Table e-1. Included studies on combined oral contraceptives (COC), hormone replacement therapy (HRT) and reproductive factors and the risk of SAH.

Study* / Period / Study type / Setting / Study Population / Age (years) / SAH criteria / Case-finding methods / Assessment of exposure / Female risk factors / Matching / adjustment variables / NOS (points)
8 / 1982 – 1983 / Case-control (P) / Auckland region, New Zealand / 115 SAH cases and 1586 controls / 35-64 / CH / ABDEHW / IPI / COC / - / - / 3
9 / 1986 – 1988 / Case-control (P) / General Practices throughout England and Wales / 296 SAH cases and 545 controls / 16-39 / CH / ACD / IPI / COC / au / bcl / 7
10 / 1987 – 1989 / Case-control (P) / King Country, Washington, USA / 104 SAH cases and 206 controls / 18+
(mean 57) / NSS / DM / IPI / COC, HRT, menopausal status / ab / aq / 3
11 / 1968 – 1990 / Nested case- control (P) / General Practices throughout the UK / 73 SAH cases and 759 controls / 21-70 / ICD / G / Data extraction from medical records. / COC / a / cd / 6
12 / 1989 – 1993 / Case-control (H) / 21 centers worldwide / 608 SAH cases and 2910 controls / 20-44 / CH / H / IPI / COC / a / ch / 5
17 / 1976 – 1992 / Retrospective cohort / Nurses’ Health Study, USA / 155 SAH cases in 662891 person-years / 30-50 at baseline / NSS / FMN / IPI at baseline and follow-up. / HRT / - / acehjlt / 8
13 / 1991 – 1995 / Case-control (P) / King, Pierce and Snohomish counties, Washington, USA / 69 SAH cases and 485 controls / 18-44 / CH / ME / IPI / COC / a / acfghi / 8
14 / 1990 – 1992 / Case-control (H) / Denmark: National Patient Register and National Person Register / 160 SAH cases and 3171 controls / 45-64 / ICD / I / Postal questionnaire. / COC, HRT / a / achlmr / 7
19 / 1994 – 1995 / Case-control (H) / WonjuChristianHospital, South-Korea / 95 SAH cases and 267 controls / 21-86 / CH / H / IPI / Age menarche, age menopause / - / ab / 4
15 / 1995 – 1998 / Case-control (P) / Adelaide, Hobart, Perth and Auckland, Australia / 268 SAH cases and 286 controls / 15+
(mean 57) / CH / BDEHRW / IPI / COC, HRT, menopausal status, age menopause, age menarche, parity, age first pregnancy / abv / acdfhjpq / 7
22 / 1992 – 1997 / Case-control (H) / 2 hospitals in Nagoya, Aichi Prefecture, Japan / 124 SAH cases and 248 controls / 30-79 / CH / H / IPI / Menopausal status, age menopause, age menarche, parity, age first pregnancy, age first birth / ab / aceh / 5
16 / 1992 - 2002 / RCT / Women’s Health Initiative, 40 centers throughout USA / 10 SAH cases in a cohort of 16608 participants / 50-79 / CH / HM / Treatment allocation at baseline. Annual telephone interview. / HRT / - / 9
20 / 1973 – 1997 / Case-control (P) / Sweden: Inpatient Register, Birth Register, and Cause of Death register / 887 SAH cases and 8512 controls / 18-38 / ICD / IN / Retrieved from Swedish Birth Register. / Parity / a / ast / 5
21 / 1987 – 2006 / Case-crossover / Area of University Medical Centre Utrecht, The Netherlands / 244 SAH cases / 18-42 / CH / H / Review of case records. / Pregnancy / - / N.A.¥
23 / 2002 – 2004 / Case-control (H) / 33 hospitals in 7 regions of South Korea / 252 SAH cases and 504 controls / 30-84 / CH / H / IPI / COC, HRT, menopausal status, parity / a / acfhklmno / 5
18 / 1994 – 1997 / Prospective cohort (P) / HUNT and Trombo
Study in Norway / 120 SAH cases in 1002148 person-years / 20+
(mean 50) / CH / HN / Self-reported questionnaire / HRT, menopausal status, age menarche, parity / - / b / 7

* Reference (s).

¥ Quality of the case crossover study could not be assessed, because quality assessments are not available for case crossover studies, but only for cohort and case-control studies.

RCT=randomized controlled trial. (P)=population based. (H)=hospital based. IPI=in-person interview. NA=not applicable.

ICD=International classification of diseases. CH=Clinical history confirmed by lumbar puncture, xanthochromia of the cerebrospinal fluid (CSF), CT or MRI scan, angiography, autopsy, or surgery. NSS=National Survey of Stroke. NOS=Newcastle-Ottawa Scale.

A=autopsy report, B=coroner, C=clinical notes, D=discharge letters, E=death certificates, F=family, G=GP, H=hospital admission/records, I=Inpatient Register, M=medical records, N=National Cause of Death Register, R=radiology list, W=ward referral.

a) age, b) sex, c) smoking, d) social class, e) educational level, f) alcohol intake, g) race, h) hypertension, i) BMI, j) cholesterol level, k) history of cerebrovascular disease, l) former use of OCP, m) use of HRT, n) menopausal status, o) age at menarche, p) center, q) respondent type, r) occupation, s) index date, t) length of follow up, u) marital status, v) city of residence.