The need for research on orthorexia nervosa: looking back, looking forward

Abstract

One of the sustainable development goals (SDGs) of the United Nations is to enhance scientific research and substantially increase the number of public and private research and development spending (SDG Target 9.5). What types of research should be carried out in the field of orthorexia nervosa (ON)? Fundamental, applied, comparative, exploratory, laboratory, longitudinal or mixed research? How can we utilize our previous misinterpretations to develop a more accurate understanding of ON? Do we need more research in this field? These and other questions encourage us to publicly debate on ON, which is why we would like to express our position on this issue.

Level of evidence: Level V: Opinions of respected authorities, based on descriptive studies, narrative reviews, clinical experience, or reports of expert committees.

A past and current research perspective in the orthorexia nervosa (ON) field

Focus on healthy foods is an important health concern in our society. However, in some individuals, (subjectively defined healthy but objectively defined unhealthy) dietary choices and diet patterns may contribute to the risk of developing ON. The first mention of the term “orthorexia nervosa” given by Steven Bratman was in the late 1990s [1]. The concept of ON was introduced to describe patterns of specific eating behaviours and excessive preoccupation with healthy eating because of maintaining health. Nowadays, ON is defined as “an overvaluation and preoccupation with food quality and its impact on health” [2; p. 3696]. The concept of ON attracts more and more interest in the scientific community, resulting in a significant increase in the number of published scientific articles in recent years. In the last 20 years, over 600 published articles (indexed in PubMed) have been published on this topic. Regarding the UNESCO Science Report [3], there has been a general trend toward more intense scientific publishing, with global output being 21% higher in 2019 than in 2015.

Some researchers [4] have noted that “Research—whether original studies or meta-analyses—based on invalid measures is invalid” (p. 2). This statement has been related to the ORTO-15 [5], which was the first widely used tool in the field of ON. It is important to recognize that at the onset of ON research, the ORTO-15 was the most frequently used questionnaire, with multiple language translations available, including Arabic, Chinese, English, French, German, Greek, Hungarian, Polish, Portuguese, Spanish, and Turkish. Furthermore, the ORTO-15 was the only available way to collect data at this time (the Eating Habits Questionnaire (EHQ) was published in 2013, the Düsseldorf Orthorexia Scale (DOS) was published in German in 2015, and Orthorexia Nervosa Inventory (ONI) was published in 2021). The ORTO-15 has significantly contributed to the rapid expansion of research in ON.

In response to concerns about the validity of the existing studies using this questionnaire, we believe it is important to adopt a more realistic view of ORTO-15. Scientists can sometimes make mistakes or misinterpret data. The mistakes that they can make are concluding that something is true when it is actually false (a type I error) or concluding that something is false when it is actually true (a type II error). These mistakes and misinterpretations of the results have been particularly noted in studies using the ORTO-15. In ORTO-15 lower scores indicate more pathological ON behaviour, meaning that the inverse correlation represents the theoretically assumed relationship between the two scales. In the majority of the previous study, the mistakes were related to the misinterpretation of correlations.

The next controversial issue pertains to the overestimation of the prevalence of ON based on ORTO-15 (e.g., 72.3% of Turkish female nutrition students [6]) and thus, the dissemination of inaccurate epidemiological data. This has resulted in disproportionately high ON prevalence, particularly among low-prevalence groups, such as the student population. This suggests that the ORTO-15 is not a tool suitable for assessing ON. To address this challenge, there is an urgent need for validated diagnostic methods to estimate the prevalence of ON accurately. The challenges associated with the ORTO-15, particularly concerning the issues of false positives that lead to misdiagnosis, as well as its sensitivity and specificity are also worth underlining. The ORTO-15 does not ensure that individuals who do have this condition are accurately identified (sensitivity) and does not permit correctly classifying those who are not at risk of ON (specificity). Researchers (e.g., [7]) have indicated that the ORTO-15 is a mediocre screening tool for ON. While it demonstrates sensitivity to diet, it lacks the necessary level of specificity to effectively identify the pathological stage of ON.

The practical application of ORTO-15 has enabled researchers to detect its errors, flaws and limitations. Unexpected results have inspired researchers to verify them. The overwhelming majority of researchers agree about the psychometric limitations of ORTO-15, including issues with reliability and validity. Its limitations are recognised among the experts in the ON field [2, 8]. Consequently, we should accept that many previously published studies using ORTO-15 hold limited value. However, we do not consider the research conducted on ON using the ORTO-15 to be entirely invalid or worthless. In science, numerous examples support our opinion. For instance, intelligence tests, such as the Stanford–Binet Intelligence Scale, were criticised for yielding various biases (e.g., cultural). Nevertheless, they significantly contributed to the advancement of intelligence research. Furthermore, the constructive criticism of the earlier versions of the tests has prompted the development of more sophisticated methods for measuring intelligence. Similarly, older versions of the Minnesota Multiphasic Personality Inventory contained many problematic questions and biases (e.g., [9]), but they led to improved versions. Literature indicates that even an “imperfect” tool can yield valuable data that contributes to the improvement of measurement methods. An example is the development of personality questionnaires—the first versions were very simplified, but they provided knowledge based on which more advanced models were created, such as the Five-Factor Model of Personality (commonly known as the Big Five). The history of science reveals that errors are an inherent part of its development. Many breakthroughs began with tools that were later turned out to be flawed. However, it was thanks to their use that researchers were able to improve them and further expand the scope of knowledge. This demonstrates that even imperfect research tools can contribute to the progress of science, despite their flaws, because they enable error detection, new method development, and theory improvement. Summing up, inappropriate questionnaires result in incorrect or imprecise conclusions; however, their use often provides the foundation for further scientific progress. Science does not develop in a vacuum, even flawed tools can contribute to its development. We can consider the studies using the ORTO-15 an initial step toward generating new research questions. Although the ORTO-15 demonstrates psychometric deficiencies, its use has contributed to the advancement of measuring the ON construct.

We are convinced that dissemination of stopping mistakes from propagating in the scientific literature must be prevented. One of the solutions could be the discontinuation of the use of the ORTO-15 for assessing the ON because of its notable limitations that reduce its clinical and research utility. The use of an unreliable tool constitutes a threat to our understanding of ON [10]. Therefore, taking into account our earlier arguments, we postulate that the use of the ORTO-15 should be abandoned and should not be propagated in the scientific literature. As researchers, we must avoid basing our conclusions on inappropriate premises. However, we cannot deny that the research using the ORTO-15 has also advanced the conceptualization and understanding of the ON construct. We are not alone in our beliefs. Other researchers also believe that “research based on imperfect measures may still be useful” [11; p. 1]. They claim that Titles such as ‘Research based on invalid measures is invalid’ can be perceived as dismissive of efforts to advance understanding in a complex and evolving field. (…) While we understand the importance of rigour, we believe that such sweeping generalisations do not adequately acknowledge the nuanced contributions of existing research to the field of ON. Hence, we advocate for a more constructive approach that encourages improvement without undermining the progress made thus far [11; p. 2).

Nowadays, we need to redefine ON measurement. Paraphrasing Thomas Edison famously said “I have not failed 700 times. I’ve succeeded in proving 700 ways how not to build a lightbulb”, we can assume which kind of mistake we can avoid by developing the (new) tool assessing ON. We need a reliable instrument that accurately reflects and measures the ON construct. Recently, the first author proposed a screening method to measure ON symptomatology, called the Orthorexia Nervosa Screening Questions (ONSQ) [10]. The seven questions address key features of ON. More precisely, they explore the following aspects: the central role that individual dietary choices play in one’s life and perception of health, the ongoing preoccupation with maintaining a healthy lifestyle, the influence of non-compliance with self-imposed dietary guidelines on an individual’s self-esteem, the influence of these dietary rules on interpersonal relationships and other areas of life, the prioritization of food purity over quantity, the experience of negative feelings related to the dietary rules one considers healthy as well as the impact of these dietary rules on overall mental and physical well-being. These questions are in line with existing diagnostic criteria for ON developed by Dunn and Bratman [12]: (1) obsessive behaviours and mental preoccupation with “healthy” eating, (2) emotional distress resulting from violations of their dietary rules, (3) escalation of dietary restrictions that may come to include elimination of entire food groups and involve progressively more frequent and/or severe “cleanses” regarded as purifying or detoxifying, (4) physical impairments resulting from nutritional deficiencies, (5) psychosocial impairments secondary to beliefs or behaviours about healthy diet and (6) positive body image, self-worth, identity and/or satisfaction being excessively dependent on compliance with self-defined “healthy” eating behaviour. The ONSQ could be a promising tool to identify the risk of ON. However, as it has been solely used in the first author’s therapeutic practice with individuals with disordered eating behaviours, it requires empirical validation.

To date, ON is not a clinically recognized disorder. Studies have reported that comorbidity between ON and other eating disorders (EDs) is associated with psychological conditions, justifying, as a result, the usefulness of assessing and treating this condition [10]. Despite a substantial similarity between ON and EDs there is still no agreement in considering ON an independent ED or a useful convincing concept. So far, a limited number study have detected similarities between ON and EDs on a neuropsychological level revealing that ON is associated with similar cognitive correlates that are seen at the intersection of anorexia nervosa (AN) and obsessive–compulsive disorder (OCD). In addition, ON symptoms were independently associated with set-shifting, self-monitoring, and working memory. ON and AN may possess an overlapping neuropsychological profile marked by deficits in executive function [13]. Both conditions involve an inflexible mindset with rigid rule-bound behaviours and thought preoccupation. Comparable to AN, ON was most strongly associated with impairments in behavioural regulation, including cognitive flexibility, inhibitory control, and emotional control. The contradictory results were also found in the literature demonstrating that although ON is associated with inflexible thoughts and behaviours specific to healthy eating, the condition does not seem to be related to cognitive inflexibility as an executive function deficit [14]. Neuroscience research is needed to investigate and understand the differences and similarities between ON and AN and to determine whether, in addition to the psychopathological (neuro-physiological) level, ON can be considered an (in)dependent ED.

A future research perspective in ON field

Since two decades the literature on ON has burgeoned. The construct of ON helps us to better understand abnormal eating patterns. Looking forward, we see an opportunity to make the coming years the milestone years for ON. Recently, we have indicated the conceptual and methodological points toward ON that have to be addressed in future research [10; Table 1]. First, despite the intensive research on ON in recent years, there is a need to achieve a consensus on adequate terminology for ON. This may help to give the accurate names of various phenomena and thereby differentiate between healthy and pathological eating behaviours and distinguish ON from normative eating behaviours that support a healthy lifestyle and help individuals achieve personal health goals [15]. Another reason justifying a need to update the term is the bi-dimensional structure of orthorexia [16], which includes “both problematic and non-problematic aspects of healthy eating” (p. 288), namely, “healthy orthorexia” and “orthorexia nervosa”. The first one represents a healthy interest in diet, which is independent of psychopathology. While “orthorexia nervosa” refers to a pathological preoccupation with a healthy diet. We wonder how the same term “orthorexia” meaning an interest in “eating right” could indicate on the one hand healthy eating behaviour not associated with a problematic approach to food, and on the other hand an unhealthy eating behaviour associated with a problematic approach to food? If we intend to describe two different eating behaviours we should use two terms having different etymology [10]. We do share the same opinion as Thom Dunn [17, p. 181] does: I believe we need to be very cautious suggesting that there is such thing as “healthy” or “non-pathological” ON (Yakin, Raynal, & Chabrol, 2021; Zickgraf & Barrada, 2021). The concept of pathologically healthful eating comes under attack by individuals who believe its proponents are anti-health. Ross Arguedas (2020), for example, writes that the diagnosis of ON “invites a re-interpretation of healthy dieting practices”. A literature of “healthy orthorexia” gives fodder for these critics and does little to advance the understanding of the topic. There is no such thing as “healthy anorexia,” for example, and suggesting so is absurd. Instead, I believe these scholars really mean “subclinical” ON. That is, somewhat extreme healthful eating that does not lead to malnutrition, nor other daily impairment. This is important to study because I have the sense (anecdotally) that it is prevalent and we need a broader understanding of this represents disordered eating that may become pathological.

Table 1.

Research opportunities and future directions in scientific inquiry in the field of ON

Conceptual issue

• Adequate terminology for ON Updating terminology and definition around ON is essential for accurately differentiating between healthy and pathological eating patterns, as well as distinguishing ON from normative eating behaviours that promote health and help individuals meet their goals A comprehensive conceptual review is necessary to establish a more precise understanding of this phenomenon

• Final nosological classification of ON Empirical evidence is required to elucidate the nature of ON and to identify whether it should be recognized as a distinct nosological entity, classified as a new eating disorder, considered a subtype of anorexia nervosa, or categorized under “other specified feeding or eating disorder" or "unspecified feeding or eating disorder.” • Validated diagnostic criteria of ON The inclusion of ON in the DSM or ICD would mark a significant advancement in both research and clinical practice

Methodological issue

• Universally accepted diagnostic tool for ON A diagnostic assessment tool for ON is crucial for differentiating this condition from normative healthy eating and anorexia nervosa. Such a tool would not only enhance effective evaluation of ON but also significantly advance clinical and research efforts by improving our understanding of the prevalence, risk factors, and treatment approaches for ON

Research issue

• Clinical case reports, observational studies, or longitudinal research To establish ON as a new diagnostic category in the Diagnostic and Statistical Manual of Mental Disorders (DSM) and/or the International Classification of Diseases (ICD), it must be recognized as a disorder (specifically, it should meet the criteria for a mental disorder). In addition, there should be strong evidence supporting its validity and clinical utility, ensuring that it can be applied reliably. A new diagnostic category must also provide significant clinical value, avoid substantial overlap with existing diagnoses or subtypes, and demonstrate a positive benefit-to-harm ratio • Population-based study Further research involving larger populations is essential to generalize findings across diverse groups

• Data sharing—open science Data sharing cultivates a collaborative environment among researchers, bolstering the reproducibility of scientific findings and facilitating a more rapid pace of discovery. In addition, it plays a crucial role in identifying and addressing research gaps, ultimately leading to more comprehensive and impactful results

Based on Brytek-Matera [10]

The use of the term “orthorexia” seems to be etymologically inaccurate, because it does not focus on eating pathology per se or on healthy food preoccupation. Per definition, the word “orthorexia” has been used because of obsessional preoccupation with healthy eating. Therefore, how the word literally meaning “right appetite” could indicate the “pathological fixation on eating proper food”? If we consider that excessive focus on healthy eating (or increasingly obsessive concerns about a healthy diet) and concern about one’s health are pathognomonic symptoms of orthorexia nervosa, we have to change the terminology to more accurately describe this phenomenon. Therefore, a new term “Salussitomania” literally meaning “a very strong interest of health and food” [10; p. 23] was proposed. To sum up, the definition of ON should have been updated and a conceptual review is essential to propose a more precise definition of this phenomenon.

Second, the current state of research does not allow the final classification of ON as a distinct diagnostic category [18]. Further empirical evidence is needed to determine the nature of this condition and whether it should be recognized as a separate nosological entity [19], classified as a new eating disorder [20], considered a subtype of anorexia nervosa [21], or categorized under other specified feeding or eating disorder or unspecified feeding or eating disorder due to its clinical and functional impact, despite lacking recognition as a distinct category [15]. The discussion on the autonomous diagnosis of ON, within or without eating disorders (EDs), remains open. The concept of ON has not been established to this day, although it can be grouped with other EDs. Classification, comparative and/or laboratory research is needed to understand this condition and the differences between ON and other EDs, such as AN and avoidant/restrictive food intake disorder (ARFID). Further studies should examine the underlying mechanisms of ON from a clinical perspective. They could help establish criteria to differentiate ON from other EDs. Etiologic models of abnormal eating behaviour suggest that genetics play a crucial role in its development. The pathogenesis of ON is currently unknown; however, a genetic contribution may be likely to be involved in ON, as with EDs [22]. Up to this point, no genetic research has been conducted in the area of ON. Therefore, research is now conducting by the first author (University of Wroclaw, grant number: BPIDUB.38.2025) on the genetic polymorphisms involved in the pathophysiology of ON, knowing that genetic polymorphisms may confer susceptibility to the onset, progress, and development of AN [23]. Orthorexia Nervosa not only overlaps with EDs but also shares connections with meat-free diets [24] motivated by personal health and/or ethical reasons (such as those focused on animal welfare or environmental preservation). Vegetarians may score higher on certain eating behaviour assessments, which could be skewed by items that address avoiding specific foods normative to their diet [25]. Their high scores may reflect challenges in finding suitable vegetarian options within existing ON assessment tools rather than genuine ON tendencies. Third, although several diagnostic proposals for ON have been developed, standardized criteria are still lacking. Establishing validated diagnostic criteria would significantly advance clinical and research efforts by improving the understanding of its prevalence, risk factors, and treatment approaches.

Fourth, although various questionnaires are available one universally accepted measure of ON should be indicated. Nowadays, two questionnaires, namely, the EHQ and the DOS, seem to be reliable measures to assess ON symptomatology; therefore, we should consider using them in the research. It should be also noted that a new version of the ORTO-15, namely, ORTO-R, was published more recently [26] with a significant improvement in psychometric properties [27]. Perhaps recently proposed a 13-item Lebanese Orthorexia Nervosa Inventory (LONI; Cronbach’s α value of 0.90 and a McDonald’s ω value of 0.90) [28] and/or the Intra- and Interpersonal Effects Scale of Orthorexia (IIESO; [29]) measuring the severity of orthorexic behavioural effects may create a new alternative for assessing ON.

Fifth, our complex understanding of orthorexia nervosa should be based on clinical case reports, observational studies, or longitudinal research, not just on the currently predominant cross-sectional studies. If we are interested in including ON to the Diagnostic and Statistical Manual of Mental Disorders (DSM) or International Classification of Diseases (ICD) as a disorder, distinct from or closely associated with EDs that requires recognition and treatment, we must ensure that it meets several important criteria. Specifically, a new diagnostic category included in the classification system must be recognized as a mental disorder (has to meet criteria for a mental disorder), have strong evidence supporting its validity or clinical utility, be applicable in a reliable manner, manifest significant clinical value, avoid substantial overlap with existing diagnoses or subtypes, and demonstrate a positive benefit-to-harm ratio [30].

Sixth, research on ON among college and university students is rapidly increasing. However, these studies limit the generalizability of findings to the broader population. Therefore, further research involving larger populations is necessary to generalise findings across different groups. Seventh, researchers should also consider making their data freely available on platforms, such as the Open Science Framework, Science Data Bank, or Zenodo, to support comparative research efforts, collect data widely in their countries, and build partnerships with researchers [31].

We recognise that a significant next step in the field of ON is needed, especially in light of the knowledge gained in the last 20 years. Advances in research have highlighted the complexities of ON, including its diagnostic challenges, overlapping features with other disorders, and the need for standardised assessment tools. It is crucial to refine the conceptualization of ON, establish validated diagnostic criteria, and develop effective tools to enhance both scientific understanding and clinical practice.

Author contributions

ABM: Conceptualisation, writing manuscript, revising, reviewing and editing. LMD: reviewing. The authors have approved the final manuscript.

Funding

This study was funded by Excellence Initiative—Research University” programme of the University of Wroclaw, Poland, BPIDUB.38.2025.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

Formal consent is not required for this type of research.