You are on page 1of 10

899664 JOO Journal of Orthodontics

Scientific Section

Journal of Orthodontics

The effect of viewing idealised smile 2020, Vol. 47(1) 55­–64


https://doi.org/10.1177/1465312519899664
DOI: 10.1177/1465312519899664
© The Author(s) 2020
images versus nature images via social Article reuse guidelines:
sagepub.com/journals-permissions
media on immediate facial satisfaction journals.sagepub.com/home/joo

in young adults: A randomised


controlled trial

Ariane Sampson1 , Huw G Jeremiah2, Manoharan Andiappan3


and J Tim Newton4

Abstract
Introduction: The objective of this randomised controlled trial (RCT) is to investigate the effect of a social networking site
(SNS) on body dissatisfaction, facial and smile dissatisfaction, and face-related discrepancy, and whether these effects differ
from the use of appearance-neutral Instagram images. We also aimed to investigate whether there are trends in increased self-
reported use of social media and increased body dissatisfaction, facial and smile dissatisfaction, and face-related discrepancy.
Methods: Undergraduate students were randomly allocated to an experimental group with idealised smile images or to
a control group with neutral nature images. They completed pre-exposure questionnaires, then perused for 5 min their
allocated images on individual Apple iPads via the Instagram application. Participants then completed the post-exposure
surveys. The main outcome was facial dissatisfaction. Body dissatisfaction and total facial and body dissatisfaction were
secondary outcomes. Simple randomisation was achieved with a computerised random number generator. Data were
analysed using repeated measures ANOVA and multivariate regression analyses.
Results: A total of 132 participants (mean age = 20.50 ± 2.21 years) were randomised to either the experimental
group with idealised smile images (n=71) or the control group with neutral nature images (n=61). Baseline characteristics
were similar between groups and no participants were lost. Exposure to ‘ideal’ facial images on social media decreases
facial satisfaction (95% confidence interval [CI] = 0.85–1.05; P < 0.0001). Individuals with high baseline self-discrepancy
scores are less satisfied with their facial features and body appearance (95% CI = 0.04–1.16; P = 0.036).
Conclusion: This study shows that viewing SNSs with high visual media reduces satisfaction with facial appearance in
the short term in men and women. This effect is greater in those with high self-discrepancy scores. Increased media
usage was not correlated with increased dissatisfaction. Wearing braces or having had braces was shown not to influence
post-exposure dissatisfaction.

Keywords
facial satisfaction, facial dissatisfaction, body satisfaction, body dissatisfaction, social media, Instagram, self-discrepancy,
internalisation

Date received: 29 March 2019; revised: 21 November 2019; accepted: 18 December 2019

1
F aculty of Dentistry, Oral & Craniofacial Sciences, King’s College
London, London, UK
Introduction 2
Orthodontic Department, Addenbrooke’s Hospital, Cambridge
University Hospital NHS Foundation Trust, Cambridge, UK
Much research has studied the effect of traditional media 3
Biostatistics and Research Methods Centre, Faculty of Dentistry, Oral
such as television and fashion publications on body image. & Craniofacial Sciences, King’s College London, London, UK
The exponential rise in Internet use has driven researchers 4
Department of Population and Patient Health, Faculty of Dentistry,
to study Internet sites, which have shown the immediate Oral & Craniofacial Sciences, King’s College London, London, UK
negative effect of social media on body image, particularly
Corresponding author:
in young women (Brown and Tiggemann, 2016; Fardouly Ariane Sampson, Faculty of Dentistry, Oral & Craniofacial Sciences,
and Vartanian, 2015; Fardouly et al., 2015; Groesz et al., King’s College London, London, UK.
2001; Haferkamp and Kramer, 2011; Tiggemann and Email: ariane_sampson@hotmail.com
56 Journal of Orthodontics 47(1)

Miller, 2010; Tiggemann et al., 2009). It was found that are more susceptible to triggers that result in dejection-
most post-Facebook-exposure body dissatisfaction was related emotions such as depression (Cattell, 1973) and
related to the photographic function of Facebook rather have lower self-esteem (Rosenberg, 1979).
than total time spent on it (Cohen et al., 2017; Meier and There is currently no theoretically driven experiment on
Gray, 2014). A recent systematic review on the effect of the effect of Instagram, or indeed any other SNS high in
social networking sites (SNSs) on body image (Holland visual media, on smile and facial dissatisfaction, and no
and Tiggemann, 2016) showed that the current pool of research to this effect in dental specialties such as ortho-
research overwhelmingly links the use of social media with dontics. The orthodontic patient base is made up predomi-
body image. nantly of teenagers and young adults, a cohort most likely
With 2.62 billion active users (eMarketer, 2018), social to be large users of SNSs high in visual media. This research
media use is inescapable and accelerating exponentially. would help us understand the effects of SNSs high in visual
Instagram, an SNS high in visual media, is a unique photo- media on mental wellbeing, smile satisfaction, perceived
sharing form of social media boasting upwards of 48,611 treatment need, motivations for treatment, patient ideals
photographs posted per minute (Instagram, 2018) and the and patient triggers.
highest rate of social engagement per post compared to all The main aim of the present study was to investigate the
other social media platforms (Leone, 2018). This platform immediate effect of an SNS high in visual media on body
can engage users in appearance-related social comparisons. dissatisfaction, facial and smile dissatisfaction, and face-
With Instagram users able to interact equally with peers related discrepancy (the extent of which one wants to
and celebrities, this creates an opportunity for regular expo- change specific aspects of their facial features), and whether
sure to idealised images. Viewing fitness inspiration (‘fit- these effects differ from the use of appearance-neutral
spiration’) images (Tiggemann and Zaccardo, 2015) or images. We also aim to investigate whether there are trends
celebrity and peer images (Brown and Tiggemann, 2016) in increased self-reported use of social media and increased
on Instagram increases negative mood and body image body dissatisfaction, facial and smile dissatisfaction, and
compared to neutral images in young women. face-related discrepancy. The null hypothesis assumes that
Some research indicates that young men are also suscep- the use of SNSs high in visual media does not result in
tible to increased body dissatisfaction, particularly in terms greater body dissatisfaction, greater facial and smile dis-
of muscular ideals (Cafri et al., 2005). Despite these find- satisfaction, or greater face-related discrepancy, and that
ings, few studies have been done on the role of social com- the effects of SNSs high in visual media is not linked to
parison in men as a mediator of idealised male media increased self-reported use of media and technology, not
images and body dissatisfaction. greater in women than in men, not less in those with a his-
To date, there have been three meta-analyses investigat- tory of orthodontic treatment and not greater compared to
ing these associations. There was significantly increased exposure to neutral images.
negative body image after viewing thin ideal media images.
Effects are stronger in women and are inversely related to
age (Grabe et al., 2008; Groesz et al., 2002; Myers and Methods
Crowther, 2009). A randomised controlled trial (RCT) was used to investi-
Festinger’s Social Comparison Theory outlines the natu- gate the effect of SNS image type (Instagram user images
ral human desire to compare themselves to others (Festinger, versus Instagram appearance-neutral images) on dependent
1954). This theory helps explain the negative effects of variables of negative body, facial and smile dissatisfaction,
media exposure on body image so often found in many and face-related discrepancy. After trial commencement,
studies (Keery et al., 2004; Myers and Crowther, 2009; van no changes to the methods occurred.
den Berg et al., 2007; Vartanian and Dey, 2013). This is The sample size calculation was carried out using a
particularly rife when women are able to compare them- medium effect size of 0.5 to detect the significant differ-
selves to those most similar to them, such as their peers ence in facial satisfaction scores between experimental and
(Heinberg and Thompson, 1995). The greater the tendency control groups with 80% power using a two-tailed test
to compare, the higher the level of body dissatisfaction (Norman et al., 2012). It showed that a sample of 128 par-
(Keery et al., 2004; Myers and Crowther, 2009; van den ticipants would be sufficient to detect a significant differ-
Berg et al., 2002, 2007; Vartanian and Dey, 2013) which ence, and there was slight over-recruitment due to subjects
may lead to the development of negative behaviours such being approached simultaneously. The final sample of 132
as disordered eating habits (Saunders and Eaton, 2018) and had sufficient power to perform all the other analyses car-
orthorexia (Turner and Lefevre, 2017). ried out in the study.
One type of self-discrepancy (Freud, 1923; Higgins Participants (n = 132) were male and female undergrad-
et al., 1985) is when a person’s ‘actual’ self does not match uate students from the university campus, recruited between
their ‘ideal’ self. Those who are high on this type of dis- September 2017 and January 2018, for what they were told
crepancy level are more likely to draw social comparisons, was a study on social media. Students of Dentistry or
Sampson et al. 57

Psychology were excluded from the study. Participants thought about appearance), compared their overall appear-
were aged 18–35 years as 90% of Instagram users are aged ance to the photographs (1 = no comparison, 7 = a lot of
18–35 years (Jin Yea Jang, 2016). comparison) and compared specific body parts / teeth or
Two sets of images were compiled for the study: the smile to photographs (1 = no comparison, 7 = a lot of
experimental group with 60 Instagram user images of comparison). Scores were averaged to give a total score for
attractive smiles; and the control group with 60 appearance- state appearance comparison. This scale was shown to have
neutral Instagram images of nature. All the images were sound inter-correlation (Tiggemann and McGill, 2004).
chosen from public Instagram images. The state variant of the Self-Discrepancy Index (SDI)
A preliminary study was carried out to choose images for (Dittmar et al., 1996; Higgins, 1985) was used to measure
the experimental group. The first author shortlisted 100 con- appearance discrepancy. Participants were asked to write
secutive public photographs of people with full, attractive out up to five aspects of their physical appearance that they
smiles and well-aligned teeth with the ‘#smile’ hashtag on would like to change, to rate how different they would like
Instagram. The photographs featured an equal proportion of to be from what they currently are (1 = a little different, 6
celebrity and non-celebrity (< 1000 followers) faces and = extremely different) and how significant this difference
men and women of a similar age group to those who would would be to them (1 = a little important, 6 = extremely
be participating in the study (18–35 years). All five male important).
and female independent raters who were Orthodontic Questions were asked to assess participants’ media and
Consultants, Specialist Orthodontic Practitioners or final technology usage using the Media and Technology Usage
year Orthodontic postgraduate students from the hospital’s and Attitudes Scale MTUAS (Rosen et al., 2013). The 46
own Orthodontic Department were asked to rate the 100 questions included frequency of use of e-mail, web
facial images on a series of 5-point Likert scales. ­browsing, photo-taking, video-watching, gaming, calling,
Orthodontists were chosen to carry out the preliminary texting, sharing files and engagement on social media
study as they are known to be the most discerning critics of sites (1 = never, 10 = all the time). Adjustments were
what constitutes an ideal smile and ideal teeth (Kokich et al., made to this questionnaire to include specific information
1999; Witt and Flores-Mir, 2011). The images were judged on Instagram use, including number of followers, number
based on attractiveness of the smile and teeth (1 = very of personal friends followed, number of celebrities
attractive, 5 = very unattractive), general appeal (1 = very ­followed and number of people regularly interacted with
appealing, 5 = not at all appealing), image quality (1 = (1 = 0, 2 = 1–50, 3 = 51–100, 4 = 101–175, 5 = 176–
good quality, 5 = not good quality) and enjoyment (1 = 250, 6 = 251–375, 7 = 376–500, 8 = 501–750, 9 = >
very enjoyable, 5 = not at all enjoyable). The top scoring 60 751). Participants could then be divided into low media
images were selected for the experimental group. The users or high media users.
repeatability of the participants’ ratings was not assessed in Following exposure to images, participants were asked
line with previous research in this area using this methodol- their age, gender, ethnicity (Office for National Statistics,
ogy (Brown and Tiggemann, 2016). 2015) and history of orthodontic treatment (yes or no).
The neutral images for the control group were chosen The primary outcome was facial dissatisfaction. The
from the top 60 consecutive images with the ‘#nature’ secondary outcomes were body dissatisfaction and total
hashtag on Instagram. These images featured only nature overall body and facial dissatisfaction. After trial com-
and had no people in them. The choice of nature was based mencement, no changes to outcome measures occurred.
on the stimuli used in previous research using this method- Participants were invited to read the introduction and
ology (Brown and Tiggemann, 2016; Tiggemann and information sheets, after which they provided informed
Zaccardo, 2015). consent and completed pre-exposure surveys. Participants
Body and facial dissatisfaction were measured using the were then randomly divided into one of two conditions
Body Satisfaction Scale (BSS) (Slade et al., 1990) and the (experimental group versus control group) by simple com-
Facial Satisfaction Scale (FSS) component within the BSS puterized list randomisation (Haahr, 2018) and asked to
(1 = very satisfied, 7 = very dissatisfied; range = 1–7), browse their allocated page for 5 min.
amounting to two questions. The BSS was scored as out- It was impossible to blind participants or operator to
lined in the original psychometric validation of the measure their group allocation as they were immediately aware of
(Slade et al., 1990). A question was included as a visual the type of images they were viewing. There was mild
analogue scale asking participants how they felt about their deception in terms of the objectives of the research as par-
smile at that precise moment in time (0 = very happy, 100 ticipants were not told the true purpose of the study until
= not happy at all). after completing the task. Participants were told at recruit-
The State Appearance Comparison Scale (Tiggemann ment that this was simply a study on social media.
and McGill, 2004) was used to determine on three 7-point The Instagram pages were presented to participants by
Likert Scales whether individuals thought about their own the first author on individual Apple iPads via the Instagram
appearance (1 = no thought about appearance, 7 = a lot of Application, a real-time social media platform. This
58 Journal of Orthodontics 47(1)

Figure 1.  CONSORT participant flow diagram.

replicates the most natural method of browsing SNSs. Multivariate regression analyses were carried out to
After perusing the Instagram page, participants completed find out the significant predictors of BSS, FSS and total
the post-exposure surveys. The protocol had not been scores. Before carrying out the multiple regression, the
published before trial commencement. This study was normality assumption of the outcomes was tested using
approved by Kings College London’s ethics committee the methods described above. Separate regression analy-
(LRU-16/17-4241). ses for each of the outcome measures, namely BSS, FSS
and total scores, were carried out by treating the outcome
measure as dependent variables of time, gender, group,
Statistical analysis braces, media use, SDI and the interaction of group X
Descriptive statistics summarised the BSS, FSS and total media use as the predictor variables. As all the predictors
scores at two time points. The association between demo- have only two categories and the interactions were not
graphic variables and the groups were compared using the significant, no multiple comparisons were carried out.
Chi-square test. Before carrying out the parametric tests, Statistical significance was assumed at a level of 0.05.
the normality assumption was tested by plotting residual The analyses were carried out using the Statistical Package
values using histograms, Shapiro–Wilk test, Kolmogorov– for Social Sciences (Release 24.0.0, SPSS Inc., IBM
Smirnov test and box plots (P = 0.12). As the data satis- Corp., Chicago, IL, USA) and Stata 13 (Stata Corp,
fied normality assumptions, parametric tests were carried College Station, TX, USA).
out. Repeated measures ANOVA was used to compare
BSS, FSS and total scores separately between groups
(experimental group versus control group), time points Results
(pre-exposure and post-exposure) and the interaction A total of 132 participants took part in this study, out of
between groups and time points. Sphericity was tested which 80 (60.6%) were women and 52 (39.4%) were men.
while using repeated measures ANOVA. As the interac- The mean age of the participants was 20.5 years (SD 2.21).
tions were significant the linear combinations were tested Participants identified as White (n = 62, 47%), Asian
using the Least Significant Difference (LSD) test to han- (n = 34, 25.8%), Mixed (n = 15, 11.3%), Black (n = 4,
dle multiple comparisons. 3%) and Other (n = 17, 12.9%). Of the participants, 53%
Sampson et al. 59

Table 1.  Demographic characteristics of study sample.

Experimental group Control group Overall χ2 P

Gender Male 27 (38) 25 (41) 52 (39) 0.12 0.73

  Female 44 (62) 36 (59) 80 (61)  

Age Mean ± SD 20.51 ± 2.34 20.49 ± 2.07 20.50 ± 2.21  

  Median (range) 20 (18–30) 20 (18–27) 20.0 (18–30)  

Ethnicity 7.5 0.73

White English 20 11 31  

  Irish 0 0 0  
  Gypsy 0 0 0  
  Other 16 15 31  

Mixed White + Black Caribbean 0 1 1  

  White + Black African 1 1 2  

  White + Asian 3 3 6  

  White + Other 3 3 6  

Asian Indian 8 9 17  
  Pakistani 3 1 4  
  Bangladeshi 0 1 1  
  Chinese 2 4 6  
  Other 3 3 6  

Black African 1 2 3  
  Caribbean 0 0 0  
  Other 1 0 1  

Other Arab 6 4 10  

  Other 4 3 7  

Braces Yes 39 31 70 0.22 0.64

  No 32 30 62  

said that they had had orthodontic braces. No patients were All the dissatisfaction scores (BSS, FSS and total scores)
lost (Figure 1). increased in the post-exposure period for the experimental
Comparison of the two groups confirmed no statistically group whereas for the control group it showed a declining
significant differences between groups on variables of age, trend (Figure 2).
gender, ethnicity or history of braces (Table 1), thus ran- Multivariate regression analyses were carried out to find
dom allocation was successful. out the significant predictors of BSS, FSS and total scores.
Values are given as n (%), mean ± SD or median (range). The results of the multivariate regression are summarized
Descriptive statistics summarised BSS, FSS and total for each of the outcome measures in Table 4.
scores. The estimated marginal means based on the repeated The dissatisfaction scores are not influenced by gen-
measures ANOVA are given in Table 2. Repeated measures der (P > 005). The FSS and the total scores differed sig-
ANOVA compared BSS, FSS and total scores separately nificantly between groups. A history of braces and
between groups (experimental and control groups), time increased use of media have no effect on dissatisfaction
points (pre- and post-exposure) and the interaction between scores.
groups and time points. The results of the comparison are In line with CONSORT reporting requirements, adverse
given in Table 3. events were monitored and none were reported.
60 Journal of Orthodontics 47(1)

Table 2.  Estimated marginal means for Body Satisfaction Scale (BSS), Facial Satisfaction Scale (FSS) and total scores according to
group and time.

Variable Mean (95% CI)

  FSS BSS Total score

Group
Experimental group 18.29 (16.57–20.01) 19.24 (17.58–20.90) 37.53 (34.51–40.55)
Control group 15.86 (14.01–17.72) 16.41 (14.62–18.20) 32.27 (29.02–35.53)

Time
Pre-exposure 16.57 (15.33–17.81) 17.69 (16.42–18.97) 34.26 (32.05–36.47)
Post-exposure 17.58 (16.22–18.94) 17.94 (16.73–19.18) 35.54 (33.23–37.84)

Group X time
1 Pre-exposure 17.06 (15.37–18.74) 18.83 (17.10–20.56) 35.89 (32.88–38.89)
  Post-exposure 19.52 (17.67–21.38) 19.65 (17.99–21.31) 39.17 (36.03–42.31)
2 Pre-exposure 16.08 (14.26–17.90) 16.56 (14.69–18.42) 32.64 (29.40–35.88)
  Post-exposure 15.64 (13.64–17.64) 16.26 (14.47–18.06) 31.90 (28.52–35.29)

CI, confidence interval.

Table 3.  Comparison of Body Satisfaction Scale (BSS), Facial Satisfaction Scale (FSS) and total scores between group and time.

Variable P value

  FSS BSS Total score

Time [pre-exposure vs. post-exposure] (p) 0.002* 0.293 0.004*

F (df) 10.08 (1–130) 1.11 (1–130) 8.84 (1–130)

Group [Control group vs. Experimental group] 0.06 0.026* < 0.0001*

F (df) 3.61 (1–130) 5.24 (1–130) 22.06 (1–130)

Interaction [Group X Time] < 0.0001* 0.024* 0.021*

F (df) 20.83 (1–130) 5.06 (1–130) 5.49 (1–130)

*Statistically significant.

Discussion Participants expressed greater facial and smile dissatis-


faction after exposure to the idealised facial images com-
The present study aimed to examine the relationship
pared to those exposed to neutral images, which is
between exposure to ‘ideal’ facial images on social
consistent with previous research (Fardouly et al., 2015).
media with facial and smile dissatisfaction. So far,
Another important finding in our study was that facial,
research has shown significant correlation between
exposure to idealised images on social media and body body and total dissatisfaction reduced in the control group.
dissatisfaction (Holland and Tiggemann, 2016). Our This suggests that viewing neutral nature images helped
study shows a statistically significant increase in facial improve participants’ mood and appearance-related satis-
and smile dissatisfaction following exposure to ‘ideal’ faction. This supports findings that recommend targeting
facial images, but not body dissatisfaction. This is likely specific SNS photo activities and to view more appear-
to be because the chosen user images for the experimen- ance-neutral images to prevent body image concerns
tal group were concentrated on the face, with the ‘ideal’ (Cohen et al., 2017).
smile being the central focus of the images. There were Overall, the present study agreed with the findings of
no images showing the whole body, so individuals were Meier and Gray (2014) and Cohen et al. (2018) that post-
unaffected in terms of their body satisfaction. This result exposure body dissatisfaction was more related to viewing
was apparent in both men and women. photographs on social media than the total time spent on
Sampson et al. 61

Figure 2.  Graph to show relationship between Facial Satisfaction Score (FSS) and Body Satisfaction Score (BSS) before and after
exposure in the experimental (smile images) group vs. the control (neutral nature images) group.

social media. There was no correlation between increased patients’ motivations to seek treatment (to gain confidence
use of social media and greater smile or facial dissatisfac- following comparisons on social media) and the impact on
tion in men and women, suggesting that any use of social their mental wellbeing (depressive states as a result of
media, not just increased use, can have an impact on facial social media).
and body dissatisfaction.
Studies have shown that adults who have had orthodon-
tic treatment have improved self-esteem and quality of life
Limitations
(de Couto Nascimento et al., 2016; Imani et al., 2018). This Our study focused on one SNS with one of the highest lev-
purported increase in self-esteem was shown in our study els of visual media. There are, however, a number of other
not to correlate in resistance to post-exposure facial or body SNSs that could be incorporated into future research, par-
dissatisfaction. Wearing braces or having had braces was ticularly when 73% of Americans actively use more than
shown not to influence post-exposure facial or body one SNS (Smith and Anderson, 2018).
dissatisfaction. Research shows distinct patterns in socioeconomic sta-
Most studies on social media effects on body dissatisfac- tus and choice of SNS (Lenhart, 2015). Users from less
tion are targeted at women. The present study shows that affluent families (< $30,000 per annum) are twice as likely
gender has no influence on the effect of social media on to regularly use Facebook than wealthier peers (> $75,000
facial or body satisfaction. Both genders are equally sus- per annum), who prefer Instagram and Snapchat (SNSs
ceptible to the effects of social media. This contrasts with higher in visual media).
other findings (Haferkamp and Kramer, 2011) but could be Differences in education were not taken into considera-
explained by evolving attitudes towards personal appear- tion in this study, as all participants were university stu-
ance and social comparison across genders, as found by dents. Smith and Anderson (2018) showed that college
Tamplin et al. (2018). graduates are more likely to use Instagram than those with
Individuals with higher self-discrepancy scores have an education level of ‘high school or less’. These differ-
been found to be more vulnerable to the adverse effects of ences in socioeconomic and educational backgrounds may
social media on body, smile and facial dissatisfaction and have relevance for future research for a truer representative
have greater propensity to make appearance-related social sample of young adults.
comparisons. Our study shows that individuals with high
self-discrepancy scores are less satisfied with their facial
features and body appearance.
Generalisability
With the growing use of social media, clinicians must be The study cohort was representative of a young UK-based
made aware of the impact this may have on patients. This student population, a population likely to be using SNSs
information may help develop a better understanding of high in visual media.
62 Journal of Orthodontics 47(1)

Table 4.  Results of multiple regression for Body Satisfaction Scale (BSS), Facial Satisfaction Scale (FSS) and total scores.

Predictors Compared with Effect 95% CI P value


FSS
  Baseline 0.95 0.85–1.05 < 0.0001*
  Gender  
   Female Male –0.52 –1.98 to 0.95 0.49
  Group  
    Control group Experimental group –2.39 –4.32 to –0.46 0.016*
  Braces  
    Braces – No Braces – Yes 0.63 –0.76 to 2.01 0.37
  Media use  
   High Low 1.15 –0.71 to 3.02 0.22
  SDI 0.60 0.04–1.16 0.036*
  Group X Media use –0.59 –3.37 to 2.19 0.67
BSS
  Baseline 0.47 0.32–0.62 < 0.0001*
  Gender  
   Female Male 0.98 –1.24 to 3.21 0.38
  Group  
    Control group Experimental group –2.74 –5.66 to 0.19 0.07
  Braces  
    Braces – No Braces – Yes –0.40 –2.50 to 1.71 0.71
  Media use  
   High Low 0.29 –2.53 to 3.12 0.84
  SDI 0.97 0.12–1.82 0.025*
  Group X Media use –0.51 –4.73 to 3.71 0.81
Total scores
  Baseline 0.94 0.86–1.01 < 0.0001*
  Gender  
   Female Male 0.27 –1.70 to 2.24 0.78
  Group  
    Control group Experimental group –3.49 –6.10 to –0.87 0.009*
  Braces  
    Braces – No Braces – Yes 0.58 –1.28 to 2.44 0.54
  Media use  
   High Low 1.08 –1.43 to 3.58 0.40
  SDI 0.63 –0.14 to 1.39 0.11
  Group X Media use –1.04 –4.78 to 2.69 0.58

*Statistically significant.
CI, confidence interval.

Conclusion this trial, viewing SNSs with high visual media signifi-
cantly reduces satisfaction with facial appearance in the
This study adds to the growing body of evidence showing short term. Individuals with high levels of self-discrepancy
the detrimental effects of SNSs on self-image. Based on have a more profound effect.
Sampson et al. 63

Declaration of conflicting interests meta-analytic review. International Journal of Eating Disorders


31: 1–16.
The author(s) declared no potential conflicts of interest with Haahr M (2018) RANDOM.ORG: True Random Number Service.
respect to the research, authorship, and/or publication of this Available at: https://www.random.org (accessed 8 September 2017).
article. Haferkamp N and Kramer NC (2011) Social comparison 2.0: exam-
ining the effects of online profiles on social-networking sites.
Funding Cyberpsychology, Behavior and Social Networking 14: 309–314.
Heinberg LJ and Thompson JK (1995) Body Image and Televised Images
The author(s) received no financial support for the research, of Thinness and Attractiveness: A Controlled Laboratory Investigation.
authorship, and/or publication of this article. Journal of Social and Clinical Psychology 14: 325–338.
Higgins ET, Klein R and Strauman T (1985) Self-concept discrepancy
theory: A psychological model for distinguishing among different
ORCID iD aspects of depression and anxiety. Social Cognition 3: 51–76.
Ariane Sampson https://orcid.org/0000-0002-5559-6185 Holland G and Tiggemann M (2016) A systematic review of the impact
of the use of social networking sites on body image and disordered
eating outcomes. Body Image 17: 100–110.
References Imani MM, Jalali A, Dinomohammadi M and Nouri P (2018) The Effect
Brown Z and Tiggeman M (2016) Attractive celebrity and peer images on of Orthodontic Intervention on Mental Health and Body Image. Open
Instagram: Effect on women’s mood and body image. Body Image Access Macedonian Journal of Medical Sciences 6: 1132–1137.
19: 37–43. Internet Live Stats, part of Real Time Statistics Project (2018). Available at:
Cafri G, Thompson JK, Ricciardelli L, Mccabe M, Smolak L and Yesalis http://www.internetlivestats.com/one-second/#instagram-band (accessed
C (2005) Pursuit of the muscular ideal: Physical and psychological 10 January 2018).
consequences and putative risk factors. Clinical Psychology Review Jang JY, Han K, Lee D, Jia H and Shih PC (2016) Teens Engage More with
25: 215–239. Fewer Photos: Temporal and Comparative Analysis on Behaviors in
Cattell RB (1973) Personality and mood by questionnaire. San Francisco, Instagram. Proceedings of the 27th ACM Conference on Hypertext
CA: Jossey-Bass Publishers. and Social Media, July 10–13, 2016. Halifax, Nova Scotia, Canada.
Cohen R, Newton-John T and Slater A (2017) The relationship between Keery H, van den Berg P and Thompson JK (2004) An Evaluation of the
Facebook and Instagram appearance-focused activities and body Tripartite Influence Model of body dissatisfaction and eating distur-
image concerns in young women. Body Image 23: 183–187. bance with adolescent girls. Body Image 1: 237–251.
Cohen R, Newton-John T and Slater A (2018) ‘Selfie’-objectification: The Kokich VO Jr, Kiyak HA and Shapiro PA (1999). Comparing the percep-
role of selfies in self-objectification and disordered eating in young tion of dentists and lay people to altered dental esthetics. Journal of
women. Computers in Human Behavior 79: 68–74. Esthetic Dentistry 11: 311–324.
De Couto Nascimento V, De Castro Ferreira Conti AC, De Almeida Lenhart A (2015) Teen, Social Media & Technology Overview 2015.
Cardoso M, Valarelli DP and De Almeida-Pedrin RR (2016) Pew Research Center. Available at: https://www.pewresearch.org/­
Impact of orthodontic treatment on self-esteem and quality of life internet/2015/04/09/teens-social-media-technology-2015/.
of adult patients requiring oral rehabilitation. Angle Orthodontist Leone C (2018) Which social media sites get the most engagement?
86: 839–845. Quintly. Available at: https://www.webstrategiesinc.com/blog/which-
Dittmar H, Beattie J and Friese S (1996) Objects, decision considera- social-media-sites-get-the-most-engagement (accessed 6 July 2018).
tions and self-image in men’s and women’s impulse purchases. Acta Mabe AG, Forney KJ and Keel PK (2014) Do you “like” my photo?
Psychologica 93: 187–206. Facebook use maintains eating disorder risk. International Journal of
Duggan M and Brenner J (2013) The Demographics of Social Media Eating Disorders 47: 516–523.
Users. Pew Research Center. Available at: https://www.pewresearch. Meier EP and Gray J (2014) Facebook photo activity associated with body
org/internet/fact-sheet/social-media/. image disturbance in adolescent girls. Cyberpsychology, Behavior
eMarketer (2018) eMarketer Updates Worldwide Social Network User and Social Networking 17: 199–206.
Figures. eMarketer. Available at: https://www.emarketer.com/ Myers TA and Crowther JH (2009) Social comparison as a predictor of
Article/eMarketer-Updates-Worldwide-Social-Network-User-Fig- body dissatisfaction: A meta-analytic review. Journal of Abnormal
ures/1016178 (accessed 17 July 2018). Psychology 118: 683–698.
Fardouly J, Diedrichs PC, Vartanian LR and Halliwell E (2015) Social Norman G, Monteiro S and Salama S (2012). Sample size calculations:
comparisons on social media: The impact of Facebook on young should the emperor’s clothes be off the peg or made to measure? BMJ
women’s body image concerns and mood. Body Image 13: 38–45. 345: e5278.
Fardouly J and Vartanian LR (2015) Negative comparisons about one’s Office for National Statistics (2015) Harmonised Concepts and Questions
appearance mediate the relationship between Facebook usage and for Social Data Sources Primary Principles Ethnic Group by the
body image concerns. Body Image 12: 82–88. Government Statistical Service version 3.3. Available at: https://gss.
Festinger L (1954). A theory of social comparison processes. Human civilservice.gov.uk/wp-content/uploads/2016/03/P3-Ethnic-Group-
Relations 7: 117–140. June-16-1.pdf.
Freud S (1923). The Ego and the Id. In: Strachey J, et al. (eds) The Rosen LD, Whaling K, Carrier LM, Cheever NA and Rokkum J (2013)
Standard Edition of the Complete Psychological Works of Sigmund The Media and Technology Usage and Attitudes Scale: An empirical
Freud. Volume 7th. London: Hogarth Press. investigation. Computers in Human Behavior 29: 2501–2511.
Gilbert P, Price J and Allan S (1995) Social comparison, social attrac- Rosenberg M (1979) Conceiving the self. New York: Basic Books.
tiveness and evolution: How may they be related? New Ideas in Saunders JF and Eaton AA (2018). Snaps, Selfies, and Shares: How Three
Psychology 13: 149–165. Popular Social Media Platforms Contribute to the Sociocultural Model
Grabe S, Ward LM and Hyde JS (2008) The role of the media in body of Disordered Eating Among Young Women. Cyberpsychology,
image concerns among women: a meta-analysis of experimental and Behavior and Social Networking 21: 343–354.
correlational studies. Psychological Bulletin 134: 460–476. Slade PD, Dewey ME, Newton T, Brodie D and Kiemle G (1990)
Groesz LM, Levine MP and Murnen SK (2002) The effect of experi- Development and preliminary validation of the body satisfaction scale
mental presentation of thin media images on body satisfaction: A (BSS). Psychology & Health 4: 213–220.
64 Journal of Orthodontics 47(1)

Smith A and Anderson M (2018). Social Media Use in 2018. Pew Tiggemann M, Polivy J and Hargreaves D (2009) The processing of
Research Center. Available at: https://www.pewresearch.org/­internet/ thin ideals in fashion magazines: A source of social compari-
2018/03/01/social-media-use-in-2018/. son or fantasy? Journal of Social and Clinical Psychology 28:
Tamplin NC, Mclean SA and Paxton SJ (2018) Social media literacy 73–93.
protects against the negative impact of exposure to appearance ideal Turner PG and Lefevre CE (2017) Instagram use is linked to increased
social media images in young adult women but not men. Body Image symptoms of orthorexia nervosa. Eating and Weight Disorders 22:
26: 29–37. 277–284.
Tiggemann M and McGill B (2004) The role of social comparison van den Berg P, Paxton SJ, Keery H, Wall M, Guo J and Neumark-Sztainer
in the effect of magazine advertisements on women’s mood and D (2007). Body dissatisfaction and body comparison with media
body dissatisfaction. Journal of Social and Clinical Psychology images in males and females. Body Image 4: 257–268.
23: 23–44. van den Berg P, Thompson JK, Obremski-Brandon K and Coovert M
Tiggemann M and Miller J (2010) The Internet and Adolescent Girls’ (2002) The Tripartite Influence model of body image and eating dis-
Weight Satisfaction and Drive for Thinness. Sex Roles 63: 79–90. turbance: a covariance structure modeling investigation testing the
Tiggemann M and Slater A (2013a) NetGirls: the Internet, Facebook, mediational role of appearance comparison. Journal of Psychosomatic
and body image concern in adolescent girls. International Journal of Research 53: 1007–1020.
Eating Disorders 46: 630–633. Vartanian LR and Dey S (2013) Self-concept clarity, thin-ideal internaliza-
Tiggemann M and Slater A (2013b) NetTweens. Journal of Early tion, and appearance-related social comparison as predictors of body
Adolescence 34: 606–620. dissatisfaction. Body Image 10: 495–500.
Tiggemann M and Zaccardo M (2015). “Exercise to be fit, not skinny”: Witt M and Flores-Mir C (2011) Laypeople’s preferences regarding frontal
The effect of fitspiration imagery on women’s body image. Body dentofacial esthetics: tooth-related factors. Journal of the American
Image 15: 61–67. Dental Association 142: 635–645.

You might also like