Universal Prevention of Eating Disorders: A Concept Analysis

Michael P. Levine

PII: S1471-0153(16)30319-1
DOI: doi:10.1016/j.eatbeh.2016.10.011
Reference: EATBEH 1127

To appear in: Eating Behaviors

Please cite this article as: Levine, M.P., Universal Prevention of Eating Disorders: A
Concept Analysis, Eating Behaviors (2016), doi:10.1016/j.eatbeh.2016.10.011

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 1

PT
Universal Prevention of Eating Disorders: A Concept Analysis

RI
Michael P. Levine*

SC
*Department of Psychology, Kenyon College, Gambier, OH, USA

NU
Correspondence should be addressed to Michael P. Levine, Ph.D., Department of
MA
Psychology, Kenyon College, Gambier, OH, USA 43022; email: Levine@kenyon.edu; phone:

740-507-9518.
D
P TE
CE
AC

Main text: 3,979 words Abstract: 165 words Tables: 1 Figures: 1

RUNNING HEAD: CONCEPT ANALYSIS OF UNIVERSAL PREVENTION

 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 2

Abstract

The definition of universal prevention is important for theoretical, research, and policy-

related reasons. The present article provides an etymological and historical look at the

PT
concept of universal prevention, in and of itself and in terms of its position on the mental

RI
health intervention spectrum involving mental health promotion, selective prevention,

SC
indicated prevention, case identification, and treatment. Following a summary of the

features commonly associated with universal prevention, these characteristics are

NU
fashioned into a family resemblance model for defining the construct. This model is applied

to four of the articles constituting the journal Eating Behaviors’ special issue on the
MA
universal prevention of eating disorders. It is argued that this family resemblance approach
D

captures the diversity of current universal approaches to reducing risk factors and/or
TE

preventing eating disorders. This type of definition, coupled with Foxcroft’s (2014)

tripartite functional analysis of universal prevention, has the potential to improve

P
CE

evaluations of universal prevention, as well as large-scale collaborative projects that seek

to integrate programs across the mental health intervention spectrum.

AC

Key words: prevention; universal; mental health intervention spectrum; eating disorders;

family resemblance definition

 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 3

Highlights

1. Universal prevention has been a complex, potentially confusing construct for over 30

years.

PT
2. Universal prevention is best understood as part of the Mental Health Intervention

RI
Spectrum.

SC
3. There are neither necessary nor sufficient conditions for defining universal prevention.

4. The family resemblance model is reasonable and practical in defining universal

NU
prevention.
MA
D
P TE
CE
AC
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 4

Universal Prevention of Eating Disorders: A Concept Analysis

It is more sensible, humane, pragmatic, and cost-effective to build

PT
psychological health and prevent maladjustment than to struggle valiantly

and compassionately to stay its awesome tide.

RI
- Emory Cowen (1983, p. 14)

SC
In the past 15 to 20 years there has been significant progress in the components of

NU
eating disorders (EDs) prevention: clarification of risk factors  design innovation 

efficacy and effectiveness research  program dissemination (Becker, Stice, Shaw, & Woda,
MA
2009; Wilksch, 2014). Nevertheless, it is still the case after well over 50 years that the

definition of prevention, the categorization of prevention philosophies and programs, and

D

the relationship, if not the distinction, between prevention and treatment can be very
TE

challenging (Caplan, 1964; Committee on the Prevention of Mental Disorders, 2009;

P

Foxcroft, 2014; Levine & Smolak, 2006). For example, as noted by Foxcroft (2014, p. 820):
CE

. . . whilst the universal-selective-indicated system for classifying prevention

AC

is a useful advance on previous notions of primary and secondary prevention,

there remains some conceptual confusion about how environmental,

community-based and individually oriented prevention approaches should

be classified and how these different types of prevention relate to the

universal-selective-indicated scheme.

Becker (2016) recently argued that the fields contributing to eating disorders

prevention will benefit greatly and reduce unproductive misunderstandings, particularly

with community stakeholders, if we strive for greater conceptual and linguistic accuracy.
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 5

“Universal prevention” is a concept (or construct) that cries out for this type of in depth

analysis. The rationale for universal prevention (the “why”) has been articulated in detail

elsewhere (Haines & Filion, 2015; Levine & Smolak, 2006, 2008), and its efficacy,

PT
effectiveness, and dissemination are the subjects of many and varied reviews (see, e.g.,

RI
Levine, 2015a; Wilksch, 2014) and of this special issue. The present article analyzes the

SC
concept of universal prevention in order to provide a working definition that captures the

breadth and complexity of this important concept.

NU
The Roots of Universal Prevention: A Selective and Targeted Review
MA
Etymology

The English word prevention (n.d.) has its roots in late medieval Latin: prae (before)
D

+ venire [to come], which when combined as a transitive verb conveyed a sense of
TE

“anticipate to hinder.” As an adjective, universal [n.d.] has Latin roots at least 200 years

older: universalis, meaning "of or belonging to all" is derived from universus, symbolizing
P
CE

"all together, whole, entire.” Even this highly simplified examination of the etymology of

universal prevention highlights some of the challenges in understanding and applying a

AC

construct that focuses on “all” or an “entirety” in order to hinder a set of disorders (e.g.,

anorexia nervosa, bulimia nervosa, binge eating disorders) that are relatively rare (Keel &

Forney, 2015).

Universal Prevention is Primary (1957 – 1983)

The etymology of universal prevention, like the declaration by Cowen (1983) that

introduces this article, reminds us that prevention is primary (Levine & Smolak, 2006,

2008). The formal distinction between primary and secondary prevention, which dates to a

1957 monograph by the Commonwealth Fund’s Commission on Chronic Illness (Mrazek &
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 6

Haggerty, 1994), was elaborated and widely disseminated through an influential book by

Caplan (1964). However, in 1983 Dr. Robert Gordon, a physician and special assistant to

the Director of the National Institutes of Health (USA), wrote a 3-page letter to the journal

PT
Public Health Reports that deftly outlined the conceptual and practical limits of the primary

RI
vs. secondary distinction (see also Cowen, 1983). Gordon (1983) argued for an improved

SC
categorization of prevention as universal or selective or indicated (these days, often

referred to as “targeted”1). It is noteworthy in a consideration of eating disorders that,

NU
according to Gordon, this typology is particularly useful in addressing disorders that are
MA
multifactorial in origin and that have a long and complex developmental trajectory.

Gordon (1983) introduces his typology by stating that:

D

Preventive measures—those which should be applied to persons not

TE

motivated by current suffering—can be operationally classified on the basis

of the population groups among which they are optimally used. The most
P
CE

generally applicable type, which we shall call universal, is a measure that is

desirable for everybody. In this category fall all those measures which can be
AC

advocated confidently for the general public and which, in many cases, can be

applied without professional advice or assistance. (p. 108; italics in the

original)

Gordon (1983) added that universal prevention is likely to require different strategies than

selective or indicated prevention, and should be undertaken only when the costs are low

and there are significant benefits for the population.

The Mental Health Intervention Spectrum

 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 7

Two influential versions of a book, published 15 years apart by representatives of

several US governmental agencies led by the National Academy of Sciences’ Institute of

Medicine (IOM), have helped to refine, but not necessarily to validate, the theoretical

PT
construct of universal prevention (Committee on the Prevention of Mental Disorders

RI
[Committee on Prevention], 2009; Mrazek & Haggerty, 1994). In the latest edition, the

SC
Committee on Prevention (2009) continued to use Gordon’s (1983) pioneering work to

define universal prevention as “preventive interventions that are targeted to the general

NU
public or a whole population group that has not been identified on the basis of individual
MA
risk. The intervention is desirable for everyone in that group” [p. xxix]. This is contrasted

with selective prevention, which focuses on “individuals or [to] a subgroup of the population
D

whose risk of developing mental, emotional, or behavioral disorders is significantly higher

TE

than average” [p. xxviii]. Determination of significant risk—but not high risk—may be done

without screening, based on research pointing to biological, psychological, or sociocultural

P
CE

variables operating at the family, community, or cultural level (Committee on Prevention,

2009).
AC

Consider a prototypical multi-lesson classroom-based eating disorders prevention

program (e.g., Happy Being Me; Dunstan, Paxton, & McLean, in press), and imagine it is to

administered to all girls ages 11 through 14 years in the Los Angeles, California, region.

This prevention project focuses on a very large (N = ~80,000, or roughly 2% of nearly 4

million people; Census Reporter, n.d) asymptomatic group who are at greater risk than

boys or younger girls because of gender roles, pubertal status, media pressures, etc.

Consequently, I would place this program between universal and selective on the IOM’s

spectrum (Levine & Smolak, 2006, 2008).

 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 8

In the more recent book the Committee on Prevention (2009) also used Mrazek and

Haggerty’s (1994) transformation of Gordon’s tripartite categorization into a continuum or

spectrum of interventions, based on the size and nature of the group for whom

PT
programming is intended. As shown in Figure 1, according to the Committee on Prevention

RI
(2009) the mental health intervention spectrum ranges from general mental health

SC
promotion  universal prevention  selective prevention  indicated prevention. The

latter overlaps with the older concept of secondary prevention (Caplan, 1964) and thus

NU
shades into the traditional steps of clinical treatment: case identification  intervention 
MA
aftercare. This spectrum maintains Gordon’s (1983) argument that (1) all people being

addressed by prevention are asymptomatic; (2) universal refers to the broadest scope in
D

the population; and (3) in contrast to indicated prevention, neither universal nor selective
TE

prevention involve screening of any sort (Committee on Prevention, 2009).

Given that its ultimate goal is healthier development, prevention subsumes

P
CE

systematic efforts to protect health. Thus, it is noteworthy that the Committee on

Prevention’s (2009) intervention spectrum juxtaposes universal prevention with mental

AC

health promotion. In other words, universal prevention is closely related to broad efforts to

promote and develop assets in the physical environment, sociocultural groupings,

individuals, and person-environment transactions, which together can foster

developmental competencies, effective coping, and other forms of resilience in the face of

inevitable stressors.

Foxcroft’s (2014) Functional Analysis

Foxcroft (2014) provides an in depth examination of the meanings of universal

prevention. Incorporating the work of Gordon (1983) and the IOM (Committee on
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 9

Prevention, 2009; Mrazek & Haggerty, 1994), Foxcroft acknowledges the utility of a

classification system in which “universal prevention takes the form of a whole population

approach, where risk of developing a disease or disorder is typically diffuse and preventive

PT
interventions are not based on level of risk” (p. 819). Foxcroft adds that universal

RI
prevention will be most relevant when Rose’s prevention paradox is in effect and when the

SC
interventions are determined to be acceptable to the population (see also Haines & Filion,

2015).

NU
The Rose paradox is a statistical phenomenon at the heart of a population-based,
MA
public health approach to prevention (Austin, 2001; Haines & Filion, 2015). To illustrate, it

is reasonable to assume that (1) the point prevalence of bulimia nervosa (BN) in females 14
D

years or older is ~2.0% (Keel & Forney, 2015); (2) there is evidence that a risk factor such
TE

as dieting or internalization of the slender beauty ideal (Austin, 2001; Becker et al., in

press; Rohde, Stice, & Marti, 2015) is normally distributed in the population; (3) those
P
CE

females who are not symptomatic but at high risk are 6 times more likely to develop BN

than those at low to moderate risk, a very generous figure for relative risk (cf. Rohde et al.,
AC

2015); and (4) 10% of the population (e.g., those with z score on the risk factor of > +1.28)

is at high risk. Given these parameters, in a hypothetical population of 1,000,000 females

ages 14 or greater, 7,800 cases (7.8% X 100,000) will emerge from the high risk-group,

while the comparable figure for the low-to-moderate-risk group is 11,700 (1.3% [i.e., 1/6

of 7.8%] X 900,000). In this instance the Rose paradox is that, although by definition the

high-risk group is at much higher risk, there are relatively very few people in this category.

Therefore, 11,700/[(.02 X 1,000,000) = 20,000] or the majority (58.5%) of the cases will
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 10

arise from the low-to-moderate risk group (Austin, 2001). This paradox indicates the need

for a significant shift downward in the population’s risk.

According to Foxcroft (2014), universal prevention will often take the form of

PT
macrolevel “environmental” changes, such as laws and regulations, or community

RI
development efforts such as mass media campaigns or changes in the educational

SC
standards for K-12, all of which apply to the entire population. However, not all planned

changes in legislation and public policy are universal in their focus. For example they may

NU
direct attention to those participating in varsity athletics at the high school level.
MA
Foxcroft (2014) then distinguishes between three functions of prevention

efforts – environmental, developmental (including positive assets in the

D

transactional sense), and informational. Briefly, universal programs with an

TE

environmental purpose are designed to reduce the opportunities to engage in

unhealthy behaviors. Universal programs with a developmental function are

P
CE

designed to promote psychosocial maturation in ways that increase resistance to

unhealthy influences. The third functional category consists of universal programs

AC

provide information in order to improve awareness, knowledge, and attitudes about

unhealthy beliefs and behaviors. As demonstrated by several articles in this special

issue (Austin, Yu, Tran, & Mayer, in press; Sánchez-Carracedo, Carretero, & Conesa,

in press), universal programs with environmental and/or informational functions

could be implemented at various levels – local, regional, national.

Universal Prevention and the Family Resemblance Model

The articles that form this special issue of Eating Behaviors demonstrate the wide

variety of approaches to which the category (or dimensional space on the Mental Health
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 11

Intervention Spectrum) of universal prevention can be applied. Based on this practice and

on the conceptual analysis above—and given that the set of “low(er) risk” girls ages 15

through 17 in Boston (USA) is at significantly higher risk than low-risk girls age 6 through 8

PT
in Boston and “low risk” girls age 15 through 17 in Kabul (Afghanistan)—it is likely

RI
impossible to specify the necessary and sufficient conditions for a categorization of

SC
universal prevention. Consequently, I believe universal prevention is a cognitive-linguistic

concept that, like “a game” or “a mental disorder,” is best understood and applied in terms

NU
of Wittgenstein’s idea of a family resemblance (Varga, 2011). In other words, there is no
MA
essential feature that all instances of universal prevention share, in contrast to, for example,

birds, all of which have feathers and are the only creatures with this characteristic. Rather,
D

as is the case with family similarities, two or more members of a family tend to have one—
TE

and often several—things in common. More precisely, defining universal prevention in

terms of family resemblance acknowledges that it is a superordinate, polythetic construct

P
CE

in which representative instances are held together by sets of overlapping similarities. This

approach is related to Lilienfeld and Marino’s (1995) application of a Roschian

AC

categorization to “mental disorder.” However, the family resemblance model does not

presume that categorization of an intervention as universal prevention is dependent on the

program’s degree of similarity to a single prototype of universal prevention (Varga, 2011).

The preceding review suggests there are nine features that determine the degree of

resemblance to the “family” concept of universal prevention (Austin, 2015; Committee on

Prevention, 2009; Foxcroft, 2014; Gordon, 1983; Haines & Filion, 2015; Levine & Smolak,

2006; Wilksch, 2014):

 Clearly not exclusively selective or indicated in nature

 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 12

 addresses, or is at least intended eventually for, extremely large groups

 focuses on risk factors, and perhaps protective factors as well, that are diffusely and

more or less normally distributed in the population

PT
 the population being addressed includes non-symptomatic people at varying

RI
degrees of risk, preferably including males

SC
 designed to transform and strengthen public institutions and policies

 involves media campaigns and/or other large group interventions designed to

NU
change normative cultural attitudes and practices
MA
 adopts an ecological approach in which risk factor research, programming, and

evaluation unfold on multiple levels, such as mass media, family, athletics, school,
D

and peers.
TE

 focuses on reducing those risk factors—and on promoting those protective

P

factors—that are likely to increase the probability of preventing multiple

CE

(sometimes comorbid) health outcomes, such as depression, eating disorders, and

the abuse of cigarettes and other substances.

AC

 designed with input from various community stakeholders to be acceptable to the

public, workable for policy makers, practical, and cost-effective in its

implementation and dissemination.

Table 1 presents the application of this family resemblance definition to four of the

studies comprising this special issue: Sánchez-Carracedo et al. (in press), from Spain;

Austin et al. (in press) from the USA; Becker et al. (in press) from the USA, England, and

Mexico; and Dunstan, Paxton, and McLean (in press) from Australia. Although it is

impractical in this conceptual analysis to consider all the studies in the special issue, these
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 13

four were selected for the following reasons. The population-based and policy-focused

collaborations described by Sánchez-Carracedo et al. (in press) and Austin et al. (in press)

are different and unique, but both are representative of a growing trend seen now in

PT
Canada, Great Britain, Australia, and Israel (see, e.g., Levine & McVey, 2015). The carefully

RI
conceived, multifaceted, multinational, and integrated dissemination projects described by

SC
Becker et al. (in press) have at their core the most successful risk factor reduction and

eating disorders prevention program to date, the Body Project, designed initially as

NU
selective and indicated prevention (see Figure 1) for late adolescent and young adult
MA
females in public school or college (Becker et al., 2009; Levine, 2015; Levine & Smolak,

2006; Stice et al., 2013). And the Happy Being Me curriculum evaluated by Dunstan et al. (in
D

press) is representative of the nature and/or the potential of many intensive prevention
TE

programs for youth ages 11 through 14 (see Levine & Smolak, 2006, 2008). Specifically,

Happy Being Me is implemented as a set of lessons for use in school classrooms, and it was
P
CE

initially designed for girls before being modified for delivery in a coeducational setting

(Dunstan et al., in press).

AC

Conclusions

No disease or disorder has ever been eliminated or even significantly reduced by

detecting and treating individual instances once they have arisen. Prevention is a necessity

and a priority for the reducing the incidence of EDs and the prevalence and intensity of risk

factors (e.g., negative body image) that are destructive in and of themselves. As noted by

Foxcroft (2014), over 20 years ago the Institute of Medicine’s book (Mrazek & Haggerty,

1994) emphasized that: “…without a system for classifying specific intervention, there is no

way to obtain accurate information on the type or extent of current activities, …and no way to
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 14

ensure that prevention researchers, practitioners, and policy makers are speaking the same

language” (p. 24; italics in the original).

There is no doubt that a family resemblance definition of the concept of universal

PT
prevention is complex. This approach resists, if not mocks, attempts to distill the essence of

RI
universal prevention into an easily digestible and reproducible summary sentence for a

SC
position paper, a press release, or a sound bite. Nevertheless, I believe it is a step in the

right direction to apply a family resemblance analysis, coupled with Foxcroft’s (2014) three

NU
categories of function, to the concept of universal prevention. Our field needs greater detail
MA
in conceptualizing and reporting the nature of universal prevention programs. This would

facilitate greater precision in evaluating claims about whether or not universal prevention
D

is effective (Levine, 2015; Levine & Smolak, 2008; Wilksch, 2014), and it would contribute
TE

to much-needed efforts to integrate universal, selective, indicated program with each other

and with health promotion and treatment (Levine & McVey, 2015).
P
CE

]
AC
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 15

References

Austin, S. B. (2001). Population-based prevention of eating disorders: An application of the

Rose prevention model. Preventive Medicine, 32, 268–283.

PT
DOI: 10.1006/pmed.2000.0797

RI
Austin, S. B., Yu, K., Tran, A., & Mayer, B. (in press). Research-to-policy translation for

SC
prevention of disordered weight and shape control behaviors: A case example

targeting dietary supplements marketed for weight loss and muscle building. Eating

NU
Behaviors.
MA
Becker, C. B. (2016). Our critics might have valid concerns: Reducing our propensity to

conflate. Eating Disorders: The Journal of Treatment & Prevention, 24, 79-89. DOI:
D

10.1080/10640266.2015.1113833
TE

Becker, C. B., Perez, M., Kilpela, L. S., Diedrichs, P. C., Trujillo, E., & Stice, E. (in press).

Engaging stakeholder communities as body image intervention partners: The Body

P
CE

Project as a case example. Eating Behaviors.

Becker, C. B., Plasencia, M., Kilpela, L. S., Briggs, M., & Stewart, T. (2014). Changing the
AC

course of comborbid eating disorders and depression: what is the role of public

health interventions in targeting shared risk factors. Journal of Eating Disorders, 2, 1-

14. Retrieved from http://www.jeatdisord.com/content/2/1/15

Becker, C. B., Stice, E., Shaw, H., & Woda, S. (2009). Use of empirically supported

interventions for psychopathology: Can the participatory approach move us beyond

the research-to-practice gap? Behaviour Research and Therapy 47, 265-274.

DOI: 10.1016/j.brat.2009.02.007

Caplan, G. (1964). Principles of preventive psychiatry. New York: Basic Books.

 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 16

Census Reporter (n.d.). Los Angeles, CA. Retrieved from

https://censusreporter.org/profiles/16000US0644000-los-angeles-ca/

Committee on the Prevention of Mental Disorders and Substance Abuse Among

PT
Children, Youth, and Young Adults [National Research Council & Institute of

RI
Medicine of the National Academies]. (2009). Preventing mental, emotional,

SC
and behavioral disorders among young people: Progress and possibilities.

Washington, DC: The National Academies Press.

NU
Cowen, E. L. (1983). Primary prevention in mental health: Past, present, and future.
MA
In R. D. Felner, L. A. Jason, J. N. Moritsugu, & S. S. Farber (Eds.), Preventive

psychology: Theory, research, and practice (pp. 11-25). New York: Pergamon.
D

Dunstan, C. J., Paxton, S. J., & McLean, S. L. (in press). An evaluation of a body image
TE

intervention in adolescent girls delivered in single-sex versus co-educational

classroom settings. Eating Behaviors.

P
CE

Foxcroft, D. R. (2014). Can prevention classification be improved by considering the

function of prevention? Prevention Science, 15, 818–822. DOI

AC

10.1007/s11121-013-0435-1

Gordon, R. S. (1983). An operational classification of disease prevention. Public Health

Reports, 98, 107-109.

Haines, J., & Filion, A. J. (2015). Why prevention? The case for upstream strategies. In L.

Smolak & M. P. Levine (Eds.), The Wiley handbook of eating disorders: Assessment,

prevention, treatment, policy, and future directions (Vol. 2, pp. 557-568). Sussex, UK:

Wiley Blackwell.

Keel, P. K., & Forney, K. J. (2015). Prevalence and incidence of eating disorders in Western
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 17

societies. In L. Smolak & M. P. Levine (Eds.), The Wiley handbook of eating disorders:

Basic concepts and foundational research (Vol. 1, pp. 53-63. Sussex, UK: Wiley

Blackwell.

PT
Levine, M. P. (2015). Does prevention work (and is this even a fair question)? Gürze-

RI
Salucore Online Catalogue. Retrieved from

SC
http://www.edcatalogue.com/does-prevention-work/

Levine, M. P. (in press). The current status of eating disorder prevention research. In T. D.

NU
Wade (Ed.), Encyclopedia of feeding and eating disorders. Sydney: Springer Science
MA
and Business Media.

Levine, M. P., & McVey, G. L. (2015). Developing an ecological approach to eating disorders
D

prevention: The Ontario Project. In L. Smolak & M. P. Levine (Eds.), The Wiley
TE

handbook of eating disorders: Assessment, prevention, treatment, policy, and future

directions (Vol. 2, pp. 639-654). Sussex, UK: John Wiley.

P
CE

Levine, M. P., & Smolak, L. (2006). The prevention of eating problems and eating disorders:

Theory, research, and practice. Mahwah, NJ: Lawrence Erlbaum Associates.

AC

Levine, M. P., & Smolak, L. (2008). “What exactly are we waiting for?” The case for

universal-selective eating disorders prevention programs. International Journal of

Child & Adolescent Health, 1, 295-304.

Lilienfeld, S. O., & Marino, L. (1995). Mental disorder as a Roschian concept: A critique of

Wakefield's "harmful dysfunction" analysis. Journal of Abnormal Psychology, 104,

411-420. DOI: 10.1037/0021-843X.104.3.411

Mrazek, P. J., & Haggerty, R. J. (Eds.). (1994). Reducing risks for mental disorders: Frontiers

for prevention intervention research (report of the Institute of Medicine’s Committee

 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 18

on Prevention of Mental Disorders). Washington, DC: National Academy Press.

Prevention. [n.d.] In Online Etymology Dictionary. Retrieved from

http://www.etymonline.com/index.php?allowed_in_frame=0&search=Prevent

PT
Rohde, P., Stice, E., & Marti, C. N. (2015). Development and predictive effects of eating

RI
disorder risk factors during adolescence: Implications for prevention efforts.

SC
International Journal of Eating Disorders, 48, 187-198. DOI: 10.1002/eat.22270

Sánchez-Carracedo, D., Carretero, C., & Conesa, A. (in press). Roundtable on the prevention

NU
of eating disorders: The Catalan Public Policy Initiative. Eating Behaviors.
MA
Stice, E., Becker, C. B., & Yokum, S. (2013). Eating disorder prevention: Current evidence-

base and future directions. International Journal of Eating Disorders, 46, 478-485.
D

DOI: 10.1002/eat.22105.
TE

Universal. (n.d.). In Online Etymology Dictionary. Retrieved from

http://www.etymonline.com/index.php?allowed_in_frame=0&search=Universal
P
CE

Varga, S. (2011). Defining mental disorder. Exploring the 'natural function' approach.

Philosophy, Ethics, and Humanities in Medicine, 6:1. DOI: 10.1186/1747-5341-6-1

AC

Wilksch, S. M. (2014). Where did universal eating disorder prevention go? Eating Disorders:

The Journal of Treatment & Prevention, 22, 184-192.

DOI: 10.1080/10640266.2013.864889
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 19

Footnotes

1For three reasons “indicated” is preferable to the commonly used “targeted”

prevention. First, “indicated” is the adjective used by Gordon (1983) and by the Committee

PT
on Prevention (2009; see Figure 1). Second, use of “indicated” avoids confusion with the

RI
more general use of “targeted” as a verb, for example, in the Committee on Prevention’s

SC
(2009) glossary definition of universal prevention (see p. X of this article) or in a statement

such as “this selective intervention targeted adolescents who were at high-risk due to

NU
participation in gymnastics and long-distance running.” Third, continuing to use “targeted”
MA
as an adjective because certain high-risk groups are the “target” of the intervention

introduces or reinforces an undesirable psychological distance between prevention experts

D

and the participants as stakeholders in the program.

P TE
CE
AC
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 20

Table 1

Application of a Family Resemblance Definition to a Sample of Studies in Eating Behaviors’

Special

PT
Issue on Universal Prevention of Eating Disorders

RI
Study in Eating Behaviors Special Issue

SC
Family Characteristic of Sánchez- Becker Austin Dunstan
Universal Prevention Carracedo et al. et al. et al. et al.

NU
Not exclusively selective
or indicated prevention Y Y Y Y
MA
Addresses extremely
large groups Y Y Y
D

Focuses on risk factor(s) with

diffuse distribution in population Y Y Y
TE

Intended for asymptomatic

P

people in mixed-risk groups Y Y Y Y

CE

Designed to transform
public institutions/policies Y Y
AC

Seeks to change normative

cultural attitudes & practices Y Y

Applies a developmental,
ecological model Y

Targets risk & protective factors

relevant to multiple disorders Y Y Y

Designed to be workable, practical,

& cost-effective for policy makers, Y Y Y
implementation, & dissemination

Note. Y = Yes, this characteristic is present in the study. A blank space means that it is
either not present, or not reported in an easily discernible manner, in the study.
 ACCEPTED MANUSCRIPT
CONCEPT ANALYSIS OF UNIVERSAL PREVENTION 21

Figure 1. The Committee on Prevention’s (2009) Mental Health Intervention Spectrum

PT
RI
SC
NU
MA
D
P TE
CE
AC