Risk of Aneurysmal Subarachnoid Hemorrhage
The Role of Confirmed Hypertension
Background and Purpose— From the studied variables in subarachnoid hemorrhage (SAH) risk, hypertension is probably the most controvertible one. The aim of this study was to determine whether confirmed hypertension, prospectively diagnosed with strict criteria, is an independent risk factor for aneurysmal SAH.
Methods— A case-control study was conducted in 2 colombian cities between July 2004 and June 2005. There were 163 new cases of SAH (mean age 51 years; 107 were women) with 2 controls per case: 1 hospital and 1 community control. Hypertension was defined according to cardiovascular criteria, based on target organ damage. In addition to hypertension, other variables were studied: present smoking, recent alcohol consumption, alcohol dependency, coffee consumption, cocaine use, and body mass index. A multivariate logistic regression model was used to determine whether hypertension was an independent risk factor.
Results— Among the studied variables, including confirmed hypertension, only present smoking became an independent risk factor for SAH.
Conclusions— Confirmed hypertension is not an independent risk factor for aneurysmal SAH.
Initial bleeding impact accounts for most deaths in aneurysmal subarachnoid hemorrhage (SAH); mortality could decrease if incidence rate is reduced, and then it is necessary to identify risk factors, of which only cigarette smoking is indisputable. All other variables, including hypertension, are still subject to debates.1–3 The study of hypertension in SAH is problematic as patients, because of Cushing reflex, frequently show elevated blood pressure, irrespective of being or not hypertensive.4 The aim of this study was to determine whether confirmed hypertension, prospectively diagnosed with target organ damage criteria, is an independent risk factor for SAH.
Subjects and Methods
A case-control study with incident SAH cases was performed, with patients hospitalized in San Vicente de Paul Hospital, Hospital General, and Clinica Leon XIII from Medellin, or in Hospital Evaristo Moreno from Cali (Colombia). Two controls per case were selected: 1 hospital and 1 community control. Hospitalized controls were patients with nontraumatic acute abdomen (appendicitis, cholecystitis, among others) and community controls were cases’ friends. Sampling size was calculated to detect an Odds Ratio of 2.0, with an expected proportion of 15.7% of hypertension among controls, and a case-control ratio of 1:2; thus, 159 cases and 318 controls were included. The studied variables were: SAH (independent), cigarette smoking (in addition with the 6-item Fagerstrom test for nicotine dependence, with a 0 to 10 score),5 recent alcohol consumption, alcohol dependency according to CAGE question-naire,2 confirmed hypertension (fulfilling 1 of these criteria: antihypertensive medication intake previously to SAH prescribed by a physician, hypertension preceding SAH confirmed by a physician; if criteria 1 and 2 were not present, patients were evaluated by a cardiologist and were diagnosed as hypertensive if complying with 1 of these criteria: hypertensive retinopathy to fundoscopy,6 hypertensive cardiopathy according to ECG-Sokolow index, or echocardiography, based on left ventricle mass7), coffee consumption, cocaine use, family history, and body mass index (BMI).
Two categories were created with those discrete variables which showed more than 2 individuals. An univariate analysis was conducted, and then a multivariate logistic regression model was used to control confusion biases. According to the stepwise forward model, variables that showed a probability value <0.25 in the univariate analysis were included in the multivariate model. Significance levels of 5% and 95% confidence intervals were applied; the odds ratio (OR) was used as the association measure. Information was processed with the SPSS version 11.5 statistical program.
There were 163 cases, 158 hospital controls, and 158 community controls. Mean age was 49.1 years, and female sex was predominant with a 3:2 women-men ratio. The demographic and clinical characteristics are shown in Table 1. An initial univariate analysis showed cigarette smoking, nicotine dependence, and coffee consumption significantly associated with SAH. A logistic regression multivariate analysis (Table 2) demonstrated that only cigarette smoking, evaluated by the Fagerstrom Index, was a risk factor for SAH, with an adjusted OR of 5.74 and a confidence interval between 3.22 and 10.23. Neither confirmed hypertension nor any of the other variables were independent risk factors for SAH.
The knowledge of risk factors for SAH is incomplete, particularly because many studies are limited by scarce number of patients and inconsistent diagnostic criteria.8 Very few studies refute cigarette-aneurysmal SAH association.9 Hypertension, alcohol, BMI, and addiction to substances have contradictory results,2,10 and hypertension is probably the most controvertible one.11,12 Moreover, blood pressure, influenced by Cushing reflex,4 can abruptly increase in patients with no hypertension as a consequence of SAH, which makes it difficult to determine whether hypertension is caused by SAH or preceded it. Another problem appears in many studies of risk factors for hemorrhagic events when no difference is made between SAH and hypertensive intracerebral hemorrhage; hypertension can falsely show an association with SAH, if we consider its more solid role as a risk factor for intracerebral hemorrhage than for SAH.12 Consequently, hypertension should only be evaluated in prospective follow-up studies, in which hy-pertension can be assessed by objective criteria, as the damage in target organs.13
Accordingly, in this study hypertension has been diagnosed mostly by the damage in target organs, relegating tensional figures. Hypertension can be diagnosed in apparently healthy individuals through procedures such as echocardiography, based on left ventricle mass study,7 ECG by Sokolow-Lyon voltage criteria or Cornell voltage duration product,4,7 or fundoscopy, according to Keith-Wagener-Barker criteria.4,6
As far as we know, this is the first study concerning the association between hypertension and SAH in which hypertension has been prospectively confirmed. According to the multivariate analysis, hypertension is not an independent risk factor for SAH.
Present cigarette smoking is the only identifiable risk factor for SAH, with a dose-response effect. Confirmed hypertension is not an independent risk factor for SAH, provided that hypertension is prospectively determined with objective criteria.
We are very grateful to Dr Seppo Juvela, from the University of Helsinki, for his valuable comments on a previous version of the manuscript. We also thank Patricia Ballesteros-Nova for her assistance in successfully collecting and processing data.
Sources of Funding
This work was funded by the Research Committee (CODI), University of Antioquia, in Medellín, Colombia.
- Received September 5, 2007.
- Accepted September 13, 2007.
Okamoto K, Horisawa R, Kawamura T, Asai A, Ogino M, Takagi T, Ohno Y. Family history and risk of subarachnoid hemorrhage: a case-control study in Nagoya, Japan. Stroke. 2003; 34: 422–426.
van der Born B, Hulsman C, Hoekstra J, Schlingemann R, Van Montfrans G. Value of routine funduscopy in patients with hypertension: systematic review. BMJ. 2005; 331: 73–77.
Feigin V, Rinkel GJ, Lawes CMM, Algra A, Bennett D, Gijn JV, Anderson C. Risk factors for subarachnoid hemorrhage. An updated systematic review of epidemiological studies. Stroke. 2005; 36: 2773–2780.
Kleinpeter G, Schatzer R, Bock F. Is blood pressure really a trigger for the circadian rhythm of subarachnoid hemorrhage? Stroke. 1995; 26: 1805–1810.
Kim HC, Nam C, Jee SH, Suh I. Comparison of blood pressure-associated risk of intracerebral hemorrhage and subarachnoid hemorrhage. Korea Medical Insurance Corporation Study. Hypertension. 2005; 46: 393–397.