NOS indicates not otherwise specified.
Snook ML, Henry LC, Sanfilippo JS, Zeleznik AJ, Kontos AP. Association of Concussion With Abnormal Menstrual Patterns in Adolescent and Young Women. JAMA Pediatr. Published online July 03, 2017. doi:10.1001/jamapediatrics.2017.1140
Is concussion associated with the development of abnormal menstrual bleeding patterns in young women?
In this cohort study of 129 adolescent and young women with a sport- or recreation-related concussion or nonhead orthopedic injury who were followed up for 120 days after their injuries, the risk of having 2 or more abnormal menstrual bleeding patterns after injury was significantly higher among patients with concussion.
Menstrual patterns should be monitored after concussion.
Brain injury may interrupt menstrual patterns by altering hypothalamic-pituitary-ovarian axis function. Investigators have yet to evaluate the association of concussion with menstrual patterns in young women.
To compare abnormal menstrual patterns in adolescent and young women after a sport-related concussion with those after sport-related orthopedic injuries to areas other than the head (nonhead).
Design, Setting, and Participants
This prospective cohort study of adolescent and young women with a sport-related concussion (n = 68) or a nonhead sport-related orthopedic injury (n = 61) followed up participants for 120 days after injury. Patients aged 12 to 21 years who presented within 30 days after a sport-related injury to a concussion or sports medicine clinic at a single academic center were eligible. Menstrual patterns were assessed using a weekly text message link to an online survey inquiring about bleeding episodes each week. The first patient was enrolled on October 14, 2014, and follow-up was completed on January 24, 2016. Inclusion criteria required participants to be at least 2 years postmenarche, to report regular menses in the previous year, and to report no use of hormonal contraception.
Sport-related concussion or nonhead sport-related orthopedic injury.
Main Outcomes and Measures
Abnormal menstrual patterns were defined by an intermenstrual interval of less than 21 days (short) or more than 35 days (long) or a bleeding duration of less than 3 days or more than 7 days.
A total of 1784 survey responses were completed of the 1888 text messages received by patients, yielding 487 menstrual patterns in 128 patients (mean [SD] age, 16.2 [2.0] years). Of the 68 patients who had a concussion, 16 (23.5%) experienced 2 or more abnormal menstrual patterns during the study period compared with 3 of 60 patients (5%) who had an orthopedic injury. Despite similar gynecologic age, body mass index, and type of sports participation between groups, the risk of 2 or more abnormal menstrual bleeding patterns after injury was significantly higher among patients with concussion than among those with an orthopedic injury (odds ratio, 5.85; 95% CI, 1.61-21.22).
Conclusions and Relevance
Adolescent and young women may have increased risk of multiple abnormal menstrual patterns after concussion. Because abnormal menstrual patterns can have important health implications, monitoring menstrual patterns after concussion may be warranted in this population. Additional research is needed to elucidate the relationship between long-term consequences of concussion and the function of the hypothalamic-pituitary-ovarian axis.
Concussion is the most common type of brain injury and is particularly prevalent in sports and recreational activities. Recent estimates suggest that as many as 1.9 million children younger than 18 years have a sport- or recreation-related concussion each year in the United States.1 This injury is biomechanically induced, with symptoms resulting from neuronal dysfunction due to functional and neurometabolic alterations rather than gross structural abnormalities.2 Compared with boys involved in similar activities, girls experience higher rates of sport-related concussion,3- 7 report more severe symptoms,8- 11 demonstrate worse cognitive impairment,8- 10,12 and take longer to recover.11 The neural mechanisms behind these postconcussion sex differences are poorly understood but have been attributed to differences in neuroanatomy and physiology,13 cerebral blood flow,14 and the female sex hormones estrogen and progesterone.15- 17
In girls, functional disruption of the neuroendocrine hypothalamic-pituitary-ovarian (HPO) axis, which governs the menstrual cycle, could lead to dysregulation of the menstrual cycle as reflected by irregular bleeding patterns. In fact, development of neuroendocrine disorders related to hypopituitarism after traumatic brain injury (TBI) is well established, with most evidence in adult men.18- 20 By contrast, a single case report21 has been published on the effect of TBI on menstrual patterns in young women that involved an adolescent who presented with amenorrhea after a severe TBI and was found to have traumatic disruption of the pituitary infundibulum. Two small retrospective studies conducted in adult women22,23 reported a high incidence of amenorrhea and oligomenorrhea after moderate and severe TBI, with the severity of injury predicting the degree of menstrual dysfunction. To date, researchers have yet to examine the role of concussion on menstrual patterns in adolescent and young women (hereinafter referred to as young women).
Early identification and management of menstrual irregularities among young women is important because menstrual irregularities such as amenorrhea and oligomenorrhea can be associated with hypoestrogenism and decreased bone mineral density,24,25 which may not be recoverable over time.26 If concussion contributes to menstrual irregularities in young women via disruption of the HPO axis, assessing menstrual status in the early postconcussive state could help clinicians identify individuals with abnormal menses and initiate or refer them for appropriate management. Although the American Academy of Pediatrics27,28 and American College of Obstetricians and Gynecologists29 advocate for clinicians to regard the menstrual cycle as a vital sign in adolescents because of the importance of estrogen in bone health and other tissues, the assessment of menstrual status is neither a standard of care nor routine practice after a concussion.
The aim of the present study was to perform a prospective comparison of menstrual patterns in young women with sport-related concussion and those with sport-related orthopedic injuries in areas other than the head (nonhead). The latter was chosen as a control group to attribute any differences in outcome to concussion specifically rather than injury in general, because injury and removal from sport and/or school are stressful events and psychological and physical stress are known to affect menstrual cyclicity.30- 32 We hypothesized that young women with concussion would be more likely to have abnormal menstrual bleeding patterns after their injuries than would those with sport-related orthopedic injuries.
Female patients aged 12 to 21 years presenting to the University of Pittsburgh Medical Center Sports Concussion Clinic or Center for Sports Medicine, Pittsburgh, Pennsylvania, for evaluation of a closed head injury or nonhead orthopedic injury that occurred during a sport- or recreation-related activity (categorized as aesthetic, endurance, or other sport types33,34 in the preceding 30 days were eligible for the study. Patients received compensation to participate in the study ($10 at enrollment and $20 at completion). A single board-certified obstetrician gynecologist (M.L.S.) performed all participant screening, consenting, and enrollment procedures. The institutional review board of the University of Pittsburgh approved the study design and consent forms before data collection. All patients younger than 18 years provided assent, and their parents provided written informed consent. Patients 18 years and older provided written informed consent.
Clinical confirmation of a concussion was made using a neurologic examination and evaluation with confirmatory follow-up assessments of balance, cognitive, vestibular, and oculomotor impairment. Concussion was defined based on guidelines from the Centers for Disease Control and Prevention35 and the Fourth International Consensus Statement on Concussion in Sport.36 According to these criteria, a concussion was defined as any mild (ie, Glasgow Coma Scale score ≥13) closed head injury involving altered cognitive functions (eg, confusion, memory loss, or disorientation), signs or symptoms (eg, headache, dizziness, balance problems, or nausea), or brief loss of consciousness lasting no more than 1 minute after a direct or indirect blow to the head. Patients with head injuries involving any structural damage or abnormality (eg, skull fracture, subdural hematoma) based on abnormal findings on computed tomographic or magnetic resonance imaging scans, if performed, were excluded.
Inclusion criteria required each patient to be at least 2 years postmenarche and to report regular menses in the previous year. This postmenarche point was chosen because irregular bleeding patterns are common immediately after menarche as the HPO axis is maturing, whereas 70% to 90% of girls will develop regular menses within 2 years.37- 39 Regular menses was defined by self-reported monthly menstrual cycles with 11 to 12 menstrual cycles in the preceding year. Inclusion criteria also required patients to be English speaking and have access to a smartphone to receive text messages. Patients were excluded from participation if they had any known condition that may affect menstrual function, including subjective report of pregnancy; any use of hormonal contraception in the preceding 6 months; a hypothalamic or pituitary disorder; a brain tumor; an endocrine disorder that may cause menstrual dysfunction (eg, thyroid disease, hyperandrogenemia); a diagnosis of delayed or precocious puberty, anxiety, depression, eating disorder, or epilepsy; or history of a severe TBI.
At the time of enrollment, each patient’s height and weight were recorded using a digital column weight scale (SECA model 769; SECA), which included a height-measuring rod (SECA model 220). Patients were asked to complete a questionnaire that inquired about demographic characteristics, such as self-identified race, injury history, menstrual and sexual history, and a preferred and reliable mobile telephone number to receive text messages.
Patients received an automated, weekly text message containing a link to an online Qualtrics survey (https://www.qualtrics.com). Text messages were sent every Sunday evening using EZtexting.com (SMS Marketing). The online survey asked patients whether in the last week they (1) experienced any bleeding episodes and if so, the day bleeding began and the duration; (2) experienced any new injury for which medical attention was received and if so, to classify the injury type; (3) started or continued to use any hormonal contraception; and (4) were or could be pregnant. After enrollment, patients received weekly text messages to 120 days after injury (range, 90-120 days). The rationale for assessing bleeding patterns for 120 days after injury was to allow for capture of approximately 3 to 4 menstrual cycles for each patient—a time frame in which amenorrhea, if identified (intermenstrual interval, >90 days), would typically prompt clinical evaluation.40
If a patient did not complete a survey after 3 weeks, the primary investigator (M.L.S.) contacted her and inquired about her desire to continue with the study. If the patient wished to continue, she was asked to verbally respond to the questions in the survey for all weeks missed. Patients were also contacted when submitted surveys were incomplete or suggested they no longer met eligibility criteria (eg, the patient used hormonal contraceptives or believed that she was pregnant). For patients who started using hormonal contraceptives (n = 2), individual data were included in the analysis until the week before they met exclusion criteria. For the 1 patient who became pregnant at the end of follow-up, data were included until approximately 4 weeks before the confirmed conception date. Menstrual diaries were then prospectively created for each patient based on her baseline questionnaire menstrual history (which included the start date and duration of the last sure menstrual period) and each weekly survey response.
The primary outcome of interest was the development of an abnormal postinjury menstrual pattern, as defined by an intermenstrual interval of less than 21 days (short) or greater than 35 days (long) or a bleeding duration of less than 3 days or greater than 7 days. These criteria were chosen because 70% to 90% of bleeding patterns in adolescents are within this range for longer than 2 years after menarche, and approximately 95% of cycles will have bleeding durations of 3 to 7 days.37- 39 Because having 1 abnormal menstrual pattern could be considered to be normal variation,38,39 we considered having 2 or more abnormal patterns within the relatively short study period (120 days, with an expected 3-4 menstrual cycles) to be indicative of dysfunction.
Sample size was determined by assuming 20% to 25% prevalence of menstrual irregularities in high school athletes.33,41- 43 A priori power analysis indicated that recruitment of 59 patients with concussion and 59 patients with nonhead orthopedic injuries would provide 80% power to detect a 50% prevalence of menstrual irregularities among patients with concussion. Recruitment was initially set at 75 patients in each group based on a dropout rate of 20%; however, retention proved to be unexpectedly high, and therefore enrollment was halted with 68 patients in the concussion group and 61 in the orthopedic injury group.
Descriptive statistics (means, SDs, and frequencies) were used to define the total patient population and compare the injury groups on demographic and injury-related factors. Comparisons were made between groups using 2-tailed independent t tests (continuous) or χ2 tests (categorical). Subsequent χ2 analyses with odds ratios and 95% CIs were performed to compare the likelihood and risk of abnormal bleeding patterns after injury in both groups. Statistical significance was considered to be P < .05. Analyses were performed using SPSS software, version 22.0 (IBM).
From October 14, 2014, through September 29, 2015, a total of 792 patients aged 12 to 21 years who presented consecutively for evaluation of an injury underwent screening for eligibility (Figure). A total of 309 patients (39%) fit eligibility criteria, but 180 of these (58.3%) were not included in the study because they did not meet 1 or more inclusion criterion, met an exclusion criterion, or declined participation. The remaining 129 patients (mean [SD] age, 16.2 [2.0] years) were enrolled in the study, with 68 (52.7%) having had concussion and 61 (47.3%) having had an orthopedic injury. Eight of 61 orthopedic injuries (13.1%) were characterized as bone or stress fractures; 41 (67.2%), muscle or ligament strains or tears; and 12 (19.7%), other injuries. Study retention was high, with 1 patient being excluded during the study period, 3 patients with fewer than 120 days of follow-up, and no voluntary withdrawals. All data from 1 patient in the orthopedic injury group were excluded when she disclosed that she inaccurately recorded bleeding episodes from study enrollment. In the concussion group, 1 patient became pregnant after 10 weeks and 2 initiated oral contraception (one at 7 weeks and another at 16 weeks after concussion).
The weekly survey response rate was 94.5% (1784 survey responses of 1888 text messages received by patients) across all weeks, yielding a total of 487 bleeding patterns recorded in 128 patients. Eighty-three of all 487 menstrual patterns were considered to be abnormal, with a prolonged intermenstrual interval being the most common type of abnormality reported (64 of 83 [77.1%]).
Comparisons of baseline and injury characteristics between groups are presented in Table 1. Patients in the concussion group were statistically younger than those in the orthopedic injury group (mean [SD] age. 15.7 [1.8] vs 16.6  years; P = .01), but gynecologic age (an indicator of HPO axis maturity, calculated as result of age of menarche subtracted from current age) was not statistically different (mean [SD], 3.9 [1.9] vs 4.1 [1.8]; P = .49). A higher proportion of patients with concussion had a prior concussion (29 of 68 [42.6%]) compared with those with an orthopedic injury (10 of 61 [16.4%]; P = .001). The concussion and orthopedic injury groups were similar with regard to body mass index, race/ethnicity, sport, and time from injury to enrollment. We found no statistical difference in demographic variables between patients who developed 2 or more abnormal menstrual patterns during the study interval and those who maintained regular menses or had 1 abnormal pattern (Table 2).
Patients reported 2 to 6 bleeding patterns after injury. Fifty-seven of 128 patients (44.5%) had at least 1 abnormal bleeding pattern during the study period, with no difference between injury groups. Of the 68 patients who had a concussion, 16 (23.5%) experienced 2 or more abnormal bleeding patterns during the study period compared with 3 of 60 patients (5%) who had orthopedic injuries. Patients with a concussion had increased risk of having 2 or more abnormal menstrual bleeding patterns after injury compared with those with an orthopedic injury (5.58; 95% CI, 1.61-21.22) (Table 3). Among patients with concussion, we found no statistical difference between those who developed 2 or more abnormal menstrual patterns and those who did not in terms of presence of prior concussion, number of prior concussions, or time to medical clearance after concussion.
The odds ratios and 95% CIs for developing the type of abnormal menstrual pattern by injury group are presented in Table 3. Patients with concussion had significantly increased risk of having short (<21 days) intermenstrual intervals compared with patients with orthopedic injuries (5.60; 95% CI, 1.19-26.38). No patients experienced prolonged bleeding for longer than 7 days or amenorrhea.
To our knowledge, this study is the first to prospectively characterize abnormal menstrual patterns after concussion in young women. Our results demonstrate that young women have significantly increased risk of multiple abnormal menstrual patterns after concussion compared with orthopedic injury. The present findings extend previous research on TBI to concussion because head injury can affect and alter menstrual cycles.
Traumatic brain injury contributing to the development of abnormal menstrual patterns has been reported.22,23,44 In a letter to the editor, Cytowic et al44 first reported lack of menses for 1 to 6 months in 11 of 84 women who sustained mild to moderate head injury after motor vehicle crashes. In a retrospective survey of 30 adult women with TBI in the preceding 1 to 3 years, Ripley et al23 reported a median duration of amenorrhea of 61 days, and 90% of women skipped menstrual cycles. The severity of TBI, determined by Glasgow Coma Scale score and posttraumatic amnesia, predicted the duration of amenorrhea. Similarly, Colantonio et al22 performed a retrospective survey of 104 women 5 to 12 years after TBI and compared results with those of a matched control group. Those authors reported that 46% of patients experienced amenorrhea and 68% developed irregular cycles after TBI vs 8% and 27% of controls, respectively, during the same period. Amenorrhea was significantly associated with lower Glasgow Coma Scale scores, higher Injury Severity index score, and longer acute hospital stay.
Although our study found significantly increased risk of multiple abnormal menstrual cycles after concussion compared with orthopedic injury, none of the patients developed amenorrhea. Discrepancy may hinge on prospective vs retrospective study designs, with associated biases and injury type and severity.
The present findings suggest that more subtle forms of brain injury, such as concussion, may adversely affect HPO axis function and therefore menstrual cycles through a number of proposed mechanisms leading to disrupted gonadotropin secretion. These mechanisms may include transient ischemic damage to the hypothalamus, infundibulum, and/or pituitary gland20,45; acute hyperprolactinema18,31; and activation of the hypothalamic-pituitary-adrenal axis31 owing to the stress response of an injury. Concussions are known to be associated with emotional responses,46 including depression and anxiety, that may further exacerbate a psychological stress response.
With these circumstances, the downstream consequences in young women include inadequate stimulation of the ovarian follicles and less than optimal estrogen production. Estrogen is especially important in young women to stimulate the development of secondary sexual characteristics and is critical for building bone mass, because peak bone mineral density is attained in late adolescence47 and may not be recoverable over time.26 Specifically, women who developed menstrual irregularity at younger than 20 years were almost 3 times more likely to have bone mineral density below normal for their age compared with those who developed menstrual irregularity at older than 20 years.48
Based on the present findings and as an extension of the recommendations from the American Academy of Pediatrics 27,28 and American College of Obstetricians and Gynecologists,29 we believe that clinicians treating patients with concussion should inquire about menstrual patterns to identify those patients who may develop abnormal menses. Early identification of menstrual abnormalities and appropriate referral for further evaluation and treatment may improve potential health concerns in adulthood.29
We recognize limitations to the present study, including that subjective report of a regular menstrual pattern does not confirm the quality of that menstrual cycle, that is, whether ovulation is occurring and/or whether ovarian estrogen production is appropriate. Given the present study design and population, we relied on self-reported menstrual history before enrollment and chose weekly surveys to minimize error in patient recall of bleeding patterns. Obtaining serial biological specimens for hormonal assays was not feasible. Therefore, weekly self-reported menstrual bleeding patterns during the 120 days after injury served as a proxy for normal and abnormal HPO axis function. We did not assess long-term menstrual status to determine any effects of concussion on more chronic end points, such as bone health. Because of the strict inclusion criteria for this study, results may not be applicable to prepubescent girls, individuals who are within the first 2 years of menarche, those with preexisting abnormal menstrual cycles, or those previously diagnosed with anxiety or depression. Perhaps the immature HPO axis, or one not functioning appropriately or under significant baseline psychological stress, is more susceptible to the metabolic and psychological effects of concussion. Along those lines, we do not have empirical measurements of anxiety, mood, or psychological stress between groups. We were also unable to account for the seasonal or training effect of the patients, because a change in training intensity could also affect menstrual patterns and disturbances.49 Moving forward, researchers should expand the present study by including a noninjured control group of young women playing sports.
The present findings suggest that young women have increased risk for multiple abnormal menstrual patterns after concussion compared with orthopedic injury. We recommend monitoring menstrual patterns after concussion. In so doing, patients with abnormal menstrual patterns could be identified and subsequently referred for appropriate evaluation and possible treatment. Larger studies with hormonal assessments and long-term follow-up are needed to better understand the effect of concussion on the HPO axis and potential implications for menstrual patterns, estrogen production, and any persistent consequences.
Corresponding Author: Anthony P. Kontos, PhD, Department of Orthopaedic Surgery, University of Pittsburgh Medical Center Sports Medicine Concussion Program, University of Pittsburgh, 3200 S Water St, Pittsburgh, PA 15203 (firstname.lastname@example.org).
Accepted for Publication: March 30, 2017.
Published Online: July 3, 2017. doi:10.1001/jamapediatrics.2017.1140
Author Contributions: Drs Snook and Kontos had full access to all the data in the study and take full responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Snook, Henry, Sanfilippo, Kontos.
Drafting of the manuscript: Snook, Sanfilippo, Zeleznik, Kontos.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Snook, Kontos.
Obtained funding: Snook, Kontos.
Administrative, technical, or material support: Snook, Sanfilippo, Kontos.
Study supervision: Henry, Sanfilippo, Zeleznik, Kontos.
Conflict of Interest Disclosures: None reported.
Funding/Support: This research was supported in part by grant 1K01DC012332-01A1 from the National Institute on Deafness and Other Communication Disorders (Dr Kontos and University of Pittsburgh) and grant CTRA-8064 from the Magee-Women’s Research Institute (Dr Snook).
Role of the Funder/Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Meeting Presentation: The contents of this paper were presented at the 63rd Annual Meeting of the American College of Sports Medicine; June 2 and 4, 2016; Boston, Massachusetts.
Additional Contributions: We thank the medical and support staff at the University of Pittsburgh Medical Center (UPMC) Sports Concussion Clinic and UPMC Center for Sports Medicine, the patients, and their parents who allowed participation in this study.