Background and purpose Subarachnoid haemorrhage (SAH) is the most devastating cerebrovascular disease. Cigarette smoking is one of the established risk factors for SAH, but the risk of SAH has not been properly elucidated in relation to smoking cessation.
Methods We performed a nationwide multicentre case control study involving 33 hospitals in Korea. A total of 426 SAH cases and 426 age and sex matched controls were enrolled. We obtained detailed information on lifestyle, medical history and, in particular, smoking habits from participants using structured questionnaires.
Results 148 SAH patients (37.4%) were current smokers compared with 103 (24.2%) controls, giving an adjusted OR of 2.84 (95% CI, 1.63 to 4.97) after controlling for possible confounders. Based on cumulative dose of smoking (pack years), the risk of SAH was found to increase in a dose–responsive fashion. Smoking cessation (≥5 years) caused a reduction in SAH to 59% (p<0.05). However, participants with a history of heavy smoking (≥20 cigarettes per day) had a 2.3 times increased risk of SAH compared with participants who had never smoked (p<0.05).
Conclusions We have demonstrated that cigarette smoking increases the risk of SAH, but smoking cessation decreases the risk in a time dependent manner, although this beneficial effect may be diminished in previous heavy smokers. To forestall tragic SAH events, our results call for more global and vigorous efforts for people to stop smoking.
Statistics from Altmetric.com
Subarachnoid haemorrhage (SAH) is the most devastating cerebrovascular disease, associated with substantial mortality and morbidity. The case death rate is about 50% in population based studies,1 and SAH is associated with 27% of all stroke related years of life lost before the age of 65 years.2 Interestingly, multi-ethnic studies reported that smoking increases the risk of SAH dose dependently.3 ,4
The beneficial effect of smoking cessation on cardiovascular diseases is well established. The excess risk of coronary heart disease is reduced by about 50% after 1 year of abstinence of smoking and continues to gradually decline so that after 5–10 years, the risk is similar to that of never smokers.5–8 Overall, among smokers with coronary heart disease, smoking cessation is associated with a 36% reduction in risk of all cause mortality.9 For ischaemic stroke, the excess risks among past smokers largely disappeared during the period from 2 to 5 years after cessation.10 ,11 However, only a few studies have focused on the relationship between smoking cessation and the development of SAH. In this context, it is necessary to assess the effect of smoking cessation on SAH more attentively, and we investigated the effect of cigarette smoking on SAH to confirm smoking as a risk factor for SAH and the impact of smoking cessation specifically in a nationwide multicentre case control study.
Subjects and methods
Our study was a nationwide multicentre matched case control study which was originally designed to investigate the risk factors for haemorrhagic stroke.12–14 The cases recruited for this study came from 33 hospitals distributed regionally, covering considerable areas of the Republic of Korea, from October 2002 to March 2004. This study was approved by the institutional review board of Seoul National University Hospital and all study participants provided written informed consent to participate (H-0912-030-303).
Selection of cases and controls
The diagnosis of SAH was based on clinical symptoms via either a brain image (CT, MRI) or evidence of xanthochromia on lumbar puncture. Eligible criteria for cases included age range 30–84 years, incident non-traumatic SAH and no history of haemorrhage prone brain lesion, as described elsewhere.12–14 A total of 940 patients with haemorrhagic stroke were included in the original studies,12 ,13 and among them, 442 SAH patients were initially enrolled in this study after excluding spontaneous intracerebral haemorrhage. Finally, after excluding 16 subjects in whom we could not obtain smoking information even from their family members, 426 SAH cases were investigated.
One control was matched to each case for sex and age at the same time when cases were enrolled. Controls were recruited from siblings, friends or neighbours of the controls, and excluded those with a history of stroke, dementia or other neurological diseases, as described previously.13
Each regional centre recruited and trained nurse interviewers, and a central steering committee held four training workshops from June 2002 to September 2002, before initiation of the study, for interviewers to improve their interview skills and revise case report forms. To increase the reliability and validity of the study, a pilot study was performed in 80 patient/control sets from all centres. After the pilot study, trained interviewers administered structured questionnaires to all participants which were evaluated for feasibility by an expert panel. Interviewers obtained information on basic characteristics, which included age, gender, height, weight, lifestyle factors (eg, alcohol consumption, dietary habits and work hours), medical history (eg, hypertension, diabetes, hyperlipidaemia and cardiovascular disease) and family history of stroke, as described elsewhere.12–14 Participants were asked about their smoking habits, including information on the amount of cigarettes (number of cigarettes per day), duration of smoking and length of time since stopping smoking (month/year). Smoking habits were categorised as never, past or current, and past smokers were defined as those who had decided to quit smoking and had stopped smoking for more than 1 week.
Differences between cases and controls were examined by McNemar's test for categorical variables or by the Student's t test for continuous variables. ORs and their 95% CIs for SAH associated with smoking status (never, past and current) and cumulative smoking dose (pack years) were estimated using conditional logistic regression analyses.
To evaluate the relationship between duration of smoking cessation and development of SAH, we calculated the ORs and their 95% CIs for SAH associated with quitting duration using restricted cubic spline. According to this analysis, stratified analyses were conducted using the duration of smoking cessation and the number of cigarettes per day in past smokers. We calculated the unadjusted and adjusted ORs and 95% CIs for SAH for each subgroup using a conditional logistic regression model. The adjusted possible confounders were: family history of stroke; history of hypertension, diabetes, hyperlipidaemia and cardiovascular diseases; body mass index; alcohol consumption; high salt intake; and work hours. All probability values were two tailed, and p values <0.05 were considered significant. Statistical analysis was performed using R 2.15.0 (http://www.r-project.org).
Among 426 SAH cases and 426 controls, there were 336 men and 516 women, and mean age was 50.7 years. Table 1 shows the characteristics of the cases and controls. The cases and controls were generally similar with regard to baseline characteristics, including: age; gender; body mass index; history of diabetes, hyperlipidaemia and cardiovascular disease; laborious work of 7 or more hours; and alcohol assumption. However, the number of current smokers among SAH patients was higher than among control subjects. The proportions with a family history of stroke or a history of hypertension were significantly greater in the SAH patients.
Effects of cigarette smoking as a risk factor
Table 2 shows the association between smoking and SAH in all subjects. A total of 148 SAH patients (34.7%) were current smokers compared with 103 (24.2%) controls, giving a crude OR of 2.79 (95% CI 1.72 to 4.51). The adjusted OR for SAH in current smokers was 2.84 (95% CI 1.63 to 4.97) compared with never smokers, after adjusting for potential confounders. According to the cumulative dose of smokers, the risk of SAH increased dose dependently (p for trend <0.01), but the risk of SAH in past smokers did not increase significantly compared with never smokers (table 2).
Effects of quitting duration and previous smoking amount
Compared with current smokers, a longer duration of smoking cessation in past smokers decreased the risk of SAH (OR per 1 year 0.80; 95% CI 0.64 to 0.99). After 5 years of smoking cessation, the lower value for 95% CI in the OR for SAH decreased below 1.0 in the restricted cubic spline model (figure 1). Compared with current smokers, past smokers whose quitting duration was 5 or more years showed a significant 59% reduction in the risk of SAH and did not have a higher risk of SAH compared with never smokers (table 3). However, past smokers who stopped smoking for less than 5 years had a 2.7 times increased risk of SAH. To identify the effect of previous smoking amount in past smokers on the risk of SAH, we dichotomised past smokers into light (<20 cigarettes per day) and heavy (≥20) smokers. Despite cessation of smoking, previously heavy smokers had an increased risk of SAH compared with never smokers.
In this nationwide matched case control study, we investigated the effect of cigarette smoking on SAH and more specifically the impact of smoking cessation. We found that the risk of SAH associated with cigarette smoking increased dose dependently and was reduced to the level of never smokers after 5 years of quitting smoking, and this beneficial effect of smoking cessation was decreased in heavy smokers.
Non-traumatic SAH is mostly caused by aneurysm rupture in cerebral blood vessels.15 Considering the mechanism of SAH associated with smoking, reversible steps may take place after smoking cessation. In terms of structural aspects, there is evidence for an association of smoking with aneurysm growth and the number of formations through a direct effect of smoking on vessel walls which may be stopped after smoking cessation.16 ,17 Moreover, in addition to structural changes, mechanisms linking cigarette smoking and SAH have also been described as short term physiological effects, such as enhanced systematic coagulability and increased blood pressure, which may be reversible after stopping smoking.18 ,19 Our result that the risk of SAH was decreased after smoking cessation time dependently could be explained by these reversible steps.
There are a lack of studies on the association between smoking cessation and risk of SAH. One case control study showed a persistent risk of SAH in past smokers but there was no investigation of its time dependent association.20 In contrast, a Nurses' Health Study showed that the risk of SAH disappeared after 15 years of smoking cessation, but the limitation of this study was that the subjects were all women, which might not reflect the general population.11 One small case control study showed that the risk of SAH among smokers disappeared 10 years after smoking cessation, but this study did not evaluate the risk within the 10 year period of smoking cessation due to the small sample size.21 In this context, our large nationwide case control study may provide information about the reduced risk of SAH after smoking cessation time dependently in the general population.
In addition, we found evidence that the effect of smoking cessation on reducing the risk of SAH was decreased in previous heavy smokers. Although the decreased risk of SAH in past smokers was time dependent, which may reflect reversible structural changes and physiological effects, our finding of the decreased effect in heavy smokers may be explained by long term irreversible effects of cigarette smoking on vessel walls.22–24 These two findings concerning smoking cessation in our study show that the impact of smoking cessation may be followed by reversible and irreversible structural changes resulting from smoking on the risk of SAH.
The present study had several limitations. First, recall bias could not be avoided in this case control study. However, we used organised questionnaires which were administered by well trained interviewers and, as indicated in a previous study, self-reports of smoking are accurate in most studies, and interviewer administered questionnaires are associated with higher sensitivity and specificity of self-reports in estimating real smoking habits.25 Second, because the control subjects were matched for age and gender, other unmatched factors such as hypertension, hyperlipidaemia, heart disease and obesity might have confounded a co-dependent smoking risk. In this context, we minimised this confounding effect by conditional logistic regression analysis to control these unmatched factors, but possible confounding effects could not be removed completely. In addition, despite efforts to reduce the confounding effects using a conditional logistic regression model, we could not completely exclude the fact that the protective effect of a longer period of smoking cessation on SAH may be influenced by unmatched potential confounders. Third, random population based case control studies may have greater validity than our hospital based study. To confirm our results, population based case control studies or prospective large cohort studies are needed.
The time dependent decreased risk of SAH since smoking cessation may indicate that smoking is a modifiable risk factor for SAH, and the dose dependent risk elevation in smokers is likely to reveal that smoking directly affects the development of SAH. In this context, vigorous efforts of government officials and health workers to promote smoking cessation should be performed to reduce the risk of SAH, a disastrous disease for patients and family members.
We are grateful to the Medical Research Collaborating Centre (MRCC) at Seoul National University Hospital for their support in the statistical analyses.
Funding This work was supported by grants from the Korean Health Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (A100331, A102065 and A111014). The funding organisation had no role in the design, conduct or analysis of this study, or in the preparation of this report.
Competing interests None.
Ethics approval Ethics approval was provided by Seoul National University Hospital.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.