Age- and maturity-associated variation in body fat distribution
Chronological age (CA) is an indicator of sidereal time whereas the body has its own clock that varies among and within individuals. Children increase in size and change in proportions and body composition with the passage of time. Maturation refers to progress towards the mature state, which varies among bodily systems. It is often viewed in terms of status (maturity status at a given CA), and timing (CA at which specific maturity events occur) .
Standard indicators of maturity status include skeletal age, which can be used from childhood through adolescence, and pubertal status, which is limited to the pubertal years. Both are considered invasive. Many current studies use self-assessments of pubertal status. Indicators of maturational timing include ages at peak height velocity (PHV) and at menarche. Both require longitudinal data for estimation in individuals. Recalled ages at menarche have major limitations with youth . Two non-invasive measures are increasingly being used. Percentage of predicted adult (mature) height attained at the time of study is an indicator of status . Predicted time before PHV (maturity offset) and in turn predicted age at PHV is an indicator of timing . Both non-invasive methods require cross-validation .
Fat distribution refers to regional variation in the accumulation of adipose tissue in the body. Initial focus was on global male (android) and female (gynoid) patterns of distribution and then contrasts of subcutaneous tissue on the trunk vs. the extremities. A common descriptor was the term pattern, android vs gynoid fat pattern or a truncal vs. extremity pattern. With advances in technology (CT, MRI), attention shifted to abdominal adiposity, specifically visceral vs subcutaneous. With widespread availability of dual-energy X-ray absorptiometry (DXA), trunk and extremity distribution of adipose tissue has received more attention [56, 57].
Available data on fat distribution during childhood and adolescence follow, in part, the development of technology and the increased emphasis on the metabolic complications of obesity. Subcutaneous fat distribution, as estimated with ratios of skin-folds measured on the trunk to those measured on the extremities, changes with age and differences between the sexes emerge during the growth spurt and sexual maturation [56, 58, 59]. Ratios of trunk to extremity adiposity measured via DXA show a similar trend [60, 61]. The ratio of abdominal VAT and SAT appears to show a similar trend, although data for normal-weight children and adolescents are less extensive than for overweight and obese youth. With increased emphasis on abdominal obesity and associated metabolic complications, waist circumference (WC) is increasingly used as an indicator of abdominal adiposity. It is often expressed relative to height (WC/Ht ratio). Allowing for discussion of the most appropriate level for measurement of WC in youth, data addressing age-, sex- and maturity-associated variation in the WC/Ht ratio and other anthropometric measurements are not yet extensive.
Fatness per se and relative fat distribution (ratio of trunk to extremity skin-folds) in late childhood and adolescence differ among individuals of contrasting maturity status. Children and adolescents advanced in maturity status compared to CA peers tend to be fatter and to have proportionally more subcutaneous fat on the trunk compared to the extremities [52, 62]. Some data indicate persistence of the trend into young adulthood and adulthood [63-66]. Longitudinal studies of fat distribution are limited. Moreover, changes in individual skin-folds are variable during the growth spurt and specifically relative to the timing of PHV, more so in boys than in girls . Such variation may influence age-associated trends.
Pubertal status, clinically and/or self-assessed, is the most commonly used maturity indicator in studies dealing with maturity effects on fat distribution. Youth are typically grouped by stage of puberty and more often combined pubertal stages, so that several CAs are represented within a stage; the effect of CA per se is generally controlled statistically. Allowing for this limitation, major changes in fat distribution based on DXA occur during puberty, specifically late puberty [60, 61]. There is a need for studies that include sufficient numbers of children of the same CA who vary in stage of puberty.
Studies of abdominal visceral and subcutaneous adipose and fat distribution based on CT and MRI tend to have relatively small sample sizes compared to those using anthropometry. Samples are generally combined across age groups and in some instances combine males and females and/or youth of different ethnic groups. Stage of puberty is often described but not systematically analyzed and other maturity indicators are not used. Moreover, recent data with CT and MRI are seemingly more focused on obese youth. Ratios of cross-sectional areas of abdominal VAT to SAT were derived from 27 studies of normal-weight and overweight/obese youth (15 based on CT, 12 on MRI; 13 at umbilical level, 13 at L4-L5, one at minimum waist) [67-93] (Fig. 2). Among normal-weight youth, there is no clear trend with age and sex in relative fat distribution during childhood into early adolescence. Subsequently, males have, on average, proportionally more visceral adiposity in later adolescence. There does not appear to be a clear trend in overweight and obese youth, indicating that excess adiposity may attenuate age and maturity differences.
Figure 2. Abdominal visceral to subcutaneous fat ratios based on cross-sectional scans in samples of girls and boys from 27 published studies described in the text and reviewed by R.M. Malina at the 2011 Pennington Biomedical Research Center symposium on adiposity in children and adolescents. Filled squares = males; filled circles = females; open squares = sexes combined.
Download figure to PowerPoint
Advances in imaging technologies have permitted the quantification of ectopic fat deposition in the liver (non-alcoholic fatty liver disease, NAFLD), heart and other organs, and in skeletal muscle (intra-myocellular fat). Such data for children and adolescents are quite rare thus far.
Sexual dimorphism in body composition during childhood
Sexual dimorphism in body composition is apparent in both total adiposity and its regional distribution, as well as lean mass and physique. Several studies have now demonstrated that body composition dimorphism is apparent at birth, with the lower average birth weight of females attributable to lower lean mass [94, 95]. Females also appear to have greater adiposity and a more central fat distribution at birth . During development, sexual dimorphism in total body size, lean mass and adiposity remains modest during childhood but increases substantially from puberty onwards. Females enter puberty earlier than males, and achieve lower final stature and lean mass; however, they gain substantially greater peripheral adiposity over the same period . Males enter puberty later and achieve greater final stature and lean mass, with less total but similar or higher central adiposity . These contrasting pubertal growth patterns lead to the well-established sexual dimorphism characteristic of all documented adult populations, with males having greater stature and upper body lean mass, and females having substantially greater total and peripheral but not central fat [98, 99].
Following similar method-specific work [100-106], new cross-sectional reference data for total and regional adiposity, obtained using the four component model, DXA, air displacement plethysmography, bioelectrical impedance analysis and anthropometry (body circumferences and skin-folds) have recently been obtained in a sample of 532 healthy individuals from the UK aged 4 to 23 years . These data are the basis for children's body composition growth charts, and enable age- and sex-specific standard deviation scores of total and regional adiposity to be calculated. The data also characterize in greater detail the age profile of adipose tissue deposition in each sex. Sex differences in arm and torso skin-fold thickness are not significant in this sample, indicating that the lower body is the primary site of adiposity dimorphism. This is supported by data on leg fat obtained by DXA.
Ethnic differences have been reported in the magnitude of sexual dimorphism for whole-body adiposity in adolescence , and for body shape in adults . A recent literature review of data from young adulthood indicates that sexual dimorphism in both lean mass and adiposity varies in relation to mid-sex adiposity (male–female average), suggesting that population energy availability influences the relative allocation of energy to lean tissue vs adiposity in males vs females, and that body composition dimorphism has a plastic component .
Longitudinal studies likewise suggest that variable rates of tissue deposition in adolescence can be predicted by growth rates during infancy , suggesting that the magnitude of adiposity dimorphism is sensitive to experience during early life. Birth weight appears most consistently associated with subsequent lean mass rather than FM; however, the ratio of fat to lean mass in childhood was found to increase more strongly in relation to ponderal index (weight/length3) at birth in females compared to males . Similar findings are emerging from studies of infant growth but few data are currently available.
Ethnic differences in body fat distribution
Despite the well-known links between obesity and metabolic diseases, and the increased rate of obesity and metabolic diseases among ethnic minority groups, the mechanisms of these observations remain elusive. Several putative explanations exist for why fat affects metabolic health and how this might vary across different ethnic groups. One such theory is based on the anatomic location of fat deposition and ectopic fat accumulation in critical organs like muscle and liver. Specifically, current literature suggests that visceral, liver and skeletal fat accumulation affect organ function and contributes to the development of insulin resistance, fatty liver and the metabolic syndrome. However, even in individuals matched for body fat and fat distribution, significant differences can exist in metabolic outcomes. In addition, ethnic differences in fat distribution and ectopic fat deposition do not fully explain ethnic disparities in metabolic diseases.
VAT has long been hypothesized to be one of the major factors linking obesity and disease risk. However, this hypothesis leads to an ethnic paradox  because African–Americans, who are at increased risk for obesity-related diseases, especially CVD, have lower VAT on average beginning early in life [80, 81]. More recent studies have also shown that ectopic fat deposition varies by ethnicity in the same way as VAT. Studies consistently show that Hispanics have a much higher prevalence of fatty liver disease than African–Americans, also beginning early in life [113, 114]. Part of this ethnic difference is driven by a genetic contribution from PNPLA3 , a single nucleotide polymorphism that is more prevalent in Hispanics than African–Americans , and influences greater liver fat in Hispanics beginning early in life , in part driven by higher sugar consumption . Pancreatic fat fraction is also higher in Hispanics than African–Americans, and the magnitude of this difference increases with age . In young obese individuals, pancreatic fat is related to visceral and liver fat but does not appear to be related to insulin resistance or beta-cell function . Thus, collectively, these studies show that African–Americans have lower visceral and ectopic fat deposition than Hispanics. This finding is in conflict with the hypothesis that greater visceral and ectopic fat drive increased risk of metabolic diseases since both these ethnic groups tend to have similarly high risk of metabolic diseases despite the clear difference in fat pattern.
There is also some evidence to suggest that ethnic differences in body fat pattern and accumulation may result from fundamental differences in adipose tissue biology and that adipose tissue biology itself drives metabolic disease risk. The increase in body fat content with obesity can occur by either an increase in adipocyte cell size or number, or by the spillover of triglycerides to ectopic tissues. When adipocyte cell size increases with progressing obesity, it is an indication of the inability of adipocytes to expand in number to accommodate the extra triglyceride accumulation. Larger adipocytes have also been shown to be associated with more lipid deposition in visceral and hepatic tissues (not muscle), and this may contribute to insulin resistance . Furthermore, it is now also evident that adipose tissue can become infiltrated with macrophages [120, 121] and this inflammatory profile drives metabolic risk. In a previous study among obese young minority adults (Hispanic and African–American), we found that approximately 40% of subjects had subcutaneous abdominal adipose tissue with crown-like structures, indicating inflammation, whereas approximately 60% of subjects had no signs of adipose tissue inflammation . Despite the two groups being identical for overall obesity and subcutaneous abdominal adipose tissue volume, those with inflamed adipose tissue had approximately 30% greater VAT and 41% greater liver fat; 53% greater fasting insulin and 23% lower beta-cell function; and 22% higher tumour necrosis factor-alpha (TNF-α). Given these observations, the disparities in metabolic diseases among obese minority individuals may be explained by the degree of chronic low-grade inflammation of adipose tissue. Therefore, targeting adipose tissue inflammation has become an important new strategy in treating the metabolic conditions typically associated with obesity.