Left Ventricular Mass in the Elderly
The Cardiovascular Health Study
Abstract Left ventricular (LV) mass, as estimated from M-mode echocardiography (echo), has previously been shown to be an independent predictor of incident cardiovascular disease morbidity and mortality. We evaluated the relationship at baseline of echo LV mass to relevant cardiovascular disease risk factors and other potential covariates in the Cardiovascular Health Study, multicenter study sponsored by the National Heart, Lung, and Blood Institute of 5201 men and women aged 65 years or older (mean, 73). Two-dimensionally directed M-mode echo LV mass measurements could be obtained in 1357 men and 2053 women (66% of this elderly cohort). Stepwise linear regression analyses of the relationship of echo LV mass to demographic and risk factor, physical activity, electrocardiographic, and prevalent disease variables resulted in a model that explained 37% of the variance for the entire cohort. In order of decreasing importance, factors positively associated with echo LV mass were body weight, male sex, systolic pressure, presence of congestive heart failure, present smoking, major and minor electrocardiographic abnormalities, treatment for hypertension, valvular heart disease, aortic regurgitation by color Doppler, and mitral regurgitation by color Doppler (in men) whereas diastolic pressure, bioresistance (a measure of adiposity), and high-density lipoprotein cholesterol were inversely related to echo LV mass. Although height and weight were both related to LV mass, height added nothing once weight was entered in multiple linear regression analyses. Furthermore, in the multiple regression models, diastolic pressure was inversely and systolic BP positively related to LV mass, with similar magnitudes for their coefficients. In consonance with these findings, pulse pressure was positively related to LV mass in bivariate analyses. Multiple linear regression analyses explained less of the variance for ventricular septal thickness (R2=.13) and LV posterior wall thickness (R2=.14) than for LV mass (R2=.37) and LV diastolic dimension (R2=.27). Intriguing findings in the elderly Cardiovascular Health Study cohort included the presence of pulse pressure as a positive correlate, and high-density lipoprotein cholesterol as an inverse correlate, of LV mass. Longitudinal studies in the Cardiovascular Health Study cohort will help to clarify the importance of demographic, risk factor, and other variables, and changes in these variables, in predicting changes in echo LV mass and its components as well as the prognostic significance of LV mass in the elderly.
Epidemiological data from the Framingham Heart Study indicate that LV hypertrophy, as measured by M-mode echocardiography (echo), is an independent predictor of mortality and morbidity from CHD during a 4-year follow-up period.1 2 After adjustment for age, diastolic BP, pulse pressure, treatment for hypertension, cigarette smoking, diabetes, obesity, and the ratio of total cholesterol to HDL cholesterol, LV mass divided by height was associated with a mean relative risk of 1.73 in men and 2.12 in women for incidence of death from cardiovascular disease. However, echo data have not been available from large multicenter, population-based studies of the free-living elderly. Furthermore, although previous studies have emphasized the important predictive value of echo LV mass for CHD mortality and morbidity, there is a paucity of published information in the elderly relating LV mass to potential explanatory factors.
Recent baseline echo data from the CHS have revealed that after adjustment for weight, BP, and other covariates, M-mode (two-dimensionally directed) LV mass is significantly higher in elderly men than elderly women, regardless of the presence of clinical heart disease (both P<.001).3 In addition, a weak relationship between age and M-mode LV mass was detected. The purpose of this report is to extend these findings by describing cross-sectional associations of baseline cardiovascular disease risk factors and other covariates with CHS baseline M-mode echo LV mass and its component variables.
CHS is a prospective, population-based study of 5201 participants (2246 men and 2955 women) aged 65 years and older. Of this cohort, 4850 participants were white, 307 were black, and 44 were classified as “other non-white.” Ages ranged from 65 to 100 years in both men and women (mean±SD, 73.3±5.8 in men and 72.4±5.4 in women). CHS participants were recruited and examined at four Field Centers: Forsyth County, NC; Sacramento County, Calif; Allegheny County, Pa; and Washington County, Md. The overall design, objectives, and recruitment strategy of CHS and characteristics of the cohort recruited have been presented in detail elsewhere.4 5
Echocardiography Performance and Reading Protocol
Echocardiography was performed during a CHS examination that included, as previously described,4 interviews related to medical history, physical activity, personal habits, cognitive function, and dietary intake; anthropometry; recumbent, sitting, and standing BP; resting and ambulatory ECG; spirometry; carotid ultrasound; and laboratory studies. The design of the echo protocol used in CHS has been described in detail elsewhere.6 Briefly, for each subject, a baseline echo was recorded onto super-VHS tape using a standardized protocol. Measurements were made at the Echocardiography Reading Center at the University of California, Irvine, from digitized images using an off-line image-analysis system equipped with customized computer algorithms. Quality-control measures included standardized training of echo technicians and readers, periodic technician observation by a trained echocardiographer, blind duplicate readings to establish interreader and intrareader measurement variabilities, periodic reader review sessions, phantom studies on the ultrasound equipment, and quality-control audits.
This report focuses on two-dimensionally directed M-mode measurements of LV mass and its three component variables: ventricular septal thickness at end diastole (VSTd), LV (internal) dimension at end diastole (LVIDd), and LV posterior wall thickness at end diastole (PWTd). M-mode measurements were made according to conventions established by the American Society of Echocardiography.7 LV mass was derived from the formula described by Devereux and associates8 : LV Mass (grams)=0.80×1.04[(VSTd+LVIDd+PWTd)3−(LVIDd)3]+0.6, where thickness and dimension measurements are expressed in centimeters.
M-mode echocardiograms of adequate quality for performing M-mode measurements of the left ventricle were available in 1357 men and 2053 women in the entire cohort. As previously described, the entire set of M-mode LV measurements could not be made in 34% of the elderly CHS cohort, with age being the strongest correlate of missing data.3
Other Independent Variables
Eligible participants giving informed consent answered standard questionnaires on personal habits, transient symptoms (such as syncope, vertigo, palpitations), family history, and medical history, including recent hospitalizations and prior cardiac diagnoses and procedures.9 A comprehensive examination included blood tests, spirometry, carotid ultrasound, and ECG. Factors considered in the current report as possible explanatory variables for LV mass and its components are listed in Table 1⇓. History of hypertension and valvular disease were defined by self-report. Diabetes was defined as fasting glucose greater than or equal to 140 mg/dL (1.4 g/L), 2-hour postglucose load greater than or equal to 200 mg/dL (2.0 g/L), a reported history of diabetes, or use of insulin or oral hypoglycemic medication. History of MI or CHF were defined as a reported history of the condition confirmed by medical records or examination.9 CHD was defined as (1) reported history of MI, angina pectoris, previous coronary artery bypass surgery, or coronary angioplasty or (2) silent MI discovered during the examination.9
Major and minor ECG abnormalities were defined by the Minnesota code.10 Major ECG abnormalities included one or more of the following: ventricular conduction defect, major Q wave abnormalities, LV hypertrophy, ST-T wave abnormalities, atrial fibrillation, or first-degree atrioventricular block. Minor ECG abnormalities were defined as the presence of one or more of the following: minor Q or QS wave abnormalities, tall R waves, minor isolated ST-T wave abnormalities, ST elevation, incomplete right bundle-branch block, long QT or short PR interval, or right axis deviation.
Descriptive statistics and multiple linear regression analyses were performed with the SPSS statistical package (SPSS for Windows, version 6.1, SPSS Inc). Bivariate correlations were calculated between each of the echo measures and each of the continuous predictors described below under multiple regression analyses. For the dichotomous predictors, means of each echo variable by level of predictor were calculated for descriptive purposes.
Multiple regression models for LV mass, ventricular-septal thickness in diastole, LV posterior wall thickness in diastole, and LV dimension in diastole were generated in two ways: by taking the natural logarithms of all continuous variables, and by using variables in their original units. The results from the two methods were similar. For ease of interpretation, only results of the models using variables in their original units are reported.
Table 1⇑ lists the variables available for entry into the multiple linear regression models. Significant variables were identified by a stepwise procedure with an entry criterion of P<.01. Because any case missing a value for any one of the potential covariates listed in the table was deleted from the model, once the significant variables were identified, they alone were entered into a regression model in order to capture the largest number of cases possible. Of the 3401 participants with LV mass measurements, 2861 (or 84%) had data for all potential covariates listed in Table 1⇑. In models for which sex was a significant predictor, sex interactions with other variables in the model were tested for significance.
Relationship of Disease Status and Other Independent Variables to Echo LV Mass and its Components: Descriptive Data
Tables 2⇓ and 3⇓ provide descriptive data relating various disease states and other independent variables to LV mass and its component variables in men and women. In general, echo LV mass, ventricular septal thickness, and LV posterior wall thickness measurements, unadjusted for variables such as age, BP, and body size, were greater in men than women and in those with than those without a history of hypertension, diabetes, MI, CHF, CHD, or valvular heart disease. In addition, participants demonstrating major and minor ECG abnormalities or using antihypertensive medications had higher unadjusted measurements for LV mass, ventricular septal thickness, LV posterior wall thickness, and LV diastolic dimension than those without these conditions. In general, the differences for LV diastolic dimension were lesser in magnitude than those for the other three echo LV measurements. However, there were larger differences in participants with than those without a history of MI or CHF.
Bivariate Analyses of Independent Variables Versus Echo LV Mass and Components
Table 4⇓ lists independent variables that either demonstrated significant bivariate correlations with LV mass or its components or appeared in the final regression models. The continuous variables most strongly correlated with LV mass and its component variables were the descriptors of body size. LV mass was most strongly correlated with weight (directly: r=.44) and bioresistance (inversely: r=−.44) (both P<.001). Significant correlations with body size variables, including height, waist and hip circumference, and body mass index, were also present for the echo components of LV mass, especially for LV diastolic dimension.
Multiple Linear Regression Analyses for Predicting LV Mass and Its Components
Table 5⇓ presents the multiple linear regression analyses for predicting LV mass and its component variables in the entire cohort. The linear model explained 37% of the variance in LV mass.
Although in addition to height, both waist and hip circumferences, as well as body mass index, were related to LV mass bivariately, these body size variables added nothing once weight was entered in the multiple linear regression models. However, bioresistance remained inversely related to LV mass, as was HDL cholesterol. Furthermore, diastolic BP was inversely and systolic BP positively related to LV mass, with similar magnitudes for their coefficients, consistent with the finding (noted on bivariate analyses) that pulse pressure was a positive predictor of LV mass. Not surprisingly, ECG abnormalities, CHF, valvular heart disease, color Doppler aortic and mitral regurgitation, and use of antihypertensive medications were all associated directly with LV mass. Present smoking was also (weakly) directly associated with LV mass.
In the model for LV mass, there were two significant sex interaction terms: major ECG abnormalities and mitral regurgitation. Since sex was coded “1” for male and “0” for female, females were described by the main effects alone and males were described by both the main effect and the interaction term. Major ECG abnormalities were a significant correlate of LV mass in both men and women: In women, it was associated with an 11.76-g increment in LV mass (all other factors held constant) and in men with an increment of 11.76+13.64=25.40 g. Mitral regurgitation was a significant correlate of LV mass only in men.
The linear model for predicting LV diastolic dimension in the overall cohort explained 27% of the variance. Height entered the model for LV diastolic dimension but not for LV mass or septal or posterior wall thicknesses. Furthermore, age was inversely related to LV diastolic dimension but unrelated to LV mass in multivariate models. Other variables related positively to LV diastolic dimension included body weight, systolic BP, ECG abnormalities, confirmed MI or CHF, and color Doppler aortic regurgitation ratio or mitral regurgitation ratio (in men). Diastolic BP and bioresistance were inversely related to LV diastolic dimension.
The overall multiple linear regression analyses explained less of the variance for ventricular septal thickness (13%) and LV posterior wall thickness (14%) than for LV mass and LV diastolic dimension. For both ventricular septal thickness and LV posterior wall thickness, the models did not differ by sex (and sex interaction terms were nonsignificant). Ventricular septal thickness was positively associated with body weight, systolic BP, history of hypertension, age, major and minor ECG abnormalities, present smoking, and color Doppler aortic regurgitation ratio and inversely related to bioresistance and HDL cholesterol. LV posterior wall thickness was positively associated with body weight, male sex, age, systolic BP, hypertensive medication use, serum albumin, major and minor ECG abnormalities, vasodilator use, and color Doppler aortic regurgitation ratio.
The results of our study reveal that LV mass and its component measurements, as derived from two-dimensionally directed M-mode echo, are related to variables in the following categories: (1) body size and obesity, (2) age, (3) sex, (4) BP, history of hypertension, and treatment for hypertension, (5) valvular heart disease by history or by color Doppler echo, (6) CHF, MI, or ECG abnormalities (major and minor), (7) history of smoking, (8) HDL cholesterol, and (9) hematocrit and clotting factors, serum albumin, and uric acid.
Body Size and Obesity
In bivariate analyses, as well as in multiple linear regression analyses, body weight was generally the strongest correlate of LV mass. In bivariate models, height, waist and hip circumferences, body mass index, and bioresistance (a measure of adiposity) were all related to LV mass and to most of its component variables. However, after weight was entered into multiple linear regression analyses for LV mass and each of its component variables, height, waist and hip circumferences, and body mass index no longer entered into the models (except for height in the model for LV diastolic dimension). For each 5-kg increase in body weight, with other factors constant, LV mass increased, on average, by 1.0 g in the CHS cohort (partial β=0.21). In contrast, bioresistance entered models (inverse relationship) for LV mass, ventricular septal thickness, and LV dimension. The relationship between total body weight, lean body mass, and body fat (or adiposity) can be expressed by the following conceptual equation: Total Body Weight=Lean Body Mass+Body Fat (or Adiposity), where total body weight is measured directly; waist and hip circumferences, or bioresistance, are surrogate measures for body fat; and height is an imperfect measure of lean body mass. This formula explains why, in multiple linear regression analyses when weight is allowed to enter the model, measures such as height, waist and hip circumferences, and body mass index become less significant or nonsignificant explanatory variables: their explanatory power is partially or completely subsumed by the total body weight variable. Furthermore, our models are consistent with the concept that the combination of total body weight (directly related to LV mass and its components) and bioresistance (inversely related to LV mass and two of its components) may actually be measuring lean body mass and reflecting its importance as a determinant of LV mass.
Our models are consistent with previous reports emphasizing the strong relationship of LV mass to measures of body size (eg, weight and height), lean body mass, and obesity.11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 In CARDIA, a multicenter study sponsored by the National Heart, Lung, and Blood Institute, in multiple linear regression analyses for LV mass in which weight was entered, height was no longer a significant explanatory variable.13 Daniels et al,29 using dual-energy x-ray absorptiometry to determine lean body mass, recently reported in 201 children and adolescents that lean body mass is a more important determinant of LV mass than are either obesity (fat mass) or systolic BP. Furthermore, even in a hypertensive population with a high prevalence of LV hypertrophy, Gottdiener et al27 have reported that body weight is a more important determinant of LV mass than is BP.
In multiple linear regression analyses, when other variables were allowed to enter the model, age was no longer a significant explanatory variable for LV mass. The absence of an association of age and LV mass may be due to the highly truncated age range; ie, all CHS subjects were at least 65 years old. In contrast, age was a significant, but relatively minor, explanatory variable for both ventricular septal thickness and LV posterior wall thickness (positive relationship) and for LV dimension (inverse relationship). We have previously shown that among CHS participants, weight-adjusted LV mass increases modestly (<1 g per year, P<.0001) with increasing age.3 Ventricular septal thickness and LV posterior wall thickness have been previously shown in smaller studies to increase with age.14 15 In these same studies, LV diastolic dimension decreased slightly with age over a wide adult age range. Not surprisingly, a positive relationship of age and LV relative wall thickness has been previously reported by a number of investigators.22 31 32 33
Sex Differences in Multivariate Associations
Multivariate analyses revealed significantly higher LV mass (by a mean of 10.6 g) and LV posterior wall thickness (by a mean of 0.3 mm) in men than women, even after adjustment for body size variables. In the model for LV mass, there were two sex interaction terms: major ECG abnormalities and mitral regurgitation. Major ECG abnormalities were significantly associated with LV mass in both men and women, whereas mitral regurgitation was significantly associated only in men.
The explanation for these sex-specific differences in association with LV mass and its components is not definitively provided by the current study. Possible explanations for sex differences in associations include actual biological differences in echo LV mass or LV posterior wall thickness measurements, or survival bias among women. Consistent with the explanation that a true sex-related biological difference may be present in the elderly CHS cohort, various workers have recently reported that heart size and the ratio of LV mass to height are larger in boys than girls34 and that LV mass is greater in younger adult men versus women.13 35 One might also observe sex-related differences if measurements detecting similar phenomena, and having similar relationships to LV mass variables, differed in prevalence or measurement accuracy by sex, thus affecting statistical power. Finally, apparent sex differences for some variables could be due to chance (when their true contribution is zero) because of the multiple testing involved in deriving the best models for LV mass and its component variables. Longitudinal follow-up of the CHS cohort and sex-specific analyses in other cohorts may help to shed further light on the etiologies for these sex-associated differences in LV mass and its components.
BP, History of Hypertension, and Treatment for Hypertension
Multiple linear regression analyses for LV mass (and LV dimension) revealed direct relationships between systolic BP and LV mass (and internal dimension) and inverse relationships, approximately equal in magnitude, between diastolic BP and LV mass (and internal dimension). For diastolic BP, this represents a reversal of the direction of correlation with LV mass (which was weakly positive) in bivariate analyses. For each 10–mm Hg increment in systolic BP, with all other factors kept constant, LV mass was greater, on average, by 3.0 g in the CHS cohort. These findings were also present in multiple linear regression analyses in a healthy subset of the CHS cohort (data not shown). In consonance with these findings, LV mass and ventricular septal and posterior wall thicknesses were all significantly (P<.001) related to pulse pressure in bivariate models. Since stroke volume is approximated by the formula (LV Diastolic Dimension)3−(LV Systolic Dimension)3, it is inherently highly positively correlated with LV diastolic dimension. In addition, stroke volume was highly positively related to LV mass in our study (P<.0001, unadjusted for other variables). Ventricular septal thickness and LV posterior wall thickness were directly related to systolic BP but unrelated to diastolic BP in multiple linear regression analyses.
One possible explanation of the direct relation of pulse pressure but not diastolic BP to LV mass and its components is that in older individuals (eg, ≥65 years), increasing systemic arterial stiffness may result primarily in increases in systolic BP as well as in LV mass and its component septal and wall thicknesses.36 At a given vascular resistance, if the rigidity of the aorta and other large arteries is augmented, the increase in systolic BP should be quantitatively greater than the decrease in diastolic BP. Stiffening of the arterial walls and alterations of their mechanical properties result in a mismatch between aortic impedance and LV ejection, with other consequences such as increases in LV mass.37 In a small subset of a large French study, increased pulse pressure was shown to have an adverse effect on cardiovascular mortality.38 Previous reports, confirmed by the findings in this large population study, lend support to recent recommendations regarding BP treatment in the elderly, which have suggested focusing primarily on systolic BP (and perhaps pulse pressure) rather than on diastolic BP.39 In the CARDIA study, systolic BP was more strongly correlated with LV mass than was diastolic BP.13 Similarly, in a Framingham substudy of 152 men and 299 women (mean age, 68±6 years), 30-year average systolic BP demonstrated significant bivariate correlations with LV mass (corrected for height) (r=.27 in men and r=.31 in women). Not only were both current and long-term values for systolic BP better predictors of LV mass and wall thickness than were current and long-term values for diastolic BP, but 30-year average systolic BP proved to be a significant independent predictor of LV mass in multiple linear regression analyses (P<.01 in men and women).40
In multiple regression analyses, treatment with antihypertensive medications was directly related to LV mass and weakly (positively) to LV posterior wall thickness, whereas vasodilator treatment was weakly related to posterior wall thickness. A history of hypertension was strongly related to ventricular septal thickness. These relationships were expected in view of the well-known relationships between LV mass and its components and a history of hypertension and BP level.41 42
Valvular Heart Disease
In multiple regression analyses, a history of valvular heart disease and color Doppler aortic regurgitant jet ratio were positively associated with LV mass, whereas aortic regurgitant jet ratio was positively associated with the components of LV mass. In men, mitral regurgitant jet ratio was associated with LV mass and diastolic dimension. These findings are consistent with the well-described LV volume-overload states in mitral and/or aortic regurgitation, which result in increased LV dimensions and mass. In a Framingham Heart Study echo examination of 4976 participants (ages 17 to 90 years), multiple linear regression analyses revealed that in addition to age, systolic BP, and obesity, the presence of valve disease and MI were each independently associated with LV hypertrophy in both men and women.28
Congestive Heart Failure, Myocardial Infarction, and ECG Abnormalities
CHF was significantly associated with LV mass but not with its components in multiple linear regression analyses. Not surprisingly, major and minor ECG abnormalities were associated with LV mass and its components since these ECG scores include indicators of LV voltage. However, somewhat surprisingly, MI was associated only with LV dimension. Since ECG abnormalities entered the models for LV mass and its components, it may be that these abnormalities explained some of the variance that would otherwise have been explained by related clinical conditions, eg, CHF and MI. In previous analyses in the CHS cohort, weight-adjusted M-mode echo LV mass was higher in participants with clinical CHD than in the healthy subgroup. However, the magnitude of this difference in weight-adjusted LV mass related to disease status was not as great as the difference related to sex.3 One possible explanation for the lesser magnitude of this disease effect is the previously reported high prevalence of subclinical CHD in elderly individuals.43
Lung Function and Smoking History
Forced expiratory volume in 1 second and forced vital capacity were directly related to LV mass and diastolic dimension in bivariate analyses—possibly a reflection of the relationship of chest cavity size to body size measures—but not in multivariate models. Multivariate analyses revealed a weak relationship between smoking history and LV mass and ventricular septal thickness. In the CARDIA study, Gidding et al44 found a modest relationship between smoking history and LV mass.
In multiple linear regression analyses, higher levels of HDL cholesterol were associated with lower LV mass and ventricular septal thickness, consistent with these individuals having a lower prevalence of clinical CHD or lower CHD risk. However, these relationships appear to be of modest importance. In the CARDIA cohort of young adults, there was a weak association between total cholesterol and LV mass in black men only.13
Hematocrit and Clotting Factors, Serum Albumin, and Uric Acid
Fibrinogen, factor VII, factor VIII, and hematocrit did not enter any of the multiple linear regression analyses for LV mass or its component variables, despite significant bivariate relationships of hematocrit and factor VII to LV mass and its component variables. Of interest, deSimone and associates45 reported a relationship between blood viscosity, as indirectly measured by hematocrit, and LV mass. This relationship did not appear to be an important one in the elderly CHS cohort. Serum albumin proved to be an explanatory variable of modest importance in the multiple linear regression model for LV posterior wall thickness. Serum albumin was modestly related to LV posterior wall thickness in multivariate models, whereas uric acid was related to LV mass and its components only in bivariate analyses. The correlations between uric acid and LV echo measurements were reduced by adjusting for sex, age, height, and weight, but they do not disappear (for LV mass: r=.06, P<.001).
Strengths and Weaknesses
Strengths of the current study include the large numbers of men and women studied as well as the well-defined and consistently measured explanatory variables. Weaknesses of the present study include the fact that in this elderly cohort, M-mode echo data were missing in 34% of the participants in the entire cohort, with the percent missing data increasing with increasing age from 29% in the 65-to-69 year age group to 50% in the 85+ year age group (P<.001).3 Another inherent limitation relates to the cross-sectional design of this analysis, so that only associations, rather than predictions, can be derived from the multiple linear regression analyses. Clearly, follow-up longitudinal studies are necessary to delineate the potential predictive value of demographic, risk factor, and other variables studied in the current report for echo changes in LV mass and component variables.
Although it is known that M-mode may have limitations compared with two-dimensional echo measurements of the left ventricle in individuals with CHD and CHF due to segmental wall motion differences, Devereux et al8 have previously shown a good correlation between M-mode echo and necropsy LV mass measurements, even in individuals with CHD. Two-dimensional echo measurements have their own limitations, eg, LV endocardial dropout in apical views46 In a pilot study in CHS, two-dimensional echo resulted in poorer measurement yield and higher reader variability for LV measurements than did two-dimensionally directed M-mode echo (data not shown). Therefore, in CHS, as in Framingham, echo measurements of LV mass, dimension, and wall thickness were carried out with the two-dimensionally directed M-mode technique.1 2 6 21 22 23 28
In conclusion, in the CHS multicenter cohort of elderly community-dwelling individuals, demographic, risk factor, and disease status variables such as body weight, male sex, systolic BP, pulse pressure, presence of CHF, present smoking, major and minor ECG abnormalities, treatment for hypertension, and mitral and aortic valve regurgitation by color Doppler are positively related, whereas diastolic BP, HDL cholesterol, and bioresistance are inversely related, to echo LV mass and to one or more of its components variables. Multiple linear regression models in this elderly cohort explain less of the variance for ventricular septal thickness (R2=.13) and LV posterior wall thickness (R2=.14) than for LV mass (R2=.37) and LV dimension (R2=.27). New cross-sectional findings in the elderly CHS cohort include the presence of pulse pressure as an independent positive correlate and HDL cholesterol as an inverse correlate of LV mass in multiple linear regression models. Longitudinal studies are currently in progress to delineate the potential value of demographic, risk factor, and other variables, and changes in these variables, in predicting changes in LV mass and its component variables as well as the prognostic significance of LV mass in the elderly.
Selected Abbreviations and Acronyms
|CARDIA||=||Coronary Artery Risk Development in Young Adults|
|CHD||=||coronary heart disease|
|CHF||=||congestive heart failure|
|CHS||=||Cardiovascular Health Study|
This work was supported by contract Nos. N01-HC-85079 to HC-85086 from the National Heart, Lung, and Blood Institute, Bethesda, Md.
Reprint requests to CHS Coordinating Center, University of Washington, JD-30, 1107 NE 4th St, Rm 530, Seattle, WA 98105.
- Received September 17, 1996.
- Revision received October 7, 1996.
- Accepted November 5, 1996.
Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Left ventricular mass and incidence of coronary heart disease in an elderly cohort. The Framingham Heart Study. Ann Intern Med. 1989;110:101-107.
Gardin JM, Siscovick D, Anton-Culver H, Smith V-E, Klopfenstein S, Bommer W, Fried LP, O’Leary D, Manolio TA. Sex, age, and disease affect echocardiographic left ventricular mass and systolic function in the free-living elderly: the Cardiovascular Health Study (CHS). Circulation. 1995;91:1739-1748.
Fried L, Borhani N, Enright P, Furberg C, Gardin J, Kronmal R, Kuller L, Manolio T, Mittelmark M, Newman A, O’Leary D, Psaty B, Rautaharju P, Tracy R, Weiler P, for the CHS Collaborative Research Group. Cardiovascular Health Study: design and rationale. Ann Epidemiol. 1991;1:263-276.
Sahn DJ, DeMaria A, Kisslo J, Weyman A. The Committee on M-mode Standardization of the American Society of Echocardiography. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic methods. Circulation. 1978;58:1072-1083.
Mittelmark MB, Psaty BM, Rautaharju P, Fried LP, Tracy RP, Gardin JM, Borhani NO, O’Leary DH, for the CHS Collaborative Research Group. Prevalence of cardiovascular diseases among older adults: the Cardiovascular Health Study. Am J Epidemiol. 1993;137:311-317.
Blackburn H, Keys A, Simonson E, Rautaharju P, Punsar S. The electrocardiogram in population studies: a classification system. Circulation. 1960;21:1160-1175.
Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Intern Med. 1991;114:345-352.
Gardin JM, Wagenknecht LE, Anton-Culver H, Flack J, Gidding S, Kurosaki T, Wong ND, Manolio TA. Relationship of cardiovascular risk factors to echocardiographic left ventricular mass in healthy young black and white adult men and women: the CARDIA Study. Circulation. 1995;92:380-387.
Henry WL, Gardin JM, Ware JH. Echocardiographic measurements in normal subjects from infancy to old age. Circulation. 1980;62:1054-1061.
Ghali JK, Liao Y, Simmons B, Castaner A, Cao G, Cooper RS. The prognostic role of left ventricular hypertrophy in patients with or without coronary artery disease. Ann Intern Med. 1992;117:831-836.
Burke GL, Arcilla RA, Culpepper WS, Webber LS, Chiang Y-K, Berenson GS. Blood pressure and echocardiographic measures in children: the Bogalusa Heart Study. Circulation. 1987;75:106-114.
Savage DD, Garrison RJ, Kannel WB, Levy D, Anderson SJ, Stokes J, Feinleib M, Castelli WP. The spectrum of left ventricular hypertrophy in a general population sample: the Framingham Study. Circulation. 1987;75(suppl I):I-26-I-32.
Mahoney LT, Schieken RM, Clarke WR, Lauer RM. Left ventricular mass and exercise responses predict future blood pressure: the Muscatine Study. Hypertension. 1988;12:206-213.
Schieken RM, Clarke WR, Lauer RM. Left ventricular hypertrophy in children with blood pressures in the upper quintile of the distribution: the Muscatine Study. Hypertension. 1981;3:669-675.
Levy DM. Echocardiographically detected left ventricular hypertrophy: prevalence and risk factors. Ann Intern Med. 1988;108:7-13.
Daniels SR, Kimball TR, Morrison JA, Khoury P, Witt S, Meyer RA. Effect of lean body mass, fat mass, blood pressure, and sexual maturation on left ventricular mass in children and adolescents: statistical, biological, and clinical significance. Circulation. 1995;92:3249-3254.
Schieken RM. Large hearts in children: biology or disease? Circulation. 1995;92:3156-3157. Editorial.
Urbina EM, Gidding SS, Bao W, Pickoff AS, Berdusis K, Berenson GS. Effect of body size, ponderosity, and blood pressure on left ventricular growth in children and young adults in the Bogalusa Heart Study. Circulation. 1995;91:2400-2406.
Marcus R, Krause L, Weder AB, Dominguez-Meja A, Schork NJ, Julius S. Sex-specific determinants of increased left ventricular mass in the Tecumseh Blood Pressure Study. Circulation. 1994;90:928-936.
Lakatta EG. Do hypertension and aging have similar effects on the myocardium? Circulation. 1987;75(suppl I):I-69-I-67.
Darne’ B, Girerd X, Safar ME, Cambien F, Guizel L. Pulsatile versus steady component of blood pressure: a cross sectional and a prospective analysis on cardiovascular mortality. Hypertension. 1989;13:392-400.
Celentano A, Galderisi M, Garofalo M, Mureddu GF, Tammaro P, Petitto M, Di Somma S, de Divitiis O. Blood pressure and cardiac morphology in young children of hypertensive subjects. J Hypertens. 1988;6(suppl 4):S107-S109.
Elveback L, Lie JT. Continued high incidence of coronary artery disease at autopsy in Olmsted County, Minnesota, 1950 to 1979. Circulation. 1984;70:345-349.
deSimone G, Devereux RB, Chien S, Alderman MH, Atlas SA, Laragh JH. Relation of blood viscosity to demographic and physiologic variables and to cardiovascular risk factors in apparently normal adults. Circulation. 1990;81:107-117.