Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.

13140 Mattioli 1885

Health Professions (Editors: Leopoldo Sarli, Giovanna Artioli)

Factors associated with eating disorders in adolescents: a

systematic review
Candy Laurine Suarez-Albor1, Maura Galletta2, Edna Margarita Gómez-Bustamante3
1
Faculty of Nursing, Universidad Popular del Cesar, Valledupar, Colombia; 2Department of Medical Sciences and Public
Health, University of Cagliari, Monserrato, Italy; 3Faculty of Nursing, University of Cartagena, Cartagena de Indias, Colombia

Abstract. Background and aim: The World Health Organization has placed eating disorders among the prior-
ity mental illnesses for children and adolescents given the risk they imply for their health. Recognizing the
risk factors associated with this problem can serve as the basis for the design of timely and effective interven-
tions. The objective of the study was to identify the factors associated with eating behavior in adolescents
through a systematic review. Methods: Systematic review. Search of the literature in the bibliographic sources
CINAHL, CUIDEN, Pubmed, Dialnet, SCIELO and Science Direct. The search was conducted in October
and November 2020. The search terms were Eating Disorders, Food Intake, and Adolescents. The evaluation
of the methodological quality was carried out using a specific guide for observational epidemiological stud-
ies. A narrative synthesis of the findings was made. Additionally, the vote counting and sign test technique
was applied. Results: 25 studies were selected. The associated factors were body dissatisfaction, female gender,
depression, low self-esteem, higher BMI that increases the risk of eating disorders. Conclusions: a high impact
of psychological factors was observed. These should be considered in the design of effective interventions to
prevent this disease, although the search needs to be broadened to identify larger and more complex studies
that allow for a more comprehensive review. (www.actabiomedica.it)

Key words: adolescent, associated factors, eating behaviour, eating disorder.

Introduction and Spain 6.2% (2); Colombia is followed by 4.5% (3), the
United Kingdom 3.7% (2) and Portugal 3.06% (2). Coun-
Eating disorders (EDs) are complex and multifacto- tries such as the United States, Italy, Costa Rica, Mexico,
rial pathologies that affect physical and mental health and Honduras, Venezuela, have numbers between 0.5% -1.5%
are life threatening. They are characterized by an excessive (4, 5). Most of these disorders are more common in wom-
preoccupation with the weight and shape of the body or en and begin in adolescence, a stage of change where body
a frank deviation of the body image, accompanied by vol- image is consolidated. This in turn generates numerous
untary restriction of the intake or the presence of episodes crises of identity, physical appearance, friendly or sexual
of binge eating that cause great suffering, impairment of requirements and a struggle for autonomy, traits of perfec-
health and quality of life (1). The prevalence of eating dis- tionism and self-demand that can lead to low self-esteem,
orders is variable; in the last two decades several studies dependence on the environment, difficulty in expressing
have been carried out, especially by the National Institute emotions or expressing aggressiveness (6, 7).
of Mental Health of the United States, which has com- The World Health Organization (WHO) has
piled cases even from European countries. The countries placed eating disorders among the priority mental ill-
with the highest cases are Switzerland 12%, Chile 8.3% nesses for children and adolescents given the risk they
2 Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140

imply for their health and the great psychiatric comor- review, experimental, intervention, or treatment studies
bidity (8). Among the most frequent, depressive disorders were excluded, as well as studies with a mixed sample
23.3%, anxiety disorders 10%, adaptive disorders 3.3% (children, adolescents, adults), and investigations with-
and negative perception of family relationships 43.3%, out statistical information of association.
which aggravate the problem and cause important com-
plications in the state of health (7, 9). For this reason, Article selection and evaluation of methodological quality
EDs have become more relevant for the interest in the
clinic, research and epidemiology (9). Various factors in- The selection of the articles was carried out in 4
tervene in the occurrence of eating disorders and show a phases. First, title and abstract were read to determine
higher attributable risk such as biological, psychological, the suitability of the study and elimination of duplicates.
family and sociocultural (2-7). Thus, scientific evidence is Second, full text was read and the inclusion and exclu-
abundant when addressing various aspects of eating dis- sion criteria were applied. Third, a reverse and forward
orders, however the state of the art revealed that in the search was performed on the included studies to locate
last five years no literature review have been published on as many documents as possible. Fourth, the risk of bias
the subject, which is relevant to design or guide effective was assessed through critical reading based on the Criti-
interventions that allow professionals to prevent these cal Reading Guide for Observational Studies in Epide-
events. In this sense, the aim of this work was to carry out miology (11, 12). A guide to assess cross-sectional stud-
an exhaustive review of the published evidence about the ies was used (11). This instrument included 31 items
factors associated with eating behaviour in adolescents. that allow for minimizing biases and the confounding
effect of internal validity. It was evaluated qualitatively
using MB: very good, B: good, A: regular, and NI: does
Methods not report. A second guide was used to assess cohort
studies and case-control studies (12). The instrument
A systematic review was carried out according to included 21 items and evaluated qualitatively the fol-
the guidelines of the PRISMA (10) statement, in the lowings aspects: selection of subjects, validation of ques-
bibliographic sources LILACS, CUIDEN, Pubmed, Di- tion, evaluation of the final outcomes, confounding fac-
alnet, SCIELO and Science Direct and MEDES. The tors, statistical analysis, general evaluation of the study,
search was carried out in October and November 2020. and description of the study, using A: adequately, B:
The search terms to be used were consulted in the DECS partially, C: improperly, and D: I don’t know. This pro-
and MESH libraries, to guarantee their standardization, cess was carried out by the first author and was audited
in English and Spanish, they were conjugated in search by the other authors.
equations with the Boolean operators AND and OR
thus: AND factors (Eating Disorders OR Food Intake Data extraction
OR Eating Behavior) AND adolescent.
The data were consolidated through a structured
Inclusion and exclusion criteria booklet in Excel based on two types of information: (i)
information about articles’ characteristics such as study
Articles were selected from cohort, cross-sectional, sample, main author, year of publication, language,
and case control studies about factors associated with country, design; (ii) information about eating disorder
EDs in adolescents. The inclusion criteria were (a) free risk factors such as biological, psychological, sociocul-
access articles in full text, (b) primary studies published tural, and family factors.
between 2009 and 2020 to ensure that as many neces-
sary and relevant studies as possible have been included Data analysis
in the review, (c) studies with a sample of adolescents
aged from 10 to 19 years, according to the classification The information was treated qualitatively and ana-
provided by the WHO (9). Dissertation, meta-analysis, lysed in a narrative way. The results were organized in
Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140 3

tables and figures according to the PRISMA statement. Results

Additionally, the found results exceeded the number of
20 articles, so the vote counting technique was applied. Methodological quality assessment
Such a technique consisted in granting a positive vote
for studies with a statistically significant relationship About the cross-sectional studies, 54.5% (n = 12)
between a risk factor and EDs, and a negative vote when obtained high methodological quality, 27.2% (n = 7)
there was no significant association. Subsequently, the medium quality, and 4.6% (n = 1) low methodological
sign test (13, 14) was applied to determine if the differ- quality. This study was excluded by the revision. With
ence in the number of positive studies was significantly regard to the cohort studies, it was assumed that stud-
greater than the opposite result. A significance value ies with adequate rating in 23-26 items were consid-
was established to be less than 0.05. It is important to ered to be of high methodological level; medium level
notice that these techniques are limited but they can was attributed to studies with adequate rating in 19-22
help to guide the results of the review in the absence of items, and low methodological level was attributed to
meta-analysis (14). studies with adequate rating in 18 items or less. In this
sense, 100% (n = 3) of the cohort studies obtained a
medium methodological quality (Tab. 1 and Tab. 2).

Fig 1. General diagram of the study

Tab 1. Critical reading and assessment of methodological quality for cross-sectional studies. 4
Statistical analysis
Participants
and confusion
(sample is adequate and similar to the
(analysis is adequate and the Summary assessment
general population, minimizing the
possibility of confusion is
Main author, year and probability of selection bias)
Study type minimized)
country
Internal validity Overall study quality
(study design allows (quality of the
2 3 4 5 6 15 16 17 18
minimizing biases and the evidence provided
confounding effect by the study):
Esteban, Et al, 2014,
Cross-sectional B MB MB NA B MB B B B high high
Spain (15)
Shahyad, Et al, 2018,
Cross-sectional MB B B MB R B B NA B high medium
Israel (16)
Yirga, Et al, 2016,
Cross-sectional MB MB MB MB B B B B B high high
Ethiopia (17)
Altamirano, Et al, 2011,
Cross-sectional MB MB MB MB B B MB R B high high
Mexico (18)
Fuentes, Et al, 2015,
Cross-sectional MB MB MB MB B MB B B B high high
Spain (19)
Lazo, Et al, 2015, Analytical
MB MB MB MB B MB MB MB B high high
Peru (20) Cross-Sectional
Moreno, Et al, 2017, Descriptive
B B MB B B B B B B high high
Colombia (21) Cross-sectional
Nuño, Et al, 2009, Analytical
B B MB B B B MB MB MB high high
Mexico (22) Cross-Sectional
Quiles, Et al, 2014,
Cross-sectional MB MB MB MB MB MB MB MB MB high high
Spain (2. 3)
Silva, Et al, 2017,
Cross-sectional MB MB MB MB MB B B B B high high
Mexico (24)
Sousa, Et al, 2013,
Cross-sectional B B MB MB MB MB MB B B high high
Acta Biomed 2022; Vol. 93, N. 3: e2022253

Brazil (25)
Cross-Sectional
Cogollo, Et al, 2012,
Analytical B B MB MB MB MB MB B B high high
Colombia (26)
Observational
Caldera, Et al, 2019,
Cross-sectional B B B B B B B NA B medium medium
Mexico (27)
Reina, Et al, 2013,
Cross-sectional B B R NA B B B NA B medium medium
USA (28)
DOI: 10.23750/abm.v93i3.13140
Tab 1. Critical reading and assessment of methodological quality for cross-sectional studies.
Statistical analysis
Participants
and confusion
(sample is adequate and similar to the
(analysis is adequate and the Summary assessment
general population, minimizing the
possibility of confusion is
Main author, year and probability of selection bias)
Study type minimized)
country
Internal validity Overall study quality
(study design allows (quality of the
2 3 4 5 6 15 16 17 18
minimizing biases and the evidence provided
confounding effect by the study):
Acta Biomed 2022; Vol. 93, N. 3: e2022253

Cross-sectional
Laporta, Et al, 2020, Quantitative,
B MB B B B B B B NI medium medium
Spain (29) Descriptive,
Retrospective
Vara, Et al, 2011,
Cross-sectional B B B B R R B B B medium medium
Spain (30)
Sousa, Et al, 2014,
Cross-sectional B B MB B B R B R R medium medium
Spain (31)
Castaño, Et al, 2012,
Cross-sectional B NI B B R B B R B medium medium
Colombia (32)
DOI: 10.23750/abm.v93i3.13140

Rutsztein, Et al, 2014, Cross-sectional,

B R B R R R B R B Low Low
Argentina (33) Descriptive
Note. Internal validity. It defines whether the study design allows minimizing biases and the confounding effect (12).
The Items used were:
2. The inclusion and exclusion criteria of participants are indicated, as well as the sources and selection methods.
3. The selection criteria are adequate to answer the question or the objective of the study.
4. The study population, defined by the selection criteria, contains an adequate spectrum of the population of interest.
5. An estimate was made of the size, the level of confidence or the statistical power of the sample to estimate the measures of frequency or association that the study intended to obtain.
6. The number of potentially eligible people is reported, those initially selected, those who accept and those who finally participate or respond; fifteen. Statistical analysis was
determined from the beginning of the study.
16. The statistical tests used are specified and appropriate
17. Participant losses, lost data or others were correctly treated
18. The main possible confounding elements were taken into account in the design and in the analysis.
Assessment: MB=very good; B=good; R=regular; NA=not applicable; NI=no information.
5
6 Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140

Tab 2. Critical reading and assessment of methodological quality for cohort studies.
Main author, Study Question Selection Evaluation Confounding Statistic Overall Study Summary
year and country type validity of subjects factors analysis rating description assessment
of the
Medium
study
Haynos, Et al, 2016,
Cohort A A A A A B A Yes
Spain (3, 4)
Maezono, Et al, 2019,
Cohort D B B A A B B Yes
Japan / Finland (35)
Batista, Et al, 2018,
Cohort B A B A A B B Yes
Croatia (36)
Note. Cohort studies allow a direct determination of relative risk and allow calculation of the interval between exposure or risk factor
and overall study disease (12). It was scored according to the validity of the question, selection of subjects, evaluation, confounding
factors, statistical analysis, general assessment and description of the study. Rating: according to the author, the items were rated
as follows: A: adequately; B: partially; C: improperly; D: I don’t know. For the purposes of this review, it was assumed that studies
with adequate rating in 23-26 items were considered to be of high methodological level; medium level was attributed to studies with
adequate rating in 19-22 items, and low methodological level was attributed to studies with adequate rating in 18 items or less.

lation, and ineffectiveness. About 14% of the studies

Characteristics of the studies analysed sociocultural factors such as alcohol use-
related problematics, internalization of the thinness
Among the selected studies, 86.4% (n = 19) were ideal, influence of media, ridicule related to weight,
cross-sectional design (15─33), 13.6% (n = 3) were and being an immigrant adolescent. Also, 7.1% of the
cohort studies (34─36). Fifty-percent (n = 11) of the studies analysed family factors such as authoritarian
studies were conducted in Latin America (18, 20─22, family style, family functioning, poor communication,
24─27, 31─33), 31.8% (n = 7) in Europe (15, 19, 23, 29, and family care. Lastly, 7.1% of the studies analysed
30, 33, 34), 9% (n = 2) in Asia (16, 34) 4.6% (n = 1) in biological factors such as being female. The complete
Africa (17), and 4,6% (n = 1) in North America (28) description of the factors is summarized in Tab. 3.
(Tab. 3).
Analysis of vote counting and sign test
Factors associated with eating disorders
It was found that there was a greater number of
Different instruments were used to measure the studies that reported statistically significant relation-
factors associated with eating disorders. In 14.2% (n ships between factors such as body dissatisfaction, fe-
= 7) of the analysed studies, authors used the Eat- male gender, depression, low self-esteem, and higher
ing Attitude Test (EAT-26) (17,18,20,21,25,30,31), body mass index (BMI) with eating disorders (Tab. 4).
11% (n = 5) the Body Shape Questionnaire (BSQ) In this sense, adolescents with those risk factors are
(21,25,27,29,31), and 9.5% (n = 4) the Sociocultural more likely developing eating disorders.
Attitudes Questionnaire towards appearance-3 (SA-
TAQ-3) (20, 28, 32, 36). About 93% (n = 14) of the
studies analysed risk factors and about 7% (n = 1) ana- Discussion
lysed correct self-image and hours of practiced sport as
protective factors (B = 0.11; p = 0.047) (30). Regarding In the present review, the results show that the
the risk factors, psychological risks were the most fre- main factors associated with eating disorders were psy-
quently analysed by the studies (71%). They included chological-type with a prevalence of the factor inher-
dissatisfaction with body image, low self-esteem, high ent the dissatisfaction with body image (16─18, 21, 25,
depression, high perfectionism, stress, impulsivity, per- 27, 29, 31, 32, 35). Literature refers that dissatisfaction
sonal and interpersonal insecurity, emotional dysregu- with body image increases significantly in adolescence
Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140 7

Tab 3. Synthesis of the studies included in the review.

Main Author, Type of Population Instrument Family Factors Biological Sociocultural Psychological
Year and study factors factors factors
Country
Esteban, Cross- 2,077 native Spanish SCOFF Eating Does not Does not Risk: immigrant Does not inform
Et al, 2014, sectional and immigrant Disorders inform inform adolescents
Spain. (15) subjects from 13 to 17 Questionnaire. living in the
years old. Madrid region
and immigrant
women.
Shahyad, Cross- 477 high school Inventory of Does not Does not Risk: Thin ideal Risk: Body
Et al, 2018, sectional
students aged 15 and eating disorders. inform inform internalization dissatisfaction
Israel. (16) 17.
Yirga, Cross- 836 high school Eating Attitude Risk: Risk: Being Does not Does not inform
Et al, 2016, sectional students between the Test-26 educational level a woman. inform
Ethiopia. (17) ages of 12 and 19. (EAT-26) of the mother
Altamirano, Cross- 1,982 women Brief Does not Does not Does not Risk: dissatisfaction
Et al, 2011, sectional between the ages of Questionnaire inform inform inform with body image and
Mexico.CO 15 and 19 of Risky Eating low self-esteem.
(18) Behaviors
(CBCAR)
Fuentes, Cross- 368 between 13 and Body Image Risk: Does not Does not Does not inform
Et al, 2015, sectional 17 years Dissatisfaction Authoritarian inform inform
Spain. (19) Assessment Scale. and negligent
family styles,
family
socialization
styles and
dissatisfaction
with body
image.
Lazo, Analytical 483 female students Eating attitude Does not Does not Risk: influence Does not inform
Et al, 2015, cross- between 12 and 17 test (EAT-26). inform inform of the media.
Peru. (20) sectional years old.
Moreno, Cross- 104 students between Abbreviated Risk: Parental Risk: female Does not Risk: Dissatisfaction
Et al, 2017, sectional 13 to 18 years old Eating Attitudes educational gender inform with adolescent body
Colombia. (21) correlation Test (EAT-26). levels image and concern
Eating behavior about weight.
questionnaire
(FBQ).
Nuño, Et al, Analytical 1,134 male and Brief Does not Risk: being Does not Risk: impulsivity,
2009, cross- female adolescents. Questionnaire inform a woman inform suicidal ideation and
Mexico. (22) sectional of Risky Eating stress
Behaviors
Quiles, Cross- 2,142 male and Eating Attitude Does not Does not Does not Risk: self-oriented
Et al, 2014, sectional female adolescents. Test (EAT-40). inform inform inform perfectionism.
Spain. (23)
Silva, Et al, Cross- 392 women between Eating Attitudes Does not Does not Risk: belonging Risk: submission
2017, Mexico. sectional the ages of 13 and 18. Test (EAT-40) inform inform to the
(24) municipality of
Pungarabato
8 Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140

Tab 3. Synthesis of the studies included in the review.

Main Author, Type of Population Instrument Family Factors Biological Sociocultural Psychological
Year and study factors factors factors
Country
Sousa, Et al, Cross- 580 adolescents of Food Does not Risk: fat Does not Risk: dissatisfaction
2013, Brazil. sectional both sexes from 10 to Attitudes Test inform percentage inform with body image
(25) 19 years Questionnaire
(COMER-26)
The EAT-26.
Body shape quiz
Cogollo, Analytical 2625 students SCOFF Does not Risk: female Does not Risk: clinically
Et al, 2012, cross- between 10 and 20 questionnaire. inform inform important depressive
Colombia. (26) sectional years old symptoms and
problematic alcohol
use.
Caldera, Et al, Cross- 988 adolescents of Brief Does not Does not Does not Risk: Body
2019, Mexico. sectional both sexes between Questionnaire inform inform inform dissatisfaction.
(27) 14 and 18 years old. of Risky Eating
Behaviors
(CBCAR).
Reina, Et al, Cross- 90 adolescents from Infant Feeding Does not Risk: being Does not Risk: Orientation
2013, USA sectional 13 to 17 years old Questionnaire inform a woman inform to Appearance,
(28) (CFQ) concern about being
overweight and
eating in the absence
of hunger.
Laporta, Et al, Descriptive 100 patients Eating Disorders Does not Risk: being Does not Risk: High
2020, Spain. Cross- diagnosed with eating Inventory-3, inform a woman inform perfectionism,
(29) sectional disorders according to EDI-3. greater severe
DSM-IV-TR, aged depressive symptoms,
between 13 and 16 body dissatisfaction
years. and lower self-
esteem.
Vara, Et al, Cross 158 adolescents of Attitude test Does not Risk: Protector: Does not inform
2011, Spain. both sexes. towards eating inform increased correct self-
(30) (EAT-26) BMI. image and
hours of sport
practiced
Sousa, Et al, Cross 562 adolescents Eating Attitudes Does not Risk: Does not Risk: body
2014, Spain. between 10 and 15 Test (EAT-26). inform increased inform dissatisfaction in
(31) years old BMI women, the degree
of psychological
commitment to
exercise.
Castaño, Cross 70 adolescents with Eating Disorders Does not Risk: Risk: Does not inform
Et al, 2012, anorexia aged 11 to Inventory-3 inform increased Internalization
Colombia. (32) 19 years (EDI-3) BMI of the slim ideal.
Haynos, 2016, Longitudinal Time I: 4,746 EAT Project Risk: Family Does not Risk: Weight- Risk: High
Spain (34) cohort students between Survey communication inform related teasing depression and low
1998-1999 from 11 and poorer care self-esteem.
to 18 years old. Time
II: 2,516 students
between 2003-2004
Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140 9

Tab 3. Synthesis of the studies included in the review.

Main Author, Type of Population Instrument Family Factors Biological Sociocultural Psychological
Year and study factors factors factors
Country
Maezono, Et Cohort 1,840 Japanese Scale developed by Does not Does not Does not Risk: Dissatisfaction
al, 2019, Japan students (2011) Koskelainen, Sour inform inform inform and concern with
/ Finland. (35) and 1,135 Finnish Ander & Helenius their bodies in
students (2014) 13-15 Japanese and Finnish
years old. women and food
distress in Finnish
women.
Batista, Et al, Cohort 35 women with Eating Disorders Does not Does not Risk: Risk: interpersonal
2018, Croatia. anorexia nervosa and Inventory-3 inform inform Internalization problems, affective
(36) 35 healthy between (EDI-3). of the slim ideal. problems and
12-18 years. excess control,
Low Self-esteem,
Personal alienation,
Interpersonal
insecurity,
Interpersonal
alienation,
Interoceptive
deficits, Emotional
dysregulation,
Perfectionism and
asceticism in women
with anorexia.

Tab 4. Analysis by vote counting and sign test. namely the higher the BMI, the higher the body dis-
Vote Counting and Sign Test
satisfaction (25, 31─33, 36). This association is more
recurrent in female gender (17, 22, 26, 28, 30) as girls
Risk factor variables Positive Negative P value n = 22
generally show greater instability of self-image, lower
Body dissatisfaction 10 0 0.4159 10
self-esteem and general dissatisfaction with their body,
Female gender 5 0 0.0085 5 if compared to boys. In most studies, the sample stud-
Depression 5 0 0.0085 5 ied was female (29, 33, 34, 36). Other psychological
lower self-esteem 3 2 0.0004 5 factors were emerged from the review. They were: ap-
Higher BMI 1 4 0.0022 5 pearance orientation (28), high level of perfectionism
(23, 29), low self-esteem (18, 29, 33, 34, 36), impulsiv-
due to environmental pressures like media (e.g., tel- ity (22), stress, suicidal idea and depression (22, 26,
evision, social networks, virtual and written press) (20, 27, 29, 34), eat in the absence of hunger (28), concern
28). They represent channels of transmission of the about being overweight, submission (24), personal and
current body aesthetic model and have a positive or interpersonal insecurity (29, 36), and emotional dys-
negative impact on an adolescent’s body image. This regulation (33). A teenager with low self-esteem shows
is more common in women, as it was the biological a negative attitude and evaluation towards himself. In
factor reported in this review. However, the findings fact, low self-esteem has been repeatedly considered
are consistent with other studies where dissatisfaction as a relevant factor of vulnerability for the develop-
with body image occurs more frequently in females ment of EDs. This evidence is supported by a previous
and is positively associated with BMI as a predictor review (39). It is also important to identify depressive
of eating disorders. (37, 38). Similarly, BMI appears and anxiety manifestations that have an impact on
directly related to dissatisfaction with one’s own body, food restriction and concerns about figure and weight.
10 Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140

The number of studies that supported the relationship and intensified and do not necessarily occur sponta-
between psychological factors and eating disorders was neously or at random. In this sense, interventions fo-
statistically significant according to the sign test. cused on strengthening those factors could be effective
Socio-cultural factors were analysed in 14.2% of to prevent eating disorders behaviours. This requires
the selected studies (15, 16, 20, 24, 26, 32, 36). The the development of research that identifies and analy-
most frequently revealed were the internalization ses the protective factors that can be strengthened in
of the thin ideal followed by the influence of media, adolescents (42─44).
weight-related bullying , and immigrant adolescents. Most of the studies included in this systematic
These sociocultural factors and the desire to con- review are cross-sectional and in a lower percentage
form to body aesthetic models promoted by media are cohort studies. Spain is the country that has done
and advertising have a greater likelihood to develop- the most research on the factors associated with eat-
ing perceptions of body dissatisfaction. Moreno (40) ing disorders in adolescents, thus showing a particular
showed a very high relationship between the influence interest in this topic. However, this review has shown
of the media and the presence of eating disorders in that there is a plurality of studies in the scientific com-
the adolescent population. This is a cultural problem munity from different sociocultural contexts. This can
that comes from long ago where the idea that a perfect explain why there is variability of risk factors for eating
body is thin and that this it is accepted by society. The behaviour, although body dissatisfaction is the most
media are very important agents in the transmission common factor emerged from the revision.
of messages about the desire for thinness that is con- The limitations of the review reflect the hetero-
stantly present in eating disorders; the media channel geneity of the study that does not allow to carry out
social pressure to be thin is obviously stronger on fe- a meta-analysis and statistic associations between fac-
males than males (40). tors. Although the vote count and the sign test allow
A few studies analysed the relationship between giving an additional value to the narrative synthesis of
family factors and eating disorders (19, 21, 34). How- the results, they are limited procedures to establish re-
ever, family functioning, poor communication, family liable statistical associations with data. In this sense,
care, and authoritarian styles are factors described in reviews around the subject with quantitative analysis
the literature as predisposing to eating disorders by im- procedures would be necessary.
pacting the way adolescents worry about the amount
of calories in food and obsessed with food and weight
gain. In this sense, parents can play a protective role, Conclusions
but they can also represent a risk factor for their chil-
dren’s eating behaviour, as adolescents regulate their Psychological factors were found to be the main
behaviour according to their parental model from early risk factors directly related to eating disorders in ado-
childhood (40, 41). lescents. The most common were: dissatisfaction with
In this review, we found only one research that body image, depression, low self-esteem and higher
addressed protective factors related to physical exercise BMI. Being a woman was also identified as the most
and correct ideas about body image. This could be due reported biological factor associated with eating disor-
to the fact that research in the last two decades has ders. These risk factors become relevant when guiding
focused on mitigating or controlling risk factors as the the creation of mental health promotion programs for
sole basis for interventions to prevent eating disorders adolescents and the prevention and early detection of
in adolescents. However, protective factors make ado- the eating disorders in adolescents.
lescents less vulnerable to the development of eating
disorders and facilitate the achievement of physical Conflict of Interest: Each author declares that he or she has
no commercial associations (e.g. consultancies, stock ownership,
and mental health, the quality of life of adolescents,
equity interest, patent/licensing arrangement etc.) that might
the development of healthy habits and social welfare. pose a conflict of interest in connection with the submitted ar-
Protective factors are susceptible to being modified ticle.
Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140 11

References 17. Yirga B, Assefa Y, Derso T, et al. Disordered eating atti-

tude and associated factors among high school adolescents
1. N
 ational Institute of Mental Health. Eating disorders: A aged 12-19 years in Addis Ababa, Ethiopia: a cross-sec-
problem that goes beyond food, 2018. tional study. BMC Res Notes 2016; 9(1):503. doi: 10.1186/
2. S
 ánchez A. [Prevalence of eating disorders in six European s13104-016-2318-6.
countries]. Metas Enferm. 2017 Jun; 20(5): 66-73. Spanish. 18. Altamirano Martínez MB, Vizmanos Lamotte B, Unikel
https://doi.org/10.35667/MetasEnf.2019.20.1003081094 Santoncini C. [Continuum of risky eating behaviors in Mexi-
3. F
 ajardo E, Méndez C, Jauregui A. [Prevalence of risk of eat- can adolescents]. Rev Panam Salud Publica 2011 Nov; 30(5):
ing disorders in a population of secondary school students, 401-7. Spanish. doi: 10.1590/s1020-49892011001100001.
Bogotá Colombia]. Revista Med 2017; 25(1): 46-57. Span- 19. Fuentes MC, García O. [Influence of family socialization
ish. https://doi.org/10.18359/rmed.2917 on satisfaction with body image in Spanish adolescents].
4. S
 mink F, Daphne V, Hans H. Epidemiology of Eating Trastornos de la Conducta Alimentaria 2015; 22: 2403-25.
Disorders; Incidence, prevalence and Mortality Rates. Curr Spanish.
Psychiatry Rep 2012; 14: 406-14. DOI: 10.1007 / s11920- 20. Lazo Y, Quenaya A, Tristán P. [Mass media influence and
012-0282-y risk of developing eating disorders in female students from
5. B
 enjet C, Borges G, Medina M, et al. The Mental Health Lima, Peru]. Arch Argent Pediatr 2015; 113(6):519-525.
Survey in Adolescents in Mexico. In: Rodríguez J, Kohn Spanish.
R, Aguilar S, editors. Epidemiology of mental disorders in 21. Londoño-Pérez C, Moreno Ruge AM. [Family and personal
Latin America and the Caribbean. Washington: Pan Amer- predictors of eating disorders in young people]. Anal Psicol
ican Health Organization; 2009. p. 107-14. 2017; 33(2): 235-42. Spanish. https://doi.org/10.6018/ana-
6. O
 rtiz Cuquejo LM, Aguiar C, Samudio Domínguez GC, lesps.33.2.236781
et al. [Eating disorders in adolescents: a growing pathol- 22. Nuño-Gutiérrez BL, Celis-de la Rosa A, Unikel-Santoncini
ogy?] Pediatr (Asunción). 2017 Apr; 44(1): 37-42. Spanish. C. [Prevalence and associated factors related to disordered
https://doi.org/10.18004/ped.2017.abril.37-42 eating in student adolescents of Guadalajara across sex]. Rev
7. G
 aete P V, López C C. [Eating disorders in adoles- Invest Clin 2009; 61(4): 286-93. Spanish.
cents. A comprehensive approach]. Rev Chil Pediatr. 23. Pamies L, Quiles Y. [Perfectionism and risk factors for the
2020 Oct;91(5):784-93. Spanish. doi: 10.32641/rchped. development of eating disorders in Spanish adolescents of
vi91i5.1534. PMID: 33399645. both genders]. Anal Psicol 2014; 30(2): 620-26. Spanish.
8. W
 orld Health Organization. Mental health in adolescents, https://dx.doi.org/10.6018/analesps.30.2.158441.
2020. 24. Silva C, Millán B, González K. [Gender role and eating
9. W
 orld Health Organization. World Report on Adolescent attitudes in adolescents from two different socio-cultural
Development, 2019. contexts: Traditional vs. non-traditional]. Rev Mex de
10. U rrutia G, Bonfill X. PRISMA statement: a proposal to Trastor Aliment 2017; 8(1) 40-8. Spanish. https://doi.
improve the publication for systematic reviews and meta- org/10.1016/j.rmta.2016.12.002.
analyzes: Rev. Medicina Clínica 2010; 1: 507-511. 25. Fortes Lde S, Cipriani FM, Ferreira ME. Risk behaviors
11. C iapponi A. [Guide to critical reading of observational for eating disorder: factors associated in adolescent students.
studies in epidemiology (first part)]. Evid Actual Práct Am- Trends Psychiatry Psychother 2013; 35(4): 279-86. doi:
bul. 2010 Oct-Dec; 13(1): 135-40. Spanish. 10.1590/2237-6089-2012-0055.
12. C iapponi A. [Critical reading guide of observational studies 26. Cogollo Z, Gómez E. [Risk of eating behavior disorders in
in epidemiology (second part)]. Evid Actual Práct Ambul. adolescents from Cartagena, Colombia]. Invest Educ En-
2011 Mar;14: 7-14. Spanish. ferm 2013; 31(3): 450-6. Spanish.
13. P ino R, Frías A, Palomino P. [The quantitative systematic 27. Caldera I, Martin P, Caldera J, et al. [Predictors of risk eat-
review in nursing]. RIDEC. 2014 Jan; 7: 24-40. Spanish. ing behaviors in high school students]. Rev Mex Trastor
14. H iggins JPT, Green S. Cochrane handbook for systematic Aliment 2019; 10(1): 22-31. http://dx.doi.org/10.22201/
reviews of interventions version 5.1.0. The Cochrane Col- fesi.20071523e.2019.1.519
laboration; 2011. 28. Reina SA, Shomaker LB, Mooreville M, et al. Sociocul-
15. Esteban L, Veiga O, Gómez S, et al. Length of residence and tural pressures and adolescent eating in the absence of
risk of eating disorders in immigrant adolescents living in hunger. Body Image 2013; 10(2): 182-90. doi: 10.1016/j.
madrid. The AFINOS study. Nutr Hosp 2014; 29(5):1047- bodyim.2012.12.004.
53. doi: 10.3305/nh.2014.29.5.7387. PMID: 24951984. 29. Laporta I, Delgado M, Reboyar S, et al. [Perfectionism in
16. Shahyad S, Pakdaman S, Shokri O, et al. The Role of In- adolescents with eating disorders]. Eur J Health Research
dividual and Social Variables in Predicting Body Dissat- 2020; 6: 97-107. Spanish. doi: 10.30552/ejhr.v6i1.205
isfaction and Eating Disorder Symptoms among Iranian 30. Rod E, Pons R, Lajara F, et al. Influence of the habits of the
Adolescent Girls: An Expanding of the Tripartite Influence adolescent population on self-image and the risk of eating
Mode. Eur J Transl Myol 2018; 28(1):72-7. doi: 10.4081/ disorder. Rev Pediatr Aten Primaria 2011; 13(51): 387-96.
ejtm.2018.7277. 31. De Sousa L, De Sousa S, Caputo E. [Influence of psy-
12 Acta Biomed 2022; Vol. 93, N. 3: e2022253 DOI: 10.23750/abm.v93i3.13140

chological, anthropometric and sociodemographic fac- 40. Moreno J, Torres C. Revisión sistemática de los determi-
tors on the symptoms of eating disorders in young ath- nantes socioculturales asociados a los trastornos de la con-
letes]. Paidéia (Ribeirão Preto) 2014; 24: 21-7. Spanish. ducta alimentaria en adolescentes latinoamericanos entre
doi: 10.1590/1982-43272457201404 2004 y 2014. Bogotá: UCA publication; University Applied
32. Castaño J, Giraldo D, Guevara J, et al. Prevalence of risk and Environmental Sciences, 2015.
of eating disorders in a female population of high school 41. Ruíz A, Vázquez R, Mancilla J, et al. [Family factors as-
students, Manizales, Colombia, 2011. Rev. Colomb. Ob- sociated with Eating Disorders: a review]. Rev Mex Trastor
stet. Ginecol 2012; 63(1): 46-56. Spanish. doi: https://doi. Aliment 2013; 4(1): 45-57. Spanish.
org/10.18597/rcog.202 42. Góngora VC. Satisfaction with life, well-being, and mean-
33. Rutsztein G, Scappatura L, Murawski B. [Perfectionism and ing in life as protective factors of eating disorder symptoms
low self-esteem across the continuum of eating disorders in and body dissatisfaction in adolescents. Eat Disord 2014;
adolescent girls from Buenos Aires]. Rev Mex Trastor Ali- 22(5): 435-49. doi: 10.1080/10640266.2014.931765.
ment 2014; 5(1): 39-49. Spanish. https://doi.org/10.1016/ 43. Levine MP, Smolak L. The role of protective fac-
S2007-1523(14)70375-1 tors in the prevention of negative body image and dis-
34. Haynos AF, Watts AW, Loth KA, et al. Factors Predicting ordered eating. Eat Disord 2016; 24(1): 39-46. doi:
an Escalation of Restrictive Eating During Adolescence. 10.1080/10640266.2015.1113826. 
J Adolesc Health 2016; 59(4): 391-6. doi: 10.1016/j.jado- 44. Haines J, Kleinman KP, Rifas-Shiman SL, et al. Examina-
health.2016.03.011. tion of shared risk and protective factors for overweight and
35. Maezono J, Hamada S, Sillanmäki L, et al. Cross-cultural, disordered eating among adolescents. Arch Pediatr Ado-
population-based study on adolescent body image and eat- lesc Med 2010; 164(4): 336-43. doi: 10.1001/archpediat-
ing distress in Japan and Finland. Scand J Psychol 2019; rics.2010.19.
60(1): 67-76. doi: 10.1111/sjop.12485.
36. Batista M, Žigić Antić L, Žaja O, et al. Predictors of eating
disorder risk in anorexia nervosa adolescents. Acta Clin Cro-
at 2018; 57(3): 399-410. doi: 10.20471/acc.2018.57.03.01. Received: 19 April 2022
37. Jiménez Flores P, Jiménez Cruz A, Bacardi Gascón M. Accepted: 6 June 2022
[Body-image dissatisfaction in children and adolescents: a Correspondence:
systematic review]. Nutr Hosp 2017; 34(2): 479-89. Span- Maura Galletta, Ph.D
ish. doi: 10.20960/nh.455. PMID: 28421808. Department of Medical Sciences and Public Health,
38. M ancilla A, Vásquez R, Mantilla J, et al. [Body dissatisfac- University of Cagliari,
tion in children and preadolescents: A systematic review]. SS554 Bivio per Sestu
Rev Mex Trastor Aliment 2012; 3(1): 62-79. Spanish. Monserrato (CA), 09042 Italy
39. B erkman ND, Lohr KN, Bulik CM. Outcomes of eating Phone: +39 070 675 3108
disorders: a systematic review of the literature. Int J Eat Dis- E-mail: [email protected]
ord 2007;40(4): 293-309. doi: 10.1002/eat.20369. ORCID number: 0000-0002-0124-4248