Childhood Obesity: A New Pandemic of the New Millennium

The major epidemiologic transition in the 20th century was the shift in mortality and morbidity from infectious diseases to chronic diseases, with cardiovascular diseases leading the list. This transition was primarily attributable to the social, economic, and public health changes that took place in the United States during the first half of the century. The availability of abundant food led not only to better overall nutriture and improved child health, but also to the current population’s state of excess positive energy balance, accelerated by an increasingly sedentary lifestyle in recent decades. At the beginning of this new millennium, a new challenge has emerged—a marked increase in obesity prevalence with a parallel increase in obesity-associated chronic diseases and their clinical onset at ever younger ages. Type 2 diabetes, a condition traditionally associated with middle-aged adults, is beginning to occur several decades earlier as obesity afflicts an ever greater number of children and adolescents in the United States.1–5A recent publication in Pediatrics presented longitudinal findings from a large multicenter cohort study—the National Heart, Lung, and Blood Institute Growth and Health Study (NGHS)—of 1213 black and 1166 white girls followed for 9 years from ages 9 to 10.6 As can be seen in the accompanying study in this issue by Kimm et al,7 even at age 9, the prevalence of overweight (≥85th percentile of body mass index BMI based on the National Health Examination Survey reference population) was about one third higher in black girls than in white girls—31% versus 22%. The prevalence of obesity (≥95th percentile) in black girls was twice as high than in white girls—18% versus 8%. A most alarming finding in this study was the approximate doubling of the rate of overweight and obesity in both groups during the 10 years between ages 9–10 years and 18–19 years. Therefore, by 19 years of age, overweight was present in more than half of black and a third of white girls, and more than a third of black girls were obese as compared with almost one fifth of white girls. Thus, these data suggest that the trend in obesity in children is continuing upward and may have even accelerated more than previously reported.8–10A “pandemic” is defined by Merriam-Webster’s dictionary as “occurring over a wide geographic area and affecting an exceptionally high proportion of the population.”11 Given the current prevalence of childhood obesity and its geographic distribution throughout the United States, the term “pandemic” is appropriate for describing the current status of childhood obesity.Although much speculation abounds, the root cause of this increase in childhood obesity remains unknown. What is evident from available national data are that this trend in childhood obesity has not been accompanied by a parallel population increase in daily energy intake.12,13 Although the information on trends in physical activity is exceedingly limited, in recent years much concern has been expressed regarding the role of television watching.14–17 The NGHS examined a biracial cohort for biological, behavioral (diet and physical activity), psychosocial, and sociologic correlates of obesity development during adolescence. Because of the study’s large sample size, disparate geographic sites, and the long duration during a critical life period, the information available from NGHS might provide some insight into this complex topic.6,18The longitudinal data from NGHS showed that pubertal maturation was associated with a significant gain in adiposity for both racial groups with the largest increases occurring at the initiation of pubertal maturation and at menarche. Starting at age 10, adiposity in black girls was significantly greater than in whites because of the more advanced pubertal maturation of black girls.6 When pubertal maturation was adjusted for in longitudinal regression analyses, the racial divergence in adiposity was manifested by age 12, around the time of menarche for the majority of black girls. Therefore, one can surmise that the nascence of the high rate of obesity in black women may be around the time of menarche.Another biological factor associated with the increased susceptibility to obesity in black women may be in part attributable to racial differences in energy balance. Several recent studies have shown that African American women display lower resting energy expenditure (REE) than do white women.19 A random sample of 150 older adolescent females from the NGHS cohort also revealed that black adolescent girls manifested REE levels lower on average by 77.5 kcal/d (adjusted for body composition) than did white girls.20 Data from 150 NGHS subjects suggest that an uncoupling protein gene may be associated with the racial difference in REE.21 Those black girls with UCP 3 (exon 5) C/C genotype showed the lowest levels of REE, and there was a significant trend toward lower REE from T/T, C/T, and C/C in black but not in white girls.In the NGHS, the role of energy intake is unclear. For instance, although caloric intake was not related to obesity when the cohort was ages 9 to 10 years, percent energy from fat (total fat in white girls, saturated fat in black girls) was associated with obesity.22 For ages 9 to 19 years, there was a significant inverse relationship between caloric intake and adiposity.6 This somewhat counterintuitive finding is most likely attributable to difficulties of accurately quantifying one’s intake, underreporting of food intake, or tendency for overweight girls to diet. The recently available doubly labeled water method—which accurately measures energy requirement and, therefore, energy intake—has shown that underreporting of energy intake can be substantial.23 Using doubly labeled water in a random sample of 150 NGHS girls aged 18 to 21, underreporting of energy intake was 11% for black girls and 23% for white girls.24 With this degree of misreporting, the quantitative relationship between caloric intake and obesity development in free-living subjects remains unresolved.As with the assessment of energy intake, the assessment of physical activity in free-living subjects poses challenges, because the most commonly used methods are based on self-reported measures whose validation studies have yielded correlation coefficients ranging from 0.17 to 0.51.25 Habitual activity had a weak inverse association with adiposity in 9- to 10-year-old NGHS black girls but not in white girls in a cross-sectional analysis.22 Preliminary data analysis revealed an inverse longitudinal association between habitual activity and obesity during adolescence in the NGHS girls.26 However, the longitudinal association between daily activity level and adiposity was not significant.6What has emerged from NGHS data are evidence of a dramatic decline in physical activity between ages 9 and 19 years. The levels of habitual activity declined by 83%, and daily activity levels also declined during this period by 35%.27 These longitudinal NGHS data on trends in physical activity corroborate what is known from national cross-sectional surveys.28Analyses of psychosocial factors in NGHS show intriguing racial differences in measures of self-perception, particularly in the domains of physical appearance and social acceptance.29 Although satisfaction with physical appearance was lower with increasing adiposity for both groups, white girls manifested greater dose-response to adiposity than did black girls. The racial difference was particularly striking for perceived social acceptance, in which black girls exhibited no variation across the entire spectrum of adiposity, whereas an inverse association between perceived social acceptance and adiposity was seen in white girls. This finding on perceived social acceptance suggests greater tolerance for obesity among African Americans.There were other interesting racial differences found in NGHS. For instance, although almost half of NGHS girls, even at ages 9 and 10, were trying to lose weight, more of the thin black girls than white girls reported trying to gain weight, suggesting a lower tolerance by black girls for being too thin.30 White girls, on the other hand, were generally more dissatisfied with their weight and body shape than black girls. In addition, for white girls, perceived physical attractiveness began to decrease even at extreme thinness and declined linearly across the entire spectrum of adiposity, suggesting the presence of an underlying drive for thinness.29 Nine- and 10-year-old black girls were more than twice as likely as white girls to engage in less favorable eating practices, such as eating big helpings and eating in the bedroom, generally believed to be associated with overeating.31Taken together, a pattern emerges that white girls are generally more concerned than black girls about physical appearance even at the lower range of adiposity. In contrast to white girls, who are sensitive about social acceptance in relationship to adiposity, black girls displayed no concern in this domain across adiposity. On the other hand, black girls tend to be more susceptible to eating behaviors that may lead to overweight and do not feel socially ostracized when obese. These findings may provide additional insight into cultural issues involving obesity.Poverty has often been considered a major contributor to the high prevalence of obesity seen in African American women. However, for 9- and 10- year-old NGHS black girls, there was no variation in the risk for obesity across household income or parental education, whereas there was an inverse association with both in white girls.32 When income, education, and the number of parents in the household were included in multivariate analyses, greater television viewing was associated with higher likelihood of being obese in black girls but not in white girls. For 9- and 10-year-old white girls, only higher parental education and having 2 parents in the household decreased the likelihood of obesity. In both races, higher levels of parental education tended to be consistently related to more favorable nutrient intakes, including lower fat intake and higher intakes of essential vitamins and minerals, and to lower likelihood of engaging in eating practices associated with overeating.31,33 The decline in physical activity during adolescence in the NGHS cohort was also inversely associated with parental education.34 In summary, based on the findings from the NGHS, parental education, not household income, tended to exert a greater effect on the salutary behavioral factors, and the effect of parental education was greater in white girls than in black girls.Because the large rise in obesity in the pediatric population has occurred primarily within the last 2 decades, several issues and questions regarding childhood obesity remain and need to be addressed.The fundamental definition of obesity in children needs to be reexamined. The traditional concept of ideal body weight in adults was based on actuarial data collected over time on a large number of insured adults, which provided the quantitative information on standard mortality ratios in relationship to relative weight (weight for height).35 Based on mortality and morbidity data compiled from other large cohort studies, the current standard definition of overweight and obesity uses specific BMI cut points of 25 and 30, respectively.36 For children, however, mortality or even disease as an adverse outcome of obesity cannot be readily assessed. The currently used cut points to categorize overweight or obesity in children, such as age- and gender-specific ≥85th and/or ≥95th percentile BMI or weight-for-height, are arbitrary statistical measures serving as reference values. They are not based on biological data linked to some estimate of adverse outcome.37–39 Perhaps, a biologically relevant outcome could be based on risk factors for disease such as elevated low-density lipoprotein cholesterol, elevated fasting blood glucose, and elevated blood pressure.40 This approach to a definition of childhood obesity would require a systematic review of available information from population studies to examine the relationship between BMI and disease risk factors in children. Some believe that children may be more vulnerable to adverse psychological, rather than biological, effects of obesity, so psychological risk factors might also be included.Unresolved issues for clinical management of obesity include the extent of diagnostic evaluation that needs to be conducted at the physician’s office when an obese child is seen. In the past, thyroid function was frequently assessed to look for underlying endocrine causes of obesity, although the yield was exceedingly low. Another unresolved issue is the extent of screening for potential morbid conditions associated with obesity such as hyperlipidemia, elevated blood pressure, and type 2 diabetes.The use of pharmacologic and surgical treatment of childhood obesity is beginning to emerge.41,42 At what degree of obesity and at what age should one entertain the use of a pharmacologic agent such as metformin hydrochloride or even gastric bypass surgery? The age range of the recent report on gastric bypass surgery was 15 to 17 years, a time when peak bone mass is being achieved.42 What would be the long-term consequence on bone health from gastric bypass surgery, which might impact calcium nutriture? What would be the criteria for undertaking such an invasive procedure in the young? Pediatric surgical services are beginning to ready themselves for providing bariatric surgery. Are pediatricians and/or family medicine physicians themselves prepared to deal with this clinical problem? Two NHLBI national surveys of primary care physicians, conducted 10 years apart, indicated that clinical management of childhood obesity was the most wanted topic for continuing medical education.43,44 Is it timely to establish bariatric pediatrics as a legitimate subspecialty? Many of these questions are unanswered because there has not yet been a multicenter, randomized, clinical trial to test the efficacy and safety of weight reduction regimens, both pharmacologic and nonpharmacologic, in a broadly representative sample of children. The randomized controlled trial of weight reduction in 6- to 12-year-old children conducted by Epstein et al45 was a single-center study of a highly selected middle-class white population. Information is urgently needed in the area of nonpharmacologic and nonsurgical approaches to the clinical management of childhood obesity. The oft-encountered reaction to childhood obesity among primary care physicians is that “nothing works.” Before pharmacologic and surgical procedures become “standard” treatments for childhood obesity, more research is urgently needed to test the efficacy of various dietary interventions for weight reduction.A question still besetting primary care physicians caring for children is the age at which one should undertake active weight reduction and the amount of caloric reduction that can be safely instituted in young growing children. For instance, for an obese 1- year-old child, should one institute a hypocaloric dietary regimen or wait until that child is older but also likely to be fatter, or should one focus more on physical activity than diet? What are appropriate parameters for safety monitoring of children on a reduced-calorie diet? Is height growth an appropriate measure of safety surveillance? Obesity is often accompanied by accelerated linear growth in prepubertal children, but height growth tends to decelerate with weight reduction in an obese child.46 If height-for-age changes from the 95th percentile to the 75th percentile after significant weight loss, should this change be regarded as undesirable and a reason to discontinue the weight reduction regimen? Or is this a biological “normalization” process for a child with an abnormally accelerated growth rate associated with obesity?Another unresolved issue is the type of health care providers who need to be involved in the clinical management of childhood obesity. For instance, dietary and physical activity counseling and behavioral therapy are considered necessary components of obesity treatment. Both treatment modalities are time-consuming. In this cost-conscious era, do we have the luxury or even an opportunity to involve a dietitian and/or behavioral therapist for every obese child seen in the primary care physician’s office, particularly given the high prevalence of this condition? What would be the necessary competency levels of these health care professionals? There is wide variation in the skills, experience, and knowledge of these professionals. A behavioral therapist well versed in smoking cessation techniques may not be as effective for weight reduction. Or a dietitian experienced with cancer patients may be less equipped to deal with childhood obesity. As we require competency testing for pediatricians, should there be competency requirements for these allied health professionals dealing with childhood obesity?The continuing increase in childhood obesity dictates that these unresolved issues be addressed. Standardized guidelines for meaningful diagnosis, clinical evaluation, and treatment of childhood obesity need to be established. To be effective, these guidelines need to be recognized and supported by third-party payers and the leaders of health maintenance organizations.The National Health and Nutrition Examination Survey (NHANES) is a periodic national survey that includes measurements of height and weight in children and adolescents. Despite the highly complex sampling strategy used to achieve a representative sample of the US population, the actual sample sizes become quite small when NHANES data are stratified by race, gender, and single-year age groups.47 Consideration should be given to the sampling strategy of these national surveys to better accommodate age- and gender-specific high-risk population groups. Although NHANES surveys serve many purposes, in light of the looming public health crisis of obesity, the scope of these surveys may need to be expanded to encompass public health issues related to obesity. For instance, as type 2 diabetes is seen in increasing numbers in children and adolescents, diabetes might be tracked as part of the surveillance of the rising tide of morbid childhood obesity. Perhaps, in the not-too-distant future, urinary glucose testing in primary schools may become widespread.Because of its widespread nature, population approaches should be considered for primary prevention of childhood obesity. The initial basis for the school lunch program was out of concern for hunger in America in the early 1960s. Given the current problem of childhood obesity, the goals and rules of the school breakfast and lunch program need to be reexamined and readdressed. For example, should chocolate milk and whole milk be removed from school menus as a national policy? As contemporary US children become more physically inactive, the goal of physical education in schools also needs to be reassessed. The architectural design of school buildings and their environment could be reexamined to provide more opportunities for physical activities. A multistory building with a carefully constructed class schedule could lead to stairclimbing during the course of a school day.The widespread nature and the continuing rise of childhood obesity border on a national crisis, because it will surely exact an enormous societal toll on health in the coming decades. Complacency needs to be overcome to begin to systematically identify and address many fundamental questions pertaining to childhood obesity. A bold vision is needed to develop novel approaches at both the population and the individual level.