitin carboxy-terminal hydrolyzing enzyme L1 (UCH-L1), glial fibrillary acidic protein, microtubule-associated protein 2, and 2′,3′-cyclic-nucelotide 3′-phosphodiesterase (CNPase) serum concentrations were measured. There were baseline biomarker differences between female athletes and male athletes; notably, female athletes had higher CNPase and lower UCH-L1 and S100 B compared with male athletes . In the same cohort, there were no differences found between the sexes in terms of previous concussions, even though male athletes had a statistically significant higher number of years of cumulative exposure to collision sports (10.2 years ± 5.2 years in male athletes compared with 8.5 years ± 6.1 years in female athletes). Biomarkers were examined in the context of reported number of prior concussions and no differences were found specific to sex. Although athletes and clinicians are eager to find the ideal concussion-related biomarker for diagnosis and recovery tracking, there is no conclusive evidence that any specific serum or blood biomarker is useful to date.

Additionally, physiologic differences from hormonal regulation have also been suggested. In women with typical menstrual cycles, estrogen is low in the early follicular phase (menses), slowly rises and peaks right before ovulation, abruptly drops, slowly increases to a lower mid–luteal phase peak, and then declines into its nadir again during menses (early follicular phase). Progesterone is low throughout the follicular phase, increases after ovulation, gradually peaks mid–luteal phase, and drops with its nadir again during the follicular phase. The average menstrual cycle is 29 days, with a normal range of 24 days to 35 days, because there is large interindividual and intraindividual variability. As with other sports injuries, it has been postulated that female athletes may be more susceptible to concussion during a particular phase of their menstrual cycles. In a preliminary report of 18 eumenorrheic female athletes who experienced an SRC, two-thirds experienced SRC during the late luteal phase or during the first 2 days of menstruation (a 9-day span), suggesting increased SRC risk when estrogen and progesterone are low. Wunderle and colleagues studied 144 premenopausal female patients (16–60 years old) who presented to the emergency department within 4 hours of a mild TBI (mTBI), assessed phase of menstrual cycle via interview and laboratory testing, and assessed recovery 1 month postinjury. Women who sustained a mTBI during the luteal phase (n = 37) had a significantly lower quality of life and were in worse health 1 month after mTBI than those who were injured during the follicular phase (n = 72) or were on hormonal contraceptives (HCs) (n = 35). The investigators proposed a “withdrawal hypothesis,” suggesting that women may fare worse if they sustained a TBI when progesterone levels were high and then decreasing or if the concussion led to an abrupt drop in progesterone, but hormone levels were not assessed to verify if there were changes in progesterone over time.

In a study of 68 adolescent and young adult female athletes post-SRC versus 61 female athletes post–other sports-related orthopedic injuries, 23.5% of the SRC group experienced 2 or more abnormal menstrual cycles during the study follow-up period (120 days) versus only 5% of the nonhead orthopedic injury group. All participants previously had been eumenorrheic, suggesting at least transient disruption of the hypothalamic-pituitary-ovarian axis in some female athletes postconcussion. This is consistent with findings in adults with TBI. In a prospective study of 46 consecutive patients with TBI (33% female), 76% had early laboratory testing consistent with hypogonadism or hypothyroidism, with long-term hypopituitarism frequent only in severe TBI.

In a study of collegiate athletes, 24 female athletes using HCs and 25 not using hormonal birth control when concussed were compared. Non-HC users had higher symptom severity than HC users but no significant difference in length of recovery. Because it was small study, with little description of type or timing of HC (1 was using NuvaRing), it makes it difficult to draw conclusions about the potential benefits of HC and various hormone levels.

In an early rodent model (1993), male rats given estrogen prior to induced brain trauma had improved concussion outcomes (eg, higher brain-free magnesium concentration, greater brain cytosolic phosphorylation potential, and better 1-week posttrauma motor function) whereas female rats post–brain trauma had higher mortality than the male rats in general, and those with estrogen treatment had even lower cytosolic phosphorylation potential postinjury. This suggests that estrogen was neuroprotective in male rats and exacerbated brain injury in female rats. In contrast, in another series of experiments, acute survival post-TBI was much greater in female rats versus male rats and female rats had a less dramatic drop in and better recovery of cortical blood flow than male rats. When rats were given 2 weeks of daily 17β-estradiol injections prior to injury, postinjury cortical blood flow was higher in both female rats and male rats compared with those given sham injections. Estrogen is known to have antioxidant effects and enhance endothelial nitric oxide synthase, improving vasodilation; thus, a microvascular benefit from estrogen is plausible. Progesterone also has potential neuroprotective properties and may enhance repair of damaged nerve cells via neurotrophic, anti-inflammatory, antiexcitotoxicity, anti–lipid peroxidase, and antiapoptotic mechanisms. Rubin and Lipton reviewed 50 publications of TBI in animal models that included male models and female models (rats, mice, and piglets). There was substantial variability in study design and results were somewhat model-dependent, but overall female models seemed to have better behavioral outcomes and less pathology after TBI compared with male models. Progesterone, in particular, was associated most commonly with beneficial effects. In the few human studies using estrogen or progesterone as a treatment post-TBI, neither hormone has shown a consistent decrease in mortality or clear clinical benefit.

More human studies are needed to determine if consistent hormone levels throughout injury and recovery are beneficial and if hormonal testing post-SRC and possible hormonal supplementation are warranted. At a minimum, it will expand knowledge in this field for medical providers managing SRC in women to ask about phases of the menstrual cycle at time of injury and to track menstrual cycle patterns more closely during recovery.

Psychosocial factors could an additional explanation to the observed sex and gender difference in concussion rates, becaus it has been suggested female athletes may be more likely to report symptoms postinjury. A study of 435 Irish adolescent athletes found no significant differences in concussion reporting behaviors between the sexes. But a large study of more than 30,000 Maine high school athletes found that nonconcussed female athletes reported concussion-like symptoms at baseline in greater numbers than male athletes. A systematic review and meta-analysis of 21 studies of 12 year olds to 26 year olds reported that female athletes consistently reported more symptoms at baseline and postconcussion on the Post-Concussion scale and the Sport Concussion Assessment Tool (SCAT) (2nd version), than male athletes, but that the difference was not clinically significant. It is unclear why symptom reporting varies between the sexes and it is likely a multifactorial reason that should be investigated further in order to better diagnosis and symptomatically managed postconcussive athletes of both genders.

Biomechanical factors

Recent literature has proposed that decreased neck strength and lower control in head motion may be a risk factor for concussion. Females’ overall smaller head circumferences, thinner skull cortices, and differences in spinal musculature all have been proposed sex-specific anatomic findings that may contribute to the sex disparity in concussion incidence. ^,

Youth female soccer players were found to have higher peak linear and rotational accelerations during a study monitoring forces during controlled soccer headers, exposing them to an increased risk of injury. In a small study of college soccer players who wore accelerometers while heading the ball, female players had lower cervical/neck strength and neck girth and greater head velocities during impact compared with male players. In youth soccer players ages 11 to 14, male athletes had more head impacts than female athletes but female athletes experienced head impacts of greater magnitude than male athletes. A small study of collegiate female hockey players showed that players had lower peak linear accelerations during head impacts compared with linear accelerations previously reported in male players, but peak rotational accelerations were similar.

An innovative, exploratory study reviewed pictures of soccer players on Google and found that there were significantly more pictures of female soccer players than male soccer players heading the ball with their eyes closed. These big data methods clearly have limitations but should prompt more investigation into the effects visual cues may contribute to increased risk of concussion in female athletes.

Additional, unexplored potential risk contributors include environmental factors, such as different coaching styles in women’s sports, game conditions, and even athletes’ hair length. Coaches may play starting athletes at rates different from male athletes, increasing exposure and leading to increased risk for starting players. Games and matches may be held in less optimal spaces (eg, poor field conditions with less upkeep), leading to an increased risk of injury. Consideration of the barriers to girls and women in sports regarding funding and accessibility to optimal playing conditions should be incorporated in future risk assessment work.

A hot topic for concussion clinicians has been whether protective headgear decreases the risk of concussive injury. The theory behind this equipment utility is that the headgear may decrease the force to the head during heading. A large randomized controlled trial of adolescent soccer players examined the effectiveness of headgear in reducing concussion. There was no protective effect of headgear for male athletes or female athletes. This may be because the most common mechanism for concussion in soccer players is from contact from other players, not because of trauma from headers.

Further research should continue to examine possible preventive biomechanical mechanisms for concussion and these interventions should be tested on both male athletes and female athletes to ensure that findings are potentially pertinent to both sexes.

Concussion testing differences

Neurocognitive testing frequently is utilized for concussion diagnosis, postconcussion symptom monitoring, and clearance for return to play. In 1 study, approximately 300 high school and collegiate athletes’ baseline and postconcussion ImPACT and Post-Concussion Symptom Scale scores were analyzed. Female athletes had significantly lower visual memory scores compared with male athletes on ImPACT testing and also reported more concussive symptoms. In a study of 79 collegiate athletes (48% female) who took ImPACT tests prior to injury and then postconcussion, there were no differences in baseline scores between the sexes. In a large study of high school athletes, there were significant differences in baseline ImPACT testing scores, with female athletes reporting more symptoms and slower reaction times on the computer-based test. Postconcussion, female athletes demonstrated lower visual memory scores. In a study of 188 collegiate athletes with a previous history of concussion (47% female), female athletes with 2 prior concussions and those with 3 or more concussions had higher visual memory scores compared with male athletes with the same concussion burden history. Additionally, male athletes scored lower on processing speed and reaction time compared with female athletes with the same concussion history.

The SCAT was developed to allow for sideline assessment of concussion in athletes. Scores range from 0 to 100 with lower scores indicating more difficulty with cognitive and physical function. More than 1000 high school athletes were given the SCAT (2nd version) prior to concussion. Female athletes had higher overall scores compared with male athletes. Examining SCAT (3rd version) in more than 2000 high school and collegiate athletes demonstrated that female athletes had higher baseline symptom scores on the testing but better Balance Error Scoring System scores compared with male athletes.

Management of symptoms

In a study of 277 high school athletes (47.5% female), adolescent girls and boys had similar SRC knowledge, but the female athletes were more likely to report their symptoms to an adult. Comparing approximately 300 female atheletes and male athletes after concussion, female athletes reported more physical and somatic symptoms than their male counterparts. In those experiencing their first lifetime concussion, female athletes reported a greater number of symptomatic days compared with male athletes regardless of age. In a robust study of more than 300 concussed adolescents who were evaluated within 10 days of injury, gender was not a predictive factor in symptom resolution (an abnormal Romberg test was the largest predictor in a multivariate analysis), but female athletes had an odds ratio of 1.37 for longer symptom duration compared with male athletes.

Female athletes have been shown to have prolonged symptoms in specific domains, including vestibular oculomotor motion screening (VOMS) and vestibular ocular reflex (VOR) scores. A small study of concussed young athletes showed that female gender explained 45% of the variance in the horizontal VOR scores postconcussion but there were no differences between sexes in the other VOMS scores. Comparing gait tasks postconcussion in adolescent male athletes and female athletes (49% female), female athletes had a greater change in cadence. Both genders had shorter stride lengths compared with controls but there was no difference in reported symptom scales between male athletes and female athletes.

Recently, more literature is highlighting that early aerobic exercise may quicken recovery from concussion. The research group that pioneered the Buffalo Concussion Treadmill Test has examined the impact of gender on this interaction. Relative rest and a stretching program were less effective than subsymptomatic threshold aerobic exercise in both male athletes and female athletes. There was a trend for female athletes to have more symptoms when prescribed rest.

Return to sport

This area of concussion has become a recent focus as more athletes and their families are looking for guidance postinjury. A large Canadian study that followed patients (ages 5–18 years) postconcussion found that most children ages 5 years to 7 years of age recovered within the initial 2 weeks after injury. For children ages 8 years old to 18 years old, recovery was closer to 4 weeks. Adolescent girls had significantly longer recovery time compared with adolescent boys, showing that most female athletes had not recovered by week 12. This group currently is conducting a randomized controlled trial to compare full physical rest until completely asymptomatic versus returning to sport within 72 hours postinjury, regardless of symptoms.

A cohort of 726 adolescent athletes were followed over a 7-step return to play protocol that included an initial return to school prior to initiation of light exercise. Female athletes took almost a full week longer to reach step 3 and step 6 compared with their male counterparts. A prospective cohort study of 355 children (ages 8–17 years) who had an active rehabilitation intervention postconcussion found that female athletes had greater symptom levels but recovered from their symptoms faster than male athletes.

A recent retrospective chart review of 579 middle school, high school, and collegiate athletes who experienced a concussion showed that female athletes took 6 days longer than their age-matched male counterparts to begin a return-to-play protocol. In a study of 170 SRCs, female hockey players took longer to return to sport compared with concussed male hockey players, but male basketball players took longer to recover than their female counterparts.

Long-term recovery from concussion also appears to have some gender-related differences. A group of approximately 200 athletes was studied more than 6 months after their concussion for cognitive and executive functions. Female athletes responded more slowly than male athletes on dual-task testing. No other sex-differences were seen in cognitive testing.

Other studies, however, have found no difference in time to resolution of concussive symptoms. These include the important high school reporting information online data that involved 100 high schools and 812 concussions, which showed no difference in return to play between the genders. A study of 215 moderate to severe SRCs also showed no difference in recovery times between female athletes and male athletes 18 years and younger. In a study of adolescent and young adult athletes in primary care sports clinic prolonged return to play in 100 SRCs, there was no difference between sexes but the study consisted primarily of male athletes (90%). Additionally, a study of 468 college athletes (57% female), found that female athletes suffered higher rates of ankle or knee injury after experiencing a concussion compared with their male counterparts postconcussion. Such work suggests the need for further research to examine potential modifiable factors to minimize injury risk during the immediate phase of return to play for athletes.

Summary

As female athlete participation continues to increase, greater awareness of prevention and management of SRC in female athletes should be emphasized because of the gender disparities detailed in this article. Women have greater risk for SRC versus male athletes participating in comparable sports and also have risks in sports in which they participate in higher numbers than male athletes (ie, cheerleading). Because SRC has been identified as an important concern for female athletes, the science must be advanced to truly understand the interplay among rules of play, equipment, biomechanics, and hormonal contribution to SRC susceptibility and recovery, because there is a paucity of published research on the topic. Large prospective studies, including female athletes from a variety of sports and with close monitoring of menstrual cycle patterns and hormonal testing, are warranted. Small studies with female athletes as a minority of subjects will not clarify the magnitude women’s physiology may play in SRC susceptibility and treatment response. With further work, it may be found that women’s physiology can be used to an advantage to prevent SRCs and improve the care of female athletes.