Definition and epidemiology
Obesity results from an interaction of genetic, environmental, developmental, and behavioral processes, and reflects a broad continuum from normal variation to a pathologic condition. This is clearly demonstrated by the limited usefulness of available definitions. During childhood and adolescence, obesity is most commonly defined as weight greater than 120% of the median weight for height, or a triceps skinfold thickness or body mass index (ie, weight/height²) greater than the 85th or 95th percentile for children of the same age and sex. However, these standards are limited by the representativeness of the samples from which they are derived. Until recently, available reference samples generally have not reflected sufficient racial, cultural, or socioeconomic diversity. The newly revised growth standards from the National Center for Health Statistics now include nationally representative samples of whites, African-Americans, and Mexican Americans, but not large numbers of children or adolescents from other racial and/or ethnic groups. Nevertheless, the greatest limitation of threshold definitions of obesity is lack of evidence for their clinical validity. Although recent data suggest that more than half of children who meet traditional definitions of obesity show evidence of associated physiological morbidities, few data confirm that children and adolescents who are “above” one or more of the previously mentioned definitions, are much worse off from a clinical standpoint than those who fall just below the cutoff points. Instead, body fatness is related to morbidity in a continuous, up-sloping, curvilinear manner, without the thresholds suggested by the common cutoff definitions.

For epidemiologic purposes, conventional definitions may provide useful information on trends in the population and differences among various groups. Data from the most recent national surveys demonstrated that the prevalence of obesity among children and adolescents has more than doubled from the 1970s to the early 1990s. The increases appear to have occurred across all ethnic groups, although the highest rates are among Mexican American boys and girls and African-American adolescent girls. More detailed analysis of the trends in being overweight has demonstrated the worrisome finding that most of the increase has occurred in the upper extremes of the distribution, resulting in larger numbers of extremely overweight children and adolescents, those who are most likely to suffer from obesity-associated morbidities.

By far, the strongest risk factor for obesity in children and adolescents is having an obese parent. Whether it is the mother or the father makes little difference; as expected, having two obese parents is a greater risk factor than having only one obese parent. Parent weight status also strongly influences the likelihood of a child becoming an obese adult. In one retrospective cohort study, for example, 3- to 5-year-olds were 3 times more likely to be overweight in their twenties if they had one overweight parent and about 15 times more likely if both parents were overweight. As children age, the child’s weight becomes a better predictor of adult obesity than parent weight. In the same study, overweight 10- to 14-year-olds were about 20 times more likely to be overweight in their twenties than their normal-weight peers, while having at least one overweight parent at age 10 to 14 years doubled the risk of obesity in young adulthood.