Smoking as a Crucial Independent Determinant of Stroke

Table 2 Smoking and the risk of subarachnoid haemorrhage: summary of studies

	No. SAH Cases (No. in cohort)	No. of controls	Smoking Status	RR/OR	95% CI	Adjustments†	Matching
Case-Control Studies
Gill et al. [12] British Midlands, UK, 1989	208	573	Nonsmoker	1.00		Age, race, social class, alcohol consumption and treatment for hypertension
			Current (Male)	4.52	2.4–8.4
			Current (Female)	2.52	1.4–4.5
Quereshi et al. [26] Maryland US, 2001	323	969	Never	1.0		Hypertension, diabetes, alcohol consumption.	Age, sex, and ethnicity.
			Current Smoker	5.2	3.6–7.5
			Previous Smoker	4.5	3.1–6.5
Isaksen et al. [16] Tromsø County, Norway, 2002	26	104	Never	1.00		Systolic and diastolic blood pressure, serum cholesterol, serum HDL, BMI, coffee consumption.	Age and sex.
			Former	2.13	1.04–4.39
			Current	4.55	1.08–19.30
Cohort Studies
Kawachi et al. [18] Nurses' Health Study, U.S. 1993	108 (117006 females)	Rest of Cohort	Never Smokers	1.00		Age, hypertension, diabetes, BMI, high cholesterol, oral contraceptive use, HRT, age of smoking commencement, and follow-up period.
			Past Smokers	2.26	1.16–4.42
			Current Smokers	4.85	2.90–8.11
			1–14/day	4.28	1.88–9.77
			15–24/day	4.02	1.90–8.54
			25–34/day	7.95	3.50–18.07
			≥ 35/day	10.22	4.03–25.94
Kurth et al. [19] U.S. Physicians' Health Study 2003	31 (22022 males)	Rest of Cohort	Never Smokers	1.00		Age, sex, and randomized treatment group
			Ex-Smokers	0.79	0.37–1.68
			Current (<20/day)	1.75	0.24–13.09
			Current (≥ 20/day	3.22	1.26–8.18

SAH, subarachnoid haemorrhage; RR, relative risk; OR, odds ratio; CI, 95% Confidence Interval; †adjustments, refers to adjustments made for either univariate or multivariate analysis. HDL, high density lipoprotein; BMI, body mass index; HRT, hormone replacement therapy.

ISSN: 1617-9625