Atrial fibrillation is increasingly common in the general population [1,2]. It is associated with serious complications such as stroke and heart failure [3–6]. However, risk of atrial fibrillation, morbidity and mortality of atrial fibrillation patients cannot be fully explained by these complications . Evidence suggests an association between vascular dysfunction and atrial fibrillation [8,9]. For this reason, vascular dysfunction might need to be considered when assessing risk of prevalent atrial fibrillation and overall health status of atrial fibrillation individuals.
Vascular dysfunction is a systemic process related to cardiovascular prognosis . As a major underlying substrate, classical cardiovascular risk factors including age, sex, smoking, obesity, hypertension, diabetes mellitus, and dyslipidemia have been identified . By creating a pro-inflammatory state and increasing oxidative stress, these risk factors contribute to increased arterial stiffness . Coherently, there is a close relation between arterial stiffness and atrial fibrillation. After treatment of atrial fibrillation with cardioversion, relapse rates differed dependent on arterial stiffness . Notably, arterial stiffness seems to be increased even in individuals with lone atrial fibrillation [14,15] and was identified as a predictor of atrial fibrillation [15,16].
Different vascular beds including conduit arteries and microvascular function play a significant role in cardiovascular homeostasis and disease. Flow-mediated dilation (FMD) of the brachial artery and peripheral arterial tonometry (PAT) measured by volume plethysmographic pulse amplitude registration are noninvasive vascular function measures that integrate structural changes, pulsatile blood flow characteristics and vascular reactivity. Both have been related to cardiovascular disease risk factors and outcome [17–19]. Furthermore, inflammation and oxidative stress are major determinants of vascular function . Thus, the measurement of FMD and PAT may provide an integrated measure of vascular impairment and morbidity and predict the presence of atrial fibrillation .
On the basis of these considerations, we hypothesized that impaired vascular reactivity may contribute to an increased risk of atrial fibrillation and examined the association of noninvasively measured vascular function parameters and atrial fibrillation in our large community-based cohort.
Fifteen thousand and ten participants of the population-based Gutenberg Health Study enrolled at the Department of Medicine 2, University Medical Center Mainz, aged 35–74 years were eligible for analysis. A detailed study outline has been published previously . In brief, during a clinic visit, comprehensive information on cardiovascular risk factors was acquired by standardized computer-assisted interview, anthropometric measures, surface electrocardiogram and noninvasive cardiovascular function testing. We categorized smoking into nonsmokers (never smokers and former smokers) and current smokers. The definition of diabetes mellitus constituted individuals who reported a physician diagnosis of diabetes and/or had an on-site fasting blood glucose concentration of at least 126 mg/dl (minimum 8-h fast) or a blood glucose level of at least 200 mg/dl at any time. Dyslipidemia was assumed when the participant knew about a physician's diagnosis of dyslipidemia and/or a low-density lipoprotein/high-density lipoprotein (LDL/HDL) cholesterol ratio of greater than 3.5 was measured on site. Hypertension was defined as antihypertensive drug treatment and/or a mean SBP of at least 140 mmHg and/or a mean DBP of at least 90 mmHg. Data on cardiac medication was collected by self-report. The diagnosis of atrial fibrillation was based on a history of atrial fibrillation and/or the 12-lead electrocardiogram (GE Cardiosoft) documentation of atrial fibrillation or atrial flutter . Atrial fibrillation was adjudicated by at least two cardiologists. In difficult cases, a third opinion by a physician trained in electrophysiology was obtained.
All participants provided written, informed consent. The study was approved by the local Ethics Committee. All authors have read and approved the manuscript as written.
Vascular function measurement
FMD was measured according to recommendations using upper arm cuff occlusion for 5 min to determine brachial artery diameter at rest and after induction of reactive hyperemia . Two-dimensional high-resolution ultrasonic images of the right brachial artery were acquired on a Philips HD11XE CV ultrasound machine (Philips, Best, The Netherlands). Loops recorded at baseline and 60 s after cuff release were saved digitally and analyzed semi-automatically on an offline workstation using a customized software package (Medical Imaging Applications LLC, Iowa City, Iowa, USA). As FMD was measured distal from the cuff occlusion, effects of ischemic dilation and endothelial stimulated dilation caused by high-flow-induced shear stress cannot completely be distinguished. Nevertheless, upper arm occlusion has been described as a routinely used technique in FMD measurement and offers several advantages compared with placement of the cuff on the forearm, for example, a greater percentage change in diameter in reactive hyperemia .
PAT was assessed by pneumatic pulse amplitude recorded by the Endo-PAT2000 device (Itamar Medical, Caesarea, Israel). Pulse amplitude was continuously registered in both index fingers with the left index finger as a control. Data were automatically transformed by a computerized algorithm.
For the present investigation, we excluded individuals with missing vascular function measures, bad image quality and technically inadequate studies (insufficient flow occlusion, occlusion for less than 5 min, incomplete data acquisition). N = 12 435 participants with FMD and 10 977 individuals with PAT measurements (84 and 74% of enrolled attendees, respectively) were included for analysis. Inter-observer intra class correlation coefficients of baseline brachial artery diameter and hyperemic brachial artery diameter were 0.87–0.95 and 0.90–0.93 suggesting a good reproducibility of vascular function parameters in our laboratory . The inter-observer intra class correlation coefficient of PAT ratio was 0.62–0.67 and thus slightly lower compared with previous reports [19,22].
Data were analyzed for the total sample and for the subsample of individuals with atrial fibrillation. Continuous data are presented as median and quartiles and binary data as absolute and relative frequencies. Logistic regressions were computed for each vascular function parameter as the independent variable of interest to relate vascular function measures to atrial fibrillation history. The models were adjusted for age, sex, and additionally for heart rate, current smoking, BMI, SBP, diabetes and LDL/HDL cholesterol ratio. Values are shown as odds ratio (OR) per standard deviation (SD) with 95% confidence interval (CI).
FMD (%) is dependent on baseline brachial artery diameter (Dbase). We performed a linear regression analysis with peak brachial artery diameter as dependent variable, baseline brachial artery diameter as independent variable, and sex as interaction variable. From this regression model, we derived separate allometric exponents for men and for women. Individual Dbase-adjusted FMD was calculated as Dpeak/Dbaseallometricexponent. This FMD measure is not dependent on Dbase anymore.
We further estimated the area under the curve using bootstrap correction with 1000 bootstrap samples each on models based on clinical variables of the Framingham risk score (age, age2, sex, male sex × age2, BMI, SBP, antihypertensive medication, congestive heart failure, congestive heart failure × age)  and vascular function parameters in multivariable-adjusted analyses to assess the additive discriminative ability of the vascular function parameters. We assumed a two-sided P < 0.05 as statistically significant.
Analyses were performed using R software, Version 3.5.1 (R Core Team, 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria).
The median age of the overall cohort was 55.0 (46.0, 65.0) years, 50.5% were men. N = 466 study participants had atrial fibrillation, 3.1% of the total sample. Prevalence of cardiovascular risk factors and vascular function measures distribution resembled a middle-aged community sample Table 1.
Boxplots of brachial artery diameter of study participants with and without atrial fibrillation revealed a statistically significant higher baseline brachial artery diameter of study participants with atrial fibrillation (Fig. 1). Median diameter measured 4.81 (4.17, 5.33) mm compared with 4.31 (3.67, 4.93) mm in study participants without atrial fibrillation (P < 0.001). Baseline pulse amplitude showed higher values in atrial fibrillation cases, 6.35 (5.76, 6.78) arbitrary units vs. 6.09 (5.36, 6.71) arbitrary units in nonatrial fibrillation cases (P < 0.001). Vascular function reactivity was significantly altered in atrial fibrillation cases: median FMD was reduced to 1.29 (1.26, 1.33) arbitrary units in atrial fibrillation individuals compared with 1.31 (1.26, 1.33) arbitrary units in nonatrial fibrillation individuals (P < 0.001). Sex-specific subgroup analysis showed no significant differences in FMD in atrial fibrillation and nonatrial fibrillation cases (Supplemental Digital Content Table 1 and 2, http://links.lww.com/HJH/B39, Supplemental Digital Content Figure 1 and 2, http://links.lww.com/HJH/B39). PAT ratio was lower in atrial fibrillation compared with nonatrial fibrillation individuals [0.42 (0.19, 0.77) vs. 0.67 (0.33, 0.94); P < 0.001].
ORs per SD of vascular function variables for multivariable-adjusted associations with atrial fibrillation are provided in Table 2. In the age-adjusted and sex-adjusted model, baseline brachial artery diameter was related to atrial fibrillation with a 19% higher odds (OR 1.19, 95% CI 1.04–1.37, P = 0.012). This association is mainly because of the female subgroup (OR 1.35, 95% CI 1.05–1.74, P = 0.018) as shown in Supplemental Digital Content Table 1, http://links.lww.com/HJH/B39, Supplemental Digital Content Figure 1, http://links.lww.com/HJH/B39. PAT ratio was inversely associated with atrial fibrillation (OR 0.83, 95% CI 0.74–0.94, P = 0.0029) in the age-adjusted and sex-adjusted model.
After adjustment for cardiovascular risk factors and heart rate statistical significance was lost for associations between study participants with and without atrial fibrillation for baseline brachial artery diameter, FMD and PAT ratio as shown in Table 2. The only vascular parameter still significantly differing after risk factor and heart rate adjustment was baseline pulse amplitude. The direction of association changed. Baseline pulse amplitude was related to atrial fibrillation with a 19% lower odds (OR 0.81, 95% CI 0.70–0.93, P = 0.0025). These findings can most likely be explained by negative confounding by sex and/or age. Additional adjustment for cardiac medication did not change the results markedly.
Subgroup analysis of individuals with atrial fibrillation during investigation revealed no difference in vascular function parameters after adjustment for cardiovascular risk factors, as shown in Supplemental Digital Content Table 3, http://links.lww.com/HJH/B39, Supplemental Digital Content Figure 3, http://links.lww.com/HJH/B39.
No improvement in the area under the curve beyond traditional risk factors combined in the Framingham Risk Score was observed. The corrected area under the curve reached 0.79 for cardiovascular risk factors alone (95% CI 0.773–0.813) as well as in the respective combination with the baseline brachial artery diameter (95% CI 0.769–0.811), baseline pulse amplitude (95% CI 0.774–0.818) or PAT ratio (95% CI 0.771–0.818) and 0.80 in combination with the FMD (95% CI 0.769–0.818).
In the community sample of the Gutenberg Health Study impaired vascular function with a higher baseline pulse amplitude, a higher baseline brachial artery diameter and a decreased FMD as well as PAT ratio was observed in participants with atrial fibrillation. These associations were driven by classical cardiovascular risk factors as statistical significance was lost after covariate adjustment. No improvement in discrimination for atrial fibrillation prediction was achieved.
Our results are in slight contrast to previous studies correlating atrial fibrillation with impaired vascular reactivity [25–27]. Whereas in our sample, we observed a correlation between reduced vascular function and atrial fibrillation, the association was no longer statistically significant after risk factor adjustment. In a prior study, individuals with persistent atrial fibrillation showed clearly reduced FMD, even in individuals without accompanying cardiovascular risk factors . Furthermore, a lower FMD was related to an increased risk of incident atrial fibrillation in addition to classical cardiovascular risk factors . Differences in study design and patient characteristics may account for different conclusions. Prior studies included patients with persistent atrial fibrillation  whereas in our study all types of atrial fibrillation in the community were used. FMD may be impaired more significantly in patients with long-lasting arrhythmia. However, restricting our analyses to individuals with atrial fibrillation on the study ECG, and thus a presumably higher atrial fibrillation burden did not change the results markedly.
It has been suggested that irregular stroke volume and consequently disturbed pulsatile blood flow as induced by atrial fibrillation may cause endothelial dysfunction . Whereas laminar shear stress on the vascular wall is required for regular nitric oxide production, a beat-to-beat variation in flow dynamics in atrial fibrillation has less favorable effects on the vascular wall, changes rheology, and reduces vascular responsiveness . In addition, atrial dysfunction in atrial fibrillation is known to create a pro-inflammatory state  and enhance oxidative stress [31,32], which both impact vascular homeostasis [33,34]. In manifest atrial fibrillation, higher levels of biomarkers of endothelial dysfunction have been reported [35,36]. Furthermore, atrial fibrillation is characterized by enhanced activity of the renin–angiotensin–aldosterone system and accompanied by elevated neurohumoral activation mirrored by natriuretic peptide concentration  that are also related to vascular function . Appropriately, restoration of sinus rhythm after cardioversion improves FMD . Similarly, catheter ablation for atrial fibrillation enhanced vascular reactivity measured by FMD [40,41].
However, in our cohort, no strong associations between vascular reactivity and atrial fibrillation can be found. Adjustment for classical risk factors, even without accounting for often co-existent cardiac dysfunction that is strongly related to vascular impairment [42–45], resulted in a loss of statistical significance of the association. Therefore, our cross-sectional associations are primarily explained by age and the underlying risk factor burden. In line with our observations, endothelial dysfunction was more pronounced in individuals with atrial fibrillation and comorbidities than in lone atrial fibrillation . Moreover, advantage of cardioversion on endothelial function correlates inversely with underlying comorbidities such as diabetes and hypertension . In the Gutenberg Health Study, our analysis of risk factors revealed that age is closely correlated with peripheral vascular function, but also with the lifetime disease atrial fibrillation. We, therefore, assume major confounding of our associations of vascular function with atrial fibrillation by age. Furthermore, as a risk factor example, hypertension is one of the strongest correlates of vascular dysfunction . It also ranges among the common risk factors for atrial fibrillation . Longstanding blood pressure elevation increases the risk for vascular stiffening and dysfunction. Vascular stiffness reflected by pulse pressure has also been identified as a major predictor of atrial fibrillation . These close pathophysiological relations show the complexity of the associations and may help to understand possible cross-correlations and confounding. A first smaller examination in a community cohort demonstrated that smaller FMD values may be associated with higher incidence rates of atrial fibrillation . Again, it cannot be excluded that vascular function may only be a surrogate for risk factor burden and comorbidities.
Limitations and strengths
In our cross-sectional data, we cannot draw conclusions on the causal pathways as we cannot exclude reverse causation of the impact of atrial fibrillation on vascular dysfunction. Due to variations of cardiac cycle length in atrial fibrillation, measurement of brachial artery diameter may be less precise. To avoid inaccurate results, measurements were repeated semi-automatically three times over three cycles and previous reports suggest sufficient reliability in ongoing atrial fibrillation . Another limitation is the lack of assessment of endothelium-independent response as application of medication to healthy individuals in our cohort was not permitted. Furthermore, we may have missed small, pathophysiologically relevant associations. However, in our large data set, we had good power to show moderate-to-strong effect sizes that would be of clinical relevance. Prospective studies need to be designed to understand the temporal relations of vascular dysfunction measured by FMD and PAT in relation to incident atrial fibrillation.
Among the strengths of the study are the large population-based sample with availability of standardized noninvasive vascular function measurements. FMD and PAT ratio were generated following standard operating procedures based on technical considerations by Flammer et al.. Further, we had the opportunity to perform careful adjustment for known confounders.
In conclusion, our study confirms impaired peripheral vascular function measured by commonly used noninvasive techniques, FMD and PAT, in atrial fibrillation individuals. The associations can be attributed to age and the underlying cardiovascular risk factor burden. Vascular dysfunction, thus, may reflect general morbidity in atrial fibrillation. On the basis of our data, FMD and PAT ratio do not appear to provide additional benefit in risk assessment for atrial fibrillation in clinical practice.
We thank the participants and dedicated study staff of the Gutenberg Health Study for their generous contribution of time and efforts.
Sources of funding: The Gutenberg Health Study is funded through the government of Rhineland-Palatinate (‘Stiftung Rheinland-Pfalz für Innovation,’ contract AZ 961-386261/733), the research programs ‘Wissen schafft Zukunft’ and ‘Center for Translational Vascular Biology (CTVB)’ of the Johannes Gutenberg-University of Mainz, and its contract with Boehringer Ingelheim and PHILIPS Medical Systems, including an unrestricted grant for the Gutenberg Health Study. P.S.W. and J.H.P. are funded by the Federal Ministry of Education and Research (BMBF 01EO1503). This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement No. 648131). This work was performed in the context of the Junior Research Alliance symAtrial project funded by the Federal Ministry of Education and Research (BMBF 01ZX1408A) e:Med – Systems Medicine program (R.B.S., T.Z.). This work was funded by Deutsche Forschungsgemeinschaft (German Research Foundation) Emmy Noether Program SCHN 1149/3-1 (R.B.S.). T.Z., S.B., P.S.W. and T.M. are principal investigators of the German Center for Cardiovascular Research (DZHK).
Conflicts of interest
There are no conflicts of interest.
1. Miyasaka Y, Barnes ME, Gersh BJ, Cha SS, Bailey KR, Abhayaratna WP, et al. Secular trends in incidence of atrial fibrillation
in Olmsted County, Minnesota, 1980 to 2000, and implications on the projections for future prevalence. Circulation
2. Stefansdottir H, Aspelund T, Gudnason V, Arnar DO. Trends in the incidence and prevalence of atrial fibrillation
in Iceland and future projections. Europace
3. Wang TJ, Larson MG, Levy D, Vasan RS, Leip EP, Wolf PA, et al. Temporal relations of atrial fibrillation
and congestive heart failure and their joint influence on mortality: the Framingham heart study. Circulation
4. Shinbane JS, Wood MA, Jensen DN, Ellenbogen KA, Fitzpatrick AP, Scheinman MM. Tachycardia-induced cardiomyopathy: a review of animal models and clinical studies. J Am Coll Cardiol
5. Friberg J, Buch P, Scharling H, Gadsbphioll N, Jensen GB. Rising rates of hospital admissions for atrial fibrillation
6. Stewart S, Hart CL, Hole DJ, McMurray JJ. Population prevalence, incidence, and predictors of atrial fibrillation
in the Renfrew/Paisley study. Heart
7. Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation
on the risk of death: the Framingham Heart Study. Circulation
8. Mitchell GF, Vasan RS, Keyes MJ, Parise H, Wang TJ, Larson MG, et al. Pulse pressure and risk of new-onset atrial fibrillation
9. Heeringa J. Subclinical atherosclerosis and risk of atrial fibrillation
. Arch Intern Med
10. Widlansky ME, Gokce N, Keaney JF, Vita JA. The clinical implications of endothelial dysfunction. J Am Coll Cardiol
11. Donnell CJO, Elosua R. Cardiovascular risk factors. Insights from Framingham Heart. Rev Española Cardiol
12. Münzel T, Sinning C, Post F, Warnholtz A, Schulz E. Pathophysiology, diagnosis and prognostic implications of endothelial dysfunction. Ann Med
13. Fumagalli S, Giannini I, Pupo S, Agostini F, Boni S, Roberts AT, et al. Atrial fibrillation
after electrical cardioversion in elderly patients: a role for arterial stiffness? Results from a preliminary study. Aging Clin Exp Res
14. Chen LY, Leening MJG, Norby FL, Roetker NS, Hofman A, Franco OH, et al. Carotid intima-media thickness and arterial stiffness and the risk of atrial fibrillation
: the Atherosclerosis Risk in Communities (ARIC) Study, Multi-Ethnic Study of Atherosclerosis (MESA), and the Rotterdam Study. J Am Heart Assoc
15. Lau DH, Middeldorp ME, Brooks AG, Ganesan AN, Roberts-Thomson KC, Stiles MK, et al. Aortic stiffness in lone atrial fibrillation
: a novel risk factor for arrhythmia recurrence. PLoS One
16. Cremer A, Lainé M, Papaioannou G, Yeim S, Gosse P. Increased arterial stiffness is an independent predictor of atrial fibrillation
in hypertensive patients. J Hypertens
17. Yeboah J, Folsom AR, Burke GL, Johnson C, Polak JF, Post W, et al. Predictive value of brachial flow-mediated dilation
for incident cardiovascular events in a population-based study
: the Multiethnic Study of Atherosclerosis. Circulation
18. Bonetti PO, Pumper GM, Higano ST, Holmes DR, Kuvin JT, Lerman A. Noninvasive identification of patients with early coronary atherosclerosis by assessment of digital reactive hyperemia. J Am Coll Cardiol
19. Schnabel RB, Schulz A, Wild PS, Sinning CR, Wilde S, Eleftheriadis M, et al. Noninvasive vascular function
measurement in the community cross-sectional relations and comparison of methods. Circ Cardiovasc Imaging
20. Camm AJ, Kirchhof P, Lip GYH, Schotten U, Savelieva I, Ernst S, et al. Guidelines for the management of atrial fibrillation
: the Task Force for the Management of Atrial Fibrillation
of the European Society of Cardiology (ESC). Europace
21. Corretti MC, Anderson TJ, Benjamin EJ, Celermajer D, Charbonneau F, Creager MA, et al. International Brachial Artery Reactivity Task Force. Guidelines for the ultrasound assessment of endothelial-dependent flow-mediated vasodilation of the brachial artery: a report of the international brachial artery reactivity task force. J Am Coll Cardiol
22. Selamet Tierney ES, Newburger JW, Gauvreau K, Geva J, Coogan E, Colan SD, et al. Endothelial pulse amplitude testing: feasibility and reproducibility in adolescents. J Pediatr
23. Atkinson G, Batterham AM, Thijssen DHJ, Green DJ. A new approach to improve the specificity of flow-mediated dilation
for indicating endothelial function in cardiovascular research. J Hypertens
24. Schnabel RB, Aspelund T, Li G, Sullivan LM, Suchy-Dicey A, Harris TB, et al. Validation of an atrial fibrillation
risk algorithm in whites and African Americans. Arch Intern Med
25. O’Neal WT, Efird JT, Yeboah J, Nazarian S, Alonso A, Heckbert SR, et al. Brachial flow-mediated dilation
and incident atrial fibrillation
the multiethnic study of atherosclerosis. Arterioscler Thromb Vasc Biol
26. Polovina M, Potpara T, Giga V, Stepanovic J, Ostojic M. Impaired endothelial function in lone atrial fibrillation
. Vojn Pregl
27. Shaikh AY, Wang N, Yin X, Larson MG, Vasan RS, Hamburg NM, et al. Relations of arterial stiffness and brachial flow-mediated dilation
with new-onset atrial fibrillation
28. Takahashi N, Ishibashi Y, Shimada T, Sakane T, Ohata S, Sugamori T, et al. Atrial fibrillation
impairs endothelial function of forearm vessels in humans. J Card Fail
29. Cai H, Li Z, Goette A, Mera F, Honeycutt C, Feterik K, et al. Downregulation of endocardial nitric oxide synthase expression and nitric oxide production in atrial fibrillation
: Potential mechanisms for atrial thrombosis and stroke. Circulation
30. Engelmann MDM, Svendsen JH. Inflammation in the genesis and perpetuation of atrial fibrillation
. Eur Heart J
31. Korantzopoulos P, Kolettis TM, Galaris D, Goudevenos JA. The role of oxidative stress in the pathogenesis and perpetuation of atrial fibrillation
. Int J Cardiol
32. Dudley SC, Hoch NE, McCann LA, Honeycutt C, Diamandopoulos L, Fukai T, et al. Atrial fibrillation
increases production of superoxide by the left atrium and left atrial appendage: role of the NADPH and xanthine oxidases. Circulation
33. Vita JA, Keaney JF, Larson MG, Keyes MJ, Massaro JM, Lipinska I, et al. Brachial artery vasodilator function and systemic inflammation in the framingham offspring study. Circulation
34. Heitzer T, Schlinzig T, Krohn K, Meinertz T, Munzel T. Endothelial dysfunction, oxidative stress, and risk of cardiovascular events in patients with coronary artery disease. Circulation
35. Ehrlich JR, Kaluzny M, Baumann S, Lehmann R, Hohnloser SH. Biomarkers of structural remodelling and endothelial dysfunction for prediction of cardiovascular events or death in patients with atrial fibrillation
. Clin Res Cardiol
36. Schnabel RB, Wild PS, Wilde S, Ojeda FM, Schulz A, Zeller T, et al. Multiple biomarkers and atrial fibrillation
in the general population. PLoS One
37. Dixen U, Ravn L, Soeby-Rasmussen C, Paulsen AW, Parner J, Frandsen E, et al. Raised plasma aldosterone and natriuretic peptides in atrial fibrillation
38. Kathiresan S, Gona P, Larson MG, Vita JA, Mitchell GF, Tofler GH, et al. Cross-sectional relations of multiple biomarkers from distinct biological pathways to brachial artery endothelial function. Circulation
39. Skalidis EI, Zacharis Ea, Tsetis DK, Pagonidis K, Chlouverakis G, Yarmenitis S, et al. Endothelial cell function during atrial fibrillation
and after restoration of sinus rhythm. Am J Cardiol
40. Shin SY, Na JO, Lim HE, Choi CU, Choi JIl, Kim SH, et al. Improved endothelial function in patients with atrial fibrillation
through maintenance of sinus rhythm by successful catheter ablation. J Cardiovasc Electrophysiol
41. Yoshino S, Yoshikawa A, Hamasaki S, Ishida S, Oketani N, Saihara K, et al. Atrial fibrillation
-induced endothelial dysfunction improves after restoration of sinus rhythm. Int J Cardiol
42. Katz SD, Biasucci L, Sabba C, Strom JA, Jondeau G, Galvao M, et al. Impaired endothelium-mediated vasodilation in the peripheral vasculature of patients with congestive heart failure. J Am Coll Cardiol
43. Kubo SH, Rector TS, Bank AJ, Williams RE, Heifetz SM. Endothelium-dependent vasodilation is attenuated in patients with heart failure. Circulation
44. Lee JF, Barrett-O’Keefe Z, Garten RS, Nelson AD, Ryan JJ, Nativi JN, et al. Evidence of microvascular dysfunction in heart failure with preserved ejection fraction. Heart
45. Maréchaux S, Samson R, van Belle E, Breyne J, de Monte J, Dédrie C, et al. Vascular and microvascular endothelial function in heart failure with preserved ejection fraction. J Card Fail
46. Guazzi M, Berti M, Belletti S, Reina G, Guazzi MD. Exercise metaboreflex activation and endothelial function impairment in atrial fibrillation
. Am J Physiol Circ Physiol
47. Guazzi M, Belletti S, Bianco E, Lenatti L, Guazzi MD. Endothelial dysfunction and exercise performance in lone atrial fibrillation
or associated with hypertension or diabetes: different results with cardioversion. Am J Physiol Circ Physiol
48. Alonso A, Krijthe BP, Aspelund T, Stepas KA, Pencina MJ, Moser CB, et al. Simple risk model predicts incidence of atrial fibrillation
in a racially and geographically diverse population: the CHARGE-atrial fibrillation
Consortium. J Am Heart Assoc
49. Flammer AJ, Anderson T, Celermajer DS, Mark A, Deanfield J, Ganz P, et al. The assessment of endothelial function - from research into clinical practice. Circulation
* DZHK (German Center for Cardiovascular Research), partner site Hamburg/Kiel/Luebeck, Germany.
† DZHK (German Center for Cardiovascular Research), partner site Rhine-Main, Mainz, Germany.
‡ Thomas Münzel and Renate B. Schnabel contributed equally to the writing of the article.