Orthorexia and Eating Disorders in Adolescents and Young Adults: A Systematic Review

Background: In recent years a new term in the field of eating behaviors has emerged, namely “orthorexia”. This behavior is associated with significant dietary restrictions and omission of food groups. The aim of the present study is to estimate the possible correlations between orthorexia and eating disorders in young adults and adolescents. Methods: A systematic review of related articles in PubMed, Google Scholar, and PsycInfo was conducted up to 30 June 2021. Results: A total of 37 studies (16,402 subjects) were considered eligible for this systematic review. Significant correlations were observed in most of the studies between orthorexia and eating disorders. However, the majority of studies adopted a cross-sectional design. Conclusions: An association between and eating disorders emerged. Prospective studies seem necessary to investigate associations and succession of orthorexia and eating disorders over time.


Introduction
Orthorexia comes from the Greek words "orthos" which means "proper" or "correct" and "orexis" which means "appetite" [1]. Eating-related behaviors that pertain to a fixation to eat healthy, a consumption of biologically pure foods regarding the personal benefits of healthy eating, describes orthorexia [2], also known as "orthorexia nervosa". Orthorectics tend not to consume food that has been processed with pesticides, herbicides or artificial substances, and they are highly worried about the techniques and materials involved in food preparation [2]. The prevalence of orthorexia among youth and young adults is estimated as high as 27% [3]; however, there are discrepancies between the results of various studies and prevalence in this age group can be found from 6.9% [4] to 45.5% [5], whereas studies performed in the more recent years tend to show higher prevalence of orthorexia in adolescents and young adults, than those conducted years ago.
It has been postulated that orthorectic individuals are engaged in four stages of daily behavior. Stage one includes excessive thoughts considering food consumption for a specific day and the following days. Next stage involves the excessive control of collecting food products and ingredients. The third stage pertains to very careful meal preparation which is compliant with healthy eating philosophy. The last stage entails following sentiment of accomplishment or failure, depending on the perceived outcomes of all previous stages [6].
A systematic literature search was performed on 30 June 2021 in PubMed, PsycINFO and Google Scholar databases. The systematic review protocol was registered in the School of Medicine, National Kapodistrian University of Athens (registration number: 29738/14-05-2021) and is available upon request. Various search terms were used, specifically combinations of the following terms as a search algorithm: (orthorexia) AND (adolescents OR "young adulthood" "young adults" OR "young adulthood") AND (anorexia OR bulimia OR "eating disorder"). Additionally, a thorough search of the reference lists of the considered eligible studies as well as those of relevant reviews was searched to identify further eligible reports.

Inclusion Criteria
Inclusion criteria encompassed the following:

1.
Reports on young adults up to 30 years old and adolescents of any age. Studies which were based on adolescents/young adults were separately presented in our systematic review from those including a subgroup of admixture between young adults and older individuals.

2.
Presentation of data about possible correlations between orthorexia and eating disorders (AN and BN); the latter defined according to DSM 5 and ICD 10.

3.
No restrictions were posed considering diagnostic tools of eating disorders. 4.
Considering study design, prospective cohorts, case-control and cross-sectional studies were included. 5.
Only articles written in English were included.
No restrictions were posed regarding publication year. All article titles and abstracts were screened by authors working in pairs, blindly to each other.

Exclusion Criteria
Case reports, animal studies, review articles, medical hypotheses, studies not disclosing age groups, as well as studies looking for a correlation between orthorexia and gender, BMI or dietary patterns in general, were excluded.

Quality Assessment of Included Studies
The quality of studies was rated with the Newcastle-Ottawa scale, which was adapted for cross-sectional studies [14].

Data Collection Process and Data Extraction
Authors worked in pairs, blindly to each other and extracted all the relevant data according to the inclusion criteria, as described above. In any case of disagreement, team consensus was followed. Collected data were categorized into tables and in any case the type of study, country, study period, study design, sample size, gender and age distribution, sample type (entirely adolescents or young adults or admixture), confounders, definitions of eating habits and orthorexia, associations studied in relation to orthorexia and potential confounding factors assessed were included.

Compliance with Ethics Guidelines
This article is based on the results of previously conducted studies. The study was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [15].

Study Characteristics
The literature search led to a sum of 1508 results, after removing the duplicates; of them 1454 were considered irrelevant according to title and abstract, while 54 full-texts were evaluated. Among the latter, 12 were excluded with reasons and a total of 37 studies (16,402 subjects) were finally included [2,3,11,. The PRISMA flowchart is presented in Figure 1. Table 1 shows the features and findings of included studies, as well as quality assessments; the majority of studies were rated as of low quality.

Assessment Tools-Observed Correlations
A number of different instruments (ORTO, EAT, EDI) were used in every study in order to define orthorexia and EDs, as presented in Table 1.
According to the majority of eligible studies, an association between orthorexia and eating disorders was reported. Specifically, Agapoyan [16], Arslantas [17], and Haddad [18] identified a significant negative correlation (p < 0.05) between the EAT-40 and ORTO-11 scores; higher scores in EAT-26 showed more eating pathology and lower scores in ORTO-11 and ORTO-15 showed more tendency for orthorexia. Likewise, Mahesh Shah's study examined the individual's scores on measures of orthorexia, eating disorder, and some other parameters, confirming the significant negative correlation between disordered eating behaviors and orthorexia nervosa (r = −0.33, p < 0.01) [19]. In 2017, Okumuşoglu using a different questionnaire to measure eating disorders and the Orto-11, found a negative correlation with eating disorder tendencies which were measured by REZZY scores (r = −0.328, p = 0.01) [20]. McInerney-Ernst collected data from self-reported responses, indicating that orthorexia shares important characteristics with established eating disorders [21]. Self-reported disordered eating patterns significantly predicted ON symptoms [21]. Bóna et al. observed an overlap between certain eating disorder traits (drive for thinness and interpersonal distrust) and orthorexia [22]. This finding was confirmed by Parra-Fernandez et al. [23] as they remarked that the highest negative correlation coefficient (−0.564, p < 0.01) was observed between the subscale "drive for thinness" and orthorexia. Moreover, Koven and Senbonmatsu noticed that two features of AN, body dissatisfaction and perfectionism symptoms increased as ORTO-15 score decreased [2]. ON was negatively predicted by eating pathology, weight concern, health orientation, and appearance orientation. Orthorexic behaviors were not found to be significant with pathways between other variables. Orthorexic behaviors were more frequent in the group with lower level of eating pathology and less common in the group reporting higher levels of eating pathology.

Influence of Age/Gender
Considering the age parameter, subjects aged between 13 and 16 years old presented the highest risk of orthorexia [24]. That was also the finding in the study of Bona et al. where orthorexia was more likely to be observed in younger and rather fit participants [22]. Moreover, the risk of displaying orthorexia tendency as well as ED symptoms was induced by the present and past use of a special diet [25,26]. Gender might also be considered as a factor which might influence the occurrence of orthorexia, but there are still no clear results from studies [27,28].

Assessment Tools-Observed Correlations
A number of different instruments (ORTO,EAT, EDI) were used in every study in order to define orthorexia and EDs, as presented in Table 1.
According to the majority of eligible studies, an association between orthorexia and eating disorders was reported. Specifically, Agapoyan [16], Arslantas [17], and Haddad [18] identified a significant negative correlation (p < 0.05) between the EAT-40 and ORTO-11 scores; higher scores in EAT-26 showed more eating pathology and lower scores in ORTO-11 and ORTO-15 showed more tendency for orthorexia. Likewise, Mahesh Shah's study examined the individual's scores on measures of orthorexia, eating disorder, and some other parameters, confirming the significant negative correlation between disordered eating behaviors and orthorexia nervosa (r = −0.33, p < 0.01) [19]. In 2017, Okumuşoğlu using a different questionnaire to measure eating disorders and the Orto-11, found a negative correlation with eating disorder tendencies which were measured by REZZY scores (r = −0,328, p = 0.01) [20]. McInerney-Ernst collected data from self-reported responses, indicating that orthorexia shares important characteristics with established eating disorders [21]. Self-reported disordered eating patterns significantly predicted ON symptoms [21].

Prognosis
As far as prognosis and features of EDs is concerned (Table 1/lower panels), Segura-Garcia et al. [11] showed that orthorexia symptoms were observed with high incidence among patients with AN and BN and tended to increase after treatment. Barthels et al. indicated that individuals with AN and pronounced orthorexic eating behavior chose to eat more often foods which are labeled as healthy [29]. "Autonomy" and "competence" was lower in AN patients who reported low orthorexic eating behavior and higher in individuals with AN and pronounced orthorexic eating behavior compared to [29]. Interestingly, results demonstrated that adults who had "normal" eating behavior had no risk of ON, while adults who had psychological and affective traits of eating disorders had ON [30].

Discussion
The present systematic review highlighted a correlation between the presence of eating disorders and orthorexia [18,27]. Higher eating disorders were significantly associated with higher tendencies and orthorexic behaviors [33]. High scores on the EAT (higher orthorexia behaviors) were significantly associated with orthorectic tendencies [16,33].
Specifically in adolescents it was found that those suffering from eating disorders had an increased risk of orthorexia [3,17,35]. Anorectic individuals are commonly preoccupied with the purity of their diet and tend to avoid certain foods which they consider to be safe. Eliminating food groups results in weight loss and nutritional deficiencies [44]. The association between AN and EDs raises concerns about whether orthorexia and anorexia nervosa are two overlapping conditions [50]. An Italian study by Dell'Osso et al. identified similarities between the two conditions and proposed that ON could be considered as an AN phenotype [51]. Adolescent orthorectics, develop common symptoms of anorexia [2]. Orthorectic and anorexic individuals have difficulty setting themselves in set-shifting. Thus, for example, individuals with orthorexia usually develop specific rules for choosing food that cannot be with preservatives, rules for preparing food (such as eating only raw foods), and they have to eat alone or in specific conditions. These rules gradually increase in number and complexity, so the orthorectic person devotes more time and energy to follow these rules. This behavior and cognitive rigidity also occur in AN [52]. In addition, orthorexia was more strongly associated with the symptoms of AN and BN sharing common symptoms such as "goal-driven" direction, "ego-syntonic" coordination, or "positively reinforcing" character with dietary restrictions [42,44]. Both ON and AN/BN are distinguished by firm diets in relation to the nutritional properties of food and a strong dread of long-term effects of the food they fear [47]. On the contrary, ON seems to be a separate condition from Avoidant/Restrictive Food Intake Disorder (ARFID). Unlike ARFID, people with ON might prefer not to limit their intake linked to an interest in food, the sensory properties of what they eat, or because of a previous deterrent food experience, but because of a tenacity to be as healthy as possible. While these people with AN can suffer serious medical consequences due to their food choices, people with ON seem not to have any problems with perceiving their weight or body shape while their self-esteem is also not affected by weight or shape.
About potential causal associations, a lot of individuals presenting a risk of orthorexia and disordered eating habits, have stated that previous diet followed their disorder [44]. It appears that people who monitor their caloric intake and are afraid of gaining weight have higher rates of eating disorders, whereas those who care about proper nutrition and are concerned about weight gain had higher scores on orthorexia [17]. People who are trained in healthy eating major report that they are interested in patients and themselves, which can lead to food obsession [53]. Thinking about how food is being prepared or spending more time thinking about food and healthy eating can be risk factors that can lead to eating disorders [54].
Regarding associations with specific groups of youth, the study by Arslandas et al. found that nursing students were 84.5% more likely to develop an eating disorder and 45.3% of the same students were at risk of developing orthorexia [17]. Aksoydan and Camcı in their research highlighted the prevalence of orthorexia between opera singers and ballet dancers [55]. Of the total sample, 56.4% were identified at risk of orthorexia. Ergin [56] conducted a study on a group of healthcare staff and observed that 60.1% of them were at risk of ON. Nutrition students in a German study had higher scores on dietary restriction than students in other curricula. This restriction concerned the tendency for moderate food intake either for weight loss or for weight maintenance [57]. Nutrition students had higher rigid control and higher flexible control of eating behavior compared to the characteristic control group that may cause some concern.
It is ascertained that orthorexia seems to be different from AN and BN. In ON, dietary restrictions are not driven by excessive fear of weight gain, or by the excessive effect that shape and weight have on their self-assessment, nor by distorted body image [58]. Body image dissatisfaction is mostly observed in other eating disorders including binge eating [59]. While overeating and cleansing or inadequate calorie compensation contributes to symptoms of anorexia and orthorexia, they are not part of the recommended ON diagnostic picture [58]. These distinctions are important, as traditional therapeutic approaches and follow-up to eating disorders such as anorexia may not be appropriate for people with ON [58]. Orthorexia and psychogenic anorexia and bulimia are associated with intense obsessions and compulsive behaviors, which link these conditions to obsessive-compulsive disorder, such as repetitive, intrusive thoughts about food and health at inappropriate and inappropriate times and a strong need to organize food in a ritual way [60]. Similar to OCD people, orthorectic people have limited time for other activities, as adhering to a strict diet affects their normal routines [13].
According to the available literature, most of the young adults who have been diagnosed with any kind of eating disorders and have received treatment in specialist eating disorder services present great results at long-term follow-up; however, even after many years, a significant number still suffer from other mental health problems, requiring therapeutic approach [61]. Nevertheless, the frequency of follow-up does not seem to reduce nor the risk of nutritional deficiencies or the use of dietary supplement [62].
The results of the present systematic review should be interpreted with some concern due to the limitations of the eligible studies. One of those limitation is the cross-sectional design, which does not permit establishment of a causal relationship between disorder eating and ON, as well as the fact that a variety of questionnaires was used. Moreover, another limitation is the low quality ratings [2,3,11,. Furthermore, factors such as the small sample in some studies, the lack of subgroup data about bulimia, and the lack of a universally accepted instrument to assess orthorexia should be taken into consideration. The fact that only studies in English were included may have limited generalizability of findings to countries with a western lifestyle.

Conclusions
In conclusion, this systematic review indicated a potential association between orthorexia and eating disorders. Further research is needed to evaluate notions of causality, as well as to document factors signaling similarities and discrepancies between these conditions. Author Contributions: Conceptualization, T.N.S. and A.T.; methodology, P.S., M.E.C., T.P., A.G. and A.T.; investigation, P.S., M.E.C., A.G. and T.N.S.; writing-original draft preparation P.S., M.E.C., E.P., T.N.S. and T.P.; writing-review and editing, A.G., G.M., T.P., T.N.S. and A.T.; visualization, G.M., T.P. and E.P.; supervision, G.M., A.T. and T.N.S. All authors have read and agreed to the published version of the manuscript.