R

ADPoOID

ACsafih

IttCtCMMcDslpQtHoaoOtpMtlDQ

DCC

A

7

RESEARCH

esearch and Professional Briefs

nthropometric, Metabolic, Psychosocial, andietary Characteristics of Overweight/Obeseostmenopausal Women with a Historyf Weight Cycling: A MONET (Montrealttawa New Emerging Team) Study

RENE STRYCHAR, EdD, RD; MARIE-ÈVE LAVOIE, MSc, RD; LYNE MESSIER, MSc, RD; ANTONY D. KARELIS, PhD; ÉRIC DOUCET, PhD;

ENIS PRUD’HOMME, MD, MSc; JONATHAN FONTAINE, BSc, RD; RÉMI RABASA-LHORET, MD, PhD

ltcpiMsvrigptcgaa2fckfmausibwbJ

Otna1aph

BSTRACTharacteristics of weight cyclers have not been fully as-essed. The objective of this study was to determine thenthropometric, metabolic, psychosocial, and dietary pro-le of postmenopausal women according to weight-cyclingistory, defined as the frequency of going on a diet and

. Strychar is a professor in the Département de Nutri-ion, Faculté de Médecine, Université de Montréal, Mon-réal, Québec, Canada, and a researcher at the Researchenter of the Centre Hospitalier de l’Université de Mon-

réal (CRCHUM) and the Montreal Diabetes Researchenter of the CRCHUM, Montréal, Québec, Canada..-È. Lavoie is a doctoral student, J. Fontaine is aaster’s student, and L. Messier is a nutrition research

oordinator in the Laboratory of Metabolic Dysfunctions,épartement de Nutrition, Faculté de Médecine, Univer-

ité de Montréal, Montréal, Québec, Canada. A. D. Kare-is is an assistant professor, Département de Kinanthro-ologie, Université de Québec à Montréal, Montréal,uébec, Canada. É. Doucet is an associate professor in

he Behavioural and Metabolic Research Unit, School ofuman Kinetics, Faculty of Health Sciences, University

f Ottawa, Ottawa, Ontario, Canada. D. Prud’homme isprofessor at the School of Human Kinetics and Dean

f the Faculty of Health Sciences, University of Ottawa,ttawa, Ontario, Canada. R. Rabasa-Lhoret is head of

he Laboratory of Metabolic Dysfunctions, an assistantrofessor in the Département de Nutrition, Faculté deédecine, Université de Montréal, and a researcher at

he Research Center of the Centre Hospitalier de’Université de Montréal (CRCHUM) and the Montrealiabetes Research Center of the CRCHUM, Montréal,uébec, Canada.Address correspondence to: Irene Strychar, EdD, RD,épartement de Nutrition, Université de Montréal, 2405hemin de la Côte Ste-Catherine, Montréal, Québec,anada, H3T 1A8. E-mail: irene.strychar@umontreal.caManuscript accepted: August 6, 2008.Copyright © 2009 by the American Dietetic

ssociation.0002-8223/09/10904-0015$36.00/0

sdoi: 10.1016/j.jada.2008.12.026

18 Journal of the AMERICAN DIETETIC ASSOCIATION

osing �10 kg: never (0 times), low (1 time), moderate (2o 3 times), or frequent (�4 times). The sample of thisross-sectional study consisted of 121 overweight/obeseostmenopausal women enrolled in a 6-month random-zed weight-loss intervention between 2003 and 2006.

easures at baseline were used to evaluate body compo-ition (fat mass by dual-energy x-ray absorptiometry andisceral fat by computed tomography); resting metabolicate by indirect calorimetry; insulin sensitivity by hyper-nsulinemic-euglycemic clamp; fasting plasma levels oflucose, lipids, leptin, ghrelin, and adiponectin; bloodressure; psychosocial profile (eg, body-esteem, self-es-eem, stress, perceived risks, perceived benefits, self-effi-acy, quality of life, dietary restraint, disinhibition, hun-er); and dietary profile (3-day food record). Differencesmong groups of weight cyclers were determined usingnalyses of variance. Among the 121 women, 15.7%,4.8%, 33.9%, and 25.6% were non-, low, moderate, andrequent cyclers, respectively. Frequent cyclers wereharacterized by higher body mass index (calculated asg/m2) (current and at 25 years of age) and percent bodyat mass, larger waist circumference, and lower restingetabolic rate/kg body weight than noncyclers (P�0.05);

nd moderate cyclers had lower plasma adiponectin val-es than noncyclers (P�0.05). For psychosocial mea-ures, frequent cyclers were characterized by greater dis-nhibition and lower body esteem after controlling forody mass index (P�0.05). In conclusion, weight cyclingas found to be associated with some unfavorable meta-olic and psychosocial parameters.Am Diet Assoc. 2009;109:718-724.

besity rates continue to escalate (1) while weight-loss maintenance rates remain low (2,3). Repeatedepisodes of weight loss and weight regain, referred

o as weight cycling, are a frequent occurrence (4,5). Inational surveys, approximately two thirds of obesedults have tried to lose weight at any one time (6,7), and0% to 40% have a history of weight cycling (8-10). Neg-tive effects of weight cycling on health risks, body com-osition, body fat distribution, and energy expenditureave been reported (11-21). Weight cycling has been as-

ociated with increased risk for metabolic syndrome, cor-

© 2009 by the American Dietetic Association

owitsg

cc(df2rdf

cieathahwr

MTfpvttMwpmtcBotawwtbsppi

ABucjiw

Botc

MRqsititotOfcmchifismmlmltmtp(ldd

PPm(totcBvcihttaad“

nary heart disease, all-cause mortality (11,15,18), andith reduced quality of life, independent of body mass

ndex (BMI) (10). Obese individuals with large fluctua-ions in body weight have greater taste preference forugar and fat (22) and are more prone to future weightain (23).In contrast, other studies have not reported adverse

onsequences of weight cycling. For example, weight cy-ling was not an independent predictor of preeclampsia24), hypertension (25), or type 2 diabetes (9). Absence ofetrimental effects of weight cycling on body composition,at distribution, energy expenditure, and blood lipids (26-8), as well as psychological health (29-32), have beeneported. Discrepancies in results might be related toifferent definitions of weight cycling and lack of controlor confounding factors.

Few studies have examined dietary aspects of weightycling (33,34). This study extends and further character-zes the anthropometric, metabolic, psychosocial, and di-tary profile of individuals with a history of losing a largemount of weight (�10 kg or 22 lb), in a group of seden-ary overweight/obese postmenopausal women. No studyas previously examined this comprehensive set of char-cteristics within the same group of women. The studyypothesis is as follows: women without a history ofeight cycling will have a more favorable anthropomet-

ic, metabolic, psychosocial, and dietary profile.

ETHODShe study sample of this cross-sectional study was drawn

rom 137 sedentary, overweight, and obese postmeno-ausal women enrolled in a 6-month weight-loss inter-ention (randomized diet with or without resistanceraining). Subjects were recruited by newspaper adver-isements. The study was approved by the Université deontréal, Faculty of Medicine Ethics Committee. Dataere collected from 2003 to 2006. Women were eligible toarticipate if they met the following criteria: cessation ofenstruation �1 year; no use of hormone replacement

herapy; sedentary (�2 hours/week of structured exer-ise); nonsmokers; free of known inflammatory disease;MI �27. Participants had no evidence of cardiovascularr peripheral vascular diseases, stroke, diabetes, hyper-ension, orthopedic limitations, and medication for met-bolic conditions. Three weeks prior to testing, bodyeight was monitored on a weekly basis to ensure that itas stable within �2 kg. Participants were then invited

o the metabolic unit to evaluate anthropometric, meta-olic, psychosocial, and dietary profiles. The data pre-ented in this article represent baseline characteristics ofarticipants. Of the initial 137 women, 121 provided com-lete information about weight-cycling history and werencluded in these analyses.

nthropometric Assessmentody weight, lean body mass, and fat mass were measuredsing dual-energy x-ray absorptiometry, and visceral fat byomputed tomography, as described previously (35). Sub-ects wore light clothing for these measurements. Stand-ng height was measured to the nearest 0.1 cm with a

all stadiometer (Perspective Enterprises, Portage, MI). “

MI was calculated as kg/m2. Waist circumference (meanf two measures) was determined using a Gulick tape athe mid-distance between the lowest rib and the iliacrest.

etabolic Assessmentesting metabolic rate (RMR) and fasting respiratoryuotient were measured by indirect calorimetry after theubjects fasted overnight for 12 hours. Subjects werenstructed to fast and drink only water for 12 hours beforeesting, consume no alcohol, restrain from physical activ-ty for 24 hours before testing, and keep physical activityo a minimum on the morning of the test. Concentrationsf CO2 and O2 were measured using the ventilated hoodechnique with a SensorMedics Delta Track II (Datex-hmeda, Helsinki, Finland). Measurements were per-

ormed during 40 minutes, the first 10 minutes wereonsidered as an acclimatization period and the last 30inutes were used for analyses. Measurement of gas

oncentrations at rest were then used to extrapolate 24-our RMR using Weir’s equation. In our laboratory, the

ntraclass correlation coefficient (two-factor random ef-ect) for RMR determined by using a test-retest conditionn 19 different subjects was 0.92 (P�0.001). Insulin sen-itivity was measured by the hyperinsulinemic-euglyce-ic clamp technique, with an insulin infusion rate of 75U/m2/min for 180 minutes: glucose disposal was calcu-

ated as described previously (35) and expressed asg�min�1 per kg free fat mass. Venous plasma glucose

evels were determined during an oral glucose toleranceest (75 g) at fasting and at 2 hours. Blood pressure waseasured and fasting plasma lipid levels of total choles-

erol, high-density lipoprotein cholesterol, low-density li-oprotein cholesterol, and triglycerides were determined35). Fasting plasma concentrations of leptin, total ghre-in, and adiponectin were measured in a subsample inuplicate with a commercial radioimmunoassay proce-ure, as described previously (36).

sychosocial Factorssychosocial factors were assessed using the followingeasures: Perceived Stress Scale (37); Self-Esteem Scale

38); Body-Esteem Scale (39), including appearance, at-ribution, and weight subscales; Perceived Risk for devel-ping heart disease and for diabetes, developed accordingo the Health Belief Model (40); Perceived Benefits forontrolling weight, developed according to the Healthelief Model (40); Self-efficacy for controlling weight, de-eloped according to Social Cognitive Theory (41); Medi-al Outcomes Study General Health Survey (42,43) qual-ty-of-life measure, including health perceptions, mentalealth, role functioning, social functioning, physical func-ioning, and pain subscales; the 3-Factor Eating Ques-ionnaire (44), including dietary restraint, disinhibition,nd hunger factors. Internal consistency reliability wasssessed using Cronbach � coefficients (45) for measureseveloped for this study with two or more items: .74 forperceived benefits for controlling weight” and .73 for

self-efficacy for controlling weight.”

April 2009 ● Journal of the AMERICAN DIETETIC ASSOCIATION 719

NFajkbmfcwSOeptwtp

WAqoc“ot

OAemamsc

ST(tmdcavtwitvfaat

RIwh

(wh1rtpoeOlTsa

slw(

nBelckrwiwyh(twlgm

ggoecRiceRirebwewtlr

ws

7

utrient Intakesood intake was assessed using a 3-day food record (46)t baseline during the weight-stabilization period. Sub-ects were instructed by a registered dietitian (RD) toeep a record of food intake, including condiments andeverages, during 2 weekdays and 1 weekend day whileaintaining their usual habits. On their return, each

ood record was reviewed by the RD with the subject toomplete missing information. Analyses were conductedith the Food Processor SQL program (Food ProcessorQL Edition, version 9.6.2, 2004; ESHA Research, Salem,R) using the 2001 Canadian Nutrient Data File. Datantry of the food records was done by an RD and inde-endently verified by a second one. Discrepancies be-ween the two RDs were discussed and modificationsere made according to their mutual decision. Mean in-

ake of 3 days for percent of energy from carbohydrates,rotein, and fat were calculated for each subject.

eight-Cycling Profilelarge weight gain (23) was assessed by asking the

uestion “How many times in your lifetime have you gonen a diet and lost more than 10 kilograms?” Responseategories included: “never,” “once,” “two to three times,”four to five times,” and “more than five times.” Becausenly nine women responded “four to five times,” the lastwo categories were regrouped as “four or more times.”

ther Measuresge, education (university education/less than universityducation), yearly household income (�$50,000/�$50,000),arital status (living with spouse-partner/living alone),

nd weight at 25 years of age (self-report) were deter-ined. BMI at 25 years of age was calculated using

elf-reported weight (kg) at 25 years of age divided byurrent height (m)2.

tatistical Analysishe Statistical Package for the Social Sciences program

SPSS, version 15, 2006, Chicago, IL) was used for sta-istical analyses. Descriptive statistics were expressed asean�standard deviation (SD). �2 were conducted to

etermine differences between education, household in-ome, and marital status with weight-cycling history. Forge, anthropometric, metabolic, psychosocial, and dietaryariables, a one-way analysis of variance was performedo analyze mean differences among the four groups ofeight cycling; a Tukey post hoc test was performed to

dentify group differences. For variables that were foundo be statistically significant, analyses of covariance (uni-ariate general linear model) were conducted to controlor BMI. Significance was accepted as P�0.05. Analysesre based on 121 subjects; cases where numbers reportedre �121 indicate that women did not answer all ques-ions.

ESULTS AND DISCUSSIONn this sample of 121 overweight/obese postmenopausalomen seeking treatment for weight loss, participants

ad a mean age of 58 years (SD�5 years), wherein 59.8% N

20 April 2009 Volume 109 Number 4

67 of 112) were university-educated, 53.5% (61 of 114)ere living with a spouse-partner, and 58.1% (61 of 105)ad a household income �$50,000. In these women,5.7% (19 of 121) were noncyclers, 24.8% (30 of 121)eported cycling one time, 33.9% (41 of 121) two to threeimes, and 25.6% (31 of 121) four or more times. Thehenomenon of weight cycling was, therefore, a commonccurrence, with approximately 60% of women havingngaged in cycling two or more times in their lifetime.nly 16% of women reported being noncyclers, a rate

ower than that found in population-based studies (6-8).his was likely a result of the composition of the studyample and the definition of weight cycling as a largemount of weight loss, ie, �10 kg per episode.For sociodemographic characteristics, there were no

ignificant differences among weight-cycling groups forevel of education, income, or age; however, noncyclersere more likely to report living with a spouse-partner

P�0.013).For anthropometric factors, there were statistically sig-

ificant differences among groups of weight cyclers forMI (P�0.001), fat mass (P�0.04), and waist circumfer-nce (P�0.003) (Table 1). Frequent cyclers were moreikely to have higher BMI and body fat mass than non-yclers, although the causal path of this association is notnown, a result of our cross-sectional study design. Self-eported BMI at 25 years of age was also associated witheight-cycling history (P�0.001). These results are sim-

lar to those of the Nurses’ Health Study II (9), whereeight cycling was associated with reported BMI at 18ears of age: noncyclers, mild cyclers, and severe cyclersad a mean BMI of 20.9 (SD�3), 22.9 (SD�4), and 24.4SD�5), respectively. Young adults with a normal BMI athe upper end of the range may be at higher risk foreight gain and weight cycling than individuals at the

ower range of normal BMI (47). Obesity-prevention pro-rams targeting such individuals in young adulthooday be warranted.For metabolic factors, no significant differences among

roups of weight cyclers were found for blood pressure,lucose values, insulin sensitivity, respiratory quotient,r RMR (even when controlling for lean body mass). How-ver, when RMR was divided by kg body weight, noncy-lers and one-time cyclers had significantly higherMR/kg body weight than frequent cyclers (P�0.003). An

nteresting finding of this study is that frequent weightyclers may be characterized by reduced relative energyxpenditure. Although absolute RMR did not differ, lowerMR/kg body weight might support the notion that some

ndividuals may become resistant to slimming followingepeated exposure to weight-loss interventions (48). How-ver, because data are cross-sectional, causality cannote established. Other studies have shown that body-eight loss causes greater than expected reduction in

nergy expenditure (49,50), which may persist in theeight-reduced state (51). High RMR of energy expendi-

ure predict long-term successful weight loss (52), whereasow RMR of energy expenditure predict weight gain andegain (53).Adiponectin, but not leptin and ghrelin, was associatedith weight cycling (P�0.04); this association remained

tatistically significant after controlling for fat mass.

oncyclers had higher plasma adiponectin values than

Table 1. Anthropometric and metabolic characteristics of overweight/obese postmenopausal women, according to weight-cycling history

FactorsTotal(n�121)

Group 1(n�19)noncycler(never)

Group 2(n�30)low cycler(1 time)

Group 3(n�41)moderate cycler(2-3 times)

Group 4(n�31)frequent cycler(>4 times) P value

Anthropometric 4™™™™™™™™™™™™™™™™™™™™™™ mean�standard deviation ™™™™™™™™™™™™™™™™™™™™™3Body weight (kg) 83.7�14.0 78.0�10.3a 77.9�8.8b 83.5�12.6c 92.9�17.1abc �0.001BMId

Current 32.6�4.7 30.4�4.7a 30.8�3.1b 32.9�4.6 35.2�5.1ab �0.001At 25 years 22.5�2.8 21.4�2.3a 21.5�1.6b 22.4�2.1c 24.1�4.0abc 0.001

LBMe (kg) 42.5�6.4 41.0�4.7a 40.3�3.9b 42.2�6.3 45.8�8.0ab 0.004LBM (%) 51.1�4.7 53.0�4.9 51.9�4.3 50.8�4.9 49.6�4.4 0.06FMf (kg) 38.8�9.4 34.6�8.2a 35.3�6.6b 38.8�8.5c 44.6�10.8abc �0.001FM (%) 45.9�4.8 44.0�5.0a 45.0�4.4 46.2�4.9 47.7�4.5a 0.04VFg (cm2) 184.9�55.3 166.0�40.6 170.6�49.8 193.7�60.2 199.2�57.2 0.06Waist circumference (cm) (n�120) 103.7�11.5 99.4�0.0a 99.8�9.4b 104.4�10.5 109.2�13.1ab 0.003MetabolicRMRh (kcal/24 hours) 1,319�181 1,267�92 1,279�148 1,320�194 1,388�217 0.06RMR/kg body weight 15.9�1.6 16.4�1.8a 16.5�1.5b 15.9�1.5 15.1�1.5ab 0.003RQi 0.86�0.07 0.86�0.08 0.87�0.05 0.86�0.08 0.88�0.06 0.60Insulin sensitivity (mg glucose/min/kg

FFMj) (n�117) 11.2�3.0 11.1�2.8 11.9�3.9 11.6�3.0 10.4�2.1 0.24Plasma glucose (mmol/L)k

Fasting (n�120) 5.3�0.5 5.2�0.4 5.4�0.5 5.3�0.6 5.2�0.5 0.302-hour (n�120) 6.4�1.8 5.9�1.9 6.3�2.0 6.4�1.7 6.7�1.7 0.49

Plasma lipidsTotal cholesterol (mg/dL)l 212.7�32.9 223.1�24.7 209.6�27.9 210.7�35.5 211.7�38.2 0.51HDL-C (mg/dL)m 57.6�12.8 58.0�13.8 61.6�14.4 57.1�12.0 54.1�10.8 0.15LDL-C (mg/dL)n 125.2�28.4 134.3�23.9 119.3�25.5 124.2�30.8 126.6�29.9 0.34TG (mg/dL)o 147.3�65.9 152.9�51.8 143.1�70.3 142.6�69.1 154.2�67.2 0.85Total cholesterol/HDL-C ratio 3.8�0.8 4.0�1.0 3.5�0.8 3.8�0.7 4.0�0.8 0.09

HormonesLeptin (ng/mL) (n�62) 27.8�9.6 24.7�5.5 27.0�8.2 27.4�11.2 31.7�9.9 0.33Ghrelin (pg/mL) (n�53) 1,178�411 960�392 1,332�409 1,244�397 1,130�414 0.16Adiponectin (�g/mL) (n�49) 9.6�5.0 13.6�9.0p 9.9�3.6 7.7�2.9p 8.8�2.9 0.04

Blood pressure (mm Hg)Systolic 121.0�14.0 118.6�8.3 118.7�13.0 125.3�16.4 119.1�13.5 0.12Diastolic 76.1�7.9 75.9�5.6 75.5�7.9 77.6�9.2 75.0�7.5 0.54

aGroup 1 is significantly different from group 4 (P�0.05).bGroup 2 is significantly different from group 4 (P�0.05).cGroup 3 is significantly different from group 4 (P�0.05).dBMI�body mass index.eLBM�lean body mass.fFM�fat mass.gVF�visceral fat.hRMR�resting metabolic rate.iRQ�respiratory quotient.jFFM�fat-free mass.kTo convert mmol/L glucose to mg/dL, multiply mmol/L by 18.0. To convert mg/dL glucose to mmol/L, multiply mg/dL by 0.0555. Glucose of 6.0 mmol/L�108 mg/dL.lTo convert mg/dL cholesterol to mmol/L, multiply mg/dL by 0.026. To convert mmol/L cholesterol to mg/dL, multiply mmol/L by 38.7. Cholesterol of 193 mg/dL�5.00 mmol/L.3mHDL-C�high-density lipoprotein cholesterol. To convert mg/dL HDL-C to mmol/L, multiply mg/dL by 0.026. To convert mmol/L HDL-C to mg/dL, multiply mmol/L by 38.7. HDL-C of193 mg/dL�5.00 mmol/L.nLDL-C�low-density lipoprotein cholesterol. To convert mg/dL LDL-C to mmol/L, multiply mg/dL by 0.026. To convert mmol/L LDL-C to mg/dL, multiply mmol/L by 38.7. LDL-C of 193mg/dL�5.00 mmol/L.oTG�triglyceride. To convert mg/dL TG to mmol/L, multiply mg/dL by 0.0113. To convert mmol/L TG to mg/dL, multiply mmol/L by 88.6. TG of 159 mg/dL�1.80 mmol/L.

pGroup 1 is significantly different from group 3 (P�0.05).

April 2009 ● Journal of the AMERICAN DIETETIC ASSOCIATION 721

mpl(aIswd

we(a0dTwermcasgd

icaftt

tb1r63rahwhn

cctss

7

oderate cyclers, a puzzling observation because adi-onectin was not associated with either plasma glucoseevels or insulin sensitivity, as might have been expected54). Lower levels of adiponectin have been reported to bessociated with the development of type 2 diabetes (55).n line with some (9,25-27), but not all (15,18), previoustudies, the current data do not support adverse effects ofeight cycling on classical variables associated with car-iovascular risk profile.For psychosocial factors, weight cycling was associatedith self-efficacy (P�0.04), perceived risk for heart dis-ase (P�0.04), body-esteem (P�0.02), and disinhibitionP�0.001) (Table 2). After controlling for BMI, differencesmong groups of weight cyclers for self-efficacy (P�.049), body-esteem appearance subscale (P�0.03), andisinhibition remained statistically significant (P�0.004).hese results are similar to the Québec Family Study,here past-dieting women had higher scores than nondi-ters for disinhibition, that is, overconsumption of food inesponse to cognitive or emotional cues (34). Because theajority of women seeking weight-loss treatment in the

urrent study had a previous history of weight cyclingnd because disinhibition scores were independently as-ociated with weight cycling, consideration needs to beiven to cognitive and behavioral aspects (56,57) to re-

Table 2. Psychosocial characteristics of overweight/obese postmeno

Factorsa Total

Group 1(n�19)noncycler(never)

4™™™™™™™™™™™™™™™™™™™™™Self-efficacy (1-4) (n�119) 2.7�0.5 2.7�0.7Perceived benefits (1-4) (n�115) 3.7�0.4 3.7�0.5Perceived risk for:

Heart disease (1-4) (n�120) 2.9�0.9 2.7�0.9Diabetes (1-4) (n�120) 2.8�0.9 2.5�1.0

Stress (0-56) (n�118) 18.9�8.1 18.1�9.6Self-esteem (1-4) (n�121) 1.8�0.5 1.7�0.5Body-esteem (0-4) (n�104) 1.4�0.5 1.7�0.5d

Appearance (n�120) 1.6�0.6 1.9�0.6d

Weight (n�119) 1.0�0.6 1.3�0.5d

Attribution (n�104) 1.9�0.6 1.9�0.53-Factor Eating Questionnaire

Restraint (0-21) (n�108) 11.1�3.8 11.7�3.9Disinhibition (0-16) (n�119) 8.3�3.5 6.4�3.0d

Hunger (0-14) (n�114) 5.4�3.0 5.4�2.6Quality of life (0-100%) (n�114) 79�14 84�12

Physical functioning (n�118) 81�24 90�13Pain (n�121) 59�31 66�31Social functioning (n�120) 96�11 99�5Role functioning (n�120) 86�25 93�16Mental health (n�121) 70�15 71�16Health perceptions (n�117) 77�18 81�17

aHigher scores indicate greater self-efficacy, greater perceived benefits, greater perceivedisinhibition, greater hunger, and greater quality of life.bGroup 3 is significantly different from group 4 (P�0.05).cGroup 2 is significantly different from group 4 (P�0.05).dGroup 1 is significantly different from group 4 (P�0.05).

uce overconsumption of foods by postmenopausal women t

22 April 2009 Volume 109 Number 4

n problematic situations. Moreover, frequent weight cy-ling was associated with lower esteem of body appear-nce, suggesting a possible link between psychologicalactors and weight cycling. Weight cycling may have long-erm negative effects on body-appearance esteem, beyondhose already noted for BMI.

Mean nutrient intakes of participants were similar inhe four study groups, where percent of energy from car-ohydrate, protein, and total fat were of the order of 48%,7%, and 32% (10.5% saturated fat, 11% monounsatu-ated fat, and 5% polyunsaturated fat), respectively. Only8% (82 of 121) of respondents volunteered to return the-day food records with complete information. This loweturn rate may explain why there were no substantialssociations between nutrient intakes and weight-cyclingistory. Provencher and colleagues (34) reported thatomen with a past history of dieting were more likely toave a higher-fat and lower-carbohydrate pattern thanondieters.There are several study limitations. Data on weight

ycling were based on self-reported episodes and wereollected retrospectively: no cause-and-effect can be es-ablished. Furthermore, the study sample is not repre-entative of postmenopausal women, but limited to aelected group seeking treatment in an intense struc-

al women, according to weight-cycling history

oup 230)cycler

time)

Group 3(n�41)moderate cycler(2-3 times)

Group 4(n�31)frequent cycler(>4 times) P value

an�standard deviation ™™™™™™™™™™™™™™™™™™™™™™3.6�0.5 2.9�0.5b 2.5�0.5b 0.04.6�0.3 3.7�0.3 3.7�0.4 0.74

.5�0.7c 3.0�0.9 3.1�1.0c 0.04

.6�0.7 2.9�1.0 3.1�1.0 0.09

.3�6.6 20.0�8.5 17.7�8.0 0.64

.9�0.5 1.8�0.4 1.9�0.5 0.59

.4�0.5 1.5�0.5 1.2�0.4d 0.002

.6�0.6 1.6�0.7b 1.2�0.5bd 0.002

.0�0.6 1.0�0.6 0.8�0.4d 0.01

.9�0.5 2.0�0.6 1.9�0.5 0.63

.8�3.8 10.9�3.7 11.3�3.9 0.90

.4�3.0c 8.1�3.7b 10.4�3.1bcd �0.001

.7�3.0 5.3�3.0 5.2�3.4 0.928�14 78�15 78�12 0.485�21 75�29 80�22 0.134�32 58�31 60�28 0.627�9 95�12 95�16 0.584�29 87�25 84�27 0.579�13 68�17 73�15 0.487�19 76�19 78�16 0.83

greater stress, lower self-esteem, higher body-esteem, greater dietary restraint, greater

paus

Gr(n�low(1

™ me23

22

1911111

10757859867

d risk,

ured weight-loss research program. However, these

snwt

CWunwctpatap

THfsrQCFfE

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

tudy results are supported by use of gold-standard tech-iques for anthropometric and metabolic assessment in aell-characterized cohort that examined large inten-

ional weight loss (�10 kg).

ONCLUSIONSeight cycling was found to be associated with some

nfavorable metabolic and psychosocial parameters, butot others. Postmenopausal women with a history ofeight cycling need to pay attention to the amount of food

onsumed in response to situational and emotional cuesriggering overeating. A problem-solving cognitive ap-roach (53,54) may be useful in assisting women in man-ging such situations. Research is needed to determinehe links between weight cycling and dietary patternss well as the effects of weight-loss methods on suchatterns.

his research was funded by a Canadian Institutes ofealth Research (CIHR) New Emerging Team Grant

rom the Institute of Nutrition and Metabolism and In-titute of Gender and Health. One of the authors is aesearch scholar of the Fonds du Recherche en Santé duuebec and has received support from the Fondation duHUM. Another author is a recipient of a CIHR/Merck-rosst New Investigator Award, a Canadian Foundation

or Innovation New Opportunities Award, and an Ontarioarly Researcher Award.

eferences1. Lau DCW for the Obesity Canada Clinical Practice Guidelines Steer-

ing Committee and Expert Panel. Synopsis of the 2006 Canadianclinical practice guidelines on the management and prevention ofobesity in adults and children. CMAJ. 2007;176:1103-1106.

2. Wing RR, Hill JO. Successful weight loss maintenance. Annu RevNutr. 2001;21:323-341.

3. Anderson JW, Konz EC, Frederich RC, Wood CL. Long-term weight-loss maintenance: A meta-analysis of US studies. Am J Clin Nutr.2001;74:579-584.

4. Burke LE, Steenkiste A, Musci E, Styn MA. A descriptive study ofpast experiences with weight-loss treatment. J Am Diet Assoc. 2008;108:640-647.

5. Nebeling L, Rogers CJ, Berrigan D, Hursting S, Ballard-Barbash R.Weight cycling and immunocompetence. J Am Diet Assoc. 2004;104:892-894.

6. Serdula MK, Mokdad AH, Williamson DF, Galuska DA, Mendlein JM,Heath GW. Prevalence of attempting weight loss and strategies forcontrolling weight. JAMA. 1999;282:1353-1358.

7. Green KL, Cameron R, Polivy J, Cooper K, Liu L, Leiter L, Heather-ton T. Weight dissatisfaction and weight loss attempts among Cana-dian adults. Canadian Heart Health Surveys Research Group. CMAJ.1997;157(suppl 1):S17-S25.

8. Lahti-Koski M, Männistö S, Pietinen P, Vartiainen E. Prevalence ofweight cycling and its relation to health indicators in Finland. ObesRes. 2005;13:333-341.

9. Field AE, Manson JE, Laird N, Williamson DF, Willett WC, ColditzGA. Weight cycling and the risk of developing type 2 diabetes amongadult women in the United States. Obes Res. 2004;12:267-274.

0. Burns CM, Tijhuis MAR, Seidell JC. The relationship between qualityof life and perceived body weight and dieting history in Dutch menand women. Int J Obes. 2001;25:1386-1392.

1. Vergnaud A-C, Bertrails S, Oppert J-M, Maillard-Teyssier L, Galan P,Hercberg S, Czernichow S. Weight fluctuations and risk for metabolicsyndrome in an adult cohort. Int J Obes. 2008;32:315-321.

2. Schulz M, Liese AD, Boeing H, Cunningham JE, Moore CG, Kroke A.Associations of short-term weight changes and weight cycling withthe incidence of essential hypertension in the EPIC-Potsdam Study. JHum Hypertens. 2005;19:61-67.

3. Marchesini C, Cuzzolaro M, Mannucci E, Dalle Grave R, Gennaro M,Tomasi F, Barantani EG, Melchionda N and the QUOVADIS Study

3

Group. Weight cycling in treatment-seeking obese persons: Data fromthe QWUOVADIS study. Int J Obes. 2004;28:1456-1462.

4. Gregg EW, Gerzoff RB, Thompson TJ, Williamson DF. Intentionalweight loss and death in overweight and obese US adults 35 years ofage and older. Ann Intern Med. 2003;138:383-389.

5. Brownell KD, Rodin J. Medical, metabolic, and psychological effects ofweight cycling. Arch Intern Med. 1994;154:1325-1330.

6. National Task Force on the Prevention and Treatment of Obesity.Weight cycling. JAMA.1994;272:1196-1202.

7. National Task Force on the Prevention and Treatment of Obesity.Dieting and the development of eating disorders in overweight andobese adults. Arch Intern Med. 2000;160:2581-2589.

8. Andres R, Muller DC, Sorkin JD. Long-term effects of change in bodyweight on all-cause mortality: A review. Ann Intern Med. 1993;119:737-743.

9. Blair SN, Shaten J, Brownell K, Collins G, Lissner L. Body weightchange, all-cause mortality, and cause-specific mortality in the Mul-tiple Risk Factor Intervention Trial. Ann Intern Med. 1993;119(7 Pt2):749-757.

0. Lissner L, Odell PM, D’Agostino RB, Stokes J III, Kreger BE, Be-langer AJ, Brownell KD. Variability of body weight and health out-comes in the Framingham population. N Engl J Med. 1991;324:1839-1844.

1. Hamm P, Shekelle RB, Stamler J. Large fluctuations in body weightduring young adulthood and twenty-five-year risk of coronary deathin men. Am J Epidemiol. 1989;129:312-318.

2. Drewnowski A, Kurth CL, Rahaim JE. Taste preferences in humanobesity: Environmental and familial factors. Am J Clin Nutr. 1991;54:635-641.

3. Korkeila M, Rissanen A, Kaprio J, Sørensen TIA, Koskenvuo M.Weight-loss attempts and risk of major weight gain: A prospectivestudy in Finnish Adults. Am J Clin Nutr. 1999;70:965-975.

4. Frederick IO, Rudra CB, Miller RS, Foster JC, Williams MA. Adultweight change, weight cycling, and prepregnancy obesity in relationto risk of preeclampsia. Epidemiology. 2006;17:428-434.

5. Field AE, Byers T, Hunter DJ, Laird NM, Manson JE, Williamson DF,Willett WC, Colditz GA. Weight cycling, weight gain, and risk ofhypertension in women. Am J Epidemiol. 1999;150:573-579.

6. Graci S, Izzo G, Savino S, Cattani L, Lezzi G, Berselli ME, Balzola F,Liuzzi A, Petroni ML. Weight cycling and cardiovascular risk factorsin obesity. Int J Obes. 2004;28:65-71.

7. Petersmarck KA, Teitelbaum HS, Bond JT, Bianchi L, Hoerr SM,Sowers MF. The effect of weight cycling on blood lipids and bloodpressure in the Multiple Risk Factor Intervention Trial Special Inter-vention Group. Int J Obes. 1999;23:1246-1255.

8. Muls E, Kempen K, Vansant G, Saris W. Is weight cycling detrimentalto health? A review of the literature in humans. Int J Obes. 1995;19(suppl 3):S46-S50.

9. Clark MM, King TK. Eating self-efficacy and weight cycling. A pro-spective clinical study. Eating Behav. 2000;1:47-52.

0. Simkin-Silverman LR, Wing RR, Plantinga P, Matthews KA, KullerLH. Lifetime weight cycling and psychological health in normal-weight and overweight women. Int J Eating Disord. 1998;24:175-183.

1. Kuehnel RH, Wadden TA. Binge eating disorder, weight cycling, andpsychopathology. Int J Eating Disord. 1994;15:321-329.

2. Wadden TA, Bartlett S, Letizia KA, Foster GD, Stunkard AJ. Rela-tionship of dieting history to resting metabolic rate, body composition,eating behavior, and subsequent weight loss. Am J Clin Nutr. 1992;56:203S-208S.

3. Li Z, Hong K, Wong E, Maxwell M, Heber D. Weight cycling in a verylow-calorie programme has no effect on weight loss velocity, bloodpressure and serum lipid profile. Diabetes Obesity Metab. 2007;9:379-385.

4. Provencher V, Drapeau V, Tremblay A, Després J-P, Bouchard C,Lemieux S. Eating behaviours, dietary profile and body compositionaccording to dieting history in men and women of the Québec FamilyStudy. Br J Nutr. 2004;91:997-1004.

5. Karelis AD, Faraj M, Bastard JP, St-Pierre DH, Brochu M,Prud’homme D, Rabasa-Lhoret R. The metabolically healthy butobese individual presents a favorable inflammation profile. J ClinEndocrinol Metab. 2005;90:4145-4150.

6. St-Pierre DH, Faraj M, Karelis AD, Conus F, Henry JF, St-Onge M,Tremblay-Lebeau A, Cianflone K, Rabasa-Lhoret R. Lifestyle behav-iours and components of energy balance as independent predictors ofghrelin and adiponectin in young non-obese women. Diabetes Metab.2006;32:131-139.

7. Cohen S, Kamarck T, Mermelstein R. A global measure of perceivedstress. JHSB. 1983;24:385-396.

April 2009 ● Journal of the AMERICAN DIETETIC ASSOCIATION 723

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

5

5

7

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

9. Mendelson BK, Mendelson MJ, White DR. Body-esteem scale foradolescents and adults. J Pers Assess. 2001;76:90-106.

0. Janz NK, Champion VL, Strecher VJ. The Health Belief Model. In:Glanz K, Rimer BK, Marcus Lewis F, eds. Health Behavior andHealth Education. Theory, Research, and Practice. 3rd ed. San Fran-cisco, CA: Jossey-Bass Inc; 2003:45-66.

1. Baranowski T, Perry CL, Parcel GS. How individuals, environments,and health behavior interact. In: Glanz K, Rimer BK, Marcus Lewis F,eds. Health Behavior and Health Education. Theory, Research, andPractice. 3rd ed. San Francisco, CA: Jossey-Bass Inc; 2003:165-184.

2. McDowell I, Newell C. Measuring Health. A Guide to Rating Scalesand Questionnaires. New York, NY: Oxford University Press; 1996:456-459.

3. Stewart AL, Greenfield S, Hays RD, Wells K, Rogers WH, Berry SD,McGlynn EA, Ware JE Jr. Functional status and well-being of pa-tients with chronic conditions. Results from the Medical OutcomesStudy. JAMA. 1989;262:907-913.

4. Stunkard AJ, Messick S. The Three-Factor Eating Questionnaire tomeasure dietary restraint, disinhibition and hunger. J PsychosomRes. 1985;29:71-83.

5. Ghiselli EE, Campbell JP, Zedeck S. Measurement Theory for theBehavioral Sciences. San Francisco, CA: W H Freeman and Company;1981.

6. Manuel de Nutrition Clinique. 3rd ed. Montréal, Québec: Ordre pro-

fessionnel des diététistes du Québec, 2000.

7. Field AE, Aneja P, Austin SB, Shrier LA, de Moor C, Gordon-Larson

24 April 2009 Volume 109 Number 4

P. Race and gender differences in the association of dieting and gainsin BMI among young adults. Obesity. 2007;15:456-464.

8. Miller DS, Parsonage S. Resistance to slimming: Adaptation or illu-sion? Lancet. 1975;1:773-775.

9. Doucet E, St-Pierre S, Alméras N, Després J-P, Bouchard C, Trem-blay A. Evidence for the existence of adaptative thermogenesis duringweight loss. Br J Nutr. 2001;85:715-723.

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

1. Astrup A, Gotzsche PC, van de Werken K, Ranneries C, Toubro S,Raben A, Buemann B. Meta-analysis of resting metabolic rate informerly obese subjects. Am J Clin Nutr. 1999;69:1117-1122.

2. Vogels N, Diepvens K, Westerterp-Plantenga MS. Predictors of long-term weight maintenance. Obes Res. 2005;13:2162-2168.

3. Pasman WJ, Saris WH, Westerterp-Plantenga MS. Predictors ofweight maintenance. Obes Res. 1999;7:43-50.

4. Owecki M, Miczke A, Pupek-Musialik D, Bryl W, Cymerys M, NikischE, Sowinski J. Circulating serum adiponectin concentrations do notdiffer between obese and non-obese Caucasians and are unrelated toinsulin sensitivity. Horm Metab Res. 2007;39:25-30.

5. Lindsay RS, Funahashi T, Hanson RL, Matsurzawa Y, Tanaka S,Tataranni PA, Knowler WC, Krakoff J. Adiponectin and development oftype 2 diabetes in the Pima Indian population. Lancet. 2002;360:57-58.

6. Fabricatore AN. Behavior therapy and cognitive-behavioral therapyof obesity: Is there a difference? J Am Diet Assoc. 2007;107:92-99.

7. Houts P, Sharada S, Klassen AC, Robinson EB, McCarthy M. A

problem solving approach to nutrition education and counseling. JNutr Educ Behav. 2006;38:254-258.