Do Women With Atrial Fibrillation Experience More Severe Strokes?
Results From the Austrian Stroke Unit Registry
Background and Purpose—Ischemic strokes associated with atrial fibrillation (AF) are more severe than those of other cause. We aim to study potential sex effects in this context.
Methods—In this cross-sectional study, 74 425 adults with acute ischemic stroke from the Austrian Stroke Unit Registry were included between March 2003 and January 2016. In 63 563 patients, data on the National Institutes of Health Stroke Scale on admission to the stroke unit, presence of AF, vascular risk factors, and comorbidities were complete. Analysis was done by a multivariate regression model.
Results—Stroke severity in general increased with age. AF-related strokes were more severe than strokes of other causes. Sex-related differences in stroke severity were only seen in stroke patients with AF. Median (Q25,75) National Institutes of Health Stroke Scale score points were 9 (4,17) in women and 6 (3,13) in men (P<0.001). The interaction between AF and sex on stroke severity was independent of age, previous functional status, vascular risk factors, and vascular comorbidities and remained significant in various subgroups.
Conclusions—Women with AF do not only have an increased risk of stroke when compared with men but also experience more severe strokes.
Ischemic stroke associated with atrial fibrillation (AF) is more severe, and, therefore, subsequently outcome is generally worse than in ischemic stroke because of other cause.1–3 Even though the association of stroke severity and AF seems to be independent of age,2 little is known whether this effect is modified by sex, vascular risk factors, or vascular comorbidities.
Since 2003, a growing network of Austrian acute stroke units is collecting data on standard characteristics and acute management of all patients with stroke admitted to one of the currently 38 participating centers. Experienced stroke neurologists are performing data collection and rating using standardized definitions for variables and scores. Evaluations were done at the time of admission to and discharge from the stroke unit and supplemented by a 3-month follow-up in person or by telephone and are immediately entered into a web-based database. Details on this registry, including definitions of clinical variables used, have previously been described.4,5 The diagnosis of AF (composite of paroxysmal or persistent AF) was made by treating physicians using patient’s self-report, documents of the healthcare system, or de novo diagnosis during stroke unit stay. Experienced stroke neurologists assessed stroke severity on admission to the stroke unit by the National Institutes of Health Stroke Scale and disability before the stroke leading to the stroke unit admission by the modified Rankin Scale. Patients classified as transient ischemic attacks by the treating physician were excluded from the analysis.
The clinical syndrome on presentation to the stroke unit was classified according to the Oxford Stroke Classification.6
The Austrian Stroke Unit Registry is part of a governmental quality assessment program for stroke care in Austria financed by the Federal Ministry of Health. It is based on the federal law promoting quality in health care (Gesundheitsqualitätsgesetz). Participation of hospitals/stroke units is mandatory. Anonymized data are centrally administered by Gesundheit Österreich GmbH, and scientific analyses are approved and supervised by an academic review board.
The relationship between stroke severity and several explanatory variables and their first-order interactions (age, sex, hypertension, diabetes mellitus, myocardial infarction, hypercholesterolemia, AF, smoking, and prestroke modified Rankin Scale) was modeled by a multivariate regression model. Additional information can be found in http://stroke.ahajournals.org.
A total of 74 425 adults with acute ischemic stroke were treated at Austrian Stroke Units between March 2003 and January 2016. In 8361 patients, information on stroke severity on admission, prestroke disability, and vascular risk factors was incomplete, and in 2501 patients, information on AF was unknown or missing, which left 63 563 patients with complete data set for the current analysis. The characteristics of the study population can be found in Table I in the online-only Data Supplement. No relevant differences were seen in patients with and without complete data set (data not shown).
AF was documented in 18 962 patients (29.8%). Characteristics of patients with and without AF can be found in Table I in the online-only Data Supplement. Among patients with AF, we detected more severe strokes in females than males (median [Q25,Q75], National Institutes of Health Stroke Scale score=9 [4,17] versus 6 [3,13]; P<0.001 [Table II in the online-only Data Supplement]). The association between AF and stroke severity was modified by sex. Figure (A) shows the association between age, sex, presence of AF, and stroke severity: stroke severity increased with age independently whether the stroke was related to AF or not. In general AF-related strokes were more severe than strokes of other causes. We detected differences in stroke severity between men and women only in AF-related strokes. In multivariate modeling, this interaction between AF and sex was independent of age. It remained significant when the population was limited to those without previous stroke or disability defined by a prestroke modified Rankin Scale score of 0 (Figure [B]).
In line with differences in stroke severity, Total Anterior Circulation Syndrome was significantly more frequent in women compared with men (27.1% versus 18.3%; P<0.001). This difference was detectable over all age groups (Table II in the online-only Data Supplement).
Stroke severity is one of the most important predictors of stroke outcome in general.7 To the best of our knowledge, we are the first to describe that sex-related differences in initial stroke severity are only present in patients with AF. Thus, female sex is not only a risk factor for the development of ischemic stroke in AF8 but also for a more severe stroke compared with males.
A previous in-depth analysis of our stroke unit population demonstrated no differences in quality of care in the acute stroke setting between men and women,5 excluding a sex-specific differences in stroke management as potential cause of the observed effect. Furthermore, other potential contributors, such as wake up stroke prevalence and differences in arrival times, could not explain the observed difference (data not shown).
From literature as well as our own data set several potential hypotheses to explain the observed effect can be deduced: first, it has been shown before that patients with AF under sufficient anticoagulation (international normalized ratio >2 in patients treated with warfarin) experience less severe strokes, and women with AF are less likely to receive oral anticoagulation in the primary8,9 and secondary10 prevention settings. The same is true in Austria.5 Yet absolute differences in anticoagulation use between sexes are low.8–10 Therefore, it seems unlikely that the observed sex effect on initial stroke severity in patients with AF is entirely accounted by differences of anticoagulation use between sexes.
Second, biological differences may contribute to sex differences in initial stroke severity in patients with AF. The higher prevalence of Total Anterior Circulation Syndrome in women with AF indicates more proximal vessel occlusions. Possible mechanisms include the fact that intracranial and extracranial vessels are significantly smaller in diameter in women11 and therefore cardiac emboli of the same size can lead to a more proximal vessel occlusion in woman compared with men.
Third, there may be an effect of sex hormones: estrogen has been shown to reduce fibrinolytic activity and to activate the coagulation system.12 This may lead to larger or denser thrombi. However, effects of oestrogen on the coagulation system or influence of postmenopausal hormonal dysbalance on cardiac remodeling13 are unlikely be the major contributor to the sex difference in AF-related stroke severity because the effect was consistent in all age groups.
Strengths of our study are the large sample size, high data quality, and consistency of findings in subgroups and sensitivity analyses. When interpreting the results, one has to keep in mind that the Austrian Stroke Unit Registry—as a governmental quality management tool—does not include information on all potential confounders, such as previous medication or concurrent cognitive deficits. Furthermore, our finding needs to be confirmed in an independent population.
In summary, we could show that the previously described association of female sex and stroke severity is only significant for AF-related stroke and does not extend to men and women without AF. Because of the fact that the magnitude of the presenting neurological deficit is the most important predictor for long-term functional disability after stroke, sex differences in patients with AF needs further elucidation.
Women with AF have not only an increased risk of stroke compared with men8 but also experience more severe stroke. Our data further support to strictly adhere to current guidelines. Underuse of oral anticoagulation must be avoided, especially in women.
↵* Part of the Master Thesis (Medical University of Vienna).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.015900/-/DC1.
- Received August 25, 2016.
- Revision received December 2, 2016.
- Accepted December 9, 2016.
- © 2017 American Heart Association, Inc.
- Steger C,
- Pratter A,
- Martinek-Bregel M,
- Avanzini M,
- Valentin A,
- Slany J,
- et al
- Kimura K,
- Minematsu K,
- Yamaguchi T
- Knoflach M,
- Matosevic B,
- Rücker M,
- Furtner M,
- Mair A,
- Wille G,
- et al
- Gattringer T,
- Ferrari J,
- Knoflach M,
- Seyfang L,
- Horner S,
- Niederkorn K,
- et al
- Bamford J,
- Sandercock P,
- Dennis M,
- Burn J,
- Warlow C
- Friberg L,
- Benson L,
- Rosenqvist M,
- Lip GY
- Friberg L,
- Rosenqvist M,
- Lindgren A,
- Terént A,
- Norrving B,
- Asplund K
- Zhao Z,
- Wang H,
- Jessup JA,
- Lindsey SH,
- Chappell MC,
- Groban L