Arterial hypertension (AH), atrial fibrillation and stroke represent a lethal triangle. Hypertension promotes the development of atherosclerotic vascular disease leading to both ischemic and hemorrhagic stroke . Moreover, AH is an independent predictor for the development of rhythm irregularities, including atrial fibrillation . On the other hand, atrial fibrillation confers a three to five-fold increase in the risk of stroke due to thromboembolism . This holds even when atrial fibrillation is ‘asymptomatic’, as is the case in about one-third of patients. Therefore, an early detection of the arrhythmia might allow the timely introduction of therapies to protect the patient, not only from the consequences of the arrhythmia, but also from atrial fibrillation progression to an utterly refractory problem.
Defining the subgroup of patients that should be screened for ‘asymptomatic’ atrial fibrillation is of outmost importance. According to the recent European Society of Cardiology (ESC) guidelines , opportunistic screening for silent atrial fibrillation seems cost-effective in elderly populations and similar effects have been reported in patients with ischemic stroke . Therefore, expanding the screening procedure to other high-risk populations would lead to increased detection of subclinical atrial fibrillation.
AH induces structural and electrophysiological changes in both left atrium and ventricle that lead to atrial fibrillation. It is well known that untreated or suboptimally treated AH leads to the development of left ventricular (LV) hypertrophy (LVH) . In the presence of LVH, LV compliance is reduced, LV stiffness and filling pressure increase, coronary flow reserve is decreased, wall stress is increased while both sympathetic nervous system and renin–angiotensin–aldosterone systems are activated. LVH also enhances connective tissue deposition and fibrosis. Finally, structural remodeling results in electrical dissociation between muscle bundles, as well as local conduction heterogeneities facilitating the initiation and perpetuation of atrial fibrillation. Over time tissue remodeling promotes and maintains atrial fibrillation by changing the fundamental properties of the atria.
Equally important to the selection of the appropriate subgroup under investigation for the asymptomatic atrial fibrillation detection is the use of a feasible monitoring device. In a recent article, Mairesse et al. grouped the atrial fibrillation detection tests into four major categories: blood pressure (BP) monitors, pulse palpation, non-12-lead ECG, and smartphone applications with similar sensitivity of all four methods. Automated BP devices that also detect atrial fibrillation (or at least indicate an ‘irregular heart rate’) based on oscillometric analysis are now widely available in the community, in contrast to smartphone applications, which are not very practical, especially among the elderly patients. In a head to head comparison study of AliveCor Heart Monitor-a smartphone-based heart monitor-versus the Microlife WatchBP Office AFIB, Chan et al.  demonstrated that the latter has higher sensitivity [83.3%, 95% confidence interval (CI): 62.6–95.3] compared with the AliveCor detector (66.7%, 95% CI: 44.7–84.4), regardless of patient age and hypertension or diabetes mellitus status.
We designed the HOME-AF study to investigate the proportion of new, asymptomatic cases with suspected atrial fibrillation detected through home BP screening among hypertensive patients. Potential association of the newly identified suspected atrial fibrillation cases with AH phenotypes (white coat/masked/uncontrolled) and the CHA2DS2-VASc score was also investigated.
PATIENTS AND METHODS
HOME-AF is a prospective, noninterventional, epidemiological study that enrolled patients aged at least 50 years old with either known and under treatment AH or newly diagnosed and untreated AH, visiting 70 private sector cardiologist offices and six outpatient hypertension hospital clinics distributed over the entire country. Patients were enrolled consecutively in study sites to avoid selection bias. Newly diagnosed hypertension was defined as: either office BP at least 140/90 mmHg in two office visits or/and home BP at least 135/85 mmHg in patients or office BP at least 140/90 mmHg in one office visit and home BP at least 135/85 mmHg in patients not treated with antihypertensive treatment. Patients with BP at least 140/90 mmHg in only one office visit were excluded. Hypertension groups were defined as follows: first, controlled with office BP less than 140/90 mmHg and home BP less than 135/85 mmHg; second, uncontrolled AH with office BP at least 140/90 mmHg and home BP at least 135/85 mmHg; third, white-coat AH with office BP at least 140/90 mmHg and home BP less than 135/85 mmHg and fourth, masked AH with office BP less than 140/90 mmHg and home BP at least 135/85 mmHg.
The recruitment period lasted for 2 years. Informed consent was obtained from all individual participants and the study was approved by the respective Institutional Research Boards. Patients unsuitable for home BP monitoring (people with physical or mental conditions), patients with pacemakers or a known history of atrial fibrillation and patients who participated in an interventional clinical trial were excluded from the study.
An automatic BP monitor (Microlife BP A6 PC) that is also capable of detecting suspected atrial fibrillation cases was provided by the attending physician to all patients. The device has an option to perform one or three consecutive BP measurements. The three BP measurement option was used in the protocol, as it increases the sensitivity (97%) and specificity (89%) for suspected atrial fibrillation screening. The procedure requires three out of three BP measurements with irregular rhythm to indicate possible atrial fibrillation presence. Regarding the method used, two studies [9,10] have demonstrated that the technology based on automated BP measurement by an oscillometric method has a high sensitivity 97–100% (low false negatives) and specificity 89% (low false positives) for atrial fibrillation detection, when compared with the ECG. The BP monitor measures the last 10 pulse intervals during cuff deflation and calculates the mean and SD of the intervals. An irregularity index is defined as the SD divided by the mean of the time intervals. To reduce the effect of premature beats on the irregularity index, a cutoff value of 25% was chosen so that each of the 10 pulse-beat intervals that is 25% greater or 25% less than the mean time interval is deleted. The remaining time intervals are used to calculate the irregularity index. If the irregularity index exceeds a threshold value of 0.06, the rhythm is considered irregular. The number of beats analyzed, and the irregularity index threshold value of 0.06 were chosen to maximize sensitivity for detecting atrial fibrillation. In a review and meta-analysis of six studies (n = 2332) regarding the diagnostic accuracy of this algorithm a pooled sensitivity and specificity of 0.98 (95% CI 0.95, 1.00) and 0.92, respectively, were detected .
All eligible patients had two medical office visits (visits 1 and 2). During visit 1, medical history was recorded, and BP was measured with the automatic device at the physician's office. The average value of these three BP measurements was recorded. A 12-lead ECG at rest was also performed at all participants. If a patient was identified with atrial fibrillation during ECG, the patient continued to be monitored at home, but was excluded from the primary point analysis. The device was then given to the patients for home BP measurements along with detailed instructions. The device was used in accordance to the recommendations of 2013 ESC/European Society of Hypertension (ESH) Guidelines  for BP management: BP was measured daily for at least 3–4 days and preferably for 7 consecutive days in the morning, as well as in the evening. Patients were instructed to take the BP measurements in a quiet room and in the seated position, with back and arm supported, after 5 min of rest. A logbook for measurement recording was not needed, as the device was equipped with internal memory for value storage. Home BP was considered the average of these readings, with the exclusion of the first monitoring day, as the first day is considered part of the familiarization period. During visit 2, patients returned the device to the physician and another triple BP measurement was recorded. The primary end point of our study was defined as the proportion of new cases of patients with suspected atrial fibrillation detected through the home BP screening among hypertensive patients.
At the end of each site's data collection period a questionnaire on physicians’ satisfaction (designed by the study team, three questions with three to five alternative responses) was administered to evaluate the added value of this screening tool for suspected atrial fibrillation detection (scale: major, important, moderate, minor and none).
Summary statistics were based on absolute and relative frequency distribution tables for the categorical variables of interest, whereas continuous variables were described via mean, SD, minimum and maximum value. All relative frequencies were based on the number of patients with available data. Univariate analysis was used to investigate the association between suspected atrial fibrillation detection (with or without ECG concordance) and the CHA2DS2-VASc items, as well as the four hypertensive groups. The concordance between BP monitoring and ECG in terms of atrial fibrillation detection was investigated by Mc Nemar's chi-squared statistic. Continuous variables with respect to atrial fibrillation groups were analyzed according to the t distribution. Multivariate analysis was employed to confirm the univariate analysis results. Only the variables with significance of P less than 0.10 in the univariate analysis [age, sex and chronic heart failure/LV dysfunction (CHF/LVD)] were entered in the model. All tests were two-sided and the level of statistical significance was set to P less than 0.05. The magnitude of association between variables was demonstrated by odds ratio followed by the corresponding 95% CIs. No imputation was performed in case of missing values. We further analyzed a subgroup of patients with positive indication for suspected atrial fibrillation by the BP monitor with descriptive statistics. In particular, the relationship between suspected atrial fibrillation episodes indicated by the device and the number (no) of days of home assessments was explored for each patient and each day, by the division of ‘Total no of suspected atrial fibrillation episodes at the specific day/Total no of measurements the specific date’. To assess if the presence of atrial fibrillation was repeated, the degree of repetitions was investigated. The following calculation was performed on each patient to extract the repetition score which could range from 0 to 1 (low atrial fibrillation repeatability to high atrial fibrillation repeatability). The overall score was calculated as the no of days of home assessments with suspected atrial fibrillation presence/total no days with home assessments. Morning, afternoon and night repetition score were similarly performed, for example, morning score as the no of morning home assessments with suspected atrial fibrillation presence/total no days with home assessments. The number of monitoring days with the BP monitor was also tested as a variable in the multivariate analysis.
We recruited 2408 patients with no known history of atrial fibrillation (mean age 66.5 ± 10, 49% males) (Table 1). From the total population, 2303 (95.6%) patients had already known AH, while 105 (4.4%) patients were newly diagnosed for AH. Regarding AH phenotypes among the patients with known hypertension, 34.2% had controlled BP, 40.3% had uncontrolled AH, 20.3% had white-coat hypertension and finally 5.2% had masked AH. Only 0.8% of patients discontinued the study at office visit two mainly due to operational problems (six patients were lost to follow-up, one withdrew consent, one had an adverse event unrelated to study procedure and 11 discontinued due to device malfunction).
The study population had the following risk factors for stroke: 186 (7.7%) patients had a history of CHF/LVD, 535 (22.2%) suffered from diabetes mellitus, 148 (6.1%) referred to a previous history of stroke/transient ischemic attack and 441 (18.3%) had known vascular disease. The whole population had a mean CHA2DS2-VASc score of 2.8 ± 1.4 (Table 1).
Of the 2408 AH patients, 2340 patients were finally analyzed for the primary endpoint [68 patients were excluded from the primary point analysis due to ECG detected atrial fibrillation (41) and incomplete data (27)]. Suspected atrial fibrillation was detected by BP monitoring in 12.5% of this group (Table 2). In those patients, CHA2DS2-VASc score was significantly higher compared with all other patients (3.3 ± 1.4 vs. 2.8 ± 1.4, P < 0.001). Suspected atrial fibrillation detection was associated with advanced age (≥75 years, P < 0.001) and female sex (P = 0.01). A nonstatistically significant association between atrial fibrillation detection and the history of CHF/LVD was observed (P = 0.06) (Table 3). However, in the multivariate analysis, only age and sex were independently associated with suspected atrial fibrillation detection (Table 4). Moreover, the risk of suspected atrial fibrillation detection was increased two-fold in older patients at least 75 years old compared with younger ones less than 64 years old (Table 4). No significant association between suspected atrial fibrillation cases and AH phenotypes (white coat hypertension/uncontrolled/masked) was identified among already known or newly detected hypertensive patients.
During visit 1 both BP monitoring and ECG were performed. Suspected atrial fibrillation detected by Microlife Afib was observed in 2.6% of patients, whereas atrial fibrillation detected by ECG was found in 1.7% of patients (Table 2). Therefore, the Microlife AFib at visit 1 detected 0.9% more suspected atrial fibrillation cases compared with the ECG, (95% CI 0.3–1.6%). Similarly, in 55.0% of the patients in whom ECG atrial fibrillation was detected, atrial fibrillation was also detected by the BP monitor (data not shown).
In an ad-hoc analysis of our data, the presence of atrial fibrillation was repeated in the 40.2% ± 31.6 of the measurements with a tendency of being more common in the morning (Suppl Table 5, http://links.lww.com/HJH/B165). However, there was no difference in the mean probability of atrial fibrillation detection between day 3 and 4 or 7 and the variable ‘number of days’ of home BP measurement did not associate with suspected atrial fibrillation detection, when entered in the multivariate analysis (Suppl Tables 3, http://links.lww.com/HJH/B165 and 4, http://links.lww.com/HJH/B165).
Finally, a questionnaire was posed to the physicians regarding their interest of implementing this screening approach in everyday clinical practice. Overall, 64 (86.5%) physicians answered positively. Almost two-thirds (64%) considered the future adoption of BP home monitoring in everyday practice in the primary care as important. More than 60% (62%) underlined that the study procedures had at least an important contribution to the better detection of suspected atrial fibrillation among AH patients.
In our study, the incidence of suspected atrial fibrillation in hypertensive patients is similar to previously reported values. Approximately one-tenth (12.5%) of hypertensive patients had suspected atrial fibrillation, as it was detected by two daily BP readings for at least 4–7 days. Similar findings were reported by Marazzi et al., who studied 503 patients referred to a hypertension clinic. Asymptomatic atrial fibrillation was present in 22% of the already known or newly detected hypertensive patients compared with 20% detected by ECG. Likewise, Wiesel et al. found that 30% of elderly and hypertensive patients had occult atrial fibrillation, as detected by a BP monitor. In addition, when BP monitoring was performed for a month in the TRIPPS trial, 19% of the patients, who met more than one criteria for stroke, had subclinical atrial fibrillation .
Detecting a significant number of asymptomatic atrial fibrillation cases among hypertensive patients is not surprising. Based on data derived from population studies , the relation between AH and atrial fibrillation has been well known. According to the Framingham Study , AH confers a 1.5-fold risk and is an independent risk factor of developing atrial fibrillation after adjusting for age and other associated conditions. Because of its high prevalence, hypertension is responsible for more atrial fibrillation incidents in the population (14%) than any other risk factor. Therefore, the routine use of an automatic BP monitor that is also capable of detecting suspected atrial fibrillation could benefit a large number of patients by preventing thousands of strokes each year. A Microlife device has been evaluated for atrial fibrillation detection in a large primary care setting compared with a reference standard of lead-I ECG . In 5969 patients Microlife WatchBP Home was an effective means to screen for atrial fibrillation, with a reasonable sensitivity of 80.6% and a high negative predictive value of 99.8%. Moreover, in a younger patient population aged less than 65 years with a lower prevalence of atrial fibrillation, Microlife WatchBP Home A demonstrated a similar diagnostic accuracy. Of note the mean CHA2DS2-VASc score of the population was 2.8 ± 1.3 and hypertension was present in 4948 patients (82.9%). Nonetheless, it seems that not all hypertensive patients carry the same risk of developing atrial fibrillation. In our study, hypertensive patients with suspected atrial fibrillation had significantly higher CHA2DS2-VASc score compared with patients with non-detected atrial fibrillation. Perhaps this finding implies that a subgroup of hypertensive patients carries a higher risk of arrhythmias. In our population, elderly patients had a two-fold increased risk of suspected atrial fibrillation detection, while as shown by the multivariate analysis, an age at least 75 years was also independently associated with suspected atrial fibrillation. Indeed, advanced age is known to correlate well with atrial structural changes (cardiac fibrosis), widespread conduction slowing, as well as anatomically determined functional conduction delay and block ; all of which could probably induce atrial fibrillation. Significantly in our study female sex was also significantly associated with suspected atrial fibrillation, a finding that comes into contradiction with the Framingham's study conclusions . However, it should be noted that in our population, female participants were older and with more comorbidities than male counterparts, which could lead to a higher propensity of developing suspected atrial fibrillation (Suppl. Tables 1, http://links.lww.com/HJH/B165, 2, http://links.lww.com/HJH/B165). In addition, since all patients were older than 50 years, a menopausal state of the female participants accompanied by subsequent hormonal changes should be implied. As Odening et al. state, the incidence of atrial fibrillation in premenopausal women is low but it increases after menopause particularly at ages over 50 years, suggesting a harmful effect of postmenopausal hormonal changes – such as the pronounced decrease in estrogen – with regards to the development of atrial fibrillation. Based on indirect evidence, estrogen levels seem to exert antiarrhythmic effects on atrial electrical features (longer atrial potential duration) and beneficial effects on both structural remodeling (attenuation of fibrosis) and diastolic function . Estrogen levels reduction leads to increased BMI, BP and LDL cholesterol levels and subsequently to the appearance of the metabolic syndrome all of which are well established risk factors for developing atrial fibrillation. In addition, in a subanalysis of the Atrial Fibrillation Follow-up Investigation of Rhythm Management trial, female sex was significantly associated with higher rates of left atrial remodeling and adverse cardiovascular endpoints . All the aforementioned, could contribute to the explanation of the female sex emergence as an independent factor for the detection of suspected atrial fibrillation in our study.
Among the common comorbidities that frequently coexist with AH, like diabetes and CHF, only the latter displayed a nonsignificant association with suspected atrial fibrillation detection. It is already known that AH causes LVH through a complex mechanism of wall stress increase and alterations on neurohormones, growth factors and cytokines levels . The primary mechanism by which heart reduces the stress on the LV wall, imposed by the AH, is the hypertrophy of cardiomyocytes, thus leading to LV wall thickening and LV mass increase. Cardiomyocyte hypertrophy results in altered heart relaxation and calcium (Ca2+) metabolism, which in turn leads to impaired diastolic filling and systolic contraction, respectively . As a consequence, CHF/LVD develops, subsequent atrial remodeling is facilitated and altered atrial conduction properties result in atrial fibrillation . However, this finding was not fully verified by our multivariate analysis.
Similarly, a very short increase in the number of monitoring days (from 3 to 7 days) did not seem to increase the probability of suspected atrial fibrillation detection, as shown in our ad hoc analysis, although our study was not initially designed to answer this question. However, the suspected atrial fibrillation was a repeated finding measured in almost half of the measurements therefore supporting our findings.
Study strengths and limitations
The study is strengthened by the usage of a validated device with a standardized method of monitoring. This method for home BP monitoring and AH management was recommended by the previous ESC/ESH 2013 guidelines  and continues to be strongly recommended by the recent ESC/ESH 2018 ones . Moreover, data from over 2400 patients at least 50 years old were prospectively collected over a week's follow-up. Data were collected by hypertension specialists covering most of the country, thus increasing the likelihood of a representative population at risk and supporting the generalization of our results in both untreated and treated hypertensive patients, as well as in several AH-related phenotypes. This design enabled capturing real world practices and improved the study's external validity. Furthermore, the study satisfaction questionnaire posed in participating physicians clearly supported the added value of the above described method regarding the atrial fibrillation screening process.
However, several limitations should also be considered concerning the current study since its cross-sectional design prevented us from inferring definite cause–effect relationships. Another limitation was observed during office visit 1 when BP monitoring detected more suspected atrial fibrillation cases than the ECG performed at the same time. Hence, there was poor agreement between ECG and the device. Although Microlife AFib device is reported to have a high sensitivity and specificity, one could not exclude either false-positive readings particularly in patients with very frequent premature complexes or false-negative readings since many episodes of asymptomatic atrial fibrillation may not have been detected by the device. In addition, since our patients were not subjected to a 24-h ECG recording, the BP device results were not verified.
A validated automatic BP monitor, useful for both home BP monitoring and suspected atrial fibrillation detection, could benefit a number of hypertensive patients, especially elderly as well as females, by stroke prevention after atrial fibrillation diagnosis confirmation and subsequent antiarrhythmic and anticoagulation therapy initiation.
The novelty of our study lies in the large sample size (2408 patients) of exclusively hypertensive patients managed by private and hospital-based cardiologists around Greece. Other studies [9–11,13–14] have also used the same method of suspected atrial fibrillation detection through BP monitor, but included either smaller samples sizes [9–11,14], unselected population [9,14] or only one office measurement [9,10,13]. We used home BP monitoring and recorded rhythm irregularity for at least 1 week giving patients an easy to use screening tool both for their BP and their heart rate. An emerging conclusion from our study was that postmenopausal, hypertensive women might be at higher risk of developing suspected atrial fibrillation.
In conclusion, the findings of the current study point to a subgroup of hypertensive patients that merits particular attention, since subclinical, suspected atrial fibrillation probably coexists. This subgroup of hypertensive patients consists mainly of elderly and female patients, who seem to be in greater risk for atrial fibrillation complications. Our results underscore the role of AH-associated atrial fibrillation as a contributor to increased thromboembolic risk and revise the need of asymptomatic suspected atrial fibrillation detection devices in the clinical practice. Those devices can only suggest the patients that need further investigation regarding occult atrial fibrillation through a 24 h or more extended ECG.
The authors would like to thank Periklis Giovas, Internist for his valuable contribution in the study design.
*HOME-AF study group investigators-Author contributors:
G. Aggelopoulos1, G. Albanis3, F. Anastasiadis4, H. Antoniou5, A. Arapogianni4, D. Arsenis4, S. Avlonitis6, V. Vachliotis4, G. Vlachou4, K. Vrogkistinos7, A. Giakoumis8, G. Giannakoulas7, A. Gountinakis4, A. Graikos4, O. Gouli7, E. Demerouti4, N. Zavos2, T. Zafeiriou4, D. Zaxarakis9, G. Zonios10, C. Kagiadaki7, M. Kalaitzakis11, A. Kalantaridou4, T. Kalos4, P. Kalpidis7, P. Kamvrogiannis12, I. Kantiloros4, P. Karidas4, T. Karonis9, D. Konstantinides4, T. Konti4, I. Kotrogiannis10, D. Kotsionis4, K. Koutrakis2, Z. Kopsida3, P. Krezias4, E. Lagoudi7, G. Lirarakis13, C. Manetos4, A. Manousakis11, A. Markomanolaki4, I. Mavrepis7, I. Mermiklis9, L. Milonadaki6, T. Mitras6, C. Mitroulas7, G. Mpaltogiannis10, N. Vatkalis6, C. Ntouloulis2, N. Oikonomidis4, E. Papadakis13, G. Papadimitriou14, E. Papadimitriou13, N. Papadopoulos7, D. Papadopoulou7, D. Papaikonomou7, C. Papastamou4, C. Parisis2, H. Petropoulos4, A. Pittaras4, D. Rovithis7, E. Rosmarakis4, A. Samothrakitis3, D. Savvalas4, A. Sikiotis4, P. Sinopoulou4, P. Staphilas7, A. Schinas4, E. Tsagka4, A. Tsapa7, I. Tsiantis4, H. Tsougkos4, G. Fagogenis15, S. Christodoulou14, I. Choursalas16.
1Aigio, 2Larisa, 3Patra, 4Attica, 5Evia, 6Kavala, 7Thessaloniki, 8Tripoli, 9Alexandroupoli, 10Ioannina, 11Chania, 12Volos, 13Hraklio, 14Lamia, 15Kerkira, 16Korinthos
Conflicts of interest
P.S., J.S. and D.M. are currently Pfizer employees. For the remaining authors none were declared. This study was sponsored by Pfizer Greece.
1. Hollander W. Role of hypertension in atherosclerosis and cardiovascular disease. Am J Cardiol
2. Kannel WB, Wolf PA, Benjamin EJ, Levy D. Prevalence, incidence, prognosis, and predisposing conditions for atrial fibrillation: population-based estimates. Am J Cardiol
3. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke
4. Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, et al. 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Europace
5. Levin LÅ, Husberg M, Sobocinski PD, Kull VF, Friberg L, Rosenqvist M, Davidson T. A cost-effectiveness analysis of screening for silent atrial fibrillation after ischaemic stroke. Europace
6. Manolis AJ, Rosei EA, Coca A, Cifkova R, Erdine SE, Kjeldsen S, et al. Hypertension and atrial fibrillation: diagnostic approach, prevention and treatment. Position paper of the Working Group ‘Hypertension Arrhythmias and Thrombosis’ of the European Society of Hypertension. J Hypertens
7. Mairesse GH, Moran P, Van Gelder IC, Elsner C, Rosenqvist M, Mant J, et al. ESC Scientific Document Group. Screening for atrial fibrillation: a European Heart Rhythm Association (EHRA) consensus document endorsed by the Heart Rhythm Society (HRS), Asia Pacific Heart Rhythm Society (APHRS), and Sociedad Latinoamericana de Estimulación Cardíaca y Electrofisiología (SOLAECE). Europace
8. Chan PH, Wong CK, Pun L, Wong YF, Wong MM, Chu DW, et al. Head-to-head comparison of the alivecor heart monitor and microlife watchBP Office AFIB for atrial fibrillation screening in a primary care setting. Circulation
9. Wiesel J, Fitzig L, Herschman Y, Messineo FC. Detection of atrial fibrillation using a modified Microlife blood pressure monitor. Am J Hypertens
10. Stergiou GS, Karpettas N, Protogerou A, Nasothimiou EG, Kyriakidis M. Diagnostic accuracy of a home blood pressure monitor to detect atrial fibrillation. J Hum Hyperten
11. Verberk WJ, Omboni S, Kollias A, Stergiou GS. Screening for atrial fibrillation with automated blood pressure measurement: research evidence and practice recommendations. Int J Cardiol
12. Mancia G, Fagard R, Narkiewicz K, Redón J, Zanchetti A, Böhm M, et al. Task Force Members. 2013 ESH/ESC guidelines for the management of arterial hypertension
. The Task Force for the management of arterial hypertension
of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). J Hypertens
13. Marazzi G, Iellamo F, Volterrani M, Lombardo M, Pelliccia F, Righi D, et al. Comparison of Microlife BP A200 Plus and Omron M6 blood pressure monitors to detect atrial fibrillation in hypertensive patients. Adv Ther
14. Wiesel J, Abraham S, Messineo FC. Screening for asymptomatic atrial fibrillation while monitoring the blood pressure at home: trial of regular versus irregular pulse for prevention of stroke (TRIPPS 2.0). Am J Cardiol
15. Benjamin EJ, Levy D, Vaziri SM, DAgostino RB, Belanger AJ, Wolf PA. Independent risk factors for atrial fibrillation in a population-based cohort. The Framingham Heart Study. JAMA
16. Chan PH, Wong CK, Pun L, Wong YF, Wong MM, Chu DW, et al. Diagnostic performance of an automatic blood pressure measurement device, Microlife WatchBP Home A, for atrial fibrillation screening in a real-world primary care setting. BMJ Open
17. Wasmer K, Eckardt L, Breithardt G. Predisposing factors for atrial fibrillation in the elderly. J Geriatr Cardiol
18. Odening KE, Deiß S, Dilling-Boer D, Didenko M, Eriksson U, Nedios S, et al. Mechanisms of sex differences in atrial fibrillation: role of hormones and differences in electrophysiology, structure, function, and remodelling. Europace
19. Pothineni NV, Vallurupalli S. Gender and atrial fibrillation: differences and disparities. US Cardiol Rev
20. Proietti M, Raparelli V, Basili S, Olshansky B, Lip GY. Relation of female sex to left atrial diameter and cardiovascular death in atrial fibrillation: the AFFIRM Trial. Int J Cardiol
21. Díez J, Frohlich ED. A translational approach to hypertensive heart disease. Hypertension
22. E Darby Md A. Management of atrial fibrillation in patients with heart failure. J Atr Fibrillation
23. Williams B, Mancia G, Spiering W, Agabiti Rosei E, Azizi M, Burnier M, et al. Authors/Task Force Members. 2018 ESC/ESH Guidelines for the management of arterial hypertension
: The Task Force for the management of arterial hypertension
of the European Society of Cardiology and the European Society of Hypertension. J Hypertens
* A list of other author contributors is listed in the ACKNOWLEDGMENTS section.