Frailty is prevalent in patients with cardiovascular disease, but few studies have evaluated relations between frailty and echocardiographically determined cardiac indexes. To assess the prevalence of frailty and its association with echocardiographic characteristics, we prospectively measured frailty in 257 patients ≥65 years who underwent echocardiography (transthoracic echocardiography [TTE]) from June 2012 to February 2013. Deficits of weight loss, exhaustion, physical activity, gait speed, and handgrip strength were used to categorize patients as frail (≥3 features), intermediately frail (1 or 2 features), or nonfrail (0 features). Pearson correlation was used to examine bivariate associations between TTE variables and frailty. Kaplan-Meier methods were used to estimate overall survival based on frailty status. A multivariable model was used to examine TTE indexes associated with frailty while accounting for age and baseline cardiac co-morbidities. Of the 257 patients studied, 40 (15.6%) were nonfrail, 167 (65.0%) intermediately frail, and 50 (19.4%) frail. Left atrial volume ( r = 0.14; p = 0.03), stroke volume ( r = −0.19; p <0.01), E/A ratio ( r = 0.26; p <0.001), and pulmonary artery systolic pressure ( r = 0.33; p <0.001) correlated with fraility. After age and baseline cardiac comorbidities were accounted for, larger left atrial volumes, lower stroke volumes, and higher pulmonary artery systolic pressures remained independently associated with frailty. Frail patients had worse survival compared with nonfrail and intermediately frail patients (p = 0.016 by log-rank). In conclusion, 1/5 of older patients who underwent clinically indicated TTE were frail, with worse survival and a unique fingerprint of TTE findings distinguishing them from nonfrail patients.
We studied the association of frailty with cardiac structure and function in a prospectively assembled cohort of patients who underwent a 2-dimensional and Doppler transthoracic echocardiographic (TTE) examination for clinical indications. Our study aim was to determine whether there were specific morphologic or hemodynamic correlations between TTE indexes and frailty after accounting for patient age and known baseline cardiac conditions. In addition, we sought to determine whether frailty assessed at a clinical TTE examination was associated with survival.
Methods
Patients ≥65 years who underwent a clinically indicated TTE at the Mayo Clinic in Rochester, Minnesota, from June 2012 to February 2013, were prospectively enrolled. Frailty was assessed on the same day as the TTE examination. Participants were required to provide written informed consent for the study. To maintain external validity, no other exclusions were applied. Patient demographic characteristics and medical history were reviewed. The Mayo Clinic Institutional Review Board approved this study.
Two-dimensional and Doppler TTE were performed for all patients, and left ventricular function, volumes, and mass were measured according to guidelines. Left atrial volume was measured in biplane views and indexed to body surface area. Estimates of pulmonary artery systolic pressure were derived from the Bernoulli equation with the tricuspid regurgitation jet velocity. Valvular heart disease was graded according to guidelines that included qualitative descriptions for valvular and annular calcification and quantitative assessment of stenosis and regurgitation. Diastolic function was assessed in all patients and included pulse-wave Doppler measurements of early and late transmitral velocities and deceleration time and tissue Doppler measurements of the medial and lateral mitral annular velocities. Elevated left ventricular filling pressure was defined as early transmitral to annular velocity ratio of ≥15, a cutoff that has 86% specificity for invasively measured mean left ventricular diastolic pressure >15 mm Hg. Noninvasively determined Doppler-derived stroke volume and cardiac output were measured in all patients. All TTE examinations were interpreted by echocardiologists at the Mayo Clinic who had completed level 3 training and were blinded to frailty status.
The frailty phenotype was assessed with a multidimensional construct as previously reported by Fried et al in the Cardiovascular Health Study. The determination of frailty incorporates abnormalities in musculoskeletal, nutritional, and functional domains. For each participant, 5 criteria were measured: unintended weight loss (>4.5 kg in the preceding year), exhaustion, physical activity, gait speed (i.e., walking time >6 m), and handgrip strength. Exhaustion was measured with the Center for Epidemiologic Studies Depression Scale. The 2 items in this scale are whether the patient believed that (1) “I felt that everything I did was an effort” and (2) “I could not get ‘going’.” Patients who answered “3 or more days in the past week” to either statement satisfied the criteria for exhaustion. The Duke Activity Status Index was used instead of the Minnesota Leisure Time Activity questionnaire to measure physical activity to determine cardiac-related functional status. Scores range from 0 to 58, with higher scores reflecting better function. The Duke Activity Status Index scores were converted to estimates of energy expenditure as metabolic equivalent task values, with a score cutoff of 4 used to fulfill criteria for an abnormal physical activity index. Gait speed was assessed as walking time >6 m in an unobstructed hallway. Handgrip strength was tested in the participant’s dominant hand with a Jamar handgrip dynamometer (Jamar, Bolingbrook, Illinois) and recorded in kilograms. Abnormal handgrip strength was determined according to values adjusted for gender and body mass index. Specially trained registered nurses performed physical testing for frailty. Patients were classified as “frail” if they had deficits in ≥3 elements, as “intermediately frail” if they had deficits in 1 or 2 elements, and as “nonfrail” if they had no apparent deficits.
Univariate continuous distributions were summarized as median (interquartile range); discrete data were summarized as frequency (percentage). We stratified groups as frail, intermediately frail, and nonfrail for comparisons and tested the association of baseline clinical and TTE parameters across frailty groups with Pearson coefficients. We also performed a separate analysis to evaluate associations between the same baseline clinical and TTE parameters with the single-item frailty components for gait speed and handgrip strength. Scatterplot matrices were visually reviewed to delineate whether TTE covariates exhibited multicollinearity. Survival estimates were computed with the Kaplan-Meier method across frailty groups. Multivariable logistic regression was used to estimate the association between TTE indexes, baseline characteristics, and the frailty phenotype. Data analysis was performed with SAS software, version 9.1 (SAS Institute, Cary, NC), and p <0.05 was considered statistically significant.
Results
Of the 257 patients studied, 40 (15.6%) were nonfrail, 167 (65.0%) were intermediately frail, and 50 (19.4%) were frail. Table 1 lists the demographic characteristics of the study population. Table 2 lists the TTE characteristics of the study cohort. Frailty was associated with larger left atrial volumes, lower stroke volume indexes, and higher estimates of pulmonary artery systolic pressure. Table 3 depicts correlations between TTE variables and both single-item frailty measures and multi-item frailty measures. Most correlations were significant but modest and varied depending on the frailty assessment used (single item vs multi-item). On multivariable analysis, left atrial volume (odds ratio [OR] 1.06; 95% confidence interval [CI] 1.00 to 1.11), stroke volume (OR 0.86; 95% CI 0.78 to 0.96), and estimates of pulmonary artery systolic pressure (OR 1.19; 95% CI 1.05 to 1.36) were independently associated with frailty, suggesting that frail adults (vs nonfrail and intermediately frail adults) were more likely to have larger left atrial volumes, lower stroke volumes, and higher pulmonary pressures ( Table 4 ). Differences in TTE characteristics remained significant after accounting for age, which was not independently associated with frailty (adjusted OR 1.22; 95% CI 0.90 to 1.67). TTE variables associated with frailty remained predictive after accounting for previous myocardial infarction and history of heart failure. During a mean ± SD follow-up of 1.0 ± 0.5 years, no patients in the nonfrail group died, but survival was estimated as 92.2% in the intermediately frail group and 77.2% in the frail group (p = 0.02; Figure 1 ).
| Variable | Nonfrail (n=40) | Intermediately Frail (n=167) | Frail (n=50) | P Value |
|---|---|---|---|---|
| Age (yrs) | 72.7±5.9 | 75.4±5.8 | 78.0±7.4 | <.001 |
| Men | 25 (62%) | 100 (60%) | 21 (42%) | .06 |
| Body mass index (kg/m 2 ) | 28.4±4.7 | 28.9±5.9 | 29.9±7.3 | .44 |
| Systolic blood pressure (mm Hg) | 135±18 | 128±19 | 124±21 | <.02 |
| Prior myocardial infarction | 1 (2%) | 19 (11%) | 12 (24%) | <.01 |
| Hypertension † | 27 (68%) | 130 (78%) | 35 (70%) | .35 |
| Hyperlipidemia ‡ | 26 (65%) | 111 (66%) | 34 (68%) | .98 |
| Heart failure | 7 (18%) | 51 (31%) | 23 (46%) | <.02 |
| Diabetes mellitus | 4 (10%) | 26 (16%) | 17 (34%) | <.01 |
| Pulmonary disease | 9 (22%) | 32 (19%) | 13 (26%) | .55 |
| Stroke or Transient ischemic attack | 2 (5%) | 12 (7%) | 2 (4%) | .68 |
| Percutaneous coronary intervention | 3 (8%) | 23 (14%) | 4 (8%) | .38 |
| Coronary artery bypass grafting | 2 (5%) | 30 (18%) | 10 (20%) | .11 |
| Smoking history | 16 (40%) | 78 (47%) | 21 (42%) | .67 |
| Medications | ||||
| β-Blocker | 22 (55%) | 108 (65%) | 32 (64%) | .62 |
| Calcium channel blocker | 8 (20%) | 30 (18%) | 7 (14%) | .71 |
| Diuretic | 13 (32%) | 66 (40%) | 30 (60%) | .07 |
| Digoxin | 1 (2%) | 16 (10%) | 6 (12%) | .27 |
| Statin | 20 (50%) | 105 (63%) | 25 (50%) | .16 |
∗ Continuous data are presented as mean ± SD; categorical data as number of patients (percentage of sample).
† Defined by treatment with any antihypertensive medication.
| Variable | Nonfrail (n=40) | Intermediately Frail (n=167) | Frail (n=50) | P Value |
|---|---|---|---|---|
| LV ejection fraction (%) | 58±12 | 58±13 | 58±11 | .97 |
| LV mass index (g/m 2 ) | 109.2±34.3 | 105.4±31.1 | 102.8±26.3 | .63 |
| LA volume (cm 3 ) | 81.2±40 | 78.9±30 | 95.4±40 | <.02 |
| Stroke volume index (cm 3 /m 2 ) | 60.0±13.1 | 48.3±9.7 | 44.2±10.5 | <.01 |
| Cardiac index (L/min/m 2 ) | 3.2±0.8 | 3.0±0.6 | 3.0±0.8 | .25 |
| E/A ratio | 1.0±0.4 | 1.0±0.5 | 1.2±0.6 | .07 |
| Deceleration time (ms) | 219.0±50 | 223.9±136 | 205.0±56 | .68 |
| E/e′ (medial) ratio | 14.5±6 | 16.1±14 | 19.1±11 | .23 |
| Pulmonary artery systolic pressure (mm Hg) | 31.2±6 | 34.2±11 | 43.2±20 | <.001 |
| Aortic stenosis (≥mild) | 3 (8%) | 15 (9%) | 8 (16%) | .30 |
| Aortic regurgitation (≥mild) | 3 (8%) | 8 (5%) | 2 (4%) | .70 |
| Mitral regurgitation (≥mild) | 6 (15%) | 18 (11%) | 12 (24%) | .06 |
| Mitral stenosis (≥mild) | 0 | 1 (1%) | 3 (6%) | <.02 |
| Mitral annular calcium | 6 (15%) | 38 (23%) | 22 (44%) | <.01 |
| Calcified aortic valve | 4 (10%) | 15 (9%) | 6 (12%) | .81 |
∗ Continuous data are presented as mean ± SD; categorical data as number of patients (percentage of sample).
| Echocardiographic Variable | Handgrip Strength | Gait Speed | Frailty Composite ∗ | |||
|---|---|---|---|---|---|---|
| r | P Value | r | P Value | r | P Value | |
| LV ejection fraction | −0.13 | .03 | −0.09 | .16 | −0.01 | .85 |
| LV mass index, g/m 2 | 0.18 | .004 | 0.02 | .79 | −0.04 | .56 |
| LA volume, cm 3 | 0.16 | .01 | 0.18 | <.01 | 0.14 | .03 |
| Stroke volume index, cm 3 /m 2 | 0.03 | .63 | −0.18 | <.01 | −0.19 | <.01 |
| Cardiac index, L/min/m 2 | −0.16 | .01 | −0.08 | .22 | −0.08 | .18 |
| E/A ratio | −0.07 | .35 | 0.12 | .10 | 0.26 | <.001 |
| E/e′ (medial) ratio | −0.12 | .06 | 0.21 | <.01 | 0.13 | .05 |
| Pulmonary artery systolic pressure, mm Hg | −0.24 | <.001 | 0.37 | <.001 | 0.33 | <.001 |
| Aortic stenosis (≥mild) | −0.03 | .60 | 0.08 | .23 | 0.03 | .68 |
| Mitral stenosis (≥mild) | −0.17 | <.01 | 0.15 | .02 | 0.14 | .03 |
| Mitral annular calcium | −0.15 | .02 | 0.18 | .02 | 0.22 | <.01 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
