Abstract
Background
An implantable loop recorder (ILR) assists in the diagnosis of unexplained syncope and atrial fibrillation (AF). Both become prevalent with age. Limited data exist describing the incidence of AF as the diagnostic rhythm underlying syncope in the elderly. This study aims to report the incidence of AF in older adults with ILRs for unexplained syncope and identify clinical characteristics associated with AF in this population.
Methods
Retrospective observational study on patients with unexplained syncope seen in syncope clinics from two Canadian centers. Participants were ≥65 years old, without a history of AF, and received an ILR for unexplained syncope. Data were collected from patient’s clinic charts. Arrhythmia diagnosis was based on ILR electrocardiogram reading during syncope (symptom–rhythm correlation). Fisher’s exact test and the Student’s t test were used to compare participants with and without AF.
Results
In our cohort of 222 patients, 124 were females and 98 were males. Mean age at implant was 80 ± 8 years. Arrhythmia was diagnosed in 98 patients (44.1%). Median time to diagnosis was 18 months. AF was diagnosed in 17 (7.7%) participants. There were fewer males in the AF group than the no AF group (11.8%, 46.8%, p = 0.01). Age, baseline EKG, and prevalence of hypertension, diabetes, stroke, or ischemic heart disease were not statistically different between patients with AF and without AF.
Conclusions
Atrial fibrillation was a common diagnostic rhythm in this cohort of adults, aged 65 and older, with ILRs for unexplained syncope. It was observed more frequently in females.
1. INTRODUCTION
Syncope and atrial fibrillation (AF) are common clinical entities (Brignole, 2018; Burkowitz, Merzenich, Grassme, & Brüggenjürgen, 2016; Chen, Shen, Mahoney, Jacobsen, & Rodeheffer, 2006; Krahn, Klein, Yee, & Skanes, 2001). Syncope is a transient loss of consciousness caused by decreased cerebral blood flow (Brignole, 2018). AF is an irregularly irregular heart rhythm and is a risk factor for stroke, cardiovascular diseases, renal disease, and death (Odutayo et al., 2016). Syncope, particularly unexplained syncope (undiagnosed after thorough clinical evaluation and diagnostic testing (Brignole, 2018)), and AF both become increasingly prevalent with age (Benjamin, 1994; Chen et al., 2006; Jansen et al., 2015; Roberto, Ronaldo, & Silvio, 2012). Thus, some elderly patients presenting with syncope may have concurrent AF.
Implantable loop recorders (ILRs) are small subcutaneous devices capable of long‐term electrocardiogram monitoring of up to 3 years. They have emerged as important diagnostic tools for unexplained syncope as they can detect infrequent and intermittent arrhythmias such as AF (Brignole et al., 2009; Burkowitz et al., 2016; Krahn et al., 2001).
Atrial fibrillation is an uncommon cause of syncope (Jansen et al., 2015). However, multiple large cohort studies have shown an association between atrial fibrillation and syncope among the general and elderly population (Jansen et al., 2015; Ruwald et al., 2013). Several proposed pathways provide a pathophysiologic link between AF and syncope (Benjamin, 1994; Brignole et al., 1993; Jansen et al., 2015; Leitch, Klein, & Tee, 1991). AF is associated with an abnormal neural response, including carotid sinus hypersensitivity, increased vagal tone at the onset of tachycardia, and increased rates of orthostatic hypotension, all of which can lead to syncope (Brignole et al., 1993; Leitch et al., 1991; Mcnicholas, Tobin, O'Callaghan, & Kenny, 2017; Wong et al., 2011). This can be exacerbated by polypharmacy, as commonly encountered in older people (Ruwald et al., 2013).
To date, only three clinical studies have shown a link between AF and syncope in older patients (Jansen et al., (2015); Mcintyre, Liebrecht, Daba, & Seifer, 2016; Ibrahim et al., 2017). As of yet, there has been no study dedicated to describing the rates of AF among older patients with ILRs for unexplained syncope. An ILR may establish a temporal relationship between AF and syncope. If AF is shown to be a common diagnostic rhythm in this population, then this may help promote further studies, particularly on the implications of ILR‐detected AF in stroke and anticoagulation for patients with unexplained syncope.
The primary aim of this study was to estimate the prevalence of AF as the diagnostic rhythm concurrent with syncope (symptom–rhythm correlation) among patients ≥65 years old who received ILRs for unexplained syncope. Our secondary aim was to identify characteristics that are associated with AF in this population.
2. METHODS
This was a two‐center, retrospective chart review of patients ≥65 years old from syncope clinics at two tertiary academic acute care facilities in Canada. Patients underwent ILR implantation for unexplained syncope between April 2005 and December 2016. All participants received comprehensive guideline‐directed assessments and were deemed to have unexplained syncope prior to implantation (Brignole, 2018; Brignole et al., 2009) Patients with a prior history of AF and who had other indications for ILR implantation (e.g., palpitations, cryptogenic stroke) were excluded. Participant demographics and baseline clinical characteristics (including hypertension, ischemic heart disease, stroke, and diabetes) were obtained from clinic charts. The diagnosis of syncope was made based on symptom–rhythm correlation, defined as the ILR rhythm reading at the time of syncope or symptoms of pre‐syncope (lightheadedness, dizziness, or feeling of fainting). All ILRs used (Medtronic™ Reveal LINQ, Plus, XT, DX, and St. Jude Medical™ DM2100) had manual patient activation capabilities and proprietary automatic AF detection algorithms. Symptom–rhythm diagnosis was made after a physician read the ILR ECG recordings and correlated these with the patient’s symptom diary, patient device triggering during symptoms, and/or self‐reported timing of symptoms at clinic visit. AF is defined as the primary arrhythmia diagnosis only when it is the rhythm recorded immediately before or during syncope or pre‐syncope. AF is further subclassified as with rapid ventricular response (heart rate >100 beats/min), with normal ventricular response (heart rate at or between 60 and 100 beats/min) or with conversion pauses (sinus pause of greater than or equal to 3 s following atrial fibrillation). Data were collected at baseline and at follow‐up appointments. We followed participants for up to 3 years or until the date of their ILR explant. We compared patients with and without AF using Fisher’s exact test for categorical variables and the Student’s t test for continuous variables.
3. RESULTS
A total of 222 patients (124 females, 98 males) were included, with an average age of 80 ± 8 years. A diagnosis of cardiac arrhythmia was made in 98 patients (44.1%), based on symptom–rhythm correlation (i.e., cardiac arrhythmia during pre‐syncope or syncope). The time from implantation to diagnosis of a cardiac arrhythmia based on symptom–rhythm correlation ranged from <1 month to over 3 years (median of 18 months). A diagnosis of AF through symptom–rhythm correlation was made in 17 (7.7%) patients. Atrial flutter was found in three (1.4%) patients (Table 1). Other arrhythmic diagnoses identified by symptom–rhythm correlation are identified in Table 1. When age was stratified, the incidence of AF increased with age (65–<75 years old: 1.4%, 75–<85 years old: 2.7%, ≥85 years old: 3.6%). AF was documented in three asymptomatic patients.
Table 1.
Symptom‐ILR rhythm diagnoses in unexplained syncope
| Diagnoses | Number of patients | % (of total) |
|---|---|---|
| Atrial fibrillation | 17 | 7.7 |
| With a rapid ventricular ratea | 10 | |
| With a normal heart rateb | 4 | |
| With a conversion pausec | 3 | |
| Atrial flutter | 3 | 1.4 |
| No events identified | 124 | 55.9 |
| Asystole | 8 | 3.6 |
| AVNRT | 3 | 1.4 |
| Type 2 AV block, Mobitz II | 3 | 1.4 |
| Junctional bradycardia | 2 | 0.9 |
| Other bradyarrhythmia | 2 | 0.9 |
| Ventricular tachycardia | 2 | 0.9 |
| Ectopic pacemaker | 1 | 0.5 |
| Sinus bradycardia | 1 | 0.5 |
| Sinus tachycardia | 1 | 0.5 |
| Supraventricular tachycardia | 1 | 0.5 |
| Type 2 AV block, Mobitz I | 1 | 0.5 |
| Total number of patients | 222 |
AV: atrioventricular; AVNRT: atrioventricular nodal re‐entry tachycardia; ILR: implantable loop recorder.
Rapid ventricular rate defined as heart rate of >100 beats/min.
Normal heart rate defined as 60–100 beats/min.
Conversion pause defined as a sinus pause for 3 s or greater following atrial fibrillation.
Of the participants with AF and symptom–rhythm correlation, 10 (58.8%) had rapid ventricular rates, while three (17%) had conversion pauses (Table 2). Eight were placed on anticoagulation, eight were started on rate‐controlling medications, and four had pacemaker implantation (Table 2).
Table 2.
Characteristics of individuals found to have atrial fibrillation using symptom‐ILR rhythm correlation
| Age | Baseline ECG | ILR activation | Rhythm | Treatment |
|---|---|---|---|---|
| 67 | RBBB | Patient | AF RVRa | BB, dabigatran |
| 70 | NSR, LVH | Patient | AF with normal HRb | N/A |
| 71 | 1° AVB | Automatic | AF with conversion pausec | PPM |
| 75 | NSR | Patient | AF RVR | BB, warfarin, PPM |
| 75 | 1° AVB | Automatic | AF RVR | Warfarin, CCB |
| 81 | NSR | Patient | AF RVR | CCB |
| 82 | LBBB | Patient | AF RVR | None |
| 82 | LBBB | Automatic | AF with conversion pause | PPM |
| 83 | Sinus bradycardia | Automatic | AF RVR | BB |
| 85 | NSR | Automatic | AF RVR | Digoxin, warfarin |
| 86 | 1° AVB | Patient | AF RVR | BB, warfarin |
| 87 | Sinus bradycardia | Patient | AF RVR | PPM |
| 87 | NSR | Patient | AF with normal HR | Warfarin |
| 87 | NSR | Patient | AF with conversion pause | BB, apixaban, PPM |
| 88 | 1° AVB | Patient | AF | N/A |
| 90 | NSR | Automatic | AF RVR | Warfarin, BB |
| 94 | NSR | Patient | AF | N/A |
1°: first degree; AF: atrial fibrillation; AVB: atrioventricular block; AVNRT: atrioventricular nodal re‐entry tachycardia; BB: beta blocker; CCB: calcium channel blocker; ILR: implantable loop recorder; LVH: left ventricular hypertrophy; N/A: not available; NSR: normal sinus rhythm; PPM: pacemaker.
RVR: rapid ventricular rate defined as a heart rate of >100 beats/min.
Normal HR: normal heart rate defined as a heart rate of 60–100 beats/min.
Conversion pause: defined as a sinus pause for 3 s or greater following atrial fibrillation.
There were significantly fewer males (11.8%) in the AF group when compared to the no AF group (46.8%; Table 3). Age and the prevalence of hypertension, diabetes, stroke, and ischemic heart disease were not statistically different between the AF and the no AF group (Table 3).
Table 3.
Baseline characteristics of syncope patients found to have atrial fibrillation compared to those in whom no atrial fibrillation was found
| Characteristics | AF (n = 17 participants) | No AF (n = 205 participants) | p‐value |
|---|---|---|---|
| Age (year ±SD) | 82 ± 8 | 80 ± 8 | 0.24* |
| Male | 2 | 96 | <0.01** |
| Hypertension | 14 | 153 | 0.57* |
| Diabetes | 2 | 46 | 0.54* |
| IHD | 6 | 70 | 1* |
| Stroke | 2 | 31 | 1* |
| Baseline EKG | |||
| NSR | 11 | 81 | 0.07* |
| LVH | 1 | 22 | 1* |
| Sinus bradycardia | 2 | 34 | 0.75* |
| 1° AVB | 4 | 57 | 1* |
| LBBB | 2 | 22 | 1* |
| RBBB | 1 | 36 | 0.32* |
1°: first degree; AF: atrial fibrillation; AVB: atrioventricular block; IHD: ischemic heart disease; NSR: normal sinus rhythm; LVH: left ventricular hypertrophy; LBBB: left bundle branch block; RBBB: right bundle branch block; SD: standard deviation.
Calculated using Student’s t test for continuous variables.
Calculated using Fisher’s exact test for categorical variables.
4. DISCUSSION
This retrospective, two‐center observational study found AF to be a common diagnostic rhythm (7.7%) for syncope among patients 65 years and older who were implanted with ILRs for unexplained syncope. AF was the third most common arrhythmia associated with syncope in this cohort, following sinus pauses and complete heart block. Atrial flutter accounted for a further 1.4% of symptom–rhythm diagnoses for syncope in these elderly patients (Table 1). AF as a diagnosis for syncope was also shown to increase in incidence in this population when age was stratified.
Comparison of patients who had AF versus those with no AF revealed a significant difference in sex, with predominantly more females in the AF group. Other characteristics, which included baseline (preimplantation) EKGs, hypertension, diabetes, ischemic heart disease, and stroke, were not significantly different between patients with and without AF.
To the best of our knowledge, this is the first cohort study to examine the incidence of AF as the diagnosed arrhythmia for unexplained syncope in ILR patients 65 years and older. Although AF, in the absence of pauses, is not classically considered a common cause of syncope, it has received attention as of late, with results from the Irish Longitudinal Study on Ageing showing that adults aged 65–74 years old with self‐reported syncope have twice‐fold likelihood of having AF (Jansen et al., 2015). The results of this study support this finding.
This study is limited by its small sample size and retrospective design. Another limitation of this study is the different models of ILR used (i.e., different AF detection algorithms) as well as non‐standardized device programming. It is also likely that we underestimated the incidence of AF in this population given that some patients might not have been able to activate their ILR during pre‐syncope/syncope and/or recall symptoms for comparison with their ILR recordings. Given the known increased prevalence of cardiac and non‐cardiac causes of syncope in older patients (Chen et al., 2006), it is also possible that some of the AF diagnosed in this study may be a bystander diagnosis (Wong et al., 2011), although establishing symptom–rhythm correlation for all our arrhythmia diagnoses and the ILRs’ retrograde memory have protected against this effect. Previous decision analysis has shown that the benefits of stroke prevention generally outweigh the risk of anticoagulation in older AF patients with predisposition to falls and syncope (Man‐Son‐Hing & Laupaci, 2003). Thus, whether treatment of ILR‐detected AF associated with syncope confers benefit, such as stroke reduction through anticoagulation, is an interesting avenue to be explored.
5. CONCLUSION
In conclusion, this study shows that atrial fibrillation and atrial flutter were common diagnostic rhythms in our cohort of older patients 65 years and older with ILRs for unexplained syncope. Atrial fibrillation was observed more frequently in females. Further studies are needed to define the need for AF treatment and assessment of stroke risk in this population.
Arcinas LA, McIntyre WF, Hayes CJ, Ibrahim OA, Baranchuk AM, Seifer CM. Atrial fibrillation in elderly patients with implantable loop recorders for unexplained syncope. Ann Noninvasive Electrocardiol. 2019;24:e12630 10.1111/anec.12630
REFERENCES
- Benjamin, E. J. (1994). Independent risk factors for atrial fibrillation in a population‐based cohort. The Journal of the American Medical Association, 271(11), 840 10.1001/jama.1994.03510350050036 [DOI] [PubMed] [Google Scholar]
- Brignole, M. (2018). Recommendations of the European Society of Cardiology Guidelines for the diagnosis and management of syncope (version 2018). Oxford Medicine Online.
- Brignole, M. , Gianfranchi, L. , Menozzi, C. , Raviele, A. , Oddone, D. , Lolli, G. , & Bottoni, N. (1993). Role of autonomic reflexes in syncope associated with paroxysmal atrial fibrillation. Journal of the American College of Cardiology, 22(4), 1123–1129. 10.1016/0735-1097(93)90426-2 [DOI] [PubMed] [Google Scholar]
- Brignole, M. , Vardas, P. , Hoffman, E. , Huikuri, H. , Moya, A. , Ricci, R. , … Botto, G. L. (2009). Indications for the use of diagnostic implantable and external ECG loop recorders. Europace, 11(5), 671–687. 10.1093/europace/eup097 [DOI] [PubMed] [Google Scholar]
- Burkowitz, J. , Merzenich, C. , Grassme, K. , & Brüggenjürgen, B. (2016). Insertable cardiac monitors in the diagnosis of syncope and the detection of atrial fibrillation: A systematic review and meta‐analysis. European Journal of Preventive Cardiology, 23(12), 1261–1272. 10.1177/2047487316632628 [DOI] [PubMed] [Google Scholar]
- Chen, L. Y. , Shen, W. , Mahoney, D. W. , Jacobsen, S. J. , & Rodeheffer, R. J. (2006). Prevalence of syncope in a population aged more than 45 years. The American Journal of Medicine, 119(12), 1088.e1–1088.e7. 10.1016/j.amjmed.2006.01.029 [DOI] [PubMed] [Google Scholar]
- Ibrahim, O. A. , Drew, D. , Hayes, C. J. , Mcintyre, W. , Seifer, C. M. , Hopman, W. , … Baranchuk, A. M. (2017). Implantable loop recorders in the real world: A study of two Canadian centers. Journal of Interventional Cardiac Electrophysiology, 50(2), 179–185. 10.1007/s10840-017-0294-y [DOI] [PubMed] [Google Scholar]
- Jansen, S. , Frewen, J. , Finucane, C. , Rooij, S. E. , Velde, N. V. , & Kenny, R. A. (2015). AF is associated with self‐reported syncope and falls in a general population cohort. Age and Ageing, 44(4), 598–603. 10.1093/ageing/afv017 [DOI] [PubMed] [Google Scholar]
- Krahn, A. D. , Klein, G. J. , Yee, R. , & Skanes, A. C. (2001). Randomized assessment of syncope trial. Circulation, 104(1), 46–51. 10.1161/01.cir.104.1.46 [DOI] [PubMed] [Google Scholar]
- Leitch, J. , Klein, G. , & Tee, R. (1991). Neurally mediated syncope and atrial fibrillation. New England Journal of Medicine, 324(7), 495–496. 10.1056/nejm199102143240718 [DOI] [PubMed] [Google Scholar]
- Man‐Son‐Hing, M. , & Laupaci, A. (2003). Anticoagulant‐related bleeding in older persons with atrial fibrillation: Physicians' fears often unfounded. Archives of Internal Medicine, 163(13), 1580–1586. 10.1001/archinte.163.13.1580 [DOI] [PubMed] [Google Scholar]
- Mcintyre, W. F. , Liebrecht, K. , Daba, L. , & Seifer, C. M. (2016). Diagnostic value of implantable loop recorders in elderly adults. Journal of the American Geriatrics Society, 64(6), 1370–1372. 10.1111/jgs.14174 [DOI] [PubMed] [Google Scholar]
- Mcnicholas, T. , Tobin, K. , O’Callaghan, S. , & Kenny, R. A. (2017). Is orthostatic hypotension more common in individuals with atrial fibrillation?—Findings from The Irish Longitudinal Study on Ageing (TILDA). Age and Ageing, 46(6), 1006–1010. 10.1093/ageing/afx096 [DOI] [PubMed] [Google Scholar]
- Odutayo, A. , Wong, C. X. , Hsiao, A. J. , Hopewell, S. , Altman, D. G. , & Emdin, C. A. (2016). Atrial fibrillation and risks of cardiovascular disease, renal disease, and death: Systematic review and meta‐analysis. British Medical Journal, 354, i4482 10.1136/bmj.i4482 [DOI] [PubMed] [Google Scholar]
- Roberto, A. F. , Ronaldo, F. R. , & Silvio, C. S. (2012). Atrial fibrillation in the elderly. Journal of Geriatric Cardiology, 9(2), 91–100. 10.3724/sp.j.1263.2011.12293 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ruwald, M. H. , Hansen, M. L. , Lamberts, M. , Hansen, C. M. , Numé, A. , Vinther, M. , … Gislason, G. H. (2013). Comparison of incidence, predictors, and the impact of co‐morbidity and polypharmacy on the risk of recurrent syncope in patients. The American Journal of Cardiology, 112(10), 1610–1615. 10.1016/j.amjcard.2013.07.041 [DOI] [PubMed] [Google Scholar]
- Wong, J. A. , Leong‐Sit, P. , Skanes, A. C. , Yee, R. , Gula, L. J. , Krahn, A. D. , & Klein, G. J. (2011). Atrial fibrillation causing syncope? Pacing and Clinical Electrophysiology, 35(1), 104–105. 10.1111/j.1540-8159.2011.03106.x [DOI] [PubMed] [Google Scholar]