Week 11 Assignment

Week10RaichEtAl2010.pdf

Home >Psychology homework help >Week 11 Assignment

RESEARCH ARTICLE

Evaluation of a School-Based Programme of Universal Eating Disorders Prevention: Is it More Effective in Girls at Risk? R. M. Raich1*, M. Portell2 & M. A. Peláez-Fernández3

1Dep. de Psicologia Clı́nica i de la Salut, Facultat de Psicologia, Universitat Autònoma de Barcelona, Bellaterra, Cerdanyola del Vallès, Spain

2Dep. de de Psicobiologia i Metodologia de les Ciències de la Salut, Facultat de Psicologia, Universitat Autònoma de Barcelona Bellaterra.

Cerdanyola del Vallès, Spain

3Dpto. de Psicologı́a Social, Antropologı́a Social, Trabajo Social y Servicios Sociales, Facultad de Psicologı́a, Universidad de Málaga, Campus

de Teatinos S/N, Málaga, Spain

Abstract

There is currently controversy surrounding the effectiveness of universal versus selective prevention in eating

disorders (ED). The present study aims at evaluating the effectiveness of universal school-based ED prevention

administered to female secondary school students (n¼ 349). Students received either the full prevention programme (learning basic concepts of nutrition, criticism of aesthetic models of beauty emphasising extreme thinness, media

literacy (ML)), a partial version of the programme (without nutritional education), or no prevention programme.

Students were also classified on the presence or absence of distinct risk factors for ED: Early menarche, overweight,

dieting, negative attitudes to food and perceived pressure to be thin. Pre-test data were collected 1 week prior to

implementation of the prevention programme, and post-test data were collected on the last day of the programme.

Results suggested that both the full and partial prevention programmes reduced perceived pressure to be thin and

improved eating attitudes and knowledge of nutrition in all the participants (regardless of risk); however, greater

effect sizes were found among particular high-risk groups (early menarche, overweight and highly influenced by

aesthetic models of beauty emphasising extreme thinness). School-based programmes of universal intervention may

have an important role to play in the prevention of ED. Copyright # 2009 John Wiley & Sons, Ltd and Eating

Disorders Association.

Keywords

eating disorders; prevention; risk factors

*Correspondence

R. M. Raich, Dep. de Psicologia Clı́nica i de la Salut, Facultat de Psicologia, Universitat Autònoma de Barcelona, 08193 Bellaterra,

Cerdanyola del Vallès, Spain.

Email: [email protected]

Published online 13 October 2009 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/erv.968

Introduction

There is currently controversy surrounding the effec-

tiveness of universal versus selective prevention in

eating disorders (ED). Selective prevention aims at

smaller groups of non-symptomatic individuals that are

at high risk, while universal prevention aims at the

general population or large groups of healthy individ-

uals (Mrazek & Haggerty, 1994). Stice and Shaw (2004)

and Stice, Shaw, and Marti (2007) argue in favour of

Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association. 49

selective interventions with adolescents and specific

groups at risk, since these may produce more significant

and lasting effects compared to universal interventions.

Other researchers (Levine & Smolak, 2006; Stewart,

Carter, Drinkwater, Hainsworth, & Fairburn, 2001)

suggest that school-based preventive programmes

should be universal and strengthened every year until

the end of high school.

Stice and colleagues (2004, 2007) maintain that

individuals at high risk of developing an ED are more

motivated to accept the content of prevention

programmes and consequently benefit more. Areas of

high risk for developing ED are, however, many.

Empirically established risk factors for ED that could

be considered when identifying a target population

for selective prevention include: early menarche,

overweight or obesity, restrictive weight-loss dieting,

distorted attitudes towards food and aesthetic models

of beauty that emphasise extreme thinness.

Longitudinal studies have found that early menarche

is associated with more unhealthy eating habits, ED

symptoms, and higher body weight (Fairburn, Welch,

Doll, Davies, & O’Connor, 1997). Kaltiala-Heino,

Rimpela, Rissanen, and Rantanen (2001) reported that

early puberty predicted the onset of bulimic behaviours.

Williams and Currie (2000) found significant associ-

ations between early maturation and body dissatisfac-

tion. On the other hand, the relationship between early

maturation and ED symptoms is confounded by

body weight. A longitudinal study by Must, Naumova,

Phillips, Blum, Dawson-Hughes, and Rand (2005)

found that higher BMI was predictive of early matu-

ration. Consistent with this finding, when body weight

was controlled, Striegel-Moore, McMahon, Biro,

Schreiber, Crawford, and Voorhees (2001) found that

there were no differences in eating problems between

women who presented with early or late menarche;

eating problems were associated with weight rather

than timing of maturation.

The alarming increase in obesity and overweight

among adolescents not only affects their general health,

but may also increase the risk of ED. These young

people are more likely to engage in unhealthy weight

control behaviours, such as skipping meals, eating less

fruit and vegetables, eating alone, avoiding food or have

strict dietary rules, and are less likely to engage in

healthy physical activity or healthy eating (Boutelle,

Neumark-Sztainer, Story, & Resnick, 2002; Neumark-

Sztainer, Story, Hannan, Perry, & Irving, 2002).

Between 20 and 40% of adolescents try to lose weight

by dieting (Daee, Robinson, Lawson, Turpin, Gregory,

& Tobias, 2002; Neumark-Sztainer & Hannan, 2000;

Neumark-Sztainer et al., 2002), and dieting has been

identified as an important risk factor for ED (Patton,

Selzer, Coffey, Carlin, &Wolfe, 1999). However, despite

its popularity as a weight-loss technique, Neumark-

Sztainer, Wall, Guo, Story, Haines, and Eisenberg,

(2006) found that dieting had the opposite effect in a

5-year follow-up longitudinal study where BMI

increased significantly among dieters.

Distorted attitudes toward food, eating and weight,

as measured by screening instruments such as the

Eating Attitudes Test (Garner and Garfinkel, 1979),

may increase the risk of developing an ED. Girls who

engaged in unhealthy behaviours to control their weight

had twice the risk of later engaging in more extreme

weight control behaviour, and six times greater risk of

binge eating during a 5 year follow-up compared to girls

who were not trying to control their weight (Neumark-

Sztainer, Wall, et al., 2006). Individuals espousing an

aesthetic model of beauty emphasising extreme thin-

ness may be more likely to engage in unhealthy

behaviour to reduce weight (Neumark-Sztainer, Pax-

ton, Hannan, Haines, & Story, 2006; Olmsted & Mc

Farlane, 2004; Stice, Kendra Orjada, & Tristan, 2006).

Mass media pressure and other social influences to be

thin may influence many women to feel ‘bad within

their body’ and to experience greater body dissatisfac-

tion, resulting in increased risk for developing an ED.

The present study investigates a universal school-

based ED prevention programme. The programme has

a multimedia format, is didactic, interactive and is

conducted in schools. It involves three components:

Learning basic concepts of nutrition (NUT); criticism

of the prevailing feminine aesthetic beauty model

(ABM) of extreme thinness and media literacy (ML)

involving a critique of information presented via mass

media. Components of the programme are designed to

modify ED risk factors, and have been used in previous

preventive interventions (Stice & Shaw, 2004; Stice

et al., 2007). The programme has recently been

published in an extended version (Raich, Sánchez-

Carracedo, & López-Guimerà, 2008), and has been

evaluated by Raich, Sánchez-Carracedo, López-

Guimerà, Portell, Moncada, and Fauquet (2008) using

a quasi-experimental design with treatment and control

groups from a representative sample of 13-year-old

girls.

50 Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association.

School-Based Universal Prevention of Eating Disorders R. M. Raich et al.

The specific aims of the present study were: (1) To

examine how Spanish adolescents at risk for ED differ

from their peers in eating attitudes, aesthetic body

models and nutritional knowledge, in relation to the

presence or absence of important risk factors (i.e. early

menarche, overweight, dieting, distorted attitudes

towards food or perceiving a high pressure to become

thin); and (2) to evaluate the differential effectiveness

of the programme in groups of students with and

without these risks. As such, we attempted to evaluate

the hypothesis of Stice and Shaw (2004) that high-risk

participants in a prevention programme for ED will

change more compared to lower risk participants.

To this end we aimed to estimate the programme’s

effect size in the presence of different risk factors,

similar to the work of Taylor et al. (2006).

Method

Participants

Participants were selected by stratified random

sampling, using type of school as a stratification base

(public and publicly subsidised private schools1), and

school as the sample unit. Using this method, 349 girls

who were registered in the second course of Obligatory

Secondary Education (O.S.E.) in 13 schools of the city

of Terrassa were selected. Although assessment and

intervention involved both males and females in order

to maintain a natural atmosphere in class, only data on

females were included in the study since the interven-

tions focused on criticizing the prevailing feminine

aesthetic beauty model of extreme thinness, and also

because ED affect mostly females.2 Of the girls initially

contacted, 288 (83%) completed the pre and post

evaluation, Seventeen per cent did not complete all

the questionnaires or were absent from school on the

days of the post-evaluation. Average age of the sample

was 13 years (SD¼ 0.47), and the average BMI was

21.31 Kg/m2 (SD¼ 3.35). Figure 1 provides more details of the selection and allocation of participants.

The study was approved and facilitated by Terrassa

Municipal Institute of Health and Social Welfare

(IMSBS), which obtained permission from the schools

and families for the study to be conducted.

Measures

Weight and height were measured using scales and a

measuring stick for calculation of BMI. A biographical

data sheetwasused to collect backgrounddata on current

age, gender, age at menarche and dieting behaviour

(participants were asked: ‘Do you diet to lose weight’?).

Eating Attitudes Test (EAT-40; Garner and Garfinkel,

1979; Spanish version developed by Castro, Toro,

Salamero, & Guimerà, 1991) was used to measure

eating attitudes. The Spanish version has acceptable

psychometric characteristics (i.e. 91% sensitivity and

69.2% specificity for a cut-off point of 20).

Questionnaire of Influences of the Aesthetic Beauty

Model (CIMEC-26; Toro, Salamero, & Martı́nez, 1994).

This instrument was created in Spanish and measures

an individual’s perception of pressure to be thin. The

CIMEC-26 has shown acceptable psychometric proper-

ties (i.e. internal consistency of 0.92 for anorexics, and

0.91 for controls; sensitivity: 83.1%; and specificity:

64%). When compared with its longer version

(i.e. CIMEC-40), the CIMEC-26 has been found to

discriminate effectively between participants with and

without ED [t(100,3)¼ 6.44; p< .001] (Raich, Sánchez- Carracedo, López-Guimerà, Portell, & Fauquet, 2007).

Questionnaire of Nutrition (NUT-Q). This instru-

ment, designed for the study, measures nutritional

knowledge in relation to particular types of food. The

questionnaire comprises 10 multiple-choice questions;

internal consistency in the pre-test was low (a¼ 0.53) but acceptable in the post-test (a¼ 0.72).

Procedure

Effects of the universal prevention programme were

evaluated using a quasi-experimental group design with

pre- and post-test measures. The independent variable

of the design had three levels: (1) complete intervention

(ABMþMLþNUT), (2) partial intervention (ABMþ ML) or (3) no intervention (control group). Based

on responses to the pre-intervention assessment, five

binary variables were defined that established the

1 The publicly subsidised private schools are centres of education

created by the Spanish Sociedad Civil, but maintained with public

funds. LOGSE: Ley Orgánica de Ordenación General del Sistema

Educativo de España, 1990. (Published in the Boletı́n Oficial del

Estado –BOE- 4th October 1990).

2 Recent ED epidemiological studies in the general population

samples have found a female/male ratio of ED of 5.63:1 (Warner,

Warner, Matuszak, Rachal, Flynn, & Grieger, 2007), and 8.42:1

(Peláez-Fernández, Labrador, & Raich, 2007).

Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association. 51

R. M. Raich et al. School-Based Universal Prevention of Eating Disorders

presence and absence of risk factors in the two

intervention groups and the control group. Risk factors

were defined according to the following criteria:

1. Early Menarche: Response to the question ‘Did you

experience menarche a year ago, or more than a year

ago’? combined with age, was used to determine

whether participants had experienced early

menarche, which was defined as menarche before

they were 12 years old.

2. Overweight: This was defined using the BMI cut-off

for overweight in 13 year-old girls established based

by Cole, Bellizzi, Flegal, and Dietz (2000); i.e.

BMI� 22.58 Kg/m2. 3. Dieting: Participants were classified according to their

answer to the question ‘Do you diet to lose weight’?

4. Distorted attitudes to food were defined as a score of

20 or more on the EAT (cut-off point established in

the Spanish version of Castro et al., 1991).

5. Perceived high pressure to become thin. This risk

factor was defined as a score of 16 or more on the

CIMEC (i.e. cut-off point on the instrument based

on Toro et al., 1994).

Data were collected between the end of 2002 and the

beginning of 2003. Pre-test assessment was carried out

one week before the intervention, and the same evalu-

ation was carried out on the last day of the programme,

with the exception of BMI, which was only measured at

pre-test. Data were analysed using SPSS (version 14.02).

Results

The f coefficient was used to examine the relation

between binary variables defined by the presence/

absence of each risk factor. The relation between the

presence/absence of each risk factor and the dependent

variables was examined using MANOVA. Differential

effects of interventions in relation to different risk

factors were also examined using MANOVA. For these

analyses dependent variables were defined as change

scores3 (generated from the difference between the pre-

Figure 1 Flow chart

3 The use of change scores allowed controlling for baseline levels.

The decision to use change scores was based on the independence

between the allocation procedure and baseline measures (Judd &

Kenny, 1981).

52 Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association.

School-Based Universal Prevention of Eating Disorders R. M. Raich et al.

test and post-test scores). In all cases, differences were

calculated so that positive values indicated that the

intervention produced the expected change. For each

one of the risk factors, a factorial MANOVAwas carried

out with intervention (control, partial or complete

intervention) and group (risk or no risk in terms of

specific risk factors) as independent variables, and

change scores obtained on the EAT, CIMEC and NUT-

Q as the dependent variable.

Table 1 describes the relation between the five risk

factors, and provides the number of participants in each

group who presented with particular risks. A statisti-

cally significant relationship was found between over-

weight, dieting, distorted attitudes to food and high

influence of the thin aesthetic model. However, there

was no statistically significant relationship between

early menarche and the other risk factors. It should

be noted that participants who presented with early

menarche had significantly higher BMIs when com-

pared to those who experienced menarche later

(M¼ 21.7 Kg/m2 vs. M¼ 20.7 Kg/m2, t(283)¼ 2.49, p¼ 0.013). Table 2 provides results of MANOVA at pre-test

for participants with and without each risk factor.

Significant between-group differences on the EAT,

CIMEC and NUT-Q were found for all risk factors

Table 2 Pre-test differences between participants with and without specific risk factors

Risk factor N EAT CIMEC NUT.Q MANOVA F p h2

M SD M SD M SD df Wilk’s L

Early menarche

No 124 17.72 11.58 14.60 11.18 2.84 1.75 3284 0.99� 1.28 0.282 0.01

Yes 164 17.68 12.23 16.37 10.83 2.99 1.75

Overweight

No 195 15.34 10.55 12.70 9.49 2.95 1.78 3281 0.86� 15.48 <0.001 0.14

Yes 90 22.31 12.97 21.47 11.47 2.89 1.68

Restrictive dieting

No 252 16.04 10.20 14.45 10.63 2.92 1.80 3284 0.86� 15.27 <0.001 0.14

Yes 36 29.31 16.22 23.69 10.23 3.00 1.35

Distorted attitudes towards food

No 197 — — 11.62 7.80 2.95 1.84 2285 0.71y 57 <0.001 0.29

Yes 91 — — 24.24 11.97 2.87 1.54

Yes 117 25.23 13.89 — — 2.81 1.72

High influence of the thin aesthetic model

No 171 12.54 6.56 — — 3.01 1.77 2285 0.73z 54.1 <0.001 0.28

Yes 117 25.23 13.89 — — 2.81 1.72

� The dependent variables are: EAT, CIMEC and NUT-Q. y The dependent variables are: CIMEC and NUT-Q. z The dependent variables are: EAT and NUT-Q.

Table 1 Association between the presence of risk factors and the sample distribution according to the type of intervention and risk group

Risk factor 1 2 3 4 5 Control ABMþML ABMþMLþNUT

1. Early menarche y — 0.084 �0.053 0.018 0.091 83(51)z 45(38) 36(35) 2. Overweight — 0.298� 0.259� 0.308� 42(92) 30(50)x 18(53)

3. Dieting — 0.285� 0.286� 16(118) 14(69) 6(65)

4. Distorted attitudes towards food — 0.441� 40(94) 26(57) 25(46)

5. High influence of the thin aesthetic model — 53(81) 36(47) 28(43)

Distribution of the recount of risk factors: 25/44/65jj 16/21/46 15/28/28

� p< 0.05. yf has been used to reflect the relation between dichotomous variables. z In parentheses: the number of cases without the risk factor. x The lack of three cases is due to the absence of values in the variable weight or size used to generate the variable. jj The first number indicates the cases without risk factors; the second one, the cases with only one; and the third, the cases with two or more risk factors.

Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association. 53

R. M. Raich et al. School-Based Universal Prevention of Eating Disorders

except early menarche. When univariate comparisons

were made between the presence and absence of risk

factors, statistically significant differences were found

on the EAT and the CIMEC, but not the NUT-Q.

The second objective of the study was to analyse the

effect of intervention in relation to each risk factor. For

each risk factor (defined as its presence or absence),

MANOVA was used to examine the main effects of

group (i.e. participants with high or no risk for a

particular risk factor) and intervention type (i.e.

control, partial and complete interventions), along

with interaction effects (i.e. risk group� intervention). Results are summarised in Table 3. The main effects of

intervention were significant across all risk groups;

however, interaction effects between group and inter-

vention did not attain statistical significance in any of

the risk groups.

Estimation of the programmes’ effect sizes in relation

to different risk factors, focused only on high-risk

participants, and is presented in Table 4. The table

presents averages and standard deviations of change

scores (defined as ‘improvement’ in post-test with

respect to the pre-test) for the control group and for

both experimental groups (partial and complete), along

with effect sizes (g) of each type of intervention with

respect to the control group, and a confidence interval

of 95%. Hedges’ g, which is similar to Cohen’s d (Kline,

2004), was used to measure effect size. Interpretation of

the measure was in keeping with the criteria proposed

by Cohen (1988): 0.2–0.5 indicating a small effect;

0.5–0.8, a medium effect; and 0.8 or above, a large

effect. Based on these principles: (1) medium effects of

the partial intervention on the EAT and the CIMEC

were found in overweight participants; (2) medium

effects of the complete intervention were displayed on

the CIMEC and NUT-Q in the early menarche group

and (3) medium effects of the complete intervention

were found on the NUT-Q among participants with a

perceived high influence of the thin aesthetic model.

Discussion

The present study aimed to examine how Spanish

adolescent girls, classified according to the presence or

absence of risk factors for ED, differ in eating attitudes,

aesthetic body models and nutritional knowledge. The

risk factors examined were early menarche, overweight,

dieting, distorted attitudes to food and perceived

pressure to be thin. The study also aimed to evaluate the

effectiveness of a universal school-based programme of

ED prevention in groups of girls with and without

particular risk factors. Strong points of the study

included examination of important risk factors

identified in the literature, and the fact that it was

carried out in schools using students who were not pre-

selected based on level of risk, i.e. using stratified

random sampling and subsequent classification based

on the presence or absence of specific risk factors.

An important weak point of the study was the fact

that statistical analyses could have been compromised

due to the small size of sub-samples, resulting from

the division of the sample into groups based on the

presence/absence of different risk factors.

As regards the first aim of the study, no relationship

was found between early menarche and the other risk

factors, which may be because ED symptoms are more

a function of overweight than early menarche (Must

et al., 2005; Striegel-Moore et al., 2001). Overweight

participants presented with more distorted eating

Table 3 MANOVA of main and interaction effects between inter-

vention (control, partial and complete interventions) and group

(participants with high or no risk for individual risk factors) on

dependent variables (change scores on EAT, CIMEC, NUT-Q)

df Wilks’ L F p h2

Early Menarche�

Group (G) 3279 0.99 1.05 0.371 0.01

Intervention (I) 6558 0.89 5.64 <0.001 0.06

Interaction (G� I) 6558 0.98 0.89 0.505 0.01 Overweight�

Group (G) 3276 0.96 3.88 0.01 0.04

Intervention (I) 6552 0.91 4.58 <0.001 0.05

Interaction (G� I) 6552 0.97 1.24 0.284 0.01 Restrictive dieting�

Group (G) 3279 0.94 5.60 0.001 0.06

Intervention (I) 3280 0.38y 3.56 0.015 0.04

Interaction (G� I) 6558 0.99 0.73 0.630 0.01 Distorted attitudes towards foodz

Group (G) 2281 0.98 3.35 0.037 0.07

Intervention (I) 4562 0.91 6.73 <0.001 0.05

Interaction (G� I) 4562 0.98 1.32 0.263 0.01 High influence of the thin aesthetic modelx

Group (G) 2280 0.93 10.17 <0.001 0.02

Intervention (I) 4560 0.91 6.66 <0.001 0.05

Interaction (G� I) 4560 0.99 0.51 0.732 0.01

‘Group (G)’: defined as the presence/absence of each risk factor. ‘Intervention (I)’: in all the analysis, it is referred to ‘Control’, ‘CMþAM’, ‘CMþAMþNUT’. � The dependent variables are: EAT, CIMEC and NUT-Q. yResults are based on the Roy test. z The dependent variables are: CIMEC and NUT-Q. x The dependent variables are: EAT and NUT-Q.

54 Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association.

School-Based Universal Prevention of Eating Disorders R. M. Raich et al.

attitudes and experienced a greater perceived pressure

to be thin than their normal or lower weight peers,

which is in keeping with the findings of Striegel-Moore

et al. (2001) and Neumark-Sztainer, Paxton et al.

(2006). Dieters had more distorted eating attitudes and

experienced a greater pressure to be thin, which is

in line with previous work (Boutelle et al., 2002;

Neumark-Sztainer, Paxton, et al., 2006). The percen-

tage of participants who reported dieting was less than

that reported in similar studies (Daee et al., 2002;

Neumark-Sztainer & Hannan, 2000; Neumark-Sztainer

et al., 2002). Our low figures may be due to the fact that

dieting was defined in terms of present behaviour only.

Participants who presented with more distorted eating

attitudes also presented greater perceived pressure to be

thin and vice versa. Nutritional knowledge was similar

across groups, but generally speaking surprisingly low,

despite instruction in this subject within the normal

school curriculum.

When the effectiveness of our full and partial school-

based intervention programmes were examined, results

tended to be at odds with the conclusions of Stice and

Shaw (2004). That is to say, changes resulting from the

intervention programmes were not significantly greater

among participants in the high-risk groups. However, it

is possible that the statistical power of the MANOVA

analyses was limited by the size of the individual groups.

Conclusions about the effectiveness of the present

intervention programmes regardless of risk factors

should, therefore, be treated with caution. To this end

we are planning to replicate the study in a larger sample

in the future.

When effect sizes of the two intervention pro-

grammes were examined among different risk groups,

greater effect sizes were found for partial interventions

(i.e. criticism of the prevailing feminine aesthetic model

of beauty and ML) among overweight participants,

who achieved greater reduction in perceived pressure

to be thin. The greatest effects on perceived pressure

to be thin (CIMEC) when the full intervention

programme was examined (i.e. the above interventions

plus nutritional education) were found in the early

menarche group, while the greatest effects on nutri-

tional knowledge for the full programme were found

among individuals at-risk in terms of restrictive dieting

and perceived pressure to be thin. On the one hand,

these results may mean that educating students to

criticise and challenge the thin aesthetic ideal for

women may be especially important in girls who are

overweight and those who experience early menarche

(and accordingly have a higher BMI than their peers).

While on the other hand, nutritional counselling and

Table 4 Effectiveness of preventative interventions (based on difference scores between pre-test and post-test) in individual groups at risk,

and effect sizes using Hedges’ g.

Risk group Control Partial

intervention

Complete

intervention

Effect of partial

intervention

(partial vs. control)

Effect of complete

intervention

(complete vs. control)

M SD M SD M SD g 95%CI g 95%CI

Early menarche

EAT 1.67 9.13 3.38 5.96 2.56 8.13 0.21 (�0.15–0.57) 0.10 (�0.29–0.49) CIMEC 1.65 5.67 2.84 5.20 4.50 6.95 0.22 (�0.15–0.58) 0.47 (0.07–0.86) NUT-Q 0.08 1.74 �0.02 2.04 1.22 2.40 �0.06 (�0.42–0.31) 0.58 (0.18–0.98)

Overweight

EAT 0.64 8.77 4.63 7.05 4.78 10.17 0.49 (0.02–0.97) 0.45 (�0.11–1.01) CIMEC 2.19 5.03 5.48 4.65 4.39 9.28 0.68 (0.19–1.16) 0.33 (�0.22–0.89) NUT-Q 0.17 1.64 0.03 2.01 1.22 2.56 �0.07 (�0.54–0.39) 0.54 (�0.02–1.10)

Restrictive dieting

EAT 5.81 13.82 6.57 8.17 10.33 5.43 0.07 (�0.65–0.78) 0.37 (�0.58–1.31) CIMEC 3.44 4.24 5.00 5.59 3.17 10.87 0.32 (�0.40–1.04) �0.04 (�0.98–0.90) NUT-Q �0.25 2.02 0.36 2.02 1.17 1.47 0.30 (�0.42–1.02) 0.75 (�0.22–1.71)

Distorted attitudes towards food

CIMEC 3.30 7.16 5.64 5.46 5.00 8.65 0.36 (�0.14–0.85) 0.22 (�0.28–0.72) NUT-Q 0.25 1.71 �0.35 2.23 0.96 1.84 �0.31 (�0.81–0.19) 0.40 (�0.10–0.91)

High influence of the thin aesthetic model

EAT 1.60 11.41 4.94 7.80 5.11 10.38 0.33 (�0.10–0.76) 0.32 (�0.14–0.78) NUT-Q �0.11 1.59 �0.06 2.06 1.04 2.03 0.03 (�0.39–0.46) 0.66 (0.19–1.13)

Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association. 55

R. M. Raich et al. School-Based Universal Prevention of Eating Disorders

education may be particularly important among girls

who are dieting and are already highly influenced by a

thin body ideal.

In the meta-analysis of Stice and Shaw (2004),

60 studies of universal and selective prevention were

analysed, and in only 20 of these studies could

the effects of internalisation of a thin body ideal be

estimated (i.e. a variable similar to that obtained in the

present study using the CIMEC). Only 11 of these

20 studies, concerning selective prevention, obtained

effect sizes that approximated those found in our work.

In the present study, results of the MANOVA analyses

indicated that the main intervention effects were

significant across all risk factors, whereas interaction

effects were non-significant. In other words, overall

changes in disordered attitudes to eating, perceived

pressure to be thin and nutritional knowledge were

greater depending on intervention generally, regardless

of risk.

Taken as a whole, our results run counter to the

conclusions of Stice and Shaw (2004) who argue in

favour of selective forms of prevention with particular

groups at risk for ED. Our work suggests that universal

forms of prevention may be effective, but once

individuals exhibit specific risk factors it may be

important to target preventative interventions to

address the particular issues related to these risk

factors. Nevertheless, it should be borne in mind that

the preventative interventions used in the present study

were not intended to be selective, but rather universal,

which could partially explain why our results appear

to be at odds with the conclusions of Stice and

Shaw (2004). However, the improvements found in

the different risk groups in the current study are

comparable to the ones found by Coughlin and

Kalodner (2006) who obtained significant positive

results in ED risk groups using interventions based

around ML. In order to more effectively combat ED

it may be important for schools to offer educational

programmes that call into question perceptions of

women in the mass media generally, and a feminine

ideal of beauty based around thinness particularly,

while working to improve knowledge of nutrition.

Acknowledgements

This study was funded partly, thanks to the research

project ‘Prevention of Eating Disorders in the School:

Ecological Intervention’ financed by the Ministry of

Education and Science of Spain: SEJ2005-07099.

Authors are grateful to the editor for his help.

References

Boutelle, K., Neumark-Sztainer, D., Story, M., & Resnick,

M. D. (2002). Weight control behaviours among obese,

overweight, and non-overweight adolescents. Journal of

Paediatric Psychology, 28, 40–50.

Castro, J., Toro, J., Salamero, M., & Guimerà, E. (1991). The

eating attitudes test: Validation of the Spanish version.

Psychological Assessment, 7, 175–190.

Cohen, J. (1988). Statistical power analysis for the beha-

vioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Cole, T. J., Bellizzi, M. C., Flegal, K. M., & Dietz, W. H.

(2000). Establishing a standard definition for child over-

weight and obesity worldwide: International survey. Brit-

ish Medical Journal, 320, 1240–1243.

Coughlin, J. W., & Kalodner, C. (2006). Media literacy as a

prevention intervention for college women at low or high

risk for eating disorders. Body Image, 3, 35–43.

Daee, A., Robinson, P., Lawson, M., Turpin, J. A., Gregory,

B., & Tobias, J. D. (2002). Psychological and physiologic

effects of dieting in adolescents. South Medicine Journal,

95, 1032–1041.

Fairburn, C. G., Welch, S. L., Doll, H. A., Davies, B. A., &

O’Connor, M. E. (1997). Risk factors for bulimia nervosa:

A community-based case-control study. Archives of Gen-

eral Psychiatry, 54, 509.

Garner, D. M., & Garfinkel, P. E. (1979). The eating attitudes

test: An index of the symptoms of anorexia nervosa.

Psychological Medicine, 9, 273–279.

Judd, C. M., & Kenny, D. A. (1981). Estimating the effects of

social interventions. Cambridge: Cambridge University

Press.

Kaltiala-Heino, R., Rimpela, M., Rissanen, A., & Rantanen,

P. (2001). Early puberty and early sexual activity are

associated with bulimic-type eating pathology in middle

adolescence. Journal of Adolescence Health, 28, 346–352.

Kline, R. N. (2004). Beyond significance testing. Washing-

ton: American Psychological Association.

Levine, M. P., & Smolak, L. (2006). Prevention of eating

problems and eating disorders: Theory, research, and

practice. Mahwah, NJ: Lawrence Erlbaum Associates.

Mrazek, P. J., & Haggerty, R. J. (Eds.). (1994). Reducing

risks for mental disorders: Frontiers for prevention inter-

vention research (report of the Institute of Medicine’s

Committee on Prevention of Mental Disorders).

Washington, DC: National Academy Press.

Must, A., Naumova, E. N., Phillips, S. M., Blum, M.,

Dawson-Hughes, B., & Rand, W. M. (2005). Childhood

overweight and maturational timing in the development

56 Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association.

School-Based Universal Prevention of Eating Disorders R. M. Raich et al.

of adult overweight and fatness: The Newton girls study

and its follow-up. Pediatrics, 116, 620–627.

Neumark-Sztainer, D., & Hannan, J. (2000). Weight-related

behaviours among adolescent girls and boys: Results from

a national study. Archieves of Pediatric Adolescent Medicine,

154, 569–577.

Neumark-Sztainer, D., Paxton, S. J., Hannan, P. J., Haines,

J., & Story, M. (2006). Does body satisfaction matter?

Five-year longitudinal associations between body satis-

faction and health behaviors in adolescent females and

males. Journal of Adolescente Health, 39, 244–251.

Neumark-Sztainer, D., Story, M., Hannan, J., Perry, C. L., &

Irving, L. M. (2002). Weight-related concerns and beha-

viours among overweight and non-overweight adole-

scents: Implications for preventing weight-related

disorders. Archives of Paediatric Adolescent Medicine,

156, 171–178.

Neumark-Sztainer, D., Wall, M., Guo, J., Story, M., Haines,

J., & Eisenberg, M. (2006). Obesity, disordered eating, and

eating disorders in a longitudinal study of adolescents:

How do dieters fare 5 years later? Journal of American Diet

Association, 106, 559–568.

Olmsted, M., & Mc Farlane, T. (2004 , 25 de agosto). Body

weight and body image. BMC Women’s Health, 4,

S5 (Recuperado el 21 de mayo de 2007) http://www.

biomedcentral.com/1472-6874 /4/S1/S5

Patton, G. C., Selzer, R., Coffey, C., Carlin, J. B., & Wolfe, R.

(1999). Onset of adolescent eating disorders: Population

based cohort study over 3 years. British Medical Journal,

318, 765–768.

Peláez-Fernández, M. A., Labrador, F. J., & Raich, R. M.

(2007). Prevalence of eating disorders among an adoles-

cent and young adult scholastic population in the Region

of Madrid (Spain). Journal of Psychosomatic Research, 62,

681–690.

Raich, R. M., Sánchez-Carracedo, D., López-Guimerà, G.,

Portell, M., & Fauquet, J. (2007). Eating disorders

prevention with a multimedia program. Cuadernos de

Medicina Psicosomática y Psiquiatrı́a de Enlace, 81,

47–71.

Raich, R. M., Sánchez-Carracedo, D., & López-Guimerà, G.

(2008). Alimentación, modelo estético femenino y medios

de comunicación (Eating, Feminine Aesthetic Model And

Mass Media). Barcelona: Gra.

Raich, R. M., Sánchez-Carracedo, D., López-Guimerà, G.,

Portell, M., Moncada, A., & Fauquet, J. (2008).

A controlled assessment of a school-based preventive

program for reducing eating-disorder risk factors in

adolescent Spanish girls. Eating Disorders: The Journal

of Treatment and Prevention, 16, 255–272.

Stewart, D. A., Carter, J. C., Drinkwater, J., Hainsworth, J., &

Fairburn, C. G. (2001). Modification of eating attitudes

and behaviour in adolescent girls: A controlled study.

International Journal of Eating Disorders, 29, 107–118.

Stice, E., Kendra Orjada, K., & Tristan, J. (2006). Trial of a

psychoeducational eating disturbance intervention for

college women: A replication and extension. International

Journal of Eating Disorders, 39, 233–239.

Stice, E., & Shaw, H. (2004). Eating disorder prevention

programs: A meta-analytic review. Psychological Bulletin,

130, 206–227.

Stice, E., Shaw, H., & Marti, N. (2007). A meta-analytic

review of eating disorder prevention programs: Encoura-

ging findings. Annual Review of Clinical Psychology, 3,

207–231.

Striegel-Moore, R. H., McMahon, R., Biro, F., Schreiber, G.,

Crawford, P. B., & Voorhees, C. (2001). Exploring the

relationship between timing of menarche and eating

disorder symptoms in Black and White adolescent girls.

International Journal of Eating Disorders, 30, 421–433.

Taylor, C. B., Bryson, S., Luce, K. H., Cunning, D., Doyle,

A. C., Abascal, L. B., et al. (2006). Prevention of eating

disorders in at-risk college age women. Archives of

General Psychiatry, 63, 881–888.

Toro, J., Salamero, M., & Martı́nez, E. (1994). Assessment of

sociocultural influences on the aesthetic body shape

model in anorexia nervosa. Acta Psychiatrica Scandina-

vica, 89, 147–151.

Warner, C., Warner, C., Matuszak, T., Rachal, J., Flynn, J., &

Grieger, T. A. (2007). Disordered eating in entry-level

military personnel. Military Medicine, 172, 147–151.

Williams, J. M., & Currie, C. (2000). Self-esteem and physical

development in early adolescence: Pubertal timing and body

image. Journal of Early Adolescence, 20, 129–149.

Eur. Eat. Disorders Rev. 18 (2010) 49–57 � 2009 John Wiley & Sons, Ltd and Eating Disorders Association. 57

R. M. Raich et al. School-Based Universal Prevention of Eating Disorders

content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's

express written permission. However, users may print, download, or email articles for individual use.