Haemorrhagic stroke, overall stroke risk, and combined oral contraceptives: results of an international, multicentre, case-control study. WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception.

BACKGROUND

The risk of haemorrhagic stroke associated with use of oral contraceptives (OCs) is less well-established than that for ischaemic stroke. We assessed the risk of haemorrhagic stroke associated with current use of modern OCs as now used throughout the world.

METHODS

In this WHO collaborative, case-control study, we assessed the association between risk of haemorrhagic stroke and use of combined OCs in 1068 cases, aged 20-44 years, and 2910 age-matched controls. We also assessed risks for all strokes combined (haemorrhagic, ischaemic, and unclassified) based on 2198 cases and 6086 controls.

FINDINGS

Overall, current use of combined OCs was associated with slightly increased risk of haemorrhagic stroke; the increase was significant in the developing countries (odds ratio 1.76 [95% CI 1.34-2.30]) but not in Europe (1.38 [0.84-2.25]). Use of OCs in women younger than 35 years did not affect risk of haemorrhagic stroke in either group of countries, whereas in women aged older than 35 years, odds ratios were greater than 2. Women who were current users of OCs and had a history of hypertension (detected before current episode of OC use, but not during pregnancy) had a substantially increased risk (ten-fold to 15-fold) of haemorrhagic stroke compared with women who did not use OCs and had no history of hypertension. Odds ratios among current OC users who were also current cigarette smokers were greater than 3. In both groups of countries, past use of OCs, dose of oestrogen, and dose and type of progestagen had no effect on risk, and risks were similar for subarachnoid and intracerebral haemorrhage. The odds ratios for any type of stroke associated with current use of low-dose (< 50 micrograms oestrogen) and higher-dose OCs were 1.41 (0.90-2.20) and 2.71 (1.70-4.32), respectively, in Europe and 1.86 (1.49-2.33) and 1.92 (1.48-2.50) in the developing countries. From these data we estimated that about 13% and 8% of all strokes in women aged 20-44 in Europe and the developing countries, respectively, are attributable to the use of OCs.

INTERPRETATION

The risk of haemorrhagic stroke attributable to OC use is not increased in younger women and is only slightly increased in older women. The estimated excess risk of all stroke types associated with use of low-oestrogen and higher-oestrogen dose OCs in Europe was about two and eight, respectively, per 100 000 woman-years of OC use. However, findings need to be considered in the context of other risks and benefits associated with OC use, as well as those associated with the use of other forms of contraception.