Objective
We sought to explore the reliability of echocardiographic parameters of diastolic function and determine their relationship with functional capacity (New York Heart Association [NYHA] class and 6-minute walk test [6MWT]) and the domains of a health-related quality of life (HR-QOL) questionnaire (Veterans RAND 36-Item Health Survey) in a cohort of ambulatory patients with atrial fibrillation (AF).
Methods
Forty-eight male veterans with persistent or permanent AF underwent clinical examination, echocardiography, and 6MWT, and filled out a Veterans RAND 36-Item Health Survey questionnaire at two visits 1 week apart. Pairwise correlation was performed to evaluate the reliability of echocardiographic parameters and multiple regression analyses to assess the association of these parameters with functional capacity and HR-QOL.
Results
E/e’ average had the highest correlation between the two visits (coefficient: 0.87) and had a significant relationship with NYHA, 6MWT, and physical functioning domain of HR-QOL even after adjusting for confounding variables (odds of NYHA >1: OR 1.5, 95% CI, 1.2–1.9; 6MWT coefficient: −31 ± 9 ft; physical functioning score coefficient: −1.7 ± 0.7).
Conclusion
Average E/e’ is a reliable diastolic function parameter that also independently relates with functional capacity and HR-QOL in ambulatory patients with atrial fibrillation.
Both diastolic heart failure (DHF) and atrial fibrillation (AF) are common diseases in the elderly. DHF refers to a clinical syndrome in patients with symptoms and signs of heart failure with preserved left ventricular (LV) systolic function and evidence of diastolic dysfunction, and constitutes approximately half of all patients with heart failure. Patients with DHF have a poor quality of life (QOL), decreased exercise tolerance, significant comorbidity, high readmission rate, and mortality rate that is similar to those with systolic heart failure. Similarly, patients with AF have a poor QOL, decreased exercise capacity, and an increased risk of death from heart failure. AF occurs in up to one quarter of patients with heart failure with preserved LV systolic function. Thus, DHF and AF frequently coexist and interact in affecting functional tolerance, QOL, morbidity, and mortality.
The main cause of DHF is presumed to be diastolic dysfunction as a result of abnormalities in relaxation or compliance (restriction) of the LV during diastole. Diastolic dysfunction in AF has been related to outcomes such as heart failure and death. Echocardiographic indicators of diastolic function that have been studied include LV wall thickness, left atrial (LA) volume, ratio of transmitral Doppler inflow velocity patterns (E) and tissue Doppler velocities of the mitral annulus (e’), deceleration time (DT) of mitral inflow velocity, and ratio of E to the color M-mode flow propagation velocity (Vp), among others. Although these parameters are well validated with clinical outcomes in patients with sinus rhythm, these conventional methodologies have not been thoroughly studied in patients with AF, who have variable cycle length that may affect the reproducibility of these measurements. Although studies have correlated some of the Doppler parameters with invasive hemodynamics, which may not necessarily reflect symptoms or clinical status, there are limited data in support of relationship of echocardiographic parameters of diastolic function with functional tolerance and QOL in patients with AF.
This study sought to explore the reliability of echocardiographic parameters of diastolic dysfunction and determine which echocardiographic parameters, if any, correlate with objective and subjective functional capacity as measured by New York Heart Association (NYHA) class, 6-minute walk test (6MWT), and QOL in a cohort of ambulatory patients with persistent or permanent AF.
Materials and Methods
Study Population
The institutional review board at the Providence VA Medical Center approved the study, and all patients signed an informed consent. Between October 2008 and May 2010, the Providence VA Medical Center electrocardiographic database was searched to identify patients with AF who presented to ambulatory care clinics, the echocardiography laboratory, the inpatient service, or the electrocardiography laboratory. Patients were eligible for the study if they were more than 18 years of age, with persistent AF (AF duration: 7 days to 1 year) or permanent AF (AF duration >1 year). Exclusion criteria included LV ejection fraction <50% on a prior echocardiogram or nuclear imaging test, a history of paroxysmal AF (self-terminating AF <7 days), greater than moderate degree of valvular heart disease, creatinine >2.5 mg/dL, constrictive or restrictive cardiomyopathy, heart rate >100 beats/min at rest, unstable angina or myocardial infarction during the previous month, and systolic blood pressure >180 mm Hg or diastolic blood pressure >100 mm Hg. A total of 1369 medical records with AF in our electrocardiographic database were reviewed. Approximately 85% of the patients screened did not meet the eligibility criteria. The most common reasons for not meeting the eligibility were presence of more than moderate valvular disease and the paroxysmal nature of AF. Of the remaining 15% eligible patients, 58 consented to participate in the study.
Clinical and Echocardiographic Data Collection
Patients had a baseline visit and a second visit 1 week apart. At each visit, patients underwent clinical examination and chart review to identify medical history and events. Electronic medical records were reviewed to identify the presence of comorbidities, such as hypertension, diabetes, hyperlipidemia, heart failure, peripheral vascular disease, valvular heart disease, chronic obstructive pulmonary disease, coronary artery disease, smoking (current or former), and any alcohol use. At each visit, patients were also graded on their functional tolerance based on the NYHA classification and health-related QOL (HR-QOL), assessed by the Veterans RAND 36-Item Health Survey (VR-36). The VR-36 was developed and modified from the original RAND version of the 36-item Health Survey version 1.0 (also known as the “MOS SF-36”). The VR-36 is a multipurpose health status survey that addresses HR-QOL from physical and mental health perspectives. It yields eight individual scales related to an overall assessment of the HR-QOL: 1) physical functioning (PF), 2) role limitations due to physical problems, 3) bodily pain, 4) general health perceptions, 5) energy or vitality, 6) social functioning, 7) role limitations due to emotional problems, and 8) mental health. The eight scales of the VR-36 have Cronbach’s alphas ranging from 0.69 to 0.89, with excellent discriminant validity of the individual scales and component summaries. Each subscale is graded from 0 to 100 points, with a score of 0 representing low functioning and 100 representing high functioning. Exercise capacity was measured at each visit using the 6-minute corridor walk test following a standardized protocol. Routine echocardiographic assessment was performed on all patients, including M-mode, two-dimensional images, and color flow Doppler recording using Philips IE-33 with a 3.5-MHz transducer (Philips Medical Systems, Andover, MA). Measurements were taken in accordance with the American Echocardiography Society/European Association of Echocardiography guidelines. Patients maintained normal respiration during the acquisition of images. The diastolic parameters assessed included LA volume index (measured in apical four-chamber view by Simpson’s method of disc summation), transmitral E-wave acceleration, transmitral E-wave velocity (E), tissue Doppler e’ at the septal and lateral mitral annulus (e’), mitral valve DT, isovolumetric relaxation time, mitral inflow Vp, and relative wall thickness ( Figure 1 ). The relative wall thickness was calculated by dividing the sum of the interventricular septal thickness and posterior wall thickness by the LV diastolic dimension in the parasternal long-axis view. Both septal and lateral wall tissue Doppler e’ signals were acquired and averaged to estimate average E/e’. Patients were not excluded if they had conditions that may affect e’, such as mitral annular calcification. Doppler recordings were recorded with a sweep speed of 150 mm/s, and a mean value of 10 cardiac cycles was used in analysis. Cardiac cycles were selected in a consecutive fashion based on presence of a discrete Doppler signal and an appropriately placed sample volume site. All recording was performed by the same operator and analyzed by two investigators in consensus using Philips Xcelera software.
Analysis
Descriptive statistics (means and SD) were used to characterize continuous variables, and proportions were used for categorical variables. The test–retest reliability of the echocardiographic measurements between visits were determined using pairwise correlation analysis, and the bias between the first and second measurements was determined through the difference of the means between the two visits. Student t test was used to compare the measurements between visits 1 and 2.
The outcome measures were NYHA class, distance walked in 6 minutes, and scores in each of the VR-36 domains. Potential clinical confounders used for multiple regression analyses were identified by a forward stepwise selection method with a P value of ≤ .20 in at least two outcome measures for entry criteria, which yielded weight, history of hypertension, peripheral artery disease, chronic obstructive pulmonary disease, and heart failure. In addition to these variables, we forced in the variables for age, heart rate, history of coronary artery disease, and valvular disease based on clinical significance. Other clinical variables that were considered but not chosen by the forward stepwise selection include oxygen saturation at rest (pulse oximetry), blood pressure, and history of diabetes, hyperlipidemia, stroke, smoking, or alcohol use.
Data from the first visit were analyzed to obtain initial estimates on the cross-sectional relationship between echocardiographic parameters of diastolic function and different outcomes using simple and multiple regression modeling techniques in a sequential fashion. We then repeated the above analyses using data from both the initial visit and the repeated visit to confirm our initial findings. Given the presence of repeated echocardiographic and outcome measures 1 week apart nested within the same patient in this confirmatory analysis, we used the general estimating equation from Liang and Zeger and Zeger and Liang to build our regression models and adjust for this clustering effect. All analyses were performed using STATA SE version 10 software (Stata Corp LP, College Station, TX).
Results
Study Population
Table 1 shows the baseline characteristics of the patients enrolled in the study. Of the 58 patients who consented, 2 were excluded because they had more than a moderate degree of valvular heart disease on baseline echocardiography, 1 had such a severe degree of chronic obstructive pulmonary disease that he could not lie down for the echocardiogram, 1 was in sinus rhythm at presentation, and 4 withdrew consent. Fifty patients who met the eligibility criteria completed the first visit, and 48 patients completed the second visit. Two patients did not complete the second visit because 1 patient spontaneously converted to sinus rhythm and 1 patient was lost to follow-up. The patients were all male with a mean age of 73 ± 9 years. The majority of the patients had hypertension, hyperlipidemia, and a history of smoking and alcohol use. Most of the patients were receiving anticoagulation (86%), rate-control medications (beta-blocker: 74%, calcium channel blocker: 40%), and HMG Co-A reductase inhibitors (72%) ( Table 1 ).
| Patient characteristics | All patients ( n = 50) |
|---|---|
| Age, y | 73.1 ± 8.8 |
| Systolic blood pressure (mm Hg) | 140.9 ± 15.7 |
| Diastolic blood pressure (mm Hg) | 84.4 ± 10.2 |
| Heart rate (beats/min) | 77.6 ± 12.2 |
| Oxygen saturation (%) | 96.7 ± 5.0 |
| Weight (lbs) | 210.7 ± 45.1 |
| NYHA class 1 | 44% |
| NYHA class II | 54% |
| NYHA class III | 2% |
| Clinical history | |
| Heart failure | 24% |
| Coronary artery Disease/MI | 22% |
| Hypertension | 80% |
| Diabetes mellitus | 36% |
| Hyperlipidemia | 82% |
| Peripheral vascular disease | 14% |
| Valvular heart disease | 26% |
| Chronic obstructive pulmonary disease | 28% |
| Alcohol | 80% |
| Smoking | 82% |
| Stroke | 14% |
| Medications | |
| Aspirin | 44% |
| ACEIs | 40% |
| ARBs | 18% |
| Beta-blockers | 74% |
| Warfarin | 86% |
| Calcium channel blocker | 40% |
| Digoxin | 20% |
| HMG CoA reductase inhibitors | 72% |
| Diuretics | 42% |
Echocardiographic Parameters of Diastolic Function
All the measured diastolic echocardiographic parameters had a high correlation when compared across a 1-week period, indicating a high test–retest reliability ( Table 2 ). The correlation coefficient was highest for E/e’ average, followed by E/e’ septal, E/e’ lateral, and left atrial area. The average difference between the measurements in visits 1 and 2 for each echocardiographic variable evaluated was within 5% of the measurement obtained during the first visit and not statistically significant ( Table 2 ). Also, the distance walked in 6 minutes, NYHA class, and PF score were measured at both visits and were not significantly different between visits ( Table 2 ).
| Echocardiographic parameters | Visit 1 ( n = 50) | Visit 2 ( n = 48) | Difference † (visit 1 to visit 2) | Correlation coefficient ∗ |
|---|---|---|---|---|
| Left atrial volume index (cm 3 /m 2 ) | 29.7 ± 9.3 | 30.5 ± 8.4 | −0.6 ± 5.9 | 0.79 |
| E velocity (cm/s) | 107.0 ± 18.5 | 106.1 ± 18.4 | 1.1 ± 11.6 | 0.80 |
| E acceleration (m/s 2 ) | 14.83 ± 4.24 | 14.42 ± 3.98 | 0.46 ± 3.36 | 0.67 |
| e’ septal (cm/s) | 8.15 ± 1.8 | 7.9 ± 1.7 | 0.004 ± 2.2 | 0.74 |
| e’ lateral (cm/s) | 11.4 ± 2.5 | 11.3 ± 2.1 | −0.1 ± 2.7 | 0.69 |
| e’ average (cm/s) | 9.8 ± 1.9 | 9.6 ± 1.7 | −0.06 ± 2.2 | 0.79 |
| E/e’ septal | 13.9 ± 4.7 | 14.2 ± 4.5 | −0.4 ± 2.7 | 0.86 |
| E/e’ lateral | 9.9 ± 3.0 | 9.9 ± 3.2 | −0.2 ± 2.1 | 0.84 |
| E/e’ average | 11.4 ± 3.3 | 11.6 ± 3.5 | −0.3 ± 2.0 | 0.87 |
| IVRT (ms) | 54.0 ± 7.6 | 53.8 ± 6.1 | 1.3 ± 9.8 | 0.54 |
| Deceleration time (ms) | 186 ± 40 | 193 ± 38 | −6.5 ± 27.9 | 0.75 |
| Vp (cm/s) | 64.5 ± 23.5 | 65.8 ± 27.8 | 0.05 ± 23.2 | 0.60 |
| E/Vp | 1.9 ± 0.9 | 1.9 ± 0.8 | 0.08 ± 0.9 | 0.79 |
| Relative wall thickness (%) | 37.7 ± 7.0 | 37.6 ± 7.0 | 0.3 ± 4.7 | 0.78 |
| Outcome variables | ||||
| NYHA class >I (n) | 28 | 23 | ||
| 6MWT distance (ft) | 1121 ± 333 | 1149 ± 344 | −28 ± 114 | 0.94 |
| PF score | 57 ± 24 | 56 ± 24 | 1 ± 14 | 0.83 |
∗ All the correlation coefficient values of the echocardiographic parameters were significant at P < .05.
† All the echocardiographic parameters and outcome variables had no significant difference between visits 1 and 2 ( P > .05).
Relationship Between Echocardiographic Parameters and Functional Tolerance
A total of 22 patients (44%) were in NYHA class I, 27 patients (54%) were in NYHA class II, and 1 patient was in NYHA class III on the first visit. Unadjusted analyses demonstrated that E-wave acceleration, all E/e’ ratios (septal, lateral, and average), and E/Vp were significantly related to NYHA functional class ( Supplementary Table 1 ), which remained significant in multiple regression analyses after adjusting for age, weight, heart rate, history of hypertension, coronary artery disease, heart failure, peripheral artery disease, chronic obstructive pulmonary disease, and valvular disease. ( Table 3 ). Although not significant in the simple regression analysis, DT showed a significant relationship with NYHA class in the multiple regression analysis. In a regression model consisting of the clinical variables and significant echocardiographic parameters identified by multivariate analyses (E-wave acceleration, E/e’ average, E/Vp, and DT), only the relationship between E/e’ average and NYHA class remained significant (OR 1.34; 95% CI, 1.03–1.75).
| Echocardiographic parameters | OR (95% CI) visit 1 | OR (95% CI) both visits | ||
|---|---|---|---|---|
| Left atrial volume index (cm 3 /m 2 ) | 0.93 | (0.85–1.00) | 0.95 | (0.89–1.00) |
| E acceleration (m/s 2 ) | 1.19 | (1.02–1.39) | 1.47 | (1.10–1.96) |
| E/e’ septal | 1.59 | (1.11–2.27) | 1.28 | (1.10–1.49) |
| E/e’ lateral | 2.11 | (1.26–3.54) | 1.60 | (1.24–2.06) |
| E/e’ average | 2.00 | (1.25–3.19) | 1.52 | (1.21–1.91) |
| IVRT (ms) | 1.14 | (1.02–1.28) | 1.06 | (0.98–1.14) |
| DT (ms) | 1.02 | (1.00–1.04) | 1.02 | (1.00–1.03) |
| E/Vp | 4.73 | (1.17–19.13) | 3.47 | (1.49–8.07) |
| Relative wall thickness (%) | 1.03 | (0.94–1.14) | 1.02 | (0.95–1.09) |
The average distance walked in 6 minutes was 1136 ± 339 ft (range: 200–1826 ft). Simple regression analysis demonstrated that the E/e’ ratios were the only echocardiographic parameters that significantly related with the distance walked in 6 minutes ( Supplementary Table 1 ). Average E/e’, septal E/e’, and lateral E/e’ remained significantly related to the 6MWT distance in multiple regression analyses after adjustment for the potential confounders previously described ( Table 4 ). Although not significant in the simple regression analysis, E-wave acceleration was significantly related to 6MWT distance in multiple regression analysis. In a regression model consisting of the clinical variables and significant echocardiographic parameters identified by multivariate analyses (E-wave acceleration, E/e’ average), only the relationship between E/e’ average and 6MWT distance remained significant (regression coefficient: −29.1 ± 10.7, P < .01).
| Echocardiographic parameters | Coefficient visit 1 | P value | Coefficient both visits | P value |
|---|---|---|---|---|
| Left atrial volume index (cm 3 /m 2 ) | 2.6 ± 5.0 | .60 | 2.3 ± 3.3 | .50 |
| E acceleration (m/s 2 ) | −19.8 ± 12.6 | .12 | −16.6 ± 8.4 | .05 |
| E/e’ septal | −23.0 ± 11.0 | .04 | −19.8 ± 7.1 | .01 |
| E/e’ lateral | −33.8 ± 15.6 | .04 | −33.5 ± 10.3 | .001 |
| E/e’ average | −32.8 ± 14.4 | .03 | −30.8 ± 9.4 | .001 |
| IVRT (ms) | −2.0 ± 6.4 | .76 | 3.2 ± 4.5 | .48 |
| DT (ms) | 0.3 ± 1.2 | .80 | −0.5 ± 0.9 | .56 |
| E/Vp | −31.3 ± 56.1 | .58 | −3.6 ± 39.3 | .93 |
| Relative wall thickness (%) | −2.0 ± 6.8 | .77 | −6.2 ± 4.3 | .15 |
Relationship Between Echocardiographic Parameters and Domains of VR-36
The results of the VR-36 questionnaire are shown in Table 5 . Of the eight domains, only PF showed a significant relationship with DT, E/e’ average, or lateral ( Supplementary Table 1 ), which remained significant after multivariate adjustment ( Table 6 ). Although not significant in simple regression analysis, E/e’ septal was significantly related to PF scores in multiple regression analysis ( Table 6 ). Also, the general health perceptions domain showed a significant independent relationship with left atrial volume index, bodily pain domain with E-wave acceleration, and social functioning and mental health with isovolumetric relaxation time. None of the other domains of HR-QOL had a significant independent relationship with any of the echocardiographic parameters. The regression coefficients between the echocardiographic parameters and the other seven domains of VR-36 are shown in Supplementary Tables 2 to 9 . In addition, the distance walked in 6 minutes significantly related to the PF score independently of the clinical confounders (coefficient = 0.05 ± 0.01, P < .001).
| VR-36 domains | Mean ± SD | Range |
|---|---|---|
| Physical functioning | 56.9 ± 24.2 | 15–100 |
| Role limitation due to physical problems | 60.6 ± 28.4 | 13–100 |
| Role limitation due to emotional problems | 70.5 ± 28.2 | 8–100 |
| Vitality | 58.5 ± 19.2 | 15–100 |
| Mental health | 74.0 ± 18.9 | 8–100 |
| Social functioning | 71.5 ± 28.5 | 13–100 |
| Bodily pain | 69.3 ± 22.8 | 20–100 |
| General health | 57.3 ± 19.4 | 20–100 |
| Echocardiographic parameters | Coefficient visit 1 | P value | Coefficient both visits | P value |
|---|---|---|---|---|
| Left atrial volume index (cm 3 /m 2 ) | 0.6 ± 0.4 | .10 | 0.5 ± 0.2 | .06 |
| E acceleration (m/s 2 ) | −1.1 ± 0.9 | .24 | −0.2 ± 0.6 | .77 |
| E/e’ septal | −1.9 ± 0.8 | .02 | −1.3 ± 0.5 | .02 |
| E/e’ lateral | −2.0 ± 1.2 | .09 | −1.6 ± 0.8 | .05 |
| E/e’ average | −2.1 ± 1.1 | .05 | −1.7 ± 0.7 | .02 |
| IVRT (ms) | −0.1 ± 0.5 | .88 | 0.1 ± 0.3 | .71 |
| DT (ms) | −0.1 ± 0.1 | .31 | −0.2 ± 0.1 | .01 |
| E/Vp | −5.1 ± 4.1 | .22 | −2.1 ± 2.9 | .47 |
| Relative wall thickness (%) | 0.2 ± 0.5 | .74 | 0.0 ± 0.3 | .98 |
Results
Study Population
Table 1 shows the baseline characteristics of the patients enrolled in the study. Of the 58 patients who consented, 2 were excluded because they had more than a moderate degree of valvular heart disease on baseline echocardiography, 1 had such a severe degree of chronic obstructive pulmonary disease that he could not lie down for the echocardiogram, 1 was in sinus rhythm at presentation, and 4 withdrew consent. Fifty patients who met the eligibility criteria completed the first visit, and 48 patients completed the second visit. Two patients did not complete the second visit because 1 patient spontaneously converted to sinus rhythm and 1 patient was lost to follow-up. The patients were all male with a mean age of 73 ± 9 years. The majority of the patients had hypertension, hyperlipidemia, and a history of smoking and alcohol use. Most of the patients were receiving anticoagulation (86%), rate-control medications (beta-blocker: 74%, calcium channel blocker: 40%), and HMG Co-A reductase inhibitors (72%) ( Table 1 ).
| Patient characteristics | All patients ( n = 50) |
|---|---|
| Age, y | 73.1 ± 8.8 |
| Systolic blood pressure (mm Hg) | 140.9 ± 15.7 |
| Diastolic blood pressure (mm Hg) | 84.4 ± 10.2 |
| Heart rate (beats/min) | 77.6 ± 12.2 |
| Oxygen saturation (%) | 96.7 ± 5.0 |
| Weight (lbs) | 210.7 ± 45.1 |
| NYHA class 1 | 44% |
| NYHA class II | 54% |
| NYHA class III | 2% |
| Clinical history | |
| Heart failure | 24% |
| Coronary artery Disease/MI | 22% |
| Hypertension | 80% |
| Diabetes mellitus | 36% |
| Hyperlipidemia | 82% |
| Peripheral vascular disease | 14% |
| Valvular heart disease | 26% |
| Chronic obstructive pulmonary disease | 28% |
| Alcohol | 80% |
| Smoking | 82% |
| Stroke | 14% |
| Medications | |
| Aspirin | 44% |
| ACEIs | 40% |
| ARBs | 18% |
| Beta-blockers | 74% |
| Warfarin | 86% |
| Calcium channel blocker | 40% |
| Digoxin | 20% |
| HMG CoA reductase inhibitors | 72% |
| Diuretics | 42% |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
