the physiology of body weight regulation: relevance - pediatrics

insulin with blood pressure in young individuals. Arch Intern Med.1993;153:323–328

58. Rocchini AP, Katch V, Kveselis D, et al. Insulin and renal sodiumretention in obese adolescents. Hypertension. 1989;14:367–374

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

60. Weisberg LA, Chutorian AM. Pseudotumor cerebri of childhood. Am JDis Child. 1977;131:1243–1248

61. Grant DN. Benign intracranial hypertension; a review of 79 cases ininfancy and childhood. Arch Dis Child. 1971;46:651–655

62. Mallory GB Jr, Fiser D, Jackson R. Sleep-associated breathing disordersin morbidly obese children and adolescents. J Pediatr. 1989;115:892–897

63. Rhodes SK, Shimoda KC, Waid LR, et al. Neurocognitive deficits inmorbidly obese children with obstructive sleep apnea. J Pediatr. 1995;127:741–744

64. Riley DJ, Santiago TV, Edelman NH. Complications of obesity-hypoventilation syndrome in childhood. Am J Dis Child. 1976;130:671–674

65. Dietz WH Jr, Gross WL, Kirkpatrick JA Jr. Blount disease (tibia vara):another skeletal disorder associated with childhood obesity. J Pediatr.1982;101:735–737

66. Kelsey JL, Keggi KJ, Southwick WO. The incidence and distribution ofslipped femoral epiphysis in Connecticut and southwestern UnitedStates. J Bone Joint Surg. 1970;52A:1203–1216

67. Sorenson KH. Slipped upper femoral epiphysis. Acta Orthop Scand.1968;39:499–517

68. Kelsey JL, Acheson RM, Keggi KJ. The body build of patients withslipped femoral capital epiphysis. Am J Dis Child. 1972;124:276–281

69. Polson DW, Wadsworth J, Adams J, Franks S. Polycystic ovaries—acommon finding in normal women. Lancet. 1988;1:870–872

70. Goldzieher JW, Green JA. The polycystic ovary. I. Clinical and histo-logic features. J Clin Endocrinol Metab. 1962;22:325–328

71. Kahn CR, Flier JS, Bar RS, et al. The syndromes of insulin resistance andacanthosis nigricans. N Engl J Med. 1976;294:739–745

72. Franks S. Polycystic ovary syndrome. N Engl J Med. 1995;333:853–86173. McKenna TJ. Pathogenesis and treatment of polycystic ovary syndrome.

N Engl J Med. 1988;318:558–56274. Yen SSC. The polycystic ovary syndrome. Clin Endocrinol. 1980;12:

177–20775. Dietz WH. Critical periods in childhood for the development of obesity.

Am J Clin Nutr. 1994;59:995–99976. Guo SS, Roche AF, Chumlea WC, et al. The predictive value of child-

hood body mass index values for overweight at age 35 y. Am J Clin Nutr.1994;59:810–819

77. DiPietro L, Mossberg H-O, Stunkard AJ. A 40-year history of over-weight children in Stockholm: life-time overweight, morbidity, andmortality. Int J Obes. 1994;18:585–590

78. Freedman DS, Shear CL, Burke G, et al. Persistence of juvenile onsetobesity over eight years: the Bogalusa Heart Study. Am J Public Health.1987;77:588–592

79. Rimm IJ, Rimm AA. Association between juvenile onset obesity andsevere adult obesity in 73,532 women. Am J Public Health. 1976;66:479–481

80. Braddon FEM, Rodgers B, Wadsworth MEJ, Davies JMC. Onset ofobesity in a 36 year birth cohort study. Br Med J. 1986;293:299–303

81. Abraham S, Nordsieck M. Relationship of excess weight in children andadults. Public Health Rep. 1960;75:263–273

82. Must A, Jacques PF, Dallal GE, et al. Long-term morbidity and mortalityof overweight adolescents; a follow-up of the Harvard Growth Study of1922 to 1935. N Engl J Med. 1992;327:1350–1355

83. Bao W, Srinivasan SR, Wattigney WA, Berenson GS. Persistence ofmultiple cardiovascular risk clustering related to Syndrome X fromchildhood to young adulthood. Arch Intern Med. 1994;154:1842–1847

84. Wattigney WA, Webber LS, Srinivasan SR, Berenson GS. The emergenceof clinically abnormal levels of cardiovascular risk factor variablesamong young adults: the Bogalusa Heart Study. Prev Med. 1995;24:617–626

The Physiology of Body Weight Regulation: Relevance to the Etiology ofObesity in Children

Michael Rosenbaum, MD, and Rudolph L. Leibel, MD

ABSTRACT. The prevalence of obesity in children andadults in the United States has increased by more than30% over the past decade. Recent studies of the physiol-ogy and molecular genetics of obesity in humans haveprovided evidence that body weight (fat) is regulated.Some of the genes encoding the molecular componentsof this regulatory system have been isolated from ro-dents. The increasing prevalence of obesity in the UnitedStates apparently represents the interaction of thesegenes with an environment that encourages a sedentarylifestyle and consumption of calories. The rapid increasein the prevalence of obesity emphasizes the role of envi-ronmental factors, because genetic changes could notoccur at this rate. Thus, understanding of the relevantgenes and how their effects are mediated by environmentand development should lead to more effective prophy-laxis and therapy of obesity. Although no clear environ-mental factors have been identified as causative of obe-sity, the rapid increases in the prevalence of obesity and

the seeming voluntary immutability of adult body fat-ness can be taken as tacit evidence that the pediatricenvironment can be altered in a way that affects adultbody weight. Pediatrics 1998;101:525–539; obesity, genet-ics, energy metabolism, body weight regulation.

ABBREVIATIONS. NHANES, National Health and Nutrition Ex-amination Surveys; C/EBP, CCAAT-enhancer binding protein;BMI, body mass index; CNS, central nervous system; ASP, agoutisignaling protein; MCH, melanin-concentrating hormone; VMH,ventromedial hypothalamus; LH, lateral hypothalamus; NPY,neuropeptide Y; PVN, paraventricular nucleus.

Storage of excess calories as fat must ultimatelyresult from a net positive energy balance (ener-gy intake greater than energy expenditure) over

time. Thus, the physiologic determinants of bodycomposition are 1) energy intake, 2) energy output,and 3) partitioning of energy stores as fat, carbohy-drate, and protein. Many physiologic systems (endo-crine, gastrointestinal, central nervous, peripheralnervous, and cardiovascular) affect these functions.Small changes in any of these determinants can, overtime, result in substantial changes in body weight.

From the Department of Pediatrics, Division of Molecular Genetics, Collegeof Physicians and Surgeons at Columbia University/Columbia PresbyterianHospital, New York, 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.

SUPPLEMENT 525 by guest on January 10, 2019www.aappublications.org/newsDownloaded from

Given the importance of energy stores to individ-ual survival and reproductive capacity,1 the ability toconserve energy as adipose tissue would have con-ferred survival advantage in the environment inwhich hominids evolved. For this reason, humansare presumably enriched for genes that promote en-ergy intake and storage and that minimize energyexpenditure. By virtue of their effects on fat stores,such genes would also enhance female fertility andthe ability to breastfeed offspring. However, in themodern industrial environment that provides easyaccess to calorically dense foods and encourages asedentary lifestyle, the metabolic consequences ofthese genes are maladaptive.2,3

In the 10 years between the National Health andNutrition Examination Survey II (NHANES II, a sur-vey that included measures of body fatness in theUnited States between 1976 and 1980) and NHANESIII (the same survey repeated between 1988 and1991), the prevalence of overweight, based on bodymass index (BMI; kg/m2) corrected for age and sex inearlier national surveys (Nutrition Health and Exam-ination Survey [NHES] II [1964 to 1965] and NHES III[1966 to 1970]) has increased by ;40%.4,5 A similarincrease in the prevalence of adult obesity (24% inNHANES II vs 33% in NHANES III) was reportedover the same period.5 The observation that the prev-alence of obesity increased substantially over a singledecade, a period much too brief for any significantchange to have occurred in the genetic makeup of thepopulation of the United States, indicates that thecurrent relative adiposity is a product of the interac-tion between genetic predisposition with regard tothe storage of body fat and an environment (lowphysical activity, high availability of caloricallydense foods) that is increasingly permissive to theexpression of that genetic tendency. Although thereare clearly strong genetic influences on susceptibilityto obesity, large changes in the prevalence of obesityover such a short time must reflect major changes innongenetic factors, providing tacit evidence thatsome instances or aspects of obesity must be respon-sive to, or preventable by, manipulation of the envi-ronment (eg, diet, physical activity).

Thus, obesity is a complex phenotype that resolvesthe influences of genes, development, and environ-ment. In any individual—and in the same individualat different times of life—the relative influence ofthese factors may vary. In this sense, obesity is aprototype of many of the diseases (eg, hypertension,dyslipidemia, type II diabetes mellitus) now con-fronting medical science. Among the most importantconcepts about these phenotypes is the idea thatrelevant genes mediate susceptibility to disease in aspecific environmental context and is not the inevi-table occurrence of the disease regardless of the en-vironment. Individuals with an otherwise potent ge-netic predisposition to obesity still will be lean in anenvironment of food deprivation or high demand forphysical activity; individuals not genetically predis-posed to obesity may still become so in an environ-ment that includes tasty, calorically dense foods andfew inducements to physical activity. Thus, in anyeffort to elucidate the genetic bases for susceptibility

to obesity, the environment in which the obesity isoccurring remains a critical factor with regard towhat sorts of genes will be identified. Because thegenes that mediate susceptibility to obesity may af-fect energy intake, energy expenditure, and parti-tioning of stored calories between lean tissues andfat, the ability to define the gross metabolic basis forobesity is very important for determining whichgenes are relevant to the phenotype. Current re-search regarding the molecular physiology of weightregulation6,7 is providing the reagents and physio-logic targets that should permit the mechanistic de-scription of specific instances of human obesity. Inthis article, we describe some of the insights pro-vided by these studies. These insights will surely berelevant to the etiology of childhood obesity. Cur-rently, however, the precise etiology of any specificinstance of human obesity remains unknown. Thisstatement is true for both the rare dysmorphic obe-sities (eg, Prader–Labhart–Willi Syndrome) and thegarden variety obesity in the population at large.

REVIEW

Biochemistry and Development of Adipose TissueWhite adipose tissue derives from mesenchyme.

Fully differentiated adipocytes are first detectable inthe human fetus at ;15 weeks of gestation. Bothadipocyte volume and number continue to increasethroughout gestation, with maximal rates of changeachieved after approximately the 30th week.8 Duringthe third trimester of pregnancy, there is an ;12-foldincrease in total body fat (percentage body fat in-creases from ;5% to 15%).9 The composition ofstored triglyceride is similar to that of maternalstores early in gestation, but fat stored during thethird trimester consists predominantly of saturatedfatty acids.9 These observations suggest that earlygestational lipid stores reflect passive transfer of ma-ternal fatty acids, whereas most newborn lipid storesare attributable to increased fetal intrahepatic lipo-genesis from carbohydrate (preferential synthesis ofsaturated free fatty acids).

During the first year of extrauterine life, most ad-ipose tissue growth occurs by enlargement (hyper-trophy) rather than by increased numbers of adipo-cytes (hyperplasia), but after 2 years there is littleadditional increase in adipocyte volume in the non-obese child.10–12 The growth of adipose tissue and thechanges in body composition with age are illustratedin the Figure. In nonobese children, there is no sig-nificant change in fat cell volume from 2 to 14 yearsof age and there is only a slight increase in fat cellnumber from 2 to 10 years of age. In obese children,there is continual enlargement of adipocytes withouthyperplasia during the same period.13 These obser-vations are consistent with a model of adipose tissuegrowth (see below) whereby adipocyte hyperplasiais triggered by achievement of a critical adipocytevolume.

As far as we know, there is a virtually unlimitedpool of preadipocytes represented in the pericytes ofthe pericapillary endothelium. These preadipocytesare visually and biochemically indistinguishable

526 SUPPLEMENT by guest on January 10, 2019www.aappublications.org/newsDownloaded from

from other fibroblasts and have no detectable lipo-genic or lipolytic enzyme activity (see below). Once acell differentiates into an adipocyte, the differentia-tion is irreversible (terminal differentiation). In vitroand in vivo studies suggest that there is local controlof adipocyte differentiation, perhaps by signals gen-erated from preadipocytes or mature adipocytes.When fibroblasts from the pericapillary epitheliumare examined in tissue culture, they multiply until astate of confluence (the culture plate is completelycovered with fibroblasts) is reached. Once confluencehas occurred, isolated nests of cells differentiate ter-minally into groups of adipocytes separated bypatches of undifferentiated cells. However, if a spaceis created around those undifferentiated cells,

thereby stimulating them to divide further, they willalso differentiate.14 In vivo, maximal adipocyte lipidcontent appears to be ;1 g of lipid per cell. Once thisdegree of hypertrophy has occurred, recruitment ofnew adipocytes begins,15 which suggests that adipo-cyte hypertrophy may generate a signal that encour-ages additional differentiation of preadipocytes.16

During therapeutic weight loss in children, the rateat which new fat cells develop is diminished. How-ever, fat cell number apparently continues to in-crease at a rate significantly greater than that ofage-matched controls on an ad libitum diet.17 Thesedata imply that fat cell hyperplasia can be restrainedtherapeutically, but that the rate of appearance ofnew fat cells may, to some extent, be programmedgenetically or developmentally.

Several transcription factors described recentlyhave powerful effects on adipogenesis. PPARg2 is amember of the perioxosome proliferator activatedreceptor subfamily of nuclear hormone receptorsthat is induced very early in adipocyte differentia-tion. PPARg2 is a transcription factor that controls orinduces adipocyte differentiation and expression ofadipocyte-specific genes.18 Inducers of PPARg2 ex-pression include insulin and insulin-sensitizing com-pounds (eg, the thiazolidinediones), glucocorticoids,certain prostaglandins, and, as the name implies,perioxosome proliferating drugs such as clofibrate.19

PPARg2 interacts in a cooperative manner with othertranscription factors in adipogenesis. CCAAT-enhancer-binding protein a (C/EBPa) is a transcrip-tion factor expressed relatively late in adipogenesis.As with PPARg2, C/EBPa activates adipocyte-spe-cific genes such as the adipocyte P2 (aP2) gene,which promotes synthesis of an intracellular fattyacid-binding protein.20,21 Deficiency of PPARg2 orC/EBPa, created by expressing them in cells thatlack specific activators for one or both of them, re-sults in significant declines in fat cell lipid contentand adipogenesis, whereas cells (eg, myoblasts) thatdo not differentiate into adipocytes will undergoadipogenesis if expression of both of these factors isinduced transgenically.22,23 The other C/EBPs,C/EBPb and C/EBPg, are expressed very early inadipogenesis. C/EBP appears to increase the expres-sion of PPAR.24 Another important factor in adipo-cyte differentiation is the transcription factor adipo-cyte determination and differentiation-dependentfactor 1 (ADD1/SREPB1), which is also expressedearly in adipogenesis. This protein induces addi-tional adipocyte differentiation as well as expressionof PPARg, fatty acid synthetase, and lipoproteinlipase.25 It is conceivable that some of these proteinsplay a role in mediating the effects of adipocytevolume on recruitment and differentiation of newadipocytes. Certain of these transcription factorsmight be targeted for future preventative treatmentsin children who are genetically at risk for obesity.

Heritability of Obesity-related Phenotypes in HumansThere are rare instances of single gene disorders

that result in human obesity (eg, Prader-Willi,Bardet-Biedl, Ahlstrom, Cohen), associated withother striking (dysmorphic) phenotypes.26 In most

Figure. Age-related changes in body composition of children.10–12


humans, however, body fatness is a continuousquantitative trait reflecting the interaction of devel-opment and environment with genotype. Studies intwins, adoptees, and families27,28 indicate that asmuch as 80% of the variance in BMI is attributable togenetic factors. That is, the genetic influences onbody weight are as potent as those on height.29 Her-itability of adipose tissue distribution (intraabdomi-nal versus subcutaneous fat), physical activity, rest-ing metabolic rate, changes in energy expenditurethat occur in response to overfeeding, certain aspectsof feeding behavior, food preferences, lipoproteinlipase activity, maximal insulin-stimulated acylglyc-eride synthesis, and basal rates of lipolysis are esti-mated to be as high as 30% to 40%.27,28

Segregation analyses in which the familiality ofobesity is examined under various genetic modelshave found evidence for segregation of major genes(allele frequencies, 0.14 to 0.26) influencing BMI (20%to 35% of variation) in human populations.28 Asmight be expected, the heritability of early-onset obe-sity appears to be considerably higher than that foradult-onset obesity.30 What is likely to be inheritedmost strongly is the rank order of body fat relative toone’s peers within a population sharing the sameenvironment.

Animal Models of Genetic ObesityBecause it is so difficult to define or control the

environment of humans experimentally, and becausethe attainment of an obese state may actually rectifythe metabolic mechanisms (increased energy intakerelative to expenditure or decreased energy expen-diture relative to intake) that predispose an individ-ual to obesity, animal models of obesity have beenlooked at intensively for clues to the relevant genesin humans. Rodent mutations in which obesity isinherited as a simple Mendelian recessive or domi-nant trait have been studied extensively in efforts toelucidate the biochemical basis for the most salientaspects of phenotypes related to adiposity: 1) in-creased food intake, 2) reduced energy expenditure,3) preferential partitioning of calories to fat, 4) auto-

nomic nervous system dysfunction, and 5) suscepti-bility to type II diabetes mellitus.31–33 Table 1 lists theextant, single gene mutations that result in obesity inrodents;33 all of these mutations are in genes forwhich human homologs exist.33,34

These rodent mutations themselves exemplifyphenotypically different subgroups of obesity thatcould have homologs in humans. For example, hy-perinsulinemia is more pronounced in Lepob andLepdb than in tub or Yellow (Ay).31,32,35 Similar to that inhumans, body fat distribution varies among theserodent mutations. Ay, fat, and tub show diffuse in-creases in body fat, whereas Lepob and Lepdb tend todeposit fat inguinally and axially.31,32 Partitioning ofstored calories as fat, carbohydrate, and protein alsois affected by these mutations. Compared with theirnonobese littermates, Lepob and Lepdb mice have re-duced fat-free mass despite their increased bodyweight.7

The complexity and redundancy of the mecha-nisms affecting body fatness also are exemplified bythese rodent models of obesity. Each mutation pro-duces a distinct obese phenotype by interruptingdifferent biochemical systems that regulate bodyweight. Lepob and Lepdb are mutations in a signalingsystem in which a protein (OB protein [leptin]) issecreted from adipocytes in proportion to fat mass(mutant in ob). The concentration of ambient leptin issensed centrally by a receptor (OB receptor or leptinreceptor, mutant in Lepdb mice and Lepfa rats) withresultant central nervous system (CNS)-mediatedchanges in food intake and energy expenditure, in-cluding hyperphagia, defective nonshivering ther-mogenesis, and preferential partitioning of excesscaloric storage as fat.7,31

Yellow mutations in the agouti gene (Ay Avy, Asy, orAiy) result in obesity and increased linear growth,hyperphagia, hyperinsulinemia, hypercortisolism,and increased lipogenesis.7 These mutations result inectopic overexpression of the agouti gene, whichencodes agouti signaling protein (ASP). In mice, ASPnormally is expressed only in hair follicles, where itcompetitively inhibits binding of melanocyte-stimu-

TABLE 1. Rodent Obesity Mutations

Gene Mutation Biochemical Effect of Mutation RodentChromosome

HumanHomologue

Physiologic Effect of Mutation

Leptin (Lep) ob112 Leptin deficiency Mouse 6 7q31.3 Central effects 31 foodintake and 2 energyexpenditure

Leptin receptor(Lepr)

Lepdb

Lepfa orLepfak126,171

Deranged leptin signaltransduction

Mouse 4Rat 5

1p31 Central effects 31 foodintake and 2 energyexpenditure

Carboxypeptidase E(Cpe)

fat37 Deficiency of carboxypeptidase E Mouse 8 4q32 Interferes with prohormone(including neuropeptide)processing and intracellulartransport

Tubby (tub) tub172 Deficiency of phosphodiesterase-like molecule

Mouse 7 11p15 Possible effects on cellularapoptosis in hypothalamus

Agouti(a) Ay173 Ectopic (brain) overexpression ofASP

Mouse 2 20q11.2 Blockade of melanocyte-stimulating hormone, andpossibly other ligands, atMC4R

Rodent obesity mutation symbols are based on current nomenclature. Earlier designations for Lepob and Lepdb were ob and db, respectively.With the exception of the dominant mutation in the agouti gene, all the mutations shown are recessive. For human genes, all capital lettersare used.


lating hormone at the melanocortin-1 receptor, re-sulting in synthesis of pheomelanin (yellow pig-ment) rather than eumelanin (black pigment). Whenexpressed centrally in the yellow mouse, ASP appar-ently competes with the normal ligand for the mela-nocortin-4 receptor. The melanocortinergic projec-tions activated by melanocortin-4 receptor arebelieved to exert a tonic inhibition of feeding. Hence,disruption of this system by ASP results in hy-perphagia.36

Fat is a mutation in a gene coding for an enzyme(carboxypeptidase E) that excises dibasic paired resi-dues at the carboxyl terminus of peptide prohormones,including processing of insulin and neuropeptidessuch as neurotensin, proopiomelanocortin, and mela-nin-concentrating hormone (MCH).37 Neurotensin andMCH are hypothalamic neuropeptides that inhibitfood intake,38 and the fat mutation may result in obesitybecause of an inability to synthesize these peptidesfully. These animals have profound hyperproinsuline-mia, which progresses to moderate obesity by the timethey are 8 weeks of age.39

Tub is a mutation in a phosphodiesterase-like genethat results in mild obesity with progressive hyper-insulinemia, loss of photoreceptors in the retina, andloss of inner ear hair cells, resulting in loss of visionand hearing. The tub gene product is highly ex-pressed in the hypothalamus, and it has been hy-pothesized that a defect in this product may result inapoptosis of specific cells within the ventromedialhypothalamus (VMH), resulting in an obese pheno-type similar to that observed when the VMH is ab-lated iatrogenically.7,32

Candidate Genes Regulating Body Fatness in HumansBecause obesity can arise only from an excess of

energy intake over expenditure or the preferentialpartitioning of stored calories to fat, the search forcandidate obesity genes should focus on those genesthat play a role in these aspects of energy metabo-lism. This strategy is being used currently to examineextended human families, sib pairs, and individualswithin populations for linkage or association of obe-sity and related phenotypes40 to molecular markerscorresponding to candidate genes based on meta-bolic physiology (eg, b3-adrenoreceptor,41 glucocor-ticoid receptor,42 Na,K-ATPase43), human syndromicobesities,44 and rodent obesity genes (eg, Lep and Leprin rodents, LEP and LEPR in humans).33 Tentativelinkage to LEP has been reported in groups of ex-tremely obese whites,45,46 but not in the Pima Indians,for whom obesity and diabetes are extremely com-mon.47 In a Hispanic population, a locus near LEPappears to be linked to some measures of adiposity.48

A recent study identified a region on human chro-mosome 2 that accounted for 47% of the variation inserum leptin concentrations and 32% of the variationin fat mass in a population of Mexican-Americans.49

Linkage of obesity50 and noninsulin-dependent dia-betes mellitus-related (acute insulin response)51 phe-notypes to a region (1p22–p31) that contains LEPR(leptin receptor) has been reported. The fine geneticstructures of all of these genes are known, enablingefforts to detect coding sequence variations in

genomic DNA, which may play a role in humanobesity.52

RESULTS

Regulation of Body WeightThe history of the scientific study of obesity began

in 1773 with the demonstration by Lavoisier that theheat of a guinea pig was derived from chemicalprocesses identical to those that occur in a piece ofburning wood, thus demonstrating the chemicalidentity of combustion and respiration and shatter-ing vitalistic views of biologic processes. In the en-suing 125 years, a distinguished succession of phys-icists, chemists, and physiologists established theprinciples of thermodynamics and their applicabilityto biological systems.53 Much of the struggle to un-derstand the biology of weight regulation has beenconducted in the presence of powerful, sometimessubtle, frequently commercially motivated, vitalistmisconceptions or misrepresentations about bioener-getics and Western societal views of obesity as thepsychological product of self-indulgence and over-gratification instead of as a disease.54 Almost 50 yearsago Mayer55,56 first proposed that the CNS sensedglucose and regulated energy intake by a glucostaticmechanism; Kennedy57 hypothesized a lipostatic sys-tem whereby body fat stores produced a signal thataffected systems of energy homeostasis, and Meyer58

proposed an aminostatic system whereby the qualityand quantity of amino acids ingested affected sys-tems of energy homeostasis. More recently, we haveamassed a substantial body of evidence that bodyweight is regulated59 by complex signaling systemsthat provide afferent signals (including glucostatic,lipostatic, and aminostatic signals) to the CNS aboutthe nutritional state of the organism, which then aretranslated into efferent signals that affect energy in-take and expenditure (see below).

Epidemiologic Evidence That Body Weight IsRegulated

Epidemiologic observations of the relative con-stancy of body composition over long periods of timesupport strongly a biological basis for the regulationof body fat. The relative stability of body weight overtime in most individuals,60 despite wide variation inenergy intake and expenditure, and the poor long-term results of weight reduction therapies (90% to95% of adults and children who lose weight return totheir previous state of fatness)27,61,62 suggest that at-tempts to maintain an altered body weight are op-posed by systems of energy homeostasis that defenda highly individualized body weight or setpoint. Ifbody weight were not regulated, then simply in-creasing caloric intake by 150 calories per day aboveweight maintenance calories per kilogram of usualbody weight (the caloric content of one 8-oz glass ofmilk) would result in a yearly caloric excess of;55 000 kcal or a weight gain of ;25 pounds. Yet,according to the Framingham study, the averageadult gains only ;20 pounds (;40 000 kcal of storedenergy versus ;20 million calories ingested over thesame period) between 35 and 55 years of age.60 Am-


ortized over 20 years, the ingestion of 5 kcal per day(;0.2% of total calories ingested) in excess of energyexpenditure is all that would be needed to gain 20pounds.

Physiologic Evidence That Energy ExpenditureChanges to Oppose Alterations in Body Weight

Energy ExpenditureRegulation of body weight in obese and never-

obese adults differs only in that these individualsdefend different body weights. Decreases in bodyweight of both obese and never-obese adults areresisted by equivalent declines in energy expendi-ture.63,64 A formerly obese individual requires 10% to15% fewer calories to maintain a normal body weightthan a never-obese individual of the same body com-position.65,66 This decline in 24-hour energy expendi-ture primarily reflects a decrease in nonresting en-ergy expenditure (energy expended in physicalactivity) and, to a lesser extent in some studies, rest-ing energy expenditure (resting cardiorespiratorywork and the maintenance of transmembrane iongradients).64,67 Maintenance of an increased bodyweight is resisted with equal metabolic force. Thus, alean or obese individual who has gained weight to10% above usual body weight requires ;15% morecalories to maintain an elevated body weight. Themetabolic forces opposing the maintenance of analtered body weight are equally potent for obese andnever-obese adults and oppose weight gain asstrongly as they oppose weight loss. It also is notablethat when weight-maintenance caloric requirementsare normalized for body composition (ie, adjusted toaccount for the fact that it requires more calories tomove a larger amount of weight), there are no sig-nificant differences in the metabolic requirements oflean and obese individuals to maintain their usualbody weight. Thus, the view that obese adults sufferfrom a primary metabolic abnormality at their usualbody weight is incorrect. However, a formerly obeseindividual requires substantially fewer calories tomaintain body weight than a never-obese person ofthe same body weight and composition.64 Studies ofadults during maintenance of a 10% reduced bodyweight have shown that the mechanical efficiency ofskeletal muscle (calories expended per unit of work)is increased after weight loss. Conversely, duringmaintenance of a 10% increased body weight, themechanical efficiency of skeletal muscle (calories ex-pended per unit of work) is diminished.68 Adultswho have maintained reduced body weight for sev-eral years continue to demonstrate the metabolic op-position to the maintenance of a reduced bodyweight.63 These studies suggest that weight alter-ation-associated long-lasting changes in metabolicefficiency constitute an important part of the mech-anism by which systems of energy homeostasis inadults oppose the maintenance of an altered bodyweight.

As indicated previously, the increasing prevalenceof obesity in the United States provides tacit evi-dence that an increase in the body weight an indi-vidual tends to maintain over time can be facilitated

by environmental factors. Currently, there is no evi-dence, however, that environmental factors can bemanipulated to lower the usual weight or setpointthat is maintained by an individual without invokingthe metabolic opposition to the maintenance of areduced body weight described above. A critical andunanswered question is whether the compensatorychanges opposing the maintenance of an alteredbody weight that occur in adults also occur in chil-dren. If they do not, aggressive treatment of obesityearly in childhood may be accompanied by perma-nent downward resetting of the regulatory systemscontrolling body weight.

Two systems in the body that are primary regula-tors of the efficiency with which energy is used arethe autonomic nervous system and the thyroid axis.The parasympathetic limb of the autonomic nervoussystem includes the vagus nerve. Increased parasym-pathetic nervous system tone slows the heart rateand increases insulin release69–72 (ie, favors weightgain). Increased output from the sympathetic limb ofthe autonomic nervous system stimulates thyroidhormone release, increases heart rate, diminishes in-sulin release, and increases thermogenesis in brownadipose tissue69–73 (ie, favors weight loss). Thyroidhormone increases energy expenditure by increasingheart rate, blood pressure, and energy expenditure.74

Obese and never-obese adults studied at their usualbody weight, during weight loss, and during main-tenance of a reduced body weight demonstrate sig-nificant increases in parasympathetic nervous sys-tem activity and decreases in sympathetic nervoussystem activity and in circulating concentrations ofthyroid hormone, compared with usual weight bothduring weight loss and during maintenance of areduced body weight.75,76 The increases in parasym-pathetic nervous system activity and mechanical ef-ficiency of skeletal muscle and the decreases in sym-pathetic nervous system activity and circulatingconcentrations of thyroid hormone during mainte-nance of a reduced body weight are consonant withthe declines observed in resting and nonresting en-ergy expenditure described previously.

An important question is whether an individualdestined to become obese demonstrates increasedmetabolic efficiency before actually becoming obese.Many of the rodent genetic models of obesity dis-cussed below demonstrate increased metabolic effi-ciency within the first few weeks of life before obe-sity occurs.31,32,77–79 Roberts et al80 examined restingmetabolic rate and 24-hour energy expenditure in3-month-old infants born to both lean and obesemothers. Twenty-four-hour energy expenditure wassignificantly lower at 3 months in those infants whowere obese by the age of 1 year. The finding that24-hour energy expenditure, but not resting meta-bolic rate, was significantly lower in infants destinedto become obese suggests that the lower 24-hourenergy expenditure is attributable to a decrease inthe quantity of energy expended in physical activity.Similarly, low energy expenditure, adjusted for bodycomposition, was associated with a fourfold in-creased risk of gaining more than 7.5 kg over 2 yearsin a study of adult Southwest Native Americans


(Pimans).81 Weight reduction in some obese individ-uals may unmask the metabolic state that predis-posed them to become obese. In this sense, the extrabody fat of the obese may mediate (perhaps via asecreted protein such as leptin)82 a metabolic correc-tion for low energy expenditure. However, otherstudies83,84 have not found that reduced energy ex-penditure was predictive of subsequent weight gain.

Energy IntakeFeeding behavior consists of decisions about initi-

ation, composition, and termination of meals; thesedecisions are influenced by many internal and exter-nal factors.85 Prepubertal children allowed to con-sume food ad libitum over 6 days from a menuconsisting of foods they liked showed wide variationin calories ingested per meal, but total daily energyintake was relatively constant,86 suggesting that, withpresumably little knowledge of nutrition, childrensense the number of calories they have consumedand tend to keep their caloric consumption relativelystable. It should be noted that energy intake mustexceed energy output in the growing child. Thisexcess of intake over output is especially pronouncedduring rapid growth periods in infancy and adoles-cence. In contrast, energy intake must equal outputin the adult, whose body composition is constant.Theoretically, it may be possible to therapeuticallyexploit the differences in systems regulating energyhomeostasis between children and adults so that animposed degree of body fatness will be defendedonce growth ceases.

If changes in energy expenditure were the solefactor that opposed maintenance of a reduced bodyweight, then it would be relatively simple to pre-scribe a diet for life for the reduced-obese person.Simply staying on this diet would maintain the re-duced weight. Many of the same hormones and areasof the brain that are involved in regulation of energyexpenditure also influence feeding behavior (see be-low). For example, the leptin-deficient Lepob mouse isboth hypometabolic and hyperphagic.33 The observa-tion that humans maintain a relatively constant bodyweight despite variations in caloric intake and phys-ical activity implies that energy output and intakeare tightly linked. However, the interlocking controlmechanisms of these regulators of energy input andoutput remain incompletely understood.

Integration of Energy Intake and ExpenditureThe ventromedial hypothalamus (VMH) and lat-

eral hypothalamus (LH) have important effects onfeeding behavior (energy intake) and autonomic reg-ulation of energy expenditure.87,88 Lesions of theVMH render rats hyperinsulinemic, hyperphagic,and hypometabolic compared with their sham-oper-ated littermates89–92 (ie, the lesioned animals behavesimilarly to an underfed rat).93 Once they havereached a certain level of increased adiposity (thedegree of obesity they achieve is directly propor-tional to the size of the hypothalamic lesion), le-sioned rats eat the same quantities of food as theirlean sham-operated littermates and no longer gainweight at a comparatively excessive rate.59,91,94,95 The

observation that these lesioned animals will consumeexcess calories and gain excess weight to achieve acertain degree of fatness and then will defend thatdegree of fatness by appropriate adjustments of foodintake suggests that the VMH lesion may have altereda setpoint for body weight.88,91,94,96 Similarly, rats withlesions of the LH become hypermetabolic and hypo-phagic97 (ie, their energy expenditure and feeding be-havior are analogous to those of a nonlesioned overfedrat and remain so only until they reach a new lowerbody weight).87,91 The VMH and LH areas are, there-fore, viewed more accurately as subserving and inte-grating complex systems regulating both feeding be-havior and energy expenditure.98–100 Evidence for thepresence of similar regulatory centers in humans isprovided by the observation that traumatic or infec-tious injury to the human hypothalamus results in asyndrome characterized by hyperphagia, hyperinsulin-emia, and overactivity of the parasympathetic nervoussystem.101,102

Molecular Mediators of Feeding Behavior and EnergyExpenditure

There are many central and peripheral hormonesand neurotransmitters that may act directly or indi-rectly on systems of energy homeostasis. Neuropep-tide Y (NPY), a potent endogenous central appetitestimulant, is synthesized by cell bodies in the arcuatenucleus of the hypothalamus and transported ax-onally to the paraventricular nucleus (PVN),103

where the highest concentrations are found.104 NPYalso is synthesized and released from the adrenalgland and sympathetic nerves, but peripherally syn-thesized NPY does not pass the blood–brain barri-er.105 In rodents, food restriction is associated with asignificant increase in production of hypothalamicNPY mRNA,106 and intracerebroventricular adminis-tration of NPY is a strong stimulator of feeding be-havior.104 In addition, injection of NPY into the ro-dent PVN increases adipose tissue lipoprotein lipaseactivity and decreases brown fat sympathetic ner-vous system activity and thermogenesis104 (ie, exertscoordinate effects on energy intake and output thatfavor weight gain). Despite the compelling evidencefor the physiologic importance of NPY in energyhomeostasis, a mouse with a genetically engineeredknockout of the NPY gene has normal body fat andfood intake and becomes normally hyperphagic withfood restriction.107 This finding highlights the ex-traordinary redundancy of the physiologic systemsregulating caloric storage.

The mixed a- and b-adrenergic agonists epineph-rine and norepinephrine may act as both appetitestimulants (a-adrenergic) and suppressants (b-ad-renergic). The net effects of these endogenous cat-echolamines are, therefore, dependent on the relativedegree of expression of a- and b-adrenergic receptorsubtypes in target tissues. Dopamine and serotoninact as appetite suppressants.108 The catecholaminer-gic and serotoninergic systems are exploited by cur-rent antiobesity medications, which act to decreaseenergy intake via the sympathetic nervous system(eg, phentermine), to decrease appetite by inhibiting


serotonin reuptake (eg, d-fenfluramine), or by affect-ing both catecholaminergic and serotoninergic sys-tems (eg, sibutramine).6

Various other peptides (corticotropin-releasingfactor, galanin, glucagon-like peptide I, growth hor-mone-releasing hormone, cholecystokinin, glucagon,MCH) may modify the feeding behavior mediatedfrom moment to moment by monoamines.6 The pro-duction and release of these neuropeptides, in turn,is probably influenced by humoral signals, whichindicate the size of adipose tissue depots, eg, insulinand leptin (see below).

Glucocorticoids play a permissive role in obesity.The increase in body fat associated with endogenousor exogenous glucocorticoid excess is well recog-nized.109 Surgical or pharmacologic adrenalectomy ofgenetically obese or VMH-lesioned animals amelio-rates the obesity. These effects are attributable in partto secondary elevations in corticotropin-releasingfactor, which reduces food intake and modulates theappetite-stimulating effects of NPY.110 Adrenalec-tomy also sensitizes rats to the anorexigenic effects ofintracerebral insulin. Intracerebral injection of corti-costerone potentiates the expression of NPY mRNAand the amount of hypothalamic NPY protein.110 Go-nadal steroids also have potent effects on bodyweight. Administration of estradiol reduces food in-take, partly by reciprocal effects on corticotropin-releasing factor (increase) and NPY (decrease) activ-ity in the hypothalamus.111

Candidate Signals From Adipose TissueThe relative long-term constancy of body weight

has led many investigators to hypothesize the exis-tence of adipose tissue-mediated signals to the brainthat reflect the amount of calories stored. By centrallymediated effects on efferent signals that affect energyintake and expenditure, the strength of the adipocytesignal induces changes in energy homeostasis, whichpromote the return of the individual to the usualbody weight. The central mechanism integrating en-ergy balance is responsive to a large number of in-puts, including psychosocial and hedonic factors,neurophysiologic signals, and concentrations ofmonoamines and neuropeptides within specific syn-apses in the hypothalamus (Tables 2 and 3).

Leptin, encoded by the LEP gene,112 is synthesizedby and released from adipose tissue and is clearly acandidate for an afferent signal relating adipose tis-sue mass to central systems of energy homeostasis.Leptin concentration in plasma is directly propor-tional to fat mass in animals113 and humans.114 Leptinreduces food intake, increases energy expenditure,and reduces body fat in Lepob (leptin-deficient) micewhen administered systemically or intracerebroven-tricularly.115–117 Similar effects (at higher doses) arefound in nonobese animals.118 Hormones that affectleptin expression in adipose tissue include insulin(increase),119 glucocorticoids (increase),119 gonadalsteroids (leptin concentrations in premenopausal fe-

TABLE 2. Endogenous Chemicals Affecting Energy Intake and Expenditure: Appetite Stimulants

Chemical Site of Action Effect on Energy Intake Effect on Energy Expenditure

NPY PVN The most potent endogenous stimulant of food intakeidentified; 1 during food restriction

1 Lipoprotein lipase activity; 1Sympathetic nervous system tone;1 thermogenesis after centraladministration104

Norepinephrine andepinephrine

PVN (a-effect)

a-Adrenergic stimulation stimulates food intake 2 Lipolysis, 2 heart rate(attributable to vasoconstriction),2 energy expenditure (?)108,174

Insulin VMH, LH,liver, fat,muscle

Central administration 3 2 NPY and 1 foodintake175

1 Acylglyceride synthesis, 2lipolysis; 2 NPY 3 2lipoprotein lipase activity,sympathetic tone andthermogenesis111

Glucocorticoids 1 Food intake; 1 expression of NPY mRNA and[NPY]; adrenalectomy decreases fat mass inrodents with genetic and surgical (VMH-lesioned)obesities; 1 brain levels of norepinephrine andfood intake stimulating effects of centralnorepinephrine are blocked by adrenalectomy108,110

2 Thermogenesis, perhapsattributable to 2 CRFsynthesis108,110,111

Galanin PVN 1 Food intake and 1 preference for fat intake, maydepend on norepinephrine because central galaninadministration 3 1 norepinephrine and galanineffect is blocked by a1-adrenergic antagonists176

2 Circulating concentrations ofinsulin and corticosterone176

Opioids(dynorphins,endorphins, andenkephalin)

1 Food intake after activation of m or d receptor isblocked by opiate antagonists176,177

Growth hormone-releasinghormone

1 Appetite178

Somatostatin 1 Food intake179

Polypeptide YY 1 Food intake176

MCH ? MCH mRNA is overexpressed in the hypothalamusof Lepob mice; 1 during fasting in normal andobese animals; 1 food intake when injectedcentrally in mice180

Neurotensin 1 Food intake181


males more than postmenopausal females more thanmales),114 and b-adrenergic agonists (decrease).120

Leptin effects on energy homeostasis apparently areconveyed, in part, by decreasing arcuate NPYmRNA.121 However, the NPY knockout mouse isresponsive to the anorexigenic effects of leptin,107,122

suggesting that leptin also may act via an NPY-independent mechanism. Leptin administration toLepob mice restores insulin-sensitive glucose dispos-al,123 but insulin resistance in obese humans occurs inthe presence of high ambient concentrations of lep-tin.124 Leptin administration to Lepob mice also re-stores fertility, which suggests that leptin also mayact as a signal relating the nutritional status of theorganism to fertility. Such a signaling system wouldact to decrease the likelihood of conception duringtimes of undernutrition (ie, when the morbidity andmortality of mother, fetus, and infant all would beincreased). As suggested by Frisch,1 a minimumamount of adipose tissue (concentration of leptin)would be required for female fertility.

Circulating insulin concentration is proportional toadipocyte volume.125 Insulin gains access to the CNSvia a saturable transport system and reduces foodintake via inhibition of NPY expression, enhance-ment of cholecystokinin effects, and inhibition ofnorepinephrine reuptake in hypothalamic neurons.Autonomic nervous traffic is also affected by centraladministration of insulin.111 Insulin does not reduceNPY mRNA in Lepfa (Zucker) rats, suggesting that

because Lepfa is a mutation in the leptin receptor,126

insulin effects on NPY occur partially via effects onleptin receptor-mediated signaling.

Considerable interest has been focused recently onthe potential role of diet composition127 and weightfluctuation (yo-yo)128,129 in regulation of body com-position. It does not appear that the metabolic effectsof diet composition130,131 or yo-yoing plays a majorrole in the pathogenesis of human obesity.64,128,132

Protein, carbohydrate, and fat are metabolically in-terconvertible, with the notable exceptions of thesynthesis of essential fatty and amino acids. There isno conclusive scientific evidence that changing therelative proportions of these fuels in the diet in theabsence of decreased caloric intake will promoteweight loss.131 However, the higher caloric density offat and its contribution to palatability of foods maypromote the ingestion of calories.

It is clear that these systems are complex and re-dundant. Central mechanisms regulating energy in-take and expenditure are affected by many afferentsignals from the gastrointestinal, endocrine, andCNS and peripheral nervous system as well as adi-pose tissue. It is thus unlikely that a pharmacologicagent acting on any single afferent limb in this sys-tem (eg, catecholaminergic or serotoninergic ago-nists) will result in prolonged maintenance of a re-duced body weight because other limbs of this bodyweight regulatory system will still actively opposemaintenance of the reduced weight.

TABLE 3. Endogenous Chemicals Affecting Energy Intake and Expenditure: Appetite Suppressants

Chemical Site of Action Effect on Energy Intake Effect on Energy Expenditure

Norepinephrine andepinephrine

LH (b-effect) b-Adrenergic stimulation is anorectic108 b-Adrenergic stimulation 3 1sympathetic nervous systemactivity, 1 [thyroid hormone], 1energy expenditure; 1 lipolysis

Dopamine LH Anorectic177

Serotonin VMH Anorectic108

Leptin VMH Arcuate nucleus 2 Food intake in Lepob mouse116,118 2 NPY mRNA, 1 thermogenesisand sympathetic nervous systemactivity111,118 (not solelyattributable to effects on NPYbecause NPY-knockout mice stillrespond to leptin)107

Cholecystokinin(CCK)

Vagus nerve, whichprojects to the PVN(type A receptor)and brain (type Breceptor)

Anorectic effect of peripherally administeredCCK is blocked by abdominal vagotomy;centrally administered CCK also hasanorectic effect, perhaps by interactionwith adrenergic receptors108,182

Corticotropin-releasing factor(CRF)

PVN Anorectic (?), perhaps by effects oncorticotropin or glucocorticoid production(see below); 2 [CRF] in hypothalamus ofLepfa rats183

1 Sympathetic nervous systemoutput; 1 thermogenesis; 1energy expenditure

Urocortin PVN Anorectic, perhaps via mechanisms similar toCRF184

Glucagon Vagus nerve, whichprojects to the PVN

Anorectic effect of peripherally administeredglucagon is blocked by abdominalvagotomy108

May 3 2 energy expenditurebecause humans given glucagonintramuscularly gain weightdespite decreasing caloric intake185

Glucagon-likepeptide-1 (GLP-1)

Hypothalamus Anorectic, may transduce the anorectic signalfrom leptin; central administration of GLP-1 antagonist 3 1 feeding in response toNPY186

g-Aminobutyricacid

VMH Anorectic108

Glucose VMH, LH, liver, fat,muscle

Anorectic187–190


Possible Environmental Influences on the Developmentof Obesity in Childhood

Prenatal InfluencesThe infant of a diabetic mother is a model for the

influences of fetal overnutrition on subsequent adi-posity. Exposure of the fetus to high concentrationsof ambient glucose stimulates fetal hyperinsulinism,increased lipogenesis, and macrosomia.133 Becausegestational diabetics often are obese, it is difficult toseparate the metabolic consequences of gestationalobesity and diabetes on subsequent adiposity of theinfant of a diabetic mother from the possibility thatthe mother has passed a genetic tendency towardobesity to her offspring. Studies of the obesity-pronePima Indians of Arizona have attempted to separategenetic pregestational and gestational influences onadiposity of offspring. Pettit et al134 reported that asignificantly higher percentage of offspring of gesta-tional diabetic Pima Indian women were obese (de-fined as .140% of ideal body weight for height) at 15to 19 years of age than offspring of nondiabeticwomen, and that these differences in the incidence ofsubsequent obesity were independent of the degreeof maternal obesity. The positive association betweenmaternal diabetes (controlled for adiposity) and ad-olescent obesity suggests that prenatal exposure ofthe fetus to the milieu of the diabetic mother mayconstitute an independent risk factor for subsequentobesity.

Rodents of mothers that were malnourished dur-ing the first 2 weeks of a 3-week gestation displayedhyperphagia and an increase in the subsequent adi-posity of the gestationally malnourished off-spring.135–138 Epidemiologic studies of the effects ofintra- and early extrauterine malnutrition on the riskof obesity in humans have suggested that such ef-fects are subtle in most instances. Males born duringor just after the Nazi-imposed Dutch “winter hun-ger” (1944 to 1945) had a slightly higher prevalenceof obesity (defined as weight/height .120% of theWorld Health Organization standards) at 19 years ofage (1.32% of subjects from the famine area vs 0.82%of controls were obese) if maternal malnutrition oc-curred during the first two trimesters, and a slightlyreduced incidence (1.45% of subjects from the faminearea vs 2.77% of controls were obese) if maternalmalnutrition occurred during the last trimester ofpregnancy or if the infant was exposed to famineconditions in the first 3 to 5 months of extrauterinelife.139 It was hypothesized that the small adipogeniceffects of early intrauterine malnutrition might beattributable to effects on hypothalamic developmentand that the small antiadipogenic effects of earlypostnatal malnutrition might be attributable to sup-pression of adipocyte generation. There is also someevidence that prenatal undernutrition or other pre-natal stresses, as evidenced by birth weight, mayaffect the prevalence of adiposity-related morbidity.Although they were not more frequently obese asadults, the prevalence of type II diabetes mellitus,hypertension, and hyperlipidemia (syndrome X) wasreported to be 10-fold greater in 50- to 70-year-oldmen and women who had low birth weights (,6.5

pounds) versus adults of the same age who had beenmore macrosomic as infants (birth weight .9.5pounds).140

Postnatal InfluencesIn 1968, Knittle and Hirsch141 reported that under-

feeding or overfeeding of preweanling rat pups wasassociated with respective decreases or increases inbody size, epididymal fat pad cell number, and cellsize, which persisted despite subsequent ad libitumfeeding of rat chow. The subsequent demonstrationthat obesity in humans may reflect adipocyte hyper-plasia as well as hypertrophy, and demonstration ofthe apparent persistence of fat cells despite weightloss,142 suggested that early nutritional factors mightinfluence later adiposity in humans. Currently, how-ever, there is no firm evidence that early infant-feeding practices significantly affect risk of later obe-sity. Although formula-fed infants tend to be longerand heavier than their breastfed peers, these differ-ences do not persist.143,144 Neither the age at whichspecific foods are introduced into the diet145–147 northe relative amount of fat, carbohydrate, or protein inthe diet131,148,149 exerts a significant influence on sub-sequent adiposity.

In the United States, obesity is most prevalentamong children raised in urban communities and insmaller families. The prevalence of obesity also var-ies by ethnic group, geographic region, and socioeco-nomic class.150–153 Davies et al154 noted significantnegative correlations between physical activity andbody fatness in preschool children and concludedthat low levels of physical activity in this age groupwere associated with increased body fat. Dietz andGortmaker155 analyzed data obtained in NHES II(1963 to 1965, children 6 to 11 years of age) andNHES III (1966 to 1970, children 12 to 17 years of age)and noted that adiposity and the amount of timespent watching television in adolescence were signif-icantly correlated, even when corrected for a historyof obesity. They concluded that television-watchingencourages inactivity and the consumption of calor-ically dense foods. Berkowitz et al156 noted that neo-natal fatness and physical activity were not signifi-cantly correlated with parental adiposity oradiposity of the same children at the age of 4 to 8years. The degree of adiposity in 4- to 8-year-oldswas significantly positively correlated with parentaladiposity and inversely correlated with daytime ac-tivity levels of the 4- to 8-year-olds. However, it isunclear whether inactivity and television-viewing inchildren cause obesity or whether obese children areless physically active because of social stigmatizationby their peers or other secondary factors.

DISCUSSION

Conclusions and Strategies for Research Into theEtiology of Obesity in Children

Obesity in adults is associated with increased riskof a wide variety of medical morbidities. Genetic andphysiologic data support the view that body fatnessis inherited and that attempts to alter body weighttherapeutically are opposed by changes in energy


metabolism. The striking increase in the prevalenceof obesity in the United States over the past fewdecades clearly indicates that body weights are de-fended in this manner, and that the effect of geneticinfluences on these body weights can be modified byenvironmental factors that favor positive caloric bal-ance.

These observations support the common-senseidea that environment has a powerful effect on bodyfatness. It is not known, however, whether an indi-vidual’s intrinsic regulatory processes for control ofbody fatness can be altered permanently by earlyenvironmental manipulation. Animal data suggestthat it may be possible to modify obesity risk inindividuals through early interventions. Identifica-tion of the child who is at risk for obesity by use ofhistorical (family history), clinical (current adiposityand growth pattern), genetic (allelic variations inspecific genes), biochemical (eg, leptin), and meta-bolic (eg, low energy expenditure) markers; identifi-cation of environmental factors that increase the ex-pression of this tendency toward obesity; anddevelopmental stages at which the environment canbe altered to prevent expression of such a tendencyshould be primary goals of future research.

Despite the fact that inheritance of most obesity inhumans is complex, the problem is tractable whenappropriate genetic strategies are used. Genetic map-ping is a process used to relate a specific phenotypeto a limited region of the genome and ultimately to asingle gene. Once a phenotype is mapped accurately,the region can be searched for known genes, or thegene can be cloned by examining directly all DNA inthe interval.33 This powerful strategy is most success-ful when the phenotype is unequivocal (affected vsunaffected, as for Huntington’s disease) and attrib-utable to the effects of a single, fully penetrant gene(phenotype always present when mutation ispresent). The existence of phenocopies (similar phe-notype not attributable to the gene) and multigenictraits (same phenotype attributable to differentgenes) can make it difficult or impossible to makesuch a map and—except in instances of extremeaffection status (eg, BMI .50) or an isolated popula-tion group (Pimans, Samoans)—obesity is not likelyattributable to variation in a single gene. Thus, forpurposes of genetic analysis, the phenotype shouldbe refined to isolate specific aspects of physiology forwhich interindividual difference based on geneticvariation could account for differences in the amountof body fat.

In addition to identifying a robust, distinct, phys-iologically relevant phenotype, it is necessary to havea map of high resolution with which to relate (map)the phenotype. Ideally, the map should be composedof molecular markers that are likely to differ betweenany two individuals. Such a requirement is fulfilledby the existence of dense (marker ;every 500 000base pairs) maps of the human157 and mouse158 ge-nomes. With such dense genetic maps, it has becomepossible to dissect a complex genetic disorder byso-called multipoint interval mapping of such quan-titative trait loci.159,160 Recent experiments on seizuresusceptibility161 and susceptibility to diet-induced

obesity in mice162,163 and hypertension164 and nonin-sulin-dependent diabetes mellitus in rats165,166 sug-gest that even complex quantitative traits can beresolved to relatively few genetic loci if those lociexert major effects on phenotype.

On the basis of considerations raised earlier, it isclear that obesity can arise only from an excess ofenergy intake over expenditure. Thus, the search forcandidate genes that might account for a geneticpredisposition to obesity should be focused on thosethat play a role in energy intake or expenditure. Thisstrategy is currently being used to examine extendedhuman families, sib pairs, and individuals withinpopulations for linkage of obesity and related phe-notypes167,168 to molecular markers corresponding tocandidate genes based on metabolic physiology (eg,b3-adrenoreceptor,41,169 glucocorticoid receptor42) androdent obesity genes (eg, Lep, Lepr).45,47,48,50

Another approach to the identification of genesrelated to the obese phenotype is to scan (for linkedgenetic intervals) the entire genome of individualswithin families segregating for an obese phenotype.The feasibility of such an approach depends on thenumber of genes responsible for the phenotype in theindividuals affected, the total number of relevantgenes in the population, and the relative risk attrib-utable to any locus. For disorders for which geneticsclearly indicates a single gene with high penetrance,this approach is a reasonable means of identifyingthe region of the genome where the responsible geneis located. However, most obesity in humans is prob-ably oligogenic, with genetic heterogeneity as well asstrong environmental influences on phenotypic ex-pression. Thus, in an ethnically heterogeneous pop-ulation, and allowing for the potent effects of age andenvironment on penetrance of the phenotype, such agenome scan (which might be successful in a genet-ically homogeneous population) would have lowpower to detect even relatively significant genes. Analternative approach is to identify candidate genesfor which biochemical or physiologic effects are con-sistent with possible molecular–physiologic mecha-nisms for the obese phenotype (eg, feeding peptidessuch as NPY, cholecystokinin, glucagon-like peptide1, MCH, galanin, and their cognate receptors; genesrelated to energy expenditure such as Na,K-ATPaseand mitochondrial uncoupling protein). These genesthen can be mapped relative to pertinent pheno-types170 or examined for variations in coding andregulatory sequences that correlate with these phe-notypes.

CONCLUSIONThe apparent increase in the fatness of adults and

children over the past few decades implies and thelack of success in maintenance of a reduced bodyweight in reduced-obese individuals suggests thatthe degree of body fatness an individual defendsmetabolically may be altered by early environmentalinfluences. Additional identification of systems thatoppose maintenance of an altered body weight, theage at which these systems become operant, and theenvironmental cues that affect the regulatory set-point of these systems may identify an optimal age at


which therapy could be instituted to prevent a childwho is genetically at risk of becoming obese fromexpressing that genetic tendency.

ACKNOWLEDGMENTThis work was supported by National Institutes of Health

Grants DK30583, DK52431, DK01983, and GCRC RR00102.

REFERENCES1. Frisch R, McArthur J. Menstrual cycles: fatness as a determinant of

minimum weight for height necessary for their maintenance or onset.Science. 1974;185:949–951

2. Neel J. Diabetes mellitus: a “thrifty” genotype rendered detrimental by“progress”? Am J Hum Genet. 1962;14:353–362

3. Mueller W. The genetics of human fatness. Yearbook Phys Anthropol.1983;26:215–230

4. Troiano RP, Flegal KM. Overweight children and adolescents: descrip-tion, epidemiology, and demographics. Pediatrics 1998;101(suppl):497–504

5. Kuczmarski R, Flegal K, Campbell S, Johnson C. Increasing prevalenceof overweight among US adults. JAMA. 1994;272:205–211

6. Rosenbaum M, Leibel R, Hirsch J. Medical progress: obesity. N EnglJ Med. 1997;337:396–407

7. Leibel R, Chung W, Chuo SC Jr. The molecular genetics of rodentsingle gene obesities. J Biol Chem. 1997;272:31937–31940

8. Knittle J. Adipose tissue development in man. In: Faulkner F, TannerJM, eds. Human Growth. Vol 2. Postnatal Growth. New York, NY: Ple-num Press; 1978

9. Rosso P. Prenatal nutrition and fetal growth and development. PediatrAnn. 1981;10:21–30

10. Johnston F. Sex differences in fat patterning in children and youth. In:Bouchard C, Johnston F, eds. Fat Distribution During Growth and LaterHealth Outcomes. New York, NY: Liss; 1988:85–102

11. Forbes G. Body composition in adolescence. In: Falkner F, Tanner JM,eds. Human Growth: A Comprehensive Treatise. Vol 2. New York, NY:Plenum; 1986

12. Leibel R, Berry E, Hirsch J. Biochemistry and development of adiposetissue in man. In: Conn HL Jr, DeFelice EA, Kuo P, eds. Health andObesity. New York, NY: Raven Press; 1983

13. Knittle J, Timmers K, Ginsberg-Fellner F, et al. The growth of adiposetissue in children and adolescents. J Clin Invest. 1979;63:239–246

14. Ailhaud G. Adipose cell differentiation in culture. Mol Cell Biochem.1982;49:17–31

15. Doglio A, Amri E, Dani C, et al. Effects of growth hormone in thedifferentiation process of preadipose cells. In: Isaksson O, Binder C,Hall K, eds. Growth Hormone—Basic and Clinical Aspects. Amsterdam,The Netherlands: Elsevier Science; 1987

16. Faust I, Miller WJ. Hyperplastic Growth of Adipose Tissue in Obesity. NewYork, NY: Raven Press; 1983:41–51

17. Hager A, Sjostrom L, Arvidsson V. Adipose tissue cellularity in obeseschool girls before and after dietary treatment. Am J Clin Nutr. 1978;31:68–75

18. Tontonoz P, Hu E, Spiegelman B. Stimulation of adipogenesis infibroblasts by PPAR gamma 2, a lipid-activated transcription factor.Cell. 1994;79:1147–1156

19. Spiegelman B, Flier J. Adipogenesis and obesity: rounding out the bigpicture. J Clin Invest. 1996;87:377–389

20. Tontonoz P, Hu E, Graves R, et al. mPPARg2: tissue specific regulatorof an adipocyte enhancer. Genes Dev. 1994;8:1224–1234

21. MacDougald O, Lane M. Transcriptional regulation of gene expressionduring adipocyte differentiation. Annu Rev Biochem. 1995;64:345–373

22. Hu E, Tontonoz P, Spiegelman B. Transdifferentiation of myoblasts bythe adipogenic transcription factors PPARg and C/EBPa. Proc NatlAcad Sci USA. 1995;92:9859–9860

23. Wang N, Finegold M, Bradley A, et al. Impaired energy homeostasis inC/EBP-a knockout mice. Science. 1995;2609:1108–1112

24. Wu Z, Bucher N, Farmer S. Induction of perioxisome proliferator-activated receptor g during conversion of 3T3 fibroblasts into adipo-cytes is mediated by C/EBP, C/EBPg, and glucocorticoids. Mol CellBiol. 1996;16:4128–4136

25. Kim J, Spiegelman B. ADD1/SRBEP1 promotes adipocyte differentia-tion and gene expression linked to fatty acid metabolism. Genes Dev.1996;10:1096–1107

26. Leibel R, Bahary N, Friedman J. Genetic variation and nutrition inobesity. In: Simopoulos A, Childs B, eds. Genetic Variation and Nutri-tion. Basel, Switzerland: Karger; 1990:90–101

27. Rosenbaum M, Leibel R. Pathophysiology of childhood obesity. AdvPediatr. 1988;35:73–137

28. Bouchard C, ed. Genetics of Obesity. Boca Raton, FL: CRC Press; 199429. Stunkard A, Foch T, Hrubec Z. A twin study of human obesity. JAMA.

1986;256:51–5430. Stunkard A, Sorensen T, Hanis C, et al. An adoption study of human

obesity. N Engl J Med. 1986;314:193–19831. Coleman D. Diabetes-obesity syndromes in mice. Diabetes. 1982;31:1–632. Coleman D, Eicher E. Fat (fat) and tubby (tub): two autosomal reces-

sive mutations causing obesity syndromes in the mouse. J Hered.1990;81:424–427

33. Leibel R, Chua S, Chung W. Animal models of genetic obesity. In:Angel A, Anderson H, Bouchard C, et al, eds. Progress in ObesityResearch, 7th International Congress on Obesity. London, UK: JohnLibbey; 1996:263–277

34. Chung W, Power-Kehoe L, Chua S, Leibel R. Mapping of the obreceptor to 1p in a region of nonconserved gene order from mouse andrat to human. Genome Res. 1996;6:431–438

35. Wolff G, Roberts D, Galbraith D. Prenatal determination of obesity,tumor susceptibility and coat color in viable Yellow (Avy/a) mice.J Hered. 1986;77:151–158

36. Fan W, Boston B, Kesterson R, et al. Role of melanocortinergic neuronsin feeding and the agouti obesity syndrome. Nature. 1997;385:119–120

37. Naggert J, Fricker L, Varlamov O, et al. Hyperproinsulinaemia in obesefat/fat mice associated with carboxypeptidase E mutation which re-duces enzyme activity. Nature Genet. 1995;10:135–142

38. Rovere C, Viale A, Nahon J, Kitabgi P. Impaired processing of brainproneurotensin and promelanin concentrating hormone in obese fat/fat mice. Endocrinology. 1996;137:2954–2958

39. Chua S, Leibel R. Obesity genes: molecular and metabolic mechanisms.Diabetes Rev. 1997;5:2–7

40. Bouchard C. Genetics of obesity: an update on molecular markers. IntJ Obes. 1995;19:S10–S13

41. Clement K, Vaisse C, Manning B, et al. Genetic variation in the beta-3-adrenergic receptor and an increased capacity to gain weight inpatients with morbid obesity. N Engl J Med. 1995;333:352–354

42. Clement K, Philippi A, Jury C, et al. Candidate gene approach offamilial morbid obesity-linkage analysis of the glucocorticoid receptorgene. Int J Obes. 1996;20:507–512

43. Deriaz O, Dionne F, Perusse L, et al. DNA variation in the genes of theNa, K-adenosine triphosphatase and its relation with resting metabolicrate, respiratory quotient, and body fat. J Clin Invest. 1994;93:838–843

44. Reed P, Ding Y, Xu W, et al. Human obesity does not segregate withthe chromosomal regions of Prader-Willi, Bardet-Biedl, Cohen, Bjore-son, or Wilson-Turner syndromes. Int J Obes. 1995;19:599–603

45. Clement K, Garner C, Hager J, et al. Indication for linkage of thehuman OB gene region with extreme obesity. Diabetes. 1996;45:687–690

46. Reed P, Ding Y, Xu W, et al. Extreme obesity may be linked to markersflanking the human ob gene. Diabetes. 1996;45:691–694

47. Norman R, Leibel R, Chung W, et al. Absence of linkage of obesity andenergy metabolism to homologs of rodent obesity genes in PimaIndians. Diabetes. 1996;45:1229–1232

48. Duggirala R, Stern M, Mitchell B, et al. Quantitative variation inobesity-related traits and insulin precursors linked to OB gene regionon human chromosome 7. Am J Hum Genet. 1996;59:694–703

49. Commuzzie A, Hixson J, Almasy L, et al. A major quantitative traitlocus determining serum leptin levels and fat mass is located onhuman chromosome 2. Nature Genet. 1997;15:1–3

50. Chagnon Y, Lamothe M, Perusse L, et al. Genetic linkages betweenmarkers on human 1p32–p22 and body fat and insulin levels. Obes Res.1997;5:115–121

51. Thompson B, Janssen R, Ossowski V, et al. Evidence for linkagebetween a region on chromosome 1p and the acute insulin response inPima Indians. Diabetes. 1995;44:478–481

52. Chung W, Power-Kehoe L, Chua M, et al. Exonic and intronic se-quence variation in the human leptin receptor gene (LEPR). Diabetes.1997;46:1509–1511

53. Kleiber M. The Fire of Life. New York, NY: Krieger; 1975:179–22254. Staffieri J. A study of social stereotype of body image in children. J

Perspec Soc Psychol. 1969;10:337–34355. Mayer J. Glucostatic mechanisms of the regulation of food intake.

N Engl J Med. 1953;249:13–1656. Mayer J. Regulation of energy intake and the body weight. The glu-

costatic theory and the lipostatic hypothesis. Ann NY Acad Sci. 1955;63:15–43

57. Kennedy G. The role of depot fat in the hypothalamic control of foodintake in the rat. Proc R Soc London [B]. 1953;140:578–591

58. Meyer J, Hargus W. Factors influencing food intake of rats fed low-


protein rations. Am J Physiol. 1959;197:1350–135259. Keesey R. Physiological regulation of body weight and the issue of

obesity. Med Clin North Am. 1989;73:15–2860. Belanger A, Cupples L, D’Agostino R. Means at each examination and

inter-examination consistency of specified characteristics: FraminghamHeart Study, 30 year follow-up. In: Kannel W, Wolf P, Garrison R, eds.The Framingham Study: An Epidemiologic Investigation of CardiovascularDisease, Section 36. Framingham Publication No 88-2970. Washington,DC: Government Printing Office; 1988

61. Wadden T. Treatment of obesity by moderate and severe caloric re-striction. Results of clinical research trials. Ann Intern Med. 1993;229:688–693

62. National Institutes of Health. Health implications of obesity: NationalInstitutes of Health development conference. Ann Intern Med. 1985;6.Supplement

63. Leibel R, Hirsch J. Diminished energy requirements in reduced-obesepatients. Metabolism. 1984;33:164–170

64. Leibel R, Rosenbaum M, Hirsch J. Changes in energy expenditureresulting from altered body weight. N Engl J Med. 1995;332:621–628

65. Weigle D, Brunzell J. Assessment of energy expenditure in ambulatoryreduced-obese subjects by techniques of weight stabilization and ex-ogenous weight replacement. Int J Obes. 1990;14:69–77

66. Weigle D, Sande K, Iverius P, et al. Weight loss leads to a markeddecrease in nonresting energy expenditure in ambulatory human sub-jects. Metabolism. 1988;37:930–936

67. Weinsier R, Nelson K, Hensrud D, et al. Metabolic predictors ofobesity. J Clin Invest. 1995;95:980–985

68. Segal K, Rosenbaum M, Chatr-Aryamontri B, et al. Impact of weightchange on muscle efficiency during exercise. FASEB J. 1995;9:A438.Abstract

69. Butera P, Bradway D, Cataldo N. Modulation of the satiety effect ofcholecystokinin by estradiol. Physiol Behav. 1993;53:1235–1238

70. Woods S, Chavez M, Park C, et al. The evaluation of insulin as ametabolic signal influencing behavior via the brain. Neurosci BiobehavRev. 1996;20:139–144

71. Bray G. Nutrient intake is modulated by peripheral peptide adminis-tration. Obes Res. 1995;3(suppl 4):569s–572s

72. Geary N. Role of gut peptides in meal regulation. In: Weston LA,Savage LM, eds. Obesity: Advances in Understanding and Treatment.Southborough, MA: International Business Communications; 1996

73. Trayhurn P. Brown adipose tissue and energy balance. In: Trayhurn P,Nicholls D, eds. Brown Adipose Tissue. Baltimore, MD: Edward Arnold;1986:299–338

74. Danforth E, Burger A. The role of thyroid hormones in the control ofenergy expenditure. Clin Endocrinol Metab. 1984;13:581–596

75. Rosenbaum M, Hirsch J, Leibel R. The effects of changes in bodyweight on carbohydrate metabolism, catecholamine excretion, andthyroid function. 1998; In press

76. Aronne L, Mackintosh R, Rosenbaum M, et al. Autonomic nervoussystem activity and energy expenditure during weight gain andweight loss. Am J Physiol. 1995;38:R222–R225

77. Bray G, York D. Genetically transmitted obesity in rodents. Physiol Rev.1971;51:598–646

78. Bray G. The inheritance of corpulence. In: Cioffi LA, ed. The BodyWeight Regulatory System Normal and Disturbed Mechanisms. New York,NY: Raven Press; 1981:185–195

79. Trayhurn P. The developmental of obesity in animals: the role ofgenetic susceptibility. Clin Endocrinol Metab. 1984;13:451–474

80. Roberts S, Savage J, Coward W, et al. Energy expenditure and intakein infants born to lean and overweight mothers. N Engl J Med. 1988;318:461–466

81. Ravussin E, Lillioja S, Knowler WC. Reduced rate of energy expendi-ture as a risk factor for body-weight gain. N Engl J Med. 1988;318:467–472

82. Ravussin E, Pratley R, Maffei M, et al. Relatively low plasma leptinconcentrations precede weight gain in Pima Indians. Nature Med. 1997;3:238–340

83. Davies P, Day J, Lucas A. Energy expenditure in early infancy and laterbody fatness. Int J Obes. 1991;15:727–731

84. Goran M, Figueroa R, McGloin A, et al. Obesity in children: recentadvances in energy metabolism and body composition. Obes Res. 1995;3:277–289

85. LeMagnen J. Neurobiology of Feeding and Nutrition. San Diego, CA:Academic Press; 1992

86. Birch L, Johnson S, Andresen G, et al. The variability of young chil-dren’s energy intake. N Engl J Med. 1991;324:232–235

87. Corbett S, Keesey R. Energy balance of rats with lateral hypothalamiclesions. Am J Physiol. 1982;242:E273–E279

88. Keesey R, Boyle P, Kemnitz J, Mitchel J. The role of the lateral hypo-thalamus in determining the body weight set-point. In: Novin D,Wyrwicka W, Bray G, eds. Hunger Basic Mechanisms and Clinical Impli-cations. New York, NY: Raven Press; 1975:243–255

89. Sclafani A. Eating rates in normal and hypothalamic and hypothalamichyperphagic rats. Physiol Behav. 1994;55:489–494

90. Vilberg T, Keesey R. Reduced energy expenditure after ventromedialhypothalamic lesions in female rats. Am J Physiol. 1984;247:R183–R188

91. Keesey R, Corbett S. Metabolic defense of the body weight set-point.In: Stunkard A, Stellar E, eds. Eating and Its Disorders. New York, NY:Raven Press; 1984:77–96

92. Vilberg T, Keesey R. Ventromedial hypothalamic lesions abolish com-pensatory reduction in energy expenditure to weight loss. Am J Physiol.1990;258:R476–R480

93. Dulloo A, Girardier L. 24 Hour energy expenditure several monthsafter weight loss in the underfed rat: evidence for a chronic decrease inwhole-body metabolic efficiency. Int J Obes. 1993;17:115–123

94. Grossman S. Role of the hypothalamus in the regulation of food andwater intake. Psychol Rev. 1975;82:200–224

95. Hoebel B, Teitlebaum P. Weight regulation in normal and hypotha-lamic hyperphagic rats. J Comp Physiol Psychol. 1966;68:536–543

96. Harris R. Role of set-point theory in regulation of body weight. FASEBJ. 1990;4:3310–3318

97. Keesey R, Corbett S, Hirvonen M, Kaufman L. Heat production andbody weight changes following lateral hypothalamic lesions. PhysiolBehav. 1984;32:309–317

98. Powley T, Laughton W. Neuronal pathways involved in the hypotha-lamic regulation of autonomic responses. Diabetologia. 1981;20:378–387

99. Sclafani A, Springer D. Dietary obesity in adult rats: similarities to hypo-thalamic and human obesity syndromes. Physiol Behav. 1976;17:461–471

100. Sclafani A. The role of hyperinsulinemia and the vagus nerve inhypothalamic hyperphagia reexamined. Diabetologia. 1981;20:402–410

101. Bray G, Inoue S, Nishizawa Y. Hypothalamic obesity: the autonomichypothesis and the lateral hypothalamus. Diabetologia. 1981;20:366–377

102. Bray G, Gallagher T. Manifestations of hypothalamic obesity in man: acomprehensive investigation of eight patients and a review of theliterature. Medicine. 1975;54:301–330

103. Frankish H, Dryden S, Hopkins D, et al. Neuropeptide Y, the hypo-thalamus, and diabetes: insights into the central control of metabolism.Peptides. 1995;16:757–771

104. Billington C, Briggs J, Harker S, et al. Neuropeptide Y in hypothalamicparaventricular nucleus: a center coordinating energy metabolism.Am J Physiol. 1994;266:R1765–R1770

105. Zukowska-Grojec Z. Neuropeptide Y: a novel sympathetic stress hor-mone and more. Ann NY Acad Sci. 1995;771:219–233

106. Chua S, Brown A, Kim J, et al. Food deprivation and hypothalamicneuropeptide gene expression: effects of strain background and thediabetes mutation. Mol Brain Res. 1991;11:291–299

107. Erickson J, Clegg K, Palmiter R. Sensitivity to leptin and susceptibilityto seizures of mice lacking neuropeptide Y. Nature. 1996;381:415–418

108. Blundell J. Pharmacological approaches to appetite suppression.Trends Physiol Sci. 1991;12:147–157

109. Foster JWD, ed. Williams Textbook of Endocrinology. 8th ed. Philadel-phia, PA: WB Saunders; 1992

110. Rohner-Jeanrenaud F. A neuroendocrine reappraisal of the dual-centrehypothesis: its implications for obesity and insulin resistance. Int JObes. 1995;19:517–534

111. Kaiyala K, Woods S, Schwartz M. New model for the regulation ofadiposity and energy balance by the central nervous system. Am J ClinNutr. 1995;62:1123S–1134S

112. Zhang Y, Proenca Y, Maffei M, et al. Positional cloning of the mouseobese gene and its human homologue. Nature. 1994;372:425–432

113. Frederich R, Hamann A, Anderson S, et al. Leptin levels reflect bodylipid content in mice: evidence for diet-induced resistance to leptinaction. Nature Med. 1995;1:1311–1314

114. Rosenbaum M, Nicolson M, Hirsch J, et al. Effects of gender, bodycomposition, and menopause on plasma concentrations of leptin. J ClinEndocrinol Metab. 1996;81:3424–3427

115. Halaas J, Gajiwala K, Maffei M, et al. Weight reducing effects of theplasma protein encoded by the ob gene. Science. 1995;269:543–546

116. Pellymounter M, Cullen M, Baker M, et al. Effects of the obese geneproduct on body weight regulation in ob/ob mice. Science. 1995;269:540–543

117. Campfield L, Smith F, Guisez Y, et al. Recombinant mouse OB protein:evidence for a peripheral signal linking adiposity and central neuralnetworks. Science. 1995;269:546–548

118. Collins S, Kuhn C, Petro A, et al. Role of leptin in fat regulation. Nature.1996;380:677


119. Rentsch J, Chiesei M. Regulation of ob gene mRNA in cultured adi-pocytes. FEBS Lett. 1996;379:55–59

120. Slieker L, Sloop K, Surface P, et al. Regulation of expression of OBmRNA proteins by glucocorticoids and cAMP. J Biol Chem. 1996;271:5301–5304

121. Zarjevski N, Cusin I, Vettor F, Rohner-Jeanrenaud F. Chronic intrace-rebroventricular neuropeptide-Y administration to normal rats mimicshormonal and metabolic changes of obesity. Endocrinology. 1993;133:1753–1758

122. Stephens T. Life without neuropeptide Y. Nature. 1996;381:377–378123. Schwartz M, Baskin D, Bukowski T, et al. Specificity of leptin action on

elevated blood glucose levels and hypothalamic neuropeptide Y geneexpression in ob/ob mice. Diabetes. 1996;45:531–535

124. Segal K, Landt M, Kline S. Relationship between insulin sensitivity andplasma leptin concentration in lean and obese men. Diabetes. 1996;45:988–991

125. Stern I, Batchelor B, Hollander N, et al. Adipose cell size and immu-noreactive insulin levels in obese and normal weight adults. Lancet.1972;II:948–951

126. Chua S, Chung W, Wu-Peng S, et al. Phenotypes of mouse diabetesand rat fatty due to mutations in the OB (leptin) receptor. Science.1996;271:994–996

127. Flatt J. Use and storage of carbohydrate and fat. Am J Clin Nutr.1995;61:952S–959S

128. Jebb S, Goldberg G, Coward W, et al. Effects of weight cycling causedby intermittent dieting on metabolic rate and body composition inobese women. Int J Obes. 1991;15:367–374

129. Rodin J, Radke-Sharpe N, Rebuffe-Scrive M, Greenwood M. Weightcycling and fat distribution. Int J Obes. 1989;14:303–310

130. Hill J, Drougas H, Peters J. Obesity treatment: can diet compositionplay a role. Ann Intern Med. 1994;120:520–523

131. Leibel R, Hirsch J, Appel B, Checani G. Energy intake required tomaintain body weight is not affected by wide variation in diet com-position. Am J Clin Nutr. 1992;55:350–355

132. NTF (National Task Force on the Prevention and Treatment of Obesity,NIH weight cycling. JAMA. 1994;272:1196–1202

133. Vohr B, Lipsitt L, Oh W. Somatic growth of children of diabeticmothers with reference to birth size. J Pediatr. 1980;97:196–199

134. Pettit D, Baird H, Allech K, Knowler W. Excessive obesity in offspringof Pima Indian women with diabetes during pregnancy. N Engl J Med.1983;08:242–245

135. Jones A, Simson E, Friedman M. Gestational undernutrition and thedevelopment of obesity in rats. J Nutr. 1984;114:1484–1492

136. Anguita R, Sigulen D, Sawaga A. Intrauterine food restriction is asso-ciated with obesity in young rats. J Nutr. 1993;123:1421–1428

137. Jones A, Assimon S, Friedman M. The effect of diet on food intake andadiposity in rats made obese by gestational undernutrition. PhysiolBehav. 1986;37:381–386

138. Aubert R, Suquet J, Lemmonier O. Long-term morphological andmetabolic effects of early under- and over-nutrition in mice. J Nutr.1980;110:649–661

139. Ravelli G, Stein Z, Susser M. Obesity in young men after famineexposure in utero and early infancy. N Engl J Med. 1976;295:349–353

140. Barker D, Hales C, Fall C, et al. Type 2 (non-insulin dependent)diabetes mellitus, hypertension, and hyperlipidemia (syndrome X):relation to reduced fetal growth. Diabetologia. 1993;36:62–67

141. Knittle J, Hirsch J. Effect of early nutrition on the development of ratepididymal fat pads: cellularity and metabolism. J Clin Invest. 1968;47:2091–2098

142. Hirsch J, Batchelor B. Adipose tissue cellularity in human obesity. ClinEndocrinol Metab. 1976;5:299–311

143. Fomon S, Thomas L, Filer L, et al. Food consumption and growth ofnormal infants fed mild-based formulas. Acta Paediatr Scand Suppl.1971;223:1–36

144. Fomon S, Rogers R, Ziegler E, et al. Indices of fatness and serumcholesterol at age eight years in relation to feeding and growth duringearly infancy. Pediatr Res. 1984;18:1233–1238

145. Borjeson M. The aetiology of obesity in children. Acta Pediatr Scand.1976;65:279–287

146. DeSwiet M, Fayers P, Cooper L. Effect of feeding habit on weight ininfancy. Lancet. 1977;II:892–894

147. Dubois S, Hill D, Beaton G. An examination of factors believed to beassociated with infantile obesity. Am J Clin Nutr. 1979;32:1997–2004

148. Enzi G, Inelman E, Rubaltelli F, et al. Postnatal development of adiposetissue in normal children on strictly controlled caloric intake. Metabo-lism. 1982;31:1029–1034

149. Butte C, Garza C, Smith E, Nichols B. Human milk intake and growthin exclusively breast fed infants. J Pediatr. 1984;104:187–195

150. Dietz W, Gortmaker S. Factors within the physical environment asso-ciated with childhood obesity. Am J Clin Nutr. 1984;39:619–624

151. Garn S, Clark D. Trends in fatness and the origins of obesity. Pediatrics.1976;57:443–456

152. Dietz W. Childhood obesity: susceptibility, cause, and management.J Pediatr. 1983;103:676–686

153. Patterson R, Typpo J, Typpo M, Krause G. Factors related to obesity inpreschool children. J Am Diet Assoc. 1986;86:1376–1381

154. Davies P, Gregory J, White A. Physical activity and body fatness inpre-school children. Int J Obes. 1995;19:6–10

155. Dietz W, Gortmaker S. Do we fatten our children at the television set?Obesity and television viewing in children and adolescents. Pediatrics.1985;75:807–812

156. Berkowitz R, Agras W, Korner A, et al. Physical activity and adiposity:a longitudinal study from birth to childhood. J Pediatr. 1985;106:734–738

157. Collins F, Galas D. A new five-year plan for the U. S. human genomeproject. Science. 1993;262:43–46

158. Genetic Map of the Mouse. Boston, MA: Whitehead Institute and MITCenter of Genome Research; 1995

159. Lander E, Botstein D. Mapping Mendelian factors underlying quanti-tative traits using RFLP linkage maps. Genetics. 1989;121:185–199

160. Lander E, Schork N. Genetic dissection of complex traits. Science.1994;265:2037–2048

161. Rise M, Frankel W, Coffin J, Seyfried T. Genes for epilepsy mapped inthe mouse. Science. 1991;253:669–673

162. Warden C, Fisler J, Shoemaker S, et al. Identification of four chromo-somal loci determining obesity in a multifactorial model. J Clin Invest.1995;95:1545–1552

163. West D, Goudey-Lefevre J, York B, Truett G. Dietary obesity linked togenetic loci on chromosomes 9 and 15 in a polygenic mouse model.J Clin Invest. 1994;84:1410–1416

164. Jacob H, Lindpaintner K, Lincoln S, et al. Genetic mapping of a genecausing hypertension in the stroke-prone spontaneously hypertensiverat. Cell. 1991;67:213–224

165. Gauguier D, Froguel P, Parent V, et al. Chromosomal mapping of lociassociated with non-insulin dependent diabetes mellitus in the GK rat.Nature Genet. 1996;12:4–6

166. Galli J, Li L, Glaser A, et al. Genetic analysis of non-insulin dependentdiabetes mellitus in the GK rat. Nature Genet. 1996;12:31–37

167. Bouchard C, Perusse L. Genetic aspects of obesity. Ann NY Acad Sci.1993;699:26–35

168. Bouchard C, Perusse L. Genetics of obesity. Ann Rev Nutr. 1993;13:337–354

169. Widen E, Lehto M, Kanninen T, et al. Association of a polymorphismin the beta-3-adrenergic receptor gene with features of insulin resis-tance syndrome in Finns. N Engl J Med. 1995;333:348–351

170. Collins F. Positional cloning moves from perditional to traditional.Nature Genet. 1995;9:347–350

171. Tartaglia L, Dembski M, Weng X, et al. Identification and cloning of aleptin receptor, OB-r. Cell. 1995;83:1263–1271

172. Noben-Trauth K, Naggert J, North M, Nishina P. A candidate for themouse mutation tubby. Nature. 1996;380:534–538

173. Bultman S, Michaud E, Woychik R. Molecular characterization of themouse agouti locus. Cell. 1992;71:1195–1204

174. Blaak E, Baak Mv, Kempen K, Saris W. Role of alpha- and beta-adrenoreceptors in sympathetically mediated thermogenesis. Am JPhysiol. 1993;264:E11–E17

175. Schwartz M, Sipols A, Marks J, et al. Inhibition of hypothalamicneuropeptide Y gene expression by insulin. Endocrinology. 1992;130:3608–3616

176. Levine A, Billington C. Peptides in regulation of energy metabolismand body weight. In: Bouchard C, Bray G, eds. Regulation of BodyWeight: Biological and Behavioral Mechanisms. West Sussex, UK: Wiley;1996:179–192

177. Morley J, Levine A. Opioids, dopamine and feeding. In: Hoebel B,Novin D, eds. The Neural Basis of Feeding and Reward. Brunswick, ME:1982:499–506

178. Dickson P, Vaccarino F. GRF-induced feeding: evidence for proteinselective and opiate involvement. Peptides. 1994;15:1343–1352

179. Feifel D, Vaccarino F. Growth hormone regulatory peptides (GHRHand somatostatin): a model for integration of central and peripheralfunction. Neurosci Biobehav Rev. 1994;18:421–433

180. Qu D, Ludwig D, Gammeltoft S, et al. A role for melatonin-concentrating hormone in the central regulation of feeding behavior.Nature. 1996;380:243–247

181. Lader M, Petursson H. Rational use of anxiolytic/sedative drugs.Drugs. 1983;25:514–528


182. Gibbs J, Young R, Smith G. Cholecystokinin decreases food intake inrats. J Comp Physiol Psychol. 1973;84:488–495

183. Rohner-Jeanrenaud F, Walker C, Perotto R, Jeanrenaud B. Chronic intra-cerebroventricular administration of corticotropin releasing factor (CRF)to genetically obese fa/fa rats arrests the further increase in their bodyweight. In: Bjorntorp P, Rossner SJ, eds. Obesity in Europe ’88: Proceedingsof European Congress on Obesity. London, UK: John Libbey; 1988

184. Spina M, Merlo-Pich E, Chan R, et al. Appetite-suppressing effects ofurocortin, a CRF-related neuropeptide. Science. 1996;273:1561–1564

185. Geary N. Pancreatic glucagon signals postprandial satiety. NeurosciBiobehav Rev. 1990;14:323–338

186. Turton M, O’Shea D, Gunn I, et al. A role for glucagon-like peptide-1

in the central regulation of feeding. Nature. 1996;379:69–72187. Kamatchi L, Veeraragavan K, Chandra D, Bapna J. Antagonism of

acute feeding response to 2-deoxyglucose and 5-thioglucose by GABAantagonists: the relative role of ventromedial and lateral hypothala-mus. Pharmacol Biochem Behav. 1986;25:59–62

188. Campfield L, Smith F. Functional coupling between transient declinesin blood glucose and feeding behavior: temporal relationships. BrainRes Bull. 1986;17:427–433

189. Oomura Y. Glucose as a regulator of neuronal activity. Adv Metab Dis.1983;10:31–65

190. VanItallie T. The glucostatic theory 1953–1988: roots and branches. IntJ Obes. 1990;14:1–10

Development of Eating Behaviors Among Children and Adolescents

Leann L. Birch, PhD, and Jennifer O. Fisher, PhD

ABSTRACT. The prevalence of obesity among childrenis high and is increasing. We know that obesity runs infamilies, with children of obese parents at greater risk ofdeveloping obesity than children of thin parents. Researchon genetic factors in obesity has provided us with estimatesof the proportion of the variance in a population accountedfor by genetic factors. However, this research does not pro-vide information regarding individual development. To de-sign effective preventive interventions, research is neededto delineate how genetics and environmental factors inter-act in the etiology of childhood obesity. Addressing thisquestion is especially challenging because parents provideboth genes and environment for children.

An enormous amount of learning about food and eat-ing occurs during the transition from the exclusive milkdiet of infancy to the omnivore’s diet consumed by earlychildhood. This early learning is constrained by chil-dren’s genetic predispositions, which include the un-learned preference for sweet tastes, salty tastes, and therejection of sour and bitter tastes. Children also are pre-disposed to reject new foods and to learn associationsbetween foods’ flavors and the postingestive conse-quences of eating. Evidence suggests that children canrespond to the energy density of the diet and that al-though intake at individual meals is erratic, 24-hour en-ergy intake is relatively well regulated. There are indi-vidual differences in the regulation of energy intake asearly as the preschool period. These individual differ-ences in self-regulation are associated with differences inchild-feeding practices and with children’s adiposity.This suggests that child-feeding practices have the po-tential to affect children’s energy balance via alteringpatterns of intake. Initial evidence indicates that impo-sition of stringent parental controls can potentiate pref-erences for high-fat, energy-dense foods, limit children’sacceptance of a variety of foods, and disrupt children’sregulation of energy intake by altering children’s respon-siveness to internal cues of hunger and satiety. This canoccur when well-intended but concerned parents assumethat children need help in determining what, when, and

how much to eat and when parents impose child-feedingpractices that provide children with few opportunitiesfor self-control. Implications of these findings for pre-ventive interventions are discussed. Pediatrics 1998;101:539–549; children, food, eating, obesity, dieting, parenting.

ABBREVIATION. BMI, body mass index.

This article addresses behavioral factors that in-fluence food preferences, food intake, and en-ergy regulation in children. Currently, the

prevalence of childhood overweight is high and hasincreased dramatically since the 1970s.1,2 This in-creased prevalence is of concern because overweightchildren are at increased risk for social stigmatiza-tion, adult obesity, and chronic disease. Obesityshows familial aggregation; the risk of obesityamong children of two obese parents is much higherthan for children in families in which neither parentis obese.3 Familial aggregation has focused researchattention on genetic factors in obesity,4 but the rapidsecular increase in the prevalence of obesity cannotbe attributable to genetic factors. The interaction ofgenes and environment influences phenotypes forintake and expenditure and suggests that a renewedfocus on the family environment may provide infor-mation about behavioral factors that contribute tofamilial aggregation of adiposity.

Between 30% and 50% of the variance in adipositywithin a population is attributable to genetic differ-ences.5 However, these heritability estimates de-scribe populations, not individuals, and do not pro-vide information about the ways genetics andenvironment interact during development to pro-duce childhood obesity. In the case of childhoodobesity, the question is especially complex because 1)genes and environments also tend to be correlated—parents typically provide children with both geneticsand environment; and 2) genetic factors can includebehavioral predispositions that affect food intakeand expenditure.4 A more complete understandingof what may characterize obesigenic environments

From the Department of Human Development and Family Studies and theGraduate Program in Nutrition, Pennsylvania State University, UniversityPark, Pennsylvania.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;525Pediatrics Michael Rosenbaum and Rudolph L. Leibel

in ChildrenThe Physiology of Body Weight Regulation: Relevance to the Etiology of Obesity

ServicesUpdated Information &

http://pediatrics.aappublications.org/content/101/Supplement_2/525including high resolution figures, can be found at:

References

BIBLhttp://pediatrics.aappublications.org/content/101/Supplement_2/525#This article cites 165 articles, 33 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 January 10, 2019www.aappublications.org/newsDownloaded from

http://http://pediatrics.aappublications.org/content/101/Supplement_2/525

http://pediatrics.aappublications.org/content/101/Supplement_2/525#BIBL

http://pediatrics.aappublications.org/content/101/Supplement_2/525#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;525Pediatrics Michael Rosenbaum and Rudolph L. Leibel

in ChildrenThe Physiology of Body Weight Regulation: Relevance to the Etiology of Obesity

http://pediatrics.aappublications.org/content/101/Supplement_2/525located 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 January 10, 2019www.aappublications.org/newsDownloaded from

http://pediatrics.aappublications.org/content/101/Supplement_2/525

the physiology of body weight regulation: relevance - pediatrics

Documents