Editorials

Female sex as a risk factor for stroke in atrial fibrillation

BMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e3789 (Published 31 May 2012) Cite this as: BMJ 2012;344:e3789
  1. Eva Prescott, professor,
  2. Rikke Sørensen, specialty registrar
  1. 1Department of Cardiology, Bispebjerg University Hospital, Bispebjerg Bakke 23, Copenhagen, Denmark
  1. epre0004{at}bbh.regionh.dk

Women are at higher risk, but in young women with no other risk factors the absolute risk is low

Stroke as a complication of atrial fibrillation is an important cause of morbidity and mortality worldwide. The incidence of atrial fibrillation is higher in men than in women in all age groups, but, as the linked paper by Friberg and colleagues confirms (doi:10.1136/bmj.e3522),1 women carry a higher risk of stroke than men, a statement that was previously questioned by the National Institute for Health and Clinical Excellence guidelines.2 Nevertheless, there is considerable evidence that women should receive special consideration in relation to preventing stroke in atrial fibrillation.

Until recently the CHADS2 score was used to classify the risk of stroke in patients with atrial fibrillation. Recently, however, the more detailed CHA2DS2-VASc score, which includes female sex as an independent risk factor (counting as 1 point), has been developed.3 The other risk factors included are congestive heart failure, hypertension, diabetes, vascular disease, age 65-74 years, which each count for 1 point, plus age 75 or more and previous stroke, which each count for 2 points. The CHA2DS2-VASc score is now recommended by the European Society of Cardiology for risk classification.4 In general the European guidelines recommend thromboembolic protection with oral anticoagulation in patients scoring 2 or more points on the CHA2DS2-VASc score, and consideration of oral anticoagulation or aspirin in patients with one “clinically relevant” risk factor. Thus, according to this recommendation all women should be considered for anticoagulation. The CHA2DS2-VASc score has been criticised on the grounds that the estimated incidence of stroke in the lowest categories is based on relatively few cases.

Friberg and colleagues evaluated the risk of stroke in men and women in a large registry study.1 This population included patients with atrial fibrillation who were not taking oral anticoagulants and the study showed a higher risk of stroke in women, with an adjusted hazard ratio of 1.18 (95% confidence interval 1.12 to 1.24) compared with men. This finding is in accordance with other large registry studies that have recently been published and provides further evidence that women have a higher risk of stroke than men.5 6 Friburg and colleagues also look at the question of whether female sex with no other risk factors is sufficient to merit oral anticoagulation in patients with atrial fibrillation. This is a clinically important question because many clinicians are reluctant to treat relatively young women merely on the basis of their sex. However, if younger women are truly at high risk of stroke, why should they not benefit from treatment?

Friberg and colleagues used high quality nationwide data on more than 100<thin>000 people with atrial fibrillation to answer this question. The authors report that among patients with atrial fibrillation alone (no vascular disease) aged below 65 years the risk of stroke was low: women had an annual incidence of 0.7% and men of 0.5%. A comparable Danish registry study found an annual risk of stroke and thromboembolism in a similarly defined group of women of 0.82 (0.68 to 1.00) during a 10 year follow-up, with female sex being by far the weakest of the risk factors contributing to risk in the group with a CHA2DS2-VASc score of 1.6

However, the data must also be looked at critically. In this study fewer women than men received oral anticoagulants, which is in contrast to surveys from Canada and across Europe.5 7 After patients taking oral anticoagulants were excluded, almost 70% of the rest were taking antiplatelet drugs at baseline. Data were not analysed according to competing risks, which may have affected the estimated sex ratio. Registry studies have the advantage of being able to include a large number of unselected real life patients. Their major drawback, however, is the unmeasured confounding inherent in the design, a drawback that is magnified in a large study. The major concern in the study by Friberg and colleagues is the selection of patients to be included and why they were given anticoagulants; clearly this was not a random choice as it would have been in a randomised trial. Whether this selection bias has a significant impact on the results and conclusions is speculative because the authors did not provide data on the patients who received oral anticoagulants during the follow-up period. The risk of stroke may have been underestimated, and additional analysis of the patients treated with oral anticoagulants who were excluded from the study might have elucidated this.

Despite these inherent weaknesses, this and other registry studies provide reassurance to clinicians. The registry data confirm overall that women are at higher risk of stroke than men, but when differences in age and risk factor profile are taken into account the excess risk is low. More importantly, the absolute risk in women below age 65 with no other risk factors is low and does not merit treatment with oral anticoagulants.

Notes

Cite this as: BMJ 2012;344:e3789

Footnotes

  • Research, doi:10.1136/bmj.e3522
  • Competing interests: Both authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.

References