Search tips
Search criteria 


Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
J Hum Hypertens. Author manuscript; available in PMC 2007 November 14.
Published in final edited form as:
PMCID: PMC2077359

Association between body composition and blood pressure in a contemporary cohort of 9-year-old children


Elevated blood pressure (BP) in children is an early risk factor for cardiovascular disease and is positively associated with body mass index (BMI). However, BMI does not distinguish between fat and lean masses, and the relationship of BP in children to different elements of body composition is not well established. BP, BMI and body composition were measured in 6863 children enrolled in the Avon Longitudinal Study of Parents and Children. Fat mass, lean mass and trunk fat were assessed using dual-energy X-ray absorptiometry. After full adjustment for confounders, total body fat and BMI were positively associated with systolic blood pressure (SBP) (β = 3.29, 95% confidence interval CI 3.02, 3.57 mm Hg/standard deviation (s.d.) and β = 3.97, 95% CI 3.73, 4.21 mm Hg/s.d., respectively) and diastolic blood pressure (DBP) (β = 1.26, 95% CI 1.05, 1.46 mm Hg/s.d. and β = 1.37, 95% CI 1.19, 1.54 mm Hg/s.d., respectively). SBP was also positively associated with lean mass (β = 3.38, 95% CI 2.95, 3.81 mm Hg/s.d.), and weakly associated with trunk fat (β = 1.42, 95% CI −0.06, 2.90 mm Hg/s.d., independent of total fat mass), which was robust in girls only. The association between lean mass and SBP remained even after accounting for fat mass. SBP in 9-year-old children is independently associated with fat mass and lean mass and, to a lesser extent, trunk fat in girls. In this analysis, because both fat and lean masses are associated with BP, BMI predicts BP at least as well as these components of body composition.

Keywords: ALSPAC, blood pressure, body composition, child, cohort study, X-ray densitometry


Elevated blood pressure (BP) during childhood and adolescence is associated with increased cardiovascular risk in later life1 and the development of early pathological lesions of atherosclerosis.2 However, it is not clear what determines BP during childhood. Although many studies have observed a positive relationship between body mass index (BMI) and BP in children and adolescents,3-7 it has been argued that BMI is a crude measure of adiposity and has been criticized for being unable to distinguish between fat and lean.8,9 Furthermore, correlations between BMI and body fat have been found to vary markedly among studies.10 This is particularly the case in children where BMI measurements are complicated by growth and development.11

Few studies have related components of body composition to BP in children, and even fewer studies have been carried out using direct measures of body composition. At present, associations reported between body fat and BP in children have been inconsistent, and the relationship with central adiposity is not clear.3,5,12-16 Additionally, there is some evidence that lean mass may be an important determinant of BP in children, however, studies have generally been small and based on skinfold estimates,3,5,14,17,18 which is not optimum. The aim of the present study was, therefore, to investigate in a large cohort, the relationship between BP and elements of body composition, using direct assessments of fat and lean mass from dual-energy X-ray absorptiometry (DXA).

Materials and methods

Study participants

The Avon Longitudinal Study of Parents and Children (ALSPAC) is a prospective cohort study investigating the health and development of children. A full description of the methodology is available elsewhere.19 Briefly, pregnant women residing in three health districts in Bristol, England with an expected date of delivery between 1 April 1991 and 31 December 1992 were invited to take part in the study. Of these women, 14 541 enroled and 13 678 had a singleton, liveborn child. Ethical approval for the study was obtained from the ALSPAC law and ethics committee, and three local research ethics committees covering the area. At 9 years of age, children were invited to the ALSPAC clinic for examination, during which BP and body composition were measured. These clinics were held between January 2001 and January 2003, and a total of 7725 children attended. Measurements for both systolic blood pressure (SBP) and diastolic blood pressures (DBP) were available for 7211 children, DXA measures were available for 6952 children, and 6863 had both BP and DXA measures.

BP measurements

SBP and DBP were measured using a Dinamap 9301 Vital Signs Monitor. The 1994 OPCS Dinamap Calibration Study20 indicated that this device was highly reliable with repeat measures yielding correlation coefficients of 0.88 for SBP and 0.83 for DBP. Corresponding correlations are similar but slightly greater for mercury sphygmomanometer although, as suggested in the OPCS report, correlations may be artificially high because of observer bias as subsequent BP measurements are likely to be influenced by previous readings, which the observer is likely to remember. Results from the 1994 OPCS study and other previous studies indicate that, compared with sphygmomanometers, BP values from the Dinamap instrument may be approximately 6-8 mm Hg greater for SBP and within 1 mm Hg difference for DBP. Additionally there is little consensus among studies on how differences between the devices vary across BP level.20 In the present study, a small adult size cuff was used for children with an upper arm circumference of less than 23 cm (97.8%) and an adult cuff for those within upper arm circumference of 23 cm or more (2.2%). Two readings were recorded for SBP and DBP and the mean for each measure was calculated. Except where indicated, all analyses were performed on the mean of the two readings. However, as it is sometimes recommended to exclude the initial measurement from oscillometric devices, the second BP reading was retained for a comparative analysis. The intra-class correlations (ICCs) for the mean SBP and DBP (N = 233 for both) were 0.58 and 0.34, respectively, reflecting the high intra-individual variability of BP measurements. Child’s demeanour during the BP measurement session (silent, talking or fidgeting), time of measurement (grouped into morning or afternoon) and room temperature were also recorded. Current age of the child was calculated from the date of clinic attendance and child’s date of birth.

Body composition measurements

Weight in underwear was measured to the nearest 0.2 kg using Tanita Body Fat Analyser (Model TBF 305; NB Body fat was not analysed using this instrument, this was used for the measurement of weight only) and height was measured to the nearest 0.1 cm using a Harpenden Stadiometer. From these measurements, BMI was calculated (weight/height2, with weight in kilograms and height in metres).

Fat mass, lean mass and trunk fat were measured to the nearest 0.01 g (minimum) using a Lunar Prodigy DXA fan beam scanner. For a detailed description of DXA technology see Nord and Payne.21 Whole body scans were conducted at the ALSPAC clinic with children wearing light clothing, ensuring that metal objects such as bracelets and watches were removed. All DXA operators were trained to the appropriate ionizing radiation medical exposure regulations standard and daily calibration checks were performed using a spine phantom according to the manufacturer’s guidelines. Scans were performed automatically and adult software was used to avoid software changes occurring at a later date. After acquisition, the whole body image was separated by the software into pre-determined regions - head, arms, legs and trunk. These images were evaluated and re-analysed as necessary, to ensure that borders between adjacent sub-regions were optimally placed. Trunk fat was estimated using the automatic region of interest that included chest, abdomen and pelvic area, as the ribs, spine and pelvis have been shown to have minimal influence on measurement of truncal soft tissue.21,22

Confounding factors

At initial enrolment during pregnancy into ALSPAC, mothers were asked to provide their date of birth, height, pre-pregnancy weight and indicate any history of hypertension. From these data mothers’ age at childbirth and maternal BMI were calculated. The date of last menstrual period was also reported at enrolment, and together with the actual date of delivery, was used to estimate gestational age. However, if this differed by more than 2 weeks from the estimate calculated from an early ultrasound assessment, then the latter was assumed to be the more accurate. Infant sex and birth weight were obtained from obstetric records and/or birth notifications. An antenatal questionnaire sent to the mother at 18 weeks requested details of all previous pregnancies resulting in either a live birth or stillbirth, which enabled parity to be derived. Mothers’ highest level of education (none/Certificate of Secondary Education (CSE), vocational, O level, A level or degree) was recorded in a questionnaire sent at 32 weeks gestation. CSE and O levels represent lower and higher levels of qualifications, respectively, taken at around 16 years of age. A levels represent the standard academic qualifications taken at around 18 years of age. Occupations of the mothers and their partners were also recorded by the mothers in the 32-week questionnaire and used for the classification of social class (I, II, III non-manual, III manual, IV and V, where I is the highest (professional) category and V (unskilled manual worker) the lowest) according to the 1991 Office and Population Censuses and Surveys standard.23 A single variable was derived from the highest class of both mother and partner. Details of puberty were gathered from child-based questionnaires sent to mothers when children were 9 years. Pubertal stage was classified based on development of pubic hair and breast development in girls, and genital development in boys.24 As the timing of the questionnaires did not necessarily coincide with the timing of attendance at the clinic, inclusion of puberty data was restricted to those collected within 16 weeks of the DXA measurement.

Statistical analyses

Means and standard deviations (s.d.) were calculated for continuous variables that were normally distributed. For variables that had skewed distributions, the geometric means and inter-quartile ranges are presented and proportions were calculated for categorical variables. Associations between confounding factors and both BP and body composition were analysed using linear regression. Skewed measures of body composition were log-transformed and all body composition measures were converted to standardized scores to enable comparisons across these measures. Standardized scores were calculated by subtracting the mean from the individual’s value and dividing by the s.d. (mean and s.d. based on the entire clinic sample). Associations with BP were explored using regression models for each body composition measure individually, in addition to simultaneous models incorporating both total fat mass and total lean mass together. Initial analyses of body composition and BP were carried out using minimally adjusted models. These models adjusted for child’s sex and child’s age, with the DXA measures further adjusted for height and height squared. Trunk fat was also adjusted for total body fat in all relevant models although separate analyses were additionally carried out without adjustments for total body fat. Fully adjusted models were carried out with additional adjustment for measurement factors (time of BP measurement, room temperature, child’s demeanour), maternal factors (age, hypertension, parity, BMI and height), social factors (social class, maternal education), birth weight and gestational age. Analyses were repeated restricting firstly to children with complete confounder information, and then to those in early puberty (all boys, and Tanner stage 1 or 2 in girls). Associations in boys and girls separately were compared by including interaction terms for gender and body composition variables in the model. In addition, the ICC was estimated using DXA measurements repeated on the same day that were available in 122 children. The ICC was then used to adjust regression coefficients for measurement error which can result in regression dilution bias.25 This adjustment was repeated using ICCs calculated for height (N = 369) and weight (N = 379), but not for ICCs calculated for SBP and DBP because, although the within-subject variability of SBP and DBP would result in reduced power, regression dilution bias occurs for measurement error in predictor variables and not outcomes. Analyses were performed using Stata version 9 and results are interpreted in relation to the strength of evidence against the null hypothesis, without reference to a P-value threshold for statistical significance (as suggested by Sterne and Davey Smith).26


The characteristics of all 6863 children (boys = 3401; girls = 3462) who attended the ALSPAC clinic are summarised in Tables Tables11 and and2.2. Although lean mass was normally distributed, the remaining measures of body composition (BMI, total fat mass and trunk fat) were positively skewed.

Table 1
Sample characteristics and potential confounders; continuous variables
Table 2
Sample characteristics and potential confounders; categorical variables

Where possible, differences in confounder information were compared between children who attended the ALSPAC clinic, and those who did not (data not shown). Children who did not attend had lower birth weight and gestational age. Mothers were slightly younger and shorter, but had similar BMI. A greater proportion of these mothers had a history of hypertension, three or more previous pregnancies, were less educated and were of lower socioeconomic status when compared with mothers of those who attended.

Based on results from tests of linearity, all categorical variables were analysed as continuous variables with exception of child’s demeanour. Associations of BP with potential confounders are summarised in Table 3. SBP was associated with all confounders, except birth weight and parity. DBP was associated with all confounders except for birth weight, gestational age, maternal height and time of measurement. For the relevant confounders, DXA measures of body composition were associated with all confounders with the exception of parity, maternal age and maternal hypertension and lean mass was additionally not associated with social class (data not shown).

Table 3
Regressions of BP on potential confounders

In minimally adjusted models, SBP was positively associated with all measures of body composition: total body fat, lean mass, trunk fat and BMI (see Table 4). The strongest association was with BMI and the weakest with trunk fat, independent of total fat. DBP (after minimal adjustment) was also positively associated with total fat mass and BMI, although with smaller effect sizes compared with SBP. The association between DBP and lean mass tended towards the null and there was no association between DBP and trunk fat after adjusting for total body fat. Adjusting for all confounders did not substantially alter the associations although there was a reduction in the association between SBP and trunk fat (independent of body fat) and the corresponding P-value.

Table 4
Regressions of BP on s.d.-scores of body composition

Associations of SBP with total fat mass and with total lean mass were of similar magnitude (minimally and fully adjusted models). In the simultaneous model, SBP was independently associated with both fat and lean and this model accounted for slightly more of the variation in SBP compared with that of BMI alone. For DBP, the simultaneous model also accounted for a similar amount of variation compared with BMI and, as DBP was not associated with lean mass, the simultaneous model did not account for more variation in BP than fat mass on its own.

Analyses were repeated after restricting to children with complete confounder information and observed associations were virtually unchanged. Results similar to the original analysis were also observed when analyses were restricted to children in early puberty. When the sexes were analysed separately, the association between SBP and total fat was greater in girls (β = 4.15, 95% CI 3.78, 4.52 mm Hg/s.d. unit compared with β = 3.04, 95% CI 2.74, 3.35 mm Hg/s.d. unit in boys, P<0.001 for interaction, minimally adjusted models). Additionally, SBP was more strongly associated with trunk fat in girls than in boys (β = 2.56, 95% CI 0.48, 4.65 mm Hg/s.d. unit compared with β = 1.26, 95% CI −0.44, 2.97 mm Hg/s.d. unit, P<0.001 for interaction, minimally adjusted models independent of total fat). Analyses were also repeated using the second BP reading instead of the mean of the first and second readings. Results remained the virtually unchanged (data not shown).

Prevalence of obesity in this population, classified according to the recent international cut-points determined by Cole et al.27 was 4.8% in girls and 3.9% in boys. Prevalences of overweight or obesity were 23.6% in girls and 17.6% in boys. However, analysis of SBP for quintiles of BMI, fat and lean mass (see Table 5) indicates that differences in mean SBP (compared with the lowest quintile of body composition) were observed across the range of quintiles and not solely in the most extreme quintile in which overweight and obese children are represented.

Table 5
Differences in mean SBP across quintiles of body compositiona

The ICCs for fat mass, lean mass and trunk fat were 0.99, indicating that there was little measurement error. Thus, there were only marginal increases in the regression coefficients after the relevant adjustment was made. (SBP: β = 3.54, 3.64, 2.16 and 4.00 mm Hg/s.d. unit for fat mass, lean mass, trunk fat (adjusted for body fat) and BMI, respectively; DBP: β = 1.40, 0.37, −0.11 and 1.38 mm Hg/s.d. unit for fat mass, lean mass, trunk fat and BMI, respectively (minimally adjusted models)). Similar results were observed (data not reported) after adjusting for measurement error related to height (ICC = 0.98) and weight (ICC = 0.99).


To our knowledge, this is the first study to assess the association between direct measures of body composition and BP in a contemporary cohort of this size. In this cohort of 6863 children aged 9 years, total body fat, lean mass, trunk fat and BMI were all positively associated with SBP and increased SBP was observed across all quintiles of body composition and not solely in the most extreme quintile in which overweight and obese children are represented. Additionally, total fat mass and total lean mass were found to predict SBP independently of each other. Associations with DBP were much weaker, and DBP was not associated with either trunk fat, when analysed independent of total fat or lean mass.

Although studies have reported associations between BP and adiposity, few studies have investigated the relationship with lean mass. The results from this study indicate that lean mass is an important predictor of BP in children. However, the reasons for this relationship are not clear. The effect of physical growth, muscle pressor reflexes and muscle composition have each been postulated as potential mechanisms.18 Furthermore, the relationship between lean mass and BP may be mediated through cardiac output (which is associated with obesity), as fat and lean mass are both independently associated with determinants of BP (cardiac output, stroke volume and peripheral resistance).16 This finding that lean mass is an independent determinant of BP is especially pertinent to pediatric research since age-related increases in body size during childhood generally reflect increases in lean mass rather than fat mass 11,28 and differences in BMI among thinner children can be largely owing to fat-free mass rather than fat mass.10 There is also evidence that lean mass may be an important contributor to adult BP,18,29 however, this requires further investigation as other studies report evidence to the contrary.30,31

Previous studies of BP and DXA measures of body composition in children have reported conflicting results. In one study of 920 children aged 5-18 years higher BP was associated with lower body fat,13 however, in another study body fat was not an independent correlate of BP (9- to 17-year-olds, N = 127).12 These studies also reported strong associations between BP and central adiposity, although in one study, this association was observed in boys only.13 In contrast, in the present study, associations between trunk fat and systolic BP were weak, and evidence for this association was robust in girls only.

Variations in sample size and participant age, pubertal stage and ethnicity may, in part, account for some of the differences observed among the previous studies and the present study. Changes in body composition occur during puberty and associations between body size and adiposity differ among children and adolescents of varying ages and sexual maturity.32,33 Thus, associations between body composition and BP may not be comparable across varying ages and stages of puberty during childhood. Additionally, whereas participants in the present sample were predominantly of European origin, those included in the previous DXA studies comprised a significant proportion of African-American and Asian children. Relationships of body size to blood pressure, and to surrogate measures of body fat, have been found to vary among ethnic groups 32,34,35 although, in the DXA studies above, the association between central adiposity and BP was not found to differ by ethnicity in one study 13 and, in the other, an independent relationship was reported with DBP but not SBP.12

In the present study, BMI was found to be the strongest predictor of SBP. This is surprising as BMI is a measure of body size that is merely a proxy for body fat. In contrast, DXA directly assesses both fat and lean mass, and the accuracy and precision of DXA has been validated using in vitro as well as in vivo techniques.36 However, in spite of its limitations, BMI was found to be at least as good as either measure of fat and lean at predicting BP in children. This may relate to the fact that BMI, and other measures of size, represents an overall measure of both fat and lean combined thus accounting, at least to some extent, for the independent effects of both elements of body composition.

In the present study, body size and composition showed stronger associations with SBP compared with DBP, and in some cases were associated with SBP only. This has also been observed in previous pediatric studies of BP and surrogate measures of body composition.3,4,6 Furthermore, observations from prospective epidemiological studies indicate that SBP in adults is a better indicator of cardiovascular risk compared with DBP.37-39 The different associations of SBP and DBP may relate to variations in the natural progression of the two measures with age. There is evidence from a study of ambulatory BP (measured using an oscillometric device) that SBP in children and adolescents is positively associated with age, independent of height and BMI, and that these age-related increases do not occur for DBP.40 This may reflect a greater sensitivity of SBP to cardiovascular risk factors (such as body size and composition) that may be evident from childhood.

In conclusion, this study aimed to determine the relationship between BP in children and different components of body composition, which are currently not well established. In this contemporary cohort of 9-year-old children, fat and lean mass were both independent predictors of SBP, and central adiposity was associated with SBP in girls. In addition, although the value and precision of BMI has been questioned, this basic index was a better predictor of SBP than either measure of body composition individually, possibly owing to BMI encompassing both fat and lean mass. Finally, differences in SBP were present across the range of quintiles for body composition and not solely in the most extreme quintile in which overweight and obese children were represented.

The positive relationship between BMI and BP is firmly established in the literature, however, the mechanisms of this association are not well understood. Although the predominant focus of these studies has generally been oriented around the effect of adiposity, the findings from the present study indicate lean mass may also play an important, independent role in determining BP. Further research on the mechanisms by which lean mass may affect BP is likely to improve the current understanding of how increased body size is related to elevated blood pressure. Additionally, as the cohort in this study were generally pre-pubertal children of approximately the same age, further studies are also required to determine how the relationship of BP with body composition in children of differs at varying ages and stages of puberty.


We thank to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laborotary technicians, clerical workers, research scientists, volunteers, managers, receptionists and nurses. The UK Medical Research Council, the Wellcome Trust and the University of Bristol provide core support for ALSPAC. This publication is the work of the authors and Marie-Jo Brion and Andy Ness will serve as guarantors for the contents of this paper. Marie-Jo Brion is jointly funded by the Overseas Research Students Awards Scheme and the University of Bristol. None of the authors had any financial or personal interest in any of the companies or organizations sponsoring this research. Contents of this article represent the views of the authors and not necessarily those of the funding bodies.


What is known about this topic

  • Body mass index is positively associated with bp in children, but does not distinguish fat and lean.3-7
  • Associations of blood pressure in children with total body fat and central adiposity are not clear.3,5,12-16
  • Lean mass may be associated with blood pressure but most of the evidence has come from surrogate measures using skinfold measurements.3,5,14,17,18

What this study adds

  • Lean mass and total body fat are each positively and independently associated with blood pressure in children and associations with adiposity are stronger in girls than in boys.
  • Body mass index predicts blood pressure at least as well as these components of body composition.
  • Differences in blood pressure are evident across the range of body composition quintiles and not solely in the extreme quintile in which overweight and obese children are represented.


1. Raitakari OT, Juonala M, Kahonen M, Taittonen L, Laitinen T, Maki-Torkko N, et al. Cardiovascular risk factors in childhood and carotid artery intima-media thickness in adulthood: the cardiovascular risk in young finns study. JAMA. 2003;290:2277–2283. [PubMed]
2. Berenson GS, Srinivasan SR, Bao W, Newman WP, III, Tracy RE, Wattigney WA. Association between multiple cardiovascular risk factors and atherosclerosis in children and young adults. The Bogalusa heart study. N Engl J Med. 1998;338:1650–1656. [PubMed]
3. Al-Sendi AM, Shetty P, Musaiger AO, Myatt M. Relationship between body composition and blood pressure in Bahraini adolescents. Br J Nutr. 2003;90:837–844. [PubMed]
4. Chen Y, Rennie DC, Reeder BA. Age-related association between body mass index and blood pressure: the humboldt study. Int J Obes Relat Metab Disord. 1995;19:825–831. [PubMed]
5. Leccia G, Marotta T, Masella MR, Mottola G, Mitrano G, Golia F, et al. Sex-related influence of body size and sexual maturation on blood pressure in adolescents. Eur J Clin Nutr. 1999;53:333–337. [PubMed]
6. Paradis G, Lambert M, O’Loughlin J, Lavallee C, Aubin J, Devlin E, et al. Blood pressure and adiposity in children and adolescents. Circulation. 2004;110:1832–1838. [PubMed]
7. Reich A, Muller G, Gelbrich G, Deutscher K, Godicke R, Kiess W. Obesity and blood pressure - results from the examination of 2365 schoolchildren in Germany. Int J Obes Relat Metab Disord. 2003;27:1459–1464. [PubMed]
8. Ellis KJ. Selected body composition methods can be used in field studies. J Nutr. 2001;131:1589S–1595S. [PubMed]
9. Wells JC. A critique of the expression of paediatric body composition data. Arch Dis Child. 2001;85:67–72. [PMC free article] [PubMed]
10. Freedman DS, Wang J, Maynard LM, Thornton JC, Mei Z, Pierson RN, Jr, et al. Relation of BMI to fat and fat-free mass among children and adolescents. Int J Obes (London) 2005;29:1–8. [PubMed]
11. Maynard LM, Wisemandle W, Roche AF, Chumlea WC, Guo SS, Siervogel RM. Childhood body composition in relation to body mass index. Pediatrics. 2001;107:344–350. [PubMed]
12. Daniels SR, Morrison JA, Sprecher DL, Khoury P, Kimball TR. Association of body fat distribution and cardiovascular risk factors in children and adolescents. Circulation. 1999;99:541–545. [PubMed]
13. He Q, Horlick M, Fedun B, Wang J, Pierson RN, Jr, Heshka S, et al. Trunk fat and blood pressure in children through puberty. Circulation. 2002;105:1093–1098. [PubMed]
14. Wilks RJ, Farlane-Anderson N, Bennett FI, Reid M, Forrester TE. Blood pressure in Jamaican children: relationship to body size and composition. West Indian Med J. 1999;48:61–68. [PubMed]
15. Mueller WH, Chan W, Meininger JC. Utility of different body composition indicators: demographic influences and associations with blood pressures and heart rates in adolescents (Heartfelt Study) Ann Hum Biol. 2003;30:714–727. [PubMed]
16. Daniels SR, Kimball TR, Khoury P, Witt S, Morrison JA. Correlates of the hemodynamic determinants of blood pressure. Hypertension. 1996;28:37–41. [PubMed]
17. Brandon LJ, Fillingim J. Body composition and blood pressure in children based on age, race, and sex. Am J Prev Med. 1993;9:34–38. [PubMed]
18. Julius S, Majahalme S, Nesbitt S, Grant E, Kaciroti N, Ombao H, et al. A ‘gender blind’ relationship of lean body mass and blood pressure in the Tecumseh study. Am J Hypertens. 2002;15:258–263. [PubMed]
19. Golding J, Pembrey M, Jones R. ALSPAC - the avon longitudinal study of parents and children. I. study methodology. Paediatr Perinat Epidemiol. 2001;15:74–87. [PubMed]
20. Bolling K. The Dinamap 8100 Calibration Study: A survey carried out by Social Survey Division of OPCS on behalf of the Department of Health. London: HMSO; 1994.
21. Nord RH, Payne RK. Body composition by dual-energy x-ray absorptiometry - a review of the technology. Asia Pacific J Clin Nutr. 1995;4:167–171. [PubMed]
22. Park YW, Heymsfield SB, Gallagher D. Are dual-energy X-ray absorptiometry regional estimates associated with visceral adipose tissue mass? Int J Obes Relat Metab Disord. 2002;26:978–983. [PubMed]
23. OPCS . Standard Occupational Classification. London: HMSO; 1991.
24. Tanner JM. Normal growth and techniques of growth assessment. Clin Endocrinol Metab. 1986;15:411–451. [PubMed]
25. Knuiman MW, Divitini ML, Buzas JS, Fitzgerald PE. Adjustment for regression dilution in epidemiological regression analyses. Ann Epidemiol. 1998;8:56–63. [PubMed]
26. Sterne JA, Davey Smith G. Sifting the evidence-what’s wrong with significance tests? BMJ. 2001;322:226–231. [PMC free article] [PubMed]
27. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320:1240–1243. [PMC free article] [PubMed]
28. Wells JC. A Hattori chart analysis of body mass index in infants and children. Int J Obes Relat Metab Disord. 2000;24:325–329. [PubMed]
29. Weinsier RL, Norris DJ, Birch R, Bernstein RS, Wang J, Yang MV, et al. The relative contribution of body fat and fat pattern to blood pressure level. Hypertension. 1985;7:578–585. [PubMed]
30. Berglund G, Ljungman S, Hartford M, Wilhelmsen L, Bjorntorp P. Type of obesity and blood pressure. Hypertension. 1982;4:692–696. [PubMed]
31. Siervogel RM, Roche AF, Chumlea WC, Morris JG, Webb P, Knittle JL. Blood pressure, body composition, and fat tissue cellularity in adults. Hypertension. 1982;4:382–386. [PubMed]
32. Daniels SR, Khoury PR, Morrison JA. The utility of body mass index as a measure of body fatness in children and adolescents: differences by race and gender. Pediatrics. 1997;99:804–807. [PubMed]
33. Deurenberg P, Weststrate JA, Seidell JC. Body mass index as a measure of body fatness: age- and sex-specific prediction formulas. Br J Nutr. 1991;65:105–114. [PubMed]
34. Deurenberg P, Yap M, van Staveren WA. Body mass index and percent body fat: a meta analysis among different ethnic groups. Int J Obes Relat Metab Disord. 1998;22:1164–1171. [PubMed]
35. Resnicow K, Futterman R, Vaughan RD. Body mass index as a predictor of systolic blood pressure in a multiracial sample of US schoolchildren. Ethn Dis. 1993;3:351–361. [PubMed]
36. Haarbo J, Gotfredsen A, Hassager C, Christiansen C. Validation of body composition by dual energy X-ray absorptiometry (DEXA) Clin Physiol. 1991;11:331–341. [PubMed]
37. Kannel WB. Elevated systolic blood pressure as a cardiovascular risk factor. Am J Cardiol. 2000;85:251–255. [PubMed]
38. Lewington S, Clarke R, Qizilbash N, Peto R, Collins R. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet. 2002;360:1903–1913. [PubMed]
39. Strandberg TE, Pitkala K. What is the most important component of blood pressure: systolic, diastolic or pulse pressure? Curr Opin Nephrol Hypertens. 2003;12:293–297. [PubMed]
40. Wuhl E, Witte K, Soergel M, Mehls O, Schaefer F. Distribution of 24-h ambulatory blood pressure in children: normalized reference values and role of body dimensions. J Hypertens. 2002;20:1995–2007. [PubMed]