Subarachnoid hemorrhage (SAH) accounts for 1% to 10%

Transcription

1 Lung Function as a Risk Factor for Subarachnoid Hemorrhage A Prospective Cohort Study Martin Söderholm, MD; Elisabet Zia, MD, PhD; Bo Hedblad, MD, PhD; Gunnar Engström, MD, PhD Background and Purpose The etiology of subarachnoid hemorrhage (SAH) is poorly understood. Reduced lung function, expressed as low forced expiratory volume in 1 second (FEV 1 ) and low forced vital capacity (FVC), is a predictor of cardiovascular disease, but whether reduced lung function is a risk factor for SAH is not known. The association between lung function and incidence of SAH was investigated in a prospective cohort study. Methods Between 1974 and 1992, men and 7237 women (mean age, 44 years) were examined in a health screening program including spirometry. The incidence of SAH was studied during a mean follow-up of 26 years in relation to age- and height-standardized FEV 1, FVC, and FEV 1 /FVC. Results One hundred forty-five subjects had a SAH (18.3 per person-years in men and 26.5 per person-years in women). The hazard ratio for SAH in the lowest compared to the highest quartile of FEV 1 and FEV 1 /FVC was 2.24 (95% CI, ; P for trend 0.014) and 1.92 (95% CI, ; P for trend 0.003), respectively, after adjustment for several confounding factors including smoking and hypertension. The results persisted when analysis was restricted to nonsmokers. FVC showed no significant association with incidence of SAH. Conclusions Baseline lung function, expressed as low FEV 1 or FEV 1 /FVC, is a risk factor for SAH, independently of smoking. (Stroke. 2012;43: ) Key Words: epidemiology forced expiratory volume forced vital capacity prospective study spirometry subarachnoid hemorrhage Subarachnoid hemorrhage (SAH) accounts for 1% to 10% of all strokes worldwide. 1 SAH is associated with higher mortality and lower age at onset than other types of stroke and therefore causes a loss of productive life years, which is comparable to all ischemic strokes. 2 The risk factors for SAH include age, female sex, family history of SAH, smoking, hypertension and excessive alcohol intake, whereas hypercholesterolemia, diabetes mellitus, and overweight might be protective factors. 3 5 In addition, high dietary intakes of fat and salt have been associated with increased incidence of SAH. 6 Studies have also shown that rigorous physical activity may trigger SAH, 7 although a high level of long-term leisure time physical activity could decrease the risk. 8 Hence, there are relatively few established risk factors for SAH. Identification of additional factors could improve prevention and also provide clues to the vascular pathology preceding SAH. Reduced lung function, expressed as forced expiratory volume in one second (FEV 1 ) and forced vital capacity (FVC), is a predictor of cardiovascular disease and mortality. 9,10 Furthermore, low lung function is related to an increased risk of all-cause stroke as well as ischemic stroke. 14,15 The relationships seem to be independent of smoking, and significant results have been reported also in studies of lifelong never smokers. 10,15 The reasons for these associations are still unclear. Whether low lung function is associated with the incidence of SAH is, to the best of our knowledge, not previously studied. We hypothesized that reduced lung function could be associated with a higher incidence of SAH, perhaps due to common pathogenic mechanisms in vessel wall degradation and destruction of lung parenchyma. The relationship between lung function and the incidence of SAH was evaluated in a large prospective cohort from an urban population. Methods Study Population A screening program with the aim to detect individuals at high risk for cardiovascular disease, the Malmö Preventive Project, was Received March 22, 2012; final revision received July 11, 2012; accepted July 12, From the Cardiovascular epidemiology research group, department of Clinical Sciences Malmö, Lund University, Lund, Sweden (M.S., E.Z., B.H., G.E.); and the Department of Neurology, Skåne University Hospital, Malmö, Sweden (M.S., E.Z.). Emmanual Touzé, PhD, was the guest editor for this article. The online-only Data Supplement is available with this article at /-/DC1. Correspondence to Gunnar Engström, MD, PhD, Cardiovascular Epidemiology Research Group, SUS Malmö, CRC, Malmö, Sweden. Gunnar.Engstrom@med.lu.se 2012 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 Söderholm et al Lung Function and Incidence of SAH 2599 conducted in Malmö, Sweden, between 1974 and Complete birth cohorts, born between 1921 and 1949, were invited to a health examination including a physical examination, a panel of laboratory tests, and a self-administered questionnaire. A total of men were examined from 1974 to 1984 and women were examined between 1977 and The participation rate was 71%. The Health Service Authority of Malmö approved the screening program. Linkage with the national cause of death and patient registers was approved by the regional ethics committee at Lund University. Spirometry was performed in complete birth cohorts during most but not all of the screening period. In total, (94%) men and 7748 (71%) women underwent spirometry. From these participants, subjects with a history of stroke (n 30), myocardial infarction (n 258), or angina pectoris (n 466) as well as those with missing information about blood pressure (n 18), alcohol consumption (n 233), body mass index (n 3), cholesterol (n 55), diabetes (n 125), physical inactivity (n 183), and erythrocyte sedimentation rate (ESR; n 81), were excluded. Additionally, subjects with very high values ( 50 mm) of ESR (n 55) were also excluded, because those high values are supposedly associated with severe current illness, for example, pneumonia or cancer. A total of individuals ( men and 7237 women) with a mean age ( SD) of 44 7 years were included. Men were somewhat younger than women (mean age, 43 7 years; range, years and 47 8 years; range, years, respectively). Baseline Examinations Blood pressure (mm Hg) was measured twice in the right arm after 10 minutes of rest. The average of 2 measurements was used. A sphygmomanometer and a rubber cuff of appropriate size were used. Height (cm) and weight (kg) were measured under standardized conditions and body mass index was calculated as weight/height 2 (kg/m 2 ). Smoking habits, alcohol consumption, use of antihypertensive medication, and physical inactivity were evaluated in a questionnaire. The subjects were categorized as smokers or nonsmokers based on their self-reported current smoking status. Alcohol abuse was assessed by means of the modified shortened version of the Michigan Alcoholism Screening Test. 17 Subjects with 2 (of 9) affirmative answers were considered to have high alcohol consumption. Physical inactivity in spare time was assessed using the question: Are you mostly engaged in sedentary activities in spare time, for example, watching TV, reading, going to the movies? Blood samples were taken after an overnight fast. Serum cholesterol and blood glucose were analyzed with standard methods at the hospital laboratory. Diabetes was defined as fasting whole blood glucose 6.1 mmol/l or self-reported treatment for diabetes. ESR was determined according to the Westergren method. FVC and FEV 1 were measured using a Spirotron apparatus (Drägerwerk AG, Lübeck, Germany) with the subjects in a standing position without nose clips. Specially trained nurses performed the tests. One acceptable maneuver with respect to the subject s performance and cooperation was required. The volumes were standardized for age and height using equations derived from linear regressions of never smokers in the present cohort. 18,19 The equations are available in the online-only Data Supplement. FEV 1, FVC, and FEV 1 /FVC are expressed as percentage of the predicted values. Retrieval and Ascertainment of Cases All subjects were followed from the baseline examination until first SAH, death, emigration, or December 31, 2008, whichever came first. Follow-up time was calculated for each individual by subtracting the day of examination from the last day of follow-up. Cases with SAH were retrieved from the Malmö stroke register. 20 To find cases that were treated for SAH in hospitals outside of Malmö, cases with SAH before 1989 and cases that died of SAH before reaching the hospital, the Swedish hospital discharge register and the causes of death register were used (International Classification of Diseases, 8 th Revision and International Classification of Diseases, 9 th Revision code 430, International Classification of Diseases, 10 th Revision code I60). These registers provide complete coverage of discharges and deaths, respectively, in Sweden. Diagnoses recorded in these registers are made by physicians at hospital discharge or death of the patient. Statistical Analysis Sex-specific quartiles of FEV 1, FVC, and FEV 1 /FVC were constructed with equal proportions of men and women in each quartile to adjust for the relationship between sex and SAH. One-way analysis of variance (for continuous variables) and logistic regression (for dichotomous variables) were used for analyses of the risk factor distribution across quartiles of FEV 1 and FVC. General linear models were used to estimate adjusted mean values and 95% CIs for FEV 1, FVC, and FEV 1 /FVC in SAH cases and non-sah subjects stratified by smoking status. The Kaplan-Meier test with the log rank statistics was used to study incidence of SAH over quartiles of FEV 1, FVC, and FEV 1 /FVC. Cox proportional hazards regression was used to calculate risk factor-adjusted hazard ratios (HRs) and 95% CIs. The initial model was unadjusted. The final model included age, sex, established risk factors for SAH 5 (ie, smoking, systolic blood pressure, antihypertensive medication, high alcohol consumption), and other potential confounders associated with lung function at baseline (ie, body mass index, cholesterol, diabetes, physical inactivity, ESR). Smoking, sex, antihypertensive treatment, high alcohol consumption, diabetes, and physical inactivity were fitted as dichotomous variables, whereas all other variables were continuous. Due to the long duration of the screening period, the Cox regression models were stratified by screening year (in 3-year groups, , , etc). The fit of the proportional hazards model was checked visually by plotting the incidence rates over time and by entering time-dependent variables into the model. Possible interactions of lung volumes with age, sex, smoking, systolic blood pressure, antihypertensive medication, hypertension (ie, blood pressure 140/90 mm Hg or taking antihypertensive medication), and high alcohol consumption, respectively, were explored by introducing interaction terms in the adjusted Cox models. The Cox regression analysis was also performed in subgroups defined by sex, smoking, and hypertension, respectively. In addition, we conducted the Cox regression analyses using the unadjusted lung volumes, instead of the standardized volumes, including height as a covariate in the full model. Results Baseline Characteristics Baseline characteristics by quartiles of FEV 1 and FVC, respectively, are shown in Table 1 and in online-only Data Supplement Table I. Low values of FEV 1 and FVC were associated with current smoking, somewhat higher systolic blood pressure, using antihypertensive medication, high alcohol consumption, diabetes, physical inactivity, and high levels of cholesterol, body mass index, and ESR. In men, FEV 1, FVC, and FEV 1 /FVC were 95.3% 17.7%, 97.0% 16.2%, and 98.2% 11.4% of predicted, respectively. In women, the corresponding values were 95.8% 17.8%, 97.4% 16.8%, and 99.7% 10.8%. Age at screening, sex, and unadjusted and adjusted mean lung volumes are presented by smoking status and SAH in online-only Data Supplement Table II. Incidence of SAH During the follow-up period (mean, years), 98 men and 47 women had a SAH, corresponding to an overall crude incidence of 20.3 per person-years (26.5 in women and 18.3 in men). Mean time from screening to SAH was years (range, years) and mean age at SAH was years (range, years).

3 2600 Stroke October 2012 Table 1. Baseline Characteristics in Relation to Quartiles of FEV 1 Q4 Q3 Q2 Q1 (low) P Value for Trend All FEV 1, % of predicted No Women, % Age, y Height, cm Current smoking, % Systolic blood pressure, mm Hg Antihypertensive medication, % High alcohol consumption, % Diabetes, % Cholesterol, mmol/l BMI, kg/m Physical inactivity, % ESR, mm Values are mean SD or percentages. Q indicates quartile; FEV 1, forced expiratory volume in 1 second; BMI, body mass index; ESR, erythrocyte sedimentation rate. The incidence of SAH by quartiles of FEV 1 is described in the Figure, and the HRs for SAH in relation to lung volumes are shown in Table 2. Low FEV 1 and low FEV 1 /FVC were significantly associated with increased incidence of SAH. For the lowest (first) compared with the highest (fourth) quartile of FEV 1, the unadjusted HR (95% CI) was 2.64 ( ). The increased risk remained significant after taking established risk factors for SAH into account (HR, 2.21; CI, ). Further adjusting for body mass index, cholesterol, diabetes, physical inactivity, and ESR did not substantially change the HR (HR, 2.24; CI, ; Table 2). FVC showed no significant association with the incidence of SAH. There was a significant interaction between FEV 1 and age (P 0.010) for incidence of SAH, indicating a stronger relationship between FEV 1 and SAH in older age groups. The final Cox models were rerun using the unadjusted volumes, with height added as a covariate, instead of the percentages of the predicted values. The results were essentially the same. Figure. Incidence of subarachnoid hemorrhage (SAH) in relation to quartiles of forced expiratory volume in 1 second (FEV 1 ). P for trend across quartiles (log-rank test). Smoking and Hypertension The relationships between incidence of SAH and quartiles of FEV 1 and FEV 1 /FVC, respectively, remained statistically significant when analyses were restricted to nonsmokers. In smokers, the HR (first versus fourth quartile) after full adjustment was 1.52 ( ) for FEV 1 (P for trend 0.145) and 1.53 ( ) for FEV 1 /FVC (P for trend 0.098; online-only Data Supplement Table III). In normotensive subjects, the HR (first versus fourth quartile) was 1.96 ( ) for FEV 1 (P for trend 0.078) and 1.66 ( ) for FEV 1 /FVC (P for trend 0.067) after full adjustment. Corresponding values in hypertensive subjects were 2.46 ( ) for FEV 1 (P for trend 0.083) and 2.61 ( ) for FEV 1 /FVC (P for trend 0.018). There were no significant interactions between lung volumes and smoking, systolic blood pressure, antihypertensive medication, or hypertension, respectively. Men and Women In the final model, HR in men (first versus fourth quartile) was 2.68 for FEV 1 (CI, ; P for trend 0.013) and 1.68 for FEV 1 /FVC ( ; P for trend 0.036). The corresponding figures in women were 1.47 ( ; P for trend 0.509) and 2.42 ( ; P for trend 0.039), respectively. There was no significant interaction between sex and lung function on incidence of SAH. Discussion In this prospective population-based study, incidence of SAH was higher in subjects with reduced lung function, expressed as a low FEV 1, and in subjects with pulmonary obstruction in terms of a low FEV 1 /FVC. The inverse associations between lung volumes and SAH remained significant after adjustment for the major risk factors for SAH. 5 Furthermore, results were consistent when analysis was restricted to nonsmokers. It is concluded that reduced lung function is a risk factor for SAH.

4 Söderholm et al Lung Function and Incidence of SAH 2601 Table 2. Hazard Ratios for SAH in Relation to Quartiles of FEV 1, FVC, and FEV 1 /FVC Q4 Q3 Q2 Q1 (low) P Value for Trend Per 10% Decrease FEV 1 No SAH, no HR (95% CI) ( ) 1.75 ( ) 2.64 ( ) ( ) HR* ( ) 1.58 ( ) 2.24 ( ) ( ) FVC No SAH, no HR ( ) 1.33 ( ) 1.21 ( ) ( ) HR* ( ) 1.22 ( ) 1.04 ( ) ( ) FEV 1 /FVC No SAH, no HR ( ) 2.33 ( ) 2.08 ( ) ( ) HR* ( ) 2.30 ( ) 1.92 ( ) ( ) SAH indicates subarachnoid hemorrhage; FEV 1, forced expiratory volume in 1 second; FVC, low forced vital capacity; Q, quartile; HR, hazard ratio. *Adjusted for sex, age, smoking, systolic blood pressure, antihypertensive medication, high alcohol consumption, cholesterol, diabetes, body mass index, physical inactivity, and erythrocyte sedimentation rate. Previous prospective studies have shown that reduced lung function is associated with an increased incidence of ischemic stroke, 14,15 stroke from all causes as well as cardiovascular disease and mortality. 9 These relationships are also proved in lifelong never-smokers. 10,15 There is also an association between asthma and the risk of all-cause stroke. 21 In the present cohort of middle-aged, healthy individuals, the risk of SAH was more than doubled in subjects with FEV 1 in the lowest quartile as compared with those with FEV 1 in the highest quartile taking several possible confounders into account. This effect on incident SAH is comparable with the effect of hypertension and smoking, respectively. 5 One question is whether residual confounding due to smoking influenced the results. However, the results were adjusted for smoking and the HR for SAH in subjects with reduced lung function was even higher when the analysis was restricted to nonsmokers. Also, a substudy from this cohort showed that the assessment of smoking had high validity, for example, with regard to concentrations of carboxylated hemoglobin. 22 This supports the view that the association between reduced lung function and higher SAH incidence is independent of smoking. Hypertension is another important risk factor for SAH. Previous studies have demonstrated inverse relationships between lung function and blood pressure both in cross-sectional and longitudinal designs. 23,24 The present results were independent of systolic blood pressure and antihypertensive treatment at baseline, and there was no significant interaction with hypertension. However, we do not know whether subjects with low lung function had higher blood pressure increase during the follow-up period. Because low FEV 1 has been associated with future blood pressure increase, 24 this could be a potential link between lung function and SAH. The association between lung volumes and SAH was significant in men. In women, incidence of SAH was related to FEV 1 /FVC but not to FEV 1. This is likely due to the lower number of events among women, causing insufficient power, and there was no significant interaction between sex and lung function. The incidence of SAH was higher in women than in men, which is consistent with previous findings. 3 Sex differences in risk factors for SAH have been proposed, 5,25 but whether this is the case for lung function cannot be concluded from the present study. The relationships between lung function and SAH were also stronger in nonsmokers than in smokers. The weaker relationships in smokers could possibly be explained by variations among smokers with respect to smoking habits and tobacco consumption. Several hypotheses may offer an explanation for a causal association between low lung volumes and SAH. Inflammation is one possible cause. It is not known whether systemic inflammation is a risk factor for SAH, but the degree of local inflammation in intracranial aneurysmal walls might predict the risk of aneurysm rupture. 26,27 Reduced lung function is associated with high levels of inflammatory markers, 28,29 which was also demonstrated in the present study. However, the associations of low FEV 1 and low FEV 1 /FVC, respectively, with incident SAH were not weakened after adjustment for ESR. Thus, the results do not support the view that inflammation explains the association between SAH and low lung volumes. Besides inflammation, degradation of extracellular matrix proteins plays an important role in lung parenchyma destruction leading to reduced elasticity of the lung and a lowered FEV In the formation of intracranial aneurysms, disruption of the extracellular matrix in the arterial wall is also supposed to be a central mechanism 31 and variations in genes coding for proteins that are important in the extracellular

5 2602 Stroke October 2012 matrix have been associated with intracranial aneurysms. 32,33 A common predisposition for extracellular matrix degradation could perhaps explain the relation between lung function and SAH. Because almost all patients with SAH are admitted to the hospital, linkage to hospital-based registers is appropriate for case retrieval. In the local stroke registry, all cases were verified by a review of the patient records. The validity of SAH diagnoses in the national hospital discharge registry was supported by information about which clinic the patient was treated in and what kind of surgical procedure that was performed. Of SAH cases identified in the causes of death registry, cause of death was based on autopsy for 92% of the cases. Furthermore, the autopsy frequency in subjects dying unexpectedly outside of hospital has been high in Malmö in the last decades. Some potential methodological shortcomings need to be considered. The prospective population-based design, the high participation rate, and the relatively high number of cases are major strengths of this study. However, despite the large size of the cohort, the number of events is still too small to perform analyses of subgroups. Furthermore, neuroimaging is performed as routine diagnostic work-up in almost all patients with suspicion of stroke in Sweden, and cerebrospinal fluid examination is made if SAH is suspected and imaging is negative. However, information about which SAHs were nonaneurysmal was not available in the registers. Because 10% to 15% of SAHs usually are nonaneurysmal, 34 our results do not represent only aneurysmal SAH. Third, we lack information on diet. Several dietary factors have shown associations with the risk of SAH, 6 and whether dietary factors could contribute to the relationship between reduced lung function and SAH is unclear. Fourth, the protocol for the health examination was developed before the now commonly used guidelines for standardization of spirometry were published, 35 and the lung function tests did not meet current standards. However, if anything, poor precision of the lung function tests should have reduced the strengths of the associations between lung volumes and SAH incidence. The validity of the lung function variables is also supported by previous studies from the present cohort, showing that FEV 1 and FVC are strongly linked to cardiovascular and pulmonary diseases. 18,19,28 Finally, the subjects were followed over a mean time of 26 years, and we do not know to what extent the risk factors changed during follow-up. Future research will be required to assess the rate of lung function decline and risk of SAH. However, changed risk factor exposure would, if anything, dilute the relationships between baseline lung function and SAH. Conclusions Baseline lung function, expressed as low FEV 1 or FEV 1 / FVC, is a risk factor for SAH, independently of smoking. The explanation for this relationship remains to be elucidated. Sources of Funding This study was supported by the Swedish Heart and Lung foundation ( ), the Swedish Stroke Foundation, Lundström s Foundation, the Swedish Research Council ( ), and funds from Lund University and Skåne University Hospital. Disclosures Prof Gunnar Engström is employed as senior epidemiologist by AstraZeneca R&D. References 1. Feigin VL, Lawes CM, Bennett DA, Barker-Collo SL, Parag V. Worldwide stroke incidence and early case fatality reported in 56 population-based studies: a systematic review. Lancet Neurol. 2009;8: Johnston SC, Selvin S, Gress DR. The burden, trends, and demographics of mortality from subarachnoid hemorrhage. Neurology. 1998;50: de Rooij NK, Linn FH, van der Plas JA, Algra A, Rinkel GJ. Incidence of subarachnoid haemorrhage: a systematic review with emphasis on region, age, gender and time trends. J Neurol Neurosurg Psychiatry. 2007;78: Bor AS, Rinkel GJ, Adami J, Koffijberg H, Ekbom A, Buskens E, et al. Risk of subarachnoid haemorrhage according to number of affected relatives: a population based case-control study. Brain. 2008;131: Feigin VL, Rinkel GJ, Lawes CM, Algra A, Bennett DA, van Gijn J, et al. Risk factors for subarachnoid hemorrhage: an updated systematic review of epidemiological studies. Stroke. 2005;36: Shiue I, Arima H, Hankey GJ, Anderson CS. Dietary intake of key nutrients and subarachnoid hemorrhage: a population-based case control study in Australasia. Cerebrovasc Dis. 2011;31: Anderson C, Ni Mhurchu C, Scott D, Bennett D, Jamrozik K, Hankey G. Triggers of subarachnoid hemorrhage: role of physical exertion, smoking, and alcohol in the Australasian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS). Stroke. 2003;34: Hu G, Sarti C, Jousilahti P, Silventoinen K, Barengo NC, Tuomilehto J. Leisure time, occupational, and commuting physical activity and the risk of stroke. Stroke. 2005;36: Sin DD, Wu L, Man SF. The relationship between reduced lung function and cardiovascular mortality: a population-based study and a systematic review of the literature. Chest. 2005;127: Hole DJ, Watt GC, Davey-Smith G, Hart CL, Gillis CR, Hawthorne VM. Impaired lung function and mortality risk in men and women: findings from the Renfrew and Paisley prospective population study. BMJ. 1996; 313: ; discussion Wannamethee SG, Shaper AG, Ebrahim S. Respiratory function and risk of stroke. Stroke. 1995;26: Engström G, Hedblad B, Valind S, Janzon L. Increased incidence of myocardial infarction and stroke in hypertensive men with reduced lung function. J Hypertens. 2001;19: Truelsen T, Prescott E, Lange P, Schnohr P, Boysen G. Lung function and risk of fatal and non-fatal stroke. The Copenhagen City Heart Study. Int J Epidemiol. 2001;30: Agnarsson U, Thorgeirsson G, Sigvaldason H, Sigfusson N. Effects of leisure-time physical activity and ventilatory function on risk for stroke in men: the Reykjavik Study. Ann Intern Med. 1999;130: Hozawa A, Billings JL, Shahar E, Ohira T, Rosamond WD, Folsom AR. Lung function and ischemic stroke incidence: the Atherosclerosis Risk In Communities Study. Chest. 2006;130: Berglund G, Nilsson P, Eriksson KF, Nilsson JA, Hedblad B, Kristenson H, et al. Long-term outcome of the Malmo Preventive Project: mortality and cardiovascular morbidity. J Intern Med. 2000;247: Kristenson H, Trell E. Indicators of alcohol consumption: Comparisons between a questionnaire (MM-MAST), interviews and serum gammaglutamyl transferase (GGT) in a health survey of middle-aged males. Br J Addict. 1982;77: Ekberg-Aronsson M, Pehrsson K, Nilsson JA, Nilsson PM, Löfdahl CG. Mortality in gold stages of COPD and its dependence on symptoms of chronic bronchitis. Respir Res. 2005;6: Engström G, Melander O, Hedblad B. Population-based study of lung function and incidence of heart failure hospitalisations. Thorax. 2010;65: Jerntorp P, Berglund G. Stroke registry in Malmo, Sweden. Stroke. 1992;23:

6 Söderholm et al Lung Function and Incidence of SAH Schanen JG, Iribarren C, Shahar E, Punjabi NM, Rich SS, Sorlie PD, et al. Asthma and incident cardiovascular disease: the Atherosclerosis Risk In Communities Study. Thorax. 2005;60: Lind P, Engström G, Stavenow L, Janzon L, Lindgärde F, Hedblad B. Risk of myocardial infarction and stroke in smokers is related to plasma levels of inflammation-sensitive proteins. Arterioscler Thromb Vasc Biol. 2004;24: Schnabel E, Nowak D, Brasche S, Wichmann HE, Heinrich J. Association between lung function, hypertension and blood pressure medication. Respir Med. 2011;105: Engström G, Wollmer P, Valind S, Hedblad B, Janzon L. Blood pressure increase between 55 and 68 years of age is inversely related to lung function: longitudinal results from the cohort study men born in J Hypertens. 2001;19: Lindekleiv H, Sandvei MS, Njolstad I, Lochen ML, Romundstad PR, Vatten L, et al. Sex differences in risk factors for aneurysmal subarachnoid hemorrhage: a cohort study. Neurology. 2011;76: Frösen J, Piippo A, Paetau A, Kangasniemi M, Niemelä M, Hernesniemi J, et al. Remodeling of saccular cerebral artery aneurysm wall is associated with rupture: histological analysis of 24 unruptured and 42 ruptured cases. Stroke. 2004;35: Kataoka K, Taneda M, Asai T, Kinoshita A, Ito M, Kuroda R. Structural fragility and inflammatory response of ruptured cerebral aneurysms. A comparative study between ruptured and unruptured cerebral aneurysms. Stroke. 1999;30: Engström G, Lind P, Hedblad B, Wollmer P, Stavenow L, Janzon L, et al. Lung function and cardiovascular risk: relationship with inflammationsensitive plasma proteins. Circulation. 2002;106: James AL, Knuiman MW, Divitini ML, Musk AW, Ryan G, Bartholomew HC. Associations between white blood cell count, lung function, respiratory illness and mortality: the Busselton Health Study. Eur Respir J. 1999;13: Yoshida T, Tuder RM. Pathobiology of cigarette smoke-induced chronic obstructive pulmonary disease. Physiol Rev. 2007;87: Ruigrok YM, Rinkel GJ, Wijmenga C. Genetics of intracranial aneurysms. Lancet Neurol. 2005;4: Ruigrok YM, Rinkel GJ, van t Slot R, Wolfs M, Tang S, Wijmenga C. Evidence in favor of the contribution of genes involved in the maintenance of the extracellular matrix of the arterial wall to the development of intracranial aneurysms. Hum Mol Genet. 2006;15: Ruigrok YM, Rinkel GJ. Genetics of intracranial aneurysms. Stroke. 2008;39: Ronkainen A, Hernesniemi J. Subarachnoid haemorrhage of unknown aetiology. Acta Neurochir (Wien). 1992;119: American Thoracic Society. Snowbird workshop on standardization of spirometry. Am Rev Respir Dis. 1979;119: