Sex and Stroke Risk in Atrial fibrillation – More Work to be Done

“All men are created equal.”

(Thomas Jefferson, 1776)

However, women with atrial fibrillation (AF) differ from men with the disease (1). In clinical practice, about half of the patients with AF are women. In general, women are older when AF manifests, have a higher prevalence of heart failure with preserved ejection fraction, and experience greater non-cardiac comorbidities compared to men (2). Since stroke is the most-feared adverse event and its risk can be reduced significantly by oral anticoagulation (OAC), there has been a strong focus on potential sex differences in stroke risk. In particular, whether female sex, is an independent risk factor for stroke incidence has been discussed intensively.

A recent meta-analysis of cohort studies examining sex differences in AF outcomes included 13 stroke studies. The authors reported a significant doubling in stroke risk in women compared to men (ratio of relative risks: 1.99; 95% confidence interval: 1.46 to 2.71). However, the meta-analysis identified significant heterogeneity across studies (I²=73%; P<0.001), as well as indicators of potential publication bias (3). Hence the meta-analysis was not definitive regarding sex differences in stroke risk.

The study by Arnson et al. in this issue [reference] adds contemporary evidence from a large Israeli health care provider database. In almost 90,000 patients with incident non-valvular AF, the authors observed a significant interaction for stroke risk by sex and age group. Compared with men, women 75 years or older revealed a higher risk of stroke. However, Arnson et al. did not observe significant sex differences in stroke risk below the age of 75 years. In this respect, the Arnson et al. study is largely confirmatory (4, 5). In patients with CHA₂DS₂-VASc score of zero or one, the authors did not detect sex differences in stroke risk. Furthermore, the authors examined patients with and without each of the CHA₂DS₂-VASc risk factors separately. Although, the details of the subgroup analyses were not shown, reportedly no sex differences in stroke risk were evident.

The increased stroke risk in women aged 75 years or older demonstrated in the Arnson et al. study has been reported rather uniformly, even after adjustment for comorbidities and stroke risk factors (4, 6, 7). However, despite vigorous accounting for potential confounders, as performed in the current and prior studies, residual confounding cannot be ruled out. In addition, sex differences in myocardial and vascular structure and function, in systemic inflammation and procoagulatory pathways, accentuated in post-menopausal women, offer many sources for pathophysiological differences in stroke risk (8).

In general, more comorbidities accumulate with advancing age that may differ by sex. Aging itself differs by sex. Since age is one of the strongest predictors of AF and its complication stroke, age-related changes also may contribute a significant proportion of the sex-related stroke risk differences in AF.

Observational studies of patients with AF, such as the Arnson et al. study, cannot exclude systematic sex differences and disparities in treatment and management that may be related to stroke risk. In the past, women showed a lower proportion of adequate OAC therapy, although women benefit similarly or even more from OAC compared to men (9). The potential for treatment biases explaining the observation of sex differences in stroke risk is lower today, due to comparable anticoagulation rates by sex in more recent data (10). In the study by Arnson et al. 40.4% of patients were on OAC, which is alarmingly low compared to other contemporary cohorts, but the percent of individuals on OAC was similar in both sexes. In the recent large direct oral anticoagulants (DOAC) trials, in which OAC was evenly distributed due to randomization, women showed a numerically higher incidence of stroke. Again, women were older on average at enrolment compared to their male counterparts.

In terms of clinical implications, the Arnson et al. study and a large body of prior evidence reinforce current guidelines that female sex is only counted as a score point in the CHA₂DS₂-VASc risk assessment in the presence of other stroke risk factors. Female sex per se does not appear to be a risk factor at younger age or in the absence of other stroke risk factors. Therefore, sex alone should not necessarily be a discriminator in anticoagulation decisions. In addition, the present study lamentably is consistent with global reports that anticoagulation is underutilized (11), which leaves much room for outcome improvement in both sexes.

Several research questions remain regarding whether women and men with AF are equal. Do truly sex-specific pathophysiologic mechanisms for stroke risk exist, in particular in older women, or does confounding, in particular comorbidities, fully explain observed differences in stroke incidence? Similarly, it needs to be clarified, is the higher residual stroke risk on OAC (12) due to sex-specific pathophysiology, residual confounding, possible bias, or even disparities in patient management of women and men? Furthermore, the predominant role of age and aging in stroke risk needs to be dissected carefully from sex differences, in particular in older adults. And finally, how can we more effectively address patient, clinician, and system barriers to improve OAC use in women and men?

Future research will need to address sex and age differences and disparities to improve management and outcome of AF. Hence, the recent directives from the National Institutes of Health on addressing sex as a biological variable and enhancing inclusion of individuals across the lifespan are welcome and will improve the evidence base for sex differences in preclinical models and in older men and women with AF.

List of abbreviations:

AF

atrial fibrillation

DOAC

direct oral anticoagulants

OAC

oral anticoagulation

Biography