RESEARCH

Cu

S itiH CLIN PhD;

ABOberysizintsatDesupSub45SetInt(hebyMaindergderbodtioStaRessizbotmo8%grogro

compamaimConpadieduanrisJ A

ChereaingcriconforassingmanaingD

cretinlonweofimcouchasocsugcatrislargely results from a sedentary lifestyle, poor nutrition,weight cycling, and/or other lifestyle habits, as opposed to

L.turDapaCApaogytheforgist, and N. L. Keim is a research chemist with the USDepartment of Agriculture, Agricultural Research Ser-vicCAA

S-8cislbaC

As0d

2solely reflecting adiposity itself.Critics of the diet to improve health model suggest a

paradigm shift in treating weight-related concerns. Theyrecommend focusing on health behavior change as op-posed to a primary focus on weight loss (6,13,14). Theirapproach is supported by increasing evidence that dis-eases associated with obesity can be reversed or mini-mized through lifestyle change, even in the absence ofweight change, and that people can improve their healthwhile remaining obese (5,10,12,15).

e, Western Human Nutrition Research Center, Davis,.ddress correspondence to: Linda Bacon, PhD, Box0, Biology Department, City College of San Fran-co, 50 Phelan Ave, San Francisco, CA 94112. E-mail:con@ccsf.eduopyright 2005 by the American Dietetic

sociation.002-8223/05/10506-0001$30.00/0oi: 10.1016/j.jada.2005.03.011

005 by the American Dietetic Association Journal of the AMERICAN DIETETIC ASSOCIATION 929rrent Research

ize Acceptance and Intuealth for Obese, Female

DA BACON, PhD; JUDITH S. STERN, ScD; MARTA D. VAN LOAN,

STRACTjective Examine a model that encourages health at ev-size as opposed to weight loss. The health at every

e concept supports homeostatic regulation and eatinguitively (ie, in response to internal cues of hunger,iety, and appetite).ign Six-month, randomized clinical trial; 2-year follow-.jects White, obese, female chronic dieters, aged 30 toyears (N78).ting Free-living, general community.erventions Six months of weekly group interventionalth at every size program or diet program), followed6 months of monthly aftercare group support.in outcome measures Anthropometry (weight, body massex), metabolic fitness (blood pressure, blood lipids), en-y expenditure, eating behavior (restraint, eating disor-pathology), and psychology (self-esteem, depression,y image). Attrition, attendance, and participant evalua-ns of treatment helpfulness were also monitored.tistical analysis performed Analysis of variance.ults Cognitive restraint decreased in the health at everye group and increased in the diet group, indicating thath groups implemented their programs. Attrition (6nths) was high in the diet group (41%), compared within the health at every size group. Fifty percent of bothups returned for 2-year evaluation. Health at every sizeup members maintained weight, improved in all out-

Bacon is an associate nutritionist with the Agricul-al Experiment Station, University of California,vis, and a nutrition professor with the Biology De-rtment, City College of San Francisco, San Francisco,. J. S. Stern is a distinguished professor with the De-rtment of Nutrition and the Division of Endocrinol-, Clinical Nutrition and Cardiovascular Medicine inDepartment of Internal Medicine, University of Cali-nia, Davis. M. D. Van Loan is a research physiolo-ve Eating Improvehronic Dieters

NANCY L. KEIM, PhD

e variables, and sustained improvements. Diet grouprticipants lost weight and showed initial improvement inny variables at 1 year; weight was regained and littleprovement was sustained.clusions The health at every size approach enabledrticipants to maintain long-term behavior change; thet approach did not. Encouraging size acceptance, re-ction in dieting behavior, and heightened awarenessd response to body signals resulted in improved healthk indicators for obese women.m Diet Assoc. 2005;105:929-936.

oncern regarding obesity continues to mount amonggovernment officials, health professionals, and thegeneral public. Obesity is associated with physical

alth problems, and this fact is cited as the primaryson for the public health recommendations encourag-weight loss (1). That dieting and weight loss are

tical to improving ones health is reinforced by a socialtext that exerts enormous pressure on women to con-m to a thin ideal. Public attention to weight and itsociated comorbidities continues to increase, and diet-is now firmly ensconced in our cultural identity. Thejority of US women are now dieting: 57% stated in ational telephone survey that they are currently engag-in weight-control behaviors (2).espite heightened attentiveness to obesity and the in-ase in dieting behavior (3), the incidence of obesity con-ues to rise (4). There are little data showing improvedg-term success for the majority of those engaged inight-loss behaviors (5). Some have challenged the abilitydiet programs to either achieve lasting weight loss or toprove health, and question the ethics and value of en-raging dieting as an obesity intervention (5-9). Othersllenge the primacy of weight loss in addressing the as-iated health risks, regardless of method (10-12). Theygest that while the epidemiologic research clearly indi-es an association between obesity and health risk, theks of obesity may be overstated, and the association

Ato(ieapthrintsizpreT

of aprevensizabexppabod

MEProApthowh3(18nodiarenInsDaE

qudeantreran

TreTwanincFofer

DieThbatudpafatfoointtheeaMafatshstrexpLE

HeaThmeactingwelestanbehresnaniqintcuestachotheadhecomactsupseelarasswatiohadocreiinfstrDiscr

EvaPawewe

AnPaha

FigAs

930n alternative obesity treatment model teaches peoplesupport homeostatic regulation and eating intuitively, in response to internal cues of hunger, satiety, andpetite) instead of cognitively controlling food intakeough dieting (16). An essential component of someuitive eating programs is to encourage health at everye (Figure 1) rather than weight loss as a necessarycondition to improved health.his study was undertaken to examine the effectivenesshealth at every size approach in improving health. [In avious report (17), we referred to this as a nondiet inter-tion. This has since been changed to health at everye to reflect the changing terminology in the field.] Met-olic fitness (blood pressure and blood lipid levels), energyenditure, eating behavior (restraint and eating disorderthology), and psychology (self-esteem, depression, andy image) were evaluated.

THODScedureplicants were recruited from the Davis, CA area, andse meeting the following inclusion criteria were enrolled:ite; female; aged 30 to 45 years; body mass index (BMI)0; nonsmoker; not pregnant or lactating; Restraint Scale) score 15 (indicating a history of chronic dieting); andrecent myocardial infarction, active neoplasms, type 1betes, type 2 diabetes, or history of cardiovascular oral disease. The research protocol was approved by thetitutional Review Board of the University of California,vis, and informed consent was obtained.nrolled participants (N78) were divided into BMI

artiles and high/low sets for dietary restraint (18),grees of flexible and rigid control of eating (19), age,d self-reported activity level to ensure balance in theatment groups. Participants in these subgroups weredomly assigned to one of two treatment groups.

atment Conditionso treatment conditions were investigated: a diet groupd a health at every size group. Both treatment groupsluded 24 weekly sessions, each 90 minutes in length.llowing this, six monthly aftercare sessions were of-ed, described as optional group support.

Accepting and respecting the diversity of body shapes andsizes.Recognizing that health and well-being are multidimensionaland that they include physical, social, spiritual, occupational,emotional, and intellectual aspects.Promoting eating in a manner which balances individualnutritional needs, hunger, satiety, appetite, and pleasure.Promoting individually appropriate, enjoyable, life-enhancingphysical activity, rather than exercise that is focused on a goalof weight loss.Promoting all aspects of health and well-being for people of allsizes.

ure 1. Basic guiding principles of the health at every size program.drafted by the Association for Size Diversity and Health (30).June 2005 Volume 105 Number 6t Groupe focus of the diet group was similar to most behavior-sed weight-loss programs: eating behaviors and atti-es, nutrition, social support, and exercise. Partici-nts were taught to moderately restrict their energy andintake, and to reinforce their diets by maintainingd diaries and monitoring their weight. Exercise at anensity within the training heart range delineated inAmerican College of Sports Medicine/Centers for Dis-

se Control and Prevention guidelines was encouraged.terial was presented on topics including how to countgrams and exchanges, understanding food labels,

opping for food, the benefits of exercise, and behaviorategies for success. The program was taught by anerienced registered dietitian and reinforced using theARN Program for Weight Control manual (20).

lth at Every Size Groupere were five aspects to the health at every size treat-nt program: body acceptance, eating behavior, nutrition,ivity, and social support. The initial focus was on enhanc-body acceptance and self-acceptance, and participants

re supported in leading as full a life as possible, regard-s of BMI. The goal was to first help participants disen-gle feelings of self-worth from their weight. The eatingavior component supported participants in letting go oftrictive eating behaviors and replacing them with inter-lly regulated eating. Participants were educated in tech-ues that allowed them to become more sensitized toernal cues and to decrease their vulnerability to externals. The nutrition component educated participants aboutndard nutrition information and the effects of foodices on well-being, and supported them in temperingir food choices with foods that honored good health (indition to their taste preferences). The activity componentlped participants identify and transform barriers to be-ing active (eg, attitudes toward their bodies) and to findivity habits that allowed them to enjoy their bodies. Theport group element was designed to help participantstheir common experiences in a culture that devalues

ge women, and to gain support and learn strategies forerting themselves and effecting change. The programs facilitated by a counselor who had conducted educa-nal and psychotherapeutic workshops and groups andd completed all of the coursework necessary to obtain atorate in physiology with a focus on nutrition. It wasnforced with a written manual that provided detailedormation and practical advice for implementing theategies (Bacon L. Hungry Nation: Why the All-Americanet Will Never Satisfy Your Appetite, unpublished manu-ipt).

luation/Outcome Measuresrticipants attended five testing sessions: baseline, 12eks (midtreatment), 26 weeks (posttreatment), 52eks (postaftercare), and 104 weeks (follow-up).

thropometric and Metabolic Fitness Measuresrticipants reported to the laboratory in the morning,ving abstained from food, beverages, or vigorous activ-

ityelemodusamlowsit

EneThadicarivthemiterrat

EatThcogDiassshdischordtru

PsyThin

surprobeh

StaPo

Fig s. aMfor at least 12 hours. Weight was measured on anctronic scale and height was measured using a wall-unted stadiometer. Blood pressure was assessed inplicate using the oscillometric technique. Fasting bloodples were analyzed for blood lipids (total cholesterol,-density lipoprotein [LDL] cholesterol, and high-den-y lipoprotein [HDL] cholesterol).

rgy Expendituree Stanford Seven-Day Physical Activity Recall (21) wasministered by interview to evaluate time spent in phys-l activity. A summary of energy expenditure was de-ed by multiplying the average time of each activity byaverage intensity in metabolic equivalents. To mini-

ze interexaminer error and reduce variability, all in-views were conducted by two examiners, who collabo-ed to achieve consistent scoring.

ing Behavior Measurese Eating Inventory (22) consists of three subscales:nitive restraint, disinhibition, and hunger. The Eatingsorder Inventory-2 (23) contains eight subscales: threeess attitudes and behaviors toward weight, bodyape, and eating (drive for thinness, bulimia, and bodysatisfaction); five measure more general psychologicalaracteristics that are clinically relevant to eating dis-ers (ineffectiveness, perfectionism, interpersonal dis-st, interoceptive awareness, and maturity fears).

chological Measurese Beck Depression Inventory (24) measures alterationsmood and self-concept. The Rosenberg Self-Esteem Mea-

ure 2. Flow chart illustrating diet vs health at every size trial proceduree (25) focuses on a self-evaluation of approval or disap-val. The Body Image Avoidance Questionnaire assessesaviors associated with negative body image (26).

tistical Methodswer analyses conducted on the Rosenberg Self-Esteemasure and Beck Depression Inventory from two healthevery size studies (27,28) determined that 20 partici-nts per treatment group (n40 total) were needed totect a difference of 0.75 standard deviations betweenups with 80% power. We attempted to recruit 80 par-ipants to allow for 50% attrition.ll analyses were conducted using Statistica (version, 1996, Statsoft, Inc, Tulsa, OK). Students t test wased to compare baseline characteristics between groups.peated measures analysis of variance with a within-bject factor of time (four levels: baseline, 26 weeks, 52eks, and 104 weeks) and a between-subject factor ofup (two levels: diet and health at every size) was runtest differences in variables. Significance was set at.05. A least significant difference post-hoc test was runany variable that indicated significant difference.

ULTSless otherwise specified, the reported results includeparticipants for whom data were available at follow-: 19 participants from the health at every size groupd 19 participants from the diet group, or 50% of eachginal sample. The 19 participants who returned forlow-up testing in the health at every size group allpleted the 26-week program, whereas the 19 partici-

nts who returned for testing in the diet group included

easurements taken at each time point. bHAEShealth at every size.MeatpadegroticA

5.1usResuwegrotoPon

RESUnallupanorifolcompaJune 2005 Journal of the AMERICAN DIETETIC ASSOCIATION 931

16drosigclustawesmfor

ParTarepenpro

AttPrhatreeve

WeAsmestude

bainicartheena p

BloAsmefolupinsigcarprodielinboB

blohefolach(Pblo

ActThincatwaatfacsuactevemosigturat

EatBostrdirinincaftsuoftwontwdeonhu(loruimme

Tplo

C

AWB

EdHSCEmNEmJoPrCTePhORMS

aBbR

932program completers, and three participants who hadpped out of the program (see Figure 2). The statisticalnificance is not different when the dropouts are ex-ded from the analysis (although average values andndard deviations are altered). Results of some aspectsre occasionally invalid or unavailable, resulting inall variation in the number of participants reported oneach measure.

ticipantsble 1 shows the sociodemographic characteristics of theorted subject population. There was no significant differ-ce in age, initial weight, or BMI. The sociodemographicfile of the completers and study dropouts was similar.

ritionogram attrition was previously reported (17). Almostlf of the diet group dropped out (42%) before the end ofatment, whereas almost all (92%) of the health atry size group completed the program.

ight-Related Measuresshown in Table 2, the health at every size groupmbers maintained weight and BMI throughout thedy and follow-up period. The diet group significantlycreased their weight posttreatment (5.2 kg7.3 from

able 1. Characteristics of white, female chronic dieters partici-ating in the health at every size program vs diet program weight-ss trial

haracteristic

Health atevery sizegroup(n19)

Diet group(n19)

4meanstandard deviation3ge (y) 41.43.0 40.04.4eight (kg) 101.113.3 101.213.8MIa 35.94.6 36.74.2

4% 3ucation

igh school or less 5 16ome college 42 21ollege graduate 53 63ployment status

ot employed 0 5ployed 100 95b categoryb

ofessional 32 63lerical 26 21chnical 11 11ysical 0 0ther 26 5elationship statusarried or domestic partnership 89.5 68ingle 10.5 32

MIbody mass index; calculated as kg/m2.efers to those currently employed.June 2005 Volume 105 Number 6seline), such that their weight loss was 5.2% of thetial weight. They maintained the weight loss postafter-e (5.3 kg6.7 from baseline), but regained some ofweight such that weight was not significantly differ-

t between baseline and follow-up (P.068). There wasarallel pattern to the change in BMI.

od Lipid Levels and Blood Pressure Measuresindicated in Table 3, the health at every size groupmbers showed an initial increase in total cholesterol,lowed by a significant decrease from baseline at follow-. The diet group members showed no significant changetotal cholesterol at any time. Both groups showed anificant decrease in LDL cholesterol levels postafter-e: the health at every size group sustained this im-vement at follow-up and LDL cholesterol levels in thet group were not significantly different between base-e and follow-up. HDL cholesterol levels decreased inth groups.oth groups showed a significant lowering of systolicod pressure posttreatment and postaftercare. Thealth at every size group sustained this improvement atlow-up (P.043), whereas the diet group did not quiteieve significance in sustaining their improvement.051). There was no significant change in diastolicod pressure in either group at any time.

ivity Measurese health at every size group demonstrated a significantrease in daily energy expenditure posttreatment andfollow-up. This was not significant at 52 weeks. Theres also an almost fourfold increase in moderate activityfollow-up, which was the only of the individual activitytors that was significantly different from baseline. Them of time spent in moderate, hard, and very hardivity was also significantly increased for the health atry size group at follow-up, such that it was slightlyre than double initial values. The diet group showed anificant increase in some aspects of energy expendi-e postaftercare; however, none of these were sustainedfollow-up.

ing Behavior Measuresth groups started with relatively low cognitive re-aint (restricted eating). This changed in oppositeections in the two groups: it significantly decreasedthe health at every size group and significantlyreased in the diet group posttreatment and post-ercare (Table 4). The health at every size groupstained this change at follow-up; the restraint scoresthe diet group were not significantly different be-een baseline and follow-up. Post-hoc analysis dem-strated a significant between-group difference be-een baseline and follow-up. Both groups alsomonstrated significant improvement posttreatmentthe two other Eating Inventory subscales; that is,nger (susceptibility to hunger) and disinhibitionss of control that follows violation of self-imposedles). The health at every size group maintained theseprovements; the diet group maintained the improve-nt in disinhibition, but not hunger.

TM

WDHBDHW

DH

aDbHcBdSeS

T

Measure

Mean ValuesStandard DeviationBaseline to Follow-upComparison (P Value)

Baseline(Time 0)

Posttreatment(24 weeks)

Postaftercare(52 weeks)

Follow-up(104 weeks)

Within-groupanalysis

Between-groupanalysis

To

DHH

DHLo

DHS

DHD

DH

aTbDcHdSeStal cholesterol(mmol/L)a

ietb (n17) 4.500.74 4.960.94 4.200.79 4.240.72 .222 .364AESc (n17) 4.610.80 5.350.77d 4.320.75 4.070.77d .026e

igh-density lipoproteincholesterol (mmol/L)

iet (n17) 1.200.27 1.230.34 1.180.32 1.010.25d .009e .404AES (n18) 1.290.29 1.230.21 1.140.23d 1.030.16d .000e

w-density lipoproteincholesterol (mmol/L)

iet (n17) 2.990.95 3.010.79 2.310.48d 2.630.57 .236 .572AES (n18) 3.010.83 3.220.55 2.550.64d 2.530.51d .038e

ystolic blood pressure(mm Hg)

iet (n18) 127.611.1 120.112.2d 116.610.9d 121.316.9 .051 .982AES (n16) 125.814.2 119.912.9d 119.915.4d 119.511.7d .043e

iastolic bloodpressure (mm Hg)

iet (n18) 73.28.0 71.69.7 69.58.1 73.310.6 .938 .403AES (n16) 70.39.0 67.87.1 69.78.4 68.38.0 .307

o convert mmol/L cholesterol to mg/dL, multiply mmol/L by 38.7. To convert mg/dL cholesterol to mmol/L, multiply mg/dL by 0.026. Cholesterol of 5.00 mmol/L193 mg/dL.ietdiet group.AEShealth at every size group.ignificant within-group difference from baseline.ignificant between-group difference.able 2. Weight-related measures of white, female, chronic dieters by treatment condition over time

easure

Mean ValuesStandard DeviationBaseline to Follow-upComparison (P Value)Baseline(Time 0)

(Dieta n19;HAESb n19)

Posttreatment(24 weeks)(Diet n16;HAES n19)

Postaftercare(52 weeks)(Diet n18;HAES n18)

Follow-Up(104 weeks)Diet n19;HAES n19)

Within-groupanalysis

Between-groupanalysis

eight (kg)iet 101.213.8 96.814.2d 95.411.9d 98.014.3 .068 .116AES 101.113.3 101.813.4 101.413.6 101.516.3 .817MIc

iet 36.74.2 35.34.1d 34.73.5d 35.54.6 .068 .786AES 35.94.6 36.14.6 36.04.5 36.05.4 .868eight changefrom baseline

iet . . . 5.27.3 5.36.7 3.27.2 .515 .116AES . . . 0.62.1e 0.64.4e 0.36.3 .841

ietdiet group.AEShealth at every size group.MIbody mass index; calculated as kg/m2.ignificant within-group difference from baseline.ignificant between-group difference.

able 3. Blood lipid and blood pressure measures of white, female, chronic dieters by treatment condition over timeJune 2005 Journal of the AMERICAN DIETETIC ASSOCIATION 933

Tcanoffor

T onic

M

eviat

fterconth

EaRDHHDHDDHEa

D

DHBDHB

DHIn

DHB

DHR

DHB

DH

aDbHcSdSeSfF

934he health at every size group demonstrated signifi-t improvement between baseline and follow-up in fourthe eight Eating Disorder Inventory-2 subscales: drivethinness, bulimia (binge eating behavior), body dis-

able 4. Eating behavior/psychological measures of white, female, chr

easure

Mean ValuesStandard D

Baseline(Time 0)(Dieta n19;HAESb n19)

Posttreatment(6 months)(Diet n16;HAES n19)

Posta(12 m(DietHAES

ting inventoryestraintiet 7.94.9 11.94.0c 10.9AES 7.64.0 5.63.7ce 5.2ungeriet 8.13.5 5.64.2c 6.3AES 8.42.9 4.43.1c 5.2isinhibitioniet 12.22.1 8.42.8c 9.7AES 12.12.5 7.64.2c 7.2ting disordersinventoryf

rive forthinness

iet 4.64.6 2.92.7 3.2AES 7.16.1 2.63.3c 2.6ulimiaiet 4.64.0 1.32.1c 1.1AES 3.83.4 1.11.5c 0.8odydissatisfaction

iet 17.55.9 15.16.2c 17.0AES 17.94.5 12.87.5c 13.6teroceptiveawareness

iet 3.54.3 2.13.2 0.9AES 4.64.5 3.33.5 2.8eckDepressionInventory

iet 7.57.2 4.56.3c 3.9AES 10.39.5 6.99.3c 6.4osenberg Self-EsteemInventory

iet 31.25.5 32.55.5 32.2AES 30.93.8 32.15.8 32.4ody imageavoidance

iet 38.38.1 36.26.3 35.9AES 38.911.2 29.69.1ce 28.4

ietdiet group.AEShealth at every size group.ignificant within-group difference from baseline.ignificant difference from baseline to follow-up.ignificant between-group difference.our Eating Disorder Index subscales did not change and are not reported (ineffectivenessJune 2005 Volume 105 Number 6isfaction, and interoceptive awareness (ability to rec-ize and respond to internal states such as emotions,nger, and satiety). The diet group showed an initialprovement in three subscales (bulimia, body dissatis-

dieters by treatment condition over time

ionBaseline to Follow-upComparison (P Value)ares)

8;8)

Follow-up(24 months)(Diet n19;HAES n19)

Within-groupanalysis

Between-groupanalysis

9.64.7 .076 .007de 5.43.3ce .047d

7.13.9 .216 .2686.14.0c .014d

10.33.1c .013d .0708.23.9c .000d

3.73.2 .354 .042d

2.63.6c .004d

2.73.7 .061 .4641.11.4c .002d

16.88.0 .634 .023d

11.96.6ce .002d

2.33.2 .123 .4252.43.1c .038d

6.65.6 .134 .0116.68.8c .001d

29.15.8c .028d .000d

33.74.5ce .001d

34.26.5 .059 .187e 30.310.0c .003d

ersonal distrust, maturity fears, perfectionism).satognhuim

n1n1

3.9c

3.1c

3.63.3c

3.0c

4.3c

4.13.9c

1.4c

1.9c

7.78.3

1.4c

4.6

4.9c

9.4c

4.85.6

6.19.9c

, interp

facthecrepropebebolow

PsyBodeatupatmestrcatevemewhcal

ParThfouspoHefeepapapeenstawhStsizPrcomofOmeLiv

DISThdifhecouenagregpamoheirrinaindB

straft

proliteoth(anhesigthetaiI

heeacexapreanofheanberhaT

allrepthethestrenteequgroheT

paconlosevewemewebolpendifimingfocO

avCoanpait ifavpaheinflatI

conma

COWeforhation, and interoceptive awareness), although none ofse improvements were sustained at follow-up. (A de-ase in interoceptive awareness score represents im-vement, because elevated scores represent a defect inrception.) Post-hoc analysis demonstrated a significanttween-group difference in the drive for thinness anddy dissatisfaction subscales between baseline and fol--up.

chological Measuresth groups demonstrated significant improvement inpression posttreatment and postaftercare; the healthevery size group sustained this improvement at follow-, whereas the diet group did not (Table 4). The healthevery size group demonstrated a significant improve-nt in self-esteem at follow-up; the diet group demon-ated a significant worsening. Post-hoc analysis indi-ed significant between-group difference. The health atry size group also demonstrated significant improve-nt in body image avoidance behavior (P.003),ereas improvement in the diet group was not statisti-ly significant.

ticipant Evaluationsere was a significant between-group difference in allr participant evaluation questions (P.000). In re-nse to the statement, My involvement with thealthy Living Project (name of study) has helped me tol better about myself, 100% of the health at every sizerticipants endorsed Agree or Strongly Agree com-red with 47% of the diet group participants. Ninety-fivercent of the health at every size group participantsdorsed Disagree or Strongly Disagree regarding thetement, I feel like I have failed in the program,ereas 53% of the diet group endorsed Agree orrongly Agree. One hundred percent of health at everye participants were hopeful that the Healthy Livingoject would have a positive life-long impact on them,pared with 37% of the diet group. Eighty-nine percent

the health at every size group endorsed Regularly orften in response to the statement: I currently imple-nt some of the tools that I learned in the Healthying Project, compared with 11% of the diet group.

CUSSIONere are two aspects of the health at every size model thatfer from the traditional treatment approach and concernalth care practitioners. First, although dieters are en-raged to increase their cognitive restraint to decreaseergy intake, health at every size participants are encour-ed to decrease their restraint, relying instead on intuitiveulation. Second, the health at every size model supportsrticipants in accepting their size, whereas in the dietdel, reduction in size (weight loss) is emphasized. Manyalth care practitioners fear that health at every size isesponsible and that these aspects will result in indiscrim-te eating and increased obesity (16). Our data at 2 yearsicate this concern is unfounded.oth groups were implementing their encouraged re-aint pattern at the conclusion of the intervention and theercare program. The diet group did not sustain this im-ved restraint at follow-up, which is consistent with therature. The health at every size group members, on theer hand, were able to sustain their decrease in restraintd interoceptive awareness) at follow-up. In other words,alth at every size participants became sensitized to bodynals regulating food intake, increased their reliance onse signals as regulators of intake, and were able to main-n this behavior change over the 104-week period.mprovements in many of the health behaviors andalth risk markers paralleled the relative success thath group had in maintaining their restraint habits. Formple, both groups showed initial improvements in de-ssion, in all of the other scales related to eating behaviord attitudes toward weight and food, and in many aspectsmetabolic functioning and energy expenditure. Thealth at every size group sustained their restraint habitsd all of these improvements at follow-up, whereas mem-s of the diet group returned to baseline in their restraintbits and all but the disinhibition scale.he health at every size group also demonstrated a par-el improvement in self-esteem, and 100% of participantsorted that their involvement in the program helpedm feel better about themselves (compared with 47% ofdiet group). The diet group, on the other hand, demon-ated initial improvement followed by a significant wors-ing of self-esteem at follow-up. This damage to self-es-m was reinforced in other of the self-evaluationestions. For example, 53% of participants in the dietup expressed feelings of failure (compared with 0% ofalth at every size participants).he diet groups change in weight exhibited the same

ttern: There was a decrease in weight at the programclusion, then a gradual regain, such that the final weights of 3.2 kg was not significant (P.068). The health atry size group members, on the other hand, maintainedight throughout the study. The fact that the improve-nts in health risk indicators occurred during relativeight stability demonstrates that improvements in meta-ic functioning can occur through behavior change, inde-dent of a change in weight. Given the well-documentedficulties in sustaining weight loss, this is a particularlyportant result, and provides further support for redirect-clients toward behavior change as opposed to a primaryus on weight.ne limitation of the study was the small sample size

ailable at follow-up (50% of the original participants).nsidering that the diet group had high attrition (42%),d that part of the attrition in diet programs is due torticipants dropping outwhen they are not successful (29),s likely that the results for the diet group may look moreorable than was actually the case had all the partici-nts been considered. Because program attrition in thealth at every size group was markedly lower (8%), thisationmay be less pronounced in the results of the healthevery size group at program end.t also should be noted that only 2-year follow-up wasducted; longer follow-up (eg, 5 years) is suggested toke more conclusive statements about long-term benefits.

NCLUSIONSight loss has been established as standard treatmentobesity. Although often successful in the short term, its shown limited long-term success in mitigating obe-June 2005 Journal of the AMERICAN DIETETIC ASSOCIATION 935

sitdieprepopromeweatianselI

apanhefroitivinganasschesta hmadieE

ansighe

Thtutcootiofel

Ref1

2

3

4

5

6

7.

8

9.

10

11

12

13

14

15

16

17

18

19

20

21

22

23.

24

25

26

27

28

29

30

936y and its associated health problems for the majority ofters. Findings in the diet group were consistent withvious literature. There was high attrition (42%), andstaftercare data (1 year after program initiation) forgram completers indicated weight loss and improve-nt in health risk indicators, although neither of thesere sustained at follow-up (2 years after program initi-on). Diet group participants additionally experiencedoverall detrimental effect on self-esteem and otherf-evaluation measures.n contrast to a diet program, the health at every sizeproach encourages persons to accept their body weight,d to rely on their body signals to support positivealth behaviors and help regulate their weight. Resultsm this randomized clinical trial were remarkably pos-e, with health at every size group participants show-sustained improvements in many health behaviors

d attitudes as well as many health risk indicatorsociated with obesity (including total cholesterol, LDLolesterol, systolic blood pressure, depression, and self-eem, but not HDL cholesterol). The data suggest thatealth at every size approach enables participants tointain long-term (2 years) behavior change, whereas at approach does not.ncouraging size acceptance, a reduction in dieting,

d a heightened awareness of and response to bodynals appears to be effective in supporting improvedalth risk indicators for obese, female chronic dieters.

is research was supported in part by National Insti-es of Health grants Nos. DK57738 and DK35747, aperative agreement with the Western Human Nutri-n Research Center, and a National Science Foundationlowship.

erences. Clinical Guidelines on the Identification, Evaluation,and Treatment of Overweight and Obesity in Adults:The Evidence Report. Bethesda, MD: National Heart,Lung, and Blood Institute; 1998.

. Neumark-Sztainer D, Rock CL, Thornquist MD,Cheskin LJ, Neuhouser ML, Barnett MJ. Weight-control behaviors among adults and adolescents: As-sociations with dietary intake. Prev Med. 2000;30:381-391.

. Serdula MK, Mokdad AH, Williamson DF, Galuska J,Mendlein M, Heath GW. Prevalence of attemptingweight loss and strategies for controlling weight.JAMA. 1999;282:1353-1358.

. Flegal K, Carroll M, Ogden C, Johnson C. Prevalenceand trends in obesity among US adults, 1999-2000.JAMA. 2002;288:1723-1727.

. Miller WC. How effective are traditional dietary andexercise interventions for weight loss?Med Sci SportsExer. 1999;31:1129-1134.

. Bacon L. Tales of mice and leptin: False promises andnew hope in weight control. Healthy Weight J. 2003;17:24-27.Position of The American Dietetic Association: Weightmanagement. J Am Diet Assoc. 1997;97:71-74.

. Garner D, Wooley S. Confronting the failure of be-havioral and dietary treatments for obesity. Clin Psy-chol Rev. 1991;11:748-754.June 2005 Volume 105 Number 6Wooley SC, Garner DM. Obesity treatment: The highcost of false hope. J Am Diet Assoc. 1991;91:1248-1251.

. Ernsberger P, Koletsky RJ. Biomedical rationale fora wellness approach to obesity: An alternative to afocus on weight loss. J Soc Issues. 1999;55:221-260.

. Gaesser G. Weight, weight loss, and health: A closerlook at the evidence. Healthy Weight J. 2003;17:8-11.

. Gaesser G. Thinness and weight loss: Beneficial ordetrimental to longevity. Med Sci Sports Exer. 1999;31:1118-1128.

. Cogan JC, Ernsberger P. Dieting, weight, and health:Reconceptualizing research and policy. J Soc Issues.1999;55:187-205.

. Robison J. Health at every size: Antidote for theobesity epidemic. Healthy Weight J. 2003;17:4-7.

. Miller WC. Fitness and fatness in relation to health:Implications for a paradigm shift. J Soc Issues. 1999;55:207-219.

. Kratina K. Health at every size: Clinical applications.Healthy Weight J. 2003;17:19-23.

. Bacon L, Keim NL, Van Loan MD, Derricote M, GaleB, Kazaks A, Stern JS. Evaluating a non-diet well-ness intervention for improvement of metabolic fit-ness, psychological well-being and eating and activitybehaviors. Int J Obes. 2002;26:854-865.

. Polivy J, Herman CP, Howard KI. Restraint Scale:Assessment of dieting. In: Hersen M, Bellack AS, eds.Dictionary of Behavioral Assessment Techniques.New York, NY: Pergamon; 1988:377-380.

. Westenhoefer J, Stunkard AJ, Pudel V. Validation ofthe flexible and rigid control dimensions of dietaryrestraint. Int J Eat Disord. 1999;26:53-64.

. Brownell KD. The LEARN program for weight con-trol. 7th ed. Dallas, TX: American Health PublishingCompany; 1997.

. Blair SN, Haskell WL, Ho P, Paffenbarger RS Jr,Vranizan KM, Farquhar JW, Wood PD. Assessmentof habitual physical activity by a seven-day recall ina community survey and controlled experiments.Am J Epidemiol. 1985;122:794-804.

. Stunkard AJ, Messick S. The Eating Inventory. SanAntonio, TX: Psychological Corporation; 1988.GarnerDM.EatingDisorder Inventory-2Manual. Odessa,FL: Psychological Assessment Resources, Inc; 1991.

. Beck AT, Rush AJ, Shaw BF, Emery G. Cognitive Ther-apy for Depression. New York, NY: Guilford; 1979.

. Rosenberg M. Society and the Adolescent Self-Image.Princeton, NJ: Princeton University Press; 1965.

. Rosen JC, Srebnik D, Saltzberg E, Wendt S. Devel-opment of a body image avoidance questionnaire.Psychol Assess. 1991;3:32-37.

. Ciliska D. Evaluation of two nondieting interventionsfor obese women. West J Nurs Res. 1998;20:119-135.

. Tanco S, Linden W, Earle T. Well-being and morbidobesity in women: A controlled therapy evaluation.Int J Eat Disord. 1998;23:325-339.

. Rothblum ED. Contradictions and confounds in cov-erage of obesity: Psychological journals, textbooks,and the media. J Soc Issues. 1999;55:355-369.

. Association for Size Diversity and Health. Available at:http://www.bgsu.edu/offices/sa/counseling/ASDAH.htm.Accessed January 10, 2004.

Size Acceptance and Intuitive Eating Improve Health for Obese, Female Chronic DietersMETHODSProcedureTreatment ConditionsDiet GroupHealth at Every Size GroupEvaluation/Outcome MeasuresAnthropometric and Metabolic Fitness MeasuresEnergy ExpenditureEating Behavior MeasuresPsychological MeasuresStatistical Methods

RESULTSParticipantsAttritionWeight-Related MeasuresBlood Lipid Levels and Blood Pressure MeasuresActivity MeasuresEating Behavior MeasuresPsychological MeasuresParticipant Evaluations

DISCUSSIONCONCLUSIONSReferences