Tracking of Left Ventricular Mass in Children: Race and Sex Comparisons
The MCV Twin Study
Background—Increased left ventricular (LV) mass is a predictor of cardiovascular disease in adults. The mechanism(s) for these observations are not fully understood.
Methods and Results—We repeatedly studied a biracial sample of children from ages 11 through 17 years. At visits 1 through 5, height, weight, and pubertal stage were determined. Blood pressure and heart rate were measured. M-mode and two-dimensional echocardiograms were performed with a 3.5-MHz transducer with the subject in the supine position. LV mass was calculated. Repeated-measures analysis using a mixed modeling approach was performed for LV mass. At all ages, boys had greater LV mass than girls. For the population as a whole, we found significant tracking correlations for LV mass between each interval of measurement and throughout the entire period of examination. The tracking correlation for the entire sample from visit 1 through visit 5 was r=.41. The LV mass in white children tracked from the youngest to the oldest. Black children tracked similarly from ages 1 to 15 years, but tracking was not significant across the widest interval, visit 1 through visit 5. Racial differences were found in the interactions of systolic blood pressure and heart rate, which magnified the differences in LV mass. During adolescence, LV mass tracks significantly in both black and white children.
Conclusions—Interactive effects such as weight, blood pressure, and heart rate magnify sex and race differences in LV mass.
Increased left ventricular (LV) mass is an accepted, independent predictor of cardiovascular morbidity and mortality in adults.1 The mechanisms for these observations are not fully understood. Although blood pressure contributes to the variability of LV mass in all age groups, there is only a weak association between the level of blood pressure in children and the amount of LV mass.2 Body size, and in particular lean body mass, explains the major portion of the variability of LV mass.3 4
Many of the coronary heart disease risk factors, such as blood pressure and cholesterol levels, track, ie, remain in a given rank order relative to peers, over time. In children, such tracking occurs most dramatically for height and weight, followed to a lesser extent by total cholesterol, LDL cholesterol,5 and blood pressure.6 The tracking coefficients depend on both the interval between measurements and the age at the initial measurement. For instance, the lowest correlation coefficients for height occur between ages 11 and 14 years, reflecting the wide variations that occur during pubertal growth spurts and the different patterns of growth between boys and girls.7
To determine whether tracking occurs for LV mass and how it may differ between boys and girls and between black and white children, we repeatedly studied a biracial sample of children from ages 11 through 17 years. We asked the following questions: (1) Does LV mass track during adolescence? (2) Is tracking consistent within age intervals despite differing rates of pubertal development? (3) Do tracking patterns differ by sex or race?
The Medical College of Virginia (MCV) Twin Study was approved by the Human Studies Institutional Review Board. Parents and children all gave informed consent before the study. From the participants in the MCV Twin Study, a twin was selected randomly from each of the pairs of twins. Measurements including echocardiographic determinants of LV mass were performed in each child at 18-month intervals from age 11 until 17 years (visit 1 through visit 5). Each twin was studied as close as possible to his or her birth date. Each cohort, therefore, had large groups of children all of whom were the same age.
Height and weight were obtained twice with a stadiometer and a calibrated electronic scale and averaged. Pubertal stage was determined by a self-report form that included drawings based on the pubic hair and breast-development phases depicted in Tanner’s five-stage scale.5 The self-report stage was validated by the technicians during the examination procedures. We obtained resting blood pressures in the sitting position with a mercury sphygmomanometer according to standard measurement criteria.8 Diastolic blood pressures were measured at the fifth Korotkoff sound. We used a cardiotachometer to record heart rates. We obtained the blood pressures and heart rates ≈30 minutes after the child’s arrival at the testing site. Blood pressures and heart rates were taken twice and averaged.
M-mode and two-dimensional echocardiograms were performed with a 3.5-MHz transducer with the subject in the supine position. LV dimensions were measured in the short-axis view with two-dimensional guided M-mode echocardiograms that were digitized and measured according to standard criteria.9 Three cycles were measured, and the average was used in subsequent analyses. LV mass was calculated by use of the anatomically validated Penn convention10: LV mass (g)=1.04[(LVDD+IVS+LVPW)3− LVDD3]−13.6, where LVDD is the LV end-diastolic dimension, IVS is the interventricular septal dimension, and LVPW is the LV posterior wall thickness.
Test-retest reliability with repeated measures of subjects and repeated digitizing of LV dimensions demonstrated that the reproducibility for the IVS (r=.88), LVDD (r=.97), and LVPW (r=.86) was high (all P<.01).
We performed pooled t tests, taking into account whether or not different variances were present, and Pearson product-moment correlations using the SAS (Cary, NC) computer package.
Repeated-measures ANOVA using a mixed modeling approach11 12 was performed for LV mass. The analysis was performed in the SAS System using PROC MIXED13 (SAS Institute, 1992). The mixed modeling approach allows for missing information over visits and allows for modeling of the variance as well as the mean. The general form of the mixed model is given by y=Xβ+Zν+e, where y is the observed response vector, X is a known design matrix for the fixed effects, β is a vector of unknown parameters for the fixed effects, Z is a known design matrix for the random effects, ν is a vector of unknown parameters for the random effects, and e is a vector of unknown random errors. The variance of y is modeled by an unstructured variance matrix, and for the mixed model, E(y) =Xβ, and thus, the mean LV mass is modeled as where y=LV mass, x1=weight, x2=sex (female or male), x3=race (black or white), x4=heart rate, x5=systolic blood pressure, x6=diastolic blood pressure, βi=unknown main fixed-effect parameters (I=1, 2, 3, 4, 5, 6), and βij=unknown interaction parameters (I=1, 2, 3, 4, 5, 6; j=2, 3, 4, 5, 6; I≠j).
The sample sizes for each of the 5 study periods (Table 1⇓) ranged from 231 subjects studied on at least 3 occasions, 203 subjects on 4 occasions, and 87 subjects on 5 occasions. Because of the number of technically acceptable echocardiograms, a larger number of subjects were studied during visit 3 than during visit 2. The small number of subjects at visit 5 reflects the number of subjects returning at age 17 years for restudy at the conclusion of the funding period. During visit 5, of the 81 subjects returning for measurement, 74 had acceptable echocardiographic tracings for analysis. We achieved an overall rate of acceptable echocardiograms during the entire study of 96%. The dropout rate of subjects between visits 1 and 4 was 8.5%.
Of all the children studied at 11 years of age, black children represented 19% of the sample. During the subsequent study periods, they continued to represent 18% to 21% of the sample population.
At all ages, boys had greater LV mass than girls (Table 1⇑). Although not shown in the table, these differences persisted after adjustment for body mass index. Prepubertal boys and girls weighed the same, but by visit 4 boys weighed more. Beginning at visit 2, boys had consistently lower heart rates than girls. During early puberty, boys had lower diastolic blood pressures than girls. In contrast, from mid to late puberty, the systolic blood pressures in boys were higher than in girls.
We compared these variables across race (Table 2⇓). In the youngest children, the LV mass of the black children exceeded that of the white children, but this difference disappeared by visit 2. Until late puberty, black children weighed more than white children. In the youngest children (visits 1 and 2) and at visit 4, the black children had higher systolic blood pressures. These differences were not found in the oldest subjects.
No heart rate or diastolic blood pressure differences were noted between the races.
We next examined differences between the races by sex. In the youngest children, black boys weighed more, had higher systolic blood pressures, and had greater LV masses than the white boys at comparable ages (Table 3⇓). At older ages, we did not observe differences for any of these variables among the boys. Only weight and systolic blood pressure were greater in the youngest black girls compared with white girls of the same age (Table 4⇓). In older girls, we did not find differences between white and black adolescents for body size, blood pressure, or LV mass.
For the population as a whole, we found significant tracking correlations for LV mass between each interval of measurement and throughout the entire period of examination (Table 5⇓). The tracking correlation for the entire sample from visit 1 through visit 5 was r=.41. Similar tracking correlations persisted after the sample was subdivided by race and sex. The LV mass of white children tracked from the youngest to the oldest (Table 6⇓). Black children tracked similarly from age 11 to 15.5 years, but the tracking was not significant across the widest interval (visit 1 through visit 5). The small sample size (n=13) for 17-year-old black children was most likely responsible for this change. Of particular interest, the oldest black children tracked highly in the interval from visit 4 to visit 5.
The repeated-measures analysis (Table 7⇓) documented that for the entire group, as weight increased, LV mass increased (Figure⇓). Race was found to interact with weight, systolic blood pressure, and heart rate. In each case, the differences in LV mass in blacks and whites were magnified as either weight, systolic blood pressure, or heart rate increased. Heart rate was the only significant interaction that occurred between boys and girls. As the heart rate increased, the LV mass differences between boys and girls increased.
In this longitudinal study of 231 normotensive subjects examined at ages 11 through 17 years, we that found persistence of peer rank order, or tracking, occurs for LV mass. Throughout adolescence, LV mass was consistently larger in boys than girls. Increases in weight, systolic blood pressure, and heart rate all increased the differences in LV mass between boys and girls. We did not observe sex differences in the degree of tracking. The tracking correlations remained constant throughout early puberty and increased in the later stages of puberty. Repeated-measures analysis showed sex and race differences in the tracking of LV mass.
The genetic control of growth that regulates both weight and LV mass is probably responsible for a major portion of tracking variability in LV mass during adolescence.14 These increases in LV mass are developmental and are the consequences of normal growth.15 In adults, increased LV mass may be a more important predictor of subsequent hypertension than resting blood pressure.16 In childhood, however, the blood pressure contribution to the variability of LV mass is quite small.17 Nonetheless, some interactions of blood pressure and LV mass affect the tracking of LV mass during childhood. In children who participated in the Muscatine School Study, echocardiograms performed earlier were excellent predictors of resting blood pressure 3 years later.18 Although we have demonstrated significant tracking during puberty, the degree of tracking between adolescence and adulthood is largely unknown. At our state of knowledge, we do not know whether or not the degree of LV mass in adolescence predicts premature atherosclerosis in the adult.
Race and genetic factors contribute to LV mass. Black adults with mild hypertension have an up to twofold greater prevalence of LV hypertrophy than white adults with similar blood pressure levels.19 In our study of randomly selected subjects, we did not find differences in the level of LV mass between black and white children. However, LV wall thickness and mass are consistently greater in children and adolescents with a positive family history of hypertension, and this appears to be independent of slight blood pressure elevations.20 Our findings of the interactions of LV mass with weight, blood pressure, and heart rate may subsequently prove to be useful predictors of future cardiovascular disease.
In summary, during adolescence, LV mass tracks significantly in both black and white children. There are interactive effects, such as weight, blood pressure, and heart rate, that magnify sex and race differences in LV mass. The importance of this tracking to predict premature cardiovascular disease is not yet known.
- Received August 25, 1997.
- Revision received January 29, 1998.
- Accepted February 4, 1998.
- Copyright © 1998 by American Heart Association
Brandao AA, Pozzan R, Albanesi Filho FM, Brandao AP. Role of anthropometric indexes and blood pressure as determinants of left ventricular mass and geometry in adolescents: the Rio de Janeiro Study. Hypertension. 1995;6:1190–1194.
Clarke WR, Schrott HG, Leaverton PE, Connor WE, Lauer RM. Tracking of blood lipids and blood pressures in school age children: the Muscatine Study. Circulation. 1978;58:626–634.
Report of the Second Task Force on Blood Pressure Control in Children. Pediatrics. 1987;79:1–25.
Schieken RM, Mahoney L, Clarke W, Lauer R. Measurement criteria for group echocardiographic studies. Am J Epidemiol. 1979;110:504–514.
Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man: anatomic validation of the method. Circulation. 1977;55:613–618.
Searle SR. Linear Models. New York, NY: John Wiley & Sons; 1971.
SAS Institute Inc. SAS® Technical Report P-229, SAS/STAT® Software Changes and Enhancements, Release 6.07. Cary, NC: SAS Institute Inc; 1992.
Schieken RM. Large hearts in children: biology or disease? Circulation. 1995;92:3156–3157.
Casale PN, Devereux RB, Milner M, Zullo G, Harshfield GA, Pickering TG, Laragh JH. Value of echocardiographic measurement of left ventricular mass in predicting cardiovascular morbid events in hypertensive men. Ann Intern Med. 1986;105:173–178.
Goble MM, Mosteller M, Moskowitz WB, Schieken RM. Sex differences in the determinants of left ventricular mass in childhood: the Medical College of Virginia Twin Study. Circulation. 1992;85:1661–1665.
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.
Malcolm DD, Burns TL, Mahoney LT, Lauer RM. Factors affecting left ventricular mass in childhood: the Muscatine Study. Pediatrics. 1993;92:703–709.