Introduction

Vegetarianism has become more popular in recent years. Over the last half century, the number of individuals adhering to a vegetarian diet has greatly increased. Globally, the total number of vegetarians is estimated to be approximately one-and-a-half billion [1]. In Western countries, the estimated prevalence of individuals following vegetarian diets oscillates between 1 and 10% [2]. However it is worth noting that these estimates are derived from polls and surveys carried out directly by vegetarian societies [3] not from national or official observational surveys.

According to the Vegetarian Society [4], a vegetarian diet is characterized by the consumption of fruits, vegetables, legumes, pulses, grains, nuts, seeds, fungi, algae, yeast and/or other non-animal-based foods, with or without dairy products, honey and/or eggs. A vegetarian diet excludes foods that contain or have been produced by means of products consisting of or created from any part of the body of a living or dead animal. However, it is worth noting that there is no single dietary pattern that describes vegetarianism [5]. Table 1 summarizes the classification of different types of vegetarian diets.

Table 1 Types of vegetarian diets (based on Phillips [5] and the Vegetarian Society [4])
Full size table

Mainly ethical, but also health-related aspects, ecological reasons, health concerns, sensory and taste preferences, philosophical reasons (e.g. Buddhism), cost, family influences, a reaction to food safety scares (e.g., salmonella) or the use of antibiotics or growth hormones in meat production are among the individual factors that contribute to the adoption of a vegetarian diet [5].

The various health benefits of a vegetarian diet have been demonstrated in both prospective cohort studies and randomized clinical trials [6]. Preliminary evidence suggests that individuals adhering to a vegetarian diet (or a vegan diet) are more likely than omnivores to state that they are health-conscious [7]. However, in some individuals, interest in healthy attitudes and behaviors towards food may show obsessive signs [8]. Preoccupation with ‘healthful’ eating may contribute to orthorexia nervosa (ON)—“a fixation on eating healthy food” [9, p. 9]. This term was coined for the first time by Bratman [10] (“ortho” comes from Greek and means “right” or “correct”; “orexia” relates to eating or appetite and “nervosa” means “obsession” or “fixation”). Orthorexia nervosa is described as an extreme preoccupation with healthy eating according to the person’s own criteria (including but not limited to a fixation on organic and biologically clean products, dietary supplements, raw foods, and high- or low-carbohydrate foods), with excessive care for one’s health, intrusive thoughts regarding being healthy, a lack of dietary flexibility, long-term planning and a negative impact of food choices on quality of life [9, 10]. According to the new guidelines [11], orthorexia nervosa is characterized by the criteria presented in Fig. 1.

Fig. 1
figure 1

Classification criteria for orthorexia nervosa [11]

Full size image

The amount of research about ON has grown in the last several years (see Fig. 2). To date, the literature on ON is dominated by descriptive and anecdotal data, frequently with inconsistent results [12].

Fig. 2
figure 2

Number of PubMed articles with the term “orthorexia”/“orthorexic” in the title from 2002 to 2019 (until 28th February)

Full size image

In the literature, there is a lack of consensus regarding the link between a vegetarian diet and orthorexia. On the one hand, there is currently some empirical evidence to suggest that vegetarians are at greater risk of orthorexic eating behaviors than are omnivores [e.g., 8, 13, 14]. On the other hand, there is some research indicating that a vegetarian diet is not related to an obsession with healthy eating [15, 16]. It is worth pointing out that a recent study [17] has shown that not a vegan diet per se but underlying motives (health, esthetics and healing) determine whether a vegan lifestyle is associated with orthorexic eating behavior. These results would suggest possible moderators (i.e., health-related motives and beliefs) of this relationship. This finding is in line with other research [8] that has found that individuals following a meat-free diet for health reasons were more prone to higher levels of orthorexic eating behavior. These results could indicate that ethical reasons for adhering to a vegan diet might serve as a protective factor regarding the development of orthorexia nervosa [8]. On the other hand, other studies have reported that ethical eating (or food ethics), exclusion of certain foods or food groups, and anxiety elicited by food consumption choices share similarities with subclinical disordered eating behaviors (e.g., skipping meals, fasting, dieting, etc.), eating disorders (e.g., anorexia nervosa) [19] as well as maladaptive eating behaviors (orthorexic eating behaviors). The various rules for healthy eating connected with vegetarian diet could prompt the development of disordered eating symptoms (e.g., restrictive eating patterns, preoccupation with food) [20,21,22]. Conversely, disordered eating behaviors could lead individuals to adhering to a vegetarian diet. Following a “healthy” diet might provide another way for individuals with orthorexia nervosa to restrict and control their food intake in a socially acceptable way (following a vegetarian diet).

To conclude, previous research findings have provided inconsistent results in the field. Therefore, the systemization of research results seems to be relevant for the current state of knowledge. The present literature review aimed to investigate the association between a vegetarian diet and orthorexia nervosa. Due to the limitations of the selection of manuscripts based on the database search which was completed on 28th February 2019, the current study does not address possible moderators of the relationship between vegetarian diet and orthorexia nervosa.

Materials and methods

In a literature review, the titles and abstracts of the search results were assessed. For each paper, the type of the study (empirical research), characteristics of the sample (vegetarians/vegans, omnivores), psychometric tests (measuring orthorexic eating behavior) and conclusion/results were defined. Case reports, comments, editorials, letters, and reviews were excluded from the selection process. Data on the first author’s surname, year of publication, country, sample characteristics, methods for assessing orthorexic eating behavior and results/conclusions were extracted from all the included studies.

Results

The search strategy performed in the PubMed/Medline database (last access on 28 February 2019), with “orthorexia”, “orthorexic”, “orthorexia and vegetarian”, “orthorexia and vegan” and “orthorexia and vegetarianism” used as search terms, identified 117 publications. First, all publications were scrutinized by the title and abstract. One hundred and three papers were excluded for the following reasons: lack of data on the vegetarian/vegan population (non-representative sample was assessed), lack of a variable for a vegetarian/vegan diet, languages different from English, and narrative articles (reviews, letters, commentaries, editorials, case reports). Finally, the full texts of 14 articles were reviewed, with 100% published within the last 5 years. The selection process of the included studies is depicted in Fig. 3.

Fig. 3
figure 3

Flowchart of study selection

Full size image

Study characteristics

Tables 2 and 3 provide a summary of each study included in the review. Table 2 presents the information with individuals adhering to a meat-free diet (vegetarian/vegan diet), while Table 3 includes the research assessing a meat-free diet (vegetarian, vegan) in the general population. The studies were performed in the United States (four studies), Germany (three studies), Poland (two studies), Italy (two studies), Spain (one study), Turkey (one study) and the United Kingdom (one study).

Table 2 Characteristics of research studies conducted amongst individuals following a vegetarian or vegan diet
Full size table
Table 3 Characteristics of studies examining a vegetarian or vegan diet in the general population
Full size table

Discussion

The present literature review investigated the association between a vegetarian diet and orthorexia nervosa. Overall, the evidence from the review indicates that vegetarians report more orthorexic behaviors than those who follow an omnivorous diet [8, 13, 23, 24]. It can be suggested that a vegetarian diet might increase the risk of developing orthorexic eating behavior [9] due to the elimination of large groups of micro- and macronutrients [30]. A vegetarian diet may also serve as a socially acceptable means to mask disordered eating behaviors [19]. Timko et al. [21] have emphasized that dietary restraint in individuals following a vegetarian diet are often interpreted as indications of disordered eating or maladaptive eating-related attitudes. The authors have hypothesized that high levels of dietary restraint are an artifact of the specific elimination of meat products from the diet, not restriction in general.

In seven studies, individuals following a vegetarian diet in the general population have been found to present more orthorexic eating behavior or to be at risk of developing ON [14, 18, 19, 25,26,27, 29]. It should also be noted that only one study [15] has shown that compared to individuals who do not restrict particular foods, individuals adopting a vegetarian diet in a general population are less likely to engage in orthorexic eating behavior. Another study [28] has shown that women with different diet types who were following social media and/or Instagram do not differ in orthorexic eating behavior. Only one study has demonstrated no difference in orthorexic eating behaviors between vegetarian and omnivorous individuals [16].

The present literature review highlights the opportunities for future research. First, as most of the studies employed a cross-sectional study design, it is currently not possible to clarify any contributions of a vegetarian diet to orthorexic eating behavior. Therefore, future research should consider a longitudinal approach to facilitate the investigation of the potential contribution of a meat-free diet on orthorexic eating behavior. A longitudinal study provides the opportunity for observing individual patterns of change; whereas in cross-sectional studies, the differences in outcomes are usually attributed to differences between individuals [30] and the effects of within-individual differences cannot be differentiated from that of between-individual differences [31]. Moreover, the longitudinal design captures relevant information for developing prediction model about within-individual differences and permits these effects to be distinguished from between-individual differences [31]. Wherefore, using a longitudinal study design would provide the possibility for identifying the prognostic factors for orthorexic behaviors. In response to researchers’ postulates that vegetarianism may be a precursor to the development of an eating disorder, Timko et al. [21] have emphasized the fact that the adoption of a vegetarian diet after the development of an eating disorder may indicate that rather than being a causal factor, a vegetarian diet may play more of a role in the maintenance of eating pathology. It can be hypothesized that this mechanism could also occur in orthorexia nervosa.

Second, there are opportunities for future research to develop validated tools that comprehensively measure orthorexic eating behavior. These tools would enhance the comparability of findings among studies and subsequently increase confidence in the interpretation of findings in the body of research in this area. The studies in this literature review incorporated a variety of methodological approaches to their study design. Furthermore, most of the studies used different measurement tools (DOS, EHQ, ORTO-15) that in some cases were not validated (BOT) and that had different levels of validity and reliability [26]. The Orthorexia Self-Test, developed by Bratman (BOT) and described by the author as “a ten-question quiz to determine if you have orthorexia” [9; p. 47], did not evaluate either the necessary psychometric properties (namely, reliability and validity) or the cut-off scores of a reference group. It was designed as a screening instrument (as an informal measure). The Orthorexia Self-Test, however, has been the basis of the ORTO-15, which is “questionnaire for the diagnosis of orthorexia nervosa” [32; p. e28]. Currently, the ORTO-15 is probably the most widely used self-report measure of orthorexia nervosa [8]. Although preliminary validation has shown that the ORTO-15 has good predictive validity [32] and that a cut-off score (of 40 or 35) was established to identify individuals with ON, the following challenges have been identified: the method used to develop the cut-off point was not well-founded; a low reliability has been established; there is a lack of clear articulation of the development of construct validity; there is a lack of discussion about the creation of an item pool; there is a lack of standardization methods; there is a lack the basic psychometric properties and the internal consistency has been criticized [16]. Furthermore, an additional problem is the poor face validity of many of its items [33]. In most of the studies, the ORTO-15 was used in spite of numerous contraindications. The overall construction problem of the ORTO-15 questionnaire indicates this method entails some basic psychometric flaws and its usage should be rethought from scratch [25] or it should be replaced by other measurement. Another method, the Eating Habits Questionnaire (EHQ) [34], evaluates cognitions (knowledge of healthy eating), behaviors (problems associated with healthy eating) and feelings (feeling positively about healthy eating) related to an extreme focus on healthy eating, which has been called orthorexia nervosa. The EHQ displayed good internal consistency and test–retest reliability in a college student sample [34]. The screening instrument, Düsseldorfer Orthorexie Skala [35], measures orthorexic eating behavior and shows good internal consistency, good test–retest reliability and good construct validity. As a preliminary cut-off score to indicate the presence of ON, a score of ≥ 30 is used. A score between 25 and 29 (95th percentiles) describes conspicuous eating behavior (indicating a risk of ON) [35]. The latest measure, Teruel Orthorexia Scale [36], assesses two related, although differentiable aspects of ON: healthy orthorexia and orthorexia nervosa. The first one is described as tendency to eat healthy food and interest in doing so. While, the second one assesses the negative social and emotional impact of trying to achieve a rigid way of eating [36].

The major limitation of the studies was to assess the vegetarian diet through self-report as opposed to through dietary data and the objective measures (such as the Food Frequency Questionnaire or dietary protocol). It should be taken into account that in the majority of studies, individuals adhering to a vegetarian diet were often combined into one group (vegetarians), even though they followed one of the types of vegetarian diets (e.g. veganism, lacto-vegetarianism). In addition, the operational definition of “vegetarianism” [21] could be ambiguous in the available studies. In the majority of studies assessing a meat-free diet, participants were asked if they were following a vegetarian diet (self-declaration). In this case, these participants could be named “self-reported” vegetarians in contrast to individuals who are “true” vegetarians (they meet all the criteria for following a free-meal diet). An absorbing study [21] has found that self-described vegetarians have significantly higher levels of restraint, external eating, hedonic hunger, and lower levels of acceptance in relation to food in comparison to true vegetarians.

The second limitation was caused by the fact that review of the literature has been mainly focused on correlational and cross-sectional studies. However, it is worth being aware that these investigations are extremely widespread in the field (the majority of research in this area has been cross-sectional). To the best of my knowledge, there is no known research that has directly examined the causal relationship between vegetarian diet and orthorexic behaviors. It is worth noting that the relationship between a vegetarian diet and orthorexia nervosa is partly artefactual due to the wording of self-report items, and motivations for being a vegetarian and adherence to the diet are both predictors of disordered eating. Timko et al. [21] have argued that vegetarians’ higher scores on traditional methods of eating behaviors appeared artificially inflated by ratings of items assessing avoidance of specific food items which should be considered normative in the context of a vegetarian diet. Individuals following a vegetarian diet only score high on specific items related to avoiding certain foods or having trouble finding foods they can eat [21]. It should be pointed out that some of the items on all of the existing orthorexia questionnaires could be interpreted as referring to difficulty finding vegetarian options, thus potentially inflating vegetarians’ scores without necessarily meaning that they have orthorexic tendencies.

To conclude, despite its limitations, this literature review provides novel insight into the relationship between a vegetarian diet and orthorexia nervosa. The present review found that following a vegetarian diet is associated with higher scores on self-report measures of orthorexic eating behaviors. It is also worth mentioning that the current research findings do not support the causal relationship between adhering to a vegetarian diet and orthorexia nervosa. For investigating the causality, which refers to the relationship between two sets of events, where one set of events (the effects) is a direct consequence of another set of events (the causes) [37], experimental and longitudinal studies are required. By searching for causality clauses (causal relationships between a vegetarian diet and orthorexia nervosa), it is possible to extract prior and posterior eating-related behaviors and food-related behaviors. Because the studies cited in the review of the literature did not assess adherence to a vegetarian diet, there is a possibility that this unexplored moderator contributes to the strength of the effects in the review of the literature.

Conclusions

Although in the most of the studies (9 out of 14) the ORTO-15 was used, the present findings based on more adequate methods (the DOS and EHQ) have shown that following a vegetarian diet was found to be associated with orthorexic eating behaviors [8, 14, 24, 29]. For having more strong evidence, we should consider ending the use of the ORTO-15 in favor of the EHQ and the DOS. With reference to the latest literature review on the various tools to assess orthorexia nervosa [38], it is worth emphasizing that the authors have stressed the need for the construction of a new diagnostic tool. Their suggestions [38] connected with development of a future diagnostic tool include: (a) constructing a modern re-conceptualization of diagnostic criteria of orthorexia nervosa (there are discrepancies between diagnostic criteria and diagnostic tools) for a better understanding of orthorexia nervosa and creating a new tool based upon this re-conceptualization (according to multiple perspectives); (b) adopting qualitative data collection techniques to gain insights into how to diagnose orthorexia nervosa; and (c) actively involving diverse stakeholders for constructing a new tool as well as during the phases of the study process.

In addition, there is a need to provide better study designs (e.g., good planning including research protocol and appropriate questionnaires) as well as experimental and longitudinal research (to establish causality between a vegetarian diet and orthorexia nervosa) to find out whether following a vegetarian diet serves as a risk factor for orthorexic eating behavior as well as whether there are moderators of this relationship (e.g. true vs. “fake” vegetarianism, self-identification as a vegetarian with or without adherence to the diet [30], motivation for being a vegetarian [17] or socioeconomic status). Proposed directions for future research include: focusing on large samples of true vegetarians, appropriate operational definitions of “vegetarianism” [21] and use of comparable tools of orthorexic behaviors.

As already mentioned previously, recent research has pointed to the bidimensional nature of orthorexia nervosa [36], with one dimension related to healthy (non-pathological) interest in diet and healthy behavior with regard to diet (a protective factor against emotional distress) named healthy orthorexia (HeOr) and another dimension related to a pathological preoccupation with eating healthily (a new variant of disordered eating related to negative affect) named orthorexia nervosa (OrNe) [36, 39, 40]. Depa et al. [39] have claimed that OrNe is more common among vegetarians and vegans, compared to people who are not adhering to a special diet. Further studies should investigate these two dimensions among individuals following a vegetarian diet for distinguishing pathological restrained eating from restrictions based on healthy food content [40].

What is already known on this subject?

ON is still not sufficiently explored phenomenon and lacks decent theoretical and empirical work. There is no clear understanding of the construct ON, especially taking into account the connection between vegetarian diet and ON. Evidence in support of this assumption has been largely inconsistent.

What does this study add?

This review literature was needed to clarify how ON manifests itself in individuals following a vegetarian diet. The present findings have shown that (a) following a vegetarian diet is associated with higher level of ON as well as (b) vegetarians report more orthorexic behaviors than those who follow an omnivorous diet.