The American Society of Echocardiography (ASE) and European Association of Echocardiography (EAE) recommend the use of quantitative estimation of left ventricular (LV) mass and defined partition values for mild, moderate, and severe hypertrophy. However, the prognostic implications associated with this categorization are unknown.
In this observational cohort study of unselected adults undergoing echocardiography for any indication, LV hypertrophy was assessed using the ASE/EAE-recommended formula and measurement convention from LV linear dimensions indexed to body surface area. Mortality and incident hospitalizations for cardiovascular disease were the outcomes of this study.
Of 2,545 subjects (mean age, 61.9 ± 15.8 years; 56.3% women), 52.9% had normal LV mass, and 15.4% had mild, 12.1% moderate, and 19.6% severe LV hypertrophy. During a mean follow-up period of 2.5 ± 1.2 years, 121 deaths and 292 incident hospitalizations for cardiovascular disease occurred. In multivariate models including age, gender, LV ejection fraction, wall motion score index, significant valvular disease, and atrial fibrillation, the adjusted hazard ratios for death were 1.81 (95% confidence interval [CI], 1.03–3.20; P = .041) for mild, 2.31 (95% CI, 1.33–4.01; P = .003) for moderate, and 2.30 (95% CI, 1.39–3.79, P = .001) for severe LV hypertrophy. The adjusted hazard ratios for incident cardiovascular hospitalizations were 1.24 (95% CI, 0.84–1.82; P = .277) for mild, 2.02 (95% CI, 1.42–2.88; P = .0001) for moderate, and 2.38 (95% CI, 1.75–3.22, P < .0001) for severe LV hypertrophy. After adjustment for known risk predictors, there was a 1.3-fold risk for death and cardiovascular disease events per category of LV mass ( P = .001).
In a cohort study of unselected adult outpatients, the categorization of LV mass according to the ASE/EAE recommendations offered prognostic information independently of age, gender, and other known predictors.
Echocardiographically detected left ventricular (LV) hypertrophy has been shown to be a strong independent marker of cardiovascular risk both in the general population and in high-risk groups. Nonetheless, these studies have used a variety of indexation methods and partition values.
The American Society of Echocardiography (ASE) and European Association of Echocardiography (EAE) recommendations for chamber quantification conclude that LV mass from the ASE-recommended formula using LV linear dimensions indexed to body surface area (BSA) is preferred in the diagnosis of LV hypertrophy over linear measurements such as septal or posterior wall thickness. In addition, they propose sex-specific cutoff values to categorize the degree of LV hypertrophy as mild, moderate, and severe on the basis of the distribution of LV mass in relation to reference limits in an ethnically heterogeneous North American population.
However, despite the widespread use of such descriptive terms in clinical practice, this categorization is, at present, inadequately substantiated by scientific data, underscoring the weaknesses inherent the use of arbitrary dichotomous threshold without providing insights into risks associated with different levels of LV mass. Therefore, aim of this study was to examine the prognostic implications of the ASE/EAE partition values of LV mass in a large group of unselected outpatients referred to a tertiary care echocardiography laboratory.
The study population comprised unselected elective adult outpatients who underwent standard Doppler echocardiography for any indication in the period from January 2005 to March 2009 at the echocardiography laboratory of Modena University Hospital. Criteria for enrollment were (1) age ≥ 18 years; (2) complete resting two-dimensional (2D) echocardiographic assessment, including real-time measurement of LV mass; and (3) residency in the province of Modena, Italy. For patients undergoing more than one echocardiographic exam during the aforementioned time frame, we considered only the first access to the echocardiography laboratory.
All exams were performed using an Acuson Sequoia ultrasound system (Siemens Medical Solutions USA, Inc., Mountain View, CA) and were performed and/or supervised by cardiologists fully trained in echocardiography with long-standing experience with the technique and intense hands-on training period with interpretation of >750 studies.
LV diameters as well as septal and posterior wall thickness were measured using the 2D-guided M-mode method in the short-axis view or the linear 2D method in the parasternal long-axis view. LV end-diastolic dimensions were measured at the onset of the QRS complex. LV ejection fraction was assessed using the biplane Simpson method or the Quinones method using LV end-systolic and end-diastolic diameters or visually estimated, a method that was documented to have accuracy comparable with that of the other methods in assessing LV ejection fraction.
Regional LV systolic function was assessed using a standard 16-segment model. Segmental scores were assigned as follows: 1 = normal or hyperkinesis, 2 = hypokinesis, 3 = akinesis, 4 = dyskinesis, and 5 = aneurysmal. The wall motion score index (WMSI) was derived as the sum of all scores divided by the number of segments visualized. Individual echocardiographic Doppler parameters (mitral inflow pattern, tissue Doppler, and Valsalva maneuver when necessary) were integrated to grade diastolic function in four stages: normal diastolic function; impaired relaxation with normal or near normal filling pressures (grade I/IV); impaired relaxation with moderate elevation of filling pressures, pseudonormal filling (grade II/IV); and impaired relaxation with marked elevation of filling pressures, restrictive filling (grades III–IV/IV), as previously described.
Left atrial volume was assessed using the modified biplane Simpson method from apical four-chamber and two-chamber views and indexed to BSA. Measurements were obtained in end-systole from the frame preceding mitral valve opening.
Each value represented the average of three consecutive beats.
Significant left-sided valve disease severity was defined as the presence of aortic or mitral prosthesis or the presence of greater than moderate native mitral or aortic valve stenosis or insufficiency, similarly to previous reports. Valve disease severity was defined according to the American Heart Association and American College of Cardiology guidelines for the management of valvular heart disease. Applying these guidelines, the cardiologist performing the exam graded valve disease as absent, mild, moderate, or severe, and this information was embedded in the echocardiographic report. The methods used included pulsed-wave and continuous-wave Doppler velocities and gradients, direct measurement of valve area planimetry, continuity equation, color Doppler to assess the jet width, or proximal isovelocity surface area for quantitative evaluation. The methods used to classify valve disease severity were at the discretion of the physician performing the exam, and often the final judgment was based on the combination of more than one method.
All measurements were performed online and entered in an electronic database at the time of the echocardiographic study. No modification from the original database was applied, and no measurement was made offline. Hence, the study consisted of a retrospective analysis of data entered in the electronic echocardiographic database.
LV mass was obtained using the ASE-recommended formula for the estimation of LV mass from LV linear dimensions on the basis of modeling the left ventricle as a prolate ellipse of revolution: LV mass (g) = 0.8[1.04(LVIDD + IVST + PWT) 3 − LVIDD 3 ] + 0.6, where LVIDD is LV internal end-diastolic dimension, IVST is end-diastolic interventricular septal wall thickness, and PWT is end-diastolic LV posterior wall thickness.
LV mass was indexed to BSA. LV mass was also indexed to height and to height 2.7 .
Cutoff Limits for LV Hypertrophy
The ASE/EAE guidelines suggest the following cutoffs for LV hypertrophy: LV mass/BSA > 95 g/m 2 (LV mass/height > 99 g/m, LV mass/height 2.7 > 44 g/m 2.7 ) for women and LV mass/BSA > 115 g/m 2 (LV mass/height > 126 g/m, LV mass/height 2.7 > 48 g/m 2.7 ) for men. Values for mild LV hypertrophy are LV mass/BSA of 96 to 108 g/m 2 (LV mass/height, 100–115 g/m; LV mass/height 2.7 , 45–51 g/m 2.7 ) for women and LV mass/BSA of 116 to 131 g/m 2 (LV mass/height, 127–144 g/m; LV mass/height 2.7 , 49–55 g/m 2.7 ) for men; values for moderate LV hypertrophy are LV mass/BSA of 109 to 121 g/m 2 (LV mass/height, 116–128 g/m; LV mass/height 2.7 , 52–58 g/m 2.7 ) for women and LV mass/BSA of 132 to 148 g/m 2 (LV mass/height, 145–162 g/m; LV mass/height 2.7 , 56–63 g/m 2.7 ) for men; values for severe LV hypertrophy are LV mass/BSA ≥122 g/m 2 for women (LV mass/height ≥ 129 g/m, LV mass/height 2.7 ≥ 59 g/m 2.7 ) and LV mass/BSA ≥ 149 g/m 2 (LV mass/height ≥ 163 g/m, LV mass/height 2.7 ≥ 64 g/m 2.7 ) for men.
Relative wall thickness (RWT) was measured using the formula (2 × PWT)/LVIDD, which permitted the categorization of an increase in LV mass as either concentric (RWT ≥ 0.42) or eccentric (RWT < 0.42) hypertrophy and allowed the identification of concentric remodeling (normal LV mass with increased RWT).
Age, sex, height, weight, BSA, body mass index, and cardiac rhythm were recorded at the time of echocardiography. History of cardiovascular diseases (prior acute coronary syndromes, including ST-segment elevation myocardial infarction and unstable angina or non–ST-segment elevation myocardial infarction, history of chronic coronary artery disease, prior acute or chronic heart failure, and prior stroke) was obtained using the hospital discharge codes of the public hospitals of Modena province. Risk factors were obtained retrospectively by manual review of electronic clinical notes of the public hospitals of the province. These electronic records allow assessment of outpatient visits as well as hospital discharge notes. To collect information on medications, we combined two methods: (1) we retrospectively reviewed through the electronic databases of Modena province public hospital discharge letters and ambulatory cardiology visits, and (2) we used the Emilia Romagna regional pharmacy centralized electronic database. This electronic database contains all prescriptions that are filled in all pharmacies of the region by all residents, regardless of the physician prescribing the medication. We examined all prescriptions filled from 60 days before to 30 days after the date of the exam.
The main end point was all-cause death. The secondary end point was cardiovascular morbidity, measured as the cumulative incidence of hospitalizations for cardiovascular disease. Included were hospitalizations for coronary disease, heart failure, ventricular arrhythmias, atrial fibrillation, lower limb critical ischemia, and stroke or transient ischemic attack. The diagnoses were derived from hospital discharge codes. In most cases, more than one code was reported, but only the first was considered to classify the cause of hospitalizations.
Follow-up information for death was obtained from the national death index, in which the status of all citizens is constantly updated and is 100% complete. In Italy, it is mandatory by law that all deceased patients be immediately registered in this national data bank. Cardiovascular morbidity requiring hospitalization was assessed using the electronic archives of the health service of Modena province. All public hospitalization records of citizens resident in Modena province are stored in a digital format and may be accessed online after obtaining permission and an access password. This archive allows nearly complete knowledge of all clinical events requiring hospitalization in Modena province since 1999. Each electronic record includes up to 30 to 50 codes reflecting various diagnoses, complications, and procedures performed while patients were hospitalized. The diagnoses were classified according to the International Classification of Disease, Ninth Revision, Clinical Modification. We obtained computerized records of all incident hospitalizations for cardiovascular reasons as the principal diagnosis after the index echocardiographic study. In case of uncertainty in adjudicating events from this electronic database, the general practitioners were contacted by telephone. To ensure completeness of follow-up, only residents in Modena province were included in the study.
Data are shown as percentages for categorical variables and as mean ± SD for continuous variables. Comparisons across groups were made using χ 2 tests for categorical variables and analysis of variance for continuous variables or Kruskal-Wallis tests for highly skewed variables.
Kaplan-Meier curves were constructed to show survival and the cumulative incidence of hospitalizations for cardiovascular diseases between patients with and without LV hypertrophy and across the LV mass partition values. Event rates ± 1 SE were estimated according to the Kaplan-Meier method, and groups were compared using a two-sided log-rank test for trend across the categories of LV mass.
Univariate and multivariate Cox regression analyses were used to estimate the relative risk for death and cardiovascular morbidity; these relative risks are presented as hazard ratios (HRs) with 95% confidence intervals (CIs). Patients were censored at the end of their event free follow-up. Only the first incident event was included for each patient.
Patients were censored at the time of their first events. LV mass index was analyzed in several ways. First, the HR, was assessed using LV mass as a continuous variable; then, the HRs for patients with LV hypertrophy were compared with those of patients without hypertrophy; finally, to assess the prognostic effect of the ASE/EAE proposed classification scheme, the risk for adverse clinical outcomes was assessed among subjects in the four sex-specific proposed categories of indexed LV mass. For this purpose, we investigated whether the risk for adverse events differed among ASE/EAE categories using different multivariate statistical models: multiple-category models, in which the risk for adverse outcome in each category was compared with that associated with normal indexed LV mass, which served as the reference group (HR, 1), and trend models, in which we investigated whether there was a stepwise increase in the risk for adverse outcomes from one category to the next higher one.
Multivariate analyses included the following covariates: age (years), gender, atrial fibrillation, LV ejection fraction, WMSI, and significant valvular disease. Ancillary analyses were performed in the subgroup of patients in whom information on risk factors was available, and these covariates were added to the multivariate model. Finally, we tested for interactions between gender and LV hypertrophy grade and between cardiac rhythm and LV hypertrophy grade.
All tests were two tailed. P values < .05 were considered statistically significant. All analyses were performed using SPSS version 15.0 for Windows (SPSS, Inc., Chicago, IL).
During the study period, 2,545 subjects (mean age, 61.9 ± 15.8 years; 56.3% women) met the inclusion criteria and were considered for the analysis. The mean LV mass indexed to BSA was 107.5 ± 37.3 g/m 2 among all subjects enrolled. According to the ASE/EAE guideline cutoffs for LV mass/BSA, 1,198 patients (47.1%) were classified as having LV hypertrophy, and 1,347 (52.9%) had normal LV mass indexes. LV hypertrophy was mild in 391 patients (15.4%), moderate in 307 (12.1%), and severe in 500 (19.6%). Patients with higher categories of indexed LV mass were older, had higher body mass indexes, and were more likely to be women and to have hypertension, diabetes mellitus, hyperlipidemia, valvular heart disease, and atrial fibrillation, while there was no difference in smoking status. Higher degrees of LV hypertrophy were associated with worse systolic function and larger left atria ( Table 1 ).
( n = 2,545)
|No LVH |
( n = 1,347)
|Mild LVH |
( n = 391)
|Moderate LVH |
( n = 307)
|Severe LVH |
( n = 500)
|P for trend|
|Baseline clinical characteristics|
|Age (y)||61.9 ± 15.8||67.6 ± 16.5||64.9 ± 14.3||66.3 ± 13.0||68.6 ± 12.9||<.001|
|Women||1434 (56.3%)||725 (53.8%)||209 (53.5%)||176 (57.3%)||324 (64.8%)||<.001|
|BSA (m 2 )||1.82 ± 0.22||1.83 ± 0.22||1.82 ± 0.21||1.83 ± 0.22||1.78 ± 0.20||<.001|
|BMI (m/kg 2 )||26.2 ± 4.6||25.4 ± 4.6||27.1 ± 4.4||26.6 ± 4.3||27.2 ± 4.6||<.001|
|Hypertension ( n = 1,851)||1325 (71.6%)||553 (60.2%)||236 (80.0%)||193 (80.8%)||343 (86.2%)||<.001|
|Diabetes mellitus ( n = 1,851)||228 (12.3%)||92 (10.0%)||36 (12.2%)||37 (15.5%)||63 (15.8%)||.001|
|Hyperlipidemia ( n = 1,850)||444 (24.0%)||194 (21.1%)||77 (26.1%)||60 (25.1%)||113 (28.4%)||.004|
|Smoking status ( n = 1,851)||187 (10.1%)||100 (10.9%)||29 (9.8%)||23 (9.6%)||35 (8.8%)||.154|
|Prior acute coronary syndromes||230 (9.0%)||98 (7.3%)||42 (10.7%)||27 (8.8%)||63 (12.6%)||.001|
|History of chronic/stable coronary artery disease||309 (12.1%)||125 (9.3%)||46 (11.8%)||51 (16.6%)||87 (17.4%)||<.001|
|History of heart failure||208 (8.2%)||75 (5.6%)||35 (9.0%)||35 (11.4%)||63 (12.6%)||<.001|
|Prior stroke||42 (1.7%)||17 (1.3%)||3 (0.8%)||7 (2.3%)||15 (3.0%)||.007|
|Significant valvular heart disease||274 (10.8%)||72 (5.3%)||42 (10.7%)||44 (14.3%)||116 (23.2%)||<.001|
|Atrial fibrillation||196 (7.7%)||76 (5.6%)||32 (8.2%)||27 (8.8%)||61 (12.2%)||<.001|
|LVEF (%)||64.6 ± 12.2||66.1 ± 10.5||64.0 ± 12.0||64.0 ± 12.7||61.6 ± 15.2||<.001|
|WMSI||1.07 ± 0.24||1.03 ± 0.14||1.06 ± 0.21||1.09 ± 0.28||1.18 ± 0.39||<.001|
|LAVI (mL/m 2 ) ( n = 1,624)||34.7 ± 17.0||29.4 ± 13.5||36.0 ± 16.7||36.2 ± 15.7||46.7 ± 19.7||<.001|
|Normal diastolic function||545 (21.4%)||366 (27.2%)||75 (19.2%)||49 (16.0%)||55 (11.0%)||<.001|
|Grade I diastolic dysfunction||411 (16.1%)||155 (11.5%)||77 (19.7%)||70 (22.8%)||109 (21.8%)||<.001|
|Grade II diastolic dysfunction||504 (19.8%)||233 (17.3%)||94 (24.0%)||64 (20.8%)||113 (22.6%)||<.001|
|Grade III or IV diastolic dysfunction||32 (1.3%)||11 (0.8%)||3 (0.8%)||7 (2.3 %)||11 (2.2 %)||<.001|
|Undetermined diastolic dysfunction||1053 (41.4%)||582 (43.2%)||142 (36.3 %)||117 (38.1%)||212 (42.4%)||<.001|
|RWT > 0.42||1007 (39.6%)||321 (23.8%)||193 (49.4%)||169 (55.0%)||324 (64.8%)||<.001|
|Medications||440 (17.3%)||231 (17.1%)||69 (17.6%)||50 (16.3%)||90 (18.0%)||.796|
|Antiplatelet agents||469 (18.4%)||250 (18.6%)||75 (19.2%)||48 (15.6%)||96 (19.2%)||.911|
|β-blockers||333 (13.1%)||156 (11.6%)||54 (13.8%)||37 (12.1%)||86 (17.2%)||.004|
|Diuretics||262 (10.3%)||140 (10.4%)||35 (9.0%)||34 (11.1%)||53 (10.6%)||.823|
|Calcium channel blockers||741 (29.1%)||379 (28.1%)||114 (29.2%)||91 (29.6%)||157 (31.4%)||.170|
|ACE inhibitors/ARBs||327 (12.8%)||174 (12.9%)||52 (13.3%)||41 (13.4%)||60 (12.0%)||.698|
|Statins||2 (0.1%)||1 (0.1%)||0||0||1 (0.2%)||.543|
|Antiarrhythmic agents (class IA, IB)||26 (1.0%)||17 (1.3%)||3 (0.8%)||2 (0.7%)||4 (0.8%)||.278|
|Antiarrhythmic agents (class IC)||18 (0.7%)||11 (0.8%)||2 (0.5%)||2 (0.7%)||3 (0.6%)||.592|
|Amiodarone||117 (4.6%)||60 (4.5%)||20 (5.1%)||13 (4.2%)||24 (4.8%)||.823|
|Insulin/oral hypoglycemic drugs||174 (6.8%)||87 (6.5%)||21 (5.4%)||18 (5.9%)||48 (9.6%)||.051|
|Anticoagulants||440 (17.3%)||231 (17.1%)||69 (17.6%)||50 (16.3%)||90 (18.0%)||.796|
After a mean follow-up period of 2.5 ± 1.2 years, death occurred in 121 patients (4.8%). Greater LV mass was associated with a linear increase in the risk for death (for each 1 g/m 2 increase in LV mass index, there was a 1.2% increase in the risk for death (HR, 1.012; 95% CI, 1.01–1.015; P < .0001), and this association remained significant on multivariate analysis adjusting for age, gender, atrial fibrillation, LV ejection fraction, WMSI, and significant valvular disease (adjusted HR, 1.007; 95% CI, 1.003–1.011; P < .0001). Patients with LV hypertrophy had significantly worse survival. Throughout the follow-up period, the presence of LV hypertrophy conferred a threefold increased risk for death (HR, 3.1; 95% CI, 2.07-4.65; P < .0001), which remained significant after multivariate adjustment for age, gender, atrial fibrillation, LV ejection fraction, WMSI, and significant valvular disease (adjusted HR, 2.14; 95% CI, 1.39–3.29; P = .001). There was a strong graded association between the severity of LV hypertrophy and survival ( P < .0001; Figure 1 ). At 3 years, Kaplan-Maier estimated survival was 97.4 ± 0.5% in patients with normal LV mass and 94.3 ± 1.5% in those with mild, 91.5 ± 2.0% in those with moderate, and 90.9 ± 1.5% in those with severe LV hypertrophy. Compared with patients with normal LV mass, the risk for death was greater than twofold in patients with mild LV hypertrophy (HR, 2.17; 95% CI, 1.23–3.81; P = .007), greater than threefold in those with moderate LV hypertrophy (HR, 3.04; 95% CI, 1.76–5.23; P < .0001), and almost fourfold increased in those with severe LV hypertrophy (HR, 3.81 95% CI, 2.43–5.97; P < .0001) ( Table 2 ). After adjusting for age, gender, atrial fibrillation, LV ejection fraction, WMSI, and significant valvular disease, the strong association between LV hypertrophy severity grade and death remained significant (for mild LV hypertrophy: adjusted HR, 1.81; 95% CI, 1.03–3.20; P = .040; for moderate LV hypertrophy: adjusted HR, 2.31; 95% CI, 1.33–4.01; P = .003; and for severe LV hypertrophy: adjusted HR, 2.30; 95% CI, 1.39–3.79; P < .0001 compared with normal LV mass) ( Table 2 ). When RWT was added to this multivariate model, the results were similar (for mild LV hypertrophy: adjusted HR, 1.63; 95% CI, 0.91–2.91; P = .099; for moderate LV hypertrophy: adjusted HR, 2.01; 95% CI, 1.13–3.57; P = .017; and for severe LV hypertrophy: adjusted HR, 1.92; 95% CI, 1.13–3.28; P = .016 compared with normal LV mass).
|Variable||Unadjusted HR||95% CI||P||Adjusted HR ∗||95% CI||P|