Aneurysmal subarachnoid hemorrhage (aSAH) is a life-threatening form of hemorrhagic stroke and is more common in women than men, typically occurring between 40 and 60 years old.1 It was anecdotally observed that women with aSAH tend to experience their menses within the first week after their aSAH, but it was unclear as to whether this was related to stress hormone physiology or central hypothalamic-pituitary axis dysfunction. Such early menses after stress is now termed anovulatory uterine bleeding (AUB). The purpose of this study was to compare the onset of menses after aneurysm rupture in women with aSAH admitted to our neuroscience intensive care unit (ICU) with a normal menstrual cycle period of 21 to 28 days.
Methods
After institutional review board approval, a retrospective medical record review was conducted of premenopausal women aged 18 to 55 years who were admitted to the neuroscience ICU from June 1, 2011, to June 30, 2012, with the diagnosis of aSAH. Their medical records were reviewed for specific documentation about the onset of menses during hospitalization. Each record was also examined for any contraceptive use such as oral, intrauterine devices, rings, implants, injections, and so forth or medication use that could affect menses, the patient's medical and surgical history (including gynecologic history such as hysterectomy), brain aneurysm location and size, Glasgow Coma Scale (GCS) score on admission, modified Fisher score for the severity of aSAH bleeding,2 and development of cerebral vasospasm. All premenopausal patients with aSAH were evaluated, counting them as both case and controls for the purposes of this study because postmenopausal patients should not theoretically experience menses and therefore could not be used as controls.
To compare the statistical likelihood of early onset of menses in our study population, it was compared against the average menstrual cycle of approximately 21 and 28 days using the 1-sample t test, comparing the means for statistical significance. P < .05 was defined as significant. The means were also compared using a non-Gaussian distribution model (Wilcoxon rank sum test and the same P value for significance) because there was uncertainty whether aSAH-associated menses behave according to bell-shaped phenomenon. It should be noted the statistical comparison was considered purely exploratory because no published data could be found on early menses in aSAH in the literature or a database.
Results
During the study period, 103 SAH cases were identified from a prospective cerebrovascular registry at our center. However, 42 were excluded because the patients were men or the origin of the SAH bleeding was nonaneurysmal (eg, pretruncal or traumatic). The remaining 61 cases were women with aSAH, and 15 met the age criterion of 18 to 55 years. Five of the 15 cases did not have documented onset of menses during their neuroscience ICU stay (all 5 were 44 years or older, 1 was older than 50 years). Four of these patients were possibly postmenopausal, but there was no documentation of postmenopausal state. Table 1 outlines patient age, aneurysm location and size, history of migraine, GCS score, modified Fisher grade, and number of days after admission when menstruation started for the remaining 10 cases. Nine patients (90%) had documented menses starting during their initial week of hospitalization after aSAH admission. Only 1 patient had menses documented after the initial week, on hospital day 12. It is also important to note that 1 patient represented a second aSAH within the study time frame, is included in the case series twice (cases 6 and 7), and had AUB documented twice. None of the patients were taking oral contraceptives or had any other contraception method (eg, intrauterine device, rings, etc).
Table 2 outlines other stress-induced conditions observed in these aSAH patients, including Cushing ulcer, Terson syndrome (intraocular bleeding), and the neurocardiologic profile including their electrocardiogram, transthoracic echocardiogram, and cardiac enzymes (B-type natriuretic peptide and troponin T). One patient (case 1) had an abnormal electrocardiogram without reduction in the left ventricular ejection fraction. Two patients (cases 3 and 9) had Takotsubo cardiomyopathy on transthoracic echocardiogram with wall motion abnormalities.
This patient population had a median age of 43.5 years, and according to a study analyzing the natural history of the menstrual cycle based on various ages, they would be expected to have a mean menstrual cycle of 25.4 days.3 Furthermore, the mean onset of menses in this study population (3.8 days) was compared with the normal expected mean onset of menses for both 21 and 28 days, assuming a Gaussian distribution. This is based on the assumption that these patients had a normal frequency of menstruation when their aSAH occurred. The comparison of means by 1-sample t test (2-tailed P values) yielded a P value less than .01 when comparing 3.8 with 21 days, which suggests that the early onset of menses after aSAH was unlikely due to chance. When the mean onset of menses of 3.8 days was compared with a mean of 28 days, it showed a similar P value less than .01. However, the 1-sample t test model does assume a Gaussian, or bell-shaped, normal distribution. When assuming a non-Gaussian distribution, we used the Wilcoxon rank-sum test yielding statistical significance versus 28 days (P < .01). This would suggest that this study population had an earlier onset of menses compared with a normal 21- to 28-day menstrual cycle.
Discussion
We theorize that an early onset of menses may occur after aSAH in women. Furthermore, this issue is an important sex-specific observation that may be previously unrecognized in the ICU. Although there is uncertainty as to the pathophysiologic origins of AUB after aSAH (AUBSAH), it is speculated that aSAH causes substantial stress hormone response in women.
In 1932, Dr Harvey Cushing first described the association of peptic ulcers occurring postoperatively in his neurosurgical patients.2 This discovery suggested a brain-body relationship that could influence different organs in the body under stress. It is plausible that brain aneurysm rupture may induce physical and psychological stress that causes some form of neurohormonal dysregulation and AUB. However, given our limited size and study design, we cannot entirely exclude other association biases inherent to this preliminary study causing AUB. Other associations of stress and its effects have been described such as Takostsubo cardiomyopathy, which has been identified in women after aSAH4-7 and was also noted in some of our patients (Figure 1). After our finding, we hypothesize that AUBSAH could be present in other brain injury models such as traumatic brain injury or traumatic SAH.
In this study, it was found that most women with aSAH aged 18 to 55 years had onset of menses documented during the first week of hospitalization after aSAH, which seems unlikely due to chance. Although frequency of menstruation varies for each individual, it is improbable that the onset of menses seen in this patient population falls within the normal mean menstrual cycle published in the literature (27.7 +/- 2.4 days or 21-28 days).8,9 According to Sherman and Korenman,3 and this population's mean age of 43.5 years, a mean menstrual cycle of 25.4 days would be expected, assuming that women with aSAH have normal menstrual cycles. However, the more extreme range of normal menses of 21 to 28 days was used against this population, and it still showed notable differences in early onset of menses. Therefore, this observation may represent a previously unrecognized phenomenon in women with aSAH in the ICU setting. Undoubtedly, this phenomenon may have been overlooked because of higher-priority critical care issues in this patient population. Nonetheless, this phenomenon is an important sex-specific issue that should be recognized going forward.
Irregular menstrual cycles are common, of varying origins, and caused by a variety of factors.3,8 After menarche, the intermenstrual intervals are long, and before menopause, the cycle duration shortens.3 As women pass their reproductive years, there is ovarian follicular depletion accompanied by a change in intermenstrual bleeding. This transition to menopause is called perimenopause and occurs on average at 47.5 years old.10 Menopause is clinically recognized after 12 consecutive months of amenorrhea. There is a wide variability in onset of presentation, but in healthy women, menopause typically occurs around 51 years old.11 The postmenopausal state starts 1 to 5 years after the final menstrual period and is further characterized by a permanent decrease in ovarian hormone function leading to accelerated bone loss12 and other sequelae. With a mean age of 43.5 years, a normal menstrual cycle should have taken place in most of our study population up until their aSAH. However, other factors may affect the frequency of menstruation including smoking, medications such as contraceptives, genetics, diet, and exercise.13 Anovulatory uterine bleeding (previously called dysfunctional uterine bleeding) is defined as excessive, prolonged, or unpredictable bleeding from the endometrium that is unrelated to structural or systemic disease, such as the premature onset of what was documented as menses in this study.14,15
A previous study suggested a lack of estrogen in Takotsubo cardiomyopathy.16 Estrogen has a cardioprotective factor against the toxic effects of catecholamines,17,18 which supports the idea that changes in hormone levels can lead to secreted substances that exert themselves at distant sites, perhaps causing otherwise unexplained brain-body syndromes.7 However, this phenomenon cannot be entirely implicated to the central nervous system stress response because one description was found in the literature describing AUB after acute musculoskeletal trauma.19 Therefore, similar to Takotsubo cardiomyopathy, it is typically the magnitude of the stress itself (eg, GCS) that correlates with the cardiac condition or postmenopausal status. Other pathophysiologic causes of this phenomenon described in the literature include transient central hypogonadotrophic hypogonadism, which can cause hormonal disruption and lead to AUB,19,20 or breakthrough menstrual bleeding in premenopausal women.
A final etiology to AUBSAH may be related to matrix metalloproteinases (MMPs), which are upregulated in most brain injury models. Matrix metalloproteinases are capable of degrading the constituents of the basal lamina and extracellular matrix. A number of MMPs are involved in both normal and abnormal endometrial bleeding.21 Matrix metalloproteinases have been implicated in a variety of intracerebral and extracerebral functions in patients after aSAH.22 Plasma levels of MMP-9 have been shown to be appreciably elevated during the first 14 days after presenting with aSAH.23 Therefore, a possible mechanism of AUBSAH may be due to the increased inflammatory cytokines after aSAH, which subsequently upregulate MMPs that can facilitate sloughing of the endometrial lining. Measuring theses inflammatory markers could be considered for future studies.
During the transition to menopause, different changes in lipid metabolism and cholesterol also occur and can lead to increased atherosclerosis and blood vessel intracranial changes. Matthews and colleagues24 reported a substantial increase in low-density lipoprotein cholesterol (LDL-C) and a modest increase in high-density lipoprotein cholesterol in women close to the final menstrual period. Patients with at least 7.32-mg/dL annual increase in LDL-C had a 50% greater chance of carotid plaque index greater than 2 in the postmenopausal years compared with those who had a lower LDL-C increase.24 The increased risk of developing atherosclerosis puts women at risk to experience cerebrovascular accidents, which is a historical term for stroke, includes SAH, and is confounded by advancing age and many other factors.
Limitations
There are several limitations of this study. First, the lack of baseline frequency of menstruation in this population is a limitation of the study because menstrual cycles do vary by individual and throughout the reproductive years and cause different physiologic changes.3,8,24 Another limitation of this study is the retrospective chart data abstraction of the menstrual cycle characteristics, although those with menses were identified prospectively by neuroscience ICU nurses and documented in the record. Another limitation was the relative lack of extensive clinical documentation of contraceptive methods, menses, and menstrual status in the medical record, including the volume of menses made (eg, milliliters or weight) and the character of the bleeding thickness and color. Nonetheless, ICU nursing was able to capture the needed information about onset, which was insightful into this potentially sex-specific issue that otherwise seemed underappreciated. It was also discovered that menopause is difficult to define because at least a year of amenorrhea has to pass and likely this event occurs more on a continuum rather than an exact date.
The relatively small sample size must be acknowledged; however, more than 100 SAH patients were screened during the period to make this observation in 61 patients with nursing documentation of their menses timing. Of this larger cohort, only 15 patients met our final age criteria of 18 to 55 years and were presumably premenopausal or perimenopausal. This age range was chosen to capture the most patients who might still be premenopausal, but that is highly variable and specific to each woman. Another item that could not be captured was the sympathetic outflow after aSAH in these patients, which can increase in response to pain or emotional stress and could be a pathophysiologic trigger of this phenomenon similar to neurohormone-induced stress-related diseases such as Takotsubo cardiomyopathy.5-7 A final limitation of this study was that a gynecologic history taken from the patients themselves was not always feasible or reliable due to depressed or altered levels of consciousness after aSAH. Two of our 10 cases (20%) presented with a GCS score of 13, making it difficult to obtain a detailed menstrual history. However, menses can be easily identified by physical examination of the vaginal area.
Strengths
The study describes a potentially important sex-specific issue occurring in women with aSAH. Second, this study may have implications in evaluating total blood loss in the ICU with so-called ICU vampirism anemia or phlebotomy-related blood loss from repeated laboratory draws, as well as providing better documentation of blood loss in menstruating women in the ICU. This study also provides benefit for designing future observational studies on this issue in terms of improved collection of gynecologic history and examination in the ICU and obtaining additional historic details of the patient's normal menstrual cycle characteristics, date of last menstrual period, methods of contraception, and other potential biomarkers such as MMP and estrogen and other hormone levels after aSAH.25 However, because we acknowledge that this work is only a retrospective study, it does not definitively prove cause and effect. In other words, it has yet to be proven that cases of aSAH in women actually have the opposite phenomenon that central nervous system hypothalamic-pituitary-ovarian axis neurohormonal disturbance occurs and leads to aneurysmal rupture and subsequent SAH bleeding and menses. Only larger prospective observational studies with biomarker analysis might help establish causal relationships.
Future studies should also consider blood sampling of hypothalamic-pituitary hormones in premenopausal women with unruptured brain aneurysms and those with aSAH for comparison. However, it was not possible to blood endocrine test every woman every day with an unruptured aneurysm as a baseline hormone change before aneurysm rupture because this would likely be too costly given relatively low rupture rates of most aneurysms. Future studies could use methods similar to the ones used in this study, albeit with better gynecologic history with measurements of drops in estrogen or progesterone levels.
Conclusion
This small, retrospective study suggests that aSAH may disrupt the normal menstrual cycle of premenopausal women, leading to earlier breakthrough menstrual bleeding or AUB. To our knowledge, this is the first description of this potentially sex-specific and stress-related phenomenon in women with aSAH. A future prospective study is suggested to evaluate the neuroendocrinologic profile in similar patients to better understand the effects of aSAH on hypothalamic-pituitary-ovarian-uterine pathophysiology.
References