International Journal of Obesity (1997) 21, 691±697 ß 1997 Stockton Press All rights reserved 0307±0565/97 $12.00 Adiposity, cardiac size and precursors of coronary atherosclerosis in 5 to 15-year-old children: a retrospective study of 210 violent deaths Department of Forensic Medicine, University of Oulu, Kajaanintie 52 D, 90220 Oulu 22, Finland OBJECTIVE: To determine the relationship between cardiac size and various measures of adiposity in children and the occurrence of coronary fatty streaks in relation to childhood obesity. DESIGN: A retrospective study based on 210 medico-legal autopsies of 5±15 y old children who had suffered violent death in the Province of Oulu, Finland, in 1970±1995. METHODS: Autopsy reports with attached clinical data and police records were recovered from the archives of the Department of Forensic Medicine, University of Oulu, Finland. All 5±15 y old individuals who had died of external trauma and were not documented as having had any chronic illnesses were chosen for investigation. The following data were extracted from the papers: age, sex, height, body weight, heart weight, liver weight, total weight of the kidneys, thickness of the abdominal subcutaneous fat and description of the coronary arteries. Heart weight was indexed to height 2.7, and liver weight and kidney weight were indexed to body weight, body surface area and height. The ponderal index (body weight/height 3 ) was calculated, and relative body weight was determined as the percentage deviation of the weight from the mean weight for height according to a growth chart developed for Finnish children. RESULTS: The average absolute heart weight and heart weight/height 2.7 were signi cantly higher in the boys than in the girls over 12 y of age. Abdominal subcutaneous fat was thicker in the girls, but there were no signi cant differences in the other indicators of obesity. After adjusting for age, there were signi cant positive correlations between the ponderal index and heart weight/height 2.7 in both the boys (P 5 0.000) and the girls (P 5 0.038), between relative body weight and heart weight/height 2.7 in the boys (P 5 0.000) and the girls (P 5 0.027), and between abdominal subcutaneous fat thickness and heart weight/height 2.7 in the boys (P 5 0.045) but not in the girls (P 5 0.234). Multiple linear regression analysis showed the ponderal index to be a signi cant predictor of heart weight/height 2.7. Coronary intimal fatty streaks were documented in six individuals (2.9%), ve of them having a ponderal index above the average and all of them having a relative weight above the average. Kidney weights were higher in the boys and were positively correlated with the measures of obesity and with heart weight/height 2.7. CONCLUSIONS: The results point to an important role of excess body weight and adiposity in the development of early myocardial and coronary changes in childhood: coronary fatty streaks were not found at all in the leanest individuals in this relatively large group and the cardiac size adjusted for body size indicated hypertrophy with increasing adiposity. Prospective autopsy studies are needed in order to determine whether adiposity and the regional distribution of fat, especially intra-abdominal fat accumulation, are associated with the earliest signs of myocardial hypertrophy and the initial phase of lesion development in the arterial intima of children. Keywords: atherosclerosis; coronary heart disease; obesity; cardiac hypertrophy Introduction Many cardiovascular risk factors are reported to be associated with early signs of arterial and cardiac pathology. Serum lipoprotein levels, blood pressure and smoking are known to be associated with the presence of early atherosclerotic lesions among adolescents and young adults, 1,2 and adiposity has been positively linked with resting blood pressure in childhood and adolescence. 1,3 The results of P ieger et al 4 indicate that overweight 6±15 y old children had higher resting systolic and diastolic blood pressures, Received 2 December 1996; revised 24 March 1997; accepted 28 April 1997 heart rates, cardiac output and stroke volumes than non-overweight children, and echocardiographic ndings have shown that overweight children have greater left ventricular mass indexed to height 2.7 than lean ones. On the other hand, Goble et al 5 reported that body weight is a strong predictor of left ventricular mass but adiposity is not, that body fat is negatively associated with left ventricular mass in 11 y old children, and that boys have a greater left ventricular mass than girls. Burke et al 6 reported that males aged from 7±22 y have a signi cantly higher left ventricular mass when adjusted for body surface area and ponderosity (body weight/height 3 ) than do females. According to the study of Daniels et al 7 both lean body mass and fat mass have a statistically signi cant independent association with left ventricular mass, but
692 the lean body mass is the strongest determinate of left ventricular mass in both sexes. Urbina et al 8 concluded that obesity is a signi cant risk factor for developing left ventricular hypertrophy in young individuals. The upper body or abdominal type of obesity is a characteristic feature of the metabolic syndrome, or insulin resistance syndrome, which is a well-known combination of cardiovascular risk factors. 9 Obesity has been reported to correlate positively with serum insulin even in childhood, 10,11 and it has been suggested that the metabolic syndrome may already exist in children. 11 Since there appears to be a lot of clinical evidence for deleterious effects of childhood obesity on the cardiovascular system, it would be of great importance to know the actual degree of obesity-related arterial and myocardial pathology to be found in children. Newman et al 1 reported that coronary or aortic fatty streaks in 35 autopsied persons with a mean age of 18 y (the majority being 15±19 y) were not related to obesity, but larger autopsy surveys on the cardiovascular system in children are few in number and have not concentrated on the effects of childhood obesity or body fat distribution patterns. Pesonen et al 12 measured narrowings in the coronary arteries of infants and 1±16 y old children who died accidentally, and concluded that intimal thickening was a morphological manifestation of a hereditary predisposition to coronary disease. They did not quote any measures of adiposity, however. Likewise, Hirvonen et al 13 reported that cholesterol esters start accumulating in the coronary arteries before the age of 10 y, but again presented no measures of body fatness. The present retrospective study was designed to investigate early signs of myocardial hypertrophy and coronary pathology in relation to body fatness in a relatively large autopsy series of children under 15 y of age. Special attention was directed towards normalization of heart size to body size in order to minimize errors in estimating the impact of obesity on the heart. In order to obtain a general picture of the relationship between childhood obesity and early signs of coronary pathology, the incidence and extent of intimal fatty streaks documented in autopsy reports was investigated. This retrospective study was based on 210 autopsies of 5±15 y old children who had died suddenly from violent causes and had been documented as otherwise healthy. Materials and methods Subjects The material was collected from the archives of medico-legal autopsies performed at the Department of Forensic Medicine, University of Oulu, Finland, during the period 1970±1995, and comprised 210 children aged from 5±15 y who had been autopsied after death from violent causes. Most of the cases were traf c accidents, followed by other mechanical traumas, drownings, homicides and a small group of suicides among older children. The individuals were not documented as having any clinical diagnoses of chronic diseases, nor were any diseases found at autopsy, and they were not reported to have used any medication or consumed alcohol or illegal drugs. The autopsy protocols with attached medical records and police records were examined and all data were collected on age, sex, height, body weight, abdominal subcutaneous fat thickness, heart weight, liver weight, sum weight of the kidneys and the state of the coronary arteries. To nd out the effect of sexual maturation on the cardiac size the male and female groups were further divided into two subgroups: from 5±11 y of age and from 12±15 y of age. Evidence of sexual maturation had not been recorded well enough in many of the autopsy reports, and therefore the individuals under 12 y of age (average time of puberty) were considered as prepubertal. Measures of obesity The ponderal index (body weight/height 3 ) was used as a measure of obesity rather than the body mass index (Quetelet index: body weight/height 2 ) because it is less closely correlated with height. 14,15 Relative body weight was determined as the percentage deviation of weight from the mean weight for height according to a growth chart designed for Finnish children by Sorva et al. 16 The thickness of the abdominal subcutaneous fat had been measured at the umbilical level according to the routine autopsy protocol and was indicated in millimetres in the autopsy reports. No other fat tissue measurements had been made at any of the autopsies. Allometric relations of heart weight Heart weight was expressed in grammes in the autopsy reports. According to the routine autopsy protocol, the heart had been removed by cutting through the aorta and the other great vessels just above the level of the aortic valve, washed and weighed fresh. The heart weight was indexed to height 2.7 for the present purposes in order to reduce the errors caused by body size when estimating the effect of overweight on heart size. According to de Simone et al 17 the left ventricular mass obtained by echocardiography and indexed to height 2.7 is the best method for normalizing left ventricular mass to body size. No chamber partition had been performed at the autopsies, but as the hearts were not documented as showing any abnormalities, and as the total heart weight is known to be associated with left ventricular mass, 18 this modi cation was considered suitable for the purposes of this study.
Liver and kidney weights indexed to body size The liver weight and the total weight of the kidneys were indexed to body weight, body surface area and height. The weights of the other internal organs were not included here because of the changes related to the cause of death, for example the increase in lung weight associated with drowning, or contusion haemorrhages and oedema in brain tissue. Liver and kidney weights were used only for those cases in which no severe injuries were documented in these organs (liver: 191 cases, kidneys: 179 cases). Results Height, weight and adiposity Statistics descriptive of the individuals are presented in Table 1. The average age was exactly the same for the boys and girls under 11 y of age and the mean values for height, weight, ponderal index and relative body weight did not show any statistically signi cant differences. Girls had signi cantly more abdominal subcutaneous fat than boys, however. In the older agegroup the girls had also more abdominal fat, but no other signi cant differences existed between the sexes. 693 Statistical analysis All the analyses were performed using Statistical Package for the Social Science software. 19 Since the distributions of most of the variables were slightly skewed, logarithmic transformations were used when needed. Student's t-test was used to compare the means for the male and female groups. Partial correlation coef cients with age as a covariate were computed between heart weight indexed to height and the various measures of obesity, including ponderal index, relative body weight and abdominal subcutaneous fat thickness. Partial correlation coef cients were also calculated between liver/kidney weight and the measures of obesity and heart weight/height 2.7. Multiple stepwise linear regression analysis was performed in order to see which of the measures of adiposity would be the best predictor of cardiac hypertrophy in the male and female subjects separately. The regression model included heart weight/height 2.7 as a dependent variable and age, ponderal index, relative body weight and abdominal subcutaneous fat thickness as independent variables. The signi cance level for entry was 0.05 and for removal 0.10. Stepwise linear regression analysis was also performed to determine whether relations of heart weight/height 2.7 and kidney/height or liver/height were independent of age, gender, height and measures of adiposity. Cardiac size The average absolute heart weight and heart weight indexed to height 2.7 were signi cantly greater in the boys than in the girls over 12 y of age, but there were no signi cant differences between the sexes in the younger agegroup (Table 1). Size of liver and kidneys The absolute and relative weights of the kidneys were signi cantly greater in the boys than in the girls over 12 y. The absolute kidney weight and the weight indexed for body surface area were signi cantly higher in the boys under 11 y, too. Liver weight indexed to body surface area was higher in the boys of the same age group, but there were no other signi cant gender differences in the liver weights (Table 1). Coronary arteries Most of the 210 children had not been reported to have any visible signs of coronary pathology, coronary intimal fatty streaks having been observed in six cases (2.9%), two girls and four boys. One 5 y old girl had fatty streaks in all the main coronary branches, while the others had at least one fatty streak in the left anterior descending. All of these individuals Table 1 Descriptive statistics for the series of autopsy cases. The results are expressed as means standard deviation Age group 5^11 y Age group12^15 y Males (n ˆ 85) Females (n ˆ 43) Males (n ˆ 60) Females (n ˆ 22) Age (y) 8 2 8 2 14 1 13 1 Height (cm) 129 11 129 14 163 13 159 12 Weight (kg) 27 6 29 10 51 13 47 11 Relative weight (% from ideal weight) 2.1 12.1 1.9 12.2 0.4 15.8 71.9 12.1 Ponderal index (kg/m 3 ) 13.2 5.0 12.7 1.7 11.7 1.8 11.7 1.4 Abdominal fat (mm) 6 4 9 5** 10 7 13 7* Heart weight (g) 142 39 136 46 258 59 206 47*** Heart weight/height 2.7 (g/m 2.7 ) 71 12 67 12 68 9 59 8*** Liver weight (g) 834 187 774 258 1313 363 1202 289 Liver weight/body weight (g/kg) 30.60 5.80 28.63 5.73 26.14 6.29 25.60 4.20 Liver weight/body surface area (g/m 2 ) 0.83 0.14 0.77 0.14* 0.86 0.18 0.82 0.13 Kidney weight (g) 136 38 118 33* 219 59 178 37** Kidney weight/body weight (g/kg) 4.99 1.22 4.54 1.13 4.34 0.87 3.83 0.69* Kidney weight/body surface area (g/m 2 ) 1.35 0.32 1.21 0.27* 1.42 0.27 1.22 0.18** *P < 0.05; **P < 0.01; ***P < 0.001. Student's t-test.
694 Table 2 Coronary arterial lesions and physical characteristics in children aged from 5±15 y A. No detectable lesions: 204 individuals (87.1%) B. Fatty streaks detectable: 6 individuals (2.9%) Girl, age 5 y, 115 cm/21 kg, ponderal index 13.8, relative weight 4%, fatty streaks in all main coronary branches Girl, age 15 y, 176 cm/62 kg, ponderal index 11.4, relative weight 4%, fatty streaks in the left anterior descending Boy, age 9 y, 139 cm/36 kg, ponderal index 13.4, relative weight 12%, fatty streaks in the left anterior descending Boy, age 12 y, 147 cm/50 kg, ponderal index 15.7, relative weight 32%, fatty streaks in the left anterior descending Figure 2 Linear relations of heart weight/height 2.7 to ponderal index in 5±15 y old girls; r ˆ 0.260, P ˆ 0.038, adjusted for age. Boy, age 13 y, 176 cm/92 kg, ponderal index 16.9, relative weight 45%, fatty streaks in the left anterior descending Boy, age 14 y, 171 cm/98 kg, ponderal index 19.5, relative weight 76%, fatty streaks in the left anterior descending had a relative weight above the mean value for their age group and ve of them had a ponderal index above the average (Table 2). Cardiac size vs measures of adiposity After adjusting for age, the ponderal index had a signi cant positive correlation with heart weight indexed to height 2.7 in both the boys (r ˆ 0.340, P < 0.001) and the girls (r ˆ 0.260, P ˆ 0.038). The linear correlations (with 95% con dence interval) between ponderal index and heart weight/height 2.7 for the boys and girls, respectively, are shown in Figures 1 and 2. Relative body weight also showed signi cant positive correlation with heart weight/height 2.7 after adjusting for age in both the boys (r ˆ 0.333, P < 0.001) and the girls (r ˆ 0.277, P ˆ 0.027). Linear correlations (with 95% con dence interval) between relative weight and heart weight/height 2.7 are presented for the boys in Figure 3 and for the girls in Figure 4. Figure 3 Linear relations of heart weight/height 2.7 to relative body weight in 5±15 y old boys; r ˆ 0.333, P ˆ 0.000, adjusted for age. Figure 4 Linear relations of heart weight/height 2.7 to relative body weight in 5±15 y old girls; r ˆ 0.277, P ˆ 0.027, adjusted for age. Figure 1 Linear relations of heart weight/height 2.7 to ponderal index in 5±15 y old boys; r ˆ 0.340, P ˆ 0.000, adjusted for age. The age-adjusted correlation between abdominal subcutaneous fat thickness and heart weight/height 2.7 also reached the level of signi cance in the boys (r ˆ 0.177, P ˆ 0.045), but not in the girls (r ˆ 0.157, P ˆ 0.234). Linear correlations (with 95% con dence interval) between abdominal subcutaneous fat and heart weight indexed for height are presented in Figures 5 and 6 for boys and girls, respectively.
girls (B ˆ 0.4351, P < 0.0001 for boys, B ˆ 0.4353, P ˆ 0.0062 for girls). 695 Figure 5 Linear relations of heart weight/height 2.7 to abdominal subcutaneous fat thickness in 5±15 y old boys; r ˆ 0.177, P ˆ 0.045, adjusted for age. Size of liver and kidneys vs measures of obesity and heart size The weights of the liver and kidneys, absolute or indexed to height, showed in general signi cant positive correlations with the measures of adiposity. Liver and kidney weights were signi cantly positively correlated with heart weight/height 2.7 in both male groups. Both absolute and relative kidney weights were also signi cantly correlated with heart weight/ height 2.7 in the girls under 11 y (Table 3). In the multiple regression analysis, kidney weight/height (B ˆ 0.2608, P ˆ 0.0069), liver weight/height (B ˆ 0.2856, P ˆ 0.0046), ponderal index (B ˆ 0.1756, P ˆ 0.0399), height (B ˆ 70.4351, P < 0.0001) and gender (B ˆ 70.1392, P ˆ 0.040) were statistically signi cant independent correlates of heart weight/ height 2.7. Discussion Figure 6 Linear relations of heart weight/height 2.7 to abdominal subcutaneous fat thickness in 5±15 y old girls; r ˆ 0.157, P ˆ 0.234, adjusted for age. When heart weight/height 2.7 was used as the dependent variable in the multiple linear regression model, the ponderal index was the only parameter which entered the regression for both boys and The present retrospective autopsy study of 5±15 y old children emphasizes the importance of body fatness as an associated factor in the development of early myocardial hypertrophy and coronary fatty streaks at an early age. Obesity has been regarded as one of the major cardiovascular risk factors, which tend to form a cluster from childhood onwards. 13 There is a considerable amount of clinical information available on the association between obesity and elevated blood pressure in children, 3 and echocardiographic studies have revealed greater left ventricular mass in overweight children than in the lean ones, 4 but obesity- Table 3 Partial correlation coef cients, adjusted for age, between the absolute and relative weights of the liver and kidneys, and measures of obesity and heart weight/height 2.7 in 5±15 y old boys and girls Ponderal index (kg/m 3 ) Relative weight (% ideal weight) Abdominal fat thickness (mm) Heart weight/ height 2.7 (g/m 2.7 ) Liver weight males 5±11 y 0.132 0.148 0.120 0.412*** females 5±11 y 0.276 0.337* 0.387* 0.161 males 12±15 y 0.208 0.269* 0.552*** 0.361** females 12±15 y 0.016 0.091 0.218 0.294 Liver weight/height males 5±11 y 0.120 0.211 0.072 0.495*** females 5±11 y 0.299 0.333* 0.331 0.251 males 12±15 y 0.243 0.291* 0.524*** 0.382** females 12±15 y 0.076 0.131 0.225 0.387 Kidney weight males 5±11 y 0.222 0.201 0.137 0.273* females 5±11 y 0.301 0.260 0.299 0.608*** males 12±15 y 0.259 0.282* 0.394** 0.317* females 12±15 y 0.147 0.175 0.370 0.324 Kidney weight/height males 5±11 y 0.215 0.248* 0.111 0.331** females 5±11 y 0.319 0.236 0.207 0.710*** males 12±15 y 0.310* 0.318* 0.339* 0.330* females 12±15 y 0.227 0.218 0.338 0.405 *P < 0.05; **P < 0.01; ***P < 0.001.
696 associated cardiac hypertrophy in children has not previously been evaluated at autopsy. The early coronary arterial changes lie beyond the limits of clinical diagnostics, of course, and the few autopsy studies on children's coronary arteries have mainly concentrated on other traditional risk factors rather than body fatness. 1,12,13,20 Since mortality from violent causes is higher in males even in childhood, 21 the majority of the cases, as expected, were boys (n ˆ 145), and although the girls' group also proved to be quite representative (n ˆ 65), its smaller size may have affected the interpretation of the results. The mean value for the absolute heart weight was higher in the boys, which is in accordance with clinical observations indicating greater left ventricular mass in boys. 5,6 Also, normalization of heart weight to height, 17 produced higher results for the boys. This gender difference in heart weight became signi cant in children over 12 y, which is in accordance with the observations on left ventricular growth in children. 22 P ieger et al 4 did not assess the effect of gender on the relationship between cardiac size and adiposity because of limited sample size, but the present autopsy results point to slightly closer correlations, that attained much higher statistical signi cance, in the larger group of boys than in the girls. Earlier autopsy ndings suggest that fatty streaks commonly appear in the coronary arteries during the second decade of life. 23 In the presents series, fatty streaks were documented in six individuals aged between 5 and 14 y, but there may have been some underestimation at the autopsies because only visual inspections had been made after cutting open the vessels. Sudan staining would probably have revealed minor fatty lesions in more than six cases. It is signi cant, however, that the individuals with relative body weights below the mean value were not documented to have any fatty streaks at all in this relatively large series of autopsies. Statistical correlations are not presented here, as the visually estimated prevalence of fatty streaks was very low and statistical analysis would require a more speci c quantitative evaluation of the lesions. This may partly be the reason why Newman et al 1 did not nd any association between obesity and fatty streaks in young individuals. Future research will require microscopic characterization and more detailed quanti cation of early coronary changes in obese and lean children in order to be able to perform more reliable statistical comparisons. The appearance of the initial lesions, type I lesions, 24 and also microscopic examination of the composition of the fatty streaks, or type II lesions, 24 would be important, as these changes may still be reversible and could possibly be in uenced by nutritional and weight reduction procedures. It was not possible to estimate the lean body mass in this survey, because no information on skeletal muscle mass was available. The sizes of the internal organs such as the liver and kidneys do provide some indication of the relationship between overweight and fat free mass, however. Excess energy intake has been observed to stimulate tissue growth and protein synthesis in the rat liver, kidneys, skeletal muscle and heart. 25 The positive associations between the measures of adiposity, liver and kidney size and heart weight/height 2.7 in the present material may point to the role of both increased lean body mass and increased fat tissue in myocardial hypertrophy in children. The relationship between heart size and kidney/liver size was also independent of age, gender, height and ponderal index. The same phenomenon has already been shown for heart size-arterial relations. 26,27 Conclusions The results of this retrospective autopsy survey point to an important role of excess body weight and adiposity in the development of myocardial and coronary changes in childhood, as coronary fatty streaks were not found at all in the leanest individuals in this relatively large series, while the cardiac size adjusted for body size indicated hypertrophy with increasing ponderal index and relative body weight. The thickness of the abdominal subcutaneous fat was the only direct measure of fat tissue which could be evaluated in this survey, but a signi cant association was, however, found in boys for heart weight/height 2.7 and abdominal fat. As the origins of the metabolic syndrome and atherosclerotic development probably lie in childhood, prospective autopsy studies of childhood adiposity and fat tissue distribution, and especially the accumulation of intra-abdominal fat, are needed in order to ascertain how early the deleterious effects of overweight on the cardiovascular system actually occur. A research group at our institution has now started a long-term project to investigate the relation between the initial phases of coronary atherosclerosis, myocardial hypertrophy and obesity and to determine fat distribution patterns, hypothesizing that adiposity may be a more important cardiovascular risk factor than was previously thought. References 1 Newman WP III, Freedman DS, Voors AW, Gard PD, Srinivasan SR, Cresanta JL, Williamson GD, Webber LS, Berenson GS. Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis: The Bogalusa Heart Study. N Engl J Med 1986; 314: 138±144. 2 Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Relationship of atherosclerosis in young men to serum lipoprotein cholesterol concentrations and smoking. JAMA 1990; 264: 3018±3023. 3 Jung FF, Ingel nger JR. Hypertension in childhood and adolescence. Ped Rev 1993; 14: 169±179. 4 P ieger KL, Treiber FA, Davis H, McCaffrey FM, Raunikar RA, Strong WB. The effect of adiposity on children's left
ventricular mass and geometry and haemodynamic responses to stress. Int J Obes 1994; 18: 117±122. 5 Goble MM, Mosteller M, Moskowitz WB, Schieken RM. Sex differences in the determinants of left ventricular mass in childhood. Circulation 1992; 85: 1661±1665. 6 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. 7 Daniels SR, Kimball TR, Morrison JA, Khoury P, Witt S, Meyer RA. Effect of lean body mass, fat mass, blood pressure, and sexual maturation of left ventricular mass in children and adolescents. Circulation 1995; 92: 3249±3254. 8 Urbina EM, Gidding SS, Bao W, Pickoff AS, Berdusis K, Berenson GS. Effect of body size, ponderosity, and blood pressure of left ventricular growth in children and young adults in the Bogalusa heart study. Circulation 1995; 91: 2400±2406. 9 Kaplan NM. The deadly quartet: upper-body obesity, glucose intolerance, hypertriglyceridemia, and hypertension. Arch Intern Med 1989; 149: 1514±1520. 10 DahlstroÈm S, Viikari J, AÊ kerblom HK, Solakin-Jaakkola T, Uhari M, Dahl M, LaÈhde PL, Pesonen E, Pietikainen M, Suoninen P. Atherosclerosis precursors in Finnish children and adolescents. II. Height, weight, body mass index, and skinfolds, and their correlation to metabolic variables. Acta Pediatr Scand Suppl 1985; 318: 65±78. 11 RoÈnnemaa TE, Knip MJ, Lautala PJ, Viikari J, Uhari M, Leino A, Kaprio EA, Salo MK, Dahl M, Nuutinen EM. Serum insulin and other cardiovascular risk indicators in Finnish children, adolescents and young adults. Ann Med 1991; 23: 67±72. 12 Pesonen E, Norio R, Hirvonen J, Karkola K, Kuusela V, Laaksonen H, MoÈttoÈnen M, Nikkari T, Raekallio J, Viikari J, YlaÈ-Herttuala S, A Ê kerblom HK. Intimal thickening in the coronary arteries of infants and children as an indicator of risk factors for coronary heart disease. Eur Heart J Suppl E 1990; 11: 53±60. 13 Hirvonen J, YlaÈ-Herttuala S, Laaksonen H, MoÈttoÈnen M, Nikkari T, Pesonen E, Raekallio J, A Ê kerblom HK. Coronary intimal thickenings and lipids in Finnish children who died violently. Acta Pediatr Scand Suppl 1985; 318: 221±224. 14 Voors AW, Webber LS, Frerichs RR, Berenson GS. Body height and body mass as determinants of basal blood pressure in children: the Bogalusa Heart Study. Am J Epidemiol 1977; 106: 101±108. 15 Bao W, Srinivasan SR, Wattigney WA, Berenson GS. Persistence of multiple cardiovascular risk clustering related to syndrome X from childhood to young adulthood. Arch Intern Med 1994; 154: 1842±1847. 16 Sorva R, Perheentupa J, Tolppanen E-M. A novel format for a growth chart. Acta Pediatr Scand 1984; 73: 527±529. 17 desimone G, Daniels SR, Devereux RB, Meyer RA, Roman MJ, dedivitiis O, Alderman MH. Left ventricular mass and body size in normotensive children and adults: Assessment of allometric relations and impact of overweight. J Am Coll Cardiol 1992; 20: 1251±1260. 18 Silver MD (ed). Cardiac Pathology. 2nd edn. Churchill Livingstone: New York, 1991. 19 SPSS Inc. SPSS for Windows 6.1. SPSS Inc: Chicago, 1994. 20 YlaÈ-Herttuala S, Nikkari T, Hirvonen J, Laaksonen H, MoÈttoÈnen M, Pesonen E, Raekallio J, A Ê kerblom HK. Biochemical composition of coronary arteries in Finnish children. Arteriosclerosis 1986; 6: 230±236. 21 Causes of Death. Statistics Finland. Helsinki 1995. 22 desimone G, Devereux RB, Daniels SR, Meyer RA. Gender differences in left ventricular growth. Hypertension 1995; 26 (part 1): 979±983. 23 Strong JP, McGill HC. The natural history of coronary atherosclerosis. Amer J Pathol 1962; 40: 37±49. 24 Stary HC, Chandler AB, Glagov S, Guyton JR, Insull W, Rosenfeld ME, Schaffer SA, Schwartz CJ, Wagner WD, Wissler RW. A de nition of initial, fatty streak, and intermediate lesions of atherosclerosis. A report from the committee on vascular lesions of the council on arteriosclerosis, American Heart Association. Circulation 1994; 89: 2462± 2478. 25 Estornel E, Cabo J, Barber T. Protein synthesis is stimulated in nutritionally obese rats. J Nutr 1995; 125: 1309±1315. 26 Roman MJ, Saba PS, Pini R, Spitzer M, Pickering TG, Rosen S, Alderman MH, Devereux RB. Parallel cardiac and vascular adaptation in hypertension. Circulation 1992; 86: 1909± 1918. 27 Roman MJ, Pickering TG, Schwartz JE, Pini R, Devereux RB. Association of carotid atherosclerosis and left ventricular hypertrophy. J Am Coll Cardiol 1995; 25: 83±90. 697