You are on page 1of 19

Body Image 32 (2020) 34–52

Contents lists available at ScienceDirect

Body Image
journal homepage: www.elsevier.com/locate/bodyimage

A systematic review on the effects of media disclaimers on young


women’s body image and mood
Sarah E. McComb ∗ , Jennifer S. Mills
Department of Psychology, York University, 4700 Keele St, Toronto, ON, M3J 1P3, Canada

a r t i c l e i n f o a b s t r a c t

Article history: The purpose of this systematic review was to determine the effectiveness of media disclaimers in
Received 23 May 2019 protecting women’s body image and mood after exposure to thin ideal media. The keywords “warn-
Received in revised form 31 October 2019 ing*” OR “disclaimer*” AND “body image” OR “body dissatisfaction” were searched in the PsycINFO and
Accepted 31 October 2019
MEDLINE/PubMed databases. Inclusion criteria included being a peer-reviewed, primary source article
available in English, which had examined the impact of media disclaimers on women’s body image.
Keywords:
Articles published prior to February 22nd , 2019 were included. In total, 15 experimental studies were
Disclaimer
included. Overall, disclaimers were ineffective at reducing women’s body dissatisfaction and negative
Warning
Body image
affect following exposure to thin ideal images, and in some cases were actually harmful to women’s
Body dissatisfaction body image. For women high in trait body dissatisfaction and thin ideal internalization, warning labels
Mood increased body dissatisfaction after exposure to thin ideal images. For women high in trait social and
Media appearance comparison specific disclaimers that outlined how the images had been altered resulted in
increased body dissatisfaction after exposure to thin ideal images. Therefore, overall, disclaimers were
ineffective at ameliorating the negative effects of exposure to thin ideal media. Future research should
examine the impact of media disclaimers on the body image of adolescents and men.
© 2019 Elsevier Ltd. All rights reserved.

1. Introduction mitter of beauty ideals (Groesz, Levine, & Murnen, 2002) through
the use of images of unrealistically thin female models.
Negative body image, particularly body dissatisfaction, is expe- Current beauty ideals emphasize thin and toned female bodies,
rienced by many people and a disproportionate number of young which over time have increasingly become thinner and more toned
women (Fallon, Harris, & Johnson, 2014; Neighbors & Sobal, 2007). in the media (Bozsik, Whisenhunt, Hudson, Bennett, & Lundgren,
Upwards of 90 % of adult women report some level of dissatisfaction 2018; Spitzer, Henderson, & Zivian, 1999; Sypeck, Gray, & Ahrens,
with their bodies (Runfola et al., 2013). The prevalence of nega- 2004). The ubiquitous depiction of the thin and fit ideal leads
tive body image and body dissatisfaction among young women is a women to believe that this body type is highly desirable and attain-
significant mental health issue, given that body dissatisfaction has able for the average woman, when in fact models represented in
been shown to precede the development of eating disorders and the media are often seriously underweight and even thinner than
is the most consistent and robust predictor of the development the weight criterion for anorexia nervosa (Rodgers, Ziff, Lowy, Yu,
of clinical eating disorders (Neumark-Sztainer, Paxton, Hannan, & Austin, 2017; Wiseman, Gray, Mosimann, & Ahrens, 1992). For
Haines, & Story, 2006; Stice, Marti, & Durant, 2011), in addition to example, a recent survey conducted with fashion models from the
being independently associated with depressed mood (Choi & Choi, 2016 New York Fashion Week found that 81 % of the models had
2016) and suicidality (Crow, Eisenberg, Story, & Neumark-Sztainer, a BMI below 18.5 and were severely underweight (Rodgers et al.,
2008). High levels of body dissatisfaction and poor body image 2017). Exposure to images of very thin models has for years been
among women are attributed to sociocultural factors, with the pointed to as a background risk factor and situational trigger for
media being the most powerful and impactful sociocultural trans- increased drive for thinness among women (Mills, Polivy, Herman,
& Tiggemann, 2002). In the past two decades, it has become rou-
tine to digitally alter photographs of models appearing in the media
to clear and lighten skin, to trim fat off body parts, and to elongate
∗ Corresponding author. body parts or overall body images (Bennett, 2008), thus making the
E-mail addresses: mccombs@yorku.ca (S.E. McComb), jsmills@yorku.ca thin ideal even thinner and less attainable for the average woman.
(J.S. Mills). According to the sociocultural model of body dissatisfaction, fre-

https://doi.org/10.1016/j.bodyim.2019.10.010
1740-1445/© 2019 Elsevier Ltd. All rights reserved.
S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52 35

quent exposure leads to the internalization of the thin ideal as effective, under what contexts, and for which types of women. This
beautiful and desirable. Social comparison to the thin ideal in the information will help to guide policy maker’s decisions about dis-
media causes women to feel dissatisfied with their own bodies claimers and will inform evidence-based body image interventions.
when they realize they cannot attain this ideal body type (Frederick,
Daniels, Bates, & Tylka, 2017; Tiggemann & Pickering, 1996). Past 2. Method
research has confirmed that exposure to thin-ideal images in mag-
azines and television has been associated with poorer body image This systematic review was conducted in accordance with
among young women (Frederick et al., 2017; Harrison & Cantor, the Preferred Reporting Items for Systematic Reviews and Meta-
1997; Naumann, Tuschen-Caffier, Voderholzer, Schafer, & Svaldi, Analyses (PRISMA) checklist (Moher, Liberati, Tetzlaff, Altman, &
2016; Tiggemann & Pickering, 1996). Group, 2009).
To address the widespread prevalence of body dissatisfaction
and its consequences, governments and policy makers around the 2.1. Eligibility criteria
world have begun to try to find cost-effective solutions that can be
easily implemented and will reach a large audience. One strategy Studies were eligible for inclusion if the following criteria were
that has been adopted is to add disclaimers to media images that met: (a) peer-reviewed article, (b) English language, (c) reported
have been digitally altered. In 2009, the Australian National Advi- results from a primary source, (d) experimental study, and (e)
sory Group on Body Image released a Voluntary Industry Code of examined the effect of a disclaimer or warning label on some type
Conduct, which outlined several strategies for reducing body dis- of media on young women’s body image. Studies were excluded if
satisfaction and negative body image caused by media images, one they did not meet all of the eligibility criteria listed above.
of which was the use of such disclaimers (Tiggemann, Slater, Bury,
Hawkins, & Firth, 2013). In 2012, Israel became the first country to 2.2. Search procedure and study selection
pass legislation that obligated the advertising industry to include
disclaimers on all photographs that had been digitally altered to The following terms were searched in the databases PsycINFO
make models appear thinner (Krawitz, 2014). In 2017, France fol- and MEDLINE/PubMed under keywords: “warning*” OR “dis-
lowed suit and made it law that all images of models that had claimer*” AND “body image” OR “body dissatisfaction.” Databases
been digitally altered to appear thinner or larger must be accompa- were searched up until February 22nd , 2019. Study screening and
nied by a disclaimer that read “photographie retouchée” (retouched selection was conducted by the first author (SM) and a gradu-
photograph; Eggert, 2017). In 2018, Canadian provincial parliament ate level research assistant. Each article’s title and keywords were
politicians also put forth a bill that would require disclaimers be dis- searched to identify eligible studies. Articles that met initial screen-
played on any advertisement that features models whose images ing criteria were then assessed at the abstract level. A full-text
have been digitally altered (Breen, 2018). examination was conducted on studies that appeared eligible for
Proponents of the use of disclaimers have assumed that dis- inclusion, or in instances where it was uncertain if a study met
claimers will be a positive strategy to prevent body dissatisfaction inclusion criteria. The reference lists of included articles were also
that results from viewing thin-ideal images. The underlying cross-referenced for other relevant articles not found in the elec-
assumption is that media disclaimers inform the consumer that the tronic literature search. Search and screening procedures can be
image, and the body within that image, are unrealistic, and there- found in Fig. 1. In total, 15 experimental studies were included in
fore are not an appropriate or realistic target of comparison for the the current review.
consumer to use. It is then assumed if the media image is perceived
to be an unrealistic target of comparison, that the consumer is less 2.3. Data extraction and risk of bias
likely to engage in social comparison to the image, and therefore
body dissatisfaction will be avoided (Charrlson, 2014; Tiggemann Data extraction was performed by the first author. The informa-
et al., 2013). While it has been assumed that media disclaimers are a tion extracted from each study was as follows: the study design;
positive strategy to protect women’s body satisfaction, it is unclear sample size; age range and mean age of sample; the sex/gender
whether there is empirical support for the idea that policies that of the sample; measures used to assess outcomes; the country
encourage the use of media disclaimers actually mitigate the neg- where participants lived; types of media images used; tools used to
ative outcomes of thin-ideal media on women’s body image and measure studied variables; a description of the experimental con-
mood. ditions in each study; and overall findings in regard to the effect of
The current study aimed to systematically review the avail- media disclaimers or warnings on young women’s body image or
able literature on media disclaimers and women’s body image mood (see Table 1). The risk of bias within each individual study
and mood. This was accomplished by examining the main effect was measured with a quality assessment that was conducted using
of disclaimers on young women’s mood and several body image the American Dietetic Association Quality Criteria Checklist (QCC).
variables, as well as examining the moderating effect of numerous This measure has been used to conduct quality assessments of body
individual difference variables. The purpose of the current review image research before (Santhira Shagar, Harris, Boddy, & Donovan,
was to determine the evidence base of media disclaimers in pro- 2018). The quality assessment was conducted independently by
tecting women’s body image and mood after exposure to thin-ideal two raters (the primary author and a graduate level research assis-
images. A systematic review was deemed to be more feasible and tant). The QCC includes 10 items that appraise the scientific validity
meaningful than a meta-analysis due to wide variability in the of each study. As per QCC scoring instructions, quality attributes
dependent variables and psychological moderators used in the pub- of each study were categorized as positive, neutral, or negative. A
lished research. This is the first paper in the literature to contribute study was given a negative overall score if six or more items on
a systematic review on the impact of media disclaimers on young the checklist were answered as ‘no.’ A study was given a positive
women’s body image and mood. Such a review is needed so that overall score if the following items were answered as ‘yes,’ plus
research about media disclaimers can be amalgamated in order to one additional item was also answered as ‘yes’: the selection of
determine whether they are an effective strategy for preventing study subjects was free from bias (Item 2), study groups were com-
body dissatisfaction caused by exposure to thin-ideal images. Fur- parable (Item 3), the study procedure was described in adequate
ther, this review is needed so that if media disclaimers are found detail (Item 6), the outcomes were clearly defined and measure-
to be effective, it is known which types of disclaimers are most ments were valid and reliable (Item 7). A neutral score was given if
36 S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52

Fig. 1. PRISMA flow-diagram of search and screening procedures of disclaimer articles.

the answers to questions 2,3,6, and 7 did not indicate that the study used photos of women from public Instagram accounts (Fardouly
was particularly strong. & Holland, 2018), one used self-taken photos of college volunteers
(Harrison & Hefner, 2014), and three used images of female swim-
suit models (Bissell, 2006; Frederick et al., 2016; Veldhuis et al.,
3. Results 2014). Of the three that used female swimsuit models, two did not
identify where the images were found from (Bissell, 2006; Veldhuis
3.1. Sample characteristics et al., 2014), and one identified the images as coming from a Google
search (Frederick et al., 2016). Nine studies used visual analogue
Sample characteristics of all studies can be found in Table 1. scales to measure body dissatisfaction or body image (Ata et al.,
All of the studies were experimental in nature. Studies were pub- 2013; Bury et al., 2016a, 2016b; Fardouly & Holland, 2018; Slater
lished between 2006-2018. Seven of the studies were conducted et al., 2012; Tiggemann & Brown, 2018; Tiggemann et al., 2013,
on samples from the United States (Ata, Thompson, & Small, 2013; 2017), three used the Multidimensional Body-Self Relations Ques-
Bissell, 2006; Fardouly & Holland, 2018; Frederick, Sandhu, Scott, tionnaire (Ata et al., 2013; Harmon & Rudd, 2016; Kwan et al., 2018),
& Akbari, 2016; Harmon & Rudd, 2016; Harrison & Hefner, 2014; two used the Eating Disorder Inventory (Bissell, 2006; Veldhuis
Kwan, Haynos, Blomquist, & Roberto, 2018), seven were conducted et al., 2014), two used the Objectified Body Consciousness Scale
on Australian samples (Bury, Tiggemann, & Slater, 2016; Bury, (Harrison & Hefner, 2014; Veldhuis et al., 2014), one used the Body
Tiggemann, & Slater, 2017; Bury, Tiggemann, & Slater, 2016; Slater, Image States Scale (Frederick et al., 2016), and one used self-made
Tiggemann, Firth, & Hawkins, 2012; Tiggemann & Brown, 2018; Likert-type questions (Harmon & Rudd, 2016).
Tiggemann et al., 2013; Tiggemann, Brown, Zaccardo, & Thomas,
2017), and one was conducted on a sample from the Netherlands 3.2. Types of disclaimers used
(Veldhuis, Konijn, & Seidell, 2014). Eleven of the studies used female
undergraduate participants (Ata et al., 2013; Bissell, 2006; Bury Altogether, six different categories of media disclaimers
et al., 2016a, 2016b, 2017; Harmon & Rudd, 2016; Kwan et al., 2018; were used in the studies presented here, which can be
Slater et al., 2012; Tiggemann & Brown, 2018; Tiggemann et al., classified as follows: generic disclaimers, specific disclaimers,
2017, 2013), two used female samples recruited through Mechani- warning/consequence disclaimers, subvertising disclaimers, infor-
cal Turk (Fardouly & Holland, 2018; Frederick et al., 2016), one used mational disclaimers, or graphic disclaimers. It is important to note
a female adolescent community sample (Veldhuis et al., 2014), and that even though the disclaimers from each study can be organized
one used an adolescent sample of both girls and boys in which par- into these categories, that each study had slight variations in how
ticipants were exposed to same-sex images (Harrison & Hefner, the disclaimers were worded for each category (see Table 3).
2014). Eight studies used magazine advertisements as the media
stimuli (Ata et al., 2013; Bury, Tiggemann, & Slater, 2014, 2016a, 3.2.1. Generic disclaimers
2016b; Harmon & Rudd, 2016; Kwan et al., 2018; Tiggemann & Altogether, 12 of the 15 studies presented here used generic
Brown, 2018; Tiggemann et al., 2013, 2017), two used fashion pho- disclaimers (see Table 3). Generic disclaimers were warnings that
tography images (Slater et al., 2012; Tiggemann et al., 2017), one simply informed participants that the image presented to them
Table 1
Summary of study characteristics and findings.

Authors Target Population Country Types of Images Outcome Measures Conditions Findings

Ata et al., 2013 342 female United States Magazine Multidimensional Body-Self Exposure to model with Exposures to models did not cause any changes
undergraduate advertisements of Relations disclaimer in body dissatisfaction, anger or depression,
students thin female models Questionnaire-Appearance either with or without a warning or disclaimer.
Evaluation Subscale (trait Body Only the car control condition experienced
dissatisfaction) changes (decreases) in body dissatisfaction.
Aged 18-37 years Visual analogue scale for state Exposure to model with Regardless of condition participants
(M = 19.58, SD = body image and mood warning experienced decreased anxiety over time; no
2.06) unique effect for disclaimers or warnings
Sociocultural Attitudes Exposure to model only No effect of disclaimers or warnings on intent
Towards Appearance (control 1) to diet post-exposure; no sig. differences
Questionnaire—Internalization- between four conditions
General
Subscale
Physical Appearance Exposure to car Higher levels of trait body dissatisfaction and
Comparison Scale advertisements (control 2) baseline internalization predicted higher
post-exposure body dissatisfaction in the
warning condition only

S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52


Social appearance comparison was not a sig.
moderator between condition and body
dissatisfaction
Bissell, 2006 124 female United States Images of female Eating Attitudes Test Exposure to swimsuit model There were no differences between the three
undergraduate swimsuit models with disclaimer conditions on bulimia, anorexia, drive for
students thinness, or body dissatisfaction scores
Aged 18-24 years Eating Disorder Inventory Exposure to swimsuit model
without disclaimer
Experimenter made Likert type Exposure to images without
questions to assess: swimsuit model or disclaimer
(control)
Feelings of similarity to models
Desire to be similar to models
Thinness and attractiveness of
each model
Bury et al., 2016a 378 female Australia Magazine Visual analogue scales for body No disclaimer, neutral No main effect of disclaimer on state
undergraduate advertisements of dissatisfaction instructions to just look at add appearance comparison or body dissatisfaction
students thin female models
Aged 18-30 years Physical Appearance No disclaimer, distractor Main effect of instructions on body
(M = 20.10, SD = Comparison Scale instructions to rate appeal of dissatisfaction, such that those in the social
2.90) advertisement comparison condition experienced more body
dissatisfaction than those in the neutral or
distractor conditions, irrespective of disclaimer
type
3 questions for state No disclaimer, social No interaction between disclaimer and
appearance comparison comparison instructions instructions for state appearance comparison
or body dissatisfaction
Generic disclaimer, neutral Trait appearance comparison did not moderate
instructions the effect of either disclaimer label or
instructions on state appearance comparison
Generic disclaimer, distractor Three-way interaction between disclaimer,
instructions instructions, and trait appearance comparison.
For women in the distractor condition who saw
specifically worded disclaimers, those high on
trait appearance comparison experienced
increased body dissatisfaction, but those low
on trait appearance comparison experienced
reduced levels of body dissatisfaction

37
38
Table 1 (Continued)

Authors Target Population Country Types of Images Outcome Measures Conditions Findings

Generic disclaimer, social


comparison instructions
Specific disclaimer, neutral
instruction
Specific disclaimer, distractor
instructions
Specific disclaimer, social
comparison instructions
Bury et al., 2016b 120 female Australia Magazine Tobii eye tracking Exposure to thin models with Participants attended to the disclaimer more in
undergraduate advertisements of no disclaimer the specific than generic disclaimer condition
students thin women
Aged 18-30 years Visual analogue scales for state Exposure to thin models with Those in the specific disclaimer condition had
(M = 20.51, SD = body dissatisfaction generic disclaimer more fixations on target body areas of the
3.03) models (arms, breasts, waist, legs) than those
in the other two conditions, and more time
spent looking at target body areas than the

S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52


generic disclaimer condition
Physical Appearance Exposure to thin models with No difference between two disclaimer
Comparison Scale specific disclaimer conditions in time until gaze first reaches
disclaimer
3 questions for state Was a higher percentage of those in the
appearance comparison specific disclaimer condition, than those in the
generic condition, who directed their gaze to
the target body areas of the models after first
looking at the disclaimer
Irrespective of disclaimer condition
participants showed a moderate increase in
body dissatisfaction after viewing ads
For those in the specific disclaimer condition
more fixations and time spent looking at target
areas on model, as well as looking at target
areas right after first looking at disclaimer, was
associated with greater increases in
post-induction body dissatisfaction. This was
not true for the generic condition.
State and trait appearance comparison both
did not moderate the relationship between
condition and body dissatisfaction
Bury et al., 2017 280 female Australia Fashion magazine Visual analogue scales for state Control pre-exposure message, Increase in body dissatisfaction over time
undergraduate advertisements of body dissatisfaction no disclaimer irrespective of whether there was a disclaimer
participants thin women or type of pre-exposure warning message; no
significant interactions so no effect of
disclaimer or message type on body
dissatisfaction
Aged 18-30 years Physical Appearance Control pre-exposure message State appearance comparison did not
(M = 20.42, SD = Comparison Scale with disclaimer moderate relationship between conditions and
2.99) body dissatisfaction
3 questions for state Digital alteration pre-exposure
appearance comparison message, no disclaimer
Four items to assess perceived Digital alteration pre-exposure
realism of models’ bodies message, disclaimer
Table 1 (Continued)

Authors Target Population Country Types of Images Outcome Measures Conditions Findings

Fardouly & 164 women United States Pre-selected Visual analogue scales for body Image of women only Those who viewed the images of attractive
Holland, 2018 recruited by photos of attractive dissatisfaction, mood, and women experienced more body dissatisfaction
Mechanical Turk women from public impressions of Instagram than control condition, but did not differ on
Instagram profiles women body dissatisfaction from each other; so
disclaimer had no effect on body dissatisfaction
Aged 18- 25 years Internalization-Thin/ Low Body Image of women + disclaimer Internalization of thin ideal did not moderate
(M = 23.09, SD = Fat subscale of Sociocultural relationship between condition and body
1.69) Attitudes Toward Appearance dissatisfaction
Questionnaire-4
Physical Appearance Travel images only (control Appearance comparison tendency moderated
Comparison Scale-Revised condition) relationship between condition and body
dissatisfaction, such that women high on
appearance comparison reported more body
dissatisfaction after exposure to image+
disclaimer and image only, relative to control
condition; no difference if disclaimer or not.
Women low in appearance comparison
reported similar body dissatisfaction
regardless of which of the three conditions

S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52


they were in.
No differences between three conditions on
changes in negative mood.
Internalization of thin ideal and appearance
comparison did not moderate relationship
between condition and negative mood
Frederick et al., 1303 women United States Slender women Drive for Thinness Scale Disclaimer condition No effect of condition on body dissatisfaction,
2016 recruited via modeling bikinis at drive for thinness, or state social comparison
Mechanical Turk the beach,
retrieved from
Google search
Aged 18-65 (M = Body Image States Scale Subvertising condition
34, SD = 11)
Physical Appearance Trait and Unaltered image condition
State Scale
State Social Comparison No image condition
measure
Harmon & Rudd, 71 female United States Fashion magazine Likert type scales for body Image only Those who viewed the disclaimer experienced
2016 undergraduate advertisements of image and mood an increase in their body satisfaction after
students thin women exposure to the images, relative to the control
group
Aged 18-24 years Rosenberg Self-Esteem Image + disclaimer Disclaimer had no effect on mood or
(M = 20, SD = 1.39) Inventory self-esteem
Body Area Satisfaction subscale
of Multi-Dimensional
Body-Self Relations
Questionnaire
Harrison & Hefner, 393 adolescents United States Photographs of Sociocultural Attitudes Unretouched image condition Regardless of gender, those in the retouched
2014 (262 females, 130 same-gender Towards Appearance aware condition experienced more objectified
males) college-aged Questionnaire body consciousness than those in the other
students taken by three conditions
the researchers and
professionally
edited
Aged 14-18 (M = Likert-type questions to asses Retouched image condition, no Regardless of gender, those in the retouched
15.46, SD = 1.32) model attractiveness awareness of retouch aware condition experienced less physical
self-esteem than those in the control or
retouched conditions

39
40
Table 1 (Continued)

Authors Target Population Country Types of Images Outcome Measures Conditions Findings

Objectified Body Retouched-aware condition Sociocultural attitudes towards appearance did


Consciousness Scale not moderate the relationships between
experimental condition and self-esteem or
body consciousness
Physical Self-Description No-exposure control condition
Questionnaire (physical
self-esteem)
Kwan et al., 2018 118 female United States Fashion magazine Total kilocalories consumed Image only Warning labels had no short-term effects on
undergraduate advertisements of negative affect, momentary body
students thin women dissatisfaction, or momentary snack
consumption.
Aged 18-25 years Eating Disorder Examination Image + warning Those exposed to the warning labels reported
(M = 20.54, SD = Questionnaire liking the images less than those who were not
2.13) exposed to a warning
Multidimensional Body-Self Restrictive eating habits moderated the
Relations Questionnaire- relationship between condition and snack

S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52


Appearance Scales consumption, such that those high on
restrictive ate fewer kilocalories in the
warning condition than the control condition
Likert type questions to assess Appearance orientation increased from
affect baseline to the 4-week follow-up for those
exposed to the warning, but did not change for
those in the control condition
Slater et al., 2012 102 female Australia Fashion shoot Visual analogue scales to assess Generic disclaimer condition Disclaimers had no effect on negative mood;
undergraduate images of thin mood and body dissatisfaction none of the conditions differed on negative
students women mood
Aged 18-35 years Internalization subscale of Specific disclaimer condition Body dissatisfaction was higher in the image
(M = 20.31, SD = Sociocultural Attitudes only condition, than in the conditions that had
3.65) Towards Appearance a disclaimer, and there was no difference in
Questionnaire body dissatisfaction between the two
disclaimer conditions
Image only Thin-ideal internalization did not moderate
the relationship between condition and body
dissatisfaction
Tiggemann & 340 female Australia Fashion magazine Visual analogue scales to Image only Label type had no effect on body
Brown, 2018 undergraduate advertisements of measure mood and body dissatisfaction; the five conditions did not sig.
students thin women dissatisfaction differ from one another in body dissatisfaction
Aged 18-30 years Perceived Realism Scale Image + generic disclaimer Those who saw a consequence or informational
(M = 20.22, SD = label perceived the images to be more realistic
2.75) than those in the no label condition
State Appearance Comparison Image + consequence label Those who saw a consequence or informational
Scale label engaged in more appearance comparison
than those in the no label condition; there was
no sig. difference between appearance
comparison in the no label, graphic label, or
disclaimer label conditions
Physical Appearance Image + informational label
Comparison Scale Revised
Image + graphic label
Tiggemann et al., 320 female Australia Fashion shoot Visual analogue scales to Product control condition Warnings had no effect on body
2017 undergraduate images of thin measure mood and body dissatisfaction; no differences between any of
students women dissatisfaction warning conditions
Table 1 (Continued)

Authors Target Population Country Types of Images Outcome Measures Conditions Findings

Aged 18-30 years Perceived Realism Scale Thin images, no warning Disclaimer labels has no sig. effect on state
(M = 20.25, SD = appearance comparison
2.61)
State Appearance Comparison Thin images, small warning Trait appearance comparison did not moderate
Scale effect of warning label on body dissatisfaction
Physical Appearance Thin images, large warning Sig. interaction between warning label and
Comparison Scale Revised internalization on body dissatisfaction, such
that internalization not related to body
dissatisfaction in the product control group,
large and very large label conditions, but for
the no label and small label conditions those
high on internalization experienced greater
body dissatisfaction
Thin-Low Body Fat subscale of Thin images, very large

S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52


Sociocultural Attitudes warning
Towards Appearance
Questionnaire
Tiggemann et al., Experiment 1: Australia Magazine Visual analogue scales for body Product image control Those who viewed thin images experienced
2013 advertisements of dissatisfaction condition more body dissatisfaction than those in the
thin women control condition, but disclaimers had no effect
on body dissatisfaction.
120 female Likert type questions for state Thin image only In experiment 1, those exposed to a disclaimer
undergraduate appearance comparison and engaged in more appearance comparison than
students perceived self-relevance those in the image only condition, with no
difference between the generic and specific
condition. In experiment 2, no sig. differences
in state appearance comparison between any
of the conditions who viewed thin images.
Aged 18-35 years Physical Appearance Thin image + generic Women higher on trait social comparison
(M = 20.90, SD= Comparison Scale disclaimer experienced increased body dissatisfaction in
5.30) the specific warning label condition.
Experiment 2: Thin image + specific Trait appearance comparison moderated
disclaimer relationship between condition and body
dissatisfaction, such that this relationship was
strongest for women in the specific disclaimer
condition. For women low on trait appearance
comparison little difference in body
dissatisfaction between the three conditions,
but for women high on trait appearance
comparison experienced most body
dissatisfaction when in the specific disclaimer
condition.
114 female
undergraduate
students
Aged 18-30 years
(M = 20.00, SD=
2.70)

41
42
Table 1 (Continued)

Authors Target Population Country Types of Images Outcome Measures Conditions Findings

Veldhuis et al., 178 adolescent Netherlands Available digital Eating Disorder Inventory- Thin model, no label Body shape of the model interacted with
2014 girls photos of media Body Dissatisfaction subscale condition, such that combining thin media
models in bikinis models with information labels resulted in less
body dissatisfaction than the thin models with
no label or a warning label (which did not
differ from one another). However, body
dissatisfaction did not differ between the label
groups for the normal weight models.
Aged 12-18 years Objectified Body Consciousness Thin model, warning label Body shape of the model, participant
(M = 14.54, SD = Scale for Preadolescent and self-esteem, and condition all interacted with
1.45) Adolescent Youth one another to predict body dissatisfaction
such that for girls with lower self-esteem in
the thin model condition who saw an
information label, their body dissatisfaction
was significantly lower than those who saw

S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52


the thin models with no label or a warning
label. For girls with high self-esteem no effect
was found for labels on thin model images.
There was no difference in body dissatisfaction
between the conditions exposed to a normal
weight model in girls with either low or high
self-esteem.
Social Comparison to Models Thin model, information label Body shape of the model, participant
and Peers Scale self-esteem, and condition all interacted with
one another to predict body objectified
consciousness, such that for girls with lower
self-esteem, using an information label on thin
media images resulted in lower levels of
objectified body consciousness, compared to
the warning label and no label (which did not
differ from one another). For normal weight
images, no differences in body consciousness
were found between the labels conditions for
girls with either low or high self-esteem.
Rosenberg’s Self-Esteem Scale Normal weight model, no label Body shape of the model, participant
self-esteem, and condition all interacted with
one another to predict body comparison, such
that for girls with low self-esteem who saw
normal weight models with an information
and warning label (no differences between the
two) body comparison was lower compared to
those who saw the images with no label. No
differences between the conditions were found
for girls with high self-esteem. No differences
between labels for girls who saw thin models,
and who had either low or high self-esteem.
Normal weight model, warning
label
Normal weight model,
information label
S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52 43

had been digitally altered or enhanced. Unlike specific disclaimers, their samples, and the majority of studies had predominantly Cau-
generic disclaimers did not specify how the images had been casian and highly educated samples. Therefore, there is a risk that
altered. samples were biased and not representative of young women in
general. All but two of the studies (Frederick et al., 2016; Harrison
3.2.2. Specific disclaimers & Hefner, 2014) had comparable experimental conditions prior to
Specific disclaimers were used in 4 of the 15 studies presented the manipulation, and therefore results can be assumed to be due
here, and explained to participants exactly how an image had to the effects of the independent variables and not pre-existing dif-
been altered. In particular, specific disclaimers explained which of ferences between groups. All of the studies blinded participants to
the model’s body parts had been lengthened, trimmed of fat, or what condition they were in, helping to reduce the likelihood of
smoothed during digital alteration. response bias (i.e., demand characteristics). All studies used valid
and reliable measures of body image. A minority of studies failed
3.2.3. Warning/consequence disclaimers to control for relevant covariates in their statistical analyses, failed
Four of the 15 studies also used a warning or consequence to indicate whether assumptions of statistical tests were met, or
disclaimer, which can be differentiated from the other types of dis- used statistical analyses even though assumptions were violated.
claimers through the inclusion of a warning about the potential It is difficult to determine whether a funding bias existed within
for harm. More specifically, these disclaimers warned participants studies, as many studies did not indicate whether the research was
about the harm or consequences that could come to their body funded.
image or health through comparing themselves to the model fea-
tured in the image. 3.4. Main effects of disclaimers on body image and mood

3.4.1. Body dissatisfaction


3.2.4. Information disclaimer
Altogether, all 15 studies examined the impact of thin-ideal
Two studies used an information disclaimer, which provided
images with disclaimers on young women’s body dissatisfaction.
weight information to the participant about the model featured in
With respect to a disclaimer versus no disclaimer comparison,
the images presented. The disclaimers either provided information
the majority of studies found that disclaimers had no impact on
to the participants that the featured model was underweight, or of
young women and girls’ level of body satisfaction. Among the seven
a normal weight.
studies that compared exposure to an image-only condition to
an image + disclaimer condition, five did not find any differences
3.2.5. Subvertising disclaimer in post-manipulation body dissatisfaction between young women
Two studies used subvertising disclaimers (Fardouly & Holland, exposed to a thin-ideal image with no disclaimer and young women
2018; Frederick et al., 2016), which are social commentary added exposed to the same thin-ideal image with a disclaimer (see Table 3
to an image with the purpose of undermining the message por- for a summary of the specific disclaimers used in each study; Bissell,
trayed by the image. In the case of thin-ideal images, the purpose 2006; Bury et al., 2017, ␩2 p < .01; Fardouly & Holland, 2018; Kwan
of the subvertising disclaimer would be to challenge the unrealistic et al., 2018; Tiggemann et al., 2017, ␩2 p < .01). However, it is difficult
nature of the image that is featured, although the two studies that to interpret whether there were any negative effects within these
utilized subvertising disclaimers differed slightly in their intent. studies for the disclaimers to actually mitigate. Three of the above
Frederick et al. (2016) used subvertising labels to achieve the fol- studies (Bissell, 2006; Fardouly & Holland, 2018; Tiggemann et al.,
lowing: (1) challenge the objectification of the women featured, (2) 2017) did not report whether there were any significant changes in
highlight that the image had been digitally manipulated, (3) high- body dissatisfaction from pre to post image exposure, and therefore
light how most women have to change their bodies to be models, it cannot be determined if the thin-ideal images caused any changes
and (4) attempt to make the model a less desirable target for com- in body dissatisfaction that the disclaimers might have had a chance
parison. Alternatively, the captions used by Fardouly and Holland of mitigating. One study conducted by Bury et al. (2017) did find
(2018) seemed to be intended to: (1) challenge the realness of the that body dissatisfaction increased from pre to post image exposure
model’s appearance and lifestyle on social media, (2) highlight how in both the image-only and the disclaimer condition, which sug-
much effort was expended on changing or improving the model’s gests that in this case the researchers can be sure that the disclaimer
appearance for a photo, and (3) ultimately making the model a less failed to mitigate the increases in body dissatisfaction since those
desirable target for comparison by highlighting how uninspiring in the disclaimer condition experienced similar amounts of body
the photos were. It is important to note that while Fardouly and dissatisfaction as those in the control condition. Conversely, Kwan
Holland (2018) did not specifically classify the disclaimers they et al. (2018) did not find any pre to post changes in momentary body
used in their study as subvertising labels, we believe that the dis- dissatisfaction in either the image-only or disclaimer condition.
claimers used in that study fit the definition of a subvertising label Since no changes in body dissatisfaction occurred in the control
better than any other type of disclaimer, and have therefore classi- (image-only) condition, it seems there was no negative effect for
fied the disclaimers as such. the disclaimer to mitigate, which likely explains why disclaimers
were found to be ineffective in the study.
3.2.6. Graphic disclaimer With respect to a generic versus specific disclaimer, no signifi-
Only Tiggemann and Brown (2018) used the graphic disclaimer. cant differences were found between young women exposed to a
This disclaimer was not a verbal disclaimer like the other types of thin-ideal image with no disclaimer, those exposed to a thin-ideal
disclaimers, but instead featured an image of a paintbrush with the image with a generic disclaimer, or those exposed to a thin-ideal
word “retouched” underneath. image with a specific disclaimer in four out of five studies (Bury
et al., 2016a, ␩2 p = .01, 2016b, ␩2 p = .01; Tiggemann & Brown, 2018,
3.3. Risk of bias within studies ␩2 p < .01; Tiggemann et al., 2013, d = 0.04-0.06). Two of these stud-
ies (Bury et al., 2016a; Tiggemann et al., 2013) did not report pre
Results of the quality assessment for each study can be found to post changes in body dissatisfaction, and therefore it cannot be
in Table 2. Overall, 13 of the 15 studies were found to be of high determined if there were any changes in body dissatisfaction to
quality, and two were rated to be of neutral quality. Several of actually mitigate. However, Bury et al. (2016b) and Tiggemann and
the studies failed to include inclusion and exclusion criteria for Brown (2018) did find that body dissatisfaction increased from pre
44
Table 2
Quality assessment of reviewed studies using the Quality Criteria Checklist (QCC).

Study Clear research Sample Comparable Participant Use of Clear Clearly defined, Appropriate Conclusions Unlikely Overall quality
question selection free study groups withdrawals or blinding description of valid, reliable statistical supported by funding bias rating
from bias response rate data collection outcomes analysis results
described procedures

S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52


Ata et al., 2013 + + + N/A + + + + + Unclear +
Bissell, 2006 + ∅ + N/A + + + ∅ ∅ Unclear +
Bury et al., 2016a + ∅ + N/A + + + + + Unclear +
Bury et al., 2016b + ∅ + N/A + + + + + + +
Bury et al., 2017 + ∅ + N/A + + + + + + +
Fardouly & + ∅ + N/A + + + + + + +
Holland, 2018
Frederick et al., + + Unclear N/A + ∅ + ∅ + Unclear ∅
2016
Harmon & Rudd, – ∅ + N/A + + + ∅ + Unclear ∅
2016
Harrison & Hefner, + + Unclear + + + + + + + +
2014
Kwan et al., 2018 + + + + + + + + + + +
Slater et al., 2012 + ∅ + N/A + + + + + Unclear +
Tiggemann & + ∅ + N/A + + + + + + +
Brown, 2018
Tiggemann et al., + ∅ + N/A + + + + + + +
2017
Tiggemann et al., + ∅ + N/A + + + + + Unclear +
2013
Veldhuis et al., + + ∅ N/A + + + + + Unclear +
2014
S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52 45

Table 3
Disclaimers used from each study organized by type of disclaimer.

Type of Disclaimer Authors that Used This Type of How Disclaimer Was Worded
Disclaimer

Ata et al., 2013 “Retouched photograph aimed at changing a person’s physical appearance.”
Bissell, 2006 “The image below has been digitally manipulated to enhance the model’s appearance.”
Bury et al., 2016a “Warning: This image has been digitally altered”
Bury et al., 2016b “Warning: This image has been digitally altered”
Bury et al., 2017 “Warning: This image has been digitally altered”
Frederick et al., 2016 “Warning: This photo has been PHOTOSHOPPED”.
Generic
Harmon & Rudd, 2016 “Warning: These images have been digitally enhanced.”
Harrison & Hefner, 2014 “After students were photographed, a professional photo retoucher refined the images
using a computer photo retouching program.”
Slater et al., 2012 “Warning: These images have been digitally altered”
Tiggemann & Brown, 2018 “Note: This image has been altered to enhance appearance”
Tiggemann et al., 2017 “Warning: These images have been digitally altered”
Tiggemann et al., 2013 “Warning: This image has been digitally altered”
Bury et al., 2016a “Warning: This image has been digitally altered to lengthen and thin legs”
Bury et al., 2016b “Warning: This image has been digitally altered to trim arms and waist”
Specific
Slater et al., 2012 “Warning: these images have been digitally altered to lengthen legs and trim inner
thighs.”
Tiggemann et al., 2013 “Warning: This image has been digitally altered to smooth skin tone and slim arms
and legs”.
Ata et al., 2013 “Warning: Trying to look as thin as this model may be dangerous to your health.”
Kwan et al., 2018 “Warning: This photograph has been altered in a manner that could promote
Warning/Consequence
unrealistic expectations of appropriate body image.”
Tiggemann & Brown, 2018 “Note: Viewing thin and unrealistic images of women can make you feel bad about
yourself”
Veldhuis et al., 2014 Thin condition: “These models are underweight. Unconsciously, exposure to media
models may negatively impact your self- image.”
Normal weight condition: “These models have a normal weight. Unconsciously,
exposure to media models may negatively impact your self-image.”
Tiggemann & Brown, 2018 “Note: This model is underweight”
Information
Veldhuis et al., 2014 Thin condition: “These models are underweight”
Normal weight condition: ‘These models have a normal weight”
Fardouly & Holland, 2018 E.g.“I had acne here, this is a lot of makeup. I was smiling because I thought I looked
Subvertising
good. Happiness based on aesthetics will suffocate your potential here on earth”
E.g. “Not real life—I didn’t pay for this outfit, took countless photos trying to look hot
for Instagram.”
Frederick et al., 2016 E.g. “Why don’t you show that she is a person with a face and personality instead of
presenting her as a sexualized body part?”
E.g. “Photoshop made me ripped”
Graphic Tiggemann & Brown, 2018 Image of a paint brush with word “Retouched” underneath

to post exposure in both the control (image-only) and disclaimer yourself.”), a graphic disclaimer (e.g., An image of a paintbrush
conditions, which suggests that the researchers can be confident with the word “Retouched” underneath), or a subvertising dis-
that the disclaimers actually failed to mitigate increases in body claimer (e.g., Photoshop made me ripped.”) (Ata et al., 2013; Frederick
dissatisfaction. et al., 2016, ␩2 p = .004; Tiggemann & Brown, 2018). However,
In the one study that did find that disclaimers mitigated results were mixed regarding the differential effects of informa-
increased body dissatisfaction relative to the no label condition, tional disclaimers on women’s body dissatisfaction after viewing
the effect size was moderate (␩2 p = .10, Slater et al., 2012), but sim- thin-ideal images. Tiggemann and Brown (2018) found that body
ilar to other studies, this study showed that specific disclaimers dissatisfaction did not differ between undergraduate women who
were no more effective than generic disclaimers (␩2 p = .002) in were exposed to thin-ideal images with an information disclaimer,
mitigating body dissatisfaction (Slater et al., 2012). Specifically, consequence disclaimer, generic disclaimer, specific disclaimer,
the researchers found that exposure to thin-ideal images caused graphic disclaimer, or no disclaimer at all. However, Veldhuis et al.
increased body dissatisfaction from pre to post for the image-only (2014) found that adolescent girls who were exposed to thin-ideal
condition, but they did not find any significant changes in pre to images with an information disclaimer had less body dissatisfaction
post scores for either the generic or specific disclaimer conditions. than those exposed to thin-ideal images with a warning disclaimer
In other words, the results indicated that both types of disclaimers or no disclaimer at all (␩2 p = .06), and this represented a medium
were effective at preventing an increase in body dissatisfaction effect. The results from Veldhuis et al. (2014) suggests that in this
caused by exposure to the thin-ideal images. Consequently, there case informational disclaimers were effective at mitigating body
appears to be no added benefit to women’s body image from includ- dissatisfaction after exposure to thin-ideal images for adolescent
ing in a disclaimer specifically how an image was altered. girls. Only two other studies found disclaimers to be effective in
With respect to other types of disclaimer labels, no significant preventing negative body image among young women, and these
differences in body dissatisfaction were found between women represented medium effects. More specifically, it was found that
who viewed thin-ideal images with no disclaimer compared to young women who were exposed to thin-ideal images with a dis-
women who viewed thin-ideal images with any type of disclaimer claimer reported less body dissatisfaction than those exposed to the
label, including a warning disclaimer (e.g., “Warning: Trying to same image without a disclaimer (Harmon & Rudd, 2016, ␩2 p = . 06;
look as thin as this model may be dangerous to your health”), a Slater et al., 2012, ␩2 p = . 10).
generic disclaimer (e.g., “Note. This image has been digitally altered to Conversely, Harrison and Hefner (2014) found that regardless
enhance appearance”), a consequence disclaimer (e.g., Note. Viewing of gender, adolescents who were exposed to retouched images of
thin and unrealistic images of women can make you feel bad about college students labelled with a disclaimer stating that the images
46 S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52

had been professionally retouched reported greater body dissatis- or a specific disclaimer, and two studies finding that exposure
faction (␩2 p = .06) and lower physical self-esteem (␩2 p = .05) than to certain disclaimers increased state appearance comparison. All
those who saw the untouched photos, those who saw the retouched studies used the State Appearance Comparison Scale (Tiggemann &
photos without a disclaimer, and those who saw no image at all; McGill, 2004) to measure the degree of state appearance compari-
these represented small-medium effects. This is the only study to son made to the models (1 = no comparison to 7 = a lot of comparison).
date that found that disclaimers could heighten body dissatisfaction. When comparing the effect of specific versus generic disclaimers to
Overall, 11 studies found that disclaimers were ineffective at no disclaimers, Bury et al. (2016a) found no significant difference
mitigating body dissatisfaction after exposure to thin-ideal images in the degree of state appearance comparison between the three
(relative to no disclaimer), three studies found medium effects that conditions (␩2 p = .01). Additionally, Tiggemann et al. (2017) found
disclaimers were effective at mitigating increased body dissatisfac- no significant differences in the degree of state appearance com-
tion, and one study found a small-medium effect that disclaimers parison between women exposed to an image with no disclaimer,
actually heightened body dissatisfaction. and women exposed to an image with either a small, large, or very
large disclaimer (␩2 p = .01). Similarly, in the first experiment of a
3.4.2. Eating behaviour set of studies, Tiggemann et al. (2013) also found no significant dif-
Eating behaviour in response to thin-ideal disclaimers has been ferences in the degree of state appearance comparison between
measured in three studies using intention-to-diet scales, measures women who were aged 18–35 years and who were exposed to
of eating pathology, and in vivo food consumption in the lab. In an image with no disclaimer versus women exposed to an image
two of three studies, it was found that viewing thin-ideal images with either a specific or generic disclaimer (d = 0.08). In this first
with disclaimers had no significant effect on young women’s eating experiment by Tiggemann et al. (2013) participants browsed the
compared to viewing an image alone. Women who viewed thin- thin-ideal images at their own pace, with no time limit on how
ideal images with a disclaimer did not differ from those who saw long they could look at the images. However, in a second experi-
the images with no disclaimer on intention to diet (␩2 p = . 009, ment with some procedural modifications, Tiggemann et al. (2013)
Ata et al., 2013) or on anorexic or bulimic behaviours, as mea- found a small-medium effect, whereby women that were exposed
sured by the Eating Attitudes Test and Eating Disorder Inventory to either a specific or generic disclaimer were more likely to engage
(Bissell, 2006). However, in regard to eating behaviour, it was found in a greater degree of appearance comparison than those exposed
that those who were high on restrictive eating habits consumed to the image with no disclaimer (d = 0.46), with no significant dif-
less kilocalories after exposure to a warning disclamer, relative ference between the generic and specific disclaimer conditions
to those who were low on restrictive eating habits, and to those (d = 0.03). In this second experiment, Tiggemann et al. (2013) used
who were exposed to the thin-ideal image alone (Kwan et al., a smaller age range for their sample (18–30 years instead of 18–35
2018). These results indicate that exposure to warning disclaimers years) and also limited exposure time to each thin-ideal image
triggered more restrictive eating among those who were already to 45 s; it is possible that these procedural modifications account
likely to exhibit restrictive eating habits. Therefore, young women’s for the mixed findings between these two experiments. Similarly,
intentions to diet and their self-reported anorexic and bulimic while Tiggemann and Brown (2018) found that there were no sig-
behaviours were not differentially impacted by exposure to thin- nificant differences in the degree of state appearance comparison
ideal images with a disclaimer, relative to seeing the image alone, between those exposed to an image with no disclaimer, a generic
but their actual in-vivo eating behaviour was negatively impacted disclaimer (␩2 p = .01), or a graphic disclaimer (␩2 p < .01), they did
by exposure to a disclaimer if they were already restrictive eaters. find that women exposed to a consequence disclaimer (␩2 p = .02) or
an informational disclaimer (␩2 p = .03) engaged in a greater degree
3.4.3. Drive for thinness of state appearance comparison as compared to the image only
Two studies found that drive for thinness did not significantly control group; however, these were very small effects. Therefore,
differ between women after viewing images of thin swimsuit mod- it remains unclear what effect media disclaimers have on state
els either with or without a disclaimer (Bissell, 2006; Frederick appearance comparison.
et al., 2016, ␩2 p = .002). Therefore, when viewing images of swim-
suit models specifically, disclaimers are neither protective against 3.4.6. Mood
nor harmful to women’s desire to lose weight. Results consistently reflected that the effect on state mood was
no different for the disclaimer groups than for the image-only
3.4.4. Appearance orientation groups. All five studies that examined the effect of disclaimers on
Only one study has examined the impact of disclaimers on mood found that there were no significant differences between the
women’s appearance orientation, which reflects the relative impor- image-only conditions and image + disclaimer conditions on nega-
tance and attention paid to physical appearance. Kwan et al. (2018) tive mood (Fardouly & Holland, 2018; Harmon & Rudd, 2016; Kwan
found a small effect, whereby women’s ratings of the importance et al., 2018; Slater et al., 2012, ␩2 p = .05) or anxiety (Ata et al., 2013).
of and attention paid to their physical appearance significantly Ata et al. (2013) and Kwan et al. (2018) both found that negative
increased over the course of four weeks when they had previously affect decreased from pre to post exposure in both the image-only
been exposed to thin-ideal images with a disclaimer (d = 0.25), and disclaimer conditions, with no differences between the con-
an effect which was not found in the image-only (no disclaimer) ditions. These results allow the researchers to be confident that
condition (d = 0.09). Therefore, preliminary research suggests that disclaimers had no greater impact on reducing negative affect than
disclaimers on thin-ideal images make women more likely to place did just seeing the images. However, Slater et al. (2012) found that
greater importance on their appearance and focus more attention the images had very little effect on negative affect pre to post expo-
on their appearance compared to viewing thin-ideal images with- sure in either the image-only or disclaimer conditions; participants’
out a disclaimer. levels of negative affect remained virtually the same as baseline lev-
els after seeing the thin-ideal images in all conditions. Since there
3.4.5. State appearance comparison were no changes in negative affect in the control (image-only) con-
Results regarding the effect of disclaimers on state appearance dition it cannot be concluded that the disclaimers were ineffective
comparison are mixed, with three studies finding no differences in at mitigating increased negative affect.
the degree of state appearance comparison between those exposed In summary, the majority of experimental research on the
to a thin-ideal image with either no disclaimer, a generic disclaimer, effects of various types of disclaimers suggests that they are nei-
S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52 47

ther protective nor harmful in terms of body image, mood, or of these studies, it remains unclear whether thin-ideal internaliza-
eating. They may in fact be harmful in terms of appearance ori- tion moderates the effect of disclaimers on body dissatisfaction and
entation and state appearance comparison. Visual inspection and for what type of disclaimer.
comparison of the variability between experimental versus control
groups in the studies reviewed above revealed equivalent indices 3.5.3. Trait appearance comparison
of variance. Nevertheless, research targeting potential individual Seven studies examined trait appearance comparison tendency
difference moderators has been conducted and is reviewed next. as a potential moderator of the effects of disclaimers. All seven
studies used a version of the Physical Appearance Comparison
3.5. Individual difference moderators Scale (Schaefer & Thompson, 2014; Thompson, Heinberg, & Tantleff,
1991) to measure how often (1 = never to 5 = always) participants
3.5.1. Trait body dissatisfaction generally compared their appearance to others in different social
One study found that levels of trait body dissatisfaction mod- contexts. Four studies found that trait appearance comparison
erated the relationship between experimental condition and body tendency did not moderate the relationship between disclaimer
dissatisfaction, such that those who were higher in trait body dis- condition and body dissatisfaction (Ata et al., 2013; Bury et al.,
satisfaction and who were exposed to a warning label reported 2016b, R2 change < .01; Bury et al., 2017, R2 change < .001; Tiggemann
higher post-exposure body dissatisfaction than the other condi- et al., 2017, R2 change = .01). However, three studies did find that
tions (␤ = −.06), whereas those low in trait body dissatisfaction in trait appearance comparison tendency moderated the relation-
the warning condition did not differ in post-exposure body dis- ship between disclaimer condition and body dissatisfaction (Bury
satisfaction from the other conditions (Ata et al., 2013). Therefore, et al., 2016a; Fardouly & Holland, 2018; Tiggemann et al., 2013).
women high in trait body dissatisfaction are at risk for experiencing Specifically, Fardouly and Holland (2018) found that for women
even greater body dissatisfaction following exposure to thin-ideal who were low on trait appearance comparison, there were no
images with warnings about the dangers of trying to look like the differences in body dissatisfaction between women exposed to a
models, while women low in trait body dissatisfaction are not at an thin-ideal image alone, a thin-ideal image + disclaimer, or travel
increased risk for greater state body dissatisfaction. control images. However, it was found that women high on trait
appearance comparison reported more body dissatisfaction after
3.5.2. Thin-ideal internalization exposure to an image + disclaimer or image only, relative to the con-
Results regarding the role of thin-ideal internalization as a trol condition, yet there were no significant differences between the
moderator of the relationship between experimental disclaimer image + disclaimer and image-only conditions. This suggests that
condition and body dissatisfaction are mixed; three studies found mere exposure to a thin-ideal image causes increased body dis-
that thin-ideal internalization did not significantly moderate the satisfaction for women high on trait appearance comparison and
relationship between disclaimer condition and body dissatisfac- that the presence of the disclaimer neither mitigates nor causes
tion, while two studies did find it to be a significant moderator. any greater body dissatisfaction than exposure to the thin-ideal
Both Slater et al. (2012) and Fardouly and Holland (2018) found image alone. Exposure to thin-ideal images had no greater effect
that thin-ideal internalization did not moderate the relationship on body dissatisfaction than control travel images for women low
between disclaimer condition and body dissatisfaction. Similarly, on trait appearance comparison. However, Tiggemann et al. (2013)
Harrison and Hefner (2014) investigated whether sociocultural found that for women high on appearance comparison, the great-
attitudes towards appearance moderated the relationship between est amount of body dissatisfaction was experienced when they
disclaimer condition and body consciousness. In this study, socio- were exposed to a specific disclaimer (B = 1.90) relative to women
cultural attitudes towards appearance consisted of attitudes that exposed to a generic disclaimer or no disclaimer. If women were
reflected not only internalization of the thin ideal, but also accep- low on appearance comparison, then no significant differences
tance of the media as a source of information about appearance and were found between the conditions on body dissatisfaction. Sim-
perceived pressure to conform to media ideals. Similar to the results ilarly, Bury et al. (2016a) found that women who were high on
found by Slater et al. (2012) and Fardouly and Holland (2018), socio- trait appearance comparison and who were instructed to rate the
cultural attitudes towards appearance were not found to moderate appeal of thin-ideal advertisements that had specific disclaimers
the relationship between disclaimer condition and body conscious- experienced increased body dissatisfaction (␤ = .20), while those
ness (Harrison & Hefner, 2014). low on trait appearance comparison experienced reduced body
However, Ata et al. (2013) found a small effect, whereby higher dissatisfaction in this condition. Therefore, it remains unclear if
baseline internalization predicted greater post-exposure body disclaimers that specify how and which body parts were altered
dissatisfaction in the warning label condition (␤ = .06), whereas put women high in trait appearance comparison at greater risk for
women low in internalization did not experience higher body increased body dissatisfaction, since two studies (Bury et al., 2016a;
dissatisfaction after exposure to the image + warning label. These Tiggemann et al., 2013) have found support for this relationship,
results by Ata et al. (2013) suggest that warning disclaimers may and one has not (Bury et al., 2016b), despite the fact that all three
put women who are high in thin-ideal internalization at greater studies used similar populations and stimuli.
risk of body dissatisfaction after looking at thin-ideal images than
generic or specific disclaimers or no disclaimers at all. Tiggemann 3.5.4. Trait self-esteem
et al. (2017) also found that thin-ideal internalization moderated Only one study examined trait self-esteem as a moderator
the relationship between size of disclaimer used and body dis- between disclaimer condition and body image. Veldhuis et al.
satisfaction, such that women high in thin-ideal internalization (2014) found that adolescent girls with low trait self-esteem who
who saw the image only or the image + small-sized disclaimer were exposed to images of thin models with an information label
experienced greater post-exposure body dissatisfaction, whereas experienced less body dissatisfaction (␩2 p = .10) and less body con-
thin-ideal internalization was not related to body dissatisfaction in sciousness (␩2 p = .05) compared to girls who viewed the images
the control condition, or in the large and very large-sized disclaimer with a warning label or no disclaimer at all, and these repre-
conditions. These results by Tiggemann et al. (2017) indicate that sented small-medium effects. No differences in body dissatisfaction
when a disclaimer is not present or is less visible, women high on or consciousness were found for girls exposed to normal weight
thin-ideal internalization experience greater body dissatisfaction images as a function of self-esteem. Therefore, for girls with
than when the disclaimer is larger in size. Given the mixed results low self-esteem, informational disclaimers were protective against
48 S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52

body dissatisfaction and body consciousness caused by exposure to in fashion shoots to make the model’s bodies seem like unreal-
thin media images, while warning labels were not. istic targets for comparison, compared to the difficulty of making
both the model and the life portrayed in magazine advertisement
unrealistic.
4. Discussion
Finally, one study found that exposure to thin-ideal images
accompanied by generic disclaimers actually increased the levels
The purpose of the current review was to systematically exam-
of body dissatisfaction and lowered physical self-esteem, relative
ine what effect media disclaimers have on women’s body image
to exposure to the unaltered or retouched image alone (Harrison
and mood following exposure to thin-ideal images. This review was
& Hefner, 2014). This was the only study to use non-models as
undertaken to determine if there is a scientific basis for including
stimuli—instead the researchers used images of older students as
media disclaimers on thin-ideal images as a public policy strategy to
stimuli, and the disclaimer specifically identified the person in the
protect women’s body image and mood after exposure to thin-ideal
image as a student. By identifying the person in the image as a
images. It was the goal of this review to amalgamate all existing
student, the disclaimer may have actually communicated to the
findings on media disclaimers and body image and mood so that
participants that the person in the image was an appropriate tar-
if disclaimers were found to be an effective strategy at preventing
get for comparison, whereas those in the image only condition may
negative body image and mood after exposure to thin-ideal images,
have assumed the person was actually a model and therefore less
it could be better understood which types of disclaimers were the
of a suitable target for comparison. If the person in the image was
most effective, under what contexts, and for whom.
perceived to be just a student and a suitable target for comparison,
then it is possible this may have increased state social comparison
4.1. Summary of results for each type of disclaimer and greater subsequent body dissatisfaction, relative to viewing the
image alone where the person in the image may have been viewed
4.1.1. Generic disclaimers as a model and therefore a less suitable target for comparison.
Of the 11 studies that included generic disclaimers, eight stud-
ies found that generic disclaimers were ineffective at preventing 4.1.2. Specific disclaimers
body dissatisfaction relative to viewing a thin-ideal image alone, Four studies included specific disclaimers as a part of their
and generic disclaimers were not found to be more effective than experimental design, and three found that overall specific dis-
any other type of disclaimer (Bissell, 2006; Bury et al., 2016a, 2016b; claimers were ineffective at preventing body dissatisfaction after
Frederick et al., 2016; Tiggemann & Brown, 2018; Tiggemann et al., exposure to thin-ideal images, relative to being exposed to the
2013, 2017). It was also found that making disclaimers larger and image alone (Bury et al., 2016a, 2016b; Tiggemann et al., 2013).
more noticeable does not seem to make the disclaimer any more Also, no differences in body dissatisfaction were found between
effective than when it is smaller in size (Tiggemann et al., 2017). The those exposed to an image with a specific or generic disclaimer,
effect of generic disclaimers on body dissatisfaction was not mod- indicating that specifically outlining which body parts have been
erated by levels of thin-ideal internalization (Slater et al., 2012), altered in an image is not superior in mitigating body dissatisfac-
state and trait appearance comparison (Bury et al., 2016b, 2017; tion to simply stating that the image has been altered (Bury et al.,
Tiggemann et al., 2017) or sociocultural attitudes towards appear- 2016a; Tiggemann et al., 2013).
ance (Harrison & Hefner, 2014). In all but one study (Tiggemann Regarding the effect of trait appearance comparison on the
et al., 2013) generic disclaimers also did not reduce engagement in relationship between specific disclaimers and body dissatisfaction,
state appearance comparison to the thin-ideal images – those who results were somewhat inconsistent. It was found that for women
saw thin-ideal images with generic disclaimers engaged in rela- low on trait appearance comparison, there was little difference in
tively equal degrees of state appearance comparison as those who body dissatisfaction between those exposed to a thin-ideal image
saw thin-ideal images without the disclaimer (Bury et al., 2016a; alone, the image with a generic disclaimer, or the image with a
Frederick et al., 2016; Tiggemann & Brown, 2018; Tiggemann et al., specific disclaimer (Tiggemann et al., 2013). However, for women
2017). Finally, exposure to generic disclaimers neither reduced nor high on trait appearance comparison tendency, exposure to the
increased negative affect post-exposure, or relative to exposure to thin-ideal image with a specific disclaimer resulted in greater felt
the image alone (Harmon & Rudd, 2016; Slater et al., 2012). body dissatisfaction than their counterparts who saw the image
However, two studies did find that generic disclaimers were alone or the image with a generic disclaimer, indicating that for
effective at mitigating body dissatisfaction resulting from exposure these women specific disclaimers were actually harmful to body
to thin-ideal images (Harmon & Rudd, 2016; Slater et al., 2012); image rather than protective. Similarly, women that were high
however, these studies differed methodologically from the other on trait appearance comparison tendency experienced increased
studies reported. Harmon and Rudd (2016) recruited a very specific body dissatisfaction when they were asked to rate the appeal of
sample of participants majoring in Fashion and Retail studies, who thin-ideal advertisements that were accompanied by specific dis-
arguably are exposed to fashion advertisements more often than claimers (Bury et al., 2016a). Conversely, women that were low
the general population, and who may serve to benefit the most from on trait appearance comparison tendency actually derived bene-
a reminder that the images they see on a regular basis are digitally fit from the specific disclaimer, and experienced reduced levels of
altered and do not reflect realistic or attainable bodies. Also, unlike body dissatisfaction. Therefore, in this particular study the benefit
most other studies that used magazine advertisements for stimuli, of specific disclaimers was dependent upon women’s degree of trait
Slater et al. (2012) used fashion shoots. As has been pointed out by appearance comparison. On the other hand, contrary to the findings
other researchers (e.g., Tiggemann et al., 2013) fashion shoots may of the two studies explained above, Bury et al. (2016b) found that
appear more natural and realistic than the highly perfected images trait appearance comparison did not moderate the effect of specific
presented in magazines advertisements, which aim to present ide- disclaimers on body dissatisfaction. A post hoc power analysis of the
als of a happy, successful, or perfect life that are due to the use of parameters of that study by the current authors suggest that it had
the advertised product (Tiggemann et al., 2013). Therefore, when adequate power to detect a medium effect size. Eye tracking studies
placed on a magazine advertisement the disclaimer not only has have revealed that specific disclaimers direct more visual attention
to make the model’s body seem like an unrealistic target for com- to target body parts mentioned in specific disclaimer compared to
parison, but also has to make the successful life portrayed seem other types of disclaimers (Bury et al., 2016b, 2014), and therefore
unrealistic as well. Therefore, it may be simpler for disclaimers it may be that for women who are already likely to compare them-
S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52 49

selves to others, the specific disclaimer actually directs a greater ing body dissatisfaction. Warning disclaimers were not found to
amount of visual attention to the body parts mentioned in the dis- be any more effective than generic disclaimers at preventing body
claimer, which has been shown to increase the likelihood of state dissatisfaction, relative to viewing the image alone (Ata et al.,
appearance comparison and increase subsequent body dissatisfac- 2013; Tiggemann & Brown, 2018). However, results were mixed
tion (Tiggemann, Brown, & Thomas, 2019). However, if women are regarding how warning labels compared to information labels in
already unlikely to compare themselves to others the specific dis- their effectiveness to prevent body dissatisfaction after exposure
claimer serves the intended purpose of making the model seem like to thin-ideal images. Tiggemann and Brown (2018) found that both
an unrealistic target of comparison by outlining all the changes that warning and information labels were equally ineffective at pre-
have been made to the model’s body. venting increased body dissatisfaction in young women. However,
Slater et al. (2012) were the only researchers to find that specific Veldhuis et al. (2014) found that exposure to information labels
(and generic) disclaimers were effective at mitigating body dissat- reduced body dissatisfaction in adolescent girls, while exposure
isfaction after exposure to thin-ideal images, relative to viewing the to warning labels did not, indicating that information labels were
image alone. As in previous studies, specific disclaimers were not superior to warning labels at preventing body dissatisfaction result-
found to be any more effective than generic disclaimers in prevent- ing from exposure to thin-ideal images. At present, it is unknown
ing body dissatisfaction for this sample. As mentioned in the section whether cross-cultural differences could explain those differences
above, the most obvious difference between Slater et al.’ (2012) (the Tiggemann study was conducted with an Australian sample
methodological design and the other studies is that Slater et al. whereas the Veldhuis study was conducted with a Dutch sample).
(2012) were one of the only studies to use fashion shoot images, The effect of a number of individual difference moderators on
while most other studies used fashion advertisements; this differ- the relationship between warning labels and body dissatisfaction
ence in stimuli may account for why they found disclaimers to be were also explored. Neither trait social appearance comparison
effective and other studies did not. nor self-esteem moderated the relationship between exposure
In regard to the effect of specific disclaimers on engagement in to a warning disclaimer and body dissatisfaction (Ata et al.,
state appearance comparison, there was also some inconsistency 2013; Veldhuis et al., 2014). However, trait body dissatisfaction
in the results. Two studies found that there were no significant dif- and thin-ideal internalization were significant moderators of the
ferences in engagement in state appearance comparison between relationship between warning labels and post-exposure body dis-
those exposed to a thin-ideal image alone, those exposed to the satisfaction, such that those who had high levels of trait body
image with a generic disclaimer, or those exposed to the image dissatisfaction or internalization and who were exposed to the
with a specific disclaimer, indicating that specific disclaimers were warning label experienced higher post-exposure body dissatisfac-
ineffective at preventing comparison to the thin-ideal models (Bury tion scores, relative to those who saw the image alone or saw the
et al., 2016a; Tiggemann et al., 2013). However, in the second exper- image with a generic disclaimer (Ata et al., 2013). Body dissatisfac-
iment of this study, Tiggemann et al. (2013) found that women tion did not significantly differ between the image only, disclaimer,
who were exposed to the thin-ideal image with either a generic and warning conditions for those who were low on trait body
or specific disclaimer, actually engaged in more appearance com- dissatisfaction or internalization. Therefore, in this case, warning
parison than those in the image only condition, and results did not labels were actually found to be harmful to young women’s body
significantly differ between generic or specific disclaimers. Once image if they were already high in trait body dissatisfaction or inter-
again, these results indicate that specific disclaimers can actually nalization. It is possible that for women who are already dissatisfied
be harmful, instead of helpful, for young women’s body image. The with their bodies, and who internalize thinness as the standard of
mixed findings in regard to engagement in state appearance com- beauty, exposure to a warning that states trying to look like the
parison may be due to procedural differences between studies. model may be dangerous to their health only increases their dissat-
In both of the studies that found no differences between condi- isfaction with their body when they are explicitly told that this body
tions in state appearance comparison, participants had a set time type is difficult and dangerous to achieve. Essentially, the warning
limit of 45 s to view each image (Bury et al., 2016a; Tiggemann label may challenge women’s hopes of being able to achieve the thin
et al., 2013). However, in studies where differences in engagement ideal, which heightens their body dissatisfaction when they realize
in state appearance comparison were found, there was no time how unattainable it is to have a body that looks like a model’s body.
limit on how long participants could look at each image, and the This is consistent with previous research showing that thinness
researchers found that time spent looking at the images was quite attainability beliefs moderate the effects of exposure to thin-ideal
variable (12–46 seconds; Tiggemann et al., 2013). Therefore, it is images, such that perceiving thinness as more attainable (i.e., that
possible that the experimental groups spent different amounts of dieting works) results in less body dissatisfaction among dieters
time looking at the images, and therefore had different amounts of (Mills et al., 2002).
time to engage in social comparison. Bury et al. (2016b) did find Similarly, warning labels were also found to increase the levels
that those exposed to specific disclaimers spent more time looking of engagement in state appearance comparison, relative to expo-
at the model’s body parts mentioned in the disclaimer than those sure to an image alone, once again indicating the harmfulness of
exposed to the image alone or to a generic disclaimer. Therefore, it warning labels (Tiggemann & Brown, 2018). Additionally, those
is possible that women exposed to a specific disclaimer spent more exposed to a warning label experienced longitudinal increases in
time looking at the images than those in the other conditions, and the importance and attention placed on appearance from baseline
therefore spent more time engaging in social comparison. to a 4-week follow up, which was not experienced by those exposed
to the thin-ideal image alone (Kwan et al., 2018), further indicat-
4.1.3. Warning or consequence disclaimers ing the harmful effects of warning labels. Additionally, those who
Four studies examined the impact of including warn- were high on restrictive eating habits were found to consume less
ing/consequence labels (hereby referred to as warning labels) on kilocalories after exposure to a warning label, relative to exposure
thin-ideal images and their impact on resulting body dissatisfac- to the thin-ideal image alone, indicating that exposure to warn-
tion (Ata et al., 2013; Kwan et al., 2018; Tiggemann & Brown, 2018; ing labels can actually trigger more restrictive eating among those
Veldhuis et al., 2014). All four studies found that body dissatisfac- who are already likely to exhibit restrictive eating habits. It is pos-
tion did not differ between those exposed to a thin-ideal image sible that the counterintuitive effects that occurred after exposure
alone or those exposed to the thin-ideal image with the warning to warning disclaimers may have been because the warning dis-
label, indicating that warning labels were ineffective at prevent-
50 S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52

claimers actually directed greater visual attention to the model’s thinness, or state social comparison, relative to seeing the image
bodies than having no disclaimer at all (Bury et al., 2016a). alone (Fardouly & Holland, 2018; Frederick et al., 2016). Internal-
ization of the thin ideal was not found to moderate the relationships
4.1.4. Information disclaimers between subvertising condition and either body dissatisfaction or
Two studies examined the impact of viewing thin-ideal images mood (Fardouly & Holland, 2018). Therefore, preliminary results
with information disclaimers on body image and found differing suggest that subvertising labels are ineffective at preventing body
effects. It was found that body dissatisfaction levels did not dif- dissatisfaction and negative mood that result from exposure to
fer between young women who saw a thin-ideal image alone, or thin-ideal images.
those who saw the image with an information, warning, generic,
or graphic disclaimer, indicating that disclaimers with information 4.1.6. Graphic disclaimers
about the weight of models were ineffective at preventing body Only one study examined the impact of thin-ideal images with a
dissatisfaction in young women, relative to seeing the image alone graphic disclaimer on women’s body image (Tiggemann & Brown,
or seeing the image with any other type of disclaimer (Tiggemann 2018). It was found that body dissatisfaction and state appearance
& Brown, 2018). A post hoc power analysis of the parameters of that comparison did not differ between those who saw the images alone
study by the current authors suggests that it had sufficient power or those who saw the images with a graphic disclaimer. There were
to detect a medium effect size. Similarly, when participants were also no differences in body dissatisfaction between the graphic,
exposed to normal weight models with information disclaimers that generic, warning, or information disclaimer conditions. Therefore,
stated “these models are normal weight” the information disclaimer graphic disclaimers were ineffective at preventing body dissatis-
was found to be ineffective at preventing body dissatisfaction in faction, and were not superior to any other type of disclaimer.
adolescent girls, relative to viewing the images alone (Veldhuis
et al., 2014). In this case, information labels were also equally as 4.2. General summary of results
ineffective as warning labels in preventing body dissatisfaction.
These information disclaimers may have been unhelpful at pre- Of the 15 studies reviewed here, 11 found that disclaimers
venting body dissatisfaction because by telling participants that the did not mitigate body dissatisfaction after exposure to thin-ideal
models were normal weight, it implies the models should be sim- images, three found that disclaimers did mitigate body dissat-
ilar to most readers’ bodies; if the participants felt like their body isfaction, and seven found that disclaimers actually heightened
was not similar to the model this may be threatening because the body dissatisfaction in women who are already at a high risk for
model should be a reasonable target for comparison since she is of body dissatisfaction, such as those high in trait body dissatisfac-
a normal and achievable weight. If participants felt their body did tion and thin-ideal internalization. Of all the types of disclaimers,
not approximate the body of the model it makes sense that viewing specific and warning disclaimers seemed to be the most harmful
the information disclaimer would not prevent body dissatisfaction and dangerous to use, as they actually increased body dissatisfac-
from increasing. tion after exposure to thin-ideal images, among those high on trait
However, when adolescent girls were exposed to thin models appearance comparison, trait body dissatisfaction, thin-ideal inter-
with an information disclaimer that explicitly stated the models nalization, and those who already exhibit restrictive eating habits.
were underweight, they felt less body dissatisfaction and body con- Also, both specific and warning disclaimers were found to increase
sciousness than those who saw the images alone or those who engagement in state appearance comparison, which is actually
saw the images with a warning disclaimer (Veldhuis et al., 2014). counterintuitive to their designed purpose.
These results indicate that information disclaimers are effective, Additionally, there seem to be few commonalities and little con-
and superior to warning disclaimers, at mitigating body dissat- sensus among the studies that did find disclaimers to be effective
isfaction when adolescent girls are exposed to thin models, but at preventing poor body image after exposure to thin-ideal images.
have little to no effect on preventing body dissatisfaction caused Harmon and Rudd (2016) found generic disclaimers to be effective
by exposure to normal weight models. Further, these results were at preventing body dissatisfaction, while Slater et al. (2012) found
moderated by self-esteem levels. More specifically, girls who were both generic and specific disclaimers to be effective, and Veldhuis
low in self-esteem and who saw thin models with information dis- et al. (2014) found information disclaimers to be effective. Further-
claimers felt less body dissatisfaction and body consciousness than more, when disclaimers have been found to be effective it seems
those who saw images of the thin models alone or those who saw to be among very unique or specific samples. As mentioned above,
the images with a warning disclaimer. Levels of body dissatisfaction Harmon and Rudd (2016) only recruited participants majoring in
and consciousness did not significantly differ between the condi- Fashion and Retail studies, who are arguably exposed to fashion
tions for those who were high in self-esteem. Arguably, information magazine advertisements much more frequently than the average
labels about weight may be needed by adolescent girls with low person, which severely limits the generalizability of their results.
self-esteem to remind them that the models are not reasonable tar- Similarly, Veldhuis et al. (2014) were the only study to recruit ado-
gets for comparison, which girls with high self-esteem may already lescent girls from the Netherlands, rather than young women from
know. Australia or the United States. There is research that suggests that
girls and women from Western Europe experience less body dis-
4.1.5. Subvertising disclaimers satisfaction than their North American counterparts (Swami et al.,
Only two studies examined the impact of subvertising labels 2010), and therefore it may have been easier for media disclaimers
and thin-ideal images on women’s body image and mood. Both to prevent body dissatisfaction among European girls, who already
studies found that there were no differences in body dissatisfac- have healthier body image. Also, Slater et al. (2012) used differ-
tion between those who saw a thin-ideal image alone and those ent, and arguably less threatening, stimuli than most other studies
who saw the image with a subvertising label (Fardouly & Holland, (Tiggemann et al., 2013), and therefore it may have been easier for
2018; Frederick et al., 2016), indicating that subvertising labels media disclaimers to prevent body dissatisfaction in this study.
were ineffective at preventing body dissatisfaction relative to see- Therefore, overall, results consistently revealed that media dis-
ing the image alone. It was also found that subvertising labels were claimers had no protective effect on body dissatisfaction, drive for
no more effective than generic disclaimers at preventing body dis- thinness, negative mood, or eating behaviour, beyond the effects
satisfaction (Frederick et al., 2016). Subvertising disclaimers were of image exposure alone. The combined results suggest that media
also found to be ineffective at preventing negative affect, drive for disclaimers in general are not effective at mitigating the negative
S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52 51

effects of exposure to thin-ideal media images on body image and has found that published literature sometimes overestimates treat-
mood among young women, and they can actually be harmful for ment effects relative to unpublished work; however, this was found
women who are already at risk for poor body image. in only a minority of cases (Schmucker et al., 2017).

4.3. Implications 4.5. Future research

The current review confirms that, without disclaimers, expo- Future research should investigate the effectiveness of media
sure to thin-ideal images often causes more negative body image disclaimers in diverse samples of adolescents, men, and women to
among female viewers. As such, there is still a strong scientific basis see if the current findings are generalizable or specific to white,
for media literacy programs that aim to reduce social comparison educated, and Westernized female samples. More longitudinal
to unrealistic images that idealize thinness and appearance per- research is needed to examine the prolonged effect of exposure to
fection. However, given that the large majority of studies found disclaimers on women’s body image. In particular, research designs
disclaimers to be ineffective at preventing negative body image, and utilizing ecological momentary assessment (EMA) methods, with
were actually harmful in some cases, it cannot be concluded that multiple follow-up points, would be informative to ascertain how
proposals or laws that advocate to include disclaimers on altered long lasting the effects of disclaimers are on women’s body image,
media images, such as French and Israeli laws regarding retouched mood, and eating habits, in a more naturalistic and ecologically
images, are justified by the existing research. Systematic evalu- valid setting. Other moderators could also continue to be explored
ation is needed to determine whether these proposals and laws such as participant’s own photo-enhancing behaviours and physi-
regarding the use of disclaimers are actually helpful in amelio- cal appearance perfectionism. Social media may be becoming even
rating the negative effects of thin-ideal images on women’s body more important than traditional media in terms of triggering body
image. Based on current research, it seems unwise to include dis- dissatisfaction, owing to social comparison among peers, which has
claimers on media images as they have been shown across different been shown to be a more powerful predictor of body image con-
samples and teams of researchers to have negative consequences cerns than comparison to celebrities or models (Carey, Donaghue,
on body image for certain groups of women. This review suggests & Broderick, 2014). Exposure to attractive peers on social media
that when disclaimers are harmful it is because they increase or has been shown to cause increased negative body image (Hogue &
exacerbate women comparing their own appearance to that of Mills, 2019) in young women, and photo retouching of selfies is very
thin, idealized models. If disclaimers are to be used, they should common on social media (Mills, Musto, Williams, & Tiggemann,
be generic in nature. Specific and warning labels have both been 2018).
shown to be harmful to women’s body image (Ata et al., 2013; Bury
et al., 2016a; Tiggemann et al., 2013). While the research has found 4.6. Conclusions
that generic disclaimers are generally not helpful in protecting
women’s body image, they have also not been found to be harmful Overall, the results consistently reflect that media disclaimers
to women’s body image. Disclaimers have been touted as an inex- are ineffective at ameliorating the negative effects of thin-ideal
pensive and easy way to prevent body dissatisfaction. However, images on women’s body image or mood. In fact, disclaimers can
given the results of the current review, government officials and actually be harmful to the body image of women who are high
policy makers would do better to identify other strategies to reduce on trait body dissatisfaction, thin-ideal internalization, and social
the negative effects of thin-ideal media on women’s body image. and appearance comparison. In particular, specific disclaimers and
Other examples of prevention policy in action include legislation warning disclaimers seem to be more harmful than generic dis-
prohibiting the use of underweight fashion models or charters claimers. Therefore, the use of media disclaimers should be avoided,
encouraging positive body image through the use of diverse body or when used should be generic in nature.
types (Gauvin & Steiger, 2012).
Declaration of Competing Interest
4.4. Limitations of the literature
None.
Despite the fact that the quality assessment revealed that most
studies were of high quality, there are still several limitations to
Acknowledgements
the existing literature on media disclaimers and body image. The
largest set of limitations is in regard to the samples most commonly
The authors would like to thank Dr. Joel Katz for his edits on early
used. Results of the studies reviewed only generalize to primarily
drafts of this work, Keisha Gobin for her assistance in conducting
white, educated, women between the ages of 18–30 years who are
the quality assessment, and Lindsay Williams for her assistance in
Australian or American. Results cannot be generalized to adoles-
proofreading.
cents, older women, women of other racial or ethnic backgrounds,
or to adult men. Additionally, results can only be generalized pri-
References
marily to magazine advertisements, as the existing research has
not studied internet advertisements, billboards, celebrity images, Ata, R. N., Thompson, J. K., & Small, B. J. (2013). Effects of exposure to thin-ideal
or cover stories of magazines as stimuli with or without disclaimers, media images on body dissatisfaction: Testing the inclusion of a disclaimer
and only one study examined social media images. Furthermore, versus warning label. Body Image, 10, 472–480. http://dx.doi.org/10.1016/j.
bodyim.2013.04.004
several studies did not include a manipulation check to ensure that Bennett, J. (2008). Picture perfect. In Newsweek.. Retrieved from. http://www.
participants had actually seen or read the disclaimers, and there- newsweek.com/2008/05/01/picture-perfect.html
fore it is difficult to judge whether null effects are due to the fact Bissell, K. L. (2006). Skinny like you: Visual literacy, digital manipulation and young
women’s drive to be thin. Simile, 6, 4. http://dx.doi.org/10.3138/sim.6.1.002
that disclaimers are ineffective, or because participants simply did
Bozsik, F., Whisenhunt, B. L., Hudson, D. L., Bennett, B., & Lundgren, J. D. (2018).
not notice them. Finally, it was a limitation that only published Thin is in? Think again: The rising importance of muscularity in the thin ideal
peer-reviewed articles were included in the present systematic female body. Sex Roles, 79, 609–615. http://dx.doi.org/10.1007/s11199-017-
review, and unpublished work was not considered. Therefore, it 0886-0
Breen, K. (2018). Toronto MPP wants disclaimers on airbrushed, altered images in
is possible that the results of this review could be biased by overes- advertising March, Retrieved from. https://globalnews.ca/news/4065445/
timating the effects of disclaimers on body image, as past research toronto-mpp-disclaimers-advertsing/
52 S.E. McComb, J.S. Mills / Body Image 32 (2020) 34–52

Bury, B., Tiggemann, M., & Slater, A. (2014). Directing gaze: The effect of disclaimer dissatisfaction in anorexia and bulimia nervosa. Journal of Psychiatric Research,
labels on women’s visual attention to fashion magazine advertisements. Body 82, 119–125. http://dx.doi.org/10.1016/j.jpsychires.2016.07.021
Image, 11, 357–363. http://dx.doi.org/10.1016/j.bodyim.2014.06.004 Neighbors, L. A., & Sobal, J. (2007). Prevalence and magnitude of body weight and
Bury, B., Tiggemann, M., & Slater, A. (2016a). The effect of digital alteration shape dissatisfaction among university students. Eating Behaviors, 8, 429–439.
disclaimer labels on social comparison and body image: Instructions and http://dx.doi.org/10.1016/j.eatbeh.2007.03.003
individual differences. Body Image, 17, 136–142. http://dx.doi.org/10.1016/j. Neumark-Sztainer, D., Paxton, S., Hannan, P., Haines, J., & Story, M. (2006). Does
bodyim.2016.03.005 body satisfaction matter? Five-year longitudinal associations between body
Bury, B., Tiggemann, M., & Slater, A. (2016b). Disclaimer labels on fashion magazine satisfaction and health behaviors in adolescent females and males. Journal of
advertisements: Impact on visual attention and relationship with body Adolescent Health, 39, 244–251. http://dx.doi.org/10.1016/j.jadohealth.2005.12.
dissatisfaction. Body Image, 16, 1–9. http://dx.doi.org/10.1016/j.bodyim.2015. 001
09.005 Rodgers, R. F., Ziff, S., Lowy, A. S., Yu, K., & Austin, S. B. (2017). Results of a strategic
Bury, B., Tiggemann, M., & Slater, A. (2017). Disclaimer labels on fashion magazine science study to inform policies targeting extreme thinness standards in the
advertisements: Does timing of digital alteration matter? Eating Behaviors, 25, fashion industry. The International Journal of Eating Disorders, 50, 284–292.
18–22. http://dx.doi.org/10.1016/j.eatbeh.2016.08.010 http://dx.doi.org/10.1002/eat.22682
Carey, R. N., Donaghue, N., & Broderick, P. (2014). Body image concern among Runfola, C. D., Von Holle, A., Trace, S. E., Brownley, K. A., Hofmeier, S. M., & Gagne,
Australian adolescent girls: The role of body comparisons with models and D. A., et al. (2013). Body dissatisfaction in women across the lifespan: Results
peers. Body Image, 11, 81–84. http://dx.doi.org/10.1016/j.bodyim.2013.09.006 of the UNC-SELF and Gender and Body Image (GABI) Studies. European Eating
Charrlson, E. (2014). Should photoshopped images have disclaimers? Retrieved from. Disorders Review, 21, 52–59. http://dx.doi.org/10.1002/erv.2201
In Daily. https://indaily.com.au/news/2014/11/28/photoshopped-images- Santhira Shagar, P., Harris, N., Boddy, J., & Donovan, C. L. (2018). The relationship
disclaimers/ between body image concerns and weight-related behaviours of adolescents
Choi, E., & Choi, I. (2016). The associations between body dissatisfaction, body and emerging adults: A systematic review. Behaviour Change, 34, 208–252.
figure, self-esteem, and depressed mood in adolescents in the United States http://dx.doi.org/10.1017/bec.2018.3
and Korea: A moderated mediation analyses. Journal of Adolescence, 53, Schaefer, L. M., & Thompson, J. K. (2014). The development and validation of the
249–259. http://dx.doi.org/10.1016/j.adolescence.2016.10.007 Physical Appearance Comparison Scale-Revised (PACS-R). Eating Behaviors, 15,
Crow, S., Eisenberg, M. E., Story, M., & Neumark-Sztainer, D. (2008). Suicidal 209–217. http://dx.doi.org/10.1016/j.eatbeh.2014.01.001
behavior in adolescents: Relationship to weight status, weight control Schmucker, C. M., Blümle, A., Schell, L. K., Schwarzer, G., Oeller, P., & Cabrera, L.,
behaviors, and body dissatisfaction. The International Journal of Eating et al. (2017). Systematic review finds that study data not published in full text
Disorders, 41, 82–87. http://dx.doi.org/10.1002/eat.20466 articles have unclear impact on meta-analyses results in medical research. PloS
Eggert, N. (2017). Is she photoshopped? In France, they now have to tell you Retrieved One, 12 http://dx.doi.org/10.1371/journal.pone.0176210
from. BBC News. http://www.bbc.com/news/world-europe-41443027 Slater, A., Tiggemann, M., Firth, B., & Hawkins, K. (2012). Reality check: An
Fallon, E. A., Harris, B. S., & Johnson, P. (2014). Prevalence of body dissatisfaction experimental investigation of the addition or warning labels to fashion
among a United States adult sample. Eating Behaviors, 15, 151–158. http://dx. magazine images on women’s mood and body dissatisfaction. Journal of Social
doi.org/10.1016/j.eatbeh.2013.11.007 and Clinical Psychology, 31, 105–122. http://dx.doi.org/10.1521/jscp.2012.31.2.
Fardouly, J., & Holland, E. (2018). Social media is not real life: The effect of 105
attaching disclaimer-type labels to idealized social media images on women’s Spitzer, B. L., Henderson, K. A., & Zivian, M. T. (1999). Gender differences in
body image and mood. New Media & Society, 20, 4311–4328. http://dx.doi.org/ population versus media body sizes: A comparison over four decades. Sex
10.1177/1461444818771083 Roles, 40, 545–565. http://dx.doi.org/10.1023/A:1018836029738
Frederick, D. A., Daniels, E. A., Bates, M. E., & Tylka, T. L. (2017). Exposure to Stice, E., Marti, C. N., & Durant, S. (2011). Risk factors for onset of eating disorders:
thin-ideal media affect most, but not all, women: Results from the Perceived Evidence of multiple risk pathways from an 8-year prospective study.
Effects of Media Exposure Scale and open-ended responses. Body Image, 23, Behaviour Research and Therapy, 49, 622–627. http://dx.doi.org/10.1016/j.brat.
188–205. http://dx.doi.org/10.1016/j.bodyim.2017.10.006 2011.06.009
Frederick, D. A., Sandhu, G., Scott, T., & Akbari, Y. (2016). Reducing the negative Swami, V., Frederick, D. A., Aavik, T., Alcalay, L., Allik, J., & Anderson, D., et al.
effects of media exposure on body image: Testing the effectiveness of (2010). The attractive female body weight and female body dissatisfaction in
subvertising and disclaimer labels. Body Image, 17, 171–174. http://dx.doi.org/ 26 countries across 10 world regions: Results of the International Body Project
10.1016/j.bodyim.2016.03.009 I. Personality & Social Psychology Bulletin, 36, 309–325. http://dx.doi.org/10.
Gauvin, L., & Steiger, H. (2012). Overcoming the unhealthy pursuit of thinness: 1177/0146167209359702
Reaction to the Quebec Charter for a Healthy and Diverse Body Image. Sypeck, M. F., Gray, J. J., & Ahrens, A. H. (2004). No longer just a pretty face: Fashion
American Journal of Public Health, 102, 1600–1606. http://dx.doi.org/10.2105/ magazines’ depictions of ideal female beauty from 1959-1999. The International
AJPH.2011.300479 Journal of Eating Disorders, 36, 342–347. http://dx.doi.org/10.1002/eat.20039
Groesz, L. M., Levine, M. P., & Murnen, S. K. (2002). The effect of experimental Thompson, J. K., Heinberg, L. J., & Tantleff, S. (1991). The Physical Appearance
presentation of thin media images on body satisfaction: A meta-analytic Comparison Scale. The Behavior Therapist, 14, 174.
review. The International Journal of Eating Disorders, 31, 1–16. http://dx.doi.org/ Tiggemann, M., & Brown, Z. (2018). Labelling fashion magazine advertisements:
10.1002/eat.10005 Effectiveness of different label formats on social comparison and body
Harmon, J., & Rudd, N. A. (2016). Breaking the illusion: The effects of adding dissatisfaction. Body Image, 25, 97–102. http://dx.doi.org/10.1016/j.bodyim.
warning labels identifying digital enhancement on fashion magazine 2018.02.010
advertisements. Fashion Style & Popular Culture, 3, 357–374. http://dx.doi.org/ Tiggemann, M., Brown, Z., & Thomas, N. (2019). (Don’t) look here!” the effect of
10.1386/fspc.3.3.357 1 different forms of label added to fashion advertisements on women’s visual
Harrison, K., & Cantor, J. (1997). The relationship between media consumption and attention. Body Image, 31, 88–95. http://dx.doi.org/10.1016/j.bodyim.2019.08.
eating disorders. The Journal of Communication, 47, 40–67. http://dx.doi.org/10. 011
1111/j.1460-2466.1997.tb02692.x Tiggemann, M., Brown, Z., Zaccardo, M., & Thomas, N. (2017). “Warning: This
Harrison, K., & Hefner, V. (2014). Virtually perfect: Image retouching and image has been digitally altered”: The effect of disclaimer labels added to
adolescent body image. Media Psychology, 17, 134–153. http://dx.doi.org/10. fashion magazine shoots on women’s body dissatisfaction. Body Image, 21,
1080/15213269.2013.770354 107–113. http://dx.doi.org/10.1016/j.bodyim.2017.04.001
Hogue, J. V., & Mills, J. S. (2019). The effects of active social media engagement with Tiggemann, M., & McGill, B. (2004). The role of social comparison in the effect of
peers on body image in young women. Body Image, 28, 1–5. http://dx.doi.org/ magazine advertisements on women’s mood and body dissatisfaction. Journal
10.1016/j.bodyim.2018.11.002 of Social and Clinical Psychology, 23, 23–44. http://dx.doi.org/10.1521/jscp.23.1.
Krawitz, M. (2014). Beauty is only photoshop deep: Legislating models’ BMIs and 23.26991
photoshopping images. Journal of Law and Medicine, 21, 859–874. Tiggemann, M., & Pickering, A. S. (1996). Role of television in adolescent women’s
Kwan, M. Y., Haynos, A. F., Blomquist, K. K., & Roberto, C. A. (2018). Warning labels body dissatisfaction and drive for thinness. The International Journal of Eating
on fashion images: Short- and longer-term effects on body dissatisfaction, Disorders, 20, 199–203. http://dx.doi.org/10.1002/(SICI)1098-
eating disorder symptoms, and eating behavior. The International Journal of 108X(199609)20:2<199::AID-EAT11>3.0.CO;2-Z
Eating Disorders, 51, 1153–1161. http://dx.doi.org/10.1002/eat.22951 Tiggemann, M., Slater, A., Bury, B., Hawkins, K., & Firth, B. (2013). Disclaimer labels
Mills, J. S., Musto, S., Williams, L., & Tiggemann, M. (2018). “Selfie” harm: Effect on on fashion magazine advertisements: Effects on social comparison and body
mood and body image in young women. Body Image, 27, 86–92. http://dx.doi. dissatisfaction. Body Image, 10, 45–53. http://dx.doi.org/10.1016/j.bodyim.
org/10.1016/j.bodyim.2018.08.007 2012.08.001
Mills, J. S., Polivy, J., Herman, C. P., & Tiggemann, M. (2002). Effects of exposure to Veldhuis, J., Konijn, E. A., & Seidell, J. C. (2014). Counteracting media’s thin-body
thin media images: Evidence of self-enhancement among restrained eaters. ideal for adolescent girls: Informing is more effective than warning. Media
Personality & Social Psychology Bulletin, 28, 1687–1699. http://dx.doi.org/10. Psychology, 17, 154–184. http://dx.doi.org/10.1080/15213269.2013.788327
1177/014616702237650 Wiseman, C. V., Gray, J. J., Mosimann, J. E., & Ahrens, A. H. (1992). Cultural
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & Group, P. (2009). Preferred expectations of thinness in women: An update. The International Journal of
reporting items for systematic reviews and meta-analyses: The PRISMA Eating Disorders, 11, 85–89. http://dx.doi.org/10.1002/1098-
statement. PLoS Medicine, 6, e1000097. 108X(199201)11:1<85::AID-EAT2260110112>3.0.CO;2-T
Naumann, E., Tuschen-Caffier, B., Voderholzer, U., Schafer, J., & Svaldi, J. (2016).
Effects of emotional acceptance and rumination on media-induced body

You might also like