Introduction

In recent years, orthorexia nervosa (ON) has emerged as a new pattern of disordered eating behaviors characterized by preoccupations related to diet quality and health concerns, rather than drive by weight and shape concerns [6]. ON is not yet established as a formal diagnosis, and indeed despite ongoing efforts to refine its characterization, definition, and assessment. However, a growing body of cross-sectional empirical data has documented associations between ON symptomatology and other indicators of disordered eating, although findings are not consistent across studies and may vary according to definitions and measures [6, 10]. So far, little attention has been paid to the potential relationship between ON behaviors and indicators of healthy eating or positive eating behaviors that may also involve an attention to diet quality. Thus, the aim of this study was to examine the relationships among ON behaviors and indices of healthy eating, namely intuitive eating [21] and eating competence [17].

One of the distinguishing features of ON, and in contrast with other types of restrictive eating, is the focus on food quality and eating for health reasons. Although the definition of ON includes the notion that the extent of the preoccupations and the resulting restrictive diet is impairing, given legitimate concerns regarding food safety and availability of fresh foods in the contemporary food environment [14, 16, 24], the types of concerns that underpin ON are not entirely unfounded. Thus, while when taken to extremes these preoccupations might be harmful and impairing, it is also possible that at lower levels, such concerns may promote food choices that can support a health-promoting diet and healthy eating. To date, the literature is largely silent on this subject.

Preliminary evidence for ON symptoms to be associated with positive dietary intake indices has emerged. For example, among a sample of undergraduates from Greece, those with higher levels of self-reported ON (although based on a tool with psychometric limitations) were less likely to consume saturated fats and, among male students, higher scores were associated with greater intake of fruit and vegetables [7]. Similarly, among German adults, orthorexic behaviors were found to be associated with adherence to the Mediterranean diet, which has been described as being related to a number of health benefits [19]. Such studies, however, have focused on the exploration of diet quality and intake, rather than broader positive eating attitudes and patterns.

Two models of healthy eating that have previously received empirical support are the intuitive eating model [21] and eating competence model [17]. Intuitive eating is grounded in the benefits of eating in response to physical cues of hunger and satiety rather than following rigid food rules, eating in response to emotional or external cues, or using eating behaviors as a means to modify weight and shape [21]. Intuitive eating has been shown to be negatively associated with a number of indices of disordered eating including dieting and a variety of unhealthy weight control behaviors and in contrast is related to positive indices of body image and emotional well-being and functioning [3]. The most well-established measure of intuitive eating [22] distinguishes four separate facets: (1) Eating for Physical Rather Than Emotional Reasons; (2) Unconditional Permission to Eat; (3) Reliance on Hunger and Satiety Cues; and (4) Body–Food Congruence, a dimension that focuses on appropriate and adaptive nutrition.

The eating competence model [17] is similarly based on eating in response to internal cues. However, it also encompasses a focus on the capacity to eat a varied diet and to be flexible and planful regarding food choices and preparation. Thus, both this model and intuitive eating are models of eating regulation, while the eating competence model also considers some of the aspects related to the food environment such as food availability and preparation [17]. The eating competence model includes four distinct facets: (1) Eating Attitudes, which describes a relaxed and comfortable attitude toward enjoying food and eating; (2) Food Acceptance, which focuses on diet flexibility and variety; (3) Food Regulation; and (4) Contextual Skills, which evaluates the capacity to integrate eating competence into daily life. Thus, both of these models are interesting to consider in relation to ON.

The aim of the present study was therefore to evaluate the relationship between ON symptomatology and the different facets of intuitive eating and eating competence among a sample of female and male college students. Given that, in the USA, emerging adults are often in a position of greater independence related to food choices once they enter college, as well as this being a critical time for the establishment of sustainable health-related behaviors; this is an important group to focus on [12]. Given the restrictive nature of ON and its documented associations with disordered eating [6], it was hypothesized that overall higher levels of ON would be negatively associated with dimensions of intuitive eating and eating competence. However, given the focus on eating for health reasons in ON, it was also hypothesized that a curvilinear relationship between ON behaviors and intuitive eating and eating competence might emerge, such that individuals with the lowest levels of ON behaviors might reveal lower levels of intuitive eating and eating competence than those with low-to-moderate levels of ON behaviors. In addition, given the overall lack of data exploring ON among men, as well as previously documented differences in intuitive eating, we explored these relationships among each gender separately [22].

Methods

Participants and procedure

Six hundred and ten undergraduate students between the ages of 18–25 and enrolled at a large, private university in the northeastern USA were recruited via flyers and online advertisements to complete an online survey on the Qualtrics survey platform. They were offered the opportunity to enter into a raffle for one of five $100 gift cards. Five participants were excluded due to large amounts of missing data resulting in a final sample of n = 605, 19% male, mean (SD) age = 19.84 (1.93) years. The study was approved by the University Institutional Review Board and participants provided consent by opting into the online survey. The survey was confidential to the researchers.

Measures

Demographic data

Participants described their year and area of study, and their parent’s highest levels of education as a proxy for socioeconomic status, as well as any chronic illnesses. In addition, they provided self-reported height and weight to calculate BMI.

Orthorexia behaviors

Orthorexia behaviors were assessed using the ORTO-7 [11], a 7-item version of the original 15-item scale [5]. The scale assesses attitudes and behaviors associated with orthorexia, including attention to calorie content, preoccupation with food and its impact on health, and the impact of food rules on psychosocial functioning, for example, “Are your eating choices conditioned by your worry about your health status?”. Items are scored on a 4-point scale from 1 (Always) to 4 (Never). Here, descriptive statistics are presented with the original scoring method, whereby higher scores indicate lower levels of concerns. However, scale scores were reversed for the bivariate and multivariate analyses such that higher scores indicated higher levels of orthorexia attitudes and behaviors for ease of interpretation. In the original sample, the ORTO-7 demonstrated good psychometric properties with the single-factor structure revealing a good fit to the data, GFI = 0.97, CFI = 0.96, RMSEA = 0.06, superior to the other models proposed to date that they tested first. In addition, good internal reliability emerged with α = 0.83 [11]. Moller et al. [11] proposed a cutoff score of 19 or lower (using the original scoring) to represent high levels of symptomatology and probable orthorexia nervosa. In the current sample, α = 0.76 and α = 0.70 in women and men, respectively.

Intuitive eating

Intuitive eating was assessed using the Intuitive Eating Scale-2 [22] that includes 23 items scored on a 5-point scale from 1 (strongly disagree) to 5 (strongly agree, distributed across four subscales. The subscales include Unconditional Permission to Eat, e.g., “If I am craving a certain food, I allow myself to have it”; Eating for Physical Rather Than Emotional Reasons, e.g., “When I am bored, I do NOT eat just for something to do”; Reliance on Hunger and Satiety Cues, e.g., “I trust my body to tell me when to eat”; and Body–Food Choice Congruence, e.g., “Most of the time, I desire to eat nutritious foods.” For all subscales, higher scores indicate higher levels of the dimension. In our sample, internal reliability was high among both women and men on all subscales: Unconditional Permission to Eat, α = 0.77 in both women and men; Eating for Physical Rather Than Emotional Reasons, α = 0.84 and 0.80 for women and men, respectively; Reliance on Hunger and Satiety Cues, α = 0.86 and 0.89 for women and men, respectively; and Body–Food Choice Congruence, α = 0.83 for both women and men.

Eating competence

The ecSatter Inventory [9] includes 16 items that assess eating patterns characterized by flexibility, internal regulation, and positive attitudes. Items are scored on a 5-point scale from 0 (Never) to 5 (Always), recoded to range from 0 to 3, that form four subscales, namely Eating Attitudes (e.g., “I am relaxed about eating”), Food Acceptance (e.g., “I eat a wide variety of food”), Food Regulation (e.g., “I eat as much as I am hungry for”), and Contextual Skills (e.g., “I make time to eat”). In the current sample, internal reliability was high for all subscales with Eating Attitudes α = 0.83 and α = 0.81, Food Acceptance α = 0.74 and α = 0.71, Food Regulation α = 0.78 and α = 0.80, Contextual Skills α = 0.79 and α = 0.80 for women and men, respectively.

Data analyses

Descriptive statistics and bivariate associations were first examined. To examine the presence of linear and quadratic relationships between ON behaviors and indices of healthy eating, hierarchical regression analyses were conducted among each gender separately, with ON behaviors entered first, followed by its quadratic term (that is, the squared value of the predictor). All analyses were conducted using SPSS 26.

Results

Descriptive statistics and bivariate relationships

The majority of participants were in their first (38%) or second (21%) year of college, with the remainder distributed somewhat equally across the 3rd, 4th, and 5th year. Just under a third (30%) of participants described being enrolled in a health-related major, 20% in natural sciences, 23% in business or political science, 17% in engineering, computing, or data sciences, 7% in humanities, and the remainder were undeclared. Mean (SD) BMI in the sample was 22.74 (3.39). A small proportion described having a chronic illness (12%). Approximately half of these (n = 35, 6%) were illness that might affect diet including celiac disease (n = 10), diabetes (n = 9), and anemia (n = 8).

Descriptive statistics for all study variables are presented in Table 1. Over half (59%) of the female and just under half (47%) of the male participants scored above the cutoff score established by Moller et al. [11] for likely orthorexia nervosa. Gender differences emerged on two dimensions of intuitive eating, with men scoring higher than women on Eating for Physical Rather Than Emotional Reasons and Reliance on Hunger and Satiety Cues. In addition, gender differences emerged on three of the four dimensions of eating competence, with men reporting higher levels of positive Eating Attitudes, Food Acceptance, and Food Regulation.

Table 1 Descriptive data, gender differences, and bivariate correlations
Full size table

Among women, higher levels of ON behaviors were significantly associated with lower levels of three of the intuitive eating subscales, namely Unconditional Permission to Eat, Eating for Physical Rather Than Emotional Reasons, and Reliance on Hunger and Satiety Cues, with small-to-large effect sizes (Table 1). Regarding eating competence, higher levels of ON behaviors were strongly associated with less positive Eating Attitudes with a large effect size, moderately associated with lower levels of Food Regulation, and weakly related to lower Contextual Skills. Among men, higher levels of ON behaviors were significantly associated with lower scores on the same three intuitive eating subscales, with small-to-large effect sizes. Regarding eating competence, higher levels of ON behaviors were associated with less positive Eating Attitudes and lower Food Regulation, with moderate effect sizes, but were not related to Food Acceptance or Contextual Skills.

For sensitivity purposes, the analyses were conducted removing all participants who reported having a chronic disease that might affect their diet. The pattern of associations was identical; therefore, these participants were retained. Second, the analyses were conducted controlled for BMI. Again, here the patterns of relationships were almost identical; therefore, the analyses were pursued without controlling for BMI.

Examination of linear and quadratic relationships

Findings among women

A summary of the findings among women is presented in Table 2. Regarding intuitive eating, findings from the regression analyses revealed that while ON behaviors were significant predictors of lower Unconditional Permission to Eat, when entered subsequently, the quadratic term was not significant. Similar patterns emerged for Eating for Physical rather Than Emotional Reasons and Reliance on Hunger and Satiety Cues. However, for Body–Food Choice Congruence, consistent with the bivariate correlations, among women neither ON behaviors nor the quadratic term emerged as significant predictors.

Table 2 Regression models among women
Full size table

Regarding eating competence, findings from the regression analyses revealed that ON behaviors were significant predictors of less positive Eating Attitudes, F(2, 546) = 177.39, p < 0.001, β = − 0.62, p < 0.001, lower Food Acceptance F(2, 546) = 4.84, p = 0.008, β = − 0.102, p = 0.018, lower Food Regulation, F(2, 546) = 71.35, p < 0.001, β = − 0.45, p < 0.001, and lower Contextual Skills, although this last model just failed to meet significance, F(2, 546) = 2.89, p = 0.056, β = − 0.102, p = 0.019. When entered subsequently, the quadratic term was not significant in any of the models.

Findings among men

A summary of the findings among men is presented in Table 3. Regarding intuitive eating, findings from the regression analyses revealed that while ON behaviors were significant predictors of lower Eating or Physical Rather than Emotional Reasons, when entered subsequently, the quadratic term was not significant. However, for the other three facets of intuitive eating a different pattern emerged. Thus, ON scores significantly predicted lower Reliance on Hunger and Satiety Cues. However, when entered in the next block, the quadratic term became the independent predictor suggesting that the relationship between ON scores and Reliance on Hunger and Satiety Cues is concave (n shaped), such that initially increasing levels of ON scores are associated with higher Reliance on Hunger and Satiety Cues, but that this relationship then inverts with higher levels of ON behaviors associated with lower Reliance on Hunger and Satiety Cues. Moreover, regarding Body–Food Choice Congruence, ON scores were not a significant predictor. However, when entered in the next block, the quadratic term was an independent predictor, suggesting that again here the relationship between ON scores and Body–Food Choice Congruence is concave (n shaped), such that initially increasing levels of ON behaviors are associated with higher Body–Food Choice Congruence, but that this relationship then inverts with higher levels of ON behaviors associated with lower Body–Food Choice. Finally, for Unconditional Permission to Eat, ON behaviors emerged as significant linear predictors of lower Unconditional Permission to Eat. When entered into the second block, the quadratic term just failed to meet significance. However, interestingly, the beta coefficient was positive suggesting a convex (u shaped) relationship, whereby the lowest and highest levels of ON are associated with high levels of Unconditional Permission to Eat, but mid-range scores are associated with low levels of Unconditional Permission to Eat.

Table 3 Regression models among men
Full size table

Regarding eating competence, findings from the regression analyses revealed that ON behaviors were significant predictors of less positive Eating Attitudes and lower lower Food Regulation; however, they did not predict Food Acceptance or Contextual Skills. When entered subsequently, the quadratic term was not significant in any of the models.

Discussion

The aim of the present study was to evaluate the relationship between ON symptomatology and the different facets of intuitive eating and eating competence. Overall, higher levels of ON symptomatology were associated with lower levels of positive eating indices among both female and male college students. These relationships emerged most strongly for the Eating Attitudes scale and Food Regulation facets of the eating competence scale, as well as for the Unconditional Permission to Eat facet of intuitive eating. Thus, our findings suggest that ON symptomatology may be particularly associated with difficulties related to food rules and restriction driven by anxiety related to certain foods. Such a finding is consistent with conceptualizations of ON as restrictive disorder in which eating is largely regulated by avoidance and preoccupation and may be accompanied by substantial weight loss due to restriction of intake [6]. These patterns were found among both college women and men, suggesting that these relationships are similarly present for both genders despite differences in overall levels.

However, in addition to the overall negative relationships found between ON symptomatology and positive eating patterns, among men some evidence emerged for curvilinear relationships between ON symptomatology and two facets of intuitive eating, namely Reliance on Hunger and Satiety Cues and Body–Food Choice Congruence, such that mid-range levels of ON symptomatology were associated with the highest levels of these two aspects of intuitive eating. This is an interesting finding as it supports the idea that, to some extent and at levels that are not rigid or impairing, concerns related to the quality of foods might be associated with positive eating outcomes. Thus, investment in healthy eating might to some extent be associated with food choices that are supportive for health as conceptualized by these two facts of intuitive eating [22]. However, consistent with findings from other areas, at high levels restrictive eating and rigid food rules, regardless of their focus, are associated with distress and impairment [1]. In contrast, no curvilinear relationship emerged regarding the other two facets of intuitive eating, namely Eating for Physical Rather Than Emotional Reasons and Unconditional Permission to Eat, nor any of the facets of eating competence. It is noteworthy that these latter aspects of positive eating are characterized by flexible attitudes toward eating a variety of foods and not restricting intake based on any criteria [8]. Given that restrictive eating behaviors, albeit due to health preoccupation, is one of the hallmark features of ON symptomatology [6], it is perhaps unsurprising that these facets did not reveal such curvilinear relationships. In addition, no evidence of curvilinear relationships emerged among women, highlighting again how overall women experience eating-related expectations and preoccupations, even when allegedly health-oriented, are not associated with positive outcomes.

In addition to these differences in the patterns of relationships among women and men, our analyses revealed gender differences in overall levels of eating attitudes and behaviors. In our sample, female participants revealed higher levels of orthorexia attitudes and behaviors compared to male participants, consistent with some previous work [15]. However, it is notable that the extant literature overall suggests more gender similarities than differences may exist [19]. This is an important finding as it suggests that even when the principal driving concern of restrictive eating patterns is not weight and shape, women still report higher levels of concerns than their male counterparts. It was also notable, however, that a larger proportion of the sample than might be expected met the cutoff score for pathological levels of ON symptoms. The criteria and assessment of ON continue to be areas of ongoing work [6], and it may be that the ORTO-7 instrument requires further adaptation to the US food context [23]. The other gender differences that emerged in terms of eating patterns are also consistent with those described in previous literature. Gender differences emerged on two of the intuitive eating subscales, with men reporting higher levels of Eating for Physical Rather Than Emotional Reasons and Reliance on Hunger and Satiety Cues [22]. Similarly, previous research has reported males scoring more highly on the three eating competence subscales that revealed such gender differences in our sample, namely Eating Attitudes, Food Acceptance, and Food Regulation [2]. Overall, female participants reported eating patterns that are likely reflective of the greater social pressure on women to control their weight and achieve unrealistically thin body sizes [15, 18].

The study presents a number of limitations. First, the gender ratio of participants was imbalanced. Future work should aim to gather more data from males regarding these issues so as to better explore gender differences and the factors driving them. Second, participants were all college students, and therefore, the sample is not generalizable to the broader population or perhaps beyond this developmental period. Third, previous work has highlighted the need to continue to develop measures of ON behaviors and symptoms that are valid and reliable across cultural contexts [11]. The tool chosen here, the ORTO-7, was found to reveal superior psychometric properties when compared to other version of this measure [11]. Nevertheless, it has been repeatedly criticized both in terms of its relatability and its validity as a measure of pathology eating attitudes and behaviors [4, 6, 13]. In particular, the tendency for the proposed cutoff values to seem over-inclusive by identifying large proportions of non-clinical samples as reporting high levels of concerns cast doubt on its validity particularly in cross-cultural settings. Finally, findings are cross-sectional and therefore do not inform the direction of relationships.

In sum, the findings from this study suggest that broadly, ON behaviors are associated with lower levels of healthful eating behaviors among both young women and men, although some evidence for curvilinear relationships also emerged among young men, such that for eating dimensions related to food choices and disinhibited eating, low levels of ON as currently conceptualized and assessed were associated with somewhat more positive patterns. Further work focused on evaluating how drive for a healthy diet can be associated with flexible and positive eating patterns is warranted.

What is already known on the subject?

Orthorexia nervosa (ON) is characterized by preoccupations related to diet quality and health concerns. ON has shown associations with other indicators of disordered eating.

What does this study add?

This study therefore evaluates the relationships between ON symptomatology and the different facets of intuitive eating and eating competence. Broadly, ON symptoms are associated with lower levels of healthful eating behaviors among both young women and men.