Hostname: page-component-59f8fd8595-gl4p7 Total loading time: 0 Render date: 2023-03-21T17:12:21.171Z Has data issue: true Feature Flags: { "useRatesEcommerce": false } hasContentIssue true

Measuring growth and obesity across childhood and adolescence

Published online by Cambridge University Press:  15 January 2014

H. David McCarthy*
Public Health Nutrition Research Group, Faculty of Life Sciences & Computing, London Metropolitan University, Holloway Road, London N7 8DB, UK
Corresponding author: Professor H. D. McCarthy, fax 020 7133 4062, email
Rights & Permissions[Opens in a new window]


The dramatic rise in childhood obesity has driven the demand for tools better able to assess and define obesity and risk for related co-morbidities. In addition, the early life origins of non-communicable diseases including type 2 diabetes are associated with subtle alterations in growth and body composition, including total and regional body fatness, limb/trunk length and skeletal muscle mass (SMM). Consequently improved tools based on national reference data, which capture these body components must be developed as the limitations of BMI as a measure of overweight and obesity and associated cardiometabolic risk are now recognised. Furthermore, waist circumference as a measure of abdominal fatness in children is now endorsed by the International Diabetes Federation and National Institute for Clinical and Health Excellence for diagnostic and monitoring purposes. The present paper aims to review the research on growth-related variations in body composition and proportions, together with how national references for percentage body fat, SMM and leg/trunk length have been developed. Where collection of these measures is not possible, alternative proxy measures including thigh and hip circumferences are suggested. Finally, body ratios including the waist:height and muscle:fat ratios are highlighted as potential measures of cardiometabolic disease risk. In conclusion, a collection of national references for individual body measures have been produced against which children and youths can be assessed. Collectively, they have the capacity to build a better picture of an individual's phenotype, which represents their risk for cardiometabolic disease beyond that of the capability of BMI.

Conference on ‘Childhood nutrition and obesity: current status and future challenges’
Copyright © The Author 2014 


bioelectrical impedance analysis


fat mass


hip circumference


International Obesity Task Force


muscle:fat ratio


National Institute for Clinical and Health Excellence


sd score


skeletal muscle mass


appendicular SMM


thigh circumference


waist circumference

Within the disciplines of child health and nutrition, growth measurement has historically performed important clinical and public health functions, being an indicator of feeding and environmental circumstances of infants and children. As a result, tools have been developed to identify and monitor children and infants displaying poor growth, failure to thrive and malnutrition. These tools, mainly in the form of growth charts and reference data are still very much in use today, with continuing improvements in their diagnostic ability( Reference Cole 1 ). However, in the 21st century, a completely different nutritional and physical environment exists, whereby in countries such as Ireland and the UK, there is less of a problem with poor growth and under nutrition, but now a near epidemic of overweight and obesity. Thus, our focus has had to switch to measuring what could be termed ‘overgrowth’ in the form of excess body fat. Consequently, the tools needed to measure obesity have also had to adapt and improve to meet the requirements for population surveillance and clinical identification. Despite this shift in focus, the requirements for assessment tools remain unchanged, in that techniques should be simple, quick and inexpensive to perform, yet still meet the demands for accuracy, reproducibility, sensitivity and specificity. There is a need to avoid the misclassification of individuals and for the measure to be strongly related to morbidity or risk for later disease especially cardiometabolic disease. Indeed the leading causes of deaths in England and Wales in men aged 35 and above, and in women aged 65 and above is attributed to CHD and in the 35–64 years age range in women it is breast cancer( 2 ). Both of these non-communicable diseases have strong links with body composition especially excess body fat, obesity and body fat distribution.

Early growth, obesity and chronic disease risk

There is an additional dimension to measuring growth and obesity since it is now recognised that growth patterns and nutritional experience during fetal development and across infancy and childhood are associated with later risk for components of the metabolic syndrome, including type 2 diabetes, dyslipidaemia, non-alcoholic steatohepatitis (fatty liver) and hypertension. Much of this evidence has stemmed from the original epidemiological studies of Professor David Barker and his team where this pioneering research demonstrated a link between low birth weight and low weight at the age of 1 and later risk for hypertension, type 2 diabetes and CHD( Reference Barker 3 ). As an alternative to the thrifty genotype hypothesis for risk for type 2 diabetes, the thrifty phenotype hypothesis stemmed from the Barker studies. This hypothesis suggests that type 2 diabetes risk could arise, in part, through a non-genetic, vertical transmission from mother to offspring, whereby metabolic adaptation occurs during fetal life and early infancy. Here, it is proposed that a sensing of a nutritionally poor environment in utero triggers the fetus to adapt its metabolism in a direction better able to aid survival and withstand a similar nutritional environment post-natally. However, a conflict between this altered setting and exposure to an environment where food energy availability is abundant and demands for physical activity are low could set the individual on a life course trajectory to type 2 diabetes and cardiometabolic disease( Reference Hales and Barker 4 ).

Body composition, body dimensions and proportionality

With respect to patterns of physical growth, such an adaptation can also be reflected in subtle changes in body composition, including body fat distribution and body proportions and dimensions, which cannot always be easily identified and quantified by some assessment tools with poor sensitivity. Over the past 20 years, however, this initial hypothesis has become refined, with one key newer suggestion being that a lower birth weight coupled with a rapid post-natal growth (or rapid post-natal growth alone) promotes the accumulation of excess visceral adipose tissue mass and its associated risk for obesity-related morbidity( Reference Wells 5 ). Thus, when assessing and characterising growth and obesity, the tools should be able to capture both elements while ensuring they are also able to associate these measurements to the phenotype associated with cardiometabolic disease risk. Up until very recently, the accurate characterisation of this phenotype has been constrained by the lack of sensitive and accurate assessment tools. For example, infant and child growth can proceed at different rates, depending largely on the post-natal diet and environment, and can be quantified using weight-for-age and height-for-age charts.

BMI and BMI centile charts

The BMI and BMI charts based on reference data are used to classify individuals across childhood and adolescence that are underweight, of a healthy weight, overweight and obese. These classifications can be statistically based, typically with the 91st and 98th centiles used to define overweight and obese children, and the 2nd centile to define underweight( Reference Cole, Freeman and Preece 6 ). This form of classification has been improved through the development of the International Obesity Task Force (IOTF) cut-offs, where the centile curves defining overweight and obesity pass through the adult BMI cut-offs of 25 and 30 at age 18 years( Reference Cole, Bellizzi and Flegal 7 ). However, such assessment tools give no indication of the partitioning of dietary energy and nitrogen across growth into fat- and fat-free tissues; a division that can determine later disease risk. Furthermore, looking more closely at what lies behind BMI indicates that this measurement reflects a composite of a number of aspects of body composition and proportion, which individually can reflect variations in patterns of growth and disease risk. For example, within the BMI the following components of body structure and composition are amalgamated: Skeletal (bone) mass; Skeletal muscle mass (SMM); Organ mass; Adipose tissue mass; Body fat distribution; Limb length; Trunk length; Hydration of the fat-free mass; Stature.

Some of these components are known to have reciprocal effects on disease risk. For example (see later for more details), skeletal muscle and adipose tissue have opposing effects on insulin sensitivity and energy disposal and a deficit in muscle mass coupled with an excess of adipose mass could balance out each other with respect to body weight, with the possibility of leading the BMI to classify a child within the healthy range. Thus, the BMI is limited at an individual level as it correlates with both fat mass and fat-free mass, it has a low sensitivity, i.e. it fails to identify relatively large numbers of children with excess fat and it can underestimate obesity prevalence by the creation of false negatives. Furthermore, the BMI gives no indication of body fat distribution. This is important as an upper body or abdominal distribution of body fat carries a greater risk for cardiometabolic disease due to the accumulation of excess visceral adipose tissue, fat tissue with greater lipolytic activity drained by the portal system and with increased release of NEFA. This raised fatty acid release is considered a key contributor to decreased insulin sensitivity. These limitations of BMI are widely acknowledged but largely ignored. Nevertheless in the UK, the formal guidelines on obesity assessment produced by the National Institute for Clinical and Health Excellence (NICE) recommend the use of BMI as a practical estimate in children, even though it concedes that it should be interpreted with caution as it is not a direct measure of adiposity( 8 ). Thus, in child obesity surveillance such as the National Child Measurement Programme where BMI is the assessment tool used, prevalence figures can be under-reported, especially in areas of the country, for example, where a large proportion of children from South Asian backgrounds reside( 9 ). This stems from the observation that the South Asian phenotype is characterised by a higher fat mass with a more abdominal distribution of body fat and a lower SMM compared with white Europeans of the same BMI for age. Consequently, the WHO has now formally lowered the BMI thresholds for defining overweight (BMI = 23) and obesity (BMI = 25) in South Asian adults( 10 ). These lowered thresholds should also translate into lower BMI centile cut-offs to define overweight and obesity in South Asian childhood populations. However, a review of the evidence up to 2010 led to the conclusion that there was insufficient evidence to support separate BMI definitions for obesity in children and adolescents from South Asian ethnic groups in the UK( Reference Viner, Fry and Gupta 11 ). With advancing technological ability to characterise further the body composition of children and youths, together with evidence arising out of clinical referrals for type 2 diabetes diagnoses in children and youths from this ethnic background, it may well be timely to revisit this position, especially as one of the aims of developing assessment tools is to make them easily understandable and as user friendly as possible. A lack of harmonisation between child/adolescent and adult systems illustrates a major limitation with the current separate systems. For example upon reaching age 18, an individual of South Asian origin with a BMI of 25 would be classified using the IOTF cut-off as borderline healthy weight/overweight before the 18th birthday; but then switching to the adult system at age 18, the individual would now be classified as obese without any change in their BMI. Such are the limitations of BMI that its use in children and youths has now been brought to the attention of the general public through profiling of the issue within the media, with one prominent child obesity researcher claiming ‘it's just not a sufficiently sensitive measure’( 12 ). This statement mirrors similar observations in adults where it has been shown that in the USA, the prevalence of obesity is drastically underestimated when BMI is used compared with the more sophisticated technology of dual energy X-ray absorptiometry( Reference Shah and Braverman 13 ).

Abdominal adiposity and waist circumference centile charts

The recognition that abdominal adiposity and visceral fat relate to cardiometabolic disease risk in children and youths dates back to the mid-1990s. Prior to that time, it was assumed that abdominal obesity was a phenomenon of adulthood, especially the accumulation of excess visceral fat. However, the correlation of waist circumference (WC) with visceral fat accumulation in children has been demonstrated( Reference Brambilla, Bedogni and Moreno 14 ) and several studies beginning around the mid-1990s demonstrated that WC was a sensitive marker for several components of cardiometabolic disease. WC has been shown to be independently related to an adverse atherogenic lipoprotein profile in 12–14 year olds( Reference Flodmark, Sveger and Nilsson-Ehle 15 ) and an adverse (raised) fasting insulin concentration in 5–17 year olds( Reference Freedman, Serdula and Srinivasan 16 ). More recent studies have extended the relationship between WC and cardiometabolic disease components in children to include non-alcoholic fatty liver disease (non-alcoholic steatohepatitis)( Reference Manco, Bedogni and Marcellini 17 ), elevated blood pressure( Reference Choy, Chan and Chen 18 ), markers of inflammation( Reference Sinzinger, Shengelia and Schmid 19 ) and a clustering of components of the metabolic syndrome( Reference Hirschler, Aranda and Calcagno Mde 20 ). Such observations have led to a demand for national references for WC in children and youths to be produced, which could be used in both clinical and epidemiological contexts. The UK WC charts were produced in 2001( Reference McCarthy, Crawley and Jarrett 21 ) followed by a global expansion of equivalent charts in more than twenty countries, with some of the most recent additions including Malaysia, Germany and Poland( Reference Poh, Jannah and Chong 22 Reference Nawarycz, Krzyzaniak and Stawińska-Witoszyńska 24 ). Despite this expansion in awareness and utility of WC in obesity and cardiometabolic disease risk assessment, there remains a reluctance to fully embrace this assessment tool by some formal authorities. For example, NICE obesity guidelines state that WC is not recommended as a routine measure in children and youths but concedes that it may be used to give additional information on risk of developing long-term health problems, preferring instead to support the continued use of BMI( 8 ). Paradoxically, the recent NICE clinical guidelines on psychosis and schizophrenia in children and young people (CG155), recommend the monitoring of WC and hip circumference (HC) and use of WC and HC centile charts for patients receiving anti-psychotic drugs, due to the known effect of this type of pharmacotherapy on weight gain( 25 ). More encouragingly, the International Diabetes Federation supports the measurement of abdominal obesity as the ‘sine qua non’ for the diagnosis of metabolic syndrome in children and adolescents. It bases this statement on the data relating WC with components of the metabolic syndrome in children and the unequivocal evidence of the dangers of abdominal obesity in adulthood. The International Diabetes Federation states the WC > 90th centile cut-off as the diagnostic criterion as several studies have shown that these children are more likely to have multiple risk factors than those with a WC below this level( 26 ). A further potential improvement to the 90th WC centile cut-off has been the recent development of centiles for UK children which pass through the adult WC cut-offs at age 18 which relate to ‘increased risk’ (94 cm men, 80 cm women) and ‘high risk’ (102 cm men, 88 cm women)( Reference McCarthy and Samani-Radia 27 ). Such centiles are the equivalent to the IOTF BMI cut-offs and research is currently being conducted to evaluate the effectiveness of these new centiles.

Body fatness, body fat centile charts and bioimpedance technology

Up to this point, both BMI and WC are used as proxy measures of total and abdominal adiposity. The ideal assessment tool should directly assess adiposity particularly since pathology is driven, in part by excess adipose mass. However, until recently it has been difficult to quantify fat mass outside the laboratory but improvements in bioelectrical impedance technology and its incorporation into step-on scales have taken the assessment of fat and lean masses into the clinical and community settings. Although slightly less accurate compared with the more sophisticated laboratory techniques, they offer practical advantages in being simple to use, with a relatively low cost. Thus in 2006, the first body fat centile charts based on bioelectrical impedance analysis (BIA) technology were produced for the UK child and youth population( Reference McCarthy, Cole and Fry 28 ). These charts clearly illustrate the gender differences in total body fat accumulation as well as the pattern of accumulation across childhood and adolescence; differences that are obscured in BMI charts. Indeed, the shape of the curves matches the expected changes in body fat patterning across human growth and clearly illustrates the pubertal influences in boys and girls. These charts show, for example, that at age 18 girls have proportionately 60% more body fat than boys, with the median per cent values being 24·6 and 15·4%, respectively. Furthermore, to define useful cut-offs for clinically and epidemiologically purposes that broadly agree with the IOTF BMI cut-offs, the 85th and 95th centiles provided a reasonably close approximation to the IOTF overweight and obese boundaries, with the 2nd centile chosen to approximate the IOTF underweight equivalent. Thus these body fat boundaries define underfat, normal, overfat and obese. We are now seeing body fat centile charts emerging in other paediatric populations, for example, in Hong Kong and Turkish children( Reference Sung, So and Choi 29 , Reference Kurtoglu, Mazicioglu and Ozturk 30 ), suggesting the wider recognition of the importance of body fat assessment. One additional advantage of body fat charts over BMI charts is the reduction in the misclassification of children who can be categorised as healthy by the BMI despite having excess body fat. Similarly, large framed or overly muscled children and youths can be classed as overweight or obese by BMI but normal or healthy when based on body fat. Indeed, we have used these body fat charts to demonstrate the extent of misclassification by BMI and in a sample of 1088 children aged between 5 and 13 years, 31% of individuals were declared false negatives (i.e. excess body fat but healthy BMI) and only 4% were rated as false positives (normal fat but overweight BMI)( Reference Samani-Radia and McCarthy 31 ). To emphasise, it is the ‘hidden’ overfat children within the healthy BMI range, which creates the under estimation of true overweight and obesity prevalence in surveillance systems such as the National Child Measurement Programme.

Components of stature, obesity and chronic disease risk

Overall stature and the height component of BMI can be divided into two simple components: trunk length and leg length. The ratio of these two elements (expressed as the leg length:height ratio) varies with age across childhood and adolescence as well as varying between the sexes and ethnic groups. Mean leg length:height ratio increases between ages 5 and the pubertal stage, at which point leg length:height ratio declines, illustrating the relative growth of the limbs and trunk at different stages of child development( Reference Pliakas and McCarthy 32 ). Proportionality is now considered an additional dimension to how body size and composition can influence chronic disease risk. For example, components of cardiometabolic disease including type 2 diabetes, hypertension, and non-alcoholic steatohepatitis as well as overall CHD risk are associated with having both absolute and relatively shorter legs as well as shorter overall stature( Reference Smith, Greenwood and Gunnell 33 , Reference Whitley, Martin and Davey Smith 34 ). Leg growth appears to display a degree of plasticity in early life, with a suggestion that a certain proportion of leg growth can be sacrificed in favour of the growth of essential organs such as the brain (whose impaired growth would have far greater negative consequences) when pre- and early post-natal energy and nutrient availability may be compromised. This plasticity of growth in response to nutritional experience comprises a key element of the thrifty phenotype hypothesis( Reference Pomeroy, Stock and Stanojevic 35 ). However, caution must be exercised when relating relative leg length to prior early nutrition as it has been shown through nutritional supplementation trials in undernourished populations, that any increased growth in height as a result of supplementation was due almost entirely to increases in trunk length growth. However, the authors concluded that it may be that the timing of dietary intervention is crucial to influencing leg growth v. trunk growth( Reference Kinra, Sarma and Hards 36 ). Nevertheless, we have recently demonstrated that overweight and obesity in children and adolescents (assessed by either BMI sd score (SDS) or WC SDS) is associated with having relatively shorter legs (as illustrated by the leg length:height ratio)( Reference Pliakas and McCarthy 32 ). The corollary of this is that overweight and obese children tend to have a greater relative trunk length, therefore suggesting a greater potential for higher truncal fat accumulation. Thus, altered limb and trunk proportions appear to be linked to measures of overweight and obesity, again illustrating that overweight/obesity, at least in part, can be a consequence of altered patterns of growth( Reference Liu, Akseer and Faught 37 ). Thus, more epidemiological studies of are incorporating measures of leg and trunk length into study design and analyses, supported by the availability of national references for leg length and sitting height in children and adolescents( Reference Dangour, Schilg and Hulse 38 ).

Child and adolescent height for age and body fatness

Aside from the alterations in the components of stature and proportionality, overweight and obese children (defined by their BMI SDS) also tend to be taller for their age compared with non-overweight children, with this observation even being seen in early childhood( Reference Pliakas and McCarthy 32 , Reference Freedman, Thornton and Mei 39 ). Thus, being tall for age could be considered an early marker of risk for overweight and obesity. It had been initially assumed that this height for age-overweight/obese relationship was simply a characteristic of using BMI as the proxy measure for overweight/obesity, with BMI tending to increase with increasing height and that BMI preferentially classified taller children as overweight or obese. However, this height-for-age-increased adiposity phenomenon has recently been confirmed in a small sample (280 children aged 7–12) in whom, body composition was measured using dual energy X-ray absorptiometry, a laboratory-based reference technique that quantifies tissue masses( Reference Metcalf, Hosking and Frémeaux 40 ). Using the simpler field-based method of BIA, conducted in the community on a larger sample (n 2298) of children and covering a wider age range (4·5–14 years) we have confirmed the findings of the laboratory-based study that children who are taller for their age also tend to have higher body fat levels for their age( Reference Navti, Samani-Radia and McCarthy 41 ). In this study, children were divided into quartiles of height SDS, and mean percentage body fat increased from 18·2% in the first quartile, to 21% in the fourth quartile. In addition, prevalence of overfat/obesity increased from 9·4% in the first quartile of height-for-age, to 29·7% in the fourth quartile of height-for-age. This height-for-age pattern of prevalence was mirrored when BMI was used as the measure of overweight/obesity as well as when WC is used as the measure of central overweight/obesity. It is possible that excess weight and fat gain in childhood is driving additional gains in height and suggests that if energy intake is more than sufficient to promote a rapid linear growth, the excess may be stored as fat, resulting in a positive relation between height and adiposity( Reference Freedman, Thornton and Mei 39 ). This drive in height growth is not a universal observation however, as it must be remembered that overweight and obese children were present (albeit to a lower extent) in the lower quartiles of height-for-age and in their cases diet was not driving height gain. However, caution must be exercised when attempting to relate height for age (cross-sectional) with growth rates (longitudinal) as genetics (parental height) is the major determinant of height growth with environment (diet) playing a minor part. In addition, rapid linear growth might also lead to advanced skeletal maturation( Reference Demerath, Jones and Hawley 42 ). Advances in skeletal age in overweight or obese individuals appear to precede increases in height, starting in mid-childhood, suggesting that factors involved with physical maturation are aetiologically linked with overweight risk( Reference Johnson, Stovitz and Choh 43 ). Thus, childhood height could be used to identify more accurately children who may be likely to become overweight or obese in later life, a suggestion previously proposed by others, where height measurement could not only improve the identification of obese children but that special attention be paid to children who have a high BMI and are also taller for their age( Reference Johnson, Stovitz and Choh 43 Reference Stovitz, Hannan and Lytle 45 ).

Socio-economic status, height-for-age and obesity

Risk for overweight and obesity is also strongly related to socio-economic status, with children from a lower income background demonstrating a higher prevalence compared with children from a higher income background. When examining socio-economic status influences on obesity prevalence, one overlooked contributory factor is the shorter height-for-age in children from a lower income background compared with those from a higher income background. Thus, at similar body weights across income groups, BMI would be greater in children from a lower income background. This shorter stature effect on BMI would be compounded when body weight-for-age is higher compared with those from a higher income background. However, as discussed earlier, overweight and obese children tend to be taller for age compared with children who are of a healthy weight. Thus, for overweight and obese children from a low-income background, there can be opposing influences on their height-for-age. Employing BIA to quantify adiposity, overfat/obese children from a low-income background have been shown to be significantly taller than their healthy weight peers, but significantly shorter than overfat/obese children from a higher income background( Reference Samani-Radia and McCarthy 46 ). Similar income group differences in height and hence BMI have also been observed in a Canadian population, indicating this phenomenon to be reproducible across population groups( Reference Moffat, Galloway and Latham 47 ). Clearly, shorter height is a second (and likely a more significant) contributor to the greater BMI, BMI z-score and overweight/obesity prevalence in lower income children. These studies illustrate how height-for-age measurement can reveal potential growth-related issues which can impact on adiposity and risk for overweight/obesity.

Skeletal muscle, cardiometabolic disease and muscle mass centile charts

Up to now, development of growth/body composition assessment charts has focused on measures of adiposity, especially as this tissue is a key driver of risk for cardiometabolic disease. However, skeletal muscle, which has been largely overlooked in epidemiological and laboratory studies, is gaining greater recognition as an independent marker of metabolic health, with muscle strength for example, being positively related to greater insulin sensitivity( Reference Benson, Torode and Singh 48 ) and low muscle fitness associated with greater metabolic risk( Reference Steene-Johannessen, Anderssen and Kolle 49 ). Indeed, skeletal muscle is a major site of insulin-mediated glucose disposal and therefore an important regulator of whole-body glucose homoeostasis. Beyond metabolic health, sarcopenia (age-related loss of muscle mass) is emerging as an important clinical and public health issue. Sarcopenia is the major cause of decreased mobility and function in old age, with quality of life diminishing due to a loss of independent living. A key to delaying age-related SMM loss is the optimal buildup of muscle across childhood and adolescence and the maintenance of this mass throughout adulthood( Reference Sayer, Syddall and Martin 50 ), a pattern that has long been recognised for maintaining bone mass and delaying the risk for osteoporosis. Thus for clinical and epidemiological purposes, it has become important to define and identify healthy (and by definition unhealthy) levels of SMM in children and youths. However, due to the difficulty in measuring SMM outside of the laboratory, there has been a lack of age- and gender-specific normative data. Using the same BIA technology used to develop the UK body fat reference charts, (which has the ability to produce regional or segmental body composition), we have been able to quantify the SMM in the arms and legs of children and youths (defined as appendicular skeletal muscle mass (SMMa)) which acts as a good proxy for whole-body SMM and to generate centile curves for SMMa( Reference McCarthy, Samani-Radia and Jebb 51 ). Three formats of SMMa curves have been constructed, which can be used to support different purposes, namely absolute SMMa (kg), SMMa expressed as a percentage of total body mass (%SMMa) and SMMa as a percentage of fat-free mass (SMMa/fat-free mass (kg) × 100). In addition, centile curves for %fat-free mass have been generated and reflect the reciprocal of %fat mass (FM). Thus, users now have the opportunity to assess body composition in children and youths against a reference population with greater detail compared with equivalent BMI charts. These charts can allow the tracking of SMMa in a clinical setting similar to the way height and weight charts are currently used. The reference data can also be used for epidemiological purposes. SMMa charts have also been produced in a Canadian population( Reference Webber and Barr 52 ) and it is likely that we shall see them appearing for different global child and youth populations in due course. However, where BIA facilities for quantifying SMMa are not available or appropriate, alternative and simpler prediction tools need to be developed.

Thigh and hip circumference measurement

There is potential for thigh circumference (TC) to act as a proxy for leg SMM. Approximately 75% of whole-body SMM is found within the arms and legs, with the group of muscles in the upper leg constituting a major fraction of total leg muscle mass. A recent observation in adults revealed that a smaller TC, which most likely reflects a low thigh muscle mass is associated with the development of cardiovascular morbidity and early mortality( Reference Heitmann and Frederiksen 53 ). With this in mind, reference data for TC across childhood and adolescence could serve a useful purpose and the first set of TC charts for the UK paediatric population have been developed( Reference McCarthy 54 ). For completion, HC centile charts have also been developed in the UK( Reference McCarthy, Rigby and Stasinopoulos 55 ), which should prove valuable in the future, especially as NICE guidelines on psychosis and schizophrenia in children and young people (CG155) also recommend the measurement of HC together with WC in patients receiving anti-psychotic pharmacotherapy( 25 ).

Muscle:fat and waist:height ratios

Now that both FM and SMMa can be quantified with reasonable accuracy in girls and boys across the age spectrum, this has facilitated the development of a new potential measure of cardiometabolic risk, the muscle:fat ratio (MFR). This measure, originally proposed by Professor Andrew Prentice at the London School of Hygiene and Tropical Medicine is based on the rationale that FM and SMM have opposing effects on insulin sensitivity and energy disposal, with the balance between them more accurately able to determine cardiometabolic risk beyond the ability of BMI. An individual having both a low SMM and high FM would have a MFR towards the low end of the MFR spectrum and may present a risk for metabolic disease greater than having excess FM alone. We now have generated reference data on the MFR distribution in boys and girls across childhood and adolescence( Reference McCarthy, Samani-Radia and Jebb 51 ). In our population, the mean MFR in boys equals 1·99 (sd 0·60) with a range between 0·69 and 4·84. The equivalent values in girls are 1·38 (sd 0·40), range 0·49–3·22. In addition, MFR is significantly and positively age related in boys and negatively related in girls. MFR also decreases with increase in BMI z-score. An MFR of 1·25 in boys, 1·10 in younger girls and 0·80 in older girls can be used to define the lower limit of normal (mean minus 2 sd). It should however be noted that since the gluteofemoral adipose mass has been shown to be protective for metabolic health, the relationships between MFR and metabolic disease risk may be non-linear( Reference Frayn 56 , Reference Manolopoulos, Karpe and Frayn 57 ) and replacing the total body fat with abdominal body fat in this ratio could further improve this measure. However, this proposition also needs to be systematically evaluated, although it is interesting to note that a ratio of SMM to visceral fat area ratio has been shown to be associated with metabolic syndrome and arterial stiffness in a Korean elderly population( Reference Kim, Park and Lim 58 ). This MFR follows a similar principle of the more established waist:height ratio, which has been shown to be effective across all ages and ethnicities as a screening tool for the prediction of cardiovascular diseases and type 2 diabetes, with a waist:height ratio of 0·50 performing as a suitable global boundary value( Reference Browning, Hsieh and Ashwell 59 ).


Given the importance of cardiometabolic disease as a major global public health and clinical condition, a need has arisen for better tools to monitor individual growth patterns, evaluate at-risk body composition and identify those most at risk of developing the metabolic components of the disease. A suite of individual measures, with national references, against which children and youths can be assessed have now been produced. Used collectively, they have the capacity to build a better picture of an individual's phenotype, which represents their risk for cardiometabolic disease beyond that of the capability of BMI. In the immediate future, most likely some of these tools will find greatest utilisation in a research capacity where they can be better evaluated and refined. In time, the associations of these body measures with components of the metabolic syndrome should be better defined through evaluation of their sensitivity, specificity and validity. Through wider use and awareness and their incorporation into clinical practice, they should be able to stimulate support by policy makers at national and international levels.


The author wishes to acknowledge the contribution of Professor Andrew Prentice, Dr Susan Jebb, Dr Dimple Samani-Radia and research students Triantafyllos Pliakas and Lifoter Navti for their contribution to data collection, analyses and publications cited in this review.

Financial Support

The financial support from Tanita Eu, the North Central London Research Consortium, University of North London Diversity & Development Fund and London Metropolitan University is gratefully acknowledged. The funding providers had no role in the design, analysis or writing of this article.

Conflicts of Interest

The author has received funding from Tanita Eu for the production of the body fat and skeletal muscle mass centile charts.


H. D. Mc. solely contributed to the writing of this manuscript.


1. Cole, TJ (2012) The development of growth references and growth charts. Ann Hum Biol 39, 382394.CrossRefGoogle ScholarPubMed
3. Barker, DJP (1998) Mothers, Babies, and Disease in Later Life. Edinburgh: Churchill Livingstone.Google Scholar
4. Hales, CN & Barker, DJ (1992) Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. Diabetologia 35, 595601.CrossRefGoogle ScholarPubMed
5. Wells, JC (2011) The thrifty phenotype: an adaptation in growth or metabolism? Am J Hum Biol 23, 6575.CrossRefGoogle ScholarPubMed
6. Cole, TJ, Freeman, JV & Preece, MA (1995) Body mass index reference curves for the UK, 1990. Arch Dis Child 73, 2529.CrossRefGoogle ScholarPubMed
7. Cole, TJ, Bellizzi, MC, Flegal, KM et al. (2000) Establishing a standard definition for child overweight and obesity worldwide: international survey. Br Med J 320, 12401245.CrossRefGoogle ScholarPubMed
8. NICE clinical guideline 43 (2006) Obesity: guidance on the prevention, identification, assessment and management of overweight and obesity in adults and children. Google Scholar
9. National Child Measurement Programme: 2010/11 school year, headline results; available at Google Scholar
10. WHO Expert Consultation (2004) Appropriate body mass index for Asian populations and its implications for policy and intervention strategies. Lancet 363(9403), 157163.CrossRefGoogle Scholar
11. Viner, RM, Fry, T, Gupta, S et al. (2010) Insufficient evidence to support separate BMI definitions for obesity in children and adolescents from south Asian ethnic groups in the UK. Int J Obes Relat Metab Dis (Lond) 34, 656658.CrossRefGoogle ScholarPubMed
13. Shah, NR & Braverman, ER (2012) Measuring adiposity in patients: the utility of body mass index (BMI), percent body fat, and leptin. PLoS ONE 7, e33308.CrossRefGoogle Scholar
14. Brambilla, P, Bedogni, G, Moreno, LA et al. (2006) Cross validation of anthropometry against magnetic resonance imaging for the assessment of visceral and subcutaneous adipose tissue in children. Int J Obes Relat Metab Dis (Lond) 30, 2330.CrossRefGoogle Scholar
15. Flodmark, CE, Sveger, T & Nilsson-Ehle, P (1994) Waist measurement correlates to a potentially atherogenic lipoprotein profile in obese 12–14-year-old children. Acta Paediatr 83, 941945.CrossRefGoogle ScholarPubMed
16. Freedman, DS, Serdula, MK, Srinivasan, SR et al. (1999) Relation of circumferences and skinfold thicknesses to lipid and insulin concentrations in children and adolescents: the Bogalusa Heart Study. Am J Clin Nutr 69, 308317.CrossRefGoogle ScholarPubMed
17. Manco, M, Bedogni, G, Marcellini, M et al. (2008) Waist circumference correlates with liver fibrosis in children with non-alcoholic steatohepatitis. Gut 57, 12831287.CrossRefGoogle ScholarPubMed
18. Choy, CS, Chan, WY, Chen, TL et al. (2011) Waist circumference and risk of elevated blood pressure in children: a cross-sectional study. BMC Public Health 11, 613.CrossRefGoogle ScholarPubMed
19. Sinzinger, H, Shengelia, N, Schmid, P et al. (2007) Waist circumference in Viennese school children is closely related to inflammatory, endothelial, and oxidation markers. Circulation 115, E243 (abstract).Google Scholar
20. Hirschler, V, Aranda, C, Calcagno Mde, L et al. (2005) Can waist circumference identify children with the metabolic syndrome? Arch Pediatr Adolesc Med 159, 740744.CrossRefGoogle ScholarPubMed
21. McCarthy, HD, Crawley, HF & Jarrett, KV (2001) Development of waist circumference percentiles in British children aged 5·0–16·0 years. Eur J Clin Nutr 55, 902907.CrossRefGoogle Scholar
22. Poh, BK, Jannah, AN, Chong, LK et al. (2011) Waist circumference percentile curves for Malaysian children aged 6–16y. Int J Pediatr Obes 6, 229235.CrossRefGoogle Scholar
23. Kromeyer-Hauschild, K, Dortschy, R, Stolzenberg, H et al. (2011) Nationally representative waist circumference percentiles in German adolescents aged 11·0–18·0 years. Int J Pediatr Obes 6, e129e137.CrossRefGoogle ScholarPubMed
24. Nawarycz, LO, Krzyzaniak, A, Stawińska-Witoszyńska, B et al. (2010) Percentile distributions of waist circumference for 7–19-year-old Polish children and adolescents. Obes Rev 11, 281288.CrossRefGoogle ScholarPubMed
25. NICE clinical guideline 155 (2013) Psychosis and schizophrenia in children and young people. Google Scholar
26.The IDF consensus definition of the metabolic syndrome in children and adolescents. International Diabetes Federation 2007. Google Scholar
27. McCarthy, HD & Samani-Radia, D (2013) Development of age- and gender-related waist circumference percentile cut-offs equating to adult cut-offs at age 18 years. Obes Facts 6, Suppl. 1, P143.Google Scholar
28. McCarthy, HD, Cole, TJ, Fry, T et al. (2006) Body fat reference curves for children. Int J Obes Relat Metab Dis 30, 598602.CrossRefGoogle ScholarPubMed
29. Sung, RY, So, HK, Choi, KC et al. (2009) Body fat measured by bioelectrical impedance in Hong Kong Chinese children. Hong Kong Med J 15, 110117.Google ScholarPubMed
30. Kurtoglu, S, Mazicioglu, MM, Ozturk, A et al. (2010) Body fat reference curves for healthy Turkish children and adolescents. Eur J Pediatr 169, 13291335.CrossRefGoogle ScholarPubMed
31. Samani-Radia, D & McCarthy, HD (2009) Bioelectrical impedance analysis exposes the overfat children classified as normal weight by the Body Mass Index. Proc Nutr Soc 69, E120.CrossRefGoogle Scholar
32. Pliakas, T & McCarthy, HD (2010) Relative leg length and total and regional body fatness in children. Ann Hum Biol 37, 1022.CrossRefGoogle Scholar
33. Smith, GD, Greenwood, R, Gunnell, D et al. (2001) Leg length, insulin resistance, and coronary heart disease risk: the Caerphilly Study. J Epidemiol Commun Health 55, 867872.CrossRefGoogle ScholarPubMed
34. Whitley, E, Martin, RM, Davey Smith, G et al. (2012) The association of childhood height, leg length and other measures of skeletal growth with adult cardiovascular disease: the Boyd-Orr cohort. J Epidemiol Commun Health 66, 1823.CrossRefGoogle ScholarPubMed
35. Pomeroy, E, Stock, JT, Stanojevic, S et al. (2012) Trade-offs in relative limb length among Peruvian children: extending the thrifty phenotype hypothesis to limb proportions. PLoS ONE 7, e51795.CrossRefGoogle ScholarPubMed
36. Kinra, S, Sarma, KV, Hards, M et al. (2011) Is relative leg length a biomarker of childhood nutrition? Long-term follow-up of the Hyderabad Nutrition Trial. Int J Epidemiol 40, 10221029.CrossRefGoogle ScholarPubMed
37. Liu, J, Akseer, N, Faught, BE et al. (2012) Use of leg length to height ratio to assess the risk of childhood overweight and obesity: results from a longitudinal cohort study. Ann Epidemiol 22, 120125.CrossRefGoogle ScholarPubMed
38. Dangour, AD, Schilg, S, Hulse, JA et al. (2002) Sitting height and subischial leg length centile curves for boys and girls from Southeast England. Ann Hum Biol 29, 290305.CrossRefGoogle ScholarPubMed
39. Freedman, DS, Thornton, JC, Mei, Z et al. (2004) Height and adiposity in children. Obes Res 12, 846852.CrossRefGoogle Scholar
40. Metcalf, BS, Hosking, J, Frémeaux, AE et al. (2011) BMI was right all along: taller children really are fatter (implications of making childhood BMI independent of height) EarlyBird 48. Int J Obes Relat Metab Dis 35, 541547.CrossRefGoogle ScholarPubMed
41. Navti, LK, Samani-Radia, D & McCarthy, HD (2013) Children's body fatness and prevalence of obesity in relation to height for age. Ann Hum Biol; Epublication 25 November 2013.Google ScholarPubMed
42. Demerath, EW, Jones, LL, Hawley, NL et al. (2009) Rapid infant weight gain and advanced skeletal maturation in childhood. J Pediatr 155, 355361.CrossRefGoogle ScholarPubMed
43. Johnson, W, Stovitz, SD, Choh, AC et al. (2011) Patterns of linear growth and skeletal maturation from birth to 18 years of age in overweight young adults. Int J Obes Relat Metab Dis 36, 535541.CrossRefGoogle ScholarPubMed
44. Stovitz, SD, Pereira, MA, Vazquez, G et al. (2008) The interaction of childhood height and childhood BMI in the prediction of young adult BMI. Obesity 16, 2336–2241.CrossRefGoogle ScholarPubMed
45. Stovitz, SD, Hannan, PJ, Lytle, LA et al. (2010) Child height and the risk of young adult obesity. Am J Prev Med 38, 7477.CrossRefGoogle ScholarPubMed
46. Samani-Radia, D & McCarthy, HD (2010) Socioeconomic status and body fatness in children: role of height growth limitation. Int J Obes Relat Metab Dis 35, 128133.CrossRefGoogle Scholar
47. Moffat, T, Galloway, T & Latham, J (2005) Stature and adiposity among children in contrasting neighborhoods in the City of Hamilton, Ontario, Canada. Am J Hum Biol 17, 355367.CrossRefGoogle ScholarPubMed
48. Benson, AC, Torode, ME & Singh, MA (2006) Muscular strength and cardiorespiratory fitness is associated with higher insulin sensitivity in children and adolescents. Int J Pediatr Obes 1, 222231.CrossRefGoogle ScholarPubMed
49. Steene-Johannessen, J, Anderssen, SA, Kolle, E et al. (2009) Low muscle fitness is associated with metabolic risk in youth. Med Sci Sports Exerc 41, 13611367.CrossRefGoogle ScholarPubMed
50. Sayer, AA, Syddall, H, Martin, H et al. (2008) The developmental origins of sarcopenia. J Nutr Health Aging 12, 427432.CrossRefGoogle ScholarPubMed
51. McCarthy, HD, Samani-Radia, D, Jebb, SA et al. (2013) New skeletal muscle mass percentile curves for children and youths aged 5–18 years. Pediatr Obes; Epublication 18 June 2013.Google Scholar
52. Webber, CE & Barr, RD (2012) Age- and gender-dependent values of skeletal muscle mass in healthy children and adolescents. J Cachexia Sarcopenia Muscle 3, 2529.CrossRefGoogle ScholarPubMed
53. Heitmann, BL & Frederiksen, P (2009) Thigh circumference and risk of heart disease and premature death: prospective cohort study. Br Med J 339, b3292.CrossRefGoogle ScholarPubMed
54. McCarthy, HD (2011) Thigh circumference percentile curves for the UK child and youth population. Proc Nutr Soc 70, E80.CrossRefGoogle Scholar
55. McCarthy, HD, Rigby, RA & Stasinopoulos, DM (2010) Hip circumference percentile curves for the British child and youth population. Proc Nutr Soc 69, E424.CrossRefGoogle Scholar
56. Frayn, KN (2010) Adipose tissue and the insulin resistance syndrome. Proc Nutr Soc 60, 375380.CrossRefGoogle Scholar
57. Manolopoulos, KN, Karpe, F & Frayn, KN (2010) Gluteofemoral body fat as a determinant of metabolic health. Int J Obes Relat Metab Dis 34, 949959.CrossRefGoogle ScholarPubMed
58. Kim, TN, Park, MS, Lim, KI et al. (2011) Skeletal muscle mass to visceral fat area ratio is associated with metabolic syndrome and arterial stiffness: the Korean Sarcopenic Obesity Study (KSOS). Diabetes Res Clin Pract 93, 285291.CrossRefGoogle Scholar
59. Browning, LM, Hsieh, SD & Ashwell, M (2010) A systematic review of waist-to-height ratio as a screening tool for the prediction of cardiovascular disease and diabetes: 0·5 could be a suitable global boundary value. Nutr Res Rev 23, 247269.CrossRefGoogle ScholarPubMed