Treatment of Pediatric Obesity

18. Pate RR, Dowda M, Ross JG. Associations between physical activity andphysical fitness in American children. Am J Dis Child. 1990;144:1123–1129

19. Ross JG, Dotson CO, Gilbert GG, Katz SJ. What are kids doing in schoolphysical education? J Phys Educ Recr Dance. 1985;56:31–34

20. Luepker RV, Perry CL, McKinlay SM, et al. Outcomes of a field trial toimprove children’s dietary patterns and physical activity. The child andadolescent trial for cardiovascular health. JAMA. 1996;275:768–776

21. Ross JG, Dotson CO, Gilbert GG, Katz SJ. After physical education—physical activity outside of school physical education programs. J PhysEduc Recr Dance. 1985;56:35–39

22. Taras HF, Sallis JF, Patterson TL, Nader PR, Nelson JA. Television’sinfluence on children’s diet and physical activity. Dev Behav Pediatr.1989;10:176–180

23. Sallis JF, Nader PR, Broyles SL, et al. Correlates of physical activity athome in Mexican-American and Anglo-American preschool children.Health Psychol. 1993;12:390–398

24. Desmond SM, Price JH, Lock RS, Smith D, Stewart PW. Urban black andwhite adolescents’ physical fitness status and perceptions of exercise. JSchool Health. 1990;60:220–226

25. Robinson TN, Hammer LD, Killen JD, et al. Does television viewingincrease obesity and reduce physical activity? Cross-sectional and lon-gitudinal analyses among adolescent girls. Pediatrics. 1993;91:273–280

26. Dietz WH, Gortmaker SL. Do we fatten our children at the televisionset? Obesity and television viewing in children and adolescents. Pedi-atrics. 1985;75:807–812

27. Klesges RC, Shelton ML, Klesges LM. Effects of television on metabolicrate: potential implications for childhood obesity. Pediatrics. 1993;91:281–286

28. Baranowski T, Thompson WO, DuRant RH, Baranowski J, Puhl J.Observations on physical activity in physical locations: age, gender,ethnicity, and month effects. Res Q Exercise Sport. 1993;64:1–7

29. Klesges RC, Eck LH, Hanson CL, Haddock CK, Klesges LM. Effects ofobesity, social interactions, and physical environment on physical ac-tivity in preschoolers. Health Psychol. 1990;9:435–449

30. Shephard RJ, Jequier JC, Lavallee H, LaBarre R, Rajic M. Habitualphysical activity: effects of sex, milieu, season, and required activity.J Sports Med Phys Fitness. 1980;20:55–66

31. Kann L, Warren CW, Harris WJ, et al. Youth risk behavior surveil-lance—United States, 1993. MMWR. 1995;44(SS-1):1–55

32. Prochaska JO, DiClemente CC. The stages and processes of self-changein smoking: towards an integrative model of change. J Consult Clin

Psychol. 1983;51:390–39533. Marcus BH, Rakowski W, Rossi JS. Assessing motivational readiness

and decision making for exercise. Health Psychol. 1992;11:257–26134. Bandura A. Social Foundations of Thought and Action: A Social–Cognitive

Theory. Englewood Cliffs, NJ: Prentice-Hall; 198635. Janz NK, Becker MH. The health belief model: a decade later. Health

Educ Q. 1984;11:1–4736. Ajzen I, Fishbein M. Understanding Attitudes and Predicting Social Behav-

ior. Englewood Cliffs, NJ: Prentice-Hall; 198037. Reynolds KD, Killen JD, Bryson SW, et al. Psychosocial predictors of

physical activity in adolescents. Prev Med. 1990;19:541–55138. Tappe MK, Duda JL, Ehrnwald PM. Perceived barriers to exercise

among adolescents. J School Health. 1989;59:153–15539. Theodorakis Y, Doganis G, Bagiatis K, Gouthas M. Preliminary study of

the ability of reasoned action model in predicting exercise behavior ofyoung children. Percept Motor Skills. 1991;72:51–58

40. Ferguson KV, Yesalis CE, Pomrehn PR, Kirkpatrick MB. Attitudes,knowledge, and beliefs as predictors of exercise intent and behavior inschool children. J School Health. 1989;59:112–115

41. Perusse L, Tremblay A, LeBlanc C, Bouchard C. Genetic and familialenvironmental influences on level of habitual physical activity. Am JEpidemiol. 1989;129:1012–1022

42. Moore LL, Lombardi DA, White MJ, Campbell JL, Oliveria SA, EllisonRC. Influence of parents’ physical activity levels on activity levels ofyoung children. J Pediatr. 1991;118:215–219

43. Freedson PS, Evenson S. Familial aggregation in physical activity. Res QExercise Sport. 1991;62:384–389

44. Godin G, Shephard RJ. Psychosocial factors influencing intentions toexercise of young students from grades 7 to 9. Res Q Exercise Sport.1986;57:41–52

45. Gottlieb NH, Chen M. Sociocultural correlates of childhood sportingactivities: their implications for heart health. Soc Sci Med. 1985;21:533–539

46. Anderssen N, Wold B. Parental and peer influences on leisure-timephysical activity in young adolescents. Res Q Exercise Sport. 1992;63:341–348

47. Greendorfer SL, Lewko JH. Role of family members in sport socializa-tion of children. Res Q Exercise Sport. 1978;49:146–152

48. Kohl HW III, Fulton JE, Caspersen CJ. Validity and repeatability ofmethods to assess physical activity among children and adolescents. Inpress

Treatment of Pediatric Obesity

Leonard H. Epstein, PhD; Michelle D. Myers, MA; Hollie A. Raynor, MS, RD; and Brian E. Saelens, MA

ABSTRACT. The primary goal of childhood obesityinterventions is regulation of body weight and fat withadequate nutrition for growth and development. Ideally,these interventions are associated with positive changesin the physiologic and psychological sequelae of obesity.To contribute to long-term weight maintenance, inter-ventions should modify eating and exercise behaviorssuch that new, healthier behaviors develop and replaceunhealthy behaviors, thereby allowing healthier behav-iors to persist throughout development and into adult-hood. This overview of pediatric obesity treatment, usingpredominantly randomized, controlled studies, high-lights important contributions and developments in pri-marily dietary, activity, and behavior change interven-tions, and identifies characteristics of successfultreatment and maintenance interventions. Potential pos-

itive (eg, reduction in blood pressure, serum lipids, andinsulin resistance) and negative (eg, development of dis-ordered eating patterns) side effects of treatment also aredescribed. Recommendations for improving implemen-tation of childhood obesity treatments, including appli-cation of behavioral choice theory, improving knowledgeof response extinction and recovery in regards to behav-ior relapse, individualization of treatment, and integra-tion of basic science with clinical outcome research, arediscussed. Pediatrics 1998;101:554–570; pediatric obesity,treatment, weight maintenance.

ABBREVIATIONS. BMI, body mass index; PSMF, protein-sparingmodified fast; HDL, high-density lipoprotein.

There has been a lot of interest in developingeffective treatments for obesity, with the over-whelming majority of this research focused on

treating adults. A salient characteristic of the adultobesity treatment is that although obese adults can

From the Department of Psychology, State University of New York atBuffalo, Buffalo, New York.Received for publication Oct 24, 1997; accepted Nov 6, 1997.PEDIATRICS (ISSN 0031 4005). Copyright © 1998 by the American Acad-emy of Pediatrics.

554 SUPPLEMENT by guest on April 3, 2019www.aappublications.org/newsDownloaded from

lose significant amounts of weight, maintenance ofweight loss has proven to be a persistent problem.1–3

There are several more hopeful signs in the treatmentof pediatric obesity by using dietary and activitychanges. For example, two research groups haveshown successful maintenance of treatment effects 5years4 and 10 years5,6 after initiation of treatment. Indirect comparisons of children and adults in familiesprovided similar family-based interventions, pread-olescent children responded with greater relativeweight loss and better maintenance of weight loss,although there was still considerable relapse.7 Al-most one third of the children treated were nonobeseafter 10 years.7 This is encouraging, but it indicatesthat a significant amount of work is required toimprove long-term maintenance of pediatric obesitytreatment.

There are behavioral and biological reasons to beoptimistic about treating pediatric obesity. For exam-ple, it may be easier to mobilize support throughfamilies for obese children than for obese adults;obese children also generally have not had the un-healthy eating or activity patterns as long as obeseadults. In addition, treatment of pediatric obesity cantake advantage of growth and increases in lean bodymass as well as weight change.5,8 Furthermore, in-stead of shrinking adipose cells, treatment at an earlyage prevents the development of excess adiposecells.

Despite these potential advantages, many pediat-ric studies show a decrement in treatment effectsover time. The goal of this report is to provide anoverview of the treatment and prevention of pediat-ric obesity, to highlight important contributions anddevelopments in treatment, and to identify charac-teristics of successful treatment. First, treatment ofpediatric obesity is reviewed. Next, the positive andnegative side effects of treatment are presented.Third, ideas for improving implementation of pedi-atric obesity treatment, including new directions andareas for future research, are discussed. This reportconcludes with a brief summary.

METHODSThe goals for treating childhood obesity are regulating body

weight through adequate nutrition for growth and development,thereby preventing interruption of linear growth, minimizing lossof lean body mass, and preventing endocrine disturbances.9 Inaddition, ideal treatments should be associated with positivechanges in physiologic and psychological sequelae of obesity.Treatments should modify eating and exercise behaviors alongwith the factors that regulate these behaviors, so that the new,healthier behaviors persist throughout development.

This review is organized according to major components oftreating pediatric obesity, including dietary, activity, and behaviorchange components. In addition, more aggressive approaches,including pharmacologic and surgical interventions, are reviewed.The quality of the study design was considered, with the focus onrandomized, controlled studies. In some areas, such as surgeryand drug studies, there are limited randomized controlled studies,and we present uncontrolled research in these areas. When studiesdid not provide sufficient detail to know how the groups weredeveloped (ie, whether subjects were allowed to self-select them-selves to groups or whether they could have been placed ingroups on the basis of factors that could bias interpretation oftreatment and control differences), we erred on the conservativeside and considered them not to be randomized studies. (To give

sufficient attention to the discussion of new ideas, many details ofintervention and supporting studies are not discussed.)

The details of the randomized studies that were reviewed arepresented in Tables 1 and 2 for clinical studies and in Tables 3 and4 for school-based studies. These tables include subject age, groupassignment, sample size, sex distribution, dietary components,exercise component, and results. There were several dependentmeasures used across studies; the most common were changes inpercent of overweight, body mass index (BMI), body weight, andpercent of body fat. Where available, we have provided the base-line values, end of treatment changes, and end of follow-upchanges. To provide a common definition of when treatment endsand follow-up begins, we considered treatment to be continuingas long as subjects were seen at least once per month. The mostrelevant information is significance in the rate of change amonggroups over time, but if this information was not available, within-group differences are presented.

Although this review is divided into contributions of diet,activity, and behavior change components, treatment efficacy isderived from integration of the components of treatment. Forexample, although most treatments include a dietary component,it is generally recognized that nutrition intervention is ineffectiveas a solitary treatment for pediatric obesity.10 Extracting the con-tribution of treatment components is complicated further by vari-ations in multiple components of treatment across studies. Differ-ent caloric ranges for the diet are in the context of different typesof exercise programs and different methods to enhance behaviorchange.

RESULTS

Dietary ComponentsDiet therapy for obesity is founded on the hypoth-

esis that obese individuals consume too much energyrelative to energy expenditure or they consume animbalance of macronutrients. Therefore, the generalgoals of most dietary interventions involve reducingand stabilizing caloric intake, reducing fat intake,and restructuring eating habits to follow moreclosely current dietary recommendations, resultingin increased nutrient density.

If positive energy balance is attributable to excessintake, then the degree of reduction in caloric intakeshould relate to treatment success. Amador et al11

demonstrated this by randomizing children to a re-stricted diet (0.17 MJ/kg of expected body weight forheight) versus a less restricted diet (0.25 MJ/kg ofexpected body weight for height) and found signifi-cantly better results after 1 year for the diet with thelower caloric intake. There is very little work on dietcomposition beyond calories. When obese childrenof elementary-school age were given 15 g of fibersupplementation combined with a reduced-energydiet for 4 weeks, they demonstrated no significantincreases in weight loss or significant decreases inenergy intake compared with the results from 4weeks of the reduced-energy diet alone.12 The caloriclevel of the diet influences weight loss. There hasbeen no research on the contribution of fat content totreatment success.

There are a number of general approaches studiedto reduce caloric intake and teach better eating hab-its. One approach is to provide individualized di-etary interventions. In a controlled study on exercise,no weight loss was observed for preadolescents whowere provided with individualized dietary recom-mendations for 16 weeks without exercise.13 Anotherapproach is the diabetic exchange system, with acaloric level calculated to produce 1 lb of weight loss

SUPPLEMENT 555 by guest on April 3, 2019www.aappublications.org/newsDownloaded from

TABLE 1. Characteristics of Child and Adolescent Obesity Treatment Studies on Clinical Samples

Source Age(y)

GroupAssignment

Between-group Variable N % Girls Diet Exercise

Amador et al11 10–13 R 1. Restrictive diet 47 47.9 Restricted to 30% of Exercise information (1,2)2. Nonrestrictive diet 47 energy requirement

(1), nutritioninformation (1,2)

Aragona et al50 5–11 R 1. Response cost 1reinforcement

5 100 Nutrition information(1,2)

Daily exercise instructionsand program for parents

2. Response cost 3 (1,2)3. Control 5

Bacon and 5–17 R 1. Fenfluramine 20 ;71.4 1000–1200 kcal (1,2) NoneLowrey41 2. Placebo

Within-subject design, orderrandomized

Becque et al14 12–13 R 1. Exercise 11 58.3 ADA exchange for 50 min, 33/wk supervised2. No exercise 11 21–2 lb/wk (1,2) aerobic activity @ 60%–3. Control 14 80% maximal heart rate

(1)Brownell et al56 12–16 RS 1. Mother and child seen

separately14 78.6 Nutrition information

(1,2,3)Exercise information (1,2,3)

2. Mother and child seentogether

15

3. Child seen alone 13Coates et al51 13–17 RS 1. Daily contact, reinforced for

weight8 68.4 Caloric goals estimated

for 21–2 lb/wkMinimal exercise

information (1,2,3,4)2. Weekly contact, reinforced

for weight8 (1,2,3,4)

3. Daily contact, reinforced forcalories

11

4. Weekly contact, reinforcedfor calories

11

Coates et al55 13–17 RS 1. Mother and child seenseparately

31Total

64.5 Nutrition information(1,2)

Exercise information (1,2)

2. Child seen aloneDuffy and 7–13 R 1. Cognitive treatment 14 ;78.6 Traffic-light diet (1,2) Lifestyle, aerobic, and

Spence30 2. Progressive relaxation 13 calisthenic activity (1,2)Epstein et al20 6–12 RS 1. Behavior modification 14 ;38.5 Traffic-light diet 900– Exercise information (1,2)

2. Nutrition education Total 1200 or 1500 kcal (1,2)Epstein et al6,29 6–12 RS 1. Mother and child targeted 76 69.6 Traffic-light diet limit Exercise information (1,2,3)

2. Child targeted Total 1200–1500 kcal (1,2,3)3. Nonspecific target

Epstein et al19 8–12 R 1. Aerobic activity 1 diet 51 ;78.4 Traffic-light diet 900– Aerobic (1,3) or lifestyle2. Lifestyle activity 1 diet Total 1200 or 1500 kcal (1,2) (2,4) activity; isocaloric3. Aerobic activity alone across groups4. Lifestyle activity alone

Epstein et al6,28 8–12 R 1. Exercise 18 NR Traffic-light diet (1,2) Lifestyle activity (200–4002. No exercise 18 kcal/day) (1)3. Control 17

Epstein et al6,26 8–12 R 1. Programmed aerobicactivity

41Total

;60.0 Traffic-light diet 1200kcal (1,2,3)

Aerobic (1), lifestyle (2), orcalisthenic (3) activity;

2. Lifestyle activity isocaloric across groups3. Calisthenics

Epstein et al25 8–12 RS 1. Exercise 23 100 Traffic-light diet 900– Supervised exercise, 3-mile2. No exercise Total 1200 kcal (1,2) walk 33/wk (1)

Epstein et al27 5–8 R 1. Behavior modification 8 100 Traffic-light diet 900– Lifestyle activity (1,2)2. Education 11 1000 or 1200 kcal (1,2)

Epstein et al6,52 8–12 R, IV 1. Parent overweight 41 NR 1200 kcal (1,2) Lifestyle activity (1,2)2. Parent normal weight Total

Epstein et al49 8–12 R 1. Mastery criteria 44 ;74.4 Traffic-light diet from Lifestyle activity (1,2)2. No mastery criteria (yoked) Total 900–1800 to 900–1200

kcal (1,2)Epstein et al21 8–12 R 1. Reinforced for 1activity 61 73.0 Traffic-light diet 1000– 1Activity (1), 2sedentary

2. Reinforced for 2sedentarybehavior

Total 1200 kcal (1,2,3) (2), or both 1activity and2sedentary (3)

3. Reinforced for both1activity 1 2sedentary

Figueroa-Colon 7–17 RS? 1. PSMF fast diet 10 57.9 10 Wk of 50% protein, 20 Min supervised aerobicet al33 2. Hypocaloric balanced diet 9 600–800 kcal (1); 10

wk of hypocaloricbalanced diet (800–1000 kcal) (2)

activity (10 min @ 70%maximal heart rate);lifestyle activity(Cooper’s aerobic pointsystem) (1,2)

Flodmark et al24 10–11 R 1. Family therapy 44 52.3 Nutrition information Exercise information (1,2)2. No family therapy Total 1500–1700 kcal, ,30%

fat kcal (1,2)


per week.14–16 In these studies, diet plus exerciseprograms were associated with better physiologicoutcomes than were the no-treatment control groups.

A third dietary approach is the traffic-light diet,which is used for preschool17 and preadolescent13,18–30

children. The traffic-light diet is a structured eatingplan (900 to 1300 kcal) used to guide participants’eating patterns to meet age recommendations of thebasic four food groups, and now the food pyramid,thereby increasing the nutrient density of the diet.The traffic-light diet groups foods into categories:green foods (go) may be consumed in unlimitedquantities; yellow foods (caution) have average nu-tritional value for the foods within their food group;and red foods (stop) provide less nutrient density percalorie because of high fat or simple carbohydratecontent.

Most interventions using the traffic-light diet as partof a comprehensive treatment have produced a signif-icant decrease in obesity18–21,23,25–30 in preadolescent chil-dren. The traffic-light diet is associated with an im-

provement in nutrient density for protein, calcium,iron, vitamin A, thiamine, and riboflavin and a de-crease in nutrient density for fat in preadolescents.18

Significant changes in eating patterns have been re-ported when comprehensive obesity treatment hasbeen combined with the traffic-light diet.29,30 Reduc-tions in “red foods” have been observed after treat-ment,23,29,30 with significant associations betweenchanges in intake of “red food” and weight loss29 ordecrease in percent of overweight.30 Finally, obese chil-dren of elementary-school age who were treated withthe traffic-light diet also showed a greater decrease inrated palatability for high-fat/low-sugar, low-fat/high-sugar, and high-fat/high-sugar foods and a greaterincrease in rated palatability for low-fat/low-sugarfoods than did comparable lean children who were nottreated.31 In addition to short-term effects, long-termobesity changes extending from 5 to 10 years afterinitiation of treatment have been observed with thetraffic-light diet in combination with behavioral, exer-cise, and familial components.5–7,32

TABLE 1. Continued

Source Age(y)

GroupAssignment

Between-group Variable N % Girls Diet Exercise

Graves et al23 6–12 RS 1. Parent problem-solving 40 NR Traffic-light diet (1,2,3) Physical activity (1,2,3)2. No parent problem-solving Total3. Education only

Gropper and 6–12 R 1. Fiber supplements (15 g) 8 62.5 ADA exchange, 500 kcal NoneAcosta12 2. Placebo less than baseline

Within-subject design, orderrandomized

intake (1,2)

Hills and Parker13 Prpb R 1. Exercise 10 NR Nutrition information 60 Min, 13/wk supervised2. No exercise 10 (1,2) aerobics; 20 min, 3–43/

wk home aerobics (1)Israel et al48 8–12 R 1. Parent training 12 69.7 Nutrition information Exercise information (1,2)

2. No parent training 12 (1,2)3. Control 9

Israel et al53 8–13 RB 1. Enhanced child self-regulation

16 NR Nutrition information(1,2)


2. No enhanced child self-regulation

18

Kirschenbaumet al54

9–13 RB 1. Mother and child seentogether

16 ;77.3 Nutrition information(1,2)


2. Child seen alone 15Mellin et al79 12–18 R 1. Treatment 37 78.8 Nutrition information (1) Supervised exercise (1)

2. Control 29Mendonca and

Brehm588–15 R 1. Perception of treatment

choice18

Total;40.0 Not described (1,2) Not described (1,2)

2. No perception of treatmentchoice

Rocchini et al15 10–17 R 1. Exercise 25 ;52.8 Calorie exchange 40 Min, 33/wk supervised2. No exercise 26 program for 21 lb/ aerobic activity @ 70–75%3. Control 22 wk (1,2) maximal heart rate (1)

Senediak andSpence22

6–13 R 1. Rapid schedule behaviormodification

12 ;33.3 Traffic-light diet, foodexchange (1,2)

30 Min, 33/wk aerobicactivity, lifestyle activity

2. Gradual schedule behaviormodification

12 (1,2)

3. Attention control 114. Control 10

Wadden et al57 12–16 RS 1. Mother and child seenseparately

14 100 1000–1500 kcal, low-fatfoods (1,2,3)

Lifestyle activity (1,2,3)

2. Mother and child seentogether

14

3. Child seen alone 19Wheeler and 2–10 R 1. Behavior modification 14 52.5 Not described (1) Not described (1)

Hess43 2. Control 14

N indicates number of subjects entering treatment (per group if available); R, subjects were randomly assigned; RS, subjects were stratifiedon key variables and then randomly assigned; NR, not reported, IV, individual variable as between-group variable; ?, unsure of degreeof randomization; Prpb, prepubertal; RB, subjects were randomized to groups balanced on key variables.


TABLE 2. Outcomes of Child Obesity Treatment Studies on Clinical Samples

Source OW,%*

BMI* BW,lb*

BF,%*

Rx,mo

OW,%†

BMI† BW,lb†

BF,%†

FU,mo

OW,%†

BMI† BW,lb†

BF,%†

Results

Amador et al11 .97 0–6 0–12 BW,Rx,FU:1.2%ile 1 218.8a 1 214.9a (analyzed by gender)

2 29.6b 2 24.8b

Aragona et al50 0–3 0–11 OW,BMI,Rx:1,2.3(29.6) (21.7) 105.6 1 (215.7) (22.4) 211.3a 1 (211.8) (21.2) 20.7 OW,BMI,FU:15253(34.1) (22.8) 104.6 2 (214.2) (22.2) 29.6a 2 (27.1) (20.3) 17.3(41.7) (23.1) 99.3 3 (21.8) (20.2) 10.9b 3 (22.5) (10.6) 15.5

Bacon and .97 0–2 lb/mo None BW,Rx:152Lowrey41 %ile 1 24.3

2 21.9Becque et al14 0–5 None BW,BF,Rx:15253

149.4 38.3 1 23.5 23.0169.8 44.0 2 20.9 23.5151.1 39.8 3 17.0 10.7

Brownell et al56 0–4 0–12 OW,Rx,FU:1.2,359.9 45.5 183.9 1 217.1a 24.7a 218.5a 1 220.5a 24.6a 216.9a BMI,BW,Rx:1.350.4 42.4 177.1 2 27.0b 23.0 211.7 2 25.5b 20.1b 16.4b BMI,BW,FU:1.2,357.4 42.0 178.4 3 26.8b 22.0b 27.3b 3 26.0b 20.1b 17.0b

Coates et al51 0–4 0–10 OW,Rx,FU:1.2,3,437.3 1 212.0‡ 1 28.0‡39.4 2 25.2 2 110.437.3 3 26.2 3 22.446.1 4 25.0 4 15.8

Coates et al55 0–5 0–18 OW,Rx,FU:15232.5 1 28.6 1 28.4‡30.9 2 25.1 2 28.2‡

Duffy and 0–2 0–8 OW,Rx,FU:152Spence30 45.5 125.7 1 27.8‡ 23.1 1 28.9‡ 17.1

51.5 122.2 2 29.1‡ 23.1 2 29.2‡ 13.6Epstein et al20 0–5 None OW,Rx:1.2

68.4 129.1 1 217.5a 28.660.9 135.4 2 26.4b 23.4

Epstein et al6,29 0–8 0–120 OW,Rx:1525339.0 24.1 106.4 1 216.0‡ 1 215.3a OW,FU:1.341.2 25.0 117.9 2 217.0‡ 2 23.045.4 25.1 119.8 3 219.0‡ 3 17.6b

Epstein et al19 0–6 0–17 OW,BMI,Rx:152535437.2 24.1 110.1 1 210.3‡ 21.3‡ 1 10.1a 11.2a OW,BMI,FU:2,4.1,340.5 24.9 116.6 2 219.0‡ 23.0‡ 2 213.8b 21.5b

38.7 25.3 126.4 3 213.9‡ 22.1‡ 3 29.7a 20.7a

34.0 24.0 115.1 4 213.2‡ 22.0‡ 4 211.2b 21.0b

Epstein et al6,28 0–6 0–120 OW,Rx:1,2.344.0 1 216.0a 1 28.4 OW,FU:15244.0 2 221.0a 2 210.044.0 3 13.0b

Epstein et al6,26 0–12 0–120 OW,BW,Rx:1525347.8 123.9 1 216.3‡ 12.4 1 210.9a OW,FU:1,2.348.3 124.0 2 216.1‡ 22.3 2 219.7a

48.0 123.9 3 217.5‡ 22.3 3 112.2b

Epstein et al25 0–12 None OW,Rx:15248.0 118.3 1 225.4‡ 28.5‡ BW,Rx:1.248.1 118.7 2 218.7‡ 23.0

Epstein et al27 41.9 22.8 0–12 None OW,BMI,Rx:1.239.2 22.7 1 226.3a 23.7a

2 211.2b 21.3b

Epstein et al6,52 20–80 0–12 0–120 OW,Rx:2.11 27.7a 1 13.1 OW,FU:1522 216.3b 2 211.1

Epstein et al49 0–12 0–24 OW,Rx:1.260.6 127.4 1 226.5a 1 215.4 OW,FU:15258.8 121.5 2 216.7b 2 210.6

Epstein et al21 51.8 33.2 0–4 0–12 OW,Rx:2.1Total Total 1 213.0a 23.5 1 29.0a 21.2a OW,FU:2.1,3

2 220.0b 25.0 2 218.0b 24.8b BF,Rx:152533 217.0 24.5 3 210.0a 22.2 BF,FU:2.1

Figueroa-Colon 0–15 None OW,Rx:1.2et al33 82.2 33.1 166.1 1 223.3‡ 22.5 12.9 BMI,BW,RX:152

78.3 31.2 148.3 2 220.3 22.7 22.0Flodmark 0–14 0–26 BMI,Rx:1.2

et al24 24.7 1 10.3a 1 11.1 BMI,FU:15225.5 2 10.6b 2 11.6


The protein-sparing modified fast (PSMF),33 amore restrictive diet, has been used to treat severepediatric obesity. The PSMF usually consists of 600 to900 kcal and 1.5 to 2.5 g of high-biological-qualityprotein per kilogram of ideal body weight per day,usually provided as lean meat; vitamin and mineralsupplementation and consumption of at least 1.5 L ofwater per day is encouraged. The PSMF is designedto maximize weight loss, preserve mineral balance,and achieve positive nitrogen balance, thereby con-serving lean muscle mass in growing individuals.PSMF diets are usually of short duration (4 to 12weeks), conducted under close medical supervision,and not commonly used with prepubertal children.Figueroa-Colon and coworkers33 reported an 11.2-kgweight loss in 10 weeks, but at 15 months the de-crease in percent of overweight was similar to the

decrease in percent of overweight achieved by a lessrestrictive dietary prescription.33

These studies show that there are several differentapproaches to improving food choice and reducingcaloric intake in obese children. There are no com-parative studies in which other aspects of treatmentare held constant as the type of dietary recommen-dations are varied. The effect of diet depends on thecontext in which it is presented in treatment; dietaryrecommendations are complemented by strong com-ponents of behavior change to enhance weight con-trol.20 Very little is known about specific componentsof dietary recommendations, such as macronutrientcontent of the diet. Likewise, no studies have con-centrated on the pattern of eating in obese childrenor on the distribution of food in meals and snacks.The degree to which the diet is rigid or flexible

TABLE 2. Continued

Source OW,%*

BMI* BW,lb*

BF,%*

Rx,mo

OW,%†

BMI† BW,lb†

BF,%†

FU,mo

OW,%†

BMI† BW,lb†

BF,%†

Results

Graves et al23 0–5 0–8 OW,BW, Rx,FU:1.2,353.0 128.9 1 224.3a 211.2a 1 224.5a 29.1a

56.3 128.9 2 213.1b 23.9b 2 210.2b 20.5b

51.8 115.8 3 210.3b 20.5b 3 29.5b 12.5b

Gropper and $20 0–2 lb/mo None BW,Rx:152Acosta12 1

220.720.1

Hills andParker13

0–4 None BW,Rx:152

113.5 1 212.1110.4 2 15.7

Israel et al48 0–2 0–12 OW,Rx:2.1.345.9 114.1 1 27.2a 24.8a 1 210.2a 111.5 OW,FU:1.253.1 126.1 2 211.6b 25.4a 2 21.3b 111.6 BW,Rx:1,2.356.0 128.0 3 20.9c 12.6b BW,FU:152

Israel et al53 0–7 0–36 OW,Rx,FU:152 (time48.1 1 215.6 1 24.8 effect significant,45.9 2 212.5 2 16.4 contrasts not reported)

Kirschenbaum 0–2 0–14 OW,Rx,FU:152et al54 58.1 122.8 1 26.3‡ 1 27.6‡

39.0 115.1 2 27.1‡ 2 26.2‡Mellin et al79 0–3 0–15 OW,Rx,FU:1.2

36.5 174.2 1 25.9‡ 26.8 1 29.9‡ 28.529.5 169.3 2 20.3 10.3 2 20.1 12.8

Mendonca and 0–3 0–9 BW,Rx:1.2Brehm58 47.4 127.7 1 27.6a 1 21.0 BW,FU:152

49.1 135.7 2 24.3b 2 15.8Rocchini et al15 0–5 None BW,BF,Rx:1,2.3

158.4 41.0 1 25.3a 26.0a

160.6 43.0 2 25.5a 24.0a

160.6 41.0 3 18.8b 11.0b

Senediak and Varied 0–7 OW,BW,Rx:2.1.3,4Spence22 34.6 114.2 1 0–1 25.3a 23.7a 1 213.0a 22.4a OW,BW,FU:1,2.3

34.9 110.1 2 0–4 213.6b 27.9b 2 219.2a 26.1a

41.7 105.2 3 0–1 21.4c 20.7c 3 25.9b 10.6b

37.6 103.6 4 0–1 12.3c 11.7c

Wadden et al57 0–10 None BMI,BW,Rx:1525336.7 223.1 40.4 1 17.732.8 191.6 39.5 2 13.735.1 211.0 41.6 3 16.6

Wheeler and 0–7 None OW,Rx:1.2Hess43 40.4 1 24.1a

38.9 2 16.3b

* Baseline.† Change.‡ Value significantly different from baseline value (not reported when between-group differences are reported).OW indicates overweight; BW, body weight; BF, body fat; Rx, treatment period; FU, follow-up period; values with different superscriptletters differ significantly from each other. Treatment period was designated as the period when meetings occurred at least monthly(except for the study by Flodmark, when meetings were more sporadic). Values in parentheses were derived for this table from dataprovided by the authors.


should be studied. If the goal is for children to de-velop a taste for fruit, should candy be removed fromtheir diet, eaten occasionally, or eaten whenever re-quested? There is very little research studying theeffects of dietary change on food preference,34 whichmay be very important if long-term compliance isdesired. Finally, research on food choice and prefer-ence has not kept up with the rapid changes in thefoods available to children, and the influence ofsugar and fat replacements should be studied withregard to the health of the children as well as theiruse in facilitating long-term dietary change.

ExerciseExercise treatments are designed to increase en-

ergy expenditure to produce or increase negativeenergy balance. In most obesity treatments, the majorimpact on energy balance occurs through decreasing

caloric intake,35 but increasing the caloric expendi-ture of obese children may accelerate weight loss andpotentiate maintenance of weight changes.36 How-ever, exercise should be combined with dietary in-tervention. For example, extra physical activity hadno significant effects on weight in the absence ofdietary intervention.37

Several studies have included diet and aerobicexercise. Hills and Parker13 found differences in skin-fold changes but not in weight in preadolescent chil-dren provided weekly supervised exercise sessionsplus a three- or four-time per week home aerobicactivity program versus diet alone. Rocchini and col-leagues14,15 explored the benefits of aerobic activitydesigned to maintain heart rate between 60% and80% of maximal exercise heart rate for 40 to 60 min-utes in 20-week interventions. In both studies, bothdiet and diet plus exercise were equal for weight and

TABLE 3. Characteristics of Child Obesity Treatment Studies on School Samples

Source Age (y) GroupAssignment

Between-group variable

N % Girls Diet Exercise

Blomquist 8–9 EO 1. Exercise 22 0.0 None reported Supervised intenseet al37 2. Control 21 physical activity, 45

min, 23/wk (1)Botvin et al44 7th, 8th Rsch 1. Treatment 119 NR Nutrition information Supervised aerobic

grades 2. Control Total (1) activity (1)Christakis 13–14 RSclass 1. Treatment 55 0.0 Nutrition information Extra exercises 1

et al46 2. Control 35 (1) basketball league (1)Figueroa-Colon 9–13 Rsch 1. Treatment 12 42.1 10 Wk of 50% protein 45 Min, 53/wk (25 min

et al85 2. Control 7 600–800 kcal then16 wk ofhypocaloricbalanced diet (1)

@ 75% maximal heartrate) (1)

Lansky and 12–15 Rclass 1. Behavior 36 54.9 Food exchange, 1200 Supervised aerobicBrownell47 modification kcal (1), nutrition activity (1,2)

2. Education 35 information (2)Lansky and ;13.1 RS 1. Treatment 30 54.5 Food exchange (1) Supervised aerobic

Vance45 (mean) 2. Control 25 activity (1)

N indicates number of subjects entering treatment (per group if available); EO, subjects ranked according to overweight, even/oddassignment to group; Rsch, schools were randomly assigned; NR, not reported; RSclass, classes were stratified on key variables and thenrandomly assigned; Rclass, classes within schools were randomly assigned; RS, subjects were stratified on key variables and thenrandomly assigned.

TABLE 4. Outcomes of Child Obesity Treatment Studies on School Samples

Source OW, %* BMI* BW, lb* Rx, mo OW, %† BMI† BW, lb† Results

Blomquist et al37 0–4 BW,Rx:15291.1 1 11.893.9 2 11.8

Botvin et al44 $20 0–31

NR NR NR OW,BW:1.2(x2 analyses)

2Christakis et al46 0–18 BW,Rx:1.2

41.1 160.9 1 210.9 15.8a OW,Rx:NR39.6 158.4 2 22.3 113.5b

Figueroa-Colon et al85 0–6 OW,BMI,BW,Rx:1.271.8 30.9 153.1 1 224.3‡ 23.8‡ 212.3‡64.0 28.8 136.8 2 20.3 10.2 16.2‡

Lansky and Brownell47 57.4 0–4 OW,BW,Rx:152Total 1 23.0 14.6

2 22.1 14.2Lansky and Vance45 0–3 OW,BW,Rx:1.2

43.9 152.1 1 25.7a 11.3a

47.5 157.8 2 12.4b 17.3b

* Baseline.† Change.‡ Value significantly different from baseline value (not reported when between-group differences are reported).OW indicates overweight; BW, body weight; Rx, treatment period; NR, not reported; values with different superscript letters differsignificantly from each other.


body fat changes, although the diet plus exercisegroups showed better improvements in other phys-iologic parameters. Epstein and colleagues demon-strated a significant effect of exercise implementedinitially in a structured camp setting beyond that ofdiet from baseline to 6 months, with a significantdifference between the groups maintained at 12months.25

Another type of exercise program that has beenstudied is lifestyle exercise, which attempts to in-crease energy expenditure in regular daily activity aswell as during periods of exercise. Lifestyle exercisewas shown to be superior to isocaloric programmedaerobic exercise after 17 months of observation, withthe addition of the traffic-light diet showing no ad-ditional effect on percent of overweight.19 In a fol-low-up study, diet plus lifestyle exercise were signif-icantly more effective than diet plus aerobic exerciseat 2 years,26 but both exercise programs were supe-rior to diet plus calisthenic control at 10 years ofobservation (219.7% and 210.9% vs 112.2%).6 How-ever, in one study, lifestyle exercise was not found tohave additional effects beyond diet over 1-year28 and10-year6 intervals.

The effects of exercise can add to dietary interven-tions to enhance weight loss and improve long-termmaintenance. Preliminary data suggest that lessstructured, more flexible lifestyle exercise may besuperior to more structured and higher intensity aer-obic exercise for weight control.6 As with diet, theeffect of an exercise program varies as a function ofthe other components of treatment. There are manyunanswered questions about the best exercise ap-proach. Are reliable differences observed as a func-tion of exercise intensity, or is the volume of expen-diture most important? Do specific types of exercisehave reliable effects on appetite or dietary compli-ance? Can the new American College of Sports Med-icine/Centers for Disease Control and Preventionguidelines for moderate activity change38 be used inchildhood obesity treatment? Should exercise be pre-sented in structured ways, as in team sports or groupactivities, or should children be taught more generalprinciples for increasing activity? Are there impor-tant age and developmental differences in coordina-tion or motor learning that would provide recom-mendations for different exercise interventions?

Surgical TreatmentsIn an uncontrolled study, jejunoileal bypass sur-

gery was used to treat 11 adolescents who were atleast 100% over their ideal body weight. At 10-yearfollow-up, they had maintained a weight loss rang-ing from 45 to 90 kg.39 However, a significant num-ber of complications were encountered by all partic-ipants, including encephalopathy, nephrolithiasis,cholelithiasis, renal cortical nephropathy, hypopro-teinemia, systemic fatty acid deficiency, and manyother nutritional deficiencies. Three of the 11 adoles-cents had complications from the bypass that weresevere enough that surgery to reverse the operationwas necessary. Thus, although this surgery did pro-duce a large amount of weight loss that was main-tained over 10 years, the side effects of the procedure

were extremely costly, and the investigators did notrecommend additional use of this operation for ad-olescents.39

Drug TreatmentsWith the surge of interest in the use of pharmaco-

therapy to treat adult obesity, the suggestion hasbeen made that the use of anorectic pharmacologicagents, specifically protocols that include fenflura-mine, should be evaluated for treatment of pediatricobesity.40 One controlled study did not find signifi-cant decreases in weight or body fat when fenflura-mine was combined with diet intervention.41 A morerecent study found that treating adolescent obesitywith an energy-restricted diet and d-fenfluramineproduced significant decreases in BMI from baseline,but these differences were not significantly differentthan the control treatment of diet and placebo.42

These studies do not provide confidence in the effi-cacy of current pharmacologic interventions for pe-diatric obesity.

Behavior Change in Pediatric Obesity TreatmentGiven the complexity of behavior change and the

difficulty in adherence to treatment protocols, anunderstanding of factors that influence behaviorchange is needed. The importance of including be-havior therapy in the treatment of childhood obesitywas demonstrated by Epstein and colleagues.20 Theyreported that the addition of the behavioral tech-niques of contingency contracting, self-monitoring ofcaloric intake and weight, praise, and stimulus con-trol to nutrition education significantly improvedreduction in percent of overweight compared withnutrition education alone over a 5-month period(217.5% vs 26.4%). Other investigators have re-ported the superiority of behavioral treatment versusno-treatment controls.43–45 However, in some school-based programs that have targeted the child alone,behavior modification has neither been necessary fortreatment gains46 nor provided treatment gains be-yond that of education.47

In an additional test of the role of behavior therapyin obesity treatment, families who have received in-tense behavioral intervention might be expected todo better than families who receive less behavioraltreatment. Israel and colleagues48 found that at1-year follow-up, children whose parents had partic-ipated in a short course in general behavior manage-ment had significantly better weight control thanchildren in an intervention that focused only onweight reduction. Families that master behaviorchange also might be expected to do better in treat-ment than families provided the same interventionwithout mastery. Epstein and colleagues49 found thatchildren in the mastery group had significantly bet-ter treatment outcome than did children in the non-mastery group at 12 months. At 24 months, childrenin the mastery condition had larger changes in per-cent of overweight than did nonmastery control chil-dren (215.4% vs 210.6%), but the results were nolonger significantly different.

Behavior therapy is designed to teach new behav-iors, and thus research showing differences in learn-


ing as a function of the interval between teachingperiods may be relevant to scheduling treatment ses-sions. Senediak and Spence22 assessed the effects ofrapid (eight sessions in 4 weeks) or gradual (eightsessions over 15 weeks) behavioral treatment versusa nonspecific control condition and a wait-list controlgroup. At 26-week follow-up, the gradual behavioralgroup had more significant weight change than didthe rapid group, who had more weight change thandid the nonspecific control group.

Aragona et al50 compared the effects of responsecost (loss of monetary deposit contingent on meetingattendance, self-monitoring, and child weight-losscriteria) with or without reinforcement versus a no-treatment control group in a small sample of 5- to11-year-old children. At follow-up, no treatment ef-fect was found, although reinforcement appeared toadd a slight benefit for the group in terms of lessweight regain. The goals targeted for reinforcementalso can influence outcome. Coates and coworkers51

contrasted groups in which changes in caloric intakeor weight were targeted and reinforced according tomeeting daily or weekly goals. At 10 months, the bestresults were obtained for adolescents reinforced forweight change on a daily basis.

Obesity has been considered a disorder of self-control, and two studies evaluated whether self-management treatment would enhance child weightcontrol. In both studies, children were providedcomprehensive behavioral family-based programs;neither the study by Epstein and colleagues6,52 northe study of Israel and colleagues53 observed signif-icant effects of child self-control training.

Research on obesity treatment in children and ad-olescents has tested the influence of including mul-tiple family members. In a long-term study, Epsteinand colleagues29 explored the influence of selectionof targeting parents and children by randomizingfamilies to one of three groups for behavior change:1) nonspecific target, 2) child target, and 3) child andparent target. At posttreatment and 2-year follow-up, children in all groups demonstrated similarchanges in percent of overweight; however, at 10-year follow-up, the parent–child group had signifi-cantly better changes in weight status than did thenonspecific control, with the child-only group in be-tween. Despite the intuitive nature of the conclusionthat including parents in treatment improves treat-ment outcome, a number of studies have not foundthat targeting parents adds to treatment effects inpreadolescent54 and adolescent55 samples.

If both parent and child are treated, the structureof the treatment sessions may be important. Twostudies explored the effect of parent participation inadolescent weight change by randomizing familiesinto one of three groups: 1) child-alone group, 2)child and mother treated together, and 3) child andmother treated in separate groups. At 1-year fol-low-up of a white sample,56 children in the child–mother separate group decreased their percent ofoverweight significantly more than did children ineither the child–mother together or the child-alonegroups, which did not differ (220.5% vs 25.5% and26.0%, respectively). However, no differences were

observed in black children by similar methods after10 months.57

Mendonca and Brehm58 evaluated the role of per-ception of choice in therapeutic outcome of behav-ioral obesity treatment. Children who perceived thatthey chose the type of treatment reduced in percentof overweight more than did children in the no-choice control at 12 weeks. No treatment effect wasfound at 9 months, but the small sample size andattrition precluded meaningful interpretation at thistime point.

On the basis of behavioral choice theory,59 Epsteinand colleagues21 evaluated the influence of treat-ments in which children are targeted and reinforcedfor greater activity, less sedentary behavior, or acombined group. The children reinforced for reduc-ing sedentary behavior showed a significantlygreater decrease in percent of overweight comparedwith the children reinforced for increased activity,with the combined group in between; at 1-year fol-low-up, the group reinforced for less sedentary be-havior maintained their weight loss, whereas thecombined group had gained weight and becomeequivalent to the group reinforced for activity.

Graves and colleagues23 found that problem-solv-ing, a component of many behavioral treatments, isnecessary for successful treatment, because familiesprovided behavioral treatments with parental prob-lem-solving had significantly better outcomes at8-month follow-up than families not provided train-ing in problem-solving. However, Duffy andSpence30 found no additional benefit of cognitivetherapy techniques such as targeting monitoring ofnegative thoughts, restructuring of maladaptivethoughts, problem-solving, and self-reinforcement inaddition to behavioral techniques for 7- to 10-year-old and 10- to 13-year-old children at 8-month fol-low-up.

Finally, one study assessed the effects of familytherapy on treatment of obesity in 10- and 11-year-old children.24 At the end of 14 months, neithergroup showed decreases in obesity, but children inthe family therapy group had significantly smallerincreases in BMI than did children in the conven-tional dietary counseling group. This difference dis-appeared 1 year later.

The use of behavior change methods in treatmentis critical to long-term success of obesity treatment,and interventions for behavior change act in concertwith the diet and exercise components to determineoutcome. Some preliminary observations on behav-ior change can be made. The research suggests thateffects are enhanced for programs that include be-havior therapy,20 with outcome related to intensity oftreatment.48 Furthermore, parental involvement canbe an important part of child obesity treatment.29,56

However, there are many unanswered questionsabout how treatment should be implemented. Howimportant is general parent training, or should theparent training be specific to eating and activitychange? How are different parent training programsadapted to the age of the child, because differentprograms are needed for preschoolers, preadoles-cents, and adolescents?60 What do we know about


parent motivation to help children change their di-etary or physical activity practices? What is the roleof specific components of parent training, includingparent use of goal-setting, parent problem-solving,and parent application of rewards for attainedchanges? How many sessions are optimal, and howshould the meetings be scheduled? Because parent-ing approaches need to change as the child growsolder, should regular booster sessions be scheduledto train parents in developmentally appropriatemethods? What is the best combination of parentmodeling and parent support of desired behavior toproduce long-term stable improvement in foodchoice and activity choice? Can school-based educa-tion teach skills needed for obesity treatment andprevention, and can school and community pro-grams be linked to clinical activities to reinforcetreatment efforts?

Side Effects of Treatment

Positive Physiologic Side Effects of TreatmentBesides the overtly obvious goal of a decrease in

percent of overweight or overfat, treatment goals forpediatric obesity should include improvement inother physiologic parameters such as blood pressure,serum lipids, and insulin resistance. Research hasfound that after treatments that produce significantdecreases in body fat, percent of overweight, orweight, adolescents demonstrated significant de-creases in systolic and diastolic blood pres-sures.15,16,33,51,56 A change in serum lipids also hasbeen documented as a positive outcome in the com-prehensive treatment of pediatric obesity.4,42,51,57,61,62

For children, significant reductions in fasting serumcholesterol and triglycerides and significant in-creases in high-density lipoprotein (HDL) serumcholesterol have been found after comprehensivetreatment, with serum HDL levels remaining signif-icantly higher than baseline levels at 5-year follow-up.4,61,62 Epstein et al62 reported that at 5-year follow-up, changes in relative weight and fitness weresignificantly associated with change in HDL levels.Sasaki and colleagues63 found that a long-term aero-bic exercise program, which resulted in significantdecreases in body fat in children, resulted in signif-icant increases in HDL concentrations. In addition,dietary treatments have been associated with signif-icant reductions in fasting serum cholesterol,64,65 trig-lycerides,64 and low-density lipoproteins.64

Other improvements in the metabolic profile ofobese children and adolescents have been docu-mented after treatment. Investigators have reporteddecreases in fasting serum insulin levels after com-prehensive treatment for obese children4 and adoles-cents.16 At 5-year follow-up, insulin levels were notsignificantly different from levels at baseline,4 butoccurrence of hyperinsulinemia had decreased, and asignificant negative correlation was found betweenserum insulin levels at the end of treatment and adecrease in relative weight over the follow-up.4 Hoff-man and coworkers66 reported significant increasesin mean insulin sensitivity when obese childrentreated with an energy-reduced diet showed signifi-

cant reductions in body weight, but Gutin and col-leagues67 found no significant changes in fasting in-sulin after weight change.

Psychological Benefits of Weight ReductionObese children seeking treatment often experience

psychological comorbidity,68–72 and obesity treat-ment may improve their psychological status. Therehas been considerable research on self-esteem73,74 andpsychopathology73 in obese children. There has beensome debate about whether psychological problemscause obesity or are caused by obesity and whetherthe prevalence of psychological problems is greaterin obese than in nonobese children and adoles-cents.73,74

Most of the research in this area has targeted self-esteem. Although self-esteem may change duringsome interventions, this change is not consistentlyassociated with decreases in percent of overweight.For example, in school-based interventions designedfor prepubertal children, increases in self-esteem oc-curred in the absence of weight change,75 and equiv-alent improvements in self-esteem were demon-strated for both experimental and control groups.76

Wadden and colleagues57 found significant increasesin self-esteem and decreases in depression in blackgirls; however, consistent with other studies, no re-lationship was found between self-esteem or depres-sion and changes in weight.77,78 Furthermore, studiesof adolescents have demonstrated equivalent in-creases in self-esteem for both treatment and controlgroups79 as well as no increase in self-esteem duringtreatment.80 Research to date suggests that improve-ment in self-esteem in obese children in treatmentmay be better accounted for by nonspecific treatmenteffects than by improvement in weight status.

Parent and child psychological problems may in-fluence treatment outcome.81 Epstein and col-leagues72 demonstrated that parental distress hadnegative effects on child weight loss during the treat-ment phase, which were mediated by their child’slevel of anxiety and depression. At 2-year follow-up,parental distress affected child weight changethrough effects on the child’s social problems; chil-dren who had greater social problems had less suc-cessful treatment. Obesity treatment also may influ-ence child psychological changes. We are in theprocess of collecting follow-up psychosocial data onobese children treated in a 16-week, family-basedbehavioral treatment program. Preliminary analysesshow an improvement in child behavior checklist82

values at 1 year, with the greatest decrease shown forsocial problems. Significant reductions also havebeen observed for anxiety and depression, with-drawn behavior, attention problems, and somaticcomplaints. The broad-band internalizing and total-problems scales also have shown significant im-provements. More importantly, the improvement inweight status from baseline to 1-year assessment isassociated with improvement in psychosocial func-tioning, with significant positive correlations be-tween weight loss and social problems, somatic com-plaints, and total problems.


Negative Physiologic Effects of TreatmentWith any dietary intervention that reduces energy

and consequential macro- and micronutrient intakeduring a period of growth, there is the potential ofimpairing linear growth. Growth failure has beenreported in the overzealous treatment of hypercho-lesterolemia in children because of inadequate en-ergy, fat, and micronutrient intake,83 and investiga-tors33,84 have reported significant reductions in heightvelocity during interventions with obese children.Amador et al11 found that more restrictive energyintakes (0.17 instead of 0.25 MJ/kg) prescribed tochildren in the early stages of puberty producedsignificantly less height gain in the children during a6-month treatment; there was no significant differ-ence in height gain between the two groups frombaseline to 12-month measures. Also, increases inlean body mass were lower in the more restrictedenergy group than in the less restricted energy intakegroup, both after treatment and from baseline to12-month measures.11 Positive growth velocity zscores in children also have been documented afterthe children were treated for obesity.85

However, these changes in growth velocity mustbe interpreted within the context that obese childrenexperience an earlier growth spurt than their non-obese counterparts;86 thus, obese children may natu-rally undergo a deceleration in height velocity laterin development.8 This hypothesis reconciles the datathat children with earlier growth spurts are not talleras adults than children who mature later.87,88 In ourlong-term, 10-year follow-up studies, we have notfound that the more successful children show lessgrowth,5 and height percentiles showed the expecteddecrease over 10 years, so that children came toresemble their same-sex parent in height. Multiple-regression analysis found that child percent of over-weight change made no contribution to predictingheight change, and that child sex, age, baselineheight, percent of overweight, midparent height, andheight change from baseline to 5 years accounted for95% of the variance in growth that occurred frombaseline to 5-year follow-up.8

Eating DisordersDietary restriction that can be part of weight-con-

trol programs is a defining characteristic of bothanorexia and bulimia nervosa.89 A number of studieshave found caloric restriction to precede binge-eat-ing. For example, in one study, 75% of bulimic pa-tients interviewed reported that the inability to main-tain a low-carbohydrate diet immediately precededbulimic symptoms.90 More importantly, prospectivestudies of adolescent girls suggest that dietary re-striction predates bulimic symptoms,91 with onestudy reporting that adolescent girls who were diet-ing had an eightfold increased risk for being diag-nosed with an eating disorder compared with non-dieters.92

However, few of the obesity treatments describedabove assessed the prevalence of disordered eatingafter obesity intervention, which would require sig-nificant long-term follow-up. Only Epstein and col-

leagues6 report the prevalence of eating disorders in158 individuals who had been treated for obesity. Atthe 10-year follow-up, 4% of the subjects (all female)reported having been treated for eating disorders,which compares favorably with population preva-lence rates #9% in studies that used self-reporteddiagnoses and prevalence rates #5% in studies thatdiagnosed eating disorders by means of structuredinterviews.93 Given that obese adults presenting forobesity treatment have higher rates of eating disor-ders than do community samples of obese adults,94 itwould not have been surprising to have found evenhigher rates of eating disorders than those in com-munity samples of children. Thus, moderate caloricrestriction in a structured behavioral interventionmay not carry the increased risk for development ofdisordered eating found by more drastic caloric re-striction.89 Clearly, more work is needed in this area;however, it would be worthwhile to determinewhether appropriate education and skills trainingduring development in obese pediatric patientsmight prevent eating disorders as well as to under-stand what characteristics of treatment, if any, areassociated with developing eating disorders.

DISCUSSION

Improving Implementation of Pediatric ObesityTreatment

Substantial progress has been made in the devel-opment of treatments for childhood and adolescentobesity. However, most pediatric obesity interven-tions are marked by small changes in relative weightor adiposity and substantial relapse, although thereis some evidence for long-term efficacy.5,6 It is pre-mature to assume that there are standardized treat-ments that are efficacious, and research is needed toimprove treatment outcome and maintenance oftreatment effects. There are many new developmentsin diet, exercise, and behavior change that should beincorporated into treating childhood obesity; how-ever, for brevity, we focus here on behavior change.Additional new findings in eating and exercise be-haviors that can inform treatment development aredescribed by both Birch and Fisher95 and Kohl andHobbs.96

First, given the central role of behavior therapy inthe treatment of child and adolescent obesity, a thor-ough understanding of behavioral principles andtheir use is a necessary component of successfulweight control programs.97–99 It is possible that inter-ventions that claim to have a behavioral componentmay be misusing behavioral techniques or misapply-ing behavioral principles. Stunkard99 notes that be-havior therapy for obesity was and sometime still isseen as a set of tools to micromanage within-mealeating behaviors such as the rate of eating and bitesize. However, there is little empirical evidence thatthese behaviors are associated with weight status orthat changes in these behaviors correlate with weightloss.100 Behavior therapy for obesity has expandedconsiderably from these roots and demands a currentknowledge of the factors that influence energy bal-ance behaviors as well as an awareness of behavioral


principles and their application to changing activityand eating habits if long-term maintenance is to beachieved.

More comprehensive assessments of the process ofbehavior change are needed. There are few studiesthat provide measures of compliance to treatmentoutcome, and these studies may be compromised bythe well-documented problems that have been notedin self-reporting of dietary and activity changes.101,102

If behavior change strategies are to be effective, it iscritical to ensure that people are in fact implementingthese programs as planned and are demonstratingmastery of new skills.49 The mediators of behaviorchange are assumed to be changes in eating andactivity. However, as Baranowski and colleagues103

have pointed out for community health interven-tions, the link between the hypothesized mediatorsand health outcomes is weak. The same could be saidfor obesity treatment, and one solution may be stud-ies that focus on modifying the proposed mediatorsinstead of assuming that changes in the mediatorswill influence outcome. Until eating and activity pat-terns can be modified reliably, investigators mustrely on weak treatments to attempt modification ofbehaviors that are quite resistant to change. Researchalso is needed to assess the short- and long-termeffects of exercise and diet interventions on regula-tion of food intake, body composition, fat distribu-tion, energy metabolism, and substrate use. In thelong term, it is more cost-effective to focus on devel-opment of methods that modify the proposed medi-ators of change reliably and test for changes in pro-posed mediators rigorously than to ignore mediatorsand focus only on weight loss.

Behavioral Economic ResearchAn example of using advances in behavior theory

to inform treatment development is application ofbehavioral choice theory.59,97 Behavioral choice the-ory provides a conceptual and methodologic frame-work for understanding behavioral choice. One im-portant choice for obese children is whether to beactive or sedentary. In general, obese children chooseto be sedentary when they are given the option ofengaging in physical or sedentary activity, becausethey have a more negative perception of physicalactivity104 and find physical activity less reinforcingthan sedentary activity relative to their nonobesepeers.105 Highly reinforcing sedentary activities com-pete with physical activity and decrease children’slevels of physical activity. Behavioral choice theorysuggests that one way to increase physical activity isto increase the relative reinforcing value of physicalactivity relative to sedentary activity, either by in-creasing the reinforcing value of physical activity orby decreasing the reinforcing value of sedentary ac-tivity. This reframing of physical activity as a choicebetween physical and sedentary activity requiresconsideration of sedentary activity when interven-tions to increase physical activity are being designed.This is particularly important in obese populations,because participation in sedentary activity has beenfound to be cross-sectionally and longitudinally re-lated to childhood and adult obesity.106,107

Both laboratory research and clinical outcome re-search show that positively reinforcing reductions inhigh-preference sedentary activities increases thephysical activity of obese children to a magnitudesimilar to that obtained by positively reinforcing in-creases in physical activity.21,108,109 In clinical research,no benefits have been observed for targeting thecombination of reduced sedentary behavior and in-creased activity.21 Children who are reinforced fordecreasing time spent in some of their high-prefer-ence sedentary activities reallocate time from thesebehaviors to engage in both lower preference seden-tary activities and physical activity.108,109

One advantage of reinforcing reductions in seden-tary activity to increase physical activity is thegreater choice and control afforded by this contin-gency compared with contingencies that reinforceincreases directly in physical activity. Maintainingchoice and control is important in establishing rein-forcing value,110 and enhancing these variables mayminimize the sense of deprivation associated withreducing common sedentary behaviors. Deprivationis a powerful way to increase the reinforcing value ofan activity or other commodity,111 and interventionsthat minimize the sense of deprivation are likely tohave more long-term behavior change and decreasedlikelihood of relapse.

Children reinforced for decreasing sedentary ac-tivities choose to replace these activities with physi-cal activity or nontargeted sedentary activities, thusmaintaining their control over the choice betweenphysical and sedentary activity. Additional supportfor the need to maintain children’s choice and controlabout activity choice comes from examination of var-ious techniques to decrease sedentary behavior.There is some evidence that different strategies forreducing sedentary activity have different effects onphysical activity. In a laboratory study that com-pared the effects of positive reinforcement, punish-ment, and restriction to reduce sedentary activity,each strategy was found to be equally effective inreducing time spent in the targeted sedentary activ-ities. However, children reinforced for decreasingsedentary activity were more physically active than anoncontingent control group, whereas children whohad their targeted sedentary activities removed fromtheir environment (restriction) did not engage in sig-nificantly more physical activity than the noncontin-gent control group. Children in the restriction groupincreased their liking for the targeted sedentary ac-tivities, whereas children reinforced for decreases inthe targeted sedentary activities decreased their lik-ing for these activities over time.109

Reinforcing a reduction in access to sedentary be-haviors that compete with being active representsone method for modifying the environment that maylead to obesity. When obesity is being treated, therole of environmental and stimulus control of activebehavior should be considered at multiple levels. Atthe individual level, families may want to increasethe cues and opportunities to be active and reducecues to be sedentary. It is obvious that the largerenvironment also influences eating and activity.Weather may influence activity, as may having safe


places to play. Children who live in an environmentwhere they can walk or bicycle to school may expendmany more calories than children who are driven orbused to school. Access to after-school and weekendactive play or sports teams also can enhance activitylevels.

Memory ResearchAnother research area that has considerable rele-

vance for obesity treatment is learning and memoryresearch designed to understand response extinctionand recovery. One of the most common and leastunderstood aspects of obesity treatment is relapse, aubiquitous occurrence in obesity treatment. Mosttherapists assume that extinction of a behavior orreplacement of an unwanted behavior with a newbehavior removes a response from the subject’s rep-ertoire. The appearance of relapse therefore impliesthat the treatment was ineffective.

Bouton and colleagues have begun to understandprocesses that account for recovery of response afterextinction. In a series of animal studies, Bouton hasexplored the influence of stimulus cues presentedwhen a response is learned (context of the learning)on extinction of that response. Bouton and col-leagues112 have demonstrated that context influencesrecovery of a response that is not reinforced and thathas been reduced in frequency. Many treatment ap-proaches attempt to replace a response with a newresponse pattern by counterconditioning. If the samecontext is present when both behaviors are learned,spontaneous recovery of the initial response is ex-pected, which demonstrates that learning the newresponse pattern does not destroy the initial learn-ing.113 If the new behavior is learned in a differentcontext from the initial behavior, presenting the con-text for the more adaptive behavior will retard re-lapse to the initial set of behaviors.114 Furthermore,the context during extinction appears to be moreimportant than the context for initial learning. Whatis learned during response extinction includes mem-ories of being reinforced for a response during learn-ing as well as memories of not being reinforced forthe response when treatment occurs. The responseobserved depends on which memory is retrieved.115

This model has many clinical implications fortreatments that attempt to influence weight status bydevelopment of positive behaviors (eg, exercise) andextinction of problematic behaviors (eg, overcon-sumption of high-fat foods). Ignoring problematicbehaviors or reinforcing competing behaviors doesnot destroy the original reinforcement associationsthat maintained the problematic behavior.116 Chang-ing contingencies for the child’s behavior has addedan association to the child’s memory and behavioralrepertoire (ie, candy request–no candy) rather thanreplaced the original association (ie, candy request–candy).

Context and memory may be very important pro-cesses for long-term behavioral regulation. Treat-ments designed to make permanent changes in be-havior must take into account that acquiredassociations may not disappear completely from therepertoire of associations that produce behavior. Re-

lapse prevention, such as expanding the contexts inwhich extinction conditioning takes place or trainingchildren to become aware of extinction cues, mayhelp increase the maintenance of treatment gains foroverweight children.

Individualization of TreatmentTreating obesity as a homogeneous condition, with

all participants receiving a common intervention,might contribute to the mixed treatment outcomesthat are reported. Research has identified many eti-ologic factors of obesity, including genetic, meta-bolic, biochemical, environmental, psychological,and physiologic variables.117,118 However, at the indi-vidual level, it is probably rare for all of these factorsto be involved in development and maintenance ofobesity; consequently, the etiology and maintenanceof obesity can be very different from one individualto another. Conceptualization of obesity in this waysuggests that interventions need to be heterogeneousand individualized; treatments included in interven-tions should depend on factors that are believed tobe involved with development and maintenance ofobesity for that individual.117–120 Research is neededto identify which treatment components are success-ful with different etiologic and maintenance factorsof obesity.

Excess intake could be attributable to individualdifferences in food craving or satiety. The most com-monly craved foods in obese individuals are choco-late, cakes, cookies, ice cream, and other desserts,121

which are composed predominantly of simple carbo-hydrates and fat. Drewnowski121 suggests that foodcravings for fat and sugar may involve endogenousopioid peptides; animal research has linked fat andsucrose intake to the endogenous opioid system, in-dicating a potential mechanism for food cravings. Ifthe mechanism for food cravings can be identified,treatment would involve identifying the critical di-mension of the food (eg, nutrient, sensory cue, etc)that is the source of the craving and substitutingmore appropriate foods that also contain the criticaldimension necessary to evoke the neurochemicalsystem linked with the food craving.

Barkeling and colleagues122 observed that obesechildren ate lunch significantly faster and did notdecelerate their eating toward the end of the meal asmuch as do children of normal weight. Because de-celeration of food intake toward the end of a meal isconsidered to be a sign of satiation, this lack ofdeceleration might indicate a deficient satiation sig-nal or an impaired behavioral response to satiationsignals.122 For obese children who show impairedsatiation responses, interventions that focus onstrengthening satiation signals through conditioningsatiation to orosensory cues or strengthening behav-ioral responses to satiation signals by training indi-viduals to focus on internal cues (learning to identifywhen hungry or full and to show appropriate behav-ioral responses such as beginning or terminating eat-ing) might be appropriate.123,124

Interventions designed to promote attending tointernal cues are consistent with a nondieting ap-proach that has been proposed as an alternative to


current treatment paradigms, which usually empha-size some amount of energy restriction combinedwith increased activity.125,126 If subjects could reliablyattend to internal cues that would help them matchenergy intake to energy requirements for a healthybody weight, there might be a reduction in restrainedeating and preoccupation with food as a result ofperceived food deprivation.124 Training in food reg-ulation would reduce the parent’s perceived need tocontrol the child’s intake, which Johnson and Birch127

have found to interfere with a child’s ability to self-regulate energy intake. In this paradigm, parentalresponsibilities might be different than in a usualweight-control program. Parents would be responsi-ble for providing nutritious and pleasing foods atpredictable and comfortable times during the day.Children would be allowed to choose from the avail-able foods; they also would choose how much to eat,so that they can respond to their internal signals ofhunger, appetite, and satiety.125

Another important factor that may introduce het-erogeneity to treatment effects are racial and ethnicfactors. Although we are not aware of any researchthat has contrasted treatment response of differentethnic groups in the same study, investigators fromone research group have shown a different pattern ofresponse to a common treatment in white56 andblack57 teenagers. Obese black and white girls differin body composition,128 and resting energy expendi-ture is different in obese black and white women.129

These physiologic differences, in addition to the in-creased prevalence of obesity in the black communi-ty,130 suggest that additional information is needed tounderstand etiologic differences as well as differ-ences in the response to treatment in different racialand ethnic populations.

If there is heterogeneity in causes for obesity andfactors that maintain the obese state, providing astandard group intervention, as is currently usuallydone, would be sufficient only for those individualswho receive the appropriate intervention by chance.With better assessment techniques, more homoge-neous groupings of obese children and adolescentscan be achieved, with interventions targeted appro-priately to these groups.

Obesity and Comorbid ConditionsAs reviewed by Dietz,131 obesity is associated with

a variety of comorbid medical conditions, includingsome that are critical in disease development, such ashypertension, hyperlipidemia, and insulin resis-tance. Treatment of obesity in these cases should beuseful in preventing morbidity and mortality associ-ated with obesity and comorbid diseases. However,there is a series of less prevalent conditions that areassociated with morbidity and mortality duringchildhood and adolescence that may require moreaggressive treatment, such as pseudotumor cerebriand sleep apnea.

There is no research designed specifically to studyobesity in association with comorbid conditions. Re-search in these areas is needed to evaluate whetherthese children are responsive to the same interven-tions as children without these often medically seri-

ous conditions. If these children do not respond tothe common dietary, activity, and behavior changecomponents of treatment, more aggressive treat-ments may be justified.

If individual difference variables such as the de-gree of obesity or comorbid medical or psychiatricconditions influence treatment outcome, it may beworthwhile to consider a graded approach to treat-ment. Children or adolescents who are more obese orhave comorbid conditions that require immediatetreatment may require more aggressive treatment,such as pharmacotherapy or PSMF. In these cases, itis important to monitor the comorbid condition care-fully and evaluate whether changes in obesity in factdo result in changes in comorbid status. If weightcontrol does not reduce comorbid problems, thenother interventions in addition to weight controlmust be considered.

One unique aspect of treating children with co-morbid medical problems may be motivation fortreatment. Most treatment outcome research thatwas reviewed used subjects who were motivated toparticipate in an obesity treatment study. However,many obese youth in treatment because of a comor-bid condition may not be motivated to participate, orif they are motivated, their parents may not be mo-tivated to take part in treatment. It may be importantto be able to assess motivation for change, takingadvantage of concepts in the stages of change mod-el.132 If it is determined that there is not sufficientmotivation for behavior change, motivational inter-viewing133 may be a useful technique for enhancingmotivation to lose weight.

Integrate Basic and Clinical ResearchAdvances in basic science should be integrated

into treatment development. Rapid increases in ourknowledge of the genetic and molecular basis forobesity provide a stimulus for development of newpharmacologic approaches to obesity treatment. Un-derstanding the influence of genes on eating behav-ior and obesity also is important for understandingthe complementary role of the environment in obe-sity as well as for identifying specific etiologic factorsthat may lead to some individuals being at high risk.For example, identifying the genetic basis for behav-ioral phenotypes, such as impairments in satiety,may be important in developing more focused treat-ments that target specific behaviors instead of theusual broad-based group methods for treatment.

As reviewed by Birch and Fisher95and Kohl andHobbs,96 basic behavioral sciences also are providingnew insights into processes that may influence eatingand exercise behavior. It is important for childhoodobesity researchers to be familiar with new develop-ments in child development, learning, and psycho-biology, to name just a few areas that can makeimportant contributions to treatment development.The basic behavioral sciences often are ignored byclinical researchers;97 consequently, new develop-ments that can inform better behavior change strat-egies are not incorporated into treatment programs.The pharmaceutics industry takes rapid advantageof new genetic and metabolic findings for drug de-


velopment and testing, but there is no industry thatuses behavioral scientists to read basic behavioralscience and neuroscience for treatment development.This responsibility falls to the individual investigatoror investigative team to translate basic science intoclinical interventions.

Investigators focusing on pharmacologic interven-tions generally do not reduce attempts to find newpharmacologic treatments when current treatmentsare not effective. Rather, the search for new drugsgoes on, and often the next generation of drug treat-ment provides better results. Hopefully, we will seethe same type of progress made with behavioral andpsychosocial interventions for obesity, and investi-gators will persist in developing more powerfultreatments.

CONCLUSIONS

Future DirectionsThis review provides an overview of contempo-

rary research in treating childhood obesity. The focusis on clinical interventions, although some school-based treatments are discussed. There are many op-portunities for treatment development outside theclinic setting—in school-based and community-based programs—although initial research onschool-based programs has not been particularlypromising.134 In addition, prevention of obesity, in-stead of treatment of the problem after it has devel-oped, may be important to public health. There isvery little research available on prevention,135 andprevention approaches are needed at the populationlevel as well as in targeted high-risk populations.Although progress has been made in treating obesechildren, research is needed that makes contact withdevelopments in nutrition, exercise, and behavioralscience to improve long-term weight regulation,maximizing the positive benefits of weight regula-tion and minimizing the negative side effects of treat-ing obese children.

ACKNOWLEDGMENTSThis work was supported in part by National Institutes of

Health Grants HD 25997, HD 20289, and HD 34284.

REFERENCES1. Brownell KD. Obesity: understanding and treating a serious prevalent,

and refractory disorder. J Consult Clin Psychol. 1982;50:820–8402. Brownell KD, Wadden TA. Behavior therapy for obesity: modern

approaches and better results. In: Brownell KD, Foreyt JP, eds. Hand-book of Eating Disorders. New York, NY: Basic Books; 1986:180–197

3. Wadden TA. The treatment of obesity: an overview. In: Stunkard AJ,Wadden TA, eds. Obesity: Theory and Therapy. New York, NY: RavenPress; 1992:197–218

4. Knip M, Nuutinen O. Long-term effects of weight reduction on serumlipids and plasma insulin in obese children. Am J Clin Nutr. 1993;54:490–493

5. Epstein LH, Valoski A, Wing RR, McCurley J. Ten-year follow-up ofbehavioral family-based treatment for obese children. JAMA. 1990;264:2519–2523

6. Epstein LH, Valoski AM, Wing RR, McCurley J. Ten year outcomes ofbehavioral family-based treatment for childhood obesity. Health Psy-chol. 1994;13:373–383

7. Epstein LH, Valoski AM, Kalarchian MA, McCurley J. Do children loseand maintain weight easier than adults: a comparison of child andparent weight changes from six months to ten years. Obes Res. 1995;3:411–417

8. Epstein LH, Valoski A, McCurley J. Effect of weight loss by obesechildren on long-term growth. Am J Dis Child. 1993;147:1076–1080

9. Rees JM. Management of obesity in adolescence. Med Clin North Am.1990;74:1275–1292

10. Coates TJ, Thorensen CE. Treating obesity in children and adolescents:a review. Am J Public Health. 1978;68:143–151

11. Amador M, Ramos LT, Morono M, Hermelo MP. Growth rate reduc-tion during energy restriction in obese adolescents. Exp Clin Endocrinol.1990;96:73–82

12. Gropper SS, Acosta PB. The therapeutic effect of fiber in treatingobesity. J Am Coll Nutr. 1987;6:533–535

13. Hills AP, Parker AW. Obesity management via diet and exerciseintervention. Child Care Health Dev. 1988;14:409–416

14. Becque MD, Katch VL, Rocchini AP, Marks CR, Moorehead C. Coro-nary risk incidence of obese adolescents: reduction by exercise plusdiet intervention. Pediatrics. 1988;81:605–612

15. Rocchini AP, Katch V, Anderson J, et al. Blood pressure in obeseadolescents: effect of weight loss. Pediatrics. 1988;82:16–23

16. Rocchini AP, Katch V, Schork A, Kelch RP. Insulin and blood pressureduring weight loss in obese adolescents. Hypertension. 1987;10:267–273

17. Epstein LH, Valoski A, Koeske R, Wing RR. Family-based behavioralweight control in obese young children. J Am Diet Assoc. 1986;86:481–484

18. Valoski A, Epstein LH. Nutrient intake of obese children in a family-based behavioral weight control program. Int J Obes. 1990;14:667–677

19. Epstein LH, Wing RR, Koeske R, Ossip DJ, Beck S. A comparison oflifestyle change and programmed aerobic exercise on weight andfitness changes in obese children. Behav Ther. 1982;13:651–665

20. Epstein LH, Wing RR, Steranchak L, Dickson B, Michelson J. Compar-ison of family based behavior modification and nutrition education forchildhood obesity. J Pediatr Psychol. 1980;5:25–36

21. Epstein LH, Valoski AM, Vara LS, et al. Effects of decreasing sedentarybehavior and increasing activity on weight change in obese children.Health Psychol. 1995;14:109–115

22. Senediak C, Spence SH. Rapid versus gradual scheduling of therapeu-tic contact in a family based behavioural weight control programmefor children. Behav Psychother. 1985;13:265–287

23. Graves T, Meyers AW, Clark L. An evaluation of problem-solvingtraining in the behavioral treatment of childhood obesity. J Consult ClinPsychol. 1988;56:246–250

24. Flodmark C, Ohlsson T, Ryden O, Sveger T. Prevention of progressionto severe obesity in a group of obese schoolchildren treated withfamily therapy. Pediatrics. 1993;91:880–884

25. Epstein LH, Wing RR, Penner BC, Kress MJ. Effect of diet and con-trolled exercise on weight loss in obese children. J Pediatr. 1985;107:358–361

26. Epstein LH, Wing RR, Koeske R, Valoski A. A comparison of lifestyleexercise, aerobic exercise and calisthenics on weight loss in obesechildren. Behav Ther. 1985;16:345–356

27. Epstein LH, Wing RR, Woodall K, Penner BC, Kress MJ, Koeske R.Effects of family-based behavioral treatment on obese 5–8 year oldchildren. Behav Ther. 1985;16:205–212

28. Epstein LH, Wing RR, Koeske R, Valoski A. Effects of diet plus exerciseon weight change in parents and children. J Consult Clin Psychol.1984;52:429–437

29. Epstein LH, Wing RR, Koeske R, Andrasik F, Ossip DJ. Child andparent weight loss in family-based behavior modification programs. JConsult Clin Psychol. 1981;49:674–685

30. Duffy G, Spence SH. The effectiveness of cognitive self-management asan adjunct to a behavioural intervention for childhood obesity: aresearch note. J Child Psychol Psychiatry. 1993;34:1043–1050

31. Epstein LH, Valoski A, Wing RR, et al. Perception of eating andexercise in children as a function of child and parent weight. Appetite.1989;12:105–118

32. Epstein LH, McCurley J, Wing RR, Valoski A. Five year follow-up offamily-based behavioral treatments for childhood obesity. J ConsultClin Psychol. 1990;58:661–664

33. Figueroa-Colon R, von Almen TK, Franklin FA, Schuftan C, SuskindRM. Comparison of two hypocaloric diets in obese children. Am J DisChild. 1993;147:160–166

34. Epstein LH, Wing RR, Valoski A, Penner B. Stability of food prefer-ences in 8–12 year old children and their parents during weightcontrol. Behav Modif. 1987;11:87–101

35. Wing RR, Jeffery RW. Outpatient treatments of obesity: a comparisonof methodology and results. Int J Obes. 1979;3:261–279

36. Brownell KD, Wadden TA. The heterogeneity of obesity: fitting treat-ments to individuals. Behav Ther. 1991;22:153–177

37. Blomquist B, Borjeson M, Larsson Y, Persson B, Sterky G. The effect of


physical activity on the body measurements and work capacity ofoverweight boys. Acta Paediatr Scand. 1965;54:566–572

38. Pate RR, Pratt M, Blair SN, et al. Physical activity and public health: arecommendation from the Centers for Disease Control and Preventionand the American College of Sports Medicine. JAMA. 1995;273:402–407

39. Silber T, Randolph J, Robbins S. Long-term morbidity and mortality inmorbidly obese adolescents after jejunoileal bypass. J Pediatr. 1986;108:318–322

40. Medeiros-Neto GA. Should drugs be used for treating obese children?Int J Obes Relat Metab Disord. 1993;17:363

41. Bacon GE, Lowrey GH. A clinical trial of fenfluramine in obese chil-dren. Curr Ther Res Clin Exp. 1967;9:626–630

42. Grugni G, Guzzaloni G, Moro D, et al. Study of the effects of d-fenfluramine on juvenile obesity. Int J Obes Relat Metab Disord. 1993;17(suppl 2):71

43. Wheeler ME, Hess KW. Treatment of juvenile obesity by successiveapproximation control of eating. J Behav Ther Exp Psychiatry. 1976;7:235–241

44. Botvin GJ, Cantlon A, Carter BJ, Williams CL. Reducing adolescentobesity through a school health program. J Pediatr. 1979;95:1060–1062

45. Lansky D, Vance MA. School-based intervention for adolescentobesity: analysis of treatment, randomly selected control, and self-selected subjects. J Consult Clin Psychol. 1983;51:147–148

46. Christakis G, Sajecki S, Hillman RW, Miller E, Blumenthal S, Archer M.Effect of a combined nutrition education and physical fitness programon the weight status of obese high school boys. Fed Proc. 1966;25:15–19

47. Lansky D, Brownell KD. Comparison of school-based treatments foradolescent obesity. J School Health. 1982;8:384–387

48. Israel AC, Stolmaker L, Andrian CA. The effects of training parents ingeneral child management skills on a behavioral weight loss programfor children. Behav Ther. 1985;16:169–180

49. Epstein LH, McKenzie SJ, Valoski A, Klein KR, Wing RR. Effects ofmastery criteria and contingent reinforcement for family-based childweight control. Addict Behav. 1994;19:135–145

50. Aragona J, Cassady J, Drabman RS. Treating overweight childrenthrough parental training and contingency contracting. J Appl BehavAnal. 1975;8:269–278

51. Coates TJ, Jeffery RW, Slinkard LA, Killen JD, Danaher BG. Frequencyof contact and monetary reward in weight loss, lipid change and bloodpressure reduction with adolescents. Behav Ther. 1982;13:175–185

52. Epstein LH, Wing RR, Koeske R, Valoski A. Effect of parent weight onweight loss in obese children. J Consult Clin Psychol. 1986;54:400–401

53. Israel AC, Guile CA, Baker JE, Silverman WK. An evaluation of en-hanced self-regulation training in the treatment of childhood obesity.J Pediatr Psychol. 1994;19:737–749

54. Kirschenbaum DS, Harris ES, Tomarken AJ. Effects of parental in-volvement in behavioral weight loss therapy for preadolescents. BehavTher. 1984;15:485–500

55. Coates TJ, Killen JD, Slinkard LA. Parent participation in a treatmentprogram for overweight adolescents. Int J Eat Disord. 1982;1:37–48

56. Brownell KD, Kelman SH, Stunkard AJ. Treatment of obese childrenwith and without their mothers: changes in weight and blood pressure.Pediatrics. 1983;71:515–523

57. Wadden TA, Stunkard AJ, Rich L, Rubin CJ, Sweidel G, McKinney S.Obesity in black adolescent girls: a controlled clinical trial of treatmentby diet, behavior modification, and parental support. Pediatrics. 1990;85:345–352

58. Mendonca PJ, Brehm SS. Effects of choice on behavioral treatment ofoverweight children. J Soc Clin Psychol. 1983;1:343–353

59. Epstein LH. Application of behavioral economic principles to treat-ment of childhood obesity. In: Allison DB, Pi-Sunyer FX, eds. ObesityTreatment: Establishing Goals, Improving Outcomes and Reviewing theResearch Agenda. New York, NY: Plenum Press; 1995:113–119

60. Epstein LH. Family-based treatment for childhood obesity. In: Wol-raich ML, Routh DK, eds. Advances in Developmental and BehavioralPediatrics. Greenwich, CT: JAI Press; 1985:1–39

61. Endo H, Takagi Y, Kuwahata K, Uemasu F, Kobayashi A. Beneficialeffects of dietary intervention on serum lipid and apoliprotein levels inobese children. Am J Dis Child. 1992;146:303–305

62. Epstein LH, Kuller LH, Wing RR, Valoski A, McCurley J. The effect ofweight control on lipid changes in obese children. Am J Dis Child.1989;143:454–457

63. Sasaki J, Shindo M, Tanaka H, Ando M, Arakawa K. A long-termaerobic exercise program decreases the obesity index and increases thehigh density lipoprotein cholesterol concentration in obese children.Int J Obes. 1987;11:339–345

64. Zwiauer K, Kerbl B, Widhalm K. No reduction of high density lipopro-

tein2 during weight reduction in obese children and adolescents. EurJ Pediatr. 1989;149:192–193

65. Fanari P, Somazzi R, Nasrawi F, et al. Haemorheological changes inobese adolescents after short-term diet. Int J Obes. 1993;17:487–494

66. Hoffman RP, Stumbo PJE, Janz KF, Nielsen DH. Altered insulin resis-tance is associated with increased dietary weight loss in obese chil-dren. Hormone Res. 1995;44:17–22

67. Gutin B, Cucuzzo N, Islam S, Smith C, Stachura ME. Physical training,lifestyle education, and coronary risk factors in obese girls. Med SciSports Exercise. 1996;28:19–23

68. Israel AC, Shapiro LS. Behavior problems of obese children enrollingin a weight reduction program. J Pediatr Psychol. 1985;10:449–460

69. Tershakovec AM, Weller SC, Gallagher PR. Obesity, school perfor-mance and behavior of black, urban elementary school children. Int JObes. 1994;18:323–327

70. Martin S, Housley K, McCoy H, et al. Self-esteem of adolescent girls asrelated to weight. Percept Mot Skills. 1988;67:879–884

71. Emery RE, Binkoff JA, Houts AC, Carr EG. Children as independentvariables: some clinical implications of child-effects. Behav Ther. 1983;14:398–412

72. Epstein LH, Wisniewski L, Weng R. Child and parent psychologicalproblems influence child weight control. Obes Res. 1994;2:509–515

73. Friedman MA, Brownell KD. Psychological correlates of obesity: mov-ing to the next research generation. Psychol Bull. 1995;117:3–20

74. French SA, Story M, Perry CL. Self-esteem and obesity in children andadolescents: a literature review. Obes Res. 1995;3:479–490

75. Sherman JB, Alexander MA, Gomez D, Kim M, Marole P. Interventionprogram for obese school children. J Commun Health Nurs. 1992;9:183–190

76. Foster GD, Wadden TA, Brownell KD. Peer-led program for the treat-ment and prevention of obesity in the schools. J Consult Clin Psychol.1985;53:538–540

77. Rohrbacher R. Influence of a special camp program for obese boys onweight loss, self-concept, and body image. Res Q. 1973;44:150–157

78. Stoner S, Fiorillo M. A program for self-concept improvement andweight reduction for overweight adolescent females. Psychology. 1976;13:30–35

79. Mellin LM, Slinkard LA, Irwin CE. Adolescent obesity intervention:validation of the SHAPEDOWN program. J Am Diet Assoc. 1987;87:333–338

80. Thomas-Dobersen DA, Butler-Simon N, Fleshner M. Evaluation of aweight management intervention program in adolescents with insulin-dependent diabetes mellitus. J Am Diet Assoc. 1993;93:535–540

81. Uzark KC, Becker MH, Dielman TE, Rocchini AP, Katch V. Perceptionsheld by obese children and their parents: implications for weightcontrol intervention. Health Educ Q 1988;15(2):185–198

82. Achenbach TM. Manual for the Child Behavior Checklist/4–18 and 1991Profile. Burlington, VT: University of Vermont Department ofPsychiatry; 1991

83. Lifshitz F, Moses N. A complication of dietary treatment of hypercho-lesterolemia. Am J Dis Child. 1989;143:537–452

84. Craighead L, Stunkard AJ, O’Brien R. Behavior therapy and pharma-cotherapy for obesity. Arch Gen Psychiatry. 1981;38:1224–1229

85. Figueroa-Colon R, Franlin FA, Lee JY, von Almen TK, Suskind RM.Feasibility of a clinic-based hypocaloric dietary intervention imple-mented in a school setting for obese children. Obes Res. 1996;4:419–429

86. Garn SM, Clark DC, Guire KE. Level of fatness and size attainment.Am J Anthropol. 1974;40:447–450

87. Abraham S, Johnson CL. Prevalence of severe obesity in the UnitedStates. Am J Clin Nutr. 1980;33:364–369

88. Lindgren G. Growth of schoolchildren with early, average and lateages of peak height velocity. Ann Hum Biol. 1978;5:253–267

89. French SA, Jeffery RW. Consequences of dieting to lose weight: effectson physical and mental health. Health Psychol. 1994;13:195–212

90. Lacey JH, Coker S, Birtchnell SA. Bulimia: factors associated with itsetiology and maintenance. Int J Eat Disord. 1986;5:475–487

91. Thompson JK, Coovert MD, Richards KJ, Johnson S, Cattarin J. Devel-opment of body image, eating disturbance, and general psychologicalfunctioning in female adolescents: covariance structural modeling andlongitudinal investigations. Int J Eat Disord. 1995;18:221–236

92. Patton GC, Johnson-Sabine E, Wood K, Mann AH, Wakeling A. Ab-normal eating attitudes in London school girls—a prospective epide-miologic study: outcome at twelve month follow-up. Psychol Med.1990;20:383–394

93. Fairburn CG, Beglin SJ. Studies on the epidemiology of bulimia ner-vosa. Am J Psychiatry. 1990;147:401–408

94. Fitzgibbon ML, Stolley MR, Kirschenbaum DS. Obese people who seektreatment have different characteristics than those who do not seek


treatment. Health Psychol. 1993;12:342–34595. Birch LL, Fisher JO. Development of eating behaviors among children

and adolescents. Pediatrics. 1998;101:539–54996. Kohl HW, Hobbs KE. Development of physical activity behaviors

among children and adolescents. Pediatrics. 1998;101:549–55497. Epstein LH. Role of behavior theory in behavioral medicine. J Consult

Clin Psychol. 1992;60:493–49898. Eisenberger R, Cameron J. Detrimental effects of reward: reality or

myth? Am Psychol. 1996;51:1153–116699. Stunkard A. “Diet, exercise and behavior therapy”: a cautionary tale.

Obes Res. 1996;4:293–294100. Spiegel TA, Kaplan JM, Tomassini A, Stellar E. Bite size, ingestion rate,

and meal size in lean and obese women. Appetite. 1993;21:131–145101. Bandini LG, Schoeller DA, Cyr HN, Dietz WH. Validity of reported

energy intake in obese and non-obese adolescents. Am J Clin Nutr.1990;52:421–425

102. Lichtman SW, Pisarska K, Berman ER, et al. Discrepancy betweenself-reported and actual caloric intake in obese subjects. N Engl J Med.1992;327:1893–1898

103. Baranowski T, Lin SL, Wetter DW, Resnicow K, Hearn MD. Theory asmediating variables: why aren’t community interventions working asdesired? Ann Epidemiol. 1998. In press

104. Worsley A, Coonan W, Leitch D, Crawford D. Slim and obese chil-dren’s perceptions of physical activities. Int J Obes. 1984;8:201–211

105. Epstein LH, Smith JA, Vara LS, Rodefer JS. Behavioral economic anal-ysis of activity choice in obese children. Health Psychol. 1991;10:311–316

106. Dietz WH, Gortmaker SL. Do we fatten our children at the televisionset? Obesity and television viewing in children and adolescents. Pedi-atrics. 1985;75:807–812

107. Gortmaker SL, Must A, Sobol AM, Peterson K, Colditz GA, Dietz WH.Television watching as a cause of increasing obesity among children inthe United States, 1986–1990. Arch Pediatr Adolesc Med. 1996;150:356–362

108. Epstein LH, Saelens BE, O’Brien JG. Effects of reinforcing increases inactive behavior versus decreases in sedentary behavior for obese chil-dren. Int J Behav Med. 1995;2:41–50

109. Epstein LH, Saelens BE, Myers MD, Vito D. The effects of decreasingsedentary behaviors on activity choice in obese children. Health Psy-chol. 1997;16:107–113

110. Wolffgramm J, Heyne A. From controlled drug intake to loss ofcontrol: the irreversible development of drug addiction in the rat.Behav Brain Res. 1995;70:77–94

111. Timberlake W, Farmer-Dougan VA. Reinforcement in applied settings:figuring out ahead of time what will work. Psychol Bull. 1991;110:379–391

112. Baker AG, Steinwald H, Bouton ME. Contextual conditioning andreinstatement of extinguished instrumental responding. Q J Exp Psy-chol. 1991;43B:199–218

113. Bouton ME, Peck CA. Spontaneous recovery in cross-motivationaltransfer (counterconditioning). Animal Learn Behav. 1992;20:313–321

114. Brooks DC, Bouton ME. A retrieval cue for extinction attenuates spon-

taneous recovery. J Exp Psychol. 1993;19:77–89115. Bouton ME. Context, ambiguity and classical conditioning. Curr Direc-

tions Psychol Sci. 1994;3:49–53116. Bouton ME, Swartzentruber D. Sources of relapse after extinction in

Pavlovian and instrumental learning. Clin Psychol Rev. 1991;11:123–140117. American Dietetic Association. Position of the American Dietetic

Association: weight management. J Am Diet Assoc. 1997;97:71–74118. Willard MD. Obesity: types and treatments. Am Fam Pract. 1991;43:

2099–2108119. Oram N. Children’s eating experiences could differ from those of

adults. Appetite. 1994;22:283–287120. Thomas J. Food choices and preferences of schoolchildren. Proc Nutr

Soc. 1991;50:49–57121. Drewnowski A. Obesity and eating disorders: cognitive aspects of food

preference and food aversion. Bull Psychonom Soc. 1991;29:261–264122. Barkeling B, Ikman S, Rossner S. Eating behaviour in obese and normal

weight 11-year-old children. Int J Obes. 1992;16:355–360123. Birch LL, McPhee L, Shoba BC, Steinberg L, Krehbiel R. “Clean up

your plate”: effects of child feeding practices on the conditioning ofmeal size. Learn Motiv. 1987;18:301–317

124. Birch LL, Deysher M. Conditioned and unconditioned caloriccompensation: evidence for self-regulation of food intake in youngchildren. Learn Motiv. 1985;16:341–355

125. Satter EM. Internal regulation and the evolution of normal growth asthe basis for prevention of obesity in children. J Am Diet Assoc. 1996;96:860–864

126. Carrier KM, Steinhardt MA, Bowman S. Rethinking traditional weightmanagement programs: a 3-year follow-up evaluation of a new ap-proach. J Psychol. 1994;128:517–535

127. Johnson SL, Birch LL. Parents’ and children’s adiposity and eatingstyle. Pediatrics. 1994;94:653–661

128. Yanovski JA, Yanovski SZ, Filmer KM, et al. Differences in bodycomposition of black and white girls. Am J Clin Nutr. 1996;64:833–839

129. Foster GD, Wadden TA, Vogt RA. Resting energy expenditure in obeseAfrican American and Caucasian women. Obes Res. 1997;5:1–8

130. Kuczmarski RJ, Flegal KM, Campbell SM, Johnson CL. Increasingprevalence of overweight among US adults. JAMA. 1994;272:205–211

131. Dietz WH. Health consequences of obesity in youth: childhood pre-dictors of adult disease. Pediatrics. 1998;101:518–525

132. Greene GW, Rossi SR, Reed GR, Willey C, Prochaska JO. Stages ofchange for reducing dietary fat to 30% of energy or less. J Am DietAssoc. 1994;94:1105–1110

133. Miller WR, Rollnick S. Motivational Interviewing: Preparing People toChange Addictive Behavior. New York, NY: Guilford Press; 1991

134. Resnicow K. School-based obesity prevention: population versus high-risk interventions. In: Williams CL, Kimm SYS, eds. Prevention andTreatment of Childhood Obesity. New York, NY: New York Academy ofSciences; 1993:154–166

135. Piscano JC, Lichter H, Ritter J, Siegal AP. An attempt at prevention ofobesity in infancy. Pediatrics. 1978;61:360–364

Childhood Obesity: Future Directions and Research Priorities

James O. Hill, PhD*, and Frederick L. Trowbridge, MD‡

ABSTRACT. The threat of obesity is greater than everfor US children and adolescents. All indications arethat the current generation of children will grow intothe most obese generation of adults in US history.Furthermore, there is every expectation that the nextgeneration of children is likely to be fatter and less fit

than the current generation. Despite the recognition ofthe severe health and psychosocial damage done bychildhood obesity, it remains low on the public agendaof important issues facing policy makers. Perhaps thisis because the most serious health effects of obesity intoday’s children will not be seen for several decades.Action must be taken now to stem the epidemic ofchildhood obesity. This action will require a prioriti-zation of research into the etiology, treatment, andprevention of childhood obesity. It is unlikely thatsufficient resources for such research will be availablefrom public and private sources until the issue ofchildhood obesity is moved higher on the public

From the *Department of Pediatrics, University of Colorado Health SciencesCenter, Denver, Colorado; and ‡Nutrition and Health Promotion Program,International Life Sciences Institute (ILSI), Atlanta, Georgia.Received for publication Oct 24, 1997; accepted Nov 6, 1997.PEDIATRICS (ISSN 0031 4005). Copyright © 1998 by the American Acad-emy of Pediatrics.


1998;101;554Pediatrics Leonard H. Epstein, Michelle D. Myers, Hollie A. Raynor and Brian E. Saelens


ServicesUpdated Information &

http://pediatrics.aappublications.org/content/101/Supplement_2/554including high resolution figures, can be found at:

References

BIBLhttp://pediatrics.aappublications.org/content/101/Supplement_2/554#This article cites 126 articles, 14 of which you can access for free at:

Subspecialty Collections

http://www.aappublications.org/cgi/collection/obesity_new_subObesityhttp://www.aappublications.org/cgi/collection/endocrinology_subEndocrinologyfollowing collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlin its entirety can be found online at: Information about reproducing this article in parts (figures, tables) or

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

by guest on April 3, 2019www.aappublications.org/newsDownloaded from

http://http://pediatrics.aappublications.org/content/101/Supplement_2/554

http://pediatrics.aappublications.org/content/101/Supplement_2/554#BIBL

http://pediatrics.aappublications.org/content/101/Supplement_2/554#BIBL

http://www.aappublications.org/cgi/collection/endocrinology_sub

http://www.aappublications.org/cgi/collection/obesity_new_sub

http://www.aappublications.org/site/misc/Permissions.xhtml

http://www.aappublications.org/site/misc/reprints.xhtml

1998;101;554Pediatrics Leonard H. Epstein, Michelle D. Myers, Hollie A. Raynor and Brian E. Saelens


http://pediatrics.aappublications.org/content/101/Supplement_2/554located on the World Wide Web at:

The online version of this article, along with updated information and services, is

1073-0397. ISSN:60007. Copyright © 1998 by the American Academy of Pediatrics. All rights reserved. Print

the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois,has been published continuously since 1948. Pediatrics is owned, published, and trademarked by Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it

by guest on April 3, 2019www.aappublications.org/newsDownloaded from

http://pediatrics.aappublications.org/content/101/Supplement_2/554

Documents

Treatment of Pediatric Obesity