Prof Juan Carlos Kaski, St George’s, University of London, UK

Adding to the current debate as to whether -like other cardiovascular risk factors- atrial fibrillation (AF) is associated with a larger risk of adverse cardiovascular events in women than in men, a recently published meta-analysis by Emdin et al (1) provided evidence for gender differences in clinical outcomes.

The large Emdin study, incorporating 30 studies with 4.37 million participants, 66,511 with AF, showed that after adjustment for confounders, AF was associated with a 12% greater risk of all-cause mortality in women compared with men. Not only mortality was higher in women with AF versus men with AF but also risk of stroke was significantly larger in women. Despite the known limitations of observational studies in general, this meta-analysis by Emdin and colleagues (1), with a large number of participants and a long term follow has substantial statistical power to give its findings clinical credibility. Its message is very important indeed and adds to the growing body of evidence showing that risk of adverse events from cardiovascular disease differs in women compared with men.

The Emdin study opens new areas for debate and for clinical and mechanistic research.  Clinicians should now debate whether current risk management strategies are truly appropriate for women, and whether stroke prediction algorithms and prevention strategies take appropriate care of gender differences, such as those revealed by this study (1).


Emdin CA, Wong CX, Hsiao AJ, et al. Atrial fibrillation as risk factor for cardiovascular disease and death in women compared with men: systematic review and meta-analysis of cohort studies BMJ 2016; DOI:10.1136/bmj.h7013.

Cardio Debate Expert Comments

This new meta-analysis confirms the many prior reports indicating that atrial fibrillation (AF) has a more adverse impact on women compared to men.  While prior reports were relatively small and so were unable to comprehensively adjust for comorbidity, this study clearly identifies a small but significantly higher mortality risk for women.

Traditional “sex-difference” explanations for the increased risk have been hypothesized. Hormonal fluctuations during the menstrual cycle that affect QT intervals are an often cited consideration when selecting antiarrhythmic drugs for premenopausal women, yet the vast majority of AF occurs in postmenopausal older women.  Women may have a higher incidence of AF because of the association with obesity, yet women are less likely to have viscerally obesity (the actual risk factor) compared to men. Women may have relatively more bleeding on anticoagulation, yet sex-specific dosing (body mass and renally adjusted) is likely relevant similar to heparin, yet not used.

The more likely “gender-difference” explanations should be explored in further investigation. Women are treated with statins less frequently than are men, possibly contributing to an increased incidence of AF in women. Women are less likely to see a cardiologist for specialty care, and gender-related reluctance among primary care physicians to use warfarin may be especially problematic in elderly women, who benefit most from it. Outcomes after catheter ablation for AF are similar between the sexes, yet women are referred later and less frequently.

It is essential that investigators pursue both “sex” and “gender” based pathways in order to optimize care for the 52% of the population that are women.

Atrial fibrillation as risk factor for cardiovascular disease and death in women compared with men: systematic review and meta-analysis of cohort studies
Emdin CA, Christopher X Wong CX, Allan J, Hsiao AJ et al BMJ 2016;352:h7013 | doi: 10.1136/bmj.h7013

Cardiovascular diseases including stroke and myocardial infarction are the most common causes of death in men and women in Europe, i.e. 40% for men and 49 % for women – versus 2 % for breast cancer (1). Recently, a EU project was completed -A Roadmap of Gender Medicine in Europe- and its firs report pointed out on the impact of sex differences in cardiovascular diseases and called for more action in this area (2). Recently published textbooks on gender cardiology and gender medicine have also pointed out that atrial fibrillation (AF) is more dangerous in women compared with men (3,4).

The Emdin meta-analysis of sex differences in AF (5) included studies published between January 1966 and March 2015 and with a minimum of 100 participants with and without atrial fibrillation that reported sex specific associations. Thirty studies, with 4 371 714 participants were identified. AF was associated with a higher risk of all-cause mortality in women, i.e. Hazard Ratio (HR) for women compared with men, 1.12. Moreover, the study reported a significantly higher risk of stroke, HR 1.99, cardiovascular mortality, HR 1.93, cardiac events, HR 1.55, and heart failure, HR 1.16.  All had acceptable confidence intervals. Results were also broadly consistent in sensitivity analyses.

Strengths of this meta-analysis are of course that the combined results have more power to detect differences than results in individual studies and that only scientifically acceptable studies, with adequate numbers of patients and adequate representation of each sex, were included in the analysis.

Among the limitations of the study, is the fact that the primary analysis focused on the ratio of relative risks, which might not reflect absolute differences in risk of death and cardiovascular disease associated with AF between sexes.  Also, many of the studies assessed had differences in design, duration of follow-up, outcome ascertainment, and population characteristics.

There are also several types of publication biases; one is that negative results tend not to be published but also, as the authors suggest, that there might be an over-publication of studies showing worse results in women.

Another issue that deserves discussion is what to adjust for to avoid the undesirable effects of confounding factors. If the natural timing of cardiovascular disease in women is that women develop the disease later in life compared to men, perhaps adjusting for age might not be the right thing to do, and a better way would be to subdivide the participants into “age classes/groups” and compare the sexes in each age class/group.  Moreover, when it comes to co-morbidities, which women more often tend to have, sometimes the results of interventions are more successful in the sickest groups. This might be true for AF as well, and then it might be wrong to adjust for co-morbidities.  A common limitation when it comes to assessing differences between men and women, is that too few women are included in the analysis because no pre-specified power-analysis has been performed. Too low upper age limit, difficulties in recruiting female patients may also contribute to the fact that many CVD studies are underpowered for women.

Emdin et al.  were unable to ascertain the underlying cause for the sex differences found between AF and increased risk of death and cardiovascular disease. However, one might speculate that women might be undertreated because of the physicians’ hesitation to prescribe anticoagulants in elderly women, as they fear that the patients would get more adverse side effects, especially major bleedings, which is not true, especially not for warfarin. Therefore many physicians treat these elderly women with aspirin instead, which they probably shouldn’t have at all. Another consideration may be that women´s adherence to oral anticoagulant medication might be worse than that of men.  Other explanations may relate to sex differences in etiology and/or pathophysiology including electrophysiology of AF (data show smaller hearts including smaller atrial appendices in women). Differences in pharmaco-kinetics and pharmaco-dynamics may also play a role.

Some risk factors for stroke are unique to women, others are more prevalent in women compared with men, and others may differentially increase the risk of stroke in women, thus women specific risk scores for stroke may need to be developed. The Emdin meta-analysis supports the development of such score, in accordance with the recommendations from The American Heart Association.

In conclusion AF is a stronger risk factor for cardiovascular disease and death in women compared with men, though further research would be needed to determine causality. Clinicians ought therefore to consider a more aggressive treatment of risk factors in women with AF than has been the case so far, because women with AF seem to be at higher proportional risk of death and cardiovascular disease than men.


  1. Townsend N et al Eur Heart J , 2015, 36, 2696-2705 Cardiovascular disease in Europe epidemiological update
  1. The EUGenMed* Cardiovascular Clinical Study Group: Regitz-Zagrosek V, Oertelt-Prigione S, Bossano-Prescott E, et al .Gender in cardio-vascular disease: impact on clinical manifestations, management and outcomes. Eur Heart J, 2016, Jan 1:37(1):24-34.DOI:10.1093/eurheart/ehv598
  1. Handbook of Clinical Gender Medicine, Editors: Schenck-Gustafsson, K. (Stockholm); DeCola, P.R.;Pfaff, D.W. (New York, N.Y.); Pisetsky, D.S. (Durham, N.C.), Karger 2012, ISBN 978–3–8055– 9929–0; e-ISBN 978–3–8055–9930–6
  1. Textbook of Gender Medicine, Editors: Oertel-Prigione S, Regitz-Zagrosek V, Springer 2012
  1. Emdin CA,1 Christopher X Wong CX,2 Allan J Hsiao,AJ et al, Atrial fibrillation as risk factor for cardiovascular disease and death in women compared with men: systematic review and meta-analysis of cohort studies, BMJ 2016;352:h7013 | doi: 10.1136/bmj.h701