This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mhurchu, C. N. Articles by Nelson, L. M. Search for Related Content PubMed PubMed Citation Articles by Mhurchu, C. N. Articles by Nelson, L. M. Pubmed/NCBI databases Medline Plus Health Information Hormones Related Collections Cerebral Aneurysm, AVM, & Subarachnoid hemorrhage Primary and Secondary Stroke Prevention Risk Factors for Stroke

(Stroke. 2001;32:606.)
© 2001 American Heart Association, Inc.

Original Contributions

Hormonal Factors and Risk of Aneurysmal Subarachnoid Hemorrhage

An International Population-Based, Case-Control Study

Cliona Ni Mhurchu, PhD; Craig Anderson, PhD; Konrad Jamrozik, Dphil; Graeme Hankey, MD; David Dunbabin, FRACP for the Australasian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS) Group

From the Clinical Trials Research Unit, University of Auckland (New Zealand) (C.N.M., C.A.); Department of Public Health, University of Western Australia, Perth (K.J., G.H.); Stroke Unit, Department of Neurology, Royal Perth Hospital (Western Australia) (G.H.); and Royal Hobart Hospital, Tasmania, Australia (D.D.).

Correspondence to Dr Cliona Ni Mhurchu, Clinical Trials Research Unit, Faculty of Medicine and Health Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand. E-mail c.nimhurchu{at}ctru.auckland.ac.nz

	Abstract

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
Background and Purpose—Subarachnoid hemorrhage (SAH) is more common in women than in men, but the role of hormonal factors in its etiology remains uncertain. The aim of this study was to examine the relationship between hormonal factors and risk of SAH in women.

Methods—This was a prospective, multicenter, population-based, case-control study performed in 4 major urban centers in Australia and New Zealand. Two hundred sixty-eight female cases of first-ever aneurysmal SAH occurred during 1995–1998. Controls were 286 frequency-matched women from the general population of each center. Outcome measures included risk of SAH associated with use of oral contraceptive pills (OCPs), hormone replacement therapy (HRT), and various endogenous hormonal factors including menstrual patterns, parity, age at birth of first child, and breast-feeding practices.

Results—Cases and controls did not differ with regard to menstrual and reproductive history except in age at birth of first child, where older age was associated with reduced risk of SAH (odds ratio [OR], 0.63; 95% CI, 0.43, 0.91). Relative to never use of HRT, the adjusted OR for ever use of HRT was 0.64 (95% CI, 0.41, 0.98), which did not alter significantly after further adjustment for possible confounding factors. Borderline evidence of an inverse association was detected for past use of HRT (adjusted OR, 0.59; 95% CI, 0.30, 1.13) and current use of HRT (adjusted OR, 0.67; 95% CI, 0.40, 1.13), but there was no evidence of an association for use of OCPs (adjusted OR, 0.97; 95% CI, 0.58, 1.60).

Conclusions—The risks of SAH are lower in women whose first pregnancy is at an older age and women who have ever used HRT but not OCPs. The findings suggest an independent etiologic role for hormonal factors in the pathogenesis of aneurysmal SAH and provide support for a protective role for HRT on risk of SAH in postmenopausal women.

Key Words: case-control studies • epidemiology • hormones • risk factors • subarachnoid hemorrhage

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
The etiology of aneurysmal subarachnoid hemorrhage (SAH) is poorly understood.^{1 2} Cigarette smoking is a strong risk factor, but only a few other factors, such as hypertension, family history, and heavy alcohol consumption, are considered credible risk factors.³ Since the incidence of SAH is highest in women after menopause,⁴ it has been suggested that sex hormones, especially estrogen, might be protective for the condition.^{1 5 6} In contrast to ischemic stroke,^{7 8} most of the epidemiological data have shown that the risk of hemorrhagic stroke (including SAH) is not substantially influenced by either hormone replacement therapy (HRT) or oral contraceptive pills (OCPs). However, a large population-based, case-control study of SAH has re- ported significant independent associations between the use of either HRT or OCPs and a reduced risk of SAH.⁶ Studies of hormone therapy and risk of SAH have been difficult to interpret because of imprecision due to small numbers of events as well as incomplete or inadequate information regarding exposures and potential confounding variables.^{3 9 10} Data pertaining to risk in relation to endogenous hormonal factors, such as menstrual pattern, are even more limited.⁶

The Australasian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS) was a large prospective, multicenter, population-based, case-control study of SAH. We present here results that pertain to the relationship between SAH and menstrual and reproductive history and use of hormonal therapy, both OCPs and HRT.

	Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
The design of ACROSS has been described.¹¹ Briefly, the study used population-based disease registers in 4 major urban centers in Australasia: Adelaide (South Australia), Hobart (Tasmania), Perth (Western Australia), and Auckland (New Zealand). The total study population (aged 15 years) was approximately 2.8 million according to the 1996 census for each city. A wide range of overlapping sources and harmonized data collection procedures were used to ascertain all cases of primary aneurysmal SAH that occurred over staggered whole-year intervals in the population between November 1995 and June 1998. SAH was defined according to standard criteria as an abrupt onset of severe headache and/or loss of consciousness, with or without focal neurological signs, and in which CT, necropsy, or lumbar puncture revealed focal or generalized blood in the subarachnoid space.¹² Patients whose hemorrhage was found definitely to originate from sources other than an intracranial aneurysm (including primary intracerebral hemorrhage, arteriovenous malformations, trauma, infections, bleeding diathesis, and neoplasms) were excluded. Those patients with proven hemorrhage in whom an aneurysm could not be identified either by cerebral angiography or at necropsy were included, as were 3 patients with acute severe headache followed by death within hours. Each event during the study period was classified as being the patient’s first-ever or a recurrent SAH.

Controls with no history of SAH were randomly selected from electoral rolls and frequency matched to cases for sex, age (10-year strata), and city of residence, on the basis of the projected incidence. Enrollment to vote is compulsory for adults in Australia and New Zealand. A postal invitation to participate in the study was followed up with a telephone call. Replacements were sought when potential control subjects could not be contacted by telephone or by personal visit (after several attempts) or when they had moved outside the study area. On the assumption that proxy interviews would be required for approximately 40% of case subjects because of early death or disability, matching also included interviews with a nominated relative, friend, or other reliable informant. Controls were enrolled in the study during the same period as that in which cases were ascertained and interviewed. Institutional ethics committees in each of the study centers approved the protocol. All participants provided informed consent, including next of kin for case subjects who were severely ill, unconscious, or deceased, and from proxy respondents for control subjects.

As soon as possible after notification, trained study nurses undertook face-to-face interviews with case subjects or, in cases in which they were deceased or disabled, with the partner or next of kin. Control subjects or their proxies were interviewed at home. A structured questionnaire was used to obtain information regarding age, ethnicity (white or other), highest level of education (secondary only or postsecondary), and occupation. Socioeconomic status was classified into 6 categories: 3 nonmanual categories (classes 1, 2, and 3) and 3 manual categories (classes 4, 5, and 6). These strata are primarily based on occupation,¹³ but level of education and profession of spouse (or partner) were also taken into account. Subjects were questioned about their lifetime use of OCP and HRT (current, past, or never) and about their menstrual and reproductive history. Limited information was obtained on the type, dose, or duration of hormonal therapy. Subjects were also questioned about their lifetime use of tobacco (never, current, or ex-smoker >12 months), about their use of alcohol in the previous 2 weeks (number of units per week), and whether a doctor had ever told them that they had hypertension, heart disease (angina, myocardial infarction), or diabetes mellitus. Usual level of activity was categorized as sedentary (<1 session per week) or active (1 session per week) on the basis of the frequency of vigorous exertion (lasting 15 minutes) in the course of work, activities about the house, or recreation.¹⁴ Body mass index (BMI) was calculated as self-reported weight (kg) divided by height (m²). A random nonfasting blood sample was taken from subjects for measurement of serum cholesterol at the laboratories of major hospitals in the study centers. For subjects who were unavailable or refused blood samples, serum cholesterol levels were estimated from other sources, either from (1) their general practitioner or hospital medical records or, when this information was not available, from (2) the most recent cholesterol level recalled by the subject.

Differences between study groups were examined with the ² test for categorical variables and independent sample t tests for continuous variables. Values of P<0.05 were considered statistically significant. Unconditional multiple logistic regression with SAS software¹⁵ was used to model the incidence of SAH in relation to various hormonal factors. The OR and 95% CI were used to estimate relative risk. To avoid problems associated with sparse data, missing values were imputed with the use of the "hot deck" approach provided by SOLAS software.¹⁶ This method adopts a strategy based on the similarity of incomplete and complete records. For each missing variable, a value was imputed from that of a subject with complete data who had the characteristics most similar to that of the subject with incomplete data, while also maintaining the appropriate measurement index for that variable (ie, categorical versus continuous). In this analysis, imputed variables were necessary for cholesterol level (missing values=111), BMI (missing values=44), smoking status (missing values=23), alcohol intake (missing values=34), and history of hypertension (missing values=19). The hot deck method was used separately for cases and controls on the basis of city, age, and social class. Imputed data were only used in the multiple logistic regression analyses to ensure stability of the models. Multiple logistic regression was performed with adjustment for matched factors (age, study center, use of proxy respondent) and confounding variables associated with both exposure and outcome. Two approaches were taken to adjust for confounders: first, confounding variables were only included in the model if they caused a meaningful change (by 10%) in the OR¹⁷ ; second, all possible confounding variables were forced into the model. For both approaches, adjusted ORs are reported.

	Results

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
Two hundred sixty-eight women with first-ever SAH identified by our surveillance effort during the 3 years of case ascertainment made up 61% of all first-ever cases. Of these, 214 women (80%) had the aneurysmal origin of the SAH diagnosed by angiography, surgery, or at autopsy. Only 6 women (2%) suffered perimesencephalic bleeds. In all, 473 (59%) of 798 potential control subjects selected at random from electoral rolls took part in the study. Two hundred eighty-six were women. Reasons for nonparticipation were death (in 9), no contact (in 159), refusal (in 118), and other factors (in 39). Table 1 shows the distribution of cases and controls by study center, demographic factors, and variables that might affect the risk of SAH or propensity to use OCPs or HRT. Controls were significantly more likely than cases to have been educated beyond secondary school, to have a higher socioeconomic status, and to be white. Variables significantly related to the risk of SAH were smoking, history of hypertension, lower BMI, lower serum cholesterol levels, and nondrinking status. Despite efforts to match by source of information, proxies were used for 173 (65%) of case subjects compared with 91 (32%) of control subjects (P<0.0001).

View this table: [in this window] [in a new window]   Table 1. Characteristics of Cases and Controls

The associations between SAH and menstrual and reproductive history are shown in Table 2. The risk of SAH declined with older age at first birth (OR, 0.63; 95% CI, 0.43, 0.91). There was no significant association with risk of SAH for any other endogenous hormonal factors. Ever use of HRT was associated with a significant 36% reduction in odds of SAH (unadjusted OR, 0.64; 95% CI, 0.43, 0.97) (Table 3). Borderline evidence of an inverse association was seen for current use of HRT (unadjusted OR, 0.70; 95% CI, 0.42, 1.13) and past use of HRT (unadjusted OR, 0.55; 95% CI, 0.29, 1.02). No evidence of an association was seen for ever use of OCPs (unadjusted OR, 0.82; 95% CI, 0.58, 1.17).

View this table: [in this window] [in a new window]   Table 2. ORs and 95% CIs for SAH Among Women According to Menstrual and Reproductive Variables

View this table: [in this window] [in a new window]   Table 3. Crude and Adjusted ORs (95% CI) of SAH According to Use of HRT and OCPs

Adjustment for age, city, and proxy respondent did not appreciably affect the association for ever use of HRT (adjusted OR, 0.64; 95% CI, 0.41, 0.98). Similar risk estimates were also observed for past use of HRT (OR, 0.59; 95% CI, 0.30, 1.13) and current use of HRT (adjusted OR, 0.67; 95% CI, 0.40, 1.13). There was no association for ever use of OCPs (adjusted OR, 0.94; 95% CI, 0.59, 1.49).

In the first approach to adjustment for potential confounding variables, in which inclusion in the model depended on a meaningful change (by 10%) in OR,¹⁷ social class, ethnicity, smoking, hypertension, cholesterol level, alcohol intake, and BMI were examined. None of these variables altered the OR by the required 10%, and therefore they were not retained in the models for ever, past, and current use of HRT. Therefore, the adjusted ORs remained unaltered (Table 3). However, the inclusion of confounding variables that altered the OR by a minimum of 10% did affect the OR for ever use of OCPs and risk of SAH (adjusted OR, 0.97; 95% CI, 0.58, 1.60).

In the second approach, potential confounding variables (social class, smoking, hypertension, cholesterol level, and alcohol intake) were forced into the models. The effect was to increase the point estimates and 95% CIs for ever use of HRT (OR, 0.75; 95% CI, 0.47, 1.18), past use (OR, 0.71; 95% CI, 0.35, 1.41), and current use of HRT (OR, 0.77; 95% CI, 0.44, 1.33). The point estimates remained well below unity, but the increased width and inclusion of unity of the 95% CIs reflected reduced precision due to small numbers.

Given that women who take hormones may have healthier lifestyles with fewer risk factors than women who do not,¹⁸ a comparison was made between subjects who had ever used HRT and those who had never used HRT. There were no significant differences between the groups in relation to age, cholesterol level, BMI, smoking habits, history of hypertension, or level of education. However, more users of HRT drank alcohol (79% versus 67%; P=0.01), and more users were classified into nonmanual social classes 1 to 3 (78% versus 67%; P=0.03). There was no significant difference in risk of SAH between premenopausal and postmenopausal women (age-adjusted OR, 0.95; 95% CI, 0.55, 1.64).

	Discussion

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
In this population-based, case-control study, SAH was more common in women, and there was an inverse association between use of HRT and the risk of SAH. The association was moderately strong for ever use of HRT, but there was only borderline evidence of an inverse relation when current or past use of HRT was considered separately. There was no evidence of an association with use of OCPs. These data are supported by the stable estimates of risk even after adjustment for confounding variables.

The strengths of the study include the large series of cases with uniform and strict diagnostic criteria for SAH. The cases were consecutive, a high proportion had CT (and/or necropsy) confirmation of the diagnosis, and they were derived from a set of population-based registries. Thus, the potential for systematic errors due to referral bias and misclassification of SAH is likely to be very small. Moreover, since enrollment to vote is compulsory in Australia and New Zealand, the controls were also representative of the population that gave rise to the cases. Finally, despite the high case fatality, we were able to obtain information regarding a wide range of exposures for most cases.

Differential error or bias is unavoidable in case-control studies, where information on exposures for cases is obtained under different circumstances from those for controls. In common with other epidemiological studies of SAH,^{6 19 20} we used proxy respondents to obtain information regarding exposures in the high proportion of the cases who were unable to be interviewed because of early death or disability. It is possible that some cases may have underreported certain exposures, such as recent alcohol intake, and proxies might have overestimated them, either of which leads to a bias. Even if efforts to frequency-match for use of proxy respondents in controls had succeeded completely, it is still possible that proxies for controls may have been more (or less) likely to give erroneous responses than proxies for cases.²¹ We have no reason to suspect, however, that any measurement error would have differed appreciably in magnitude between cases and controls. Most cases would not have had any prior knowledge about the hypotheses being investigated, and they were interviewed early after the event, reducing the likelihood for habits to change significantly subsequent to diagnosis. In addition, studies have reported high agreement between index and proxy respondents with regard to personal habits and medical conditions,^{22 23} such as those used in these analyses. Adjustment for use of proxy was made in all multivariate analyses.

Another limitation of this study, however, was that the hormonal data were based on self-report and often proxy report, and information was not confirmed from medical record review, which may have resulted in misclassification of exposures. However, there appears to be high agreement between self-report and medical records regarding the use of HRT.²⁴

The results of earlier epidemiological analyses of the use of HRT and incident SAH have not been consistent. Our data support those of another large population-based, case-control study undertaken by Longstreth et al,⁶ who found that ever use of HRT was associated with a 53% reduction in odds of SAH (OR, 0.47; 95% CI, 0.26, 0.86). Ever use of OCPs also had an inverse association (OR, 0.52; 95% CI, 0.25, 1.10), although this was not significant. Interestingly, they found that the reduction in risk associated with HRT was greatest in women who had smoked compared with those who had never smoked. However, adjustments were made only for age and source of information (index subject or proxy) in the analyses; when smoking was included in the model, the association of use of HRT with risk of SAH was significantly attenuated. Other case-control^{25 26 27} and cohort^{28 29} studies have found no clear association for use of hormonal therapy and SAH, although the estimates were imprecise, as indicated by generally wide 95% CIs. Among the limitations of these studies are the use of select population groups and broader criteria for the diagnosis of SAH.

If estrogen (either alone or in combination with progestogen) reduces the risk of SAH, our results suggest that age at exposure is important. Apart from a later age of first pregnancy, there was no association of menstrual or reproductive variables or use of OCPs with the risk of SAH. The negative association between use of HRT and risk of SAH, however, supports the hypothesis of a protective role only in postmenopausal women. Because of the sample size, it is likely that lack of power is the explanation for the failure to detect a significant association for past and current users despite that seen for ever users of HRT. A number of potential beneficial vascular effects of estrogen, including those on lipid profiles, have been proposed,³⁰ but none is as yet entirely satisfactory as an explanation regarding the propensity toward formation and rupture of cerebral aneurysms.

Since the vast majority of strokes in women occur after the menopause, there has been much hope that HRT would alleviate the burden of stroke, as it may do for cardiovascular disease.³¹ Many studies have been devoted to the relationships between HRT and stroke. Conflicting results have been produced for nonfatal stroke, yet consistent findings have been obtained of decreased risk for fatal stroke .^{31 32} Stratification by fatal and nonfatal strokes may be important in providing indirect evidence concerning HRT and the risks of ischemic and hemorrhagic stroke because the great majority of nonfatal strokes are ischemic, while between approximately one and two thirds of fatal strokes are hemorrhagic, including SAH.³³ There is growing evidence that ischemic and hemorrhagic strokes have qualitatively different associations with certain risk factors such as cholesterol.³⁴ A similar relationship may well hold for HRT, given the present data for SAH together with those from the Nurses’ Health Study⁸ that indicate an increased risk of ischemic stroke among users of HRT (adjusted relative risk, 1.4; 95% CI, 1.02, 1.92).

These data, together with those from other well-designed and -conducted epidemiological studies, provide the best evidence currently available to support a key role for hormonal therapy in the prevention of SAH among postmenopausal women.

	Appendix 1

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
The following are the committee members, principal investigators, and study coordinators of ACROSS: Steering Committee: C. Anderson (Study Chair), N. Anderson, R. Bonita, D. Dunbabin, G. Hankey, K. Jamrozik. Data Management and Statistics: D. Bennett, J. Duncan, S. Vanderhoorn. Study Coordinators: J. Bennett (Study Manager), D. Healy, S. Rubenach (Adelaide); J Sansom, J. Flecker (Hobart); J. Harvey, J. Linto, G. Mann, K. White (Perth); S. Hawkins, C. Mulholland (Auckland). Neurosurgical Investigators: B. Brophy (Flinders Medical Center), J. Liddell (Royal Hobart Hospital), E. Mee (Auckland Hospital), G. MucCulloch (Queen Elizabeth Hospital), N. Knuckey (Sir Charles Gairdner Hospital), P. Riley (Royal Adelaide Hospital). Clinical Centers: Ashford Hospital, Flinders Medical Center, Memorial Hospital, Repatriation General Hospital, Queen Elizabeth Hospital, Royal Adelaide Hospital (Adelaide, South Australia); Royal Hobart Hospital, Calvary Hospital, St Helen’s Private Hospital (Hobart, Tasmania); Fremantle Hospital, Royal Perth Hospital, St John of God Hospital, Sir Charles Gairdner Hospital (Perth, Western Australia); Auckland Hospital, North Shore Hospital, Middlemore Hospital, Waitakare Hospital (Auckland, New Zealand).

	Acknowledgments

 
This study was supported by grants from the National Health and Medical Research Council of Australia, the Health Research Council of New Zealand, and the Sylvia and Charles Viertel Charitable Foundation of Queensland, Australia. Dr Ni Mhurchu undertook this analysis as part of a fellowship awarded by the National Heart Foundation of New Zealand. We are indebted to the study investigators and coordinators for their dedication and performance; Janet Bennett for her efforts; the support of the Coroner’s Department in each center; the assistance of the Australian Bureau of Statistics and Statistics New Zealand; and the help provided from nursing, administration, and medical records staff of the clinical centers.

	Footnotes

 
Members of the study group are listed in the Appendix.

Received July 18, 2000; revision received November 23, 2000; accepted December 6, 2000.

	References

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
1. Shievink WI. Intracranial aneurysms. N Engl J Med. 1997;336:28–40.[Free Full Text]

2. Mayberg MR, Batjer HH, Dacey R, Diringer M, Haley EC, Heros RC, Sternau LL, Torner J, Adams HP, Feinberg W, Thies W. Guidelines for the management of aneurysmal subarachnoid haemorrhage: a statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke. 1994;25:2315–2328.[Medline] [Order article via Infotrieve]

3. Teunissen LL, Rinkel GJE, Algra A, van Gijn J. Risk factors for subarachnoid haemorrhage: a systematic review. Stroke. 1996;27:544–549.[Abstract/Free Full Text]

4. Linn FHH, Rinkel GJE, Algra A, van Gijn J. Incidence of subarachnoid haemorrhage: role of region, year, and rate of computed tomography: a meta-analysis. Stroke. 1996;27:625–629.[Abstract/Free Full Text]

5. Stober T, Sen S, Anstätt T, Freier G, Schimrigk K. Direct evidence of hypertension and the possible role of post-menopausal oestrogen deficiency in the pathogenesis of berry aneurysms. J Neurol. 1985;232:67–72.[Medline] [Order article via Infotrieve]

6. Longstreth WT Jr, Nelson LM, Koepsell TD, van Belle G. Subarachnoid haemorrhage and hormonal factors: a population-based case-control study. Ann Intern Med. 1994;121:168–173.[Abstract/Free Full Text]

7. Thorogood M. Stroke and steroid hormonal contraception. Contraception. 1998;57:157–167.[Medline] [Order article via Infotrieve]

8. Grodstein F, Stampfer MJ, Manson JE, Colditz GA, Willett WC, Rosner B, Speizer PE, Hennekens CH. Postmenopausal estrogen and progestin use and the risk of cardiovascular disease. N Engl J Med. 1996;335:453–461.[Abstract/Free Full Text]

9. Paganini-Hill A. Estrogen replacement therapy and stroke. Prog Cardiovasc Dis. 1995;38:223–242.[Medline] [Order article via Infotrieve]

10. Hannaford PC, Owen-Smith V. Using epidemiological data to guide clinical practice: review of studies on cardiovascular disease and use of combined oral contraceptives. BMJ. 1998;316:984–987.[Abstract/Free Full Text]

11. Australasian Cooperative Research on Subarachnoid Haemorrhage Study Group. The epidemiology of aneurysmal subarachnoid haemorrhage in Australia and New Zealand: incidence and case fatality from the Australasian Cooperative Research on Subarachnoid Haemorrhage Study (ACROSS). Stroke. 2000;31:1843–1850.[Abstract/Free Full Text]

12. Vermeulen M, van Gijn J. The diagnosis of subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry. 1990;53:365–372.[Free Full Text]

13. Australian Bureau of Statistics. Classification and Classified List of Occupations. Canberra, Australia: ABS; 1980.

14. Commonwealth of Australia. The Fitness of Australians. Canberra, Australia: Australian Government Publishing Service; 1992. Publication ISBN0644245867.

15. Statistical Analysis System Institute Inc. SAS Version 6.12. Cary, NC: SAS Inc; 1997.

16. Statistical Solutions LTD. SOLAS for Missing Data Analysis 1.0. Cork, Ireland: Statistical Solutions Inc; 1997.

17. Rothman KJ. Modern Epidemiology. Boston, Mass: Little, Brown & Co; 1998.

18. Goldstein F. Invited commentary: can selection bias explain the cardiovascular benefits of hormone replacement therapy? Am J Epidemiol. 1996;143:979–982.[Free Full Text]

19. Longstreth WT Jr, Nelson LM, Koepsell TD, Schievink WI, van Belle G. Cigarette smoking, alcohol use, and subarachnoid hemorrhage. Stroke. 1992;23:1242–1249.[Abstract/Free Full Text]

20. Juvela S, Hillbom M, Numminen H, Koskinen P. Cigarette smoking and alcohol consumption as risk factors for aneurysmal subarachnoid haemorrhage. Stroke. 1993;24:639–646.[Abstract/Free Full Text]

21. Walker AM, Velema JP, Robins JM. Analysis of case-control data derived in part from proxy respondents. Am J Epidemiol. 1988;127:905–914.[Free Full Text]

22. Tepper A, Connally LB, Haltmeier P, Smith E, Sweeney MH. Knowledge of medical history information among proxy respondents for deceased study subjects. J Clin Epidemiol. 1993;46:1243–1248.[Medline] [Order article via Infotrieve]

23. Nelson LM, Longstreth WT Jr, Koepsell TD, Checkoway H, van Belle G. Completeness and accuracy of interview data from proxy respondents: demographic, medical and lifestyle factors. Epidemiology. 1994;5:204–217.[Medline] [Order article via Infotrieve]

24. Jain MG. Rohan TE, Howe GR. Agreement of self-reported use of menopausal hormone replacement therapy with physician reports. Epidemiology. 1999;10:260–263.[Medline] [Order article via Infotrieve]

25. Pedersen AT, Lidegaard Ø, Kreiner S, Ottesen B. Hormone replacement therapy and risk of non-fatal stroke. Lancet. 1997;350:1277–1283.[Medline] [Order article via Infotrieve]

26. WHO Collaborative Study on Cardiovascular Disease and Steroid Hormone Contraception. Haemorrhagic stroke, overall stroke risk, and combined oral contraceptives: results of an international, multicentre, case-control study. Lancet. 1996;348:505–510.[Medline] [Order article via Infotrieve]

27. Schwarts SM, Petitti DB, Siscovick DS, Longstreth WT Jr, Sidney S, Raghunathan TE, Quesenberry CP, Kelaghan J. Stroke and use of low-dose oral contraceptive in young women: a pooled analysis of two US studies. Stroke. 1998;29:2277–2284.[Abstract/Free Full Text]

28. Falkeborn M, Persson I, Terént A, Adami H, Lithell H, Bergström R. Hormone replacement therapy and the risk of stroke: follow-up of a population-based cohort in Sweden. Arch Intern Med. 1993;153:1201–1209.[Abstract/Free Full Text]

29. Stampfer MJ, Colditz GA, Willett WC, Manson JE, Rosner B, Speizer FE, Hennekens CH. Postmenopausal estrogen therapy and cardiovascular disease: ten-year follow-up from the Nurses’ Health Study. N Engl J Med. 1991;325:756–762.[Abstract]

30. Anderson HV. Estrogen therapy, atherosclerosis, and clinical cardiovascular events. Circulation. 1996;94:1809–1811.[Free Full Text]

31. Bousser MG. Stroke in women. Circulation. 1999;99:463–467.[Free Full Text]

32. Sourander L, Rajala T, Räihä I, Mäkinen J, Erkkola R, Helenius H. Cardiovascular and cancer morbidity and mortality and sudden cardiac death in postmenopausal women on oestrogen replacement therapy (ERT). Lancet. 1998;352:1965–1969.[Medline] [Order article via Infotrieve]

33. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. A prospective study of acute cerebrovascular disease in the community: the Oxfordshire Community Stroke Project—1981–86, II: incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intracerebral and subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry. 1990;53:16–22.[Abstract/Free Full Text]

34. Eastern Stroke and Coronary Heart Disease Collaborative Research Group. Blood pressure, cholesterol, and stroke in eastern Asia. Lancet. 1998;352:1801–1807.[Medline] [Order article via Infotrieve]

Editorial Comment

The Gender Gap in Aneurysmal Subarachnoid Hemorrhage

W.T. Longstreth, Jr, MD, MPH, Guest Editor

Departments of Neurology and Epidemiology, University of Washington, Seattle, Washington

Lorene M. Nelson, PhD, Guest Editor

Division of Epidemiology, Department of Health Research & Policy, Stanford University Stanford, California

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
Why does aneurysmal subarachnoid hemorrhage affect women more than men? The answer to this question may provide clues about the pathophysiology and prevention of this deadly condition. The occurrence of a disease can be reduced if etiologic risk factors can be identified and modified. Although successful clipping or coiling of an aneurysm that has spilled blood into the subarachnoid space can eliminate the risk for rebleed, such treatments are imperfect with the majority of patients with an aneurysmal subarachnoid hemorrhage being left dead or disabled in most series.^R1 The high morbidity and mortality is made worse by the fact that most of those who suffer aneurysmal subarachnoid hemorrhage are under age 65, unlike most other types of stroke.

The search for etiologic risk factors for aneurysmal subarachnoid hemorrhage is complicated because different factors may influence the risk of aneurysm formation, rupture, or both.^R2 To date, some of the strongest evidence exists for cigarette smoking, hypertension and alcohol use.^R3 These risk factors are more common in men than women so these factors alone do not explain the excess of aneurysmal subarachnoid hemorrhage in women. If the effect of these risk factors were more potent in women than men, the excess of aneurysmal subarachnoid hemorrhage in women could be explained. The studies of risk factors that have sought evidence for such effect modification have not found it.^R3 Nonetheless, one of the many benefits of controlling these risk factors would likely be a reduced risk of subarachnoid hemorrhage for both women and men.

Hormonal factors deserve careful attention in a disease that affects women more than men. Observational studies have addressed the use of oral contraceptive pills and hormone replacement therapy. The studies performed to date have not resolved the issue, and thus the work from Dr. Mhurchu and colleagues is a welcome addition. In a population-based study of subarachnoid hemorrhage from Australia and New Zealand, the investigators examined the risk associated with oral contraceptive pills and hormone replacement therapy. They found no association with ever use of oral contraceptive pills (adjusted odds ratio 0.97 with a 95% confidence interval (CI) of 0.58 to 1.60). A recent meta-analysis of oral contraceptive pills indicated a small but significant increased risk of subarachnoid hemorrhage (relative risk estimate 1.42, 95% CI 1.12 to 1.80).^R4 The risk was less when analyses were limited to studies of oral contraceptive pills containing low doses of estrogens, as are most commonly used nowadays. The results of the current study when added to the meta-analysis will edge the risk estimate even closer to 1.0. Of note, in a pooled analysis from two recent US case-control studies, too few users of the low-dose oral contraceptive pills were cigarette smokers or hypertensive to allow the investigators to examine whether these factors modified the small risk associated with oral contraceptive pills, as was suggested for high-dose preparations commonly used in the past.^R5 The investigators recommended maintaining cigarette smoking and hypertension as contraindications to prescribing oral contraceptive pills, until evidence suggests otherwise.

Less work has been done with hormone replacement therapy, which may be more pertinent than oral contraceptive pills because the incidence of subarachnoid hemorrhages is higher after menopause than before.^R6 Dr. Mhurchu and colleagues found that ever use of hormone replacement therapy was associated with a reduced risk of subarachnoid hemorrhage (adjusted odds ratio 0.64, 95% CI 0.41 to 0.98). The findings were consistent with a similarly designed study in which the reduced risk with hormone replacement therapy was limited to those who had ever smoked.^R7 Other observational studies have not confirmed these findings but have potential shortcomings as reviewed by Dr. Mhurchu and colleagues. Concern remains that a protective effect of hormone replacement therapy may be an artifact resulting from residual confounding by some unmeasured or poorly measured factor. Such concerns are heightened by the results of a recent clinical trial in which hormone replacement therapy did not reduce the risk of coronary heart disease, despite the preponderance of evidence from observational studies indicating that such treatment should be beneficial.^R8

We are left with an incomplete understanding of why aneurysmal subarachnoid hemorrhage affects women more than men, or for that matter, why women are more likely than men to have multiple aneurysms.^R9 Perhaps the loss of some beneficial effect of estrogens on cerebral blood vessels increases the risk of subarachnoid hemorrhage. Men would be at less risk because they do not experience such a dramatic estrogen withdrawal, as do women. Alternatively, some interaction of other environmental and genetic factors could result in an increased risk. Genetic factors have just started to be explored but will still need to be able to explain why this disease affects women more than men. Because it is a relatively common cause of stroke in women under age 65 and because of its high morbidity and mortality, the excess of aneurysmal subarachnoid hemorrhage in women remains a pressing question in stroke research.

Received July 18, 2000; revision received November 23, 2000; accepted December 6, 2000.

	References

Top Abstract Introduction Subjects and Methods Results Discussion Appendix 1 References Introduction  References

 
1. Hop JW, Rinkel GJ, Algra A, van Gijn J. Case-fatality rates and functional outcome after subarachnoid hemorrhage: a systematic review. Stroke. 1997;28:660–664.[Abstract/Free Full Text]

2. Torner JC. Epidemiology of subarachnoid hemorrhage. Semin Neurol. 1984;4:354–369.

3. Teunissen LL, Rinkel GJ, Algra A, van Gijn J. Risk factors for subarachnoid hemorrhage: a systematic review. Stroke. 1996;27:544–549.

4. Johnston SC, Colford JM Jr, Gress DR. Oral contraceptives and the risk of subarachnoid hemorrhage: a meta-analysis. Neurology. 1998;51:411–418.[Abstract/Free Full Text]

5. Schwartz SM, Petitti DB, Siscovick DS, Longstreth WT Jr, Sidney S, Raghunathan TE, Quesenberry CP Jr, Kelaghan J. Stroke and use of low-dose oral contraceptives in young women: a pooled analysis of two US studies. Stroke. 1998;29:2277–2284.

6. Linn FH, Rinkel GJ, Algra A, van Gijn J. Incidence of subarachnoid hemorrhage: role of region, year, and rate of computed tomography: a meta-analysis. Stroke. 1996;27:625–629.

7. Longstreth WT Jr, Nelson LM, Koepsell TD, van Belle G. Subarachnoid hemorrhage and hormonal factors in women. A population-based case-control study. Ann Intern Med. 1994;121:168–173.

8. Hulley S, Grady D, Bush T, Furberg C, Herrington D, Riggs B, Vittinghoff E. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/Progestin Replacement Study (HERS) Research Group JAMA. 1998;280:605–613.[Abstract/Free Full Text]

9. Ostergaard JR, Hog E. Incidence of multiple intracranial aneurysms. Influence of arterial hypertension and gender. J Neurosurg. 1985;63:49–55.[Medline] [Order article via Infotrieve]

This article has been cited by other articles:

				M. S. Sandvei, P. R. Romundstad, T. B. Muller, L. Vatten, and A. Vik Risk Factors for Aneurysmal Subarachnoid Hemorrhage in a Prospective Population Study: The HUNT Study in Norway Stroke, June 1, 2009; 40(6): 1958 - 1962. [Abstract] [Full Text] [PDF]

				P. Appelros, B. Stegmayr, and A. Terent Sex Differences in Stroke Epidemiology: A Systematic Review Stroke, April 1, 2009; 40(4): 1082 - 1090. [Abstract] [Full Text] [PDF]

				N K de Rooij, F H H Linn, J A van der Plas, A Algra, and G J E Rinkel Incidence of subarachnoid haemorrhage: a systematic review with emphasis on region, age, gender and time trends J. Neurol. Neurosurg. Psychiatry, December 1, 2007; 78(12): 1365 - 1372. [Abstract] [Full Text] [PDF]

				C.-Y. Yang, C.-C. Chang, H.-W. Kuo, and H.-F. Chiu Parity and risk of death from subarachnoid hemorrhage in women: Evidence from a cohort in Taiwan Neurology, August 8, 2006; 67(3): 514 - 515. [Abstract] [Full Text] [PDF]

				V. L. Feigin, G. J.E. Rinkel, C. M.M. Lawes, A. Algra, D. A. Bennett, J. van Gijn, and C. S. Anderson Risk Factors for Subarachnoid Hemorrhage: An Updated Systematic Review of Epidemiological Studies Stroke, December 1, 2005; 36(12): 2773 - 2780. [Abstract] [Full Text] [PDF]

				C. S. Anderson, V. Feigin, D. Bennett, R.-B. Lin, G. Hankey, K. Jamrozik, and for the Australasian Cooperative Research on Subar Active and Passive Smoking and the Risk of Subarachnoid Hemorrhage: An International Population-Based Case-Control Study Stroke, March 1, 2004; 35(3): 633 - 637. [Abstract] [Full Text] [PDF]

				D. Gaist, L. Pedersen, S. Cnattingius, and H. T. Sorensen Parity and Risk of Subarachnoid Hemorrhage in Women: A Nested Case-Control Study Based on National Swedish Registries Stroke, January 1, 2004; 35(1): 28 - 32. [Abstract] [Full Text] [PDF]

				A. S. Dumont and N. F. Kassell Editorial Comment--Parity and Risk of Subarachnoid Hemorrhage: An Emerging Association Stroke, January 1, 2004; 35(1): 32 - 33. [Full Text] [PDF]

				S. Wills, A. Ronkainen, M. van der Voet, H. Kuivaniemi, K. Helin, E. Leinonen, J. Frosen, M. Niemela, J. Jaaskelainen, J. Hernesniemi, et al. Familial Intracranial Aneurysms: An Analysis of 346 Multiplex Finnish Families Stroke, June 1, 2003; 34(6): 1370 - 1374. [Abstract] [Full Text] [PDF]

				J. P. Broderick, C. M. Viscoli, T. Brott, W. N. Kernan, L. M. Brass, E. Feldmann, L. B. Morgenstern, J. L. Wilterdink, and R. I. Horwitz Major Risk Factors for Aneurysmal Subarachnoid Hemorrhage in the Young Are Modifiable Stroke, June 1, 2003; 34(6): 1375 - 1381. [Abstract] [Full Text] [PDF]

				H. Ohkuma, H. Tabata, S. Suzuki, and M. S. Islam Risk Factors for Aneurysmal Subarachnoid Hemorrhage in Aomori, Japan Stroke, January 1, 2003; 34(1): 96 - 100. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mhurchu, C. N. Articles by Nelson, L. M. Search for Related Content PubMed PubMed Citation Articles by Mhurchu, C. N. Articles by Nelson, L. M. Pubmed/NCBI databases Medline Plus Health Information Hormones Related Collections Cerebral Aneurysm, AVM, & Subarachnoid hemorrhage Primary and Secondary Stroke Prevention Risk Factors for Stroke