Developing a stage of change measure for assessing recovery from anorexia nervosa

Developing a stage of change measure for assessing

recovery from anorexia nervosa

Patricia J. Jordana,*, Colleen A. Reddinga, Nicholas A. Troopb,Janet Treasurec, Lucy Serpellc

aCancer Prevention Research Center, University of Rhode Island, 2 Chafee Road, Kingston, RI 02881, USAbDepartment of Psychology, London Guildhall University, London, UK

cSouth London and Maudsley NHS Trust Eating Disorders Unit,

Institute of Psychiatry, Kings College, London, UK

Abstract

The purpose of this study was to advance understanding of the self-change process in recovery from

anorexia nervosa (AN). This included the development of a measure for assessing readiness to change

behaviors and/or cognitions associated with recovery from AN across five stages of change, based on

the transtheoretical model of behavior change. Two-hundred and seventy-eight anorexics,

predominantly female, completed questionnaires that measured all constructs of the TTM, as well

as the EAT-26, demographic items, and other self-reported recovery measures. Based on a quantitative

and qualitative comparison of five staging definitions in this relatively large sample of anorexics, it

was concluded that the most meaningful staging measure was one that measured progress through the

stages by readiness to stop restricting/bingeing/purging behaviors. The development of an algorithm to

measure stages of change for recovery from AN has the potential to accelerate clinical research and to

augment available treatments in this area.

D 2003 Elsevier Science Ltd. All rights reserved.

Keywords: Stages of change; Anorexia nervosa; Transtheoretical model; Eating disorders

1. Introduction

Extensive research in this field of eating disorders has focused on a variety of psycho-

logical, biological, and sociocultural treatment models (Compas, Haaga, Keefe, Leitenberg, &

1471-0153/03/$ – see front matter D 2003 Elsevier Science Ltd. All rights reserved.

PII: S1471 -0153 (02 )00087 -9

* Corresponding author. Tel.: +1-401-874-2830; fax: +1-401-874-5562.

E-mail address: [email protected] (P.J. Jordan).

Eating Behaviors 3 (2003) 365–385

Williams, 1998; Greben & Kaplan, 1995; Herzog, Keller, Strober, Yeh, & Pai, 1992), no one

of which has proven wholly successful in the prevention or treatment of anorexia nervosa

(AN). Several researchers have stated the need for treatment of AN that recognizes multiple

factors of causation and utilizes a range of treatment modalities (Bemis, 1978; Fisher, 1996;

Greben & Kaplan, 1995), yet no integrative approaches exist that capitalize on the modest

success rates of different specific interventions.

A substantial body of research has already developed in the treatment of various problem

behaviors, suggesting that treatment outcome is much improved when treatment is tailored to

the individual’s stage of readiness to change (CDC, 1999; Marcus et al., 1998; Prochaska et

al., 1993, 2001; Rakowski et al., 1998; Redding et al., 1999; Rossi, Weinstock, Redding,

Cottrill, & Maddock, 1997; Velicer, Prochaska, Fava, Laforge, & Rossi, 1999). It has been

previously suggested that the therapeutic process for a patient with AN may occur in stages

(Fisher, 1996), with different interventions most effectively applied at different stages

(Greben & Kaplan, 1995). Since tailoring treatments to an individual’s level of readiness

or stage of change improves treatment outcome (progress to action) across many other health

behavior changes, this may be a useful strategy to test in the treatment of eating disorders.

The transtheoretical model of behavior change (TTM; for review, see Prochaska & Velicer,

1997) explains intentional behavior change along a temporal dimension that utilizes both

cognitive and performance-based components. Based on more than two decades of research,

the TTM has found that individuals move through a series of stages—precontemplation (PC),

contemplation (C), preparation (PR), action (A), and maintenance (M)—in the adoption of

healthy behaviors or cessation of unhealthy ones (Prochaska & Velicer, 1997). The TTM uses

the stages of change to integrate cognitive and behavioral processes and principles of change,

including 10 processes of change (i.e., how one changes; Prochaska, 1979; Prochaska,

Velicer, DiClemente, & Fava, 1988), pros and cons (i.e., the benefits and costs of changing;

Janis & Mann, 1977; Prochaska, Redding, Harlow, Rossi, & Velicer, 1994; Prochaska,

Velicer, et al., 1994), and self-efficacy (i.e., confidence in one’s ability to change; Bandura,

1977; DiClemente, Prochaska, & Gibertini, 1985)—all of which have demonstrated reliability

and consistency in describing and predicting movement through the stages (Prochaska &

Velicer, 1997). Initial applications of the TTM to eating disorders have also demonstrated

encouraging results for understanding change in illnesses such as AN (Ward, Troop, Todd, &

Treasure, 1996) and bulimia nervosa (Levy, 1999; Stanton, Robert, & Zinn, 1986).

The purpose of this study was to advance understanding of the self-change process in

recovery from AN through the development of a stage of change measure for assessing

readiness to change behaviors and/or cognitions associated with recovery from AN.

1.1. The stages of change

The TTM or stages of change model has its origins in psychotherapy (Prochaska, 1979)

and was elaborated in smoking cessation research (DiClemente & Prochaska, 1982;

Prochaska & DiClemente, 1983). The stage model is best conceived as both linear and

cyclical in nature (Prochaska, DiClemente, & Norcross, 1992). Individuals are described as

progressing in a spiral fashion from PC to C, C to PR, and so on. Although rare, linear

P.J. Jordan et al. / Eating Behaviors 3 (2003) 365–385366

progression is possible; however, most individuals attempting a health-behavior change will

relapse and recycle through previous stages, gradually learning how to successfully progress

to maintenance (Prochaska et al., 1992).

PC is the stage in which an individual has no intent to change behavior in the near future,

usually measured as the next 6 months. Precontemplators are often characterized as resistant

or unmotivated and tend to avoid information, discussion, or thought with regard to the

targeted health behavior (Prochaska et al., 1992). Individuals in the C stage openly state their

intent to change within the next 6 months. They are more aware of the benefits of changing,

but remain keenly aware of the costs (Prochaska, Redding, & Evers, 1997). Contemplators

are often seen as ambivalent to change or as procrastinators (Prochaska & DiClemente, 1984).

PR is the stage in which individuals intend to take steps to change, usually within the next

month (DiClemente et al., 1991). PR is viewed as a transition rather than stable stage, with

individuals intending progress to A in the next 30 days (Grimley, Prochaska, Velicer, Blais, &

DiClemente, 1994). The A stage is one in which an individual has made overt, perceptible

lifestyle modifications for fewer than 6 months (Prochaska et al., 1997), while those in M are

working to prevent relapse and consolidate gains secured during A (Prochaska et al., 1992).

Maintainers are distinguishable from those in the A stage in that they report the highest levels

of self-efficacy and are less frequently tempted to relapse (Prochaska & DiClemente, 1984).

Current research utilizing the TTM assesses each participant’s readiness to change a

specific behavior with the use of a staging measure or algorithm. The purpose of the

algorithm is to unambiguously classify an individual into one of the stages of change.

Previous research applying the TTM to AN (Rieger et al., 1998; Ward et al., 1996) utilized

adaptations of the University of Rhode Island Change Assessment (URICA) (McConnaughy,

DiClemente, Prochaska, & Velicer, 1989; McConnaughy, Prochaska, & Velicer, 1983) as a

continuous change assessment scale. The primary purpose of the URICA is to identify

specific stage profiles characteristic of transitions between PC, C, A, and M, or to identify

subtypes of individuals within a stage (Rossi, Rossi, Velicer, & Prochaska, 1995). However,

this measure is lengthy (32 items), and is general in format—subjects respond to each item

with respect to a ‘‘problem for treatment,’’ not specific behaviors associated with AN per se—

and its use is often discouraged (Rossi et al., 1995).

Studies using staging measures created specifically for AN have thus far provided

inconsistent behavioral definitions upon which participants must base their responses (Blake,

Turnbull, & Treasure, 1997). One study in the area of exercise addressed both of these issues,

concluding that a good staging algorithm should include: (1) a complete definition of the

criterion behavior and (2) a true/false or five-choice response format (Reed, Velicer,

Prochaska, Rossi, & Marcus, 1997). The main difficulty with developing a staging algorithm

in this area is that to-date there is no agreed upon definition for recovery in AN (Herzog et al.,

1993), nor is there one that has been used consistently in the literature. Restoration and

maintenance of weight are frequently utilized indicators of improvement; however, some

argue that weight gain is not as critical as other factors, such as improvement in intrapsychic

and interpersonal conflicts, menstrual regularity, adequate sexual functioning, or normal

eating (Garfinkel, Garner, & Molodofsky, 1977). Such issues also make study comparisons

difficult.

P.J. Jordan et al. / Eating Behaviors 3 (2003) 365–385 367

The lack of consensus about what constitutes an appropriate outcome by which to assess

recovery from AN has significant implications for effective utilization of the TTM.

Quantitative assessment of the remaining TTM constructs is partially based on their

comparative levels of utilization at each stage. In order to accurately determine an

individual’s stage of recovery from AN, it is important to create a valid instrument that

can assess a person’s readiness to change one or more specific behaviors that exemplify

recovery from AN.

Studies requiring a symptom-free state for recovery may present an overly pessimistic

view of recovery from AN by not permitting the observation of less dramatic improvements

in the course of the disorder (Herzog et al., 1993; Schacter, 1982). Prochaska et al. (1992)

found that the vast majority of addicted people are not in the A stage when they enter

treatment, which may account for the high rates of recidivism in these areas. Furthermore, the

amount of progress individuals make following intervention tends to be a direct function of

their pretreatment stage of readiness for change (Prochaska et al., 1992). This finding mirrors

that found in patients with AN who often deny having a problem (Bemis Vitousek, Daly, &

Heiser, 1991) and are notoriously ambivalent about treatment (Ward et al., 1996).

This study compared five different definitional approaches to stages of change for recovery

from AN and determined an optimal staging measure, based on theoretical and empirical

criteria. This is a necessary first step in the application of the TTM to recovery from AN.

2. Method

2.1. Procedure

Participants across all five stages of change for recovery from AN or bulimia nervosa were

recruited in several ways: (1) through postings on eating disorders sites on the internet

(n= 119); (2) through an announcement in the National Eating Disorders Organization (USA)

newsletter (n = 20); (3) from treatment centers, university counseling centers and mental

health professionals throughout New England (n = 25); (4) via direct mail to a volunteer

register at the Institute of Psychiatry (IoP), composed of participants/patients from previous

studies, and those who responded to an advertisement in the magazine of the Eating Disorders

Association (UK), following a feature article on the Eating Disorders Unit (EDU; shared by

the IoP, the Maudsley and Bethlem Royal Hospitals in London) (n = 441); or (5) from patients

prior to commencing treatment at the Maudsley (n= 70).

In total, 99 of the 164 questionnaires sent via the US researchers (options 1, 2, or 3) were

returned (60%), and an additional 357 were returned from those sent by the UK researchers

(options 4 and 5) (70%). This study included only participants who reported that: (1) they

were currently suffering from AN, recovering from AN, or had recovered from AN; and/or

(2) their healthcare professional had diagnosed them with AN, now or in the past—for a total

study sample of n= 278.

Institutional approvals were obtained, and participants’ anonymous and voluntary par-

ticipation implied informed consent. Participants under the age of 18 were required to include


a signed parental consent form with their returned questionnaire. Participants were also given

the choice of one of three charitable organizations to which a donation would be made on

their behalf as an incentive to return the completed questionnaire. The set of questionnaires

took approximately 30–40 min to complete.

Due to an unforeseen clerical error, approximately 80% of the questionnaires distributed by

the IoP did not include the Eating Attitudes Test (EAT-26). All nonresponding volunteers

were mailed a reminder notice, along with a complete version of the EAT-26. Unfortunately,

very few of these were returned. As a result, the analyses specific to the EAT-26 were

conducted on a smaller proportion of the sample (n= 131).

2.2. Measures

2.2.1. Visual Analogue Recovery Scale (VARS)

A continuous visual analogue scale allowed individuals to rate their own advancement in

their recovery process. Each participant placed an ‘‘x’’ along a 7-in. line at the spot that best

represented their current recovery status. Measurements were recorded to the nearest 1/16th

inch.

2.2.2. Readiness to Change (RTC)

This eight-item scale, developed by the researchers, assessed readiness to change many

specific behaviors and attitudes associated with AN. Items were based on those used in the

EAT-26, except that respondents were asked to indicate their readiness to change on a five-

point Likert scale, ranging from not at all ready to change (1) to no longer or never was a

problem (5). Cronbach’s (1951) internal consistency for the scale was very good (a=.91) (seeJordan, 1999 for scale development).

2.2.3. Pros and cons

Development of this 10-item measure followed the sequential method of scale construction

(Comrey, 1988; Jackson, 1970, 1971) and involved a sequence of steps to ensure content and

internal validity. Respondents were asked to rate the importance of each item in their decision

to recover from AN on a five-point Likert scale, ranging from not important at all (1) to

extremely important (5). Cronbach’s (1951) internal consistency was strong for both five-item

scales (pros, a=.82; cons, a=.86) (see Jordan, 1999 for scale development and psychometrics).

2.2.4. EAT-26

This 26-item inventory was designed to measure behaviors and attitudes symptomatic of

AN (Garner, Olmsted, Bohr, & Garfinkel, 1982). Higher scores on this scale coincide with a

diagnosis of clinical AN (Garner & Garfinkel, 1979), moderate scores correspond to

subclinical AN (Button & Whitehouse, 1981), and low scores generally reflect normal

dieting behaviors (Button & Whitehouse, 1981). Coefficient a for the EAT has been reported

to be .79 for anorexic subjects and .94 for pooled samples of anorexic and normal controls

(Garner & Garfinkel, 1979). Cronbach’s (1951) coefficient a within this sample (.92) was

consistent with previous findings.


2.2.5. Stages of change (SoC)

A series of five discrete five response choice measures were used. Each measure focused

on specific criteria that were considered indicative of recovery from AN, based on current

clinical and research literature. The measures assessed a participant’s readiness to:

1. Recover from anorexia/bulimia, which was defined as ‘‘cessation of restricting, bingeing,

and/or purging behaviors associated with the eating disorder, resumption of regular

menstrual cycles, and maintenance of a healthy body weight that is average for weight and

height’’;

2. Eat normally, which was defined as ‘‘digestion of an average of three meals per day, with a

minimum daily caloric intake of 1200 cal’’;

3. Stop restricting/bingeing/purging, which defined restricting as ‘‘attempts to lose weight by

self-imposed starvation or through severe dieting,’’ bingeing as ‘‘consumption of at least

1000 cal at one time and fear that the eating was uncontrollable,’’ and purging as ‘‘any

extreme weight control activity, such as self-induced vomiting, excessive exercise, use of

cathartics (e.g., laxatives and purgatives), or diuretics’’;

4. Stop losing weight, which was defined as ‘‘the use of extreme weight control behaviors,

such as excessive dieting or exercise, which results in a body weight that is below average

for weight and height and/or negatively impacts one’s physical and emotional health’’;

and

5. Gain weight, which was defined as ‘‘steady weight gain of approximately 1 or 2 lb/month

until a body weight that is average for age and height is reached.’’

Participants were placed into one of five stages based on their single choice response to the

five alternatives presented. Precontemplators were not thinking about and not planning to

recover from anorexia/bulimia in the next 6 months. Contemplators were thinking about

trying to recover in the next 6 months. Preparers intended to start trying to recover in the next

30 days. Those in Action had been actively recovering from their eating disorder for fewer

than 6 months, while those in Maintenance had been in active recovery (or recovered) for

more than 6 months.

2.3. Analyses and evaluation criteria

A combination of quantitative and qualitative analyses was used to determine which was

the best SoC measure for recovery from AN. Algorithms were judged according to the extent

to which they met the following criteria:

1. Discriminant and concurrent validity. Prior to evaluating the algorithms, it was determined

that: (1) the pros and cons should behave inversely across the stages of change (Prochaska,

Redding, et al., 1994; Prochaska, Velicer, et al., 1994) and (2) RTC and VARS would

mirror each other across the stages of change.

2. Theoretical consistency should be maintained. First, it was assumed that RTC and VARS

would increase across the stages of change. Second, the decisional balance constructs


should behave consistently with other TTM research—the pros should go up and the cons

should come down across the stages of change (Prochaska, Redding, et al., 1994;

Prochaska, Velicer, et al., 1994). Thirdly, BMI should be highest and approximately 20 in

the M stage. Pursuant to that, it was also important that current-minus-lowest BMI (DBMI)

be significantly greater in the post-action versus pre-action stages of change. Finally, and

most importantly, EAT-26 score should be greater than 21 in PC, C, and PR, but less than

21 in M (based on Garner & Garfinkel, 1979).

3. Sensitivity of the measure. A critical component of any staging algorithm is its ability to

detect changes between stages of change. Sensitivity of each algorithm was evaluated by

the number of significant mean differences that were detected between stages of change for

the various measures.

Table 1

Correlations between all major constructs

RTC DBMI Pros Cons EAT-26 SoC01 SoC02 SoC03 SoC04 SoC05

VARS .76** .39** .22** � .31** � .63** .68** .65** .73** .58** .25**

RTC .37** .35** � .53** � .76** .59** .65** .72** .67** .44**

DBMI .05 � .11 � .35** .39** .37** .38** .29** .14 *

Pros � .21** � .19 * .18** .21** .26** .25** .29**

Cons .51** � .18** � .31** � .33** � .35** � .29**

EAT-26 � .55** � .58** � .62** � .59** � .37**

SoC01 .60** .66** .55** .25**

SoC02 .75** .60** .37**

SoC03 .66** .34**

SoC04 .48**

VARS=Visual Analogue Recovery Scale, RTC=Readiness to Change, DBMI = current minus lowest BMI,

SoC01 = readiness to recover from AN, SoC02 = readiness to eat normally, SoC03 = readiness to stop restricting/

bingeing/purging, SoC04 = readiness to stop losing weight, SoC05 = readiness to gain weight.

* P < .05.

** P < .01.

Table 2

Stage distributions across the five different staging measures

Staging algorithma Percent in stage

PC C PR A M

SoC01 (n = 277) 9.4 21.3 6.9 19.1 43.3

SoC02 (n = 277) 18.1 27.8 9.7 17.0 27.4

SoC03 (n = 276) 14.1 33.0 11.6 17.4 23.9

SoC04 (n = 273) 16.8 23.1 4.4 20.1 35.5

SoC05 (n = 274) 51.1 19.7 3.6 10.6 15.0

PC = precontemplation, C = contemplation, PR= preparation, A= action, M=maintenance, SoC01 = readiness to

recover from AN, SoC02 = readiness to eat normally, SoC03 = readiness to stop restricting/bingeing/purging,

SoC04 = readiness to stop losing weight, SoC05 = readiness to gain weight.a Slight variations in sample size are due to missing data.


4. Overall percent explained variance. The percent of explained variance by each of the

major variables used in this study, as well as the overall average effect size was considered.

5. Percent correctly classified in the prediction of stage membership. Correct classification of

stage membership, using only pros, cons, and EAT-26 score, was also reviewed.

Multivariate and univariate analyses of variance, graphs of relevant constructs by each

staging alternative, and discriminant function analyses (DFA) were conducted to inform this

choice.

Fig. 1. T-scores for major constructs across stages of change for five staging measures.


3. Results

3.1. Participants

According to self-reports, the sample was composed of 40.3% anorexics (49.1% restricter

type, 50.9% binge-purge type), 44.3% recovering anorexics, 13.2% recovered anorexics, and

2.2% other. Based on self-reported professional diagnosis, the sample included 66.3% who

Fig. 2. Raw EAT-26 scores across stages of change for the five measures.


Table 3

(M)ANOVA and Tukey follow-up results across stages of change for the five measures

Source Stage mean (S.D.) df F h2 Tukey’s HSD

PC C PR A M

SoC01—readiness for recovery from anorexia nervosaa (Wilks’ �=.35, P< .001, h2=.23)Pros 17.17 (4.90) 19.29 (4.60) 21.21 (4.22) 19.42 (4.22) 20.53 (4.03) 4271 4.04** .06 PC < PR, M

Cons 21.38 (4.57) 20.12 (4.91) 19.74 (6.19) 19.29 (4.99) 18.40 (5.18) 4272 2.42 * .03

EAT-26 51.85 (13.92) 44.67 (12.69) 44.40 (14.85) 34.30 (14.15) 25.43 (16.05) 4128 14.73*** .32 PC>A, PC, C, PR>M

BMI 15.70 (2.34) 16.44 (2.03) 17.20 (1.97) 18.34 (2.51) 19.02 (2.44) 4251 17.61*** .22 PC, C <A, PC, C, PR<M

DBMI 2.68 (1.89) 2.47 (1.61) 2.76 (1.75) 3.67 (2.17) 5.17 (3.30) 4248 12.68*** .17 PC, C, PR, A<M

RTC 16.79 (8.34) 21.85 (7.06) 25.12 (7.03) 27.68 (6.59) 32.73 (8.04) 4253 33.60*** .35 PC < PR, PC, C <A, PC, C,

PR, A <M

VARS 4.49 (3.33) 7.46 (2.91) 8.48 (1.67) 10.09 (2.92) 12.82 (3.08) 4266 61.37*** .48 PC <C, PR, PC, C <A, PC,

C, PR, A<M

SoC02—readiness for normal eatinga (Wilks’ �=.33, P< .001, h2=.24)Pros 17.60 (4.39) 19.92 (4.27) 19.78 (4.99) 20.22 (3.81) 20.86 (4.21) 4271 4.58** .06 PC <C, A, M

Cons 21.78 (3.35) 20.36 (4.30) 18.56 (5.92) 18.53 (5.27) 17.38 (5.76) 4272 7.34*** .10 PC>PR, A, PC, C>M

EAT-26 45.18 (15.22) 44.94 (12.51) 40.31 (14.70) 28.38 (13.60) 21.11 (15.17) 4128 18.15*** .36 PC, C>A, PC, C, PR>M

BMI 17.19 (2.52) 16.78 (2.23) 17.31 (2.15) 19.18 (2.86) 18.98 (2.36) 4251 11.66*** .16 PC, C, PR<A, M

DBMI 3.50 (2.08) 2.47 (1.81) 2.93 (1.63) 4.79 (3.85) 5.50 (2.79) 4248 14.52*** .19 PC <M, C, PR <A, M

RTC 19.64 (8.43) 23.07 (6.60) 26.22 (5.82) 31.64 (6.81) 35.39 (7.17) 4253 46.83*** .43 PC < PR, PC, C, PR <A, PC,

C, PR, A<M

VARS 6.58 (3.89) 8.15 (2.76) 9.35 (2.24) 11.83 (3.06) 13.51 (3.10) 4271 50.77*** .43 PC <C, PR, PR, C, PR<A,

PC, C, PR, A<M

SoC03—readiness to stop restricting/bingeing/purginga (Wilks’ �=.26, P< .001, h2=.28)Pros 16.56 (4.69) 19.84 (4.25) 19.94 (4.69) 20.44 (3.64) 21.09 (3.89) 4271 7.68*** .10 PC <C, PR, A, M

Cons 22.97 (2.37) 19.85 (4.92) 19.13 (5.54) 18.78 (4.76) 16.92 (5.62) 4271 9.84*** .13 PC>C, PR, A, PC, C>M

EAT-26 46.80 (11.03) 43.28 (13.02) 43.94 (16.21) 27.45 (11.20) 18.56 (15.18) 4128 23.34*** .42 PC, C, PR>A, M

BMI 17.32 (2.64) 16.89 (2.19) 17.49 (2.42) 18.96 (2.65) 19.16 (2.54) 4249 10.27*** .13 PC, C <A, PC, C, PR<M

P.J.

Jordanet

al./Eatin

gBehavio

rs3(2003)365–385

374

DBMI 3.24 (2.34) 2.85 (1.84) 3.21 (2.03) 4.70 (3.10) 5.66 (3.39) 4246 12.23*** .14 C<A, PC, C, PR<M

RTC 16.78 (6.07) 23.40 (6.69) 26.83 (6.37) 31.77 (6.01) 36.69 (6.48) 4253 71.22*** .53 PC <C, PR, PC, C, PR<A,

PC, C, PR, A<M

VARS 5.83 (3.86) 8.05 (2.71) 9.50 (2.29) 11.96 (2.77) 14.25 (2.44) 4270 76.08*** .53 PC <C, PR, PC, C, PR<A,

PC, C, PR, A<M

SoC04—readiness to stop losing weighta (Wilks’ �=.36, P< .001, h2=.23)Pros 17.22 (4.21) 19.59 (4.68) 20.75 (4.76) 20.20 (3.80) 20.81 (4.17) 4267 5.90*** .08 PC <C, A, M

Cons 21.93 (3.87) 20.72 (4.45) 19.67 (6.65) 18.98 (4.63) 17.16 (5.53) 4268 9.26*** .12 PC <A, PC, C>M

EAT-26 46.48 (12.88) 46.10 (13.19) 49.80 (10.90) 34.52 (13.16) 22.02 (15.31) 4125 20.72*** .40 PC, C>A, PC, C, PR, A>M

BMI 17.68 (2.50) 16.97 (2.16) 17.15 (2.71) 17.84 (3.02) 18.73 (2.51) 4247 4.54** .07 C<M

DBMI 3.38 (2.23) 2.73 (2.07) 2.83 (1.76) 3.71 (2.59) 5.16 (3.31) 4244 8.30*** .12 PC, C, A<M

RTC 19.80 (8.31) 21.28 (5.98) 23.91 (7.40) 28.54 (6.12) 35.16 (6.75) 4249 55.64*** .47 PC, C <A, PC, C, PR, A<M

VARS 6.96 (4.45) 7.91 (3.06) 8.02 (2.24) 10.49 (2.69) 12.91 (3.26) 4267 35.91*** .35 PC, C <A, PC, C, PR, A<M

SoC05—readiness to start gaining weighta (Wilks’ �=.47, P< .001, h2=.17)Pros 18.55 (4.54) 20.87 (4.09) 20.40 (4.22) 21.07 (3.79) 21.98 (3.23) 4268 7.34*** .10 PC <C, A, M

Cons 20.44 (4.63) 19.22 (5.33) 17.10 (6.99) 19.20 (3.71) 15.68 (6.00) 4269 7.65*** .10 PC, C>M

EAT-26 37.16 (17.17) 45.78 (13.21) 35.67 (14.67) 30.75 (16.53) 18.29 (12.23) 4126 9.16*** .23 PC, C>M

BMI 18.66 (2.62) 16.39 (2.14) 15.62 (1.98) 17.00 (2.23) 18.57 (2.42) 4249 11.46*** .16 PC <C, PR, A, C, PR<M

DBMI 4.12 (3.08) 2.37 (1.83) 2.63 (1.19) 3.70 (2.57) 5.59 (2.52) 4246 8.69*** .12 PC <C, PC, PR <M

RTC 25.33 (9.66) 23.84 (7.26) 27.80 (5.67) 31.54 (6.03) 36.55 (5.76) 4250 18.29*** .23 PC, C <A, PC, C, PR<M

VARS 9.82 (4.56) 8.40 (2.74) 8.69 (1.77) 10.86 (2.83) 12.94 (3.39) 4268 8.91*** .12 C<A, PC, C, PR<M

PC= precontemplation, C = contemplation, PR= preparation, A= action, M=maintenance, DBMI = current minus lowest BMI, RTC=Readiness to Change,

VARS=Visual Analogue Recovery Scale.a Variations in sample size due to missing data.

* P< .05.

** P < .01.

*** P < .001.

P.J.

Jordanet

al./Eatin

gBehavio

rs3(2003)365–385

375

had been diagnosed with AN, 13.2% had been diagnosed with an eating disorder, 11.4%

reported no current diagnosis, and 9.2% were diagnosed with bulimarexia.

The sample was primarily female (98.2%), Caucasian (92.3%), single (59%), educated

(M = 15.4 years, S.D. = 2.9), and from a middle- or upper-middle-class background (72.2%).

Participants had a mean age of 30.7 years (S.D. = 10.5, range = 13–66), older than most clinical

samples; a mean BMI of 17.9 (S.D. = 2.6); average weight and height were 106.7 lb (S.D. =

18.2) and 64.6 in. (S.D. = 3.1).

A multivariate analysis of variance (MANOVA) compared participants collected by the US

researchers (ED1, n= 65) to those collected by the UK researchers (ED2, n= 208) on 10 study

variables (Wilks’ �=.85, P < .05, h2=.08). Follow-up ANOVAs detected significant mean

differences between the groups on weight [F(1,255) = 6.63, P < .05, h2=.03 (MED1 = 111.8 lb,

S.D. = 19.9;MED2 = 105.1 lb, S.D. = 17.3)], age [F(1,275) = 7.30,P < .01, h2=.03 (MED1 = 27.5,

S.D. = 9.6; MED2 = 31.5, S.D. = 10.6)], and BMI [F(1,255) = 4.93, P< .05, h2=.02 (MED1 =

18.6, S.D. = 2.8; MED2 = 17.7, S.D. = 2.5)]. There were no significant differences detected

between the two groups on education, height, age of onset, EAT-26 score, current-minus-lowest

BMI, or on two separate RTC measures.

Correlation coefficients were calculated for each of the major constructs (Table 1). Stage

distributions of each SoC measure are presented in Table 2. Standardized scores (T-scores,

M = 50, S.D. = 10) were then calculated for pros, cons, RTC, and VARS, and plotted across

the stages of change for each of the five proposed measures (Fig. 1). Raw scores for the EAT-

26 were also plotted across the stages of change (Fig. 2).

Five separate multivariate analyses of variance (MANOVA) determined mean differences

across the stages of change (independent variable) with pros, cons, EAT-26 score, BMI,

DBMI, RTC, and VARS as the dependent variables. All MANOVAs were significant,

P < .001 (Bonferroni corrected). Follow-up ANOVAs found significant mean differences

between the stages of change on all of the variables tested, while Tukey’s HSD compared the

means of specific stage pairs on each significant variable. Overall and post-hoc results for

each stage instrument are shown in Table 3. Effect sizes across the stage measures are

presented in Table 4.

Table 4

Summary table of effect sizes (h2) of variables by stage of change across each of the five staging measures

Variable SoC01 SoC02 SoC03 SoC04 SoC05

Pros .06 .06 .10 .08 .10

Cons .03 .10 .13 .12 .10

EAT-26 score .32 .36 .42 .40 .23

BMI .22 .16 .13 .07 .16

DBMI .17 .19 .14 .12 .12

RTC .35 .43 .53 .47 .23

VARS .48 .43 .53 .35 .12

Mean h2 (S.D.) .23 (.16) .25 (.16) .28 (.20) .23 (.17) .15 (.06)

DBMI = current minus lowest BMI, RTC=Readiness to Change, VARS =Visual Analogue Recovery Scale,




Five separate DFA were performed to test the strength of the continuous TTM measures

(pros and cons) and the validation measure, EAT-26 score, in their combined ability to

differentiate and/or predict stage of change for each SoC measure. It is important to note that

conditions for these DFAs were not ideal because of the large number of cases that had to be

excluded from analysis as a result of missing EAT-26 scores from the EDU.

3.1.1. SoC01 (readiness to recover from AN)

DFA using pros, cons, and EAT-26 score as predictors for stage of change were performed

using listwise deletion (n = 133). One of three discriminant functions was significant, with a

combined c2(12) = 63.07, P < .001, Wilks’ �=.61, and accounted for about 84.3% of the

between-group variability (canonical r=.58). The predictors correctly classified 42.9% of the

total cases, and most accurately predicted group membership in PC (53.8%, n= 13) and M

(51.7%, n= 60). These variables were poor predictors of C (n = 30), where they correctly

classified only 20.0% of the membership and misclassified 56.7% into either PR or A. A

summary of correctly classified cases is shown in Fig. 3.

3.1.2. SoC02 (readiness to eat normally)

DFA using pros, cons, and EAT-26 score as continuous predictors for stage of change were

performed using listwise deletion (n = 133). One of three discriminant functions was

significant, with a combined c2(12) = 66.97, P< .001, Wilks’ �=.59, accounting for 88.4%

of the between-group variability (canonical r=.60). The predictors correctly classified 46.6%

of the total cases, and accurately predicted group membership in C (57.6%, n= 33) and M

(63.2%, n = 37). These variables were particularly poor predictors of the PR stage (n = 13),

where they correctly classified only 7.7% and misclassified 30.8% into PC and 46.2% into C

and A (see Fig. 3).

3.1.3. SoC03 (readiness to stop restricting/bingeing/purging)


performed using listwise deletion (n = 133). Two of the three discriminant functions were

significant. The first had a combined c2(12) = 85.61, P< .001, Wilks’ �=.51, while the secondhad a combined c2(6) = 13.56, P < .05, Wilks’ �=.90. The functions accounted for 87.3%

(canonical r=.66) and 10.0% (canonical r=.28) of the between-group variability, respectively.

The predictors correctly classified 46.6% of the total cases, and accurately predicted group

membership in PC (60%, n= 20), A (45.5%, n = 22), and M (68.8%, n = 32). The variables

were least accurate at predicting membership in the PR stage (12.5%, n = 16) (see Fig. 3).

3.1.4. SoC04 (readiness to stop losing weight)


performed using listwise deletion (n= 130). One of the three discriminant functions was

significant, with a combined c2(12) = 73.87, P < .001, Wilks’ �=.55, accounting for 91.4% of

the between-group variability (canonical r=.64). The predictors correctly classified 45.4% of

the total cases, and accurately predicted group membership in M (68.8%, n= 48). The three

variables were poor predictors of PR, correctly classifying only one of the five cases (see Fig. 3).


3.1.5. SoC05 (readiness to gain weight)

DFA using pros, cons, and EAT-26 score as continuous predictors for stage of change

were performed using listwise deletion (n= 131). Two of the three discriminant functions

were significant. The first had a combined c2(12) = 54.63, P < .001, Wilks’ �=.65, while thesecond had a combined c2(6) = 14.56, P< .05, Wilks’ �=.89. The functions accounted for

75.4% (canonical r=.52) and 24.0% (canonical r=.33) of the between-group variability,

respectively. The predictors correctly classified 45.0% of the total cases, and performed

Fig. 3. A comparison of stage membership correctly classified by each staging measure.


particularly well in C (52.2%, n = 23) and M (66.7%, n = 21). The three variables were

extremely poor predictors of PR, where they failed to correctly classify any of the cases

(see Fig. 3).

3.2. Choice of SoC measure

The results of these analyses established that SoC03 (restricting/bingeing/purging) was the

best measure of stage of change for recovery from AN based on the previously outlined

criteria. A summary of this evaluation is shown in Table 5.

4. Discussion

The main purpose of this study was to develop an SoC measure for assessing readiness to

change behaviors and/or cognitions associated with AN. This is an important first step to

introducing TTM measures developed specifically for AN that can be further investigated and

applied to clinical practice.

In order to select the best SoC measure, the following evaluation criteria were considered:

(1) discriminant and concurrent validity, (2) theoretical consistency, (3) sensitivity of the

measure, (4) overall percent explained variance, and (5) percent correctly classified in the

prediction of stage membership. Several analyses led to the determination that SoC03

(readiness to stop restricting/bingeing/purging) was the best overall measure. Following is

a brief summary of the considerations made for each of the five algorithms (see Section 2 for

elaboration of these criteria).

4.1. SoC01 (readiness to recover from anorexia/bulimia)

This stage distribution for SoC01 placed most participants in the M stage, despite a mean

EAT-26 score of 25.4, above the cutoff of 21, and a mean BMI of 19. In addition, pros and cons

Table 5

Summary of evaluation criteria for each of the five staging measures

Criterion SoC01 SoC02 SoC03 SoC04 SoC05

Discriminant validity no yes yes no no

Concurrent validity yes yes yes yes no

Theoretical consistency no no yes no no

Total number of mean 30 40 42 28 24

Differences across stages

Average effect size 0.23 0.25 0.28 0.23 0.15

% Correctly classified by pros, cons, EAT-26 score 42.9 46.6 45.4 44.9 45.0

DBMI = current minus lowest BMI, RTC=Readiness to Change, VARS=Visual Analogue Recovery Scale,




did not demonstrate the inverse relationship that theory would suggest, therefore, little

discriminant validity can be concluded. Finally, while it seemed sensitive to changes between

the early and M stages, it showed little differentiation between adjacent stages.

4.2. SoC02 (readiness to eat normally)

This instrument performed well in terms of concurrent validity and demonstrated some

theoretical consistency. SoC02 was sensitive to differences between the early and late stages,

but did not detect many differences between adjacent stages. This measure did, however,

detect significant increases in DBMI between the pre-action and post-action stages of change.

It performed well in terms of its overall predictive ability, but did not outperform any other

algorithm on any of the criteria for evaluation.

4.3. SoC03 (readiness to stop restricting/bingeing/purging)

This measure outperformed all others at most criterion levels. It demonstrated both

discriminant and concurrent validity. The mean EAT-26 score at M (18.6) was well below

the cutoff of 21. SoC03 also detected significant increases in DBMI between the pre-action

and post-action stages of change. Furthermore, the measure proved to be sensitive enough to

detect changes in RTC and VARS between most stages of change, including some of the

subtle changes occurring between PC and C. In addition, SoC03 had the highest average

effect size.

4.4. SoC04 (losing weight)

The number of participants who placed themselves in the PR stage was extremely small,

making it difficult to draw definitive conclusions about the utility of SoC04. It is possible that

participants did not read the behavioral definition preceding the staging items carefully

enough. Nonetheless, this instrument demonstrated good discriminant and concurrent

validity, but was not theoretically consistent considering the mean EAT-26 score (22.0)

and BMI (18.7) of those who placed themselves in M. Unfortunately, this measure

demonstrated little sensitivity in detecting mean differences between stages.

4.5. SoCO5 (gaining weight)

The large proportion of participants who placed themselves in PC with SoC05 may have

been due to a misinterpretation of its wording. This appeared to be the case given the

similarity of mean BMIs for both PC and M stages. Although the mean EAT-26 score in M

was well below the cutoff (18.3), EAT-26 score showed no meaningful decrease until after the

A stage. The measure also failed to demonstrate any real discriminant or concurrent validity

and had the smallest average effect size of the five algorithms.

The selection of SoC03 (readiness to stop restricting/bingeing/purging) is consistent with

what we know about the course of eating disorders. This measure provided the most specific


behavioral criteria, suggesting that individuals with AN view it in terms of tangible behaviors,

rather than the associated physical or affective symptoms.

The pattern of EAT-26 score and BMI across the stages of change deserves further

attention. While it would be expected that an individual in M to score lower on the EAT-26

and have a BMI in the normal range, it was not necessarily assumed that there would be a

detectable difference in such overt measures between PC and C. The finding that early stage

movement corresponded to slight decreases in EAT-26 score and BMI seemed to indicate a

complex interaction between eating disorder symptomology and readiness to change. The

individuals in this study reported a reduction in eating disordered attitudes and behaviors

(based on EAT-26), an increase in readiness to change, but a decrease in BMI. These results

have important implications for the use of weight or BMI as an outcome measure of changes

in eating disorder symptomology.

4.6. Limitations

As with all studies, there are limitations that moderate the strength of the conclusions.

First, the cross-sectional nature of the data does not allow causal inferences to be drawn from

these results. Furthermore, the lack of longitudinal data precludes our ability to add predictive

validity to the strengths of this measure.

Second, the self-report nature of the data introduces a few complicated issues. For example,

anorexics are ‘‘notoriously protective of their private experience’’ (Bemis Vitousek et al.,

1991, p. 647). AN is a highly secretive illness and participants may have had a tendency to

underreport their behavioral patterns or overreport their physical characteristics. Anorexics in

the early stages of change may also have been more likely to deny the severity of their

behavioral and cognitive symptoms. In addition, the lack of objective diagnostic validation for

many of these participants makes it difficult to determine the extent of their recovery in the A

and M stages. It is important to reiterate here that there were objective diagnoses made for

those who were at the beginning of their treatment with the Maudsley (n= 25). The majority

(90%) of these participants placed themselves in either C or PR, as would be anticipated.

Thirdly, given the complex nature of AN, it may be simplistic to conclude that a single

instrument fully captures the true nature of recovery from this illness. Each of the remaining

four SoC measures had one or more viable strengths. It may be that a selective combination of

these algorithms could provide a more subtle assessment of stage of change.

Finally, the voluntary nature of sample may have implications for the staging distribution.

For example, one might expect a voluntary sample to overrepresent those in A and M and to

underrepresent those in early stages, especially PC. Nevertheless, this sample clearly included

individuals at all stages of readiness to change that encompassed a broad spectrum of

anorexic symptomology. Furthermore, although the sample collected via the internet was

reactively recruited and the sample collected via the IoP was more proactively recruited, these

samples showed no significant differences across any of the recovery process measures.

Moreover, an independent study comparing this internet sample to that of other published

samples in the field showed no significant differences in anorexic symptomology (Jordan &

Redding, 1999).


4.7. Implications, applications, and future directions

Despite the progress made in recent years developing effective treatments for bulimia

nervosa (Keel & Mitchell, 1997; Treasure et al., 1999), treatments for AN are far behind

(Hsu, 1990; Peterson & Mitchell, 1999)—due in part to its relatively low prevalence and the

resistance of most anorexics to engage in treatment. The TTM provides a comprehensive

approach to change by incorporating temporal, motivational, and contextual variables that can

be used to both describe and drive the change process. The TTM also has the ability to serve

as a unifying model that integrates theoretically distinct cognitive and behavioral interven-

tions into a flexible framework—one that adapts to an individual’s readiness to receive

specific treatments at any particular time.

Additional work on the temporal relationship between stages is important. The stages of

change rely heavily on a 6-month time interval, primarily for the distinction between A and

M. While this time period has been shown to be appropriate for a wide variety of health

behaviors, there is currently no evidence to support this time interval for recovery from AN.

Further research should include the development of measures for all TTM constructs—

processes of change, pros and cons, confidence, temptations—to help better understand and

describe the self-change process for AN and other eating disorders. These dimensions of the

TTM provide rich information for the development of treatment plans and can quickly inform

a therapist about their client’s progress. Such measures can also serve to further enhance newer

treatment models that utilize the TTM, like cognitive analytic therapy (Treasure, Schmidt, &

Troop, 2000), coping strategies therapy (Tobin, 2000), and the readiness and motivation

interview for eating disorders (Geller & Drab, 1999).

Incorporating the TTM into eating disorders research and treatment requires a new way of

thinking about the recovery process. The TTM measures success by progress through the

stages, not only weight gain or increased calorie intake, for example. This may mean that,

while a client moves through the first three stages of change, demonstrating potentially

dramatic changes in readiness for recovery and self-understanding, there might be no visible

changes in weight or caloric intake. Traditional methods of measuring success often result in

demoralization for both the patient and the therapist when no indication of recovery is

evident. When the recovery process is viewed from a TTM standpoint, both the patient and

therapist can take satisfaction in the gains made by a client who progresses from one stage to

the next. The TTM framework also understands the relapse or recycling process as a natural

part of the successful change process, not as a type of failure. Efforts such as this one to better

describe the process of recovery from AN will enhance our treatments.

References

Bandura, A. (1977). Self-efficacy: toward a unifying theory of behavioral change. Psychological Review, 84,

191–215.

Bemis, K. M. (1978). Current approaches to the etiology and treatment of anorexia nervosa. Psychological

Bulletin, 85, 593–617.


Bemis Vitousek, K., Daly, J., & Heiser, C. (1991). Reconstructing the internal world of the eating-disordered

individuals: overcoming denial and distortion in self-report. International Journal of Eating Disorders, 10,

647–666.

Blake, W., Turnbull, S., & Treasure, T. (1997). Stages and processes of change in eating disorders: implications for

therapy. Clinical Psychology and Psychotherapy, 4, 186–191.

Button, E. J., & Whitehouse, A. (1981). Subclinical anorexia nervosa. Psychological Medicine, 11, 509–516.

Compas, B. E., Haaga, D. A. F., Keefe, F. J., Leitenberg, H., & Williams, D. A. (1998). Sampling of empirically

supported psychological treatments from health psychology: smoking, chronic pain, cancer, and bulimia

nervosa. Journal of Consulting and Clinical Psychology, 66, 89–112.

Comrey, A. L. (1988). Factor-analytic methods of scale development in personality and clinical psychology.

Journal of Consulting and Clinical Psychology, 56, 754–761.

Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297–334.

DiClemente, C. C., & Prochaska, J. O. (1982). Self-change and therapy change of smoking behavior: a compar-

ison of processes of change in cessation and maintenance. Addictive Behaviors, 7, 133–142.

DiClemente, C. C., Prochaska, J. O., Fairhurst, S. K., Velicer, W. F., Velasquez, M. M., & Rossi, J. S. (1991). The

process of smoking cessation: an analysis of precontemplation, contemplation, and preparation stages of

change. Journal of Consulting and Clinical Psychology, 59, 295–304.

DiClemente, C. C., Prochaska, J. O., & Gibertini, M. (1985). Self-efficacy and the stages of self-change in

smoking. Cognitive Therapy and Research, 9, 181–200.

Fisher, R. (1996). Anorexia nervosa and bulimia nervosa: the thin line between lifestyle and disorder. Current

Therapeutics, 37, 61–67.

Garfinkel, P. E., Garner, D. M., & Molodofsky, H. (1977). The role of behavior modification in the treatment of

anorexia nervosa. Journal of Pediatric Psychology, 2, 113–121.

Garner, D. M., & Garfinkel, P. E. (1979). The Eating Attitudes Test: an index of the symptoms of anorexia

nervosa. Psychological Medicine, 9, 273–279.

Garner, D. M., Olmsted, M. P., Bohr, Y., & Garfinkel, P. E. (1982). The Eating Attitudes Test: psychometric

features and clinical correlates. Psychological Medicine, 12, 871–878.

Geller, J., & Drab, D. L. (1999). The readiness and motivation interview: a symptom-specific measure of readiness

for change in the eating disorders. European Eating Disorders Review, 7, 259–278.

Greben, D. H., & Kaplan, A. S. (1995). Conceptual models and integrative therapy: anorexia nervosa as a

prototype. Canadian Journal of Psychiatry, 40, 584–592.

Grimley, D., Prochaska, J. O., Velicer, W. F., Blais, L. M., & DiClemente, C. C. (1994). The transtheoretical model

of change. In T. M. Brinthaupt, & R. P. Lipka (Eds.), Changing the self: philosophies, techniques and

experiences (pp. 201–227). Albany, NY: State University of New York Press.

Herzog, D. B., Keller, M. B., Strober, M., Yeh, C., & Pai, S. -Y. (1992). The current status of treatment for

anorexia nervosa and bulimia nervosa. International Journal of Eating Disorders, 12, 215–220.

Herzog, D. B., Sacks, N. R., Keller, M. B., Lavori, P. W., von Ranson, K. B., & Gray, A. B. (1993). Patterns and

predictors of recovery in anorexia nervosa and bulimia nervosa. Journal of the American Academy of Child

and Adolescent Psychiatry, 32, 835–842.

Hsu, L. K. G. (1990). Eating disorders. New York: Guilford Press.

Jackson, D. N. (1970). A sequential system for personality scale development. Current Topics in Clinical and

Community Psychology, 2, 61–96.

Jackson, D. N. (1971). The dynamics of structured personality tests. Psychological Review, 78, 229–248.

Janis, I. L., & Mann, L. (1977). Decision making: a psychological analysis of conflict, choice and commitment.

New York: Free Press.

Jordan, P. J. (1999). Development and preliminary validation of stages of change and decisional balance measures

for recovery from anorexia nervosa. Unpublished master’s thesis, University of Rhode Island, Kingston, RI.

Jordan, P. J., & Redding, C. A. (1999, March). Collecting eating disorders data through the internet: implications

for research. Annals of Behavioral Medicine, 21, S148.

Keel, P. K., & Mitchell, J. E. (1997). Outcome in bulimia nervosa. American Journal of Psychiatry, 154, 313–321.


Levy, R. K. (1999). The transtheoretical model of change: an application to bulimia nervosa. Psychotherapy, 34,

278–285.

Marcus, B. H., Bock, B. C., Pinto, B. M., Forsyth, L. H., Roberts, M. B., & Traficante, R. M. (1998). Efficacy of

an individualized, motivationally-tailored physical activity intervention. Annals of Behavioral Medicine, 20,

174–180.

McConnaughy, E. A., DiClemente, C. C., Prochaska, J. O., & Velicer, W. F. (1989). Stages of change in

psychotherapy: a follow-up report. Psychotherapy, 26, 494–503.

McConnaughy, E. A., Prochaska, J. O., & Velicer, W. F. (1983). Stages of change in psychotherapy: measurement

and sample profiles. Psychotherapy, 20, 368–375.

Peterson, C. B., & Mitchell, J. E. (1999). Psychosocial and pharmacological treatment of eating disorders: a

review of research findings. Journal of Clinical Psychology, 55, 685–697.

Prochaska, J. O. (1979). Systems of psychotherapy: a transtheoretical analysis. Homewood, IL: Dorsey Press.

Prochaska, J. O., & DiClemente, C. C. (1983). Stages and processes of self-change of smoking: toward an

integrative model of change. Journal of Consulting and Clinical Psychology, 51, 390–395.

Prochaska, J. O., & DiClemente, C. C. (1984). Self-change processes, self-efficacy and decisional balance across

five stages of smoking cessation. Advances in cancer control—1983. New York: Alan R. Liss.

Prochaska, J. O., DiClemente, C. C., & Norcross, J. (1992). In search of how people change. American Psychol-

ogist, 47, 1102–1114.

Prochaska, J.O., DiClemente, C.C., Velicer, W.F., & Ross, J.S. (1993). Standardized, individualized, interactive

and personalized self-help programs for smoking cessation. Health Psychology, 12, 399–405.

Prochaska, J. O., Redding, C. A., & Evers, K. E. (1997). The transtheoretical model and the stages of change.

In K. Glanz, F. Marcus Lewis, & B. K. Rimer (Eds.), Health behavior and health education: theory

research and practice (2nd ed.) (pp. 60–84). San Francisco, CA: Jossey-Bass.

Prochaska, J. O., Redding, C. A., Harlow, L. L., Rossi, J. S., & Velicer, W. F. (1994). The transtheoretical model

and human immunodeficiency virus prevention: a review. Health Education Quarterly, 21, 471–486.

Prochaska, J. O., & Velicer, W. F. (1997). The transtheoretical model of health behavior change. American Journal

of Health Promotion, 12, 38–48.

Prochaska, J. O., Velicer, W. F., DiClemente, C. C., & Fava, J. (1988). Measuring processes of change: applica-

tions to the cessation of smoking. Journal of Consulting and Clinical Psychology, 56, 520–528.

Prochaska, J. O., Velicer, W. F., Fava, J. L., Ruggiero, L., Laforge, R. G., & Rossi, J. S. (2001). Counselor and

stimulus control enhancements of a stage-matched system intervention for smokers in a managed care setting.

Preventive Medicine, 32, 66–72.

Prochaska, J. O., Velicer, W. F., Rossi, J. S., Goldstein, M. G., Marcus, B. H., Rakowski, W., Fiore, C., Harlow,

L. L., Redding, C. A., Rosenbloom, D., & Rossi, S. R. (1994). Stages of change and decisional balance for 12

problem behaviors. Health Psychology, 13, 39–46.

Rakowski, W., Ehrich, B., Goldstein, M. G., Rimer, B. K., Pearlman, D. N., Clark, M. A., Velicer, W. F., &

Woolverton, H. (1998). Increasing mammography among women aged 40–74 by use of a stage-matched,

tailored intervention. Preventive Medicine, 27, 748–756.

Redding, C. A., Prochaska, J. O., Pallonen, U. E., Rossi, J. S., Velicer, W. F., Rossi, S. R., Greene, G. W., Meier,

K. S., Evers, K. E., Plummer, B. A., & Maddock, J. E. (1999). Transtheoretical individualized multimedia

expert systems targeting adolescents’ health behaviors. Cognitive Behavioral Practice, 6, 144–153.

Reed, G., Velicer, W. F., Prochaska, J. O., Rossi, J. S., & Marcus, B. H. (1997). What makes a good staging

algorithm: examples from regular exercise. American Journal of Health Promotion, 12, 57–66.

Rieger, E., Touyz, S. W., Schotte, D. E., Beumont, P. J. V., Russell, J., & Griffiths, R. (1998, November).

Development of an instrument to assess motivation for recovery from anorexia nervosa. Poster presented at

the Eating Disorders Research Society Annual Meeting, Cambridge, MA.

Rossi, J. S., Rossi, S. R., Velicer, W. F., & Prochaska, J. O. (1995). Motivational readiness to control weight.

In D. B. Allison (Ed.), Handbook of assessment methods for eating behaviors and weight-related problems:

measures, theory and research (pp. 387–430). Thousand Oaks, CA: Sage.

Rossi, J. S., Weinstock, M. A., Redding, C. A., Cottrill, S. D., & Maddock, J. E. (1997, April). Effectiveness of


stage-matched interventions for skin cancer prevention: a randomized clinical trial of high-risk beach bathers.

Paper presented at the 18th Annual Meeting of the Society of Behavioral Medicine, San Francisco, CA.

Schacter, S. (1982). Recidivism and self-cure of smoking and obesity. American Psychologist, 37, 436–444.

Stanton, A. L., Robert, W. M., & Zinn, L. M. (1986). Self-change in bulimia: a preliminary study. International

Journal of Eating Disorders, 5, 917–924.

The CDC AIDS Community Demonstration Projects Research Group (1999). Community-level HIV intervention

in 5 cities: final outcome data from the CDC AIDS Community Demonstration Projects. American Journal of

Public Health, 89, 336–345.

Tobin, D. L. (2000). Coping strategies therapy for bulimia nervosa. Washington, DC: American Psychological

Association.

Treasure, J. L., Katzman, M., Schmidt, U., Troop, N., Todd, G., & De Silva, P. (1999). Engagement and outcome

in the treatment of bulimia nervosa: first phase of a sequential design comparing motivation enhancement

therapy and cognitive behavioral therapy. Behavior Research and Therapy, 37, 405–418.

Treasure, J. L., Schmidt, U. H., & Troop, N. A. (2000). Cognitive analytic therapy and the transtheoretical frame-

work. In K. J. Miller, & J. Scott Mizes (Eds.), Comparative treatment of eating disorders (pp. 283–308). New

York, NY: Springer.

Velicer, W. F., Prochaska, J. O., Fava, J. L., Laforge, R. G., & Rossi, J. S. (1999). Interactive versus non-

interactive interventions and dose–response relationships for stage matched smoking cessation programs in

a managed care setting. Health Psychology, 18, 21–28.

Ward, A., Troop, N., Todd, G., & Treasure, J. (1996). To change or not to change—‘how’ is the question? British

Journal of Medical Psychology, 69, 139–146.


Documents

Developing a stage of change measure for assessing recovery from anorexia nervosa