Elsevier

Appetite

Volume 108, 1 January 2017, Pages 303-310
Appetite

Orthorexia nervosa: Assessment and correlates with gender, BMI, and personality

https://doi.org/10.1016/j.appet.2016.10.021Get rights and content

Abstract

This study investigated whether orthorexia nervosa (ON; characterized by an obsessive fixation on eating healthy) may be predicted from the demographics variables of gender and BMI, and from the personality variables of self-esteem, narcissism, and perfectionism. Participants were 459 college students, who completed several online questionnaires that assessed these variables. A principal components analysis confirmed that the Eating Habits Questionnaire (Gleaves, Graham, & Ambwani, 2013) assesses three internally-consistent ON components: healthy eating behaviors, problems resulting from those behaviors, and positive feelings associated with those behaviors. A MANOVA and its tests of between subjects effects then revealed significant interactions between gender and BMI, such that for men but not women, a higher BMI was associated with greater symptomatology for all ON components. Partial correlation analyses, after controlling for gender and BMI, revealed that both narcissism and perfectionism were positively correlated with all aspects of ON symptomatology.

Introduction

The benefits of healthy eating on physical and mental health are seemingly endless. In particular, eating a healthy vegetarian or semi-vegetarian diet (i.e., a nutrient-rich diet that consists largely of fruits and vegetables) may decrease one's risk of such life-threatening diseases as cancer and cardiovascular disease (Bazzano, 2006, Trichopoulou et al., 2003, Van Duyn and Pivonka, 2000). Moreover, this kind of diet is also associated with an enhanced quality of life with greater life satisfaction (Blanchflower et al., 2012, Grant et al., 2009, Lengyel et al., 2009), greater happiness (Blanchflower et al., 2012, Piqueras et al., 2011, White et al., 2013), greater self-esteem (Brug et al., 1995, Elfhag et al., 2008, Steptoe et al., 2003), greater optimism (Boehm et al., 2013, Giltay et al., 2007, Kelloniemi et al., 2005), and a lower incidence of depression (Jacka et al., 2010, Kulkarni et al., 2015, McMartin et al., 2013). Ironically, however, healthy eating may become unhealthy for select individuals who develop orthorexia nervosa (ON), and the aforementioned benefits become replaced with a host of negative physical and mental health consequences.

ON, characterized by obsessive fixation on eating healthy, was first introduced by Steven Bratman (1997), an M.D. with practices in alternative medicine and occupational medicine. According to Bratman and Knight (2000), ON begins innocently, often with a desire to treat or prevent an illness (e.g., asthma, arthritis, cancer, heart disease, diabetes), to lose weight, or to simply break free from some of the bad habits of the typical American diet that often includes foods high in sugar, fat, and various artificial preservatives and chemicals. Over time, however, this healthy eating may transform itself into an unhealthy obsession for some individuals, whereby the time spent planning, purchasing, preparing, and eating their meals becomes extensive, consumes them, and interferes with other aspects of their life. These select individuals will self-praise successful resisting of temptation, often feeling superior to others, and any dietary lapses will be followed by extreme feelings of guilt and self-condemnation. Quality of life will diminish as they forego both intimate and larger social gatherings that involve food, socially isolating themselves and experiencing corresponding feelings of depression and anxiety. At an extreme, if they become too restrictive in the types of foods that they allow themselves to consume, malnutrition and even death may result.

Past research on ON has largely focused on its association with disordered eating, obsession, and lifestyle. Regarding disordered eating, multiple studies reveal that increased ON symptomatology corresponds to higher scores on questionnaires assessing disordered eating behaviors that include items related to calorie restriction and preoccupation with weight (Asil and Surucuoglu, 2015, Fidan et al., 2010, Gleaves et al., 2013, McInerney-Ernst, 2011, Segura-Garcia et al., 2012, Segura-Garcia et al., 2015). Although these findings suggest that ON is similar to anorexia nervosa (AN), the reasons for such eating behaviors may differ, with AN individuals trying to overcome a negative body image and ON individuals trying to lose excess weight that is associated with negative health risks. Regarding obsession, studies reveal that increased ON symptomatology corresponds to greater obsessive-compulsive tendencies (Arusoglu et al., 2008, Asil and Surucuoglu, 2015, Gezer and Kabaran, 2013, Gleaves et al., 2013, Koven and Senbonmatsu, 2013). These findings are expected, considering the ON individuals' obsession with healthy eating that extends to what, when, and how they eat. Regarding lifestyle, consistent with ON individuals' pursuit of good physical health, research reveals that increased ON symptomatology corresponds to increased exercise frequency (Eriksson, Baigi, Marklund, & Lindgren, 2008) and sport participation (Segura-Garcia et al., 2012, Varga et al., 2014), and is not associated with cigarette smoking or alcohol use (Aksoydan and Camci, 2009, McInerney-Ernst, 2011, Varga et al., 2014).

Research on the personality correlates of ON is limited to two studies that investigated the relation between ON and the big five personality factors, and both found that increased ON symptomatology corresponded to higher levels of neuroticism (Forester, 2014, Gleaves et al., 2013). Neuroticism is a trait with a tendency toward a negative emotional state that encompasses feelings of depression, anxiety, and anger. Consistent with the finding above, increased ON symptomatology has been found to correspond to increased depression, negative affect, and suicidal thinking (Gleaves et al., 2013). Given that ON becomes an unhealthy time-consuming obsession that interferes with other aspects of their life and leads to social isolation along with feelings of guilt and self-condemnation for dietary lapses, feelings of depression are not surprising.

The present research further explores the personality correlates of ON, by investigating self-esteem, perfectionism, and narcissism. Gatward (2007) proposed that food restriction in AN is a response to perceived threats of social exclusion, and the same may be true for at least the initial food restriction in ON. Given the uncommonly thin ideals that are glamorized in mass media, individuals who are low in self-esteem may feel extra pressure to become thin and conform to the culture's ideals of beauty. Supporting this statement is research revealing that individuals with AN also suffer from lower self-esteem than controls (Brockmeyer et al., 2013, Hartman et al., 2014). Furthermore, in an attempt to overcome their low self-esteem and vulnerable ego, some may pursue perfection and create a deceptive allure of self-assurance, consistent with the finding that individuals with AN score higher than controls on measures of both perfectionism and narcissism (Steiger et al., 1997, Waller et al., 2007). These studies have shown that losing weight causes AN sufferers to feel power and superiority, causing the continuation of the disorder. Similar findings would be expected in ON individuals who take pride in their self-control, resisting temptation, and eating foods that they believe are superior to others' diets.

Two demographic variables that have been repeatedly linked to disordered eating, and should thus be considered in the present research too, are gender and body-mass index (BMI). Regarding gender differences in disordered eating, past research reveals that women outnumber men in prevalence of AN with an average ratio of 9:1 (Fisher et al., 2014, Forman et al., 2014, Nicely et al., 2014, Norris et al., 2014). Any gender differences in ON are more difficult to detect, partially due to the lack of research on clinically diagnosed individuals, with ON not being recognized as a disorder in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5; American Psychiatric Association, 2013). The existing research typically compares men and women on the basis of their scores on a single ON questionnaire. Three of these studies found that ON symptomatology was significantly greater in women than men (Arusoglu et al., 2008, Keller and Konradsen, 2013, Koven and Senbonmatsu, 2013), two studies found that ON symptomatology was significantly greater in men than women (Donini et al., 2004, Fidan et al., 2010), and seven studies found no significant gender differences in ON symptomatology (Aksoydan and Camci, 2009, Bosi et al., 2007, Brytek-Matera et al., 2015, Lewis, 2012, McInerney-Ernst, 2011, Ramacciotti et al., 2011, Valera et al., 2014). Taken together, these findings may suggest no or minimal gender differences, which is in contrast to AN. However, given the mixed results, further research is needed to clarify any potential differences.

Regarding BMI, part of being physically healthy is maintaining an appropriate body weight. Two studies have found that increased ON symptomatology corresponds to greater BMI (Asil and Surucuoglu, 2015, Fidan et al., 2010). Although others studies found no significant relationship, the trend of increased ON symptomatology corresponding to greater BMI held (Aksoydan and Camci, 2009, Donini et al., 2004, McInerney-Ernst, 2011, Ramacciotti et al., 2011, Varga et al., 2014). Perhaps ON first develops out of an attempt by overweight individuals to lose weight, such that these individuals have an accurate rather than distorted body image and turn to healthy eating in an attempt to achieve and maintain an ideal weight for proper physical health. As with gender, these mixed findings create the need for further research into the impact of BMI on ON. Nevertheless, the fact that no study has found a negative correlation between BMI and ON suggests that ON is distinct from AN, which is marked by an excessively low BMI.

The vast majority of research studies have assessed ON with the ORTO-15 (Donini Marsili, Graziani, Imbriale, & Cannella, 2005). Donini et al. first diagnosed participants as ON or non-ON on the basis of an extensive battery of the healthy foods they ate, the unhealthy foods they avoided, and their obsessive-compulsive personality. Participants then completed the ORTO-15, a 15-item questionnaire based off of the self-test proposed by Bratman and Knight (2000), with lower scores representing greater ON symptomatology. This questionnaire demonstrated the greatest predictive capability with a threshold of 40, such that a score below 40 would indicate the presence of ON. However, this predictive capability was still limited. Of the 27 individuals previously diagnosed with ON using the more comprehensive means, only 56% would have been diagnosed with ON using the ORTO-15. Further, of the 366 individuals not diagnosed as ON using the more comprehensive means, 28% would have been misdiagnosed with ON using the ORTO-15.

The validity and reliability of the ORTO-15 have varied quite drastically across different research studies. For example, prevalence rates using a threshold of 40 have ranged from a reasonable 4% in Donini et al.'s (2005) original study to an extreme 86% (Aksoydan and Camci, 2009, Asil and Surucuoglu, 2015, Bosi et al., 2007, Koven and Senbonmatsu, 2013, Lewis, 2012, McInerney-Ernst, 2011, Meister, 2010, Ramacciotti et al., 2011, Valera et al., 2014, Varga et al., 2014). The mean prevalence rate for these studies combined was 55% (SD = 24%), which seems unreasonably high for a debilitating condition. Moreover, Cronbach's alpha as a measure of internal consistency for the ORTO-15 has ranged from an unacceptable 0.14 to an acceptable 0.83 (Fidan et al., 2010, Koven and Senbonmatsu, 2013, Lewis, 2012, McInerney-Ernst, 2011, Meister, 2010, Varga et al., 2014). The mean alpha for those studies combined was 0.55 (SD = 0.27), which is below the standard 0.7 for research studies and 0.9 for diagnostic uses. No information is available on test-retest reliability.

The questionable validity and reliability of the ORTO-15 may be due to multiple factors. First, the original ON diagnosis was based on being a health fanatic with an obsessive-compulsive personality, whereas an additional key component of ON is a negative impact on one's life in terms of physical or mental health. Additionally, some questions may apply more to AN than to ON (e.g., “When eating, do you pay attention to the calories of the food?”), some questions may apply more to thoughts than behaviors (e.g., “Do you think that consuming healthy food may improve your appearance?”), some questions may not be relevant to all geographic regions (e.g., “Do you think that on the market there is also unhealthy food?”), and some questions may simply be confusing in meaning (e.g., “When you go in a food shop do you feel confused?”).

Given the limited validity and reliability of data from the ORTO-15, an alternative measure of ON is needed. Independent of the development of the ORTO-15, Gleaves et al. (2013) developed a 21-item Eating Habits Questionnaire (EHQ). They began with an initial pool of 160 items, which included the items from Bratman and Knight's (2000) self-test. Of these 160 items, 59 were unanimously agreed upon as representing ON by four out of four ON-trained graduate students in clinical psychology. This questionnaire was administered to 174 undergraduates. Exploratory factor analyses revealed three factors (healthy eating behaviors, problems associated with healthy eating, and feeling positively about healthy eating), each with at least four items with factor loadings greater than 0.60 representing stable components. Of the 59 items, 24 were deleted based on similar loadings on multiple factors, theoretical inconsistency of item content, or lack of interpretation ease. The remaining 35 items were administered to a second sample of 213 undergraduates. Based on the modification indices and standardized residual scores from confirmatory factor analyses, researchers eliminated 14 items that were redundant and overlapped in content with other items. Fit indices of the final 21-item EHQ were good (GFI = 0.85, TLI = 0.90, CFI = 0.91, RMSEA = 0.07). Gleaves et al. subsequently presented additional reliability and validity data for this final 21-item scale (see the Method section below for a summary of these findings), revealing great promise for the EHQ as a measure of ON symptomatology. However, to the best of the current authors' knowledge, no other publications include the use of the EHQ measure beyond that one study by Gleaves et al. More research with further reliability and validity data is needed.

The current study has three goal sets. First, it assesses the factor structure and internal consistency of the EHQ that has been used in the United States with promising results (Gleaves et al., 2013). Consistent with that original study, we predict that EHQ data in the current study will further demonstrate three subscales (behaviors, problems, and feelings), each with high internal consistency. Second, it explores the ON correlates with gender and BMI. Regarding gender, the preceding literature review indicates that three studies found that ON symptomatology was significantly greater in women than men, two found the opposite, and seven found no significant gender differences; based on these findings taken together, we predict no or minimal gender differences in the current study. Regarding BMI, the preceding literature review indicates that two studies found that increased ON symptomatology corresponds to greater BMI, and five found the same but insignificant trend; based on these findings taken together, we predict greater ON symptomatology with greater BMI. Given the mixed findings from past research on gender and BMI differences, perhaps the relation of ON to BMI depends on gender, a question that the current study will also address; we have no a priori directional predictions for this interaction. Third, the current study explores the ON correlates with the personality traits of self-esteem, narcissism, and perfectionism. No studies have yet investigated these relationships. However, we may reasonably predict that ON symptomatology is positively correlated with each of these variables if high-ON individuals take pride in their self-control, resisting temptation, and eating a near-perfect diet that they believe is superior to others' diets.

Section snippets

Participants

This study's participants were 459 students (80.8% women, 19.2% men) enrolled in undergraduate psychology courses at a large university in the southern region of the United States. Their ages ranged from 16 to 48 years (M = 19.85, SD = 2.79). Based on self-reported ethnicity, 38.1% were European American, 35.9% were Hispanic or Latino, 17.0% were African American, 3.1% were Asian American, 3.5% were bi-ethnic, and 2.4% were of another ethnicity. All participants received a small amount of

Assessment of ON

As shown in Table 1, results of the principal components analysis with varimax rotation revealed the same three factors (i.e., three principal components with eigenvalues greater than 1) as the original study (Gleaves et al., 2013): healthy eating behaviors (EHQ-Behaviors), problems associated with healthy eating (EHQ-Problems), and feeling positively about healthy eating (EHQ-Feelings). The only difference was that three of the items that initially loaded onto the EHQ-Problems subscale in the

Assessment of ON

This study first sought to assess whether data in this study would support the use of the EHQ (Gleaves et al., 2013) in other studies measuring ON symptomatology. As research investigating ON becomes more prevalent, the need for a good measure of ON symptomatology likewise becomes increasingly more important. While the vast majority of past research used the ORTO-15 (Donini et al., 2005), the internal consistency of that measure has been greatly questioned with a mean Cronbach's alpha of 0.55 (

Conclusions

In summary, the current research found that the EHQ produced reliable and valid data for measuring ON symptomatology, that high BMI is associated with greater ON symptomatology for men, and that narcissism and perfectionism are associated with greater ON symptomatology for both men and women. Future research should investigate these relationships with individuals from different regions of the country and with different ages represented, and should investigate the relationships between ON

References (65)

  • S.B. Murray et al.

    A comparison of eating, exercise, shape, and weight related symptomatology in males with muscle dysmorphia and anorexia nervosa

    Body Image

    (2012)
  • H.-L. Pai et al.

    Parental perceptions, feeding practices, feeding styles, and level of acculturation of Chinese Americans in relation to their school-age child's weight status

    Appetite

    (2014)
  • M.L. Reyes-Rodriguez et al.

    ‘Las penas con pan duelen menos’: The role of food and culture in Latinas with disordered eating behaviors

    Appetite

    (2016)
  • M.A.S. Van Duyn et al.

    Overview of the health benefits of fruit and vegetable consumption for the dietetics professional: Selected literature

    Journal of the American Dietetic Association

    (2000)
  • E. Aksoydan et al.

    Prevalence of orthorexia nervosa among Turkish performance artists

    Eating and Weight Disorders

    (2009)
  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders

    (2013)
  • G. Arusoglu et al.

    Orthorexia nervosa and adaptation of ORTO-11 into Turkish

    Turkish Journal of Psychiatry

    (2008)
  • E. Asil et al.

    Orthorexia nervosa in Turkish dieticians

    Ecology of Food and Nutrition

    (2015)
  • D.G. Blanchflower et al.

    Is psychological well-being linked to the consumption of fruit and vegetables?

    Social Indicators Research

    (2012)
  • J.K. Boehm et al.

    Association between optimism and serum antioxidants in the Midlife in the United States Study

    Psychosomatic Medicine

    (2013)
  • A.T.B. Bosi et al.

    Prevalence of orthorexia nervosa in resident medical doctors in the faculty of medicine

    Appetite

    (2007)
  • S. Bratman

    Health food junkie: Obsession with dietary perfection can sometimes do more than good, says one who has been there

    Yoga Journal

    (1997)
  • S. Bratman et al.

    Health food junkies: Overcoming the obsession with healthful eating

    (2000)
  • T. Brockmeyer et al.

    The thinner the better: Self-esteem and low body weight in anorexia nervosa

    Clinical Psychology and Psychotherapy

    (2013)
  • A. Brytek-Matera et al.

    Orthorexia nervosa and self-attitudinal aspects of body image

    Journal of Eating Disorders

    (2015)
  • L.M. Donini et al.

    Orthorexia nervosa: A preliminary study with a proposal for diagnosis and an attempt to measure the dimension of the phenomenon

    Eating and Weight Disorders

    (2004)
  • L.M. Donini et al.

    Orthorexia nervosa: Validation of a diagnosis questionnaire

    Eating and Weight Disorders

    (2005)
  • K. Elfhag et al.

    Consumption of fruit, vegetables, sweets and soft drinks are associated with psychological dimensions of eating behaviour in parents and their 12-year-old children

    Public Health Nutrition

    (2008)
  • L. Eriksson et al.

    Social physique anxiety and sociocultural attitudes toward appearance impact on orthorexia test in fitness participants

    Scandinavian Journal of Medicine and Science in Sports

    (2008)
  • D.S. Forester

    Examining the relationship between orthorexia nervosa and personality traits

    (2014)
  • R.O. Frost et al.

    The dimensions of perfectionism

    Cognitive Therapy and Research

    (1990)
  • N. Gatward

    Anorexia nervosa: An evolutionary puzzle

    European Eating Disorders Review

    (2007)
  • Cited by (0)

    View full text