Subclinical Atrial Fibrillation and the Risk of Stroke

Editor’s Note: This article is based on Healey JS, Connolly SJ, Gold MR, et al. Subclinical atrial fibrillation and risk of stroke. N Engl J Med. 2012; 366:120-9. 


Almost 15% of ischemic strokes are attributable to previously documented atrial fibrillation.(1,2) In turn, given that in many patients atrial fibrillation can be asymptomatic, there may be a high rate of patients with otherwise undiagnosed atrial fibrillation.(3,4) Studies regarding the clinical implication of subclinical atrial fibrillation are particularly important in the context of stroke prevention, especially given that almost 25% of ischemic strokes never have a cause identified.(1,5,6) Thus, identifying whether subclinical atrial fibrillation carries similarly increased stroke risk to symptomatic atrial fibrillation may be important to understanding how to manage patients with incidentally discovered atrial fibrillation. Patients with implantable devices, including pacemakers and defibrillators, provide a unique cohort in whom to identify subclinical atrial fibrillation given that when an atrial lead is in place, the device may be programmed to identify atrial tachyarrhythmias based on the occurrence of atrial rates above a pre-specified cutoff.

Furthermore, some studies have suggested that dual chamber devices (ie, devices with both an atrial and ventricular lead) may be used to suppress atrial arrhythmias via atrial overdrive pacing, during which the device continuously paces the atrium at a rate slightly higher than the patient’s sinus rate.(7,8) Given that patients with implantable devices already have some evidence or risk of conduction system disease, it is possible that use of algorithms to attenuate atrial arrhythmia risk in this population may be of use.

ASSERT (Asymptomatic Atrial Fibrillation and Stroke Evaluation in Pacemaker Patients and the Atrial Fibrillation Reduction Atrial Pacing Trial, NCT00256152) was designed to address the principal study objective of whether subclinical atrial fibrillation as incidentally diagnosed by implanted pacemakers or defibrillators was associated with an increased risk of stroke or systemic embolism and secondarily whether atrial overdrive pacing could prevent atrial fibrillation.(9,10)


ASSERT was a randomized trial which included patients 65 years of age or older with a history of hypertension requiring medical therapy and who had undergone implantation of either a dual chamber St. Jude pacemaker (for sinus node or atrioventricular node disease) or defibrillator (for any indication) in the preceding 8 weeks. Patients were excluded if there was any history of atrial fibrillation or atrial flutter lasting more than 5 minutes or required treatment with a vitamin K antagonist for any reason. All devices were programmed according to standardized settings with atrial tachycardia detection set to rates above 190 beats per minute and atrial fibrillation suppression algorithms turned off.(9)

Patients were then seen 3 months later and underwent device interrogation. Patients were classified according to whether a subclinical (ie, otherwise asymptomatic) atrial tachyarrhythmia had occurred. A subclinical atrial tachyarrhythmia consisted of all episodes of at least six minutes in length stored by the pacemaker or defibrillator. At the 3 month visit patients were also randomly assigned to having atrial overdrive pacing programmed either on or off.

Patients were then followed for a mean of 2.5 years. The primary outcome related to the principal study objective and included ischemic stroke or systemic embolism. Secondary outcomes included vascular death, myocardial infarction, stroke from any cause, and atrial tachyarrhythmias documented by surface electrocardiography (i.e., clinical atrial tachyarrhythmias). All electrograms suggesting presence of an atrial tachyarrhythmia and clinical events underwent blinded adjudication by expert committees. The primary outcome of the randomization portion of the study was symptomatic or asymptomatic atrial tachyarrhythmias lasting more than 6 minutes.


Between December 2004 and September 2009, 2451 patients with newly implanted pacemakers and 129 patients with newly implanted defibrillators were enrolled. The average patient age was 77+7 years with almost 58% male and an average body mass index of 27+5.

Over the first 3 months of follow-up, at least one atrial tachyarrhythmia was detected by the device in 261 (10.1%) patients while clinical atrial tachyarrhythmias (i.e., atrial arrhythmias documented on surface electrocardiogram) occurred in only 7. There was no significant difference in the percentage of patients with prior stroke between those with and without subclinical atrial tachyarrhythmias in the first 3 months. The prevalence of sinus nodal disease was higher and the resting heart rate was lower in patients with subclinical atrial tachyarrhythmias.

In the ensuing 3 month to 2.5 year follow-up period, 14 patients (0.5%) were lost to follow-up and 194 patients received a vitamin K antagonist, 47 of whom had a subclinical atrial tachyarrhythmia in the first 3 months (18.0%). Over the mean 2.5 year follow-up, subclinical atrial tachyarrhythmias occurred in 633 (24.5%) patients.

Patients with subclinical atrial tachyarrhythmias in the first 3 months had a significantly higher rate of ischemic stroke or systemic embolism (11/261, 4.2% vs. 40/2319, 1.7%) with a hazard ratio after adjusting for baseline stroke risk of 2.50 (95% CI 1.28-4.89). No association was noted between the presence of subclinical atrial tachyarrhythmias and any of the secondary outcomes. Over the mean 2.5 year follow-up period, patients with atrial tachyarrhythmias of at least 6 minutes in duration detected by their device tended to have an increased risk of ischemic stroke or systemic embolism (hazard ratio 1.76, 95% CI 0.99-3.11). This risk was also seen when accounting for atrial tachyarrhythmias lasting longer than 6 hours (hazard ratio 2.00, 95% CI 1.13-3.55) or 24 hours (hazard ratio 1.98, 95% CI 1.11-3.51). The relative risk of ischemic stroke or systemic embolism increased with CHADS2 score, reaching a rate of 3.78% per year in patients with a CHADS2 score greater than 2. The annual rate of stroke or systemic embolism was highest in patients with longer episodes of atrial tachyarrhythmias but did not necessarily correlate with the number of episodes of atrial tachyarrhythmias over the follow-up period.

The randomization portion of the study showed no significant effect of atrial overdrive pacing on the incidence of subclinical atrial tachyarrhythmias.


Subclinical atrial tachyarrhythmias occur in more than a third of patients with implantable devices and are associated with a significantly increased risk of ischemic stroke or systemic embolism. Atrial overdrive pacing did not prevent atrial fibrillation.


The ASSERT study suggested both that there is a high prevalence of subclinical atrial tachyarrhythmias amongst patients older than 65 years of age, with a history of hypertension and who had undergone pacemaker or defibrillator implantation and that there is a concomitant increased risk of ischemic stroke or systemic embolism. However, the incidence of subclinical atrial tachyarrhythmias did not necessarily portend the identification of atrial fibrillation in routine clinical practice in these patients, given that only 15.7% of patients with episodes noted via their device had clinical atrial fibrillation. These data support the hypothesis that even subclinical atrial tachyarrhythmias may carry increased risk of systemic thromboembolism, in particular ischemic stroke.

The population stroke risk associated with subclinical atrial tachyarrhythmias occurring in the first 3 months of the study was 13%, consistent with risk associated with clinical atrial fibrillation reported in Framingham.(11) Also, consistent with prior studies, the risk of stroke correlated with the CHADS2 score. Perhaps most importantly, the risk of the primary outcome was present with or without the presence of clinical atrial fibrillation as documented by surface electrocardiogram.

The noted relationship between ischemic stroke and subclinical atrial fibrillation is especially relevant to patients with stroke of unknown etiology, or cryptogenic stroke. Had the patient cohort included in this study not had devices, then atrial fibrillation may not have been identified via routine clinical monitoring and thus the strokes that occurred in this population may have otherwise been termed “cryptogenic,” though this cannot be stated absolutely since the management of these patients including work-up for stroke may have been different had they not already had an implantable device. The most important finding of the study is that it may not be the burden of atrial fibrillation but the mere presence or absence of it over any given period of time that determined stroke risk. While the risk of the primary outcome increased with increasing duration of atrial fibrillation, there was still an increased risk of stroke or systemic embolism seen in the general cohort. This speaks to the fact that documentation of the lack of atrial fibrillation on limited monitoring, such as transtelephonic monitoring, Holter monitoring, or routine surface electrocardiograms, may not be sufficient to state that a patient has no or rare atrial fibrillation and thus no associated increase in stroke risk.

The results of this study, however, may not be attributable in the opposite direction. Namely, while patients with otherwise asymptomatic atrial fibrillation incidentally discovered in the setting of having had a cardiac device implanted for other reasons are at increased risk of stroke or systemic embolism, the population risk of subclinical atrial fibrillation amongst patients with cryptogenic stroke cannot be gleaned from this study. Furthermore, the population included here were relatively high risk, with an average CHADS2 risk greater than 2 and of at least 1 in all patients. There was also heterogeneous treatment of these patients with antiplatelet or anticoagulant agents and further study is needed to determine if treatment with antiplatelet or anticoagulant agents attenuates risk of stroke similar to what has been seen in prior studies.(12) Finally, the cohort consisted of patients with devices in place for other reasons and, thus, the prevalence of subclinical atrial tachyarrhythmias may not be generalizable to those patients who do not have implantable devices.

There are also several limitations to the study. First, more than half of the patients were on aspirin at the start of the study and 18% of patients with subclinical atrial tachyarrhythmias were started on a vitamin K antagonist which may have attenuated risk of stroke, thus decreasing the incidence of the primary outcome. Furthermore, since all patients had implantable defibrillators or pacemakers, there was already some element of conduction system disease in all patients which may have modified risk of atrial tachyarrhythmias. For example, it is known that sinus node dysfunction is associated with an increased risk of atrial fibrillation.(13,14) Finally, there was no differentiation made in the study regarding atrial fibrillation versus regular atrial tachycardias or atrial flutters and, thus, it is possible that the risk of stroke may have been different if the population only consisted of those patients with atrial fibrillation as adjudicated by device electrograms.

The importance of the study lies in its demonstration of incidentally discovered atrial fibrillation by implantable devices as carrying increased risk of stroke and in its showing that this risk exists even in the setting of rare, brief (as short as 6 minutes) episodes of atrial fibrillation over a relatively long follow-up period (2.5 years). Further study is needed to determine if these same findings are generalizable to patients without implantable devices and what the actual risk is of atrial tachyarrhythmias amongst patients without pre-existing conduction disease.

Supported in part by the F. Harlan Batrus Research Fund and Murray and Susan Bloom Research Fund.


  1. Wolf PA, Dawber TR, Thomas HE Jr, Kannel WB. Epidemiologic assessment of chronic atrial fibrillation and risk of stroke: the Framingham Study. Neurology 1978; 28:973-7.
  2. Wolf PA, Abbot RD, Kannel WB. Atrial fibrillation: a major contributor to stroke in the elderly. Arch Intern Med 1987; 147:1561-4.
  3. Flaker GC, Belew K, Beckman K, et al. Asymptomatic atrial fibrillation: demographic features and prognostic information from the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study. Am Heart J 2005; 149: 657-63.
  4. Israel CW, Gronefeld G, Ehrlich JR, Li YG, Hohnloser H. Long-term risk of recurrent atrial fibrillation as documented by an implantable monitoring device: implications for optimal patient care. J Am Coll Cardiol 2004; 43:47-52.
  5. Tayal AH, Tian KM, Kelly M, et al. Atrial fibrillation detected by mobile cardiac outpatient telemetry in cryptogenic TIA or stroke. Neurology 2008; 71:1696- 701.
  6. Jabaudon D, Sztajzel J, Sievert K, Landis T, Sztajzel R. Usefulness of ambulatory 7-day ECG monitoring for the detection of atrial fibrillation and flutter after acute stroke and transient ischemic attack. Stroke 2004; 35:1647-51.
  7. Carlson MD, Ip J, Messenger J, et al. A new pacemaker algorithm for the treatment of atrial fibrillation: results of the Atrial Dynamic Overdrive Pacing Trial (ADOPT). J Am Coll Cardiol 2003; 42:627-33.
  8. Gold MMR, Adler S, Fauchier L, et al. Impact of atrial prevention pacing on atrial fibrillation burden: primary results of the Study of Atrial Fibrillation Reduction (SAFARI) trial. Heart Rhythm 2009; 6:295-301.
  9. Hohnloser SH, Capucci A, Fain E, et al. A Symptomatic atrial fibrillation and Stroke Evaluation in pacemaker patients and the atrial fibrillation Reduction atrial pacing Trial (ASSERT). Am Heart J 2006; 152:442-7.
  10. Healey JS, Connolly SJ, Gold MR, et al. Subclinical atrial fibrillation and risk of stroke. N Engl J Med 2012; 366:120-9.
  11. Wolf PA, Abbot RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke 1991; 22:983-8.
  12. Lip GY, Tse HF, Lane DA. Atrial fibrillation. Lancet 2012; 379:648-61.
  13. Kerr CR, Connolly SJ, Abdollah H, et al. Canadian Trial of Physiological Pacing: effects of physiological pacing during long-term follow-up. Circulation 2004; 109: 357-62.
  14. Skanes AC, Krahn AD, Yee R, et al. Progression to chronic atrial fibrillation after pacing: the Canadian Trial of Physiologic Pacing. J Am Coll Cardiol 2001; 38:167-72.

Keywords: Aspirin, Atrial Fibrillation, Atrial Flutter, Defibrillators, Electrocardiography, Heart Rate, Hypertension, Myocardial Infarction, Stroke, Tachycardia, Thromboembolism, Vitamin K

< Back to Listings