Non-Invasive Pulmonary Capillary Wedge Pressure Assessment on Speckle Tracking Echocardiography as a Predictor of New-Onset Non-Valvular Atrial Fibrillation　― Four-Year Prospective Study (NIPAF Study) ―

Abstract

Background: Although new-onset atrial fibrillation (AF) increases with ageing, the prediction of new-onset AF is complicated. We previously reported that pulmonary capillary wedge pressure (ePCWP) estimated by the combination of left atrial volume index (LAVI) and active left atrial emptying function (aLAEF) had a strong relationship with PCWP on catheterization (r=0.92): ePCWP=10.8−12.4×log (aLAEF/minimum LAVI). We sought to determine the usefulness of ePCWP to predict new-onset AF.

Methods and Results: We measured LAVI, aLAEF and ePCWP on speckle tracking echocardiography (STE) in 566 consecutive elderly patients (72±6 years) without a history of AF. A total of 63 patients (73±6 years) developed electrocardiographically confirmed AF during a mean follow-up period of 50 months. Baseline aLAEF was significantly lower in patients with than without new-onset AF (17.9±6.5 vs. 28.2±7.5%), whereas ePCWP was significantly higher (14.8±3.7 vs. 10.3±3.1 mmHg). In multivariate logistic regression analysis, ePCWP and aLAEF were strong independent predictors of AF. Using ePCWP >13 mmHg or aLAEF ≤22% on univariate Cox regression analysis, the HR for new-onset AF were 3.53 (95% CI: 1.68–7.44, P<0.001) and 4.06 (95% CI: 1.90–8.65, P<0.001), respectively. By combining these 2 criteria (>13 mmHg and ≤22%), the HR increased to 11.84 (95% CI: 6.85–20.5, P<0.001).

Conclusions: ePCWP and aLAEF measured on STE are useful predictors of new-onset AF. ePCWP provides added value for risk stratification of new-onset AF.

Atrial fibrillation (AF) is associated with an increased risk of cerebral infarction and is the most common sustained cardiac arrhythmia.^1–3 Risk stratification for the development of AF has a major public health impact. Pulmonary capillary wedge pressure (PCWP) or left ventricular (LV) filling pressure is useful for the stratification of LV diastolic dysfunction that can lead to AF.⁴ There have been few prospective studies, however, on whether PCWP predicts the risk of new-onset AF because the measurement of PCWP requires an invasive method. Recently, we developed a novel index (kinetics-tracking index; KT index) to estimate PCWP based on the combination of active left atrial emptying function (LAEF) and minimum LA volume index (LAVI) obtained from the time-LAV curve using speckle tracking echocardiography (STE).⁵ The PCWP estimated with the KT index (ePCWP) had a strong correlation with PCWP obtained on right heart catheterization (r=0.92) in patients with normal sinus rhythm.⁵ The aim of the present study was to evaluate the usefulness of ePCWP and its optimal cut-off to predict new-onset AF.

Methods

Patients and Study Protocol

This was a prospective multicenter study that determined the ability of Non-Invasive ePCWP assessed on STE to Predict new-onset non-valvular Atrial Fibrillation (NIPAF study). We performed echocardiography in 675 consecutive elderly subjects (≥65 years of age) without a history of AF to screen for cardiac disease in patients with cardiac murmur, palpitation, chest pain, chest discomfort or electrocardiogram (ECG)-documented arrhythmia. Echocardiography was also performed in patients who were diagnosed with coronary artery disease (CAD), hypertension or valvular heart disease to evaluate the severity of disease. We included patients on medication for dyslipidemia, CAD, hypertension, type 2 diabetes mellitus (T2DM), mild-moderate mitral regurgitation or other valvular heart disease, heart failure or arrhythmia other than atrial arrhythmia, and patients with ECG abnormalities such as bundle branch block, non-specific ST-T change or first- or second-degree atrioventricular block that did not require specific medication. We excluded patients with mitral stenosis; severe mitral regurgitation; previous mitral valve surgery; permanent cardiac pacemaker implantation; and atrial arrhythmia such as AF, atrial flutter or atrial tachycardia on ECG or Holter ECG. Medical therapy and risk factors for atherosclerotic disease (medication dependent only) were evaluated in each patient, including T2DM, dyslipidemia, hypertension and current smoking. Monthly ECG was carried out for all patients regardless of symptoms of arrhythmia during the follow-up period. Both ECG and Holter ECG were performed to confirm the new-onset AF, including paroxysmal AF, in all patients who had symptoms of arrhythmia and/or palpitation. Final confirmation of AF was done on Holter ECG. We defined the time point of new-onset AF as the month when Holter ECG confirmed the presence of AF. The present study was approved by the institution ethics committee, and informed consent was obtained from all patients before enrollment.

Echocardiography

Echocardiography was performed using an ACUSON sequoia 512 ultrasound system (Siemens, Mountain View, CA, USA) with a 4V1c ultrasound transducer (1.5–4.25 MHz) to evaluate LA function. LAV was measured during a single cardiac cycle on velocity vector imaging (VVI) using on-line software (Syngo Velocity Vector Imaging; Siemens).

This software allowed editing of the initial trace, and the endocardial velocity was derived as the ratio between frame-to-frame displacement and the time interval.⁵ First, we visually identified the endocardial border of the LA, and then manually outlined the border. Manual tracing of an endocardial border over 1 frame was automatically tracked throughout the cardiac cycle with this system. The velocity vectors in the 2-D plane were displayed throughout the cardiac cycle, representing displacement of the speckles in relation to each other along the endocardial contour of the LA (Figure 1A), and a time-LAV curve and strain can be promptly provided (Figure 1B) as well as a previous study.⁶ Maximum LAV, minimum LAV and LAV just before atrial contraction were obtained from the apical 4-chamber view with a frame rate of 55–60 frames/s using Simpson’s method. Total LAEF (reservoir function), passive LAEF (conduit function) and active LAEF (booster pump function) were calculated to evaluate phasic LA function. Total, passive and active LAEF were defined during a cardiac cycle as (maximum LAV−minimum LAV)/maximum LAV×100%; (maximum LAV−pre-atrial contraction LAV)/maximum LAV×100%; and (pre-atrial contraction LAV−minimum LAV)/pre-atrial contraction LAV×100%, respectively. The following conventional echocardiographic parameters were also measured according to standard methods of echocardiography:⁷ LA dimension (LAD); LV ejection fraction (LVEF); LV mass; and the ratio of early diastolic transmitral inflow velocity to annular tissue velocity (E/e’). Pre-atrial contraction index (LAVI), maximum LAVI and minimum LAVI were indexed to body surface area. LA peak strain was calculated as the average of LA mid-septal and mid-lateral wall strain.

Figure 1.

(A) Representative velocity vector imaging (VVI) in the patients who did not develop new-onset atrial fibrillation (AF). (B) Time-left atrial volume (time-LAV) curve (orange) constructed using VVI. Blue line, time-dV/dt curve; white line below the time-LAV curve, electrocardiogram. AC, atrial contraction; ePCWP, estimated pulmonary capillary wedge pressure; LAEF, left atrial emptying function.

Reliability and Reproducibility of LAV and LAEF on VVI

In our previous study, the interobserver correlation coefficients for maximum LAV, minimum LAV and total LAEF measured on VVI were 0.98, 0.99 and 0.99, respectively; the relative interobserver differences in these 3 parameters were 0.94±6.8%, 8.0±11.0%, and 2.2±15.9%, respectively.⁸ The intraobserver correlation coefficients for maximum LAV, minimum LAV and total LAEF on VVI were 0.98, 0.98 and 0.90, respectively; the relative intraobserver differences in these 3 parameters were 2.6±6.2%, 8.1±11.6%, and 4.0±12.4%, respectively.⁸ There were significant correlations between VVI and enhanced computed tomography for the evaluation of maximum LAV (r=0.95, P<0.001), minimum LAV (r=0.96, P<0.001) and total LAEF (r=0.97, P<0.001).⁸

Statistical Analysis

The normality of data distributions was tested using Kolmogorov-Smirnov test. The data are expressed as mean±SD. Categorical data are presented as the number of patients and were compared using chi-squared test. The differences in continuous variables were evaluated using unpaired Student’s t-test. We performed multivariate logistic analysis using 4 models to identify the independent predictors of new-onset AF. Model 1 included only clinical variables with P<0.001 on univariate analysis. Model 2 included clinical and conventional echocardiographic variables. Models 3 and 4 included clinical, conventional and STE variables. Receiver operating characteristic (ROC) curve analysis was performed to determine the optimal cut-offs of several echocardiographic parameters to predict new-onset AF; the sensitivity and specificity, positive predictive value (PPV) and negative predictive value using these cut-offs were calculated. Differences in area under the curve (AUC) in the echocardiographic parameters were tested using the method established by Hanley and McNeil.⁹ Survival curves were plotted using the Kaplan-Meier method and hazard ratios were calculated on univariate Cox regression analysis to compare 2 survival curves. All statistical analysis was performed using Stat View version 5.0 (SAS Institute, Cary, NC, USA) and StatFlex version 6.4 (Artech Co, Osaka, Japan). P<0.05 was considered significant.

Results

Subject Characteristics

Of the 675 subjects, 63 withdrew from follow-up due to the inability to walk, 12 withdrew due to bone fracture and 22 withdrew due to dementia. We included 566 subjects in the final analysis after excluding 12 more because of poor echocardiographic recordings due to severe obesity or emphysema. Baseline patient clinical characteristics are listed in Table 1. During a mean follow-up period of 50 months (median, 48 months; IQR, 46–60 months), 63 subjects (11.1%) developed ECG-confirmed new-onset AF. There were no significant differences between the 2 groups (AF and non-AF) in age because we included only elderly subjects (≥65 years of age). The frequency of hypertension was significantly higher in the AF group than in the non-AF group. There were no significant differences between the AF group and the non-AF group in concomitant medication use except for β-blockers and calcium channel blockers. LV mass, E/e’, LAD and LAVI were significantly greater in the AF group than in the non-AF group, despite the lack of difference in LVEF.

Table 1. Baseline Patient Characteristics

	Non-AF group (n=503)	New-onset AF group (n=63)	P-value
Clinical characteristics
Men	257 (51.0)	38 (60.3)	0.18
Age (years)	71.4±5.8	73.0±6.2	0.058
Symptom or findings at enrollment
Chest discomfort	36 (7.2)	5 (8.0)	0.80
Chest pain	21 (4.2)	3 (4.7)	0.74
Palpitations	23 (4.6)	4 (6.3)	0.53
Cardiac murmurs	30 (6.0)	4 (6.3)	0.78
ECG-documented arrhythmia	18 (3.6)	3 (3.6)	0.72
SBP (mmHg)	129±8	133±9	<0.001
Hypertension	303 (60.2)	49 (77.8)	0.008
Diabetes mellitus	41 (8.2)	6 (9.5)	0.63
Dyslipidemia	161 (32.0)	14 (22.2)	0.15
Coronary artery disease	79 (15.7)	12 (19.0)	0.47
Valvular heart disease	50 (9.9)	8 (12.7)	0.51
Chronic heart failure	29 (5.8)	6 (9.5)	0.26
BBB or first- or second-degree AVB	41 (8.2)	8 (12.7)	0.23
Arrhythmia except atrial arrhythmia	22 (4.4)	2 (3.2)	>0.99
Current smoking	82 (16.3)	10 (15.9)	>0.99
Medication
ARB or ACEI	136 (27.0)	20 (31.7)	0.46
β-blockers	56 (11.1)	14 (22.2)	0.023
Calcium channel blockers	243 (48.3)	39 (61.9)	0.046
Dihydropyridine	203 (40.3)	34 (54.0)	0.040
Benzothiazepine	40 (8.0)	5 (7.9)	>0.99
Phenylalkylamine	0 (0)	0 (0)	>0.99
Diuretics	44 (8.7)	6 (9.5)	0.81
Ventricular anti-arrhythmia medication	11 (2.2)	2 (3.2)	0.65
Statins	178 (35.4)	11 (17.5)	0.004
Conventional echocardiography
LVMI (g/m2)	134±30	143±37	<0.001
LVEF (%)	65±8	65±7	0.54
E/e’	8.5±1.7	9.6±1.8	<0.001
LAD (mm)	40.3±6.0	44.2±5.8	<0.001
Speckle tracking echocardiography
ePCWP (mmHg)	10.3±3.1	14.8±3.7	<0.001
Maximum LAVI (mL/m2)	47.2±15.4	56.9±15.4	<0.001
Pre-atrial contraction LAVI (mL/m2)	38.0±14.2	45.7±13.9	<0.001
Minimum LAVI (mL/m2)	27.4±11.2	37.5±12.9	<0.001
LA peak strain	24.7±7.6	19.2±5.9	<0.001
Total LAEF (%)	43.4±7.5	35.1±8.1	<0.001
Passive LAEF (%)	20.6±8.3	19.5±6.1	0.33
Active LAEF (%)	28.2±7.5	17.9±6.5	<0.001

Data given as n (%) or mean±SD. ACEI, angiotensin-converting enzyme inhibitor; AF, atrial fibrillation; ARB, angiotensin II receptor blocker; AVB, atrioventricular block; BBB, bundle branch block; E/e’, ratio of early diastolic transmitral inflow velocity to annular tissue velocity; ECG, electrocardiogram; ePCWP, estimated pulmonary capillary wedge pressure; LA, left atrial; LAD, left atrial dimension; LAEF, left atrial emptying function; LAVI, left atrial volume index; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; SBP, systolic blood pressure.

Predictors of New-Onset AF

We performed multivariate logistic regression analysis for parameters with P<0.001 on unpaired Student’s t-test or chi-squared test. In model 1, we excluded the history of hypertension from the analysis because this is related to systolic blood pressure (SBP). In model 1, SBP was a positive predictor and statin use was a negative predictor of new-onset AF (Table 2). In model 2, we excluded parameters that were not significant in model 1 (β-blockers and calcium channel blockers) from the analysis. In model 2, SBP, statin use, LAD and E/e’ were independent predictors of new-onset AF. In model 3, we excluded the parameter that was not significant in model 2 (LV mass index) from the analysis. We also excluded LAD because this is related to the LAV used in model 3 (minimum LAVI and maximum LAVI). In model 3, SBP, statin use and ePCWP were independent predictors of new-onset AF. We did not include both ePCWP and active LAEF in any of these models due to multicollinearity because there was a strong correlation between ePCWP and active LAEF (r=−0.81, P<0.001). Therefore, active LAEF instead of ePCWP was included in model 4. In model 4, SBP, statin use and active LAEF were independent predictors of new-onset AF.

Table 2. Multivariate Age-Adjusted Independent Predictors of New-Onset AF

Variables	OR	95% CI	P-value
Model 1
SBP (mmHg)	1.054	1.010–1.101	0.015
Statin use	0.379	0.192–0.749	0.005
β-blockers	1.923	0.976–3.790	0.059
Calcium channel blockers	1.303	0.728–2.331	0.37
Model 2
SBP (mmHg)	1.054	1.010–1.101	0.015
Statin use	0.379	0.192–0.749	0.005
LAD (mm)	1.088	1.032–1.146	0.002
E/e’	1.303	1.106–1.535	0.002
LVMI (g/m2)	0.997	0.987–1.007	0.56
Model 3
SBP (mmHg)	1.058	1.008–1.111	0.023
Statin use	0.335	0.148–0.758	0.009
LAD (mm)	1.042	0.973–1.114	0.23
E/e’	1.222	0.984–1.516	0.069
Minimum LAVI (mL/m2)	0.814	0.658–1.007	0.058
Maximum LAVI (mL/m2)	1.001	0.871–1.151	0.98
Pre-atrial contraction LAVI (mL/m2)	1.019	0.857–1.211	0.83
Total LAEF (%)	1.002	0.885–1.134	0.98
ePCWP (mmHg)	2.804	1.839–4.274	<0.001
Model 4
SBP (mmHg)	1.053	1.005–1.104	0.030
Statin use	0.400	0.183–0.864	0.020
LAD (mm)	1.058	0.992–1.129	0.085
E/e’	1.159	0.944–1.438	0.15
Minimum LAVI (mL/m2)	0.961	0.800–1.156	0.68
Maximum LAVI (mL/m2)	1.082	0.961–1.217	0.19
Pre-atrial contraction LAVI (mL/m2)	0.926	0.780–1.099	0.38
Total LAEF (%)	0.930	0.836–1.034	0.18
Active LAEF (%)	0.817	0.731–0.914	<0.001

Abbreviations as in Table 1.

ROC Curve and Kaplan-Meier Analysis

We performed ROC curve analysis to identify the optimal cut-offs of ePCWP, active LAEF and E/e’ to predict AF (Figure 2). The AUC for ePCWP (0.84) and active LAEF (0.85) were better than that for E/e’ (0.67, P<0.001). There was no significant difference, however, between the AUC for ePCWP and active LAEF (P=0.64). Table 3 presents the predictive accuracy for new-onset AF at the optimal cut-offs based on ROC curve analysis.

Figure 2.

Receiver operating characteristic curves for the prediction of new-onset atrial fibrillation. AUC, area under the curve; E/e’, ratio of early diastolic transmitral inflow velocity to annular tissue velocity; ePCWP, estimated pulmonary capillary wedge pressure; LAEF, left atrial emptying function.

Table 3. Prediction of New-Onset AF: Accuracy of Optimal Echocardiography Cut-Offs

	Sensitivity % (95% CI)	Specificity % (95% CI)	PPV % (95% CI)	NPV % (95% CI)
ePCWP (>13 mmHg)	76 (85–91)	81 (75–81)	33 (16–22)	96 (98–100)
Active LAEF (≤22%)	79 (85–91)	77 (78–84)	30 (18–24)	97 (98–100)
E/e’ (>8)	83 (96–98)	51 (56–64)	18 (9–15)	51 (98–100)
ePCWP (>13 mmHg) plus active LAEF (≤22%)	71 (96–98)	86 (56–64)	39 (9–15)	86 (98–100)

NPV, negative predictive value; PPV, positive predictive value. Other abbreviations as in Table 1.

The difference in AF-free survival (including hazard ratios) after stratification by ePCWP (≤13 mmHg vs. >13 mmHg), active LAEF (≤22% vs. >22%), E/e’ (≤8 vs. >8) or the combination of ePCWP and active LAEF is shown on Kaplan-Meier curves (Figure 3). Using the cut-offs of ePCWP >13 mmHg and active LAEF ≤22%, the hazard ratios for new-onset AF over 5 years were 3.53 (95% CI: 1.68–7.44, P<0.001) and 4.06 (95% CI: 1.90–8.65, P<0.001), respectively. By combining these criteria (>13 mmHg and ≤22%), the hazard ratio increased to 11.84 (95% CI: 6.85–20.5, P<0.001).

Figure 3.

Kaplan-Meier curves (including hazard ratios) of cumulative survival free from atrial fibrillation (AF) according to (A) estimated pulmonary capillary wedge pressure (ePCWP) >13 mmHg and ≤13 mmHg, (B) active left atrial emptying function (LAEF) ≤22% and >22%, (C) ratio of early diastolic transmitral inflow velocity to annular tissue velocity (E/e’) >8 and ≤8, and (D) ePCWP >13 mmHg plus active LAEF ≤22% combined, and ePCWP ≤13 or active LAEF >22%.

Discussion

We examined STE parameters in 566 elderly subjects to prospectively follow for new-onset AF. We showed that baseline ePCWP was increased and active LAEF was decreased in the AF group compared with the non-AF group. We also showed that increased ePCWP (>13 mmHg) and reduced active LAEF (≤22%) were the most useful predictors of new-onset AF.

Clinical Usefulness of ePCWP

We previously reported that PCWP can be non-invasively estimated using the KT index, which is defined as log₁₀ [active LAEF/minimum LAVI].⁵ ePCWP had a strong correlation with PCWP obtained on right heart catheterization in patients with normal sinus rhythm (r=0.92),⁵ in those with AF (r=0.77),⁵ and in those with mitral regurgitation (r=0.70).¹⁰ The reliability and clinical usefulness of ePCWP have been demonstrated in many studies. There have been no previous studies on PCWP in healthy subjects in a relatively large population, because the measurement of PCWP requires an invasive method. We measured ePCWP in healthy subjects and reported that ePCWP was maintained at 8.3±1.8 mmHg in male subjects and 8.2±2.3 mmHg in female subjects due to compensation by an increase in active LAEF with advancing age.¹¹ We also demonstrated that ePCWP was independently associated with hypertensive heart failure and may be useful to non-invasively detect the early stage of hypertensive heart failure.¹²

Echocardiographic Parameters to Predict AF

LA pressure and volume overload cause LA structural and functional remodeling. LV diastolic dysfunction with aging is caused by a progressive deterioration in LV filling.^13–15 Therefore, LA enlargement has been proposed as a predictor of AF.^16–18 There have been a few studies that used LA function to predict AF.^19,20 Minimum LAV has been reported an independent predictor of new-onset AF or atrial flutter.²¹ The predictive ability of minimum LAV to predict AF, however, was moderate (AUC, 0.65–0.69). Another study reported that manually measured total LVEF was a risk factor for new-onset AF independent of LAV.¹⁷ Given that the manual measurement of active LAEF was too time consuming, active LAEF could not be measured in that study. In the present study we have shown that active LAEF but not total LAEF is an independent predictor of new-onset AF (AUC, 0.92).

LA wall degeneration evaluated on integrated backscatter using transesophageal echocardiography has been shown to be more advanced in patients with than without paroxysmal AF despite a similar LA diameter.²¹ Taken together, this suggests that active LAEF may have a relationship with LA pathological degeneration.

Clinical Implications

Although guidelines for the management of AF were established by the committees of the European Society of Cardiology, American College of Cardiology and American Heart Association, thromboembolic disease such as cerebral infarction caused by paroxysmal AF or chronic AF results in significant mortality.²² Anticoagulant therapy can reduce thromboembolic disease in patients with AF, but most patients with AF will not develop thromboembolic disease. These patients will be exposed to the risks of bleeding due to anticoagulant therapy.²³ Therefore, useful predictors for new-onset AF were required to stratify the balance between thromboembolic and bleeding risks. The present study suggests that patients with ePCWP >13 mmHg and active LAEF ≤22% should be frequently monitored on Holter ECG. Angiotensin II receptor blockers (ARB) or angiotensin-converting enzyme inhibitors (ACEI) have been reported to delay the progression of paroxysmal AF to chronic AF.^24–26 The present study suggest that it might be possible to delay new-onset AF by identifying the patients with ePCWP >13 mmHg and active LAEF ≤22% and then initiating medication with ARB or ACEI.

We enrolled elderly subjects (≥65 years of age) in the present study because elderly subjects have a higher risk of AF and more elderly patients visit the cardiology department than younger patients. To apply the cut-offs of ePCWP and active LAEF to younger patients, a study involving younger subjects needs to be carried out.

Study Limitations

There are several limitations in the present study. First, we enrolled a relatively small number of patients and performed echocardiography only at baseline in all patients. The study may have been underpowered to determine the relationship between the incidence of new-onset AF and the baseline or follow-up parameters. A study in a larger population is needed to definitively define the relationship between the incidence of new-onset AF and the baseline or follow-up parameters. Second, the number of incident AF cases may have been underestimated because only AF episodes that were confirmed on Holter ECG were considered an endpoint, and some patients with AF had no symptoms. In the present study, 68% of patients developed AF without symptoms. This limitation applies to most previous studies that included AF as an endpoint. In a subanalysis of the ALLHAT and PROSPER studies that investigated the effect of statins on new-onset AF, annual or biannual ECG was recorded to detect new-onset AF.^27,28 To detect new-onset AF precisely, more frequent Holter ECG are required. Third, the PPV of active LAEF and ePCWP for new-onset AF was low (39%) because the overall incidence of AF in primary prevention patients was not high (2.2%/years). It might be necessary to combine other variables to improve the PPV for AF. Finally, VVI is based on manual tracing and represents a 2-D measurement. Three-dimensional STE is the latest technique and is now available to assess LA structure and function, including LAV and strain without any assumptions of LA geometry.²⁹ The optimal cut-offs and their predictive accuracy obtained in the present study need to be reassessed on 3-D STE in the future. Although there was also a good correlation between PCWP and ePCWP in both primary and secondary mitral regurgitation (r=0.70 and r=0.67, P<0.01, respectively),¹¹ patients with severe mitral regurgitation were excluded from the present study.

Conclusions

Baseline ePCWP and active LAEF assessed in sinus rhythm were deteriorated in patients who developed new-onset AF compared with those who did not develop new-onset AF. Non-invasive echocardiography of LA phasic function, particularly ePCWP and active LAEF, may provide added value for risk stratification and be a more useful predictor than any other LA function for new-onset AF.

Acknowledgments

The authors acknowledge the help of Mr. Satoshi Yamajima, Mr. Noriyuki Onishi, and Ms. Tomomi Endo for ultrasound investigation, and Ms. Ritsuko Tanaka and Mr. Kenichiro Ikeda for preparation of the manuscript.

Disclosures

We have no financial or other relations that could lead to conflict of interest.

References

• 1.   Yamauchi T, Sakata Y, Miura M, Onose T, Tsuji K, Abe R, et al. Prognostic impact of atrial fibrillation and new risk score of its onset in patients at high risk of heart failure: A report from the CHARY-2 study. Circ J 2017; 81: 185–194.
• 2.   Kokubo Y, Watanabe M, Higashiyama A, Nakao YM, Kusano K, Miyamoto Y. Development of a basic risk score for incident atrial fibrillation in a Japanese general population: The Suita study. Circ J 2017; 81: 1580–1588.
• 3.   Alonso A, Norby FL. Predicting atrial fibrillation and its complications. Circ J 2016; 80: 1061–1066.
• 4.   Baicu CF, Zile MR, Aurigemma GP, Gaasch WH. Left ventricular systolic performance, function, and contractility in patients with diastolic heart failure. Circulation 2005; 111: 2306–2312.
• 5.   Kawasaki M, Tanaka R, Ono K, Minatoguchi S, Watanabe T, Iwama M, et al. A novel ultrasound predictor of pulmonary capillary wedge pressure assessed by the combination of left atrial volume and function: A speckle tracking echocardiography study. J Cardiol 2015; 6: 253–262.
• 6.   Hayashi T, Yamada S, Iwano H, Nakabachi M, Sakakibara M, Okada K, et al. Left ventricular global strain for estimating relaxation and filling pressure: A multicenter study. Circ J 2016; 80: 1163–1170.
• 7.   Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al; Chamber Quantification Writing Group; American Society of Echocardiography’s Guidelines and Standards Committee; European Association of Echocardiography. Recommendations for chamber quantification: A report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005; 18: 1440–1463.
• 8.   Hirose T, Kawasaki M, Tanaka R, Ono K, Watanabe T, Iwama M, et al. Left atrial function assessed by speckle tracking echocardiography as a predictor of new-onset non-valvular atrial fibrillation: Results from a prospective study in 580 adults. Eur Heart J Cardiovasc Imaging 2012; 13: 243–250.
• 9.   Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 1983; 148: 839–843.
• 10.   Kawase Y, Kawasaki M, Tanaka R, Nomura N, Fujii Y, Ogawa K, et al. Noninvasive estimation of pulmonary capillary wedge pressure in patients with mitral regurgitation: A speckle tracking echocardiography study. J Cardiol 2016; 67: 192–198.
• 11.   Kawasaki M, Tanaka R, Ono K, Minatoguchi S, Watanabe T, Arai M, et al. Impact of gender and healthy aging on pulmonary capillary wedge pressure estimated by the kinetics-tracking index using two-dimensional speckle tracking echocardiography. Hypertens Res 2016; 39: 327–333.
• 12.   Minatoguchi S, Kawasaki M, Tanaka R, Yoshizane T, Ono K, Saeki M, et al. Evaluation of systolic and diastolic properties of hypertensive heart failure using speckle tracking echocardiography with high volume rates. Heart Vessels 2017; 32: 1202–1213.
• 13.   Redfield MM, Jacobsen SJ, Borlaug BA, Rodeheffer RJ, Kass DA. Age- and gender-related ventricular-vascular stiffening: A community-based study. Circulation 2005; 112: 2254–2262.
• 14.   Tamaki K, Koshiyama M, Ohsawa M. Trend in the prevalence of atrial fibrillation during the past 15 years in Iwate (Northeastern area of Japan). Circ J 2017; 81: 1537–1539.
• 15.   Gottdiener JS, Kitzman DW, Aurigemma GP, Arnold AM, Manolio TA. Left atrial volume, geometry, and function in systolic and diastolic heart failure of persons >65 years of age (the cardiovascular health study). Am J Cardiol 2006; 97: 83–89.
• 16.   Yu HT, Yang PS, Kim TH, Uhm JS, Kim JY, Joung B, et al. Poor rhythm outcome of catheter ablation for early-onset atrial fibrillation in women: Mechanistic insight. Circ J 2018; 82: 2259–2268.
• 17.   Tani T, Tanabe K, Ono M, Yamaguchi K, Okada M, Sumida T, et al. Left atrial volume and the risk of paroxysmal atrial fibrillation in patients with hypertrophic cardiomyopathy. J Am Soc Echocardiogr 2004; 17: 644–648.
• 18.   Fatema K, Barnes ME, Bailey KR, Abhayaratna WP, Cha S, Seward JB, et al. Minimum vs. maximum left atrial volume for prediction of first atrial fibrillation or flutter in an elderly cohort: A prospective study. Eur J Echocardiogr 2009; 10: 282–286.
• 19.   Abhayaratna WP, Fatema K, Barnes ME, Seward JB, Gersh BJ, Bailey KR, et al. Left atrial reservoir function as a potent marker for first atrial fibrillation or flutter in persons >65 years of age. Am J Cardiol 2008; 101: 1626–1629.
• 20.   Kuppahally SS, Akoum N, Burgon NS, Badger TJ, Kholmovski EG, Vijayakumar S, et al. Left atrial strain and strain rate in patients with paroxysmal and persistent atrial fibrillation: Relationship to left atrial structural remodeling detected by delayed enhancement-MRI. Circ Cardiovasc Imaging 2010; 3: 231–239.
• 21.   Kubota T, Kawasaki M, Takasugi N, Takeyama U, Ishihara Y, Okubo M, et al. Evaluation of left atrial degeneration for the prediction of atrial fibrillation: Usefulness of integrated backscatter transesophageal echocardiography. JACC Cardiovasc Imaging 2009; 2: 1039–1047.
• 22.   Estes NA 3rd, Halperin JL, Calkins H, Ezekowitz MD, Gitman P, Go AS, et al. ACC/AHA/Physician Consortium 2008 Clinical Performance Measures for Adults with Nonvalvular Atrial Fibrillation or Atrial Flutter: A report of the American College of Cardiology/American Heart Association Task Force on Performance Measures and the Physician Consortium for Performance Improvement (Writing Committee to Develop Clinical Performance Measures for Atrial Fibrillation) Developed in Collaboration with the Heart Rhythm Society. J Am Coll Cardiol 2008; 51: 865–884.
• 23.   Connolly SJ, Eikelboom J, O’Donnell M, Pogue J, Yusuf S. Challenges of establishing new antithrombotic therapies in atrial fibrillation. Circulation 2007; 116: 449–455.
• 24.   Vermes E, Tardif JC, Bourassa MG, Racine N, Levesque S, White M, et al. Enalapril decreases the incidence of atrial fibrillation in patients with left ventricular dysfunction: Insight from the Studies of Left Ventricular Dysfunction (SOLVD) trials. Circulation 2003; 107: 2926–2931.
• 25.   Wachtell K, Lehto M, Gerdts E, Olsen MH, Hornestam B, Dahlöf B, et al. Angiotensin II receptor blockade reduces new-onset atrial fibrillation and subsequent stroke complication to atenolol. J Am Coll Cardiol 2005; 45: 712–719.
• 26.   Madrid AH, Bueno MG, Rebollo JM, Marín I, Peña G, Bernal E, et al. Use of irbesartan to maintain sinus rhythm in patients with long-lasting persistent atrial fibrillation: A prospective and randomized study. Circulation 2002; 106: 331–336.
• 27.   Haywood LJ, Ford CE, Crow RS, Davis BR, Massie BM, Einhorn PT, et al. Atrial fibrillation at baseline and during follow-up in ALLHAT (Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial). J Am Coll Cardiol 2009; 54: 2023–2031.
• 28.   Macfarlane PW, Murray H, Sattar N, Stott DJ, Ford I, Buckley B, et al. The incidence and risk factors for new onset atrial fibrillation in the PROSPER study. Europace 2011; 13: 634–639.
• 29.   Pérez de Isla L, Balcones DV, Fernández-Golfín C, Marcos-Alberca P, Almería C, Rodrigo JL, et al. Three-dimensional-wall motion tracking: A new and faster tool for myocardial strain assessment: Comparison with two-dimensional-wall motion tracking. J Am Soc Echocardiogr 2009; 22: 325–330.