Elevated blood pressure (BP) is the leading cause of the global burden of cardiovascular disease (CVD).1 Before menopause, women have lower BP levels than men of the same age.2,3 The incidence of hypertension is also higher in men than in women.4 Although these sex differences in BP may be associated with higher CVD morbidity and mortality among men,5 it is not well understood when or how the sex differences in BP appear in the life course.6 BP increases during childhood and adolescence.7 Recent evidence suggests that sex differences in the patterns of increases in systolic BP with age8,9 and differences in the prevalence of high systolic BP10 are already evident in adolescence.
The increase in BP with age is most pronounced during adolescence.11 This increase may be related to the rapid growth (height, weight, and body fat) associated with puberty (Fig. 1).12 Although the mechanisms responsible for BP changes in youth are currently not well understood,13,14 it is believed that obesity or excess weight promotes higher BP through an increase in the sympathetic nervous activity, insulin resistance, and arterial stiffness—all of which increase cardiac output and systemic vascular resistance.15,16 Weight loss in obese youth improves insulin sensitivity and sympathetic nervous activity and decreases cardiac output and systemic vascular resistance,17–19 promoting restoration of normal BP.13
It remains unclear whether sex differences in the patterns of increase in systolic BP with age during adolescence may be due to the differences between boys and girls in gains in body size, adiposity, hormones, or some other factors.6 The relationship between increase in BP with age and simultaneous changes in height, weight, and body fat has important clinical and public health implications,1 particularly in the context of the current childhood obesity epidemic.20,21 However, longitudinal cohort studies on BP changes with age have not considered the effect of changes in height, weight, or other anthropometric characteristics. The objective of the present study was to assess the sex-specific impact of changes in anthropometric characteristics, such as height, body mass index (BMI), waist circumference, and triceps and subscapular skinfold thickness on systolic BP in a cohort of adolescents.
The Natural History of Nicotine Dependence in Teens study is an ongoing prospective cohort of 1293 students who were in grade 7 (aged 12-13 years) at baseline, recruited in Fall 1999 in a convenience sample of 10 Montreal secondary schools.22 Schools were selected to include a mix of French and English schools, schools in urban, suburban, and rural areas, and schools in high, medium, and low socioeconomic neighborhoods. All students provided assent, and their parents or guardians provided signed informed consent. The Institutional Review Boards at McGill University and Centre de recherche de Centre hospitalier de l'Université de Montréal approved the study protocol.
During the first 5 years of follow-up, systolic BP and anthropometry were measured 3 times (Fall 1999/Winter 2000, Spring 2002, and Spring 2004).23 A total of 1192 students in 1999, 953 in 2002, and 799 in 2004 had systolic BP measurements available for analysis. The mean number of repeated BP measurements per student was 2.5. (718 [57%] students had BP measured on all 3 occasions; 250 [20%] on 2 occasions; and 290 [23%] on 1 occasion). Students were, on average, 12.8 years of age in 1999, 15.2 in 2002, and 17.0 in 2004.
BP was assessed in the sitting position, on the right arm, after the student had voided and then had a 5-minute rest period. Technicians who had been trained and certified according to a standardized protocol24 took the measurements with an automated oscillometric device (Dinamap XL, model CR9340; Critikon Co., Tampa, FL). Oscillometric devices were calibrated against a mercury sphygmomanometer before each data collection period. A minimum of 3 BP readings were recorded for each student at 1-minute intervals. If the difference between the second and third readings exceeded 20 mm Hg and 10 mm Hg for systolic and diastolic BP, respectively, then fourth and fifth readings were taken. To reduce BP reactivity or habituation, the first reading was not considered.25 The mean value of the 2 readings with the least difference was calculated from the remaining readings.
Anthropometric measurements were collected with students dressed in light clothing without shoes. Study personnel used a stadiometer (model 214 Road Rod; Seca Corp., Hanover, MD) for height measurement, a scale (floor model 761, Seca Corp.) for weight measurement, Lange-type baseline skinfold calipers (AMG Medical Inc., Montreal, QC, Canada) for skinfold thickness measurement, and a standard tape measure just above the uppermost border of the right ilium at the end of normal expiration for waist circumference. Measurements were recorded to the nearest 0.1 cm for height, 0.2 kg for weight, 0.1 cm for waist circumference, and 0.5 mm for skinfold thickness. Two measurements were obtained for each participant. A third measurement was taken if the difference between the first 2 measurements exceeded 0.5 cm for height, 0.2 kg for weight, 0.5 cm for waist circumference, and 1 mm for skinfold thickness. All anthropometric measurements were repeated on every 10th student to assess inter-rater reliability (intraclass correlation coefficient = 0.99 for height and weight, 0.98 for waist circumference, and 0.95 for triceps and 0.96 for subscapular skinfold thickness). For all anthropometric measurements, the mean value of 2 measurements with the least difference was calculated. Body mass index (BMI) was computed as weight/height2 (kg/m2).
Sex-specific change patterns in mean BP and anthropometric characteristics (height, weight, BMI, waist circumference, and triceps and subscapular skinfold thickness) during 5 years were assessed using the Lowess curve smoothing technique, on the basis of a series of locally weighted means.26 We analyzed the repeated systolic BP measurements in 1999/2000, 2002, and 2004 using sex-specific individual growth modeling (SAS 9.1 PROC MIXED; SAS Institute, Cary, NC). This analytic approach assumes that each student has a unique linear change or increase in systolic BP with age. It estimates the average change over time (fixed effect of time) and the variation in initial levels of systolic BP and rates of change, which vary randomly among students (variance components). It can also account for time-varying predictors of the linear change in systolic BP, such as height and other anthropometric characteristics. Student's age represented time, and height was included in the models as a crude proxy measure for maturation.
Model 1 included the effect of time and 2 sources of variation—between- and within-student variation. Variance between students was further decomposed into variance in initial levels and in rates of SBP change. To assess whether changes in systolic BP during 5 years and variation in change patterns between students could be attributed to changes in anthropometric characteristics, model 2 included the time-varying effect of height. The time-varying effect of other anthropometric characteristics were then added to model 2 separately (BMI in model 3, waist circumference in model 4, and triceps and subscapular skinfold thickness in models 5 and 6, respectively). The effects of BMI, waist circumference, and skinfold thickness could not be tested simultaneously due to multicollinearity. All associations were adjusted for physical activity (self-reported number of physical activity sessions ≥5 minutes in the week preceding data collection, averaged over follow-up) and smoking (self-reported number of cigarettes smoked in the past week, averaged over follow-up). To formally test sex differences in these associations, we estimated models with interaction terms for sex and age, height, BMI, waist circumference, triceps, and subscapular skinfold thickness.
For all models, age and anthropometric characteristics were “centered” on their respective minimum values (closest integer) to facilitate the interpretation of the results (age was centered on 12 years, height on 130 cm, BMI on 13 kg/m2, triceps and subscapular skinfold thickness on 4 mm and 3 mm, respectively, and waist circumference on 45 cm). “Centering” a variable about its minimum value yields an intercept that is interpreted as an estimate of systolic BP at that value (eg, baseline values refer to systolic BP at age 12), rather than at 0.
Characteristics of the students at baseline (when they were 12-13 years old, on average) are presented in Table 1. Increases in mean systolic BP during 5 years were substantial in boys and modest in girls (Fig. 2). On average, systolic BP increased by 11.1 mm Hg (103.5 to 114.6 mm Hg) in boys and 3.2 mm Hg (103.1 to 106.3 mm Hg) in girls. Diastolic BP increased by 4.0 mm Hg (55.3 to 59.3 mm Hg) in boys and 2.6 mm Hg (56.0 to 58.6 mm Hg) in girls. Boys had substantial growth in height (21.3 cm; 154.0 to 175.3 cm). Height increased by 8.9 cm (154.0 to 162.9 cm) in girls, with the most substantial increases occurring before the age of 15, after which growth in height slowed. BMI increased by 3 units (kg/m2) in both boys and girls. Increases in waist circumference were greater in boys than in girls (8.4 cm vs. 7.0 cm). Girls experienced greater increases in triceps and subscapular skinfold thickness during 5 years than boys (7.9 and 6.7 mm vs. 0.5 and 3.9 mm).
Among boys, the average systolic BP level at age 12 was 103.6 mm Hg and systolic BP increased on average by 2.2 mm Hg with each year of age (Table 2, model 1), and thus, 11 mm Hg during 5 years (illustrated in Fig. 2). When individual changes in height during 5 years were taken into account (model 2), the rate of systolic BP change with each year of age was reduced by half (1.0 mm Hg). Systolic BP increased by 0.25 mm Hg with each centimeter increase in height (model 2). Because boys' height increased by 21 cm during 5 years, the increase in systolic BP of 0.25 mm Hg with each centimeter of height corresponds to an increase in systolic BP of 5.3 mm Hg during 5 years. Hence, half of the overall increase in systolic BP (11.1 mm Hg during 5 years) was explained by changes in boys' height. When individual changes in other anthropometric characteristics during 5 years were taken into account (models 3-6), an increase over time in 1 BMI unit, 1 cm waist circumference, or 1 mm triceps and subscapular skinfold thickness was associated with an increase in systolic BP of 0.7 mm Hg (model 3), 0.24 mm Hg (model 4), and 0.3 and 0.4 mm Hg (models 5 and 6, respectively).
There was no clear linear pattern in systolic BP changes during 5 years in girls. The parameter estimates of the rate of change with each year of age reflected the modest changes in systolic BP in girls (observed in Fig. 2). Although the average systolic BP level at age 12 in girls was similar to that of boys (104.1 mm Hg), systolic BP increased, on average, by 0.3 mm Hg with each year of age (Table 2, model 1). The estimate for the rate of change in systolic BP with age decreased to 0 (model 2), suggesting that virtually all increases in systolic BP were explained by changes in girls' height. The effect of height in girls was slightly lower than that observed in boys; systolic BP increased by 0.2 mm Hg with each centimeter increase in height (model 2).
The effect of changes in other anthropometric characteristics on changes in SBP in girls was similar to that observed in boys. An increase over time in 1 BMI unit, 1 mm of triceps and subscapular skinfold thickness, or 1 cm of waist circumference was associated with an increase in systolic BP of 0.7 mm Hg (model 3), 0.3 mm Hg (model 4), or 0.3 and 0.4 mm Hg (models 5 and 6, respectively). No important sex differences were observed in the association of systolic BP with height, BMI, waist circumference, and triceps and subscapular skinfold thickness (Table 2). Because there were no sex differences in age-related mean change patterns in diastolic BP, we did not explore the effect of changes in anthropometric characteristics on individual changes in diastolic BP over time.
There was substantial variation among boys in their initial levels of systolic BP (SD = 7.5), indicating unexplained heterogeneity in boys' systolic BP levels at age 12 (Table 2, model 1). However, there was little heterogeneity in boys' rates of systolic BP change (SD = 0.8, model 1). There was also substantial unexplained heterogeneity in girls' SBP levels at age 12 and little heterogeneity in their rates of change (SD = 6.9 and 0.6, respectively, model 1; Table 3). Variation in systolic BP levels at age 12 among boys was partly due to their differences in height and other anthropometric characteristics; it was reduced with the inclusion of height and BMI (SD = 6.7, model 3) and of height and waist circumference (SD = 6.8, model 4; Table 2). Similar to boys, the variation in initial levels among girls was reduced most substantially with the inclusion of height and BMI (SD = 6.4, model 3), and of height and waist circumference (SD = 6.5, model 4).
The within-student variance, which reflects the extent to which each student's observed systolic BP values deviate from his or her own estimated linear change pattern, was also substantial (SD = 7.3 in boys and 6.4 in girls), representing nearly half of the total variation in systolic BP. This component of the total variance represents the extent of measurement error (“noise”) in estimated linear change patterns. Only a small proportion of the within-student variance was systematically associated with changes in anthropometric characteristics.
We examined systolic BP changes during 5 years in a cohort of adolescents in relation to changes in anthropometric characteristics. Although boys and girls had similar mean systolic BP levels at age 12, systolic BP increased relatively rapidly with age in boys, but only modestly in girls. These sex differences in mean systolic BP progression are consistent with previous studies, which show that systolic BP increases were greater in girls until age 12. However, systolic BP increases level off and decline in late teens in girls, whereas BP continues to increase at a steeper rate for boys until age 18.8,27 In this study, steep increases in systolic BP in boys during 5 years were due in large part to changes in height, with half of the increase in systolic BP attributable to simultaneous changes in boys' height. Although increases in girls' height and systolic BP were modest, the effect of changes in girls' height on systolic BP change was of similar magnitude to boys and explained virtually all systolic BP increases in girls.
Increases in mean BMI were similar in boys and girls during 5 years. However, triceps and subscapular skinfold thickness (a crude measure of subcutaneous body fat) increased substantially in girls but not in boys, whereas increases in waist circumference (a crude measure of abdominal fat) were slightly higher in boys than in girls. These trends in anthropometric characteristics during this age period are consistent with recent reports.9,28 Although previous studies highlighted sex differences in mean change patterns in BP and anthropometric characteristics during adolescence,9 they did not directly attribute changes in BP to simultaneous changes in anthropometric characteristics. Results of our study show that despite differences in mean change patterns in systolic BP and anthropometric characteristics between boys and girls, the magnitude of the effect of changes in height, BMI, waist circumference, and skinfold thickness over time on systolic BP changes was similar in boys and girls. Thus, regardless of sex differences in mean trends in BP and adiposity indicators, the effect of gaining weight or body fat on BP change was the same in both sexes.
Although both abdominal and upper-body fat are strongly related to CVD risk factors in youth,29,30 the relationship with BP is stronger for abdominal adiposity than for subcutaneous adiposity.31,32 Waist circumference, in particular, has recently been advocated as being better than BMI as an adiposity indicator and as a predictor of BP and cardiovascular risk in youth.33–35 However, BMI, waist circumference, and subscapular skinfold thickness had similar effects on changes in systolic BP in our study. Although the parameter estimates differed (these variables were measured on different scales), the estimates of effect size were similar. The partial Eta-squared value, a standardized measure of the magnitude of the effect in regression analysis, was 0.6 and 0.7 for BMI in boys and girls, 0.5 and 0.6 for waist circumference in boys and girls, and 0.5 for subscapular skinfold thickness in both boys and girls. Our study showed no superiority of waist circumference or any other anthropometric characteristic in predicting BP changes in adolescents. Lastly, although waist circumference and skinfold thickness have been recently shown to have a weaker association with systolic BP in girls than in boys,8 our study did not find sex differences in the association of systolic BP with each of the anthropometric characteristics.
We found substantial between-person variation in BP at age 12, but little variation in rates of change (slopes) during adolescence. These findings indicate that, although BP levels may vary widely among adolescents at age 12, the course of BP changes or increases after age 12 does not vary, and each adolescent may already follow his or her own change trajectory. Although within-person variability was high in our study, this finding is consistent with the study by Shea et al36 and several cross-sectional studies37,38 that described within-person variability in youth's BP using mixed regression models.
The key strength of this study was the use of individual growth modeling. This approach has the advantage over analytic methods that rely on population-average measures (eg, generalized estimating equations [GEE]) because, in addition to providing an estimate of an average change pattern, it provides an estimate of between-student variation.39 Although BP change patterns vary among individuals,36 only a small number of longitudinal studies have taken into account this variability in BP.8,9,36 Using individual growth modeling of 6 repeated measurements of systolic BP in prepubertal children, Shea et al36 reported substantial variability in individual patterns of BP changes among children. They concluded that models that assume no individual variation in BP change patterns are not well suited for understanding BP progression in youth. Other strengths of our study include the longitudinal follow-up, repeated BP measurements during 5 years, and multiple readings of BP at each data collection.
Limitations of our study include the use of a convenience sample. However, although this could have implications for a prevalence or incidence study of elevated BP, it should not affect the magnitude of the relations between longitudinal changes in BP and its predictors. Other major studies of BP and cardiovascular health in youth (eg, Bogalusa Heart Study) were also based on convenience samples. Because BP levels in populations reported in other studies are not directly comparable (as they depend on BP measurement methods, the number of readings, time of day, etc.),40 the external validity of a convenience sample is better assessed using behavioral characteristics. Previous studies using this data set showed that study participants have health-related behaviors similar to those of a representative sample of Quebec youth.41
BMI, waist circumference, and triceps and subscapular skinfold thickness are crude indicators of adiposity in youth, representing overall, abdominal, and upper-body subcutaneous fat, respectively. Changes in body composition in youth over time are poorly reflected by changes in BMI32,42 because, unlike in adults, increases in BMI in children aged 8-18 years are primarily related to increases in lean (fat-free) mass rather than fat mass.28,43 In another longitudinal study, systolic BP was independently associated with fat mass and fat-free mass.8 Even so, the correlation between BMI and fat mass was r = 0.9, indicating that BMI is closely associated with fat mass.44 Although waist circumference and measures of skinfold thickness may offer an improvement over BMI in assessing changes in body composition in youth over time, they still may not reflect important changes in fat mass and fat-free mass during adolescence. Height was included in the models as a crude proxy measure of maturation. The growth spurt in height begins earlier in girls than in boys. This aspect of height would not be captured in the BMI values because they are intended to measure a different underlying construct—weight adjusted for height.45
Sex and age are the main determinants of systolic BP levels in youth.46 We conducted sex-specific analyses to minimize the possibility of residual confounding by sex when investigating systolic BP increases with age (exogenous variable).47 The majority of adolescents in our cohort were aged 12-13 years at baseline, but the age distribution ranged from 11 to 16 years. We centered age on 12 years in individual growth models to reflect changes in BP from age 12 to 17. Adjusting the association between systolic BP and age for intermediary variables, such as height and BMI, may not remove confounding effects of these factors because their association with systolic BP may be confounded by unobserved variables such as diet and physical activity (Fig. 1).48 The growth spurt in height begins earlier in girls than in boys, with almost all girls completing pubertal maturation by age 15.8,49
Difficulties in measuring BP in youth may lead to a misclassification of a student's “true” BP.50,51 Automated measurement devices, such as the Dinamap used in our study, provide several key advantages. Compared with mercury sphygmomanometer methods, they require relatively little training; they avoid observer bias due to digit preference or knowledge of prior readings; and they are relatively easy to use with small children because there is no need for auscultation. The systolic BP readings obtained with a Dinamap device correlate well with intra-arterial readings obtained by direct auscultatory method,52,53 and have been shown to provide accurate readings in children.54 A validation study in 52 youth, 8-16 years old with a BMI range of 15-45 kg/m2, found no systematic differences between the Dinamaps and mercury sphygmomanometers.55 Reliability of BP measurements can be improved by increasing the number of readings on each measurement occasion; in this study, each child had a minimum of 3 and up to 5 readings. Although automated devices can still lead to bias, it is likely to be of the same magnitude on different measurement occasions. Individual growth modeling takes into account within-student variability inherent in BP (ie, measurement error). Within-person variability has also been reported to be substantially lower (approximately half) in older children (13-18 years old) than in younger children (8-12 years old).38
Modeling longitudinal changes with 3 data points required an assumption that BP increases linearly with age, which precluded inclusion of a quadratic term. Although Shea et al36 suggest that the quadratic form may be more appropriate for modeling BP change in children, inclusion of a quadratic term may have been necessary in their study of prepubertal children aged 5 to 8 years to reflect the sharp increases in BP as children entered puberty. However, adolescents in our cohort were, on average, 12-13 years old at baseline, and most had already entered puberty. We had no measure of pubertal maturation; changes in height were used as a crude proxy for maturation.56 The effect of changes in height on SBP remained stable across models that included BMI, waist circumference, and skinfold thickness; and the correlation between height and other anthropometric characteristics in each model was low (eTable, http://links.lww.com/EDE/A376). Changes in height then explained a proportion of variance uncorrelated with the variance explained by other anthropometric characteristics.
The relationship between changes in weight and body fat during adolescence and simultaneous systolic BP increases has important clinical and public health implications for the burden of hypertension, particularly in the context of the current childhood obesity epidemic. The evidence is only beginning to emerge on sex differences in the patterns of SBP increases with age and their relationship to changes in anthropometric characteristics. Results of our study suggest that, although sex differences in mean BP changes during adolescence were largely attributable to differences in individual gains in height between boys and girls, the effect of gaining weight or body fat on BP change was similar in boys and girls.
1. Lawes CM, Vander Hoorn S, Law MR, Elliott P, MacMahon S, Rodgers A. Blood pressure and the global burden of disease 2000. Part II: estimates of attributable burden. J Hypertens
2. Burt VL, Whelton P, Roccella EJ, et al. Prevalence of hypertension in the US adult population. Results from the Third National Health and Nutrition Examination Survey, 1988-1991. Hypertension
3. Martins D, Nelson K, Pan D, Tareen N, Norris K. The effect of gender on age-related blood pressure changes and the prevalence of isolated systolic hypertension among older adults: data from NHANES III. J Gend Specif Med
. 2001;4:10-13, 20.
4. Wolf-Maier K, Cooper RS, Banegas JR, et al. Hypertension prevalence and blood pressure levels in 6 European countries, Canada, and the United States. JAMA
5. Mackay J, Mensah G. Atlas of Heart Disease and Stroke
. Geneva: World Health Organization; 2004.
6. Pilote L, Dasgupta K, Guru V, et al. A comprehensive view of sexspecific issues related to cardiovascular disease. CMAJ
7. Rodriguez BL, Labarthe DR, Huang B, Lopez-Gomez J. Rise of blood pressure with age. New evidence of population differences. Hypertension
8. Labarthe DR, Dai S, Fulton JE, Harrist RB, Shah SM, Eissa MA. Systolic and fourth-and fifth-phase diastolic blood pressure from ages 8 to 18 years: Project HeartBeat! Am J Prev Med
. 2009;37(1 suppl):S86–S96.
9. Moran A, Jacobs DR Jr, Steinberger J, et al. Changes in insulin resistance and cardiovascular risk during adolescence: establishment of differential risk in males and females. Circulation
10. Dasgupta K, O'Loughlin J, Chen S, et al. Emergence of sex differences in prevalence of high systolic blood pressure: analysis of a longitudinal adolescent cohort. Circulation
11. Maggisano V, Chiarotti F, Botunac I, Campanella C, Galietta G, Loizzo A. Adolescence as possible critical temporal window for blood pressure short-term monitoring in boys and girls. Eur J Epidemiol
12. Whelton PK. Epidemiology of hypertension. Lancet
13. Steinberger J. Diagnosis of the metabolic syndrome in children. Curr Opin Lipidol
14. Steinberger J, Daniels SR. Obesity, insulin resistance, diabetes, and cardiovascular risk in children: an American Heart Association scientific statement from the Atherosclerosis, Hypertension, and Obesity in the Young Committee (Council on Cardiovascular Disease in the Young) and the Diabetes Committee (Council on Nutrition, Physical Activity, and Metabolism). Circulation
15. Sorof J, Daniels S. Obesity hypertension in children: a problem of epidemic proportions. Hypertension
16. Davy KP, Hall JE. Obesity and hypertension: two epidemics or one? Am J Physiol Regul Integr Comp Physiol
17. Rocchini AP, Katch V, Anderson J, et al. Blood pressure in obese adolescents: effect of weight loss. Pediatrics
18. Kasa-Vubu JZ, Lee CC, Rosenthal A, Singer K, Halter JB. Cardiovascular fitness and exercise as determinants of insulin resistance in postpubertal adolescent females. J Clin Endocrinol Metab
19. Schmitz KH, Hong CP, Jacobs DR, Steinberger J, Moran A, Sinaiko AR. Physical activity is associated with insulin resistance in children. Circulation
20. Katzmarzyk PT. Obesity in Canadian children. CMAJ
. 2001;164:1563-1564; author reply 1564–1565.
21. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser
. 2000;894:i–xii, 1–253.
22. O'Loughlin J, DiFranza J, Tyndale RF, et al. Nicotine-dependence symptoms are associated with smoking frequency in adolescents. Am J Prev Med
23. Webber LS, Osganian V, Luepker RV, et al. Cardiovascular risk factors among third grade children in four regions of the United States. The CATCH Study. Child and Adolescent Trial for Cardiovascular Health. Am J Epidemiol
24. Labarthe DR. Project Heartbeat! Blood Pressure Manual of Operations
. Houston, TX: University of Texas Health Science Center at Houston; 1992.
25. Parati G, Mancia G. Assessing the white-coat effect: which blood pressure measurement should be considered? J Hypertens
26. Cleveland WS, Devlin SJ. Locally weighted regression: an approach to regression analysis by local fitting. J Am Stat Assoc
27. Brotons C, Singh P, Nishio T, Labarthe DR. Blood pressure by age in childhood and adolescence: a review of 129 surveys worldwide. Int J Epidemiol
28. Eissa MA, Dai S, Mihalopoulos NL, Day RS, Harrist RB, Labarthe DR. Trajectories of fat mass index, fat free-mass index, and waist circumference in children: Project HeartBeat! Am J Prev Med
. 2009;37(1 suppl):S34–S39.
29. Shear CL, Freedman DS, Burke GL, Harsha DW, Berenson GS. Body fat patterning and blood pressure in children and young adults. The Bogalusa Heart Study. Hypertension
30. Srinivasan SR, Myers L, Berenson GS. Rate of change in adiposity and its relationship to concomitant changes in cardiovascular risk variables among biracial (black-white) children and young adults: the Bogalusa Heart Study. Metabolism
31. Despres JP, Lemieux I. Abdominal obesity and metabolic syndrome. Nature
32. Prentice AM, Jebb SA. Beyond body mass index. Obes Rev
33. Maffeis C, Pietrobelli A, Grezzani A, Provera S, Tato L. Waist circumference and cardiovascular risk factors in prepubertal children. Obes Res
34. Maffeis C, Corciulo N, Livieri C, et al. Waist circumference as a predictor of cardiovascular and metabolic risk factors in obese girls. Eur J Clin Nutr
35. Genovesi S, Antolini L, Giussani M, et al. Usefulness of waist circumference for the identification of childhood hypertension. J Hypertens
36. Shea S, Rabinowitz D, Stein AD, Basch CE. Components of variability in the systolic blood pressures of preschool children. Am J Epidemiol
37. Burke GL, Webber LS, Shear CL, Zinkgraf SA, Smoak CG, Berenson GS. Sources of error in measurement of children's blood pressure in a large epidemiologic study: Bogalusa Heart Study. J Chronic Dis
. 1987; 40:83–89.
38. Rosner B, Cook NR, Evans DA, et al. Reproducibility and predictive values of routine blood pressure measurements in children. Comparison with adult values and implications for screening children for elevated blood pressure. Am J Epidemiol
39. Singer JD, Willet JB. Applied longitudinal data analysis: modeling change and event occurrence. New York: Oxford University Press; 2003.
40. Chiolero A, Bovet P, Paradis G, Paccaud F. Has blood pressure increased in children in response to the obesity epidemic? Pediatrics
41. O'Loughlin J, Tarasuk J, Difranza J, Paradis G. Reliability of selected measures of nicotine dependence among adolescents. Ann Epidemiol
42. Demerath EW, Schubert CM, Maynard LM, et al. Do changes in body mass index percentile reflect changes in body composition in children? Data from the Fels Longitudinal Study. Pediatrics
43. Maynard LM, Wisemandle W, Roche AF, Chumlea WC, Guo SS, Siervogel RM. Childhood body composition in relation to body mass index. Pediatrics
44. Fulton JE, Dai S, Steffen LM, Grunbaum JA, Shah SM, Labarthe DR. Physical activity, energy intake, sedentary behavior, and adiposity in youth. Am J Prev Med
. 2009;37(1 suppl):S40–S49.
45. Michels KB, Greenland S, Rosner BA. Does body mass index adequately capture the relation of body composition and body size to health outcomes? Am J Epidemiol
46. National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics
. 2004;114(2 suppl 4th report):555– 576.
47. Cole SR, Hernan MA. Fallibility in estimating direct effects. Int J Epidemiol
48. Kaufman JS, Maclehose RF, Kaufman S. A further critique of the analytic strategy of adjusting for covariates to identify biologic mediation. Epidemiol Perspect Innov
49. Grumbach MM, Styne DM. Puberty: ontogeny, neuroendocrinology, physiology, and disorders. In: Larsen PR, Kronenberg HM, Melmed S, Polonsky KS, eds. Williams Textbook of Endocrinology
. 10th ed. Philadelphia: Saunders, An Imprint of Elsevier; 2003.
50. Cook NR, Rosner BA, Chen W, Srinivasan SR, Berenson GS. Using the area under the curve to reduce measurement error in predicting young adult blood pressure from childhood measures. Stat Med
51. Menard SW, Park MK. Blood pressure measurement in children: a brief review. Semin Perioper Nurs
52. Park MK, Menard SM. Accuracy of blood pressure measurement by the Dinamap monitor in infants and children. Pediatrics
53. Gillman MW, Cook NR. Blood pressure measurement in childhood epidemiological studies. Circulation
54. Wattigney WA, Webber LS, Lawrence MD, Berenson GS. Utility of an automatic instrument for blood pressure measurement in children. The Bogalusa Heart Study. Am J Hypertens
55. Paradis G, Lambert M, O'Loughlin J, et al. Blood pressure and adiposity in children and adolescents. Circulation
56. Shankar RR, Eckert GJ, Saha C, Tu W, Pratt JH. The change in blood pressure during pubertal growth. J Clin Endocrinol Metab