Overweight is present if body weight adjusted for height exceeds a specified cutoff; obesity implies an excessive fraction of total body weight comprised of fat mass. In theory at least, obesity can only be diagnosed if fat and fat-free mass are determined. Various models and indirect methods for estimation of body composition have been developed, all of which are imperfect and require a number of assumptions, including age-specific considerations. For research purposes, body composition in children has been determined using techniques such as total body water, dual energy X-ray absorptiometry (DEXA), total body electrical conductivity, and total body potassium. By contrast, in clinical studies bioelectrical resistance and skinfold measurements have been employed instead (1).

In clinical practice, however, the BMI, defined as weight in kilograms divided by height in meters squared (kg/m²), has become widely accepted as the optimal weight-height index throughout childhood and adolescence and as a surrogate measure of adiposity (2). Indeed, the correlations between BMI and percent body fat as determined by DEXA are high (3,4). Nevertheless, a high fat free mass can, as in the case of some athletic individuals, also entail a high BMI.

The developmental pattern of BMI is similar to that of fat mass. Typically BMI increases steeply during the first 6 months of life primarily due to an increase of fat mass, to then decline until age 6 to 8 years as a result of a strong relative increase in height (Figure 5.7.3.1 a,b). The time point during childhood at which BMI again increases has been termed adiposity rebound; the earlier it occurs, the higher the risk of subsequent development of obesity and vice versa (5). BMI rises throughout the


rest of childhood and adolescence; this rise continues— albeit more slowly— during adulthood. In males the rise during adolescence is on average largely due to an increase in fat-free mass; percent body fat remains stable at approximately 15 to 18%, whereas in females percent body fat increases from a prepubertal value similar to that of boys to a value of 20 to 25% after puberty (6).

Birth weight is minimally correlated with BMI at age 35 (r = 0.1). Accordingly, an overweight infant is only at a minimally elevated risk for obesity in adulthood provided that parental obesity is not present (Figure 5.7.3.2). The greatest variation in rates of weight gain is seen in the first 1 to 2 years of life, when infants may show significant “catch-up” or “catch-down” growth as a consequence of intrauterine restraint or enhancement of fetal growth. By age two, growth usually follows the genetic trajectory (7). Small-for-date babies who rapidly gain weight after birth (catch-up growth) are at an elevated risk for adult obesity and cardiovascular disorders (8). At age 10 BMI is correlated with BMI at age 35 in the magnitude of 0.35; during late adolescence BMI tracks well into adulthood (r = 0.7) (9). If a child is obese the risk of persistence into adulthood increases with the degree of adiposity (10), age of the child, and parental obesity (Figure 5.7.3.2).

The strong age dependency of BMI implies that an adequate interpretation of the index of an individual child requires knowledge of the BMI distribution of same aged and sexed children. Frequently, centile curves are based on BMI values of a large and ideally representative sample; older children and parents are readily able to grasp this epidemiological concept as having good face value for educatory purposes. BMI standard deviation scores (BMI-SDS) or Z-scores are commonly employed to further differentiate extreme overweight or underweight (> 99th and < 1st centile).

In adults, overweight, obesity and extreme obesity are diagnosed if BMI is ≥ 25, ≥ 30 and ≥ 40 kg/m², respectively; (11) these cutoffs were chosen because they mark increments in risks for complications of elevated BMI. For children and adolescents, however, there is no clear-cut consensus as to what cutoff should be used to define overweight and obesity; all existing definitions are statistically rather than risk-based. Himes and Dietz (12) have suggested use of the 85th and 95th centiles; in other countries the 90th and 97th centiles serve as the respective cutoffs. The 85th and 95th centiles, based on nationally representative data from the 2000 growth curves of the Centers for Disease Control and Prevention (CDC) (13) have been recommended for classification of persons as being overweight or at risk of overweight in the United States (Figure 5.7.3.1 a,b; CDC; http://www.cdc.gov/growthcharts). However, these cutoffs are not consistent with those used in adulthood; for example, the centile of an 18-year-old female with a BMI of 30.1 kg/m²— thus fulfilling the adult criterion for obesity— is below the 95th centile.

Use of specific centiles as cutoffs for the definition of obesity additionally requires a consensus as to the reference population. International comparison of the absolute BMI values constituting centiles in the lower to middle range typically reveals only a slight degree of divergence among different Western countries. However, large differences exist in the upper centile range; BMI values corresponding to the 90th or 97th centile are for example substantially higher in the United States of America in comparison to most European countries. Because of secular trends for both height and weight, centiles based on a formerly representative population-based survey usually do not reflect the true BMI distribution of the current population. For instance, the measurement data used to construct the nationally representative CDC Growth Charts were obtained from a series of national health examination surveys conducted by the National Center for Health Statistics from 1963 to 1994 and from supplemental data sources; more recent data were excluded to avoid an upward shift in the BMI-for-age curves (http://www.cdc.gov.growthcharts; see link interactive training modules).

In an attempt to define internationally acceptable centile cutoffs for overweight and obesity, the International Obesity Task Force (IOTF) averaged the national curves from six different countries passing through BMIs of 25 and 30 kg/m² at age 18 years (14). It remains to be seen if the cutoffs are adopted sufficiently to indeed reach an international consensus of the definition of childhood obesity.

Physical growth and development are sensitive indicators of the quality of the psychosocial, economic, and political environment. Child growth in terms of height, weight and body composition are widely used indicators of nutritional and health status for both the individual child and the community (15). Three major secular trends have consistently emerged at the population level in industrialized countries: Mean height and BMI have increased, and mean age at puberty has decreased. Increments in height and decrements in age at menarche can be dated back several decades; the increment in BMI roughly began three decades ago. Both body height and age at menarche have recently reached plateaus in several industrialized countries; within a society the achievement of these plateaus typically occurred first among individuals with a higher socioeconomic status (16). These plateau effects indicate that the full genetic potential has been realized under an advantageous environment, or that social conditions have ceased to improve.

Obesity has only recently reached epidemic proportions in several countries worldwide. Currently, approximately 66%, 31%, and 5% of adults over 20 in the United States are overweight, obese, and extremely obese, respectively; (17) it has been estimated that if the recent secular trend continues unabated, obesity prevalence will reach 50% by 2025. Similar trends have been observed in other countries in Asia, Europe, North and South America (18). However, it should be pointed out that slight increments in mean BMI of a population can
translate into substantial elevations of prevalence rates for overweight and obesity.

Children and adolescents have also been affected strongly by this secular trend. The National Health Examination Surveys have revealed a four-fold increase in the prevalence of overweight among children aged 6 to 11 and a three-fold increase for adolescents aged 12 to 19 between the 1960s and the most recent survey, conducted between 1999 and 2002 (17). In the 1999–2002 survey 16% of the children aged 6 through 19 years were overweight as defined via a BMI ≤ 95th centile of the surveys in the 1960s. An additional 16% of children were classified as at risk of overweight. In the United States of America, African Americans and Mexican Americans are bearing the brunt of the obesity epidemic (19). Worldwide mean BMI and, as a consequence, prevalence rates for overweight and obesity in children and adolescents have been on the rise in both industrialized and developing countries (20). The most pronounced secular increments in absolute BMI have been detected in the overweight and obese range (21).

A detailed review of the medical consequences of childhood and adolescent obesity is beyond the scope of this chapter (22,23,24). In brief, cardiovascular disorders including hypertension, type 2 diabetes mellitus, dyslipidemia and metabolic syndrome are becoming increasingly prevalent during adolescence as a consequence of the obesity epidemic [e.g., 25]. Childhood and adolescent obesity has been associated with increased mortality even after adjustment for adult BMI (24). Child and adolescent psychiatrists need to be aware of medical complications associated with obesity, including nocturnal enuresis and sleep apnea, both of which can occur in up to 16% of obese children and adolescents (23). Although far less common, pseudotumor cerebri can occur in young obese subjects.

Obesity is a truly complex disorder that is caused by several genetic and nongenetic risk factors. In historic terms, the etiology of childhood obesity has been perceived differently over the past 100 years; the pendulum has swung from biological to psychological explanations and back. In the early part of the 20th century, pituitary/hypothalamic dysfunction was assumed to underlie obesity (26). From the 1940s through to the 1970s psychological and psychodynamic aspects were thought to play the most prominent role (27,28). Prader, Labhart and Willi delineated the first syndromal form of obesity in 1956. As of the 1970s the relevance of (psycho)social factors has been discussed (29,30,31,32). At the end of the 1980s and 1990s milestone twin and adoption studies substantiated that genetic factors play a prominent role in body weight regulation (33,34,35). The cloning of the leptin gene in 1994 (36) led to a virtual explosion of biomedical research and marked the introduction of large-scaled molecular genetic studies. The detection of children with leptin deficiency (37) and their successful treatment with recombinant leptin (38) for the first time proved that mutations in a single gene can lead to hyperphagia and obesity in individuals of normal intelligence. The most recent years have viewed a strengthening of the hypothesis that obesity is a neuroendocrine disorder, which results if “obesogenic” environmental factors and a polygenic predisposition act in concert. The genetic predisposition is thought to affect both metabolic and behavioral features. Both public and scientific interest in obesity has witnessed an unparalleled boom over the past 20 years, reflecting the obesity epidemic, the discovery of novel pathways involved in body weight regulation, greater health concerns, the perception of social inequality, and a stronger awareness of the societal norms for body weight.