Relationship of Cardiovascular Risk Factors to Echocardiographic Left Ventricular Mass in Healthy Young Black and White Adult Men and Women
The CARDIA Study
Background The objective of this study was to describe the distribution of echo left ventricular (LV) mass and its association with demographic and cardiovascular risk factors in a large race- and sex-balanced cohort of young adults. Recent epidemiological data have suggested that M-mode echocardiographically determined LV hypertrophy is an independent predictor of mortality and morbidity from coronary heart disease (CHD) in older adults. Echocardiographic LV mass has been associated in middle-aged and older adults with multiple factors including age, arterial blood pressure, body mass, and sex. However, there are few data describing the distribution of echo LV mass among black and white young adult men and women and relating LV mass to cardiovascular disease risk factors within race-sex subgroups.
Methods and Results CARDIA (Coronary Artery Risk Development in Young Adults) is a multicenter study of young adults, including approximately equal proportions of black and white men and women aged 23 to 35 years at the time of echo examination (1990 through 1991). Two-dimensionally guided M-mode echocardiograms were attempted in 4243 participants with recordings deemed acceptable for calculation of LV mass, that is, of at least fair quality score, obtained in 3840 (90.5% of the 1990-1991 cohort). M-mode LV mass was calculated from the formula of Devereux and Reichek, adapted for use with measurements made according to the American Society of Echocardiography Standards. LV mass was greater in men than in women and greater in blacks than in whites (P<.001) (mean±SD): black men, 176±42 g; white men, 169±40 g; black women, 135±38 g; and white women, 125±33 g. In all race-sex groups, LV mass was positively correlated (P<.0001) in bivariate analyses with body weight, subscapular skinfold thickness, height, and systolic blood pressure. In multivariate analyses, LV mass remained independently and positively related to body weight and systolic blood pressure and, when body weight was not considered, with subscapular skinfold thickness and height. In addition, the multivariate models allowed us to infer a direct relation between LV mass and both fatness and lean body mass. Weaker positive associations were noted of LV mass with pulse pressure in white participants and with physical activity in men. After adjustment for subscapular skinfold thickness, height, systolic and diastolic blood pressures, alcohol consumption, pulmonary function, smoking history, physical activity, total serum cholesterol, and family history of hypertension, LV mass remained higher in men than in women (P<.0001), in black men (167±43 g) than in white men (156±50 g, P<.0001), and in black women (142±49 g) than in white women (137±43 g, P<.002).
Conclusions In the healthy young adults of the CARDIA cohort, LV mass was highly correlated with body weight, subscapular skinfold thickness, height, and systolic blood pressure across race and sex subgroups. Furthermore, after adjustment for anthropometric, blood pressure, and other covariates, LV mass remained higher in men than in women and in blacks than in whites. Longitudinal studies are necessary to delineate the possible roles of these factors in the genesis of LV hypertrophy.
Recent epidemiological data have suggested that M-mode echocardiographically determined left ventricular (LV) hypertrophy is an independent predictor for mortality and morbidity from coronary heart disease (CHD) in older adults.1 2 Echocardiographic LV mass has been associated in middle-aged and older adults with multiple factors, including age, arterial blood pressure, body mass, and sex.3 4 5 6 7 8 9 However, there are few data relating these factors as well as race, blood pressure, anthropometric measurements, and other factors to echocardiographic LV mass in young adults.
CARDIA (Coronary Artery Risk Development in Young Adults) is a prospective multicenter epidemiological study of young adults aged 23 to 35 years at the time of their first echocardiographic examination. The purpose of the current study was to evaluate the relation of race, sex, arterial blood pressure, anthropometric measures, and measures of physical activity, smoking, alcohol consumption, history of high blood cholesterol, and family history of hypertension to LV mass as determined by M-mode echocardiography in the CARDIA young adult population.
The CARDIA cohort initially comprised 5115 participants who were aged 18 to 30 years at the time of enrollment (1985 through 1986), 5 years before the echocardiography examination. The CARDIA cohort included participants recruited and examined at four field centers located in Birmingham, Ala; Chicago, Ill; Minneapolis, Minn; and Oakland, Calif. An echocardiography reading center was located at the University of California, Irvine, and a coordinating center at the University of Alabama in Birmingham. The overall design and objectives of the CARDIA study have been presented in detail elsewhere.10 11 Echocardiograms were initially performed in 4243 participants, aged 23 to 35 years, as part of the third CARDIA examination in 1990 through 1991. This cohort comprised 874 black men, 1034 white men, 1176 black women, and 1159 white women.
Two-dimensionally directed M-mode echocardiograms were attempted in all 4243 participants using a protocol similar to that used in the Cardiovascular Health Study, a National Heart, Lung, and Blood Institute–sponsored, prospective multicenter study of free-living elderly subjects above the age of 65 years.12 13 Briefly, for each subject, a baseline echocardiogram was recorded onto super-VHS tape using an Acuson cardiac ultrasound machine and a standardized recording protocol. Thirty minutes was allotted to each subject for obtaining M-mode, two-dimensional, spectral, and color Doppler studies. Videotapes were forwarded from the field centers to the echocardiography reading center, where images were displayed and digitized. Measurements were made from digitized images using a Dextra D-200 off-line image analysis system, equipped with customized computer algorithms. Quality control measures for echocardiography included standardized training of echo technicians and readers, technician observation by a trained echocardiographer, blind duplicate readings to establish interreader and intrareader measurement variability, periodic reader review sessions, phantom studies on the ultrasound equipment, and quality control audits.12
This report focuses on measurements of LV mass derived from two-dimensionally guided M-mode echocardiograms. M-mode measurements were made according to conventions established by the American Society of Echocardiography.14 LV mass was derived from the formula described by Devereux and associates15 :
where VSTd is ventricular septal thickness at end diastole, LVIDd is LV internal dimension at end diastole, and PWTd is LV posterior wall thickness at end diastole.
With respect to M-mode LV image quality, if right and left septal and endocardial and epicardial LV posterior wall interfaces could be seen along the entire cycle from which measurements were made, studies were scored as “excellent” (publication quality) or “good” (slightly less optimal edge definition). A score of “fair” implied that at least 5 mm of continuous interface could be seen for each of these four interfaces on contiguous beats but not necessarily on the same beat. A score of “poor” implied that portions of each interface sufficient to make a measurement could be extrapolated from three consecutive cycles. The LV was scored as “unmeasurable” if one or more of the four (septal or LV posterior wall) interfaces could not be identified.
Factors previously shown to have an association with LV mass or with cardiovascular risk were measured contemporaneously with the echocardiographic study. The following covariates, chosen on an a priori basis of suspected association with LV mass, were analyzed in relation to LV mass: (1) race, (2) sex, (3) age (in years), (4) systolic and diastolic blood pressures (in millimeters of mercury) corresponding to the average of the second and third resting, seated cuff blood pressures in systole and diastole, respectively, measured after a 5-minute rest, (5) height (in centimeters), (6) body weight (in pounds), (7) subscapular skinfold thickness—the average of two skinfold measurements (in millimeters) made using Harpenden calipers (with skinfold measurements ≥50 mm set equal to 50), (8) physical activity—the total of the heavy and moderate intensity scores calculated from responses to the CARDIA Physical Activity Questionnaire, a modification of the Minnesota leisure time questionnaire, which included intensity and duration of various activities16 but not occupational physical activity, (9) alcohol consumption—the average weekly consumption of alcohol (in milliliters), (10) cigarette smoking history coded as current, former, and never, with current smokers being defined as those who smoke regularly, at least five cigarettes per week, almost every week, (11) pulmonary forced expiratory volume in 1 second (FEV1 in liters) and forced vital capacity (FVC in liters), (12) total serum cholesterol (in milligrams per deciliter), and (13) family history of hypertension—report of natural mother or natural father or sibling having high blood pressure.
The technical error defined by the formula
was calculated to assess intratechnician and intertechnician performance and reader variability, where di is the difference in paired measurements, n is sample size, and X̄ is the mean of the distribution of measurements for the variable. The technical error can be viewed as a coefficient of variation and is similar to the standard deviation of the measurement differences divided by the mean of the measurements.
Mean unadjusted values (and standard deviations) were initially calculated for LV mass in black and white men and women. Pearson correlation coefficients were then determined individually for the bivariate associations for each race-sex group separately between LV mass and each of the following variables: age, body weight, height, subscapular skinfold thickness, systolic and diastolic blood pressures, FEV1 and FVC, physical activity, alcohol consumption, and total blood cholesterol. ANCOVA was performed to obtain mean values of LV mass among race and sex strata, adjusting for the other covariates noted above. The independent association with LV mass was determined for each of these potential covariates: race, sex, and the race-sex interaction by multiple linear regression. Finally, linear regression coefficients were calculated in each race-sex group describing the independent relation of LV mass to each of the continuous covariates included in the multiple linear regression models. These β-coefficients were used in the following general formula to estimate LV mass from a designated set of values for each of the covariates.
where each of the β-coefficients was multiplied by the designated covariate and β0 is a constant. β-Coefficients were not provided for the smoking history or history of hypertension variables because the β-coefficient estimates for these parameters were not linearly related to LV mass. (BP indicates blood pressure.)
In each of the four race-sex subgroups, body weight and subscapular skinfold thickness were highly related (range of r, .65 to .75). Because of the strong colinearity of body weight and subscapular skinfold thickness as well as our desire to evaluate the relation of measures of lean body mass and fatness to LV mass, race-sex–specific multivariate models were constructed that considered subscapular skinfold thickness (a measure of body fatness) and height with and without body weight included.
Measurement Yield and Variability
M-mode echocardiograms of at least minimal quality to calculate LV mass were obtainable in 94.4% of the CARDIA cohort (4110 participants). For the purposes of this analysis, studies considered to be of poor quality were eliminated: Only the 3840 studies (90.5% of the cohort) scored as fair, good, or excellent were included.
Technical errors for components of variability for LV mass measurements were for intratechnician performance, 10% (from 60 paired studies); for intertechnician performance, 10% (from 44 paired studies); for intrareader, 8% (from 158 paired studies); and for inter-reader, 14% (from 350 paired studies).
Unadjusted Left Ventricular Mass Values by Race-Sex Subgroup
Fig 1⇓ displays unadjusted LV mass in each of the four race-sex subgroups. LV mass was highest in black men, followed by white men, black women, and white women. Differences within sex and within race for unadjusted LV mass were significant (P<.001).
Correlates of Left Ventricular Mass by Race-Sex Subgroups
When analyzed individually by race-sex subgroup, weight was the strongest bivariate correlate of LV mass in each subgroup (Table 1⇓). Subscapular skinfolds, height, systolic blood pressure, FEV1, and FVC were significant but weaker bivariate correlates of LV mass in each race-sex subgroup. Subscapular skinfolds and systolic blood pressure exhibited their strongest correlations with LV mass in black women, followed by white women. FEV1 and FVC were most strongly correlated with LV mass in men, whereas diastolic blood pressure was most strongly correlated with LV mass in women.
As an example of the relations, Fig 2⇓ displays, in each of the four race-sex groups, mean LV mass as a function of systolic blood pressure quartile. In all four race-sex groups, LV mass was significantly higher in the highest (versus the lowest) quartile of systolic blood pressure (range of P, <.005 to .0001).
Distribution of Left Ventricular Mass Adjusted for Covariates Among Race-Sex Subgroups
In each of the race-sex subgroups, LV mass was adjusted for age, height, body weight, subscapular skinfold thickness, systolic blood pressure, diastolic blood pressure, physical activity, alcohol consumption, smoking history, FEV1, FVC, total blood cholesterol, and family history of hypertension. LV mass (adjusted) remained significantly higher in men than in women (P<.0001) and in black men (162±45 g) than in white men (157±45 g, P<.0032). Black and white women exhibited similar adjusted LV mass (mean, 139 to 140 g). When weight was excluded from the variables used to adjust LV mass, adjusted LV mass remained significantly higher in men than in women and in blacks than in whites (both P<.0001). In addition, adjusted LV mass remained significantly higher in black versus white men (167±43 versus 156±50 g, P<.0001) as well as in black versus white women (142±49 versus 137±43 g, P<.002).
Regression Coefficients Relating Left Ventricular Mass to Covariates Within Each Race-Sex Subgroup
In Table 2⇓, regression (β) coefficients are provided describing the relation between LV mass and each of the covariates in multiple linear regression models. These coefficients are expressed separately in each race-sex subgroup, assuming a linear relation between LV mass and each covariate. In multivariate analyses across sex and racial subgroups, body weight and systolic blood pressure were significantly positively correlated, whereas subscapular skinfold thickness was inversely correlated, with LV mass. With body weight and subscapular skinfold thickness in the model, height was not statistically significant. The overall variance (R2) explained by the models in each race-sex subgroup ranged from .24 to .30.
In white men and women, diastolic blood pressure was inversely associated with LV mass, suggesting that there was also a positive relation between LV mass and pulse pressure (defined as the difference between systolic and diastolic blood pressures) in the white subgroup. Physical activity was also positively associated with LV mass in men.
Of interest, when weight was not included in the multiple linear regression model (Table 3⇓), subscapular skinfold thickness was strongly positively associated with LV mass, as was systolic blood pressure. With weight not in the model, height showed a significant positive association with LV mass in all race-sex subgroups. However, the overall variance (R2) explained by the models with weight not included (Table 3⇓) was less in each race-sex subgroup (range of R2, .15 to .21) than in the models with weight included (Table 2⇑). Diastolic blood pressure continued to be weakly and inversely related (and, by inference, pulse pressure was positively related) to LV mass in white men and white women; physical activity showed a weak positive association with LV mass in men.
LV hypertrophy measured from M-mode echocardiography has been demonstrated to be a powerful independent predictor for mortality and morbidity in adults. In a 4-year follow-up of 1141 original Framingham participants (mean age, 69 years) who were clinically free of apparent CHD, Levy et al2 reported that after adjustment for age, systolic blood pressure, smoking, obesity, and cholesterol, increased LV mass (highest versus lowest quartile) conferred a relative risk of 7.8 in men (P<.004) and 3.4 in women (P<.01) for a CHD event. After adjusting for age, systolic blood pressure, smoking, and the ratio of total to high-density lipoprotein cholesterol, the relative risk for a coronary event per 50-g/m increment in LV mass/height was 1.67 in men (95% confidence interval [CI], 1.24 to 2.23) and 1.60 in women (95% CI, 1.10 to 2.32).2 Additional evidence that echocardiographic LV mass predicts morbidity and mortality independent of standard risk factors also has been obtained in studies of patients with hypertension,17 renal failure,18 congestive heart failure,19 and in patients with or without significant coronary artery obstruction at cardiac catheterization.20 21 The relation between LV mass in young adults—for example, in the age range of the CARDIA cohort—and the development of potentially pathological LV hypertrophy in an older cohort, as in Framingham, is unknown. Nonetheless, higher LV mass in children and young adults has been related to a number of covariates that are cardiovascular risk factors, as described below.
The current study has demonstrated that in the healthy young adults of the CARDIA cohort, LV mass is generally correlated, in multivariate analyses across sex and black-white racial subgroups, with body weight and systolic blood pressure. In addition, in the multivariate model not including body weight, LV mass is correlated with subscapular skinfold thickness and height. After adjustment for age, systolic and diastolic blood pressures, height, subscapular skinfold thickness, physical activity, alcohol consumption, smoking history, FEV1, FVC, and family history of hypertension, LV mass remains, on average, 19 and 25 g higher in white men versus white women and black men versus black women, respectively, and 5 to 11 g higher in black versus white women and black versus white men, respectively (all differences statistically significant).
After adjustment for previously described covariates and weight, LV mass in black men was 16% (on average) greater than in black women and in white men was 13% greater than in white women. Adjusted LV mass in black men was slightly greater (by 3%) than in white men. In women, adjusted LV mass was similar in both racial groups.
Various workers have found that echocardiographic measurements of LV mass are lower in female than in male children and adults. Daniels and associates,22 in a study of 334 subjects aged 6 to 23 years who were free of cardiovascular disease by history, physical examination, and echocardiography, found a strong relation between LV mass and sex (P<.001). These workers suggested the following sex-specific echocardiographic criteria for LV hypertrophy in children and adolescents based on the 95th percentile of their measurements for LV mass, LV mass divided by body surface area (BSA), and LV mass divided by height, respectively: 184.9 g, 103.0 g/m2, and 99.8 g/m for male subjects and 130.2 g, 84.2 g/m2, and 81.0 g/m for female subjects.22
In a study of 654 young, healthy subjects from the Bogalusa Heart Study, aged 7 to 22 years, Burke and associates23 also noted that both black and white male subjects demonstrated significantly higher LV mass after adjustment for BSA and ponderosity than did black and white female subjects. Similarly, Gardin and associates6 demonstrated in 136 healthy men and women without clinical heart disease, aged 20 to 97 years, that for any given age and BSA, women had, on average, a 7.2% smaller LV mass than did men (P<.05). Savage et al24 also demonstrated a clear sex difference in LV mass. On the other hand, Devereux and associates25 reported that although indexing for BSA did not eliminate sex differences for LV mass, indexing for lean body mass eliminated these differences. In the CARDIA cohort, adjusting LV mass for height did not eliminate sex differences.
Studies in cohorts younger than the CARDIA cohort (with smaller sample sizes) have not shown significant associations between race and LV mass. Specifically, there was no significant relation between LV mass and race in the study of Daniels and associates.22 Burke et al23 reported no race-related differences in the increase in LV mass noted with increasing age and weight in the Bogalusa Heart Study.
Age had a weak direct bivariate relation (range of r, .08 to .12) with LV mass in the relatively narrow age range (23 to 35 years) of the CARDIA cohort in all race-sex subgroups except for white men. However, these relations were generally nonsignificant in the multivariate analyses. A number of other studies in growing children and young adults have shown a significant relation between age and LV mass. For example, Mahoney et al26 report that in children aged 6 to 15 years from the Muscatine Study, age was strongly correlated (r=.62) with unadjusted LV mass. Similarly, in the Bogalusa Heart Study, Burke et al23 reported that in young subjects aged 7 to 22 years, the correlation of unadjusted LV mass with age was r=.52 (P<.0001). However, after adjusting for BSA and ponderosity, adding age did not significantly improve the relation between blood pressure and echo LV mass (and its component variables), suggesting that these correlations were related to linear growth.
Gardin et al6 7 reported a 15% increase in LV mass in healthy, older adults older than age 70 years compared with adults in the 21- to 30-year age group (P<.01). However, in a subgroup of 862 healthy adults (aged 18 to 79 years) in the Framingham Study, multivariate analyses demonstrated that age was significantly related to LV mass in women (P<.001) but not in men.27 de Simone and associates28 also reported an increase in LV mass with age in normal women but not in men. Devereux et al25 concluded that after indexing LV mass measurements for lean body mass, there was no significant relation between age and LV mass.
In the CARDIA study, body weight, subscapular skinfold thickness, and height were among the strongest correlates of LV mass in race-sex subgroup analyses. In this study, we did not have a direct measure of lean body mass. Specifically, measures such as creatinine clearance, bioelectric impedance, or body water displacement were not available. Nonetheless, we used height as an imperfect measure of lean body mass and inferred lean body mass from the following conceptual equation.
where total body weight is measured directly and subscapular skinfold thickness is a surrogate measure for body fat.
This formula explains why in multivariate models in which weight, subscapular skinfold thickness, and height are entered as candidate variables, weight is strongly positively associated with LV mass, whereas subscapular skinfold thickness exhibits negative β–partial correlation coefficients in all race-sex subgroups. This is true because weight, which is a composite of lean body mass and fat, not only helps explain the variance due to lean body mass but a portion of the positive relation that is due to subscapular skinfold thickness. The fact that body weight “overcompensates” for this positive relation of subscapular skinfold thickness with LV mass results in a negative partial correlation coefficient for subscapular skinfold thickness when it enters the model. After body weight and subscapular skinfold thickness enter the model, height adds nothing significant to explaining LV mass (see Table 2⇑).
In contrast, when weight is excluded as a variable for consideration in the multivariate model, subscapular skinfold thickness is strongly positively associated with LV mass in all race-sex subgroups (see Table 3⇑). Height also is positively associated with LV mass and is a statistically significant covariate in all race-sex subgroups. Consequently, our data suggest that both fatness, as reflected in subscapular skinfold thickness, and lean body mass, as inferred from either total body weight minus subscapular skinfold thickness in LV mass, or from height, have a role in explaining LV mass. It should be noted that the overall variance explained by models with weight included (range of R2, .24 to .30) was greater in each race-sex subgroup than that explained by models with weight not included (range of R2, .15 to .21).
Our findings are consistent with previous reports emphasizing the strong relation of LV mass to measures of body size, lean body mass, and obesity.6 7 8 9 20 21 22 23 24 26 27 28 29 30 31 32 33 In the Bogalusa Heart Study, there were strong relations between unadjusted LV mass and height, weight, and BSA (r=.71, .76, and .78, respectively).23 Similarly, in the Muscatine Study, unadjusted LV mass was strongly correlated with height, weight, and BSA (r=.73, .75, and .77, respectively, all P<.001) and was weakly correlated with triceps skinfold thickness (P<.05). In their study of 334 subjects aged 6 to 23 years, Daniels and associates22 found that LV mass was strongly correlated with both height (r=.82 for male subjects and r=.71 for female subjects) and BSA (r=.83 for male subjects and r=.74 for female subjects). In multivariate analyses of the Framingham Study healthy subgroup, height, lean body mass, and body mass index (a measure of obesity) were all significant independent correlates of LV mass in men and in women (P<.001).9 27
Our race-sex–specific multivariate models (weight included as a covariate) compare favorably with the multivariate models for prediction of LV mass in normotensive subjects previously reported by Hammond et al33 in a group of 162 working adults with a mean age of 44 years. When body mass index and age (in women) or height (in men) were included as independent variables, the “multiple R” for both normotensive men and women was equal to .51, corresponding to an R2 of approximately .25. These workers also demonstrated independent relations of systolic blood pressure, height, and obesity (as measured by body mass index) to LV mass in the normotensive, borderline hypertensive, and hypertensive subgroups.33
Striking evidence of the relation between obesity and LV mass was provided in a controlled, randomized trial of the effects of weight reduction on M-mode echocardiographic measurements (including LV mass) performed in 41 overweight patients with hypertension. LV mass decreased by 20% (16% when adjusted for BSA) in the weight reduction group after 21 weeks of therapy.23 This decrease in LV mass was associated with an average weight loss of 8.3 kg and an average decrease in blood pressure of 14/13 mm Hg (compared with a 9/4-mm Hg reduction in the placebo group).
In a report from Framingham, Lauer and associates34 demonstrated that hypertension and obesity each had significant independent associations with LV mass and wall thickness (all P<.001 in men and women) and that the strengths of association were additive but not synergistic.
Blood Pressure and Family History of Hypertension
In the current study, systolic blood pressure was more strongly correlated to LV mass than was diastolic blood pressure both in the entire cohort and in each race-sex subgroup. In the Framingham substudy of healthy participants, there was no significant relation between LV mass and systolic blood pressure in either sex.27 However, in the Bogalusa Heart Study of young, healthy subjects, systolic and diastolic blood pressures were both significantly related to LV mass (r=.34 and r=.21, respectively; both P<.0001). Similarly, in the Muscatine Study, LV mass, corrected for age, sex, and measures of body size, was significantly greater in children with blood pressures in the highest quintile than in those in lower quintiles (P<.02).32
Some investigators have reported findings supporting the concept that increased LV mass may precede the development of hypertension.26 35 36 37 38 39 Nishio and associates35 tracked blood pressure, height, weight, and LV muscle volume in normotensive children who were 6 to 9 years old at the initial examination period. Children who remained in the highest systolic blood pressure quintile group over the two examinations, separated by 3 years, had the largest LV muscle volume index. These findings are consistent with the findings in the current study, in which in all four race-sex groups, LV mass was significantly higher in subjects in the highest compared with the lowest quartile of systolic blood pressure.
A family history of hypertension also has been reported to be associated with increased LV mass. Specifically, Celentano and associates36 reported higher LV mass index for BSA as well as increased ventricular septal and left ventricular posterior wall thicknesses in children of hypertensive parents compared with children of normotensive parents (P<.01). Furthermore, in a study of normotensive adolescents, Radice et al40 reported that boys with at least one hypertensive parent had significantly greater LV mass/BSA than did boys with normotensive parents (125.0±29.1 versus 109.2±25.4 g/m2, P<.005). No significant differences were found among adolescent male subjects with at least one hypertensive parent and those who exhibited borderline hypertension. These data are consistent with the hypothesis that increases in LV mass and ventricular septal and posterior wall thicknesses might precede elevations in arterial blood pressure. However, in the current study of the CARDIA cohort, family history of hypertension was not significantly associated with LV mass. Of interest, the study of Ravogli et al41 suggested that the increased LV mass in offspring of hypertensive patients was due at least in part to higher ambulatory pressures in these offspring.
Physical activity, calculated as “exercise units” derived from a validated questionnaire, was significantly related to LV mass only in CARDIA men. In the Framingham Offspring Study, leisure-time physical activity, derived as an average of kilocalories expended in leisure activity per week over the previous year, was significantly correlated with LV mass both in men (P<.001) and in women (P<.01).20 29
Alcohol has been reported to have both positive and detrimental effects with respect to cardiovascular disease. In a report from Framingham, Manolio and associates42 reported alcohol intake versus LV mass measured by M-mode echocardiography in 1980 men and 2511 women, ages 17 to 90 years, who were free of cardiovascular disease. This report found that alcohol intake was positively associated with LV mass in men (P<.01) but not in women. Estimated quantity of beer and wine consumption in both men and women and liquor consumption in men were positively related to LV mass. These workers concluded that alcohol use is independently associated with LV mass and that this association may vary by beverage type.
In the CARDIA cohort, a weak relation between alcohol consumption and LV mass was significant only in white men (r=.07) and in black women (r=.08). Alcohol consumption was not an independent predictor of LV mass in any race-sex subgroup.
Our study of healthy, young adults in the CARDIA cohort has shown that LV mass as measured by M-mode echocardiography is independently related to anthropometric measures—body weight and/or subscapular skinfold thickness—as well as with height (if body weight is not considered in the model), systolic blood pressure and, to a lesser extent, with diastolic blood pressure and physical activity. Of additional interest, after adjustment for these and other covariates, LV mass remains higher in men than in women and in black men than in white men. Since measures of body size, obesity, and blood pressure appear to have important relations with LV mass, longitudinal studies are necessary to delineate their possible roles in the genesis of LV hypertrophy.
This study was supported by contracts NO1-HC 48047, 48048, 48049, 48050, 95095, and 95100 from the National Heart, Lung, and Blood Institute.
Reprint requests to Julius M. Gardin, MD, Division of Cardiology, University of California Irvine Medical Center, PO Box 14091, Orange, CA 92613-1491.
Presented in part at the 64th Annual Scientific Sessions of the American Heart Association, Anaheim, Calif, November 1991.
- Received October 5, 1994.
- Revision received January 3, 1995.
- Accepted January 17, 1995.
- Copyright © 1995 by American Heart Association
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.
Devereux RB, Pickering TG, Harschfield GA, Kleiner THD, Denby L, Clark L, Pregibon D, Jason M, Kleiner B, Borer JS, Laragh JH. Left ventricular hypertrophy in patients with hypertension: importance of blood pressure response to regularly recurring stress. Circulation. 1983;68:470-476.
Gerstenblith G, Fredericksen J, Yin FCP, Fortuin NJ, Lakatta EG, Weisfeldt MD. Echocardiographic assessment of a normal adult aging population. Circulation. 1977;56:273-278.
Gardin JM, Savage DD, Ware JH, Henry WL. Effect of age, sex and body surface area on echocardiographic left ventricular wall mass in normal subjects. Hypertension. 1987;9(suppl II):II-36-II-39.
Henry WL, Gardin JM, Ware JH. Echocardiographic measurements in normal subjects from infancy to old age. Circulation. 1980;62:1054-1061.
Cutter GR, Burke GL, Dyer AR, Friedman GD, Hilner JE, Hughes GH, Hulley SB, Jacobs DR Jr, Liu K, Manolio TA. Cardiovascular risk factors in young adults: the CARDIA baseline monograph. Control Clin Trials. 1991;12(suppl 1)1S-77S.
Gardin JM, Wong ND, Bommer W, Klopfenstein HS, Smith V-E, Tabatznik B, Siscovick D, Lobodzinski S, Anton-Culver H, Manolio TA. Echocardiographic design of a multi-center investigation of free-living elderly subjects: the Cardiovascular Health Study. J Am Soc Echocardiogr. 1992;5:63-72.
Fried LP, Borhani NO, 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, CHS Collaborative Research Group. The 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.
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.
Quinones MA, Weiner DH, Shelton BJ, Greenberg BH, Limacher MC, Koilpillai C, Shindler DM, Yusuf S, SOLVD Investigators. Echocardiographic predictors of one-year clinical outcome in the study of left ventricular dysfunction (SOLVD) trial and registry: an analysis of 1172 patients. Circulation. 1993;88(suppl I):I-304. Abstract.
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-25-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.
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.
de Simone G, Devereux RB, Roman MJ, Schlussel Y, Alderman MH, Laragh JH. Echocardiographic left ventricular mass and electrolyte intake predict arterial hypertension. Ann Intern Med. 1991;114:202-209.
Post WS, Larson MG, Levy D. Impact of left ventricular structure on the incidence of hypertension. Circulation. 1994;90:179-185.
Iso H, Kiyama M, Doi M, Nakanishi N, Kitamura A, Naito Y, Sato S, Iida M, Konishi M, Shimamoto T, Komachi Y. Left ventricular mass and subsequent blood pressure changes among middle-aged men in rural and urban Japanese populations. Circulation. 1994;89:1717-1724.
Ravogli A, Trazzi S, Villani A, Mutti E, Cuspidi C, Sampieri L, de Ambroggi L, Parati G, Zanchetti A, Mancia G. Early 24-hour blood pressure elevation in normotensive subjects with parental hypertension. Hypertension. 1990;16:491-497.