C Pediatric Obesity

Beatrice S. Kanders, Ed.D., M.P.H., R.D.

Pediatric obesity is a significant public health issue and a prevalent nutritional problem in the United States. The prevalence of obesity among children and adolescents has increased markedly since 1964, possibly affecting 20–27 percent of them (Gortmaker et al., 1987). Recent data from the Third National Health and Nutrition Examination Survey (NHANES III) indicate that these trends have continued for adolescents despite national efforts to reduce overweight (Harlan, 1993). Although it can be explained in part by environmental and economic factors, the cause of the rising incidence of pediatric obesity over the past three decades is not readily apparent. The current literature on childhood and adolescent obesity is reviewed here, focusing on its causes, consequences, treatment, and prevention. As there is still much to be learned about pediatric obesity, research needs are also identified.

OVERVIEW

Obesity is characterized by an excess of adipose tissue in relation to lean body mass. However, a standard clinical definition of childhood (6–10

COMMITTEE'S NOTE: Time constraints prevented us from being able to address the important subject of pediatric obesity. However, given its importance and the fact that obesity among children and adolescents is increasing as it is among adults, we asked obesity specialist Beatrice Kanders to prepare the following background paper on the subject. We hope it helps stimulate further study and research towards preventing and treating obesity during the formative years of childhood and adolescence.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs C Pediatric Obesity Beatrice S. Kanders, Ed.D., M.P.H., R.D. Pediatric obesity is a significant public health issue and a prevalent nutritional problem in the United States. The prevalence of obesity among children and adolescents has increased markedly since 1964, possibly affecting 20–27 percent of them (Gortmaker et al., 1987). Recent data from the Third National Health and Nutrition Examination Survey (NHANES III) indicate that these trends have continued for adolescents despite national efforts to reduce overweight (Harlan, 1993). Although it can be explained in part by environmental and economic factors, the cause of the rising incidence of pediatric obesity over the past three decades is not readily apparent. The current literature on childhood and adolescent obesity is reviewed here, focusing on its causes, consequences, treatment, and prevention. As there is still much to be learned about pediatric obesity, research needs are also identified. OVERVIEW Obesity is characterized by an excess of adipose tissue in relation to lean body mass. However, a standard clinical definition of childhood (6–10 COMMITTEE'S NOTE: Time constraints prevented us from being able to address the important subject of pediatric obesity. However, given its importance and the fact that obesity among children and adolescents is increasing as it is among adults, we asked obesity specialist Beatrice Kanders to prepare the following background paper on the subject. We hope it helps stimulate further study and research towards preventing and treating obesity during the formative years of childhood and adolescence.

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs years) and adolescent (11–21 years) obesity does not exist. Generally, obesity is defined as weight-for-height above the 90th percentile on the growth charts from the National Center for Health Statistics (NCHS) or weight in excess of 120 percent of the median weight for a given height (Leung and Robson, 1990). The 85th percentile of triceps skinfolds is also used (Garn and Clark, 1976; Gortmaker et al., 1987), and skinfolds correlate well with percentage of body fat in children of both sexes (Forbes and Amirhakimi, 1970; Roche et al., 1981). It may be difficult to measure skinfolds, as evidenced by the mere 8 percent of pediatricians and family practitioners who take these measures (Greecher, 1993). The Quetelet index or body mass index (BMI) is an indirect measure of body fat, but it is easily and reliably measured, correlating well with more precise estimates of subcutaneous and total body fat (Deurenberg et al., 1991; Roche et al., 1981). In 1985, the National Institutes of Health (NIH) Consensus Conference on Obesity recommended that BMI plus relative weight be used in clinical and public health settings (Burton et al., 1985). Although arbitrary, the 85th and 95th percentiles of BMI have been used to define obesity and superobesity, respectively (Garn and Clark, 1975; Harlan, 1993; Must et al., 1991), and these definitions are used in Table C-1. However, use of the 85th percentile may incorrectly label a large group of children obese, and a BMI in the 90th or 95th percentile may be more appropriate (Dietz and Robinson, 1993; Williams et al., 1993). The Committee on Clinical Guidelines for Overweight in Adolescent Preventive Services recently proposed specific criteria for defining overweight and obesity in adolescents (Himes and Dietz, 1994), recommending routine use of BMI in screening. The committee also recommended TABLE C-1 Body Mass Index (BMI) Cutoff Values for Obesity for Children 6–10 Years of Age by Sex   Males   Females   Age 85th Percentile 95th Percentile 85th Percentile 95th Percentile 6 16.6 18.0 16.2 17.5 7 17.4 19.2 17.2 18.9 8 18.1 20.3 18.2 20.4 9 18.9 21.5 19.2 21.8 10 19.6 22.6 20.2 23.2   SOURCE: Adapted from Must et al., 1991. Reprinted with permission of the authors and the American Society for Clinical Nutrition.

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs TABLE C-2 Recommended BMI Cutoff Values for Adolescents Who Are at Risk of Overweight or Are Overweight   At Risk of Overweight   Overweight   Age (years) Males Females Males Females 10 20 20 23 23 11 20 21 24 25 12 21 22 25 26 13 22 23 26 27 14 23 24 27 28 15 24 24 28 29 16 24 25 29 29 17 25 25 29 30 18 26 26 30 30 19 26 26 30 30 20–24 27 26 30 30   SOURCE: Adapted from Himes and Dietz, 1994, and Must et al., 1991. Reprinted with permission of the authors and the American Society for Clinical Nutrition. that BMI be used routinely to screen for overweight adolescents. Adolescents with either a BMI in the 95th percentile and higher or a BMI greater than 30 (whichever is smaller) should be considered overweight and referred for in-depth medical follow-up. Adolescents whose BMI is greater than or equal to the 85th percentile but less than the 95th percentile or whose BMI is equal to 30 (whichever is smaller) should be considered at risk for overweight and referred for further screening. The recommended cutoff values are shown in Table C-2. Similar guidelines have yet to be developed for younger children. Prevalence Table C-3 shows estimates of the prevalence of obesity and superobesity among children 6–17 years old using data gathered during 1976–1980. In 1987, the prevalence of obesity and superobesity in children was estimated at 27.1 and 11.7 percent (6–11 years) and 21.9 and 9.0 percent (12–17 years) (Gortmaker et al., 1987). Obesity and superobesity were defined as 85th and 95th percentile of triceps skinfolds, respectively. A comparison of the 1976–1980 levels and the 1987 levels shows that, depending on the age, sex, and race of the group studied, the prevalence of obesity has increased by 54 percent in 6- to 11-year-olds and by 64 percent

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs TABLE C-3 Estimate of Prevalence of Obesity in the United States by Race and Sex in Children 6–17 Years of Age (1976–1980)   85th Percentile 95th Percentile Males 17.9 5.8 Females 17.3 6.1 Blacks 8.8 2.7 Whites 19.1 6.5 Total 17.6 7.1 NOTE: Data are from the Second National Health and Nutrition Examination Survey (1976–1980) (NHANES II). SOURCE: Adapted from Gortmaker et al., 1987. in 12- to 17-year-olds, and the prevalence of superobesity has increased 98 percent in 6- to 11-year-olds and 64 percent in 12to 21-year-olds. The 1987 data also indicated that not only was the pediatric population getting fatter, but the fatter members were becoming more obese (Gortmaker et al., 1987). Sex and racial differences in obesity have also been noted. Girls had an increased prevalence of obesity compared with boys in all age groups and degrees of obesity (Harlan, 1993). More recent data from NHANES III indicate that these trends for adolescents have continued (DHHS, 1991; Harlan, 1993). Similarly, with respect to racial differences, obesity and superobesity are less prevalent among African-American than white children (see Table C-3); however, higher than average rates of obesity have been reported in other minority populations, including Native Americans, Puerto Ricans, and Cuban-Americans (Kumanyika, 1993). Ethnicity The prevalence of obesity varies among children of different ethnic groups. In this country, approximately 20 percent of the population is made up of ethnic minorities. A high prevalence of obesity has been reported among children of certain minority groups, although this finding depends on the age and sex of the group studied as well as the data used to obtain the information (Kumanyika, 1993). A high prevalence of obesity has been reported in preschool and school-aged Native American children. Increases in the prevalence of obesity among preschoolers have

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs also been reported among Puerto Rican males and females and Cuban-American females. Among school-aged children, increases in obesity have been reported among African-American and Puerto Rican girls; Mexican, Puerto Rican, and Cuban boys; and Native Hawaiian boys and girls. Additional information on ethnicity and pediatric obesity is provided by Kumanyika (1993). ETIOLOGY Energy Intake Obesity results from an energy imbalance in which energy intake exceeds energy expenditure. This imbalance can result from excessive energy intake, low energy expenditure, or some combination of both. The relative importance of each of these factors in children and adolescents in causing and sustaining the obese state remains a matter of debate (Gutin and Manos, 1993). Although there has been considerable debate over the issue of whether the obese consume more calories than the nonobese (Griffiths et al., 1987; Rolland-Cachera and Bellisle, 1986; Stefanick et al., 1959; Sunnegardh et al., 1986), there is little evidence to support this theory, and no general consensus exists. For example, several studies have reported a lack of association between energy intake and body weight in adults (Braitman et al., 1985; Romieu, 1988). With regard to children and adolescents, energy intake (as measured by self-report) does not appear to differ significantly among obese, obesity-prone, and normal-weight youths (Corbin and Fletcher, 1968; Eck et al., 1992; Johnson et al., 1956; Stefanick et al., 1959; Wilkinson et al., 1977). Excessive fat intake has been proposed as a determinant of obesity, and data indicate that obese children may consume more calories as fat than do nonobese children (Eck et al., 1992), although these results remain equivocal (Obarzanek et al., 1994; Rose and Mayer, 1986). In addition, self-report data on dietary intake may not be a valid measure of habitual energy intake (Bandini et al., 1990b; Lichtman et al., 1992). Indeed, using doubly-labeled water techniques, Bandini et al. (1987) found that both obese and nonobese adolescents underreport habitual energy intake. While the self-report estimates of intake of the obese and lean subjects were similar, the doubly-labeled water results showed that obese adolescents underreported caloric intake by 40 percent, while nonobese adolescents underreported caloric intake by only 20 percent. The doubly-labeled water technology allows researchers to assess energy expenditure of free-living individuals accurately over repeated days. Since obesity can result from even a small imbalance in energy intake, a

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs variety of accurate and readily accessible methods for measuring caloric intake must be developed. Additionally, further research is needed to clarify our understanding of the role of energy intake and macronutrient composition in the development of childhood and adolescent obesity. Energy Expenditure There has been much interest in the role of energy expenditure in the development and maintenance of obesity. Total energy expenditure (TEE) is divided into three main components: resting metabolic rate, thermogenesis, and physical activity. If any of these components were reduced, it theoretically could predispose an individual to obesity. Three prospective investigations have examined the role of low energy expenditure in the development of obesity in children (Griffiths and Payne, 1976; Griffiths et al., 1990; Roberts et al., 1988). In one, 18 infants of lean and obese mothers were studied during their first year of life (Roberts et al., 1988). Total energy expenditure and metabolizable energy intake were measured using doubly-labeled water over a period of 7 days when the infants were 3 months of age; postprandial metabolic rate was measured by indirect calorimetry when the infants were 3 days and 3 months of age. No significant differences with respect to weight, length, skinfold thicknesses, metabolic rate (at 0.1 and 3 months), and metabolizable energy intake (at 3 months) were observed between infants who became overweight by the age of 1 year (50 percent of infants born to overweight mothers) and those who did not. However, total energy expenditure at 3 months was 20.7 percent lower in the infants who became overweight than in infants who remained normal weight. These findings indicate that at 3 months of age, low energy expenditure (rather than high energy intake) was the principal cause of rapid weight gain, which is consistent with prospective studies of weight gain in adult Pima Indians (Ravussin et al., 1988). The authors attributed this energy difference to diminished physical activity and/or arousal rather than resting metabolic rate. Two additional studies support the hypothesis that low energy expenditure contributes to weight gain in infancy and childhood. In the first study, 3- to 5-year-old children of obese parents were found to have lower resting metabolic rates and energy intake (on average 22 percent lower) per kilogram of body weight than did children of normal-weight parents, even though the children were matched for size and fatness (Griffiths and Payne, 1976). These results suggested that children who have a family history of obesity maintain normal weight by consuming fewer calories per day than do children without a family history. Although these results do not demonstrate any relationship between low

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs energy expenditure and the development of obesity (since the children were normal weight at the time of the study), they might be predictive of obesity at some later time. A recent follow-up investigation of this same population at ages 15 and 16 years showed that the children of the obese parents gained substantially more weight than did the children of normal-weight parents (Griffiths et al., 1990), suggesting that low energy expenditure earlier in life did contribute to weight gain. Boys with obese parents were taller and heavier (but not fatter) than boys with nonobese parents. However, resting metabolic rate per kilogram of body weight was significantly lower in the boys with obese parents compared with those with nonobese parents; no differences were noted between girls with obese and nonobese parents, though body fat was greater in the girls with the obese parents (not statistically significant, possibly owing to the small sample size) (Griffiths et al., 1990). Physical activity represents a significant variable in total energy expenditure and the only component that can be directly controlled by the individual. Several studies have reported lower levels of physical activity among obese infants (Roberts et al., 1988; Rose and Mayer, 1986), children (Johnson et al., 1956), and adolescents (Bullen et al., 1964) but have been contradicted by other studies (Bandini et al., 1990a; Berkowitz et al., 1985). The issue of physical activity and obesity is complicated by several factors that make it difficult to draw conclusions. First, free-living physical activity is difficult to measure (Gutin and Manos, 1993). In addition, obese children use more energy for the same amount of movement. Defining physical activity as energy used may lead to the conclusion that the obese child engages in more activity than a lean child, while defining physical activity as movement may lead to the conclusion that the obese child engages in less physical activity (Dietz, 1992). Finally, levels of physical activity may vary over a lifetime. Just as periods of overeating may precede the development of obesity, periods of reduced activity may likewise precede the development of obesity. Currently, it appears that fatter children move less than leaner children. With the use of more precise measurement techniques, the role of physical activity in the development and/or maintenance of obesity will become clearer. One hypothesis that needs further testing is that a low energy output is in some way corrected by an increase in food intake and subsequent weight gain (Hirsch and Leibel, 1988). The short-term imbalance of energy intake and expenditure required for the development of obesity may be small, making it difficult to identify as the cause. Measures of inactivity (e.g., time spent in a car or watching TV) may also be related to obesity independent of activity (Gortmaker et al., 1990). Further investigations are needed to identify the mechanisms that cause energy

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs expenditure to be reduced early in life and to characterize more accurately the role of energy intake in promoting obesity. Genetic Factors Convincing evidence now shows that body size and fatness have a strong genetic component (Stunkard et al., 1986). If both parents are overweight, approximately 80 percent of the offspring will be overweight. If neither parent is overweight, fewer than 10 percent of the children will be overweight (Bray, 1987). If children remain obese as they age, a greater percentage of them will become obese adults, especially among teenage children (Abraham and Nordsieck, 1960). In addition, the weight status of other family members may affect childhood obesity; obese children are more likely to have obese siblings (Garn et al., 1980, 1981). The weight status of the parents may also affect the response of the obese child to treatment. Epstein et al. (1987) have reported that obese children of heavy parents do not lose weight as successfully as obese children of thin parents. In addition, in the absence of any treatment effort, a young obese child with two obese parents is much more likely to remain obese than an obese child of lean parents (Garn et al., 1981). Studies of nuclear families and adoption data report heritability levels for BMI of about 20 to 30 percent (Bouchard and Pérusse, 1993a). However, twin studies have yielded higher heritability estimates of 60 to 90 percent (Feinleib et al., 1977; Stunkard et al., 1986, 1990). Taken together, the research to date suggests an important genetic component in the development of obesity. This genetic component interacts with environmental agents, primarily diet and level of physical activity, to produce obesity in susceptible people. Environmental Factors Several environmental factors have been linked to obesity in children. Childhood obesity has been associated with season, geographic region, and population density. Prevalence is higher in the winter and spring and is lower in the summer and fall. The highest geographic rate of obesity occurs in the Northeast, followed by the Midwest, South, and West. Finally, the prevalence of obesity in urban areas is greater than in rural areas (Dietz, 1983; Dietz and Gortmaker, 1984). Childhood obesity has also been associated with family characteristics (Dietz and Gortmaker, 1984) such as socioeconomic class (Rolland-Cachera and Bellisle, 1986; Rolland-Cachera, et al. 1988), family size (Ravelli and Belmont, 1979), and education (Sobal and Stunkard, 1989). Sobal and Stunkard (1989) showed that weight was inversely related to

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs socioeconomic status for women but directly related to socioeconomic status for men and children. Jacoby et al. (1975) have shown the prevalence of obesity to be inversely related to family size. Children from high-income families have likewise been found to be fatter than those from lower-income families until late adolescence, when low-income girls become fatter while the girls from high-income families become leaner (Garn and Clark, 1976). As role models and food providers, parents and other family members offer the first and most powerful influence on a child's eating and activity behavior (Garn et al., 1984). This area has been extensively reviewed by Ray and Klesges (1993). However, nonshared familial influences (defined as certain influences on eating and activity that are not shared among all family members) on body weight and obesity have also been noted (Grilo and Pogue-Geile, 1991). Waxman and Stunkard (1980) have reported that parents will feed obese children more than their thin siblings, claiming that obese children need more food because they are bigger. In another study, encouragement to eat from parents was more frequently directed at obese children compared with their lean siblings (Klesges et al., 1983). Thus, shared and nonshared family influences can affect the development of obesity in children. Television represents another environmental factor with significant influence on the development of obesity. Dietz and Gortmaker (1985) reported a direct, positive relationship between television viewing and childhood obesity in a national sample of children and adolescents from diverse socioeconomic backgrounds. The average child 6–11 years old watches more than 20 hours of television a week (A. C. Nielsen Company, 1990), more time than attending school (Dorr, 1986). Children will be exposed to hundreds of commercials during this time, the majority of which are for food products (Ray and Klesges, 1993). Television may contribute to obesity by displacing more vigorous activities (Ross et al., 1987; Tucker, 1986), thereby reducing physical activity (or promoting inactivity) and increasing food and snack consumption, particularly of the nonnutritious foods frequently advertised (Clancy-Hepburn et al., 1974; Taras et al., 1989). Television viewing has been shown to be associated with significant reductions in resting metabolic rate among obese and nonobese girls (Klesges et al., 1993), although these results were not replicated in a more recent study (Dietz et al., 1994). Restriction of television viewing has been suggested to reduce body fat and prevent the development of obesity (Gortmaker et al., 1990). Clearly more research is needed to improve our understanding of how television influences both eating and activity behavior as well as body weight.

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs Dieting Behaviors and Weight Preoccupation Children as young as elementary school age appear to be responding to the societal and health pressures against obesity and are dieting to lose weight. Thus, dieting in childhood has become a very common behavior (Lifshitz et al., 1993). In one survey, almost 40 percent of high school children in an upper-class suburban area were dieting on the day of the survey, and almost two-thirds reported dieting during the preceding 4 to 8 weeks (Moses et al., 1989). Children not only diet but also worry about their appearance. Children who are concerned about their weight and who undertake dieting can adversely affect their physical as well as psychological well-being. Inappropriate nutrient intake may lead to nutritional dwarfing and to a variety of other nutrition problems, as has been described elsewhere (Lifshitz, 1993). Children and adolescents can develop eating disorders such as anorexia and bulimia as a consequence of their weight and dieting preoccupation, which can lead to a variety of medical complications (Palla and Litt, 1988). HEALTH CONSEQUENCES OF OBESITY Medical Obesity is an independent risk factor for the development of a number of chronic diseases that are the leading causes of morbidity and mortality in the United States, including atherosclerotic cardiovascular disease, hypertension, diabetes mellitus, gallbladder disease, and some cancers (NRC, 1989a). Childhood obesity is accompanied by significant morbidity and is acknowledged as a precursor to several risk factors for adult chronic disease. In addition, compared with adult-onset obesity, the persistence of child-onset obesity has been associated with higher rates of morbidity and mortality (Mellin, 1993). Thus, childhood obesity is a public health concern owing both to its immediate impact on health status and to its potential impact on adult body weight. Childhood obesity is a major determinant of elevated blood pressure levels (Clarke et al., 1986; Lauer et al., 1991) and serum lipids and lipoproteins (Aristimuno et al., 1984; Wattigney et al., 1991). Childhood obesity is also associated with lower levels of high-density lipoprotein cholesterol, high levels of insulin, and increased heart rate and cardiac output (Freedman et al., 1987; Soto et al., 1989; Srinivasan et al., 1993). In addition, obese children are at greater risk for psychosocial dysfunction (Kimm et al., 1991; Wallace et al., 1993), respiratory disease (Simpser et al., 1977; Tracey et al., 1971), certain orthopedic problems (Figueroa-Colon

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs et al., 1992), and possibly diabetes (Deschamps et al., 1978). Data from longitudinal studies suggest that persistent childhood obesity can predict later health risks and adult mortality (Javier Nieto et al., 1992; Mossberg, 1989; Must et al., 1992). For example, school-age children, initially 13 to 18 years of age, were followed for up to 55 years. Follow-up mortality was greater among those who were overweight in adolescence; except for diabetes, the risks appear to be independent of overweight in adulthood (Must et al., 1992). Psychosocial The most prevalent consequences of childhood obesity are psychosocial. Obese children have been shown to suffer depression (Sheslow et al., 1993) and problems with self-esteem (Sallade, 1973). Children as young as 5 years old learn to associate obesity with a variety of negative characteristics, using adjectives like ''ugly," "lazy," "stupid," and "dirty" to describe the obese body type. By the time children reach the first grade, they prefer other disabilities over obesity (Richardson et al., 1967). In addition, parents themselves may have negative perceptions of their obese child (Waxman and Stunkard, 1980). Although the literature in this area is sparse and inconsistent, at least some children classified as obese perceive themselves negatively (Sheslow et al., 1993; Wadden et al., 1990a). Additional research is necessary to clarify our understanding of the psychosocial aspects of this disease in children and adolescents. At the present time, however, obesity should be viewed as a heterogeneous disease that may require a psychological evaluation to rule out associated social and emotional problems. Longitudinal Several longitudinal studies tracking obesity in children indicate that overweight children are more likely than nonobese children to become obese as adults (Mossberg, 1989; Stark et al., 1981; Zack et al., 1979). The relative risk of an obese child becoming an overweight adult increases with the child's age. Approximately one-fourth of obese infants will become obese adults, whereas 80 percent of obese adolescents will become obese adults (Charney et al., 1976; Garn et al., 1989; Mossberg, 1989). Severity of childhood obesity also predicts adult obesity. In one study, one-third of children who were 120 percent of ideal body weight were normal weight by follow-up (approximately 9 years later); but none of those children who were greater than 160 percent of ideal body weight were normal weight at follow-up (Börjeson, 1962).

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs at the reduced body weight. In moderately overweight young children, the goal may simply be to maintain body weight since future growth will normalize the body weight. While treatment combines dietary change, an increase in exercise and physical activity, and behavior modification, what is unique to childhood treatment programs is the inclusion of parent training. Effect of Age Because of the wide variation in developmental capabilities between the ages of 1 and 18, the age of the obese child must be considered when planning the treatment (Epstein, 1985). The child's capacity for understanding the disease is limited by his or her cognitive and motor abilities, which vary with age (Epstein, 1993a, b). In the very young child (1–5 years of age), parents have major control over the child's eating and activity and therefore represent an important focus of treatment in terms of food selection, preparation, and availability. In later years, the responsibility shifts from parents to the child, particularly as the adolescent strives to achieve autonomy. In fact, research among adolescents has shown that the outcome improves when parents and children are treated separately (Brownell et al., 1983). To date, little research has been devoted to age-appropriate intervention programs, though Epstein (1985) has developed a multistage model for the treatment of childhood obesity in which the responsibility for habit change is shifted from the parent to the child on the basis of the child's age and stage of development. Most intervention studies have focused on treatment of preadolescent children (6–12 years of age) as compared to younger children (1–5 years) or adolescents (13–19 years) (Epstein, 1993b; Epstein and Wing, 1987). Future studies must focus on younger children and adolescents to fill in these gaps in our understanding of treatment and its outcome, to identify an optimal or preferred age to initiate treatment, and to develop treatment strategies that would be age appropriate. Parent Training Since obesity is a familial disorder (Garn and Clark, 1976), the family, particularly the parents, is important in the treatment of the disease. As previously discussed, family influence is both environmental (e.g., modeling, food availability, and food and activity habits) as well as genetic. Epstein and Wing (1987) have demonstrated in numerous investigations the importance of family involvement in the treatment of obesity. The most impressive of these studies showed that targeting both obese children and their parents during treatment results in lower body weights

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs and significantly less obesity at 5 and 10 years of follow-up than in a comparison group (Epstein et al., 1990c). In addition, more participants in the parent-plus-child group approached or achieved normal weight after 10 years than in the comparison group. Although participants were treated for a relatively short period of time (eight weekly visits and six monthly visits), these data provide the first evidence that a behavioral, family-based treatment program initiated when the child is 6 to 12 years of age can produce long-term effects that persist into young adulthood. Exercise Physical activity has an important role in the treatment and prevention of obesity in children (Epstein et al., 1984a). Despite the controversy on the effect of inactivity in the development of obesity, physical activity in combination with dietary change has been shown to be effective in reducing obesity in children (Epstein, 1992; Epstein et al., 1985b). The inclusion of exercise has been shown to help preserve fat-free mass and to retard the reduction in metabolic rate that occurs among adults who are dieting (Whatley and Poehlman, 1994). To date, however, the results are inconclusive in children and require further study. Some researchers have reported greater weight loss when exercise is combined with a low-calorie diet versus diet treatment alone (Epstein et al., 1985b; Reybrouck et al., 1990). Exercise may exert its greatest influence by promoting a more active lifestyle throughout adolescence and into adulthood (DHHS, 1991). Epstein et al. (1984b, 1985a) have evaluated two types of exercise programs (aerobic versus lifestyle) in conjunction with a hypocaloric diet in obese children. Both programs were isocaloric and included equivalent amounts of energy expenditure through exercise. The goal of the lifestyle program was to increase energy expenditure by increasing the amount of the participant's less structured, less intense daily activities (e.g., walking and climbing stairs). The aerobic exercise program included activities (e.g., running, cycling, swimming, or walking) of a specific duration and intensity. The results demonstrated that the lifestyle exercise program was associated with better weight change over the long term (17 to 24 months) but not over the short term (in which both programs were equal). The lifestyle program may have been more effective because of increased long-term adherence. Research in adults indicates that many people who begin exercise programs drop out (Oldridge, 1982). Epstein (1992) and colleagues (1984a) have shown that exercise intensity is an important determinant in predicting exercise nonadherence (i.e., the greater the intensity, the greater the dropout rate). Thus, exercise programs must always be evaluated with respect to adherence as well as changes in weight, fitness, and body composition.

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs Finally, data also demonstrate the importance of family involvement with regard to exercise adherence. Family-based programs in which parents are trained to reinforce their children's physical activity have reported increases in both activity and fitness levels in obese children (Epstein et al., 1987, 1990c). Diet As with obesity treatment in adults, diet remains the cornerstone of therapy for the obese child since caloric restriction produces far greater energy deficits than exercise alone. Debate continues, however, about the degree of calorie restriction that should be used in children. Most diet programs for children fall into one of two fundamental types: a balanced hypocaloric diet or the more restrictive protein-sparing modified fast. Clearly, the diet should be matched to the severity of the problem. A balanced hypocaloric diet, which is low in fat, high in complex carbohydrates, and contains adequate protein, is appropriate for most children and adolescents. This diet should be sufficient to promote growth while concurrently decreasing excess body fat. A reasonable goal on such a program is to lose 1 to 4 pounds per month, with maximal loss of fat and preservation of lean tissue. The diet must provide adequate protein, calories, and micronutrients to ensure normal growth and development (Dietz, 1983; Dietz and Robinson, 1993; Epstein, 1986; Williams et al., 1993). Because pediatric patients who undergo weight loss are in a period of growth and development, it is important to evaluate weight change as a function of stature. When administered properly, these diets pose no recognized hazard for continued growth during weight reduction (Epstein et al., 1990c, 1993) and have been used by children as young as 5 years of age with good success (Epstein, 1986). Caloric intake may range from 1,200 to 2,000 calories per day or a deficit of 30 to 40 percent of usual intake (Figueroa-Colon et al., 1992), but should be adjusted to promote gradual weight loss (Williams et al., 1993). Children and parents must be taught portion size and food exchanges to provide balance and variety within the restricted intake. The balanced hypocaloric diet programs can produce a 5-kilogram weight loss over 10 weeks (Figueroa-Colon et al., 1992). In a review of the literature, Epstein and Wing (1987) reported reductions in percentage of overweight that ranged from 3.2 to 28 percent over 2 to 12 months among published controlled studies. Variations in treatment outcome could be explained by the different components of the programs (i.e., family-based, treatment contingencies, or exercise). The more restrictive protein-sparing modified fast (PSMF) should be

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs limited for use in more serious cases of childhood and adolescent obesity, for which rapid weight reduction is essential. Use of a PSMF must always be carried out under the careful supervision of a physician (Merritt et al., 1980, 1981, 1983; NTF, 1993; Wadden et al., 1990b). In adults, very-low-calorie diets are generally recommended for individuals who are at least 30 percent or more over ideal body weight (Wadden et al., 1990b). In children, although the data are limited, the PSMF has been used on children as young as 6 years of age and by children whose body weight ranges from 120 percent to greater than 200 percent of ideal body weight (Suskind et al., 1993). In England, the PSMF is not recommended for use by children under the age of 13 (Department of Health and Social Security, 1987), though in the United States no consensus exists on what represents appropriate indications for its use in children. Recently, the National Task Force on the Prevention and Treatment of Obesity (1993) suggested that use of the PSMF in children and adolescents should be considered experimental and treatment should be administered by experienced medical staff. The PSMF is contraindicated in children who have renal, hepatic, or cardiac disease, and all patients should be monitored closely for sustained nitrogen losses, cardiac arrhythmias, and cholelithiasis (Dietz and Robinson, 1993). The PSMF for children and adolescents typically ranges from 600 to 800 calories per day. The diet provides 1.5 to 2.0 grams of high-quality protein per kilogram of ideal body weight per day, which is higher than the level recommended for adults because of the potential for complications from increased nitrogen losses in children (Dietz and Robinson, 1993; Dietz and Wolfe, 1985; Merritt et al., 1980). In addition, at least 2 liters of water or calorie-free fluids are consumed daily, along with 2–4 cups of low-starch vegetables, one multivitamin tablet containing iron, 800 milligrams of calcium, and 25 milliequivalents of potassium. Details of the dietary protocol for the PSMF are given by Dietz (1983), Figueroa-Colon et al. (1992), and Suskind et al. (1993). An average weight loss of about 10 kilograms in 10 weeks has been reported for use of the PSMF (Figueroa-Colon et al., 1992; Suskind et al., 1993). Long-term data are lacking, though recently Figueroa-Colon et al. (1993) reported that children maintained a 10-kilogram weight loss produced on a PSMF for 14.5 months. In another study, 12 of 17 obese adolescents treated with a PSMF diet were available for follow-up at 1 year (Stallings et al., 1988); of these, 48 percent had maintained their weight loss. Although not frequently used in children or adolescents, a liquid very-low-calorie diet (VLCD) has been shown in one study to have poor compliance and a high dropout rate in eight adolescents (Brownell et al., 1983). Clearly, more short- and long-term studies (with at least 2 years or more of follow-up) are needed to characterize treatment outcome in this

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs population, particularly with regard to the effect of severe caloric restriction on linear growth and development. Behavior Modification Behavior modification facilitates weight-loss efforts by helping patients develop appropriate eating and exercise behaviors and cope with disturbing weight-related thoughts and emotions (Stunkard, 1987). The effectiveness and importance of this treatment component has been documented in more than 100 controlled clinical trials (Brownell and Kramer, 1994). Standard behavioral programs include self-monitoring, positive reinforcement, stimulus control, cognitive restructuring, and nutrition education. The details of the standard behavior modification program have been reviewed extensively by Epstein (1986) and Kramer et al. (1989). Long-Term Outcome The true effectiveness of treatment can be measured only by examining data on long-term outcome. While few studies have attempted to follow patients beyond the initial treatment, results reported to date have been promising, especially compared with 5- and 10-year outcomes in adults (Kramer et al., 1989). Epstein et al. (1990b) evaluated the 5-year outcome results of four prospective, randomized, controlled, multidisciplinary treatment programs that differed with respect to how family-based treatment or exercise was implemented. Subjects were 6–12 years of age at the start of the studies and were randomized to alternative treatments that lasted 8–12 weeks, with monthly meetings continuing for another 6–12 months. One subgroup within each of two of the studies achieved a 10 percent reduction in relative weight at the end of 5 years. The best results were achieved by reciprocal reinforcement of the children and their parents compared with a program that reinforced only the child for success. Ten-year follow-up data have been reported for a three-arm study of multidisciplinary treatment (Epstein et al., 1990c). The groups differed only in the method of reinforcement for weight and behavior change. Children in the group that reinforced both parents and children showed significantly greater reductions in percentage of overweight after 5 and 10 years (11.2 and 7.5 percent, respectively). In addition, a greater percentage of children in the parent-plus-child treatment group achieved or approached normal weight for height (33 percent) than of the children treated without their parents (5 percent). Nuutinen and Knip (1992) conducted a 5-year follow-up study of 48

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs obese children who participated in a multidisciplinary weight-reduction program. They defined a successful weight loser as a child with at least a 10 percent reduction in relative weight after 2 years of active treatment. At 5 years' follow-up, 49 percent (n = 22) of subjects who had been treated and defined as successful weight losers after 2 years were still able to maintain at least a 10 percent decrease in relative weight, and 31 percent (n = 14) of children were not obese after 5 years. Favorable changes in cardiovascular risk factors (e.g., reduced total cholesterol, triglycerides, and plasma insulin levels and increased HDL cholesterol levels) were also reported. School-Based Studies Few obesity intervention studies have been conducted in the schools. However, schools provide an excellent opportunity for preventing and treating this disease, as children spend a significant amount of their first two decades of life in school and more than 95 percent of American youth aged 5–17 are enrolled in school (NCES, 1990). In contrast to clinic programs, school programs require no financial commitment from parents and may reach children who might otherwise not seek treatment. In addition, children eat one or two meals at school, which offers the opportunity to expose them to nutritious foods and to teach healthful eating habits. Over the past 12 years, only four large controlled studies have been conducted in schools (Brownell and Kaye, 1982; Foster et al., 1985; Lansky and Brownell, 1982; Lansky and Vance, 1983). Each program included nutrition education, exercise, and behavior modification strategies, and interventions lasted from 10 to 18 weeks. The programs were administered only to overweight or obese children. Three of the studies compared the intervention group with a no-treatment control (Brownell and Kaye, 1982; Foster et al., 1985; Lansky and Brownell, 1982); one study compared the intervention with a standard health-education program as the comparison group (Lansky and Brownell, 1982). In all four studies, the intervention group had a significantly greater reduction in percentage of overweight compared with the control group (see Table C-4). A more in-depth summary of these studies is provided by Epstein and Wing (1987) and Resnicow (1993). Unfortunately, few follow-up data are available from these studies, and the long-term effects remain unknown. School-wide programs that would include both nonobese and obese children may also have an impact on children not currently obese but at risk (e.g., thin children of two obese parents). Thus the school represents a potentially useful setting for

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs TABLE C-4 Summary of Multidisciplinary School-Based Intervention Studies Study Age (years) Treatment Group Randomized Number Treatment Duration (months) Change in % Overweight Brownell and Kaye, 1982 5–12 beh mod yes 63 10 -15.4     control   14 — -2.8 Foster et al., 1985 5–12 beh mod no 43 3 (6)a -5.4 (-3.6)a     control   41 — +0.3 (-0.2)a Lansky and Brownell, 1982 12–15 beh mod yes 36 3.5 -3.0     health ed   35 — -2.1 Lansky and Vance, 1983 11–14 beh mod no 30 12 -5.7     control   25 — +2.4     ss control   59 — -1.5 NOTE: Beh mod = behavior modification; control = no treatment control; health ed = health education program which includes nutrition and exercise education; ss control = self-selected control. a 6-month follow-up data are available and denoted in parentheses. SOURCE: Epstein, L.H., and R.R. Wing. 1987. Behavioral treatment of childhood obesity. Psychology Bulletin 101:331–342. Copyright 1987 by the American Psychological Association. Adapted by permission.

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs the primary and secondary prevention of obesity, and more research needs to be initiated in this setting. PREVENTION Because of the refractory nature of obesity, prevention remains the treatment of choice (Council on Scientific Affairs, 1988). Preventing childhood obesity may, in fact, be an effective way to prevent adult obesity. However, to date little research has addressed the prevention of pediatric obesity. In the only controlled study specifically investigating the prevention of obesity in children, Pisacano et al. (1978) gave parents of infants health education emphasizing a low-fat, prudent diet. Compared with a usual-care control group, the prevalence of obesity was significantly reduced for intervention group children from 3 months to 3 years. Available epidemiologic data make it possible to identify individuals at increased risk for obesity in whom preventive measures could be implemented. Parental obesity is an important risk factor for childhood obesity (Garn and Clark, 1976). In children of obese parents, careful monitoring of weight change could help to identify children for whom alterations in diet and physical activity are needed to prevent the disease. Recently, specific periods for the development of obesity and its health sequels have been identified (Dietz, 1994a). These include the period of adiposity rebound that occurs between the ages of 5 and 7, adolescence, and possibly the period of gestation and early infancy. However, it is unclear whether these periods represent critical periods of the onset of persistent obesity. Further research is needed to identify the most effective time to initiate treatment of childhood obesity and possibly prevent adult obesity; the most effective time to initiate preventive efforts for childhood obesity could also be the focus of research. Prevention interventions could be adapted from existing treatment programs (Epstein and Wing, 1987; Epstein et al., 1990c). With respect to prevention, two important interventions are altering diet and increasing physical activity. Decreasing the amount of inactivity, possibly by limiting television exposure, may also be effective (Gortmaker et al., 1990). A recent study has suggested that early and frequent (more than four visits) intervention among preschool children (ages 1–5 years) was effective at reducing the degree of obesity over a 1-year period (Davis et al., 1993a). These results, while provocative, suggest that future research must address the effect of frequency and timing of the intervention on the long-term outcome of preventive treatment. School-based prevention programs would likewise offer an excellent setting for intervention and education (Resnicow, 1993). Although many gaps exist in our current knowledge of the etiology and treatment of

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs obesity, the foundation is in place to develop and evaluate the primary prevention of obesity in the pediatric population. SUMMARY AND DIRECTIONS FOR FUTURE RESEARCH In its Healthy People 2000 report, the federal government proposed the goal of reducing obesity to a prevalence of no more than 15 percent among adolescents aged 12 to 19 years (DHHS, 1991). From a health perspective, obesity treatment in children and adolescents seeks a reduction in body fat that is safe; involves minimal hunger; preserves lean body mass; and allows for normal growth, development, and activity. As with adult programs, comprehensive treatment for childhood and adolescent obesity includes a hypocaloric diet, an exercise program, and a behavior management program. The program must be appropriate for the child's age and stage of development and include the parents and family in treatment. Few data address the etiology, treatment, and prevention of childhood and adolescent obesity. More studies must be conducted to develop a consensus in a number of areas. Six general areas have been identified for future studies: definition, etiology, treatment, eating behavior, weight maintenance, and prevention. Although not exhaustive, the following observations summarize the most salient research agendas in each of these areas. Definition Although new guidelines have been issued by the Expert Committee on Clinical Guidelines for Overweight in Adolescent Preventive Services, the expected sensitivity, specificity, and predictive value of the two-level screening remains unknown. In addition, no consensus exists for equivalent values in preschool (ages 1–5 years) and preadolescent (ages 6–9 years) children. Further research is clearly needed to provide standard guidelines for the screening of at-risk and overweight children of all ages. Etiology Despite several theories, the causes of obesity in children remain unclear, particularly those related to energy imbalance and age of onset. Fundamental to answering these questions is the development of improved techniques for measuring subtle changes in energy expenditure (including total energy expenditure, physical activity, resting metabolic rate, thermogenesis, and energy intake). Some evidence suggests that

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs obese children consume a greater percentage of calories as fat than do normal weight children, thus making the effect of macronutrient composition on the development of obesity another area of potential interest. Treatment What is the optimal approach to treating childhood obesity? Standard programs tend to be modeled after adult treatment programs and include dietary change, increasing physical activity, and behavior modification along with parent training. However, future work must continue to examine the efficacy of each component and to test variations in intervention in different populations. For example, continued investigation of the short- and long-term (into adulthood) effects of use of a PSMF or liquid VLCD is needed, as is research into the role of duration of treatment. A large body of work has come from Leonard Epstein and his colleagues at the University of Pittsburgh, who have explored different behavioral strategies and their effect on treatment outcome. This work needs to be continued and effective treatment models refined. Other important research questions include the identification of appropriate treatment goals for children and adolescents as well as identifying the optimal age for treatment. Interventions must also be tailored to the child's age and level of cognitive, social, and emotional development. Because of the psychological consequences of adolescent-onset obesity and the disproportionate share of adult obesity that originates in adolescence, treatment for this group deserves special emphasis. Research must be undertaken to determine the most effective methods for intervention during adolescence, bearing in mind the vulnerability of adolescents to eating disorders. Treatment programs also need to be culturally and ethnically appropriate. Much research is needed in the area of minority food preference, preparation habits, and parental influence to develop effective treatment programs for different ethnic groups. Eating Behavior More research must explore the effect of family, social/cultural, and environmental factors on eating behavior. Epstein et al. (1990c) have shown that children can lose and maintain weight over an extended period of time and that parental participation is essential to a positive outcome. It is likely, however, that the motivation structures within the family that affect behavior change may be as important as the behaviors being modified (Epstein, 1993a). Future research must identify methods that mobilize family support for behavior change. In addition, descriptive

OCR for page 210
Weighing the Options: Criteria for Evaluating Weight-Management Programs research is needed to clarify our understanding of how the family contributes to or can help mitigate the development of obesity. Research into how environmental factors (e.g., television, video games, cars, and readily available snack foods) affect behavior and body weight must also continue. Weight Maintenance Strategies for promoting maintenance of weight loss and preventing relapse have been largely ignored in the pediatric population. Future cstudies must address both maintenance and relapse prevention in children and adolescents. Prevention Almost no research has been done in the area of prevention, although prevention of pediatric obesity may be the only effective treatment of adult obesity. By adapting existing treatment protocols, programs directed at primary prevention of obesity can be developed. Issues to be resolved include who should be targeted (all children or just those at risk) and where these programs would be administered (e.g., in schools, in communities, through physicians' offices). Schools clearly represent a cost-effective setting, but intervention programs must be developed and tested in other locations as well. At least two or three critical periods have been identified during childhood for the development of obesity and its putative health consequences (diabetes, hypertension, hypercholesterolemia, and cardiovascular disease). These periods represent potential treatment targets for the prevention of adult obesity, and additional research is needed to identify the most effective time to target efforts to prevent and treat childhood obesity.