Avoidant/restrictive food intake disorder (ARFID) was recently added to the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5
) Feeding and Eating Disorders section to replace and expand upon the former DSM-IV
diagnosis of feeding disorder of infancy and early childhood. Individuals with ARFID have a unique psychological profile manifesting as sensory sensitivity, fear of aversive consequences (e.g., choking or vomiting), and/or lack of interest in food or eating. Yet, they often present with dysfunctional feeding behaviors similar to anorexia nervosa (AN), including restrictive eating [1
]. Restrictive eating in ARFID is typically characterized not only by limited volume, but also limited variety. Case reports suggest that individuals with ARFID often limit their intake to highly palatable carbohydrate-rich processed foods, and—in more extreme cases—only eat foods within two food groups (e.g., grains and dairy), or even just one or two foods total (e.g., pasta and bread) [2
]. This severely selective diet can lead to nutrition-related medical issues such as electrolyte imbalances, fat-soluble vitamin deficiencies, B vitamin deficiencies (including vitamin B12 and folate deficiencies), and mercury toxicity [2
]. Unique to this population is the prevalence of nutritional deficiencies that exist independent of low weight or even in the absence of faltering growth [5
It is well known that a nutritionally adequate and balanced diet is critical for optimal growth and development in children and adolescents. The U.S. Department of Agriculture (USDA) Dietary Guidelines for Americans promote the intake of a diverse diet including nutrient-dense foods such as varied fruits and vegetables, whole grains, and lean proteins. Specific nutrient reference values for over 40 nutrients are described in the Dietary Reference Intakes (DRI) developed by the Institute of Medicine of the National Academies. The DRI are intended to guide nutrition recommendations and target adequate intake of both macronutrients and micronutrients [8
]. These recommendations further support the need for a varied diet as individual foods even within a food group can differ with respect to nutritional composition [11
]. Current estimates of nutrient status in the U.S. population indicate that vitamins A, C, D, E, folate, B6, B12, calcium, zinc, iron, and magnesium are often under-consumed relative to their DRI [12
]. Coupling this general estimate of under-consumption with disordered eating may increase risk of deficiencies in the ARFID population.
The USDA Dietary Guidelines for Americans recommend that both children and adults limit consumption of saturated fat, added sugar, and added salt, which are most often found in processed foods that may displace healthier alternatives. Processed foods—defined as any food that is altered during preparation (i.e., baking, drying, freezing)—include cereals, bread, snacks (i.e., chips, pies, pastries), convenience foods such as ready-to-eat meals, and sugar-sweetened beverages [14
]. Globally (and in the U.S.), processed foods are the predominant source of nutrients and energy accounting for 50%–90% (and more than 70%, respectively) of intake [16
]. Greater intake of processed foods is associated with health consequences including obesity, type 2 diabetes, and cardiovascular disease [18
], supporting the importance of limiting processed foods and promoting diet quality with nutrition education and lifestyle interventions.
Clinical observations suggest that individuals with ARFID often rely heavily on processed foods and limit their intake of fruits and vegetables, which may contribute to failure to meet adequate nutritional and/or energy needs with associated negative health implications. Few studies have evaluated the role of picky eating, a unique restrictive eating pattern that can lead to symptoms of ARFID, in children or adults [20
]. Among individuals with picky eating, severity of picky eating has been inversely correlated with diet variety, more specifically, intake of fruits and vegetables [22
]. However, dietary intake has yet to be systematically examined in the ARFID population. To empirically characterize dietary intake patterns associated with ARFID, we evaluated differences in four-day food record data in children and adolescents with full or subthreshold ARFID compared to healthy controls. This is the first study to assess dietary patterns, food intake by food groups, and usual intake of energy, macronutrients, and micronutrients of individuals with full or subthreshold ARFID. We hypothesized that, consistent with our clinical observations, individuals with full or subthreshold ARFID would commonly report intake of foods that are defined as processed foods, leading to greater consumption of added sugars and fat. We also hypothesized that greater consumption of processed foods would displace healthier food groups, such that individuals with full or subthreshold ARFID would report lower intake of fruits, vegetables, and fibrous grains compared to healthy controls. We further predicted that these differences in intake among food groups would be reflected in differential macronutrient intake such that individuals with full or subthreshold ARFID would have a diet higher in carbohydrates and fat (but lower in protein) compared to healthy controls. Lastly, we hypothesized that individuals with full or subthreshold ARFID would have significantly lower intake of fat-soluble (vitamins A, D, E, and K) and water-soluble vitamins (vitamins B6, folate, B12, and C) and minerals (calcium, iron, magnesium, zinc) compared to healthy controls, with a greater percentage of full or subthreshold ARFID not meeting the current dietary recommendations for micronutrient intake.
This is the first report systematically characterizing dietary intake in youth with full or subthreshold ARFID. We found that, compared to heathy children and adolescents, the diet of youth with full or subthreshold ARFID is higher in refined-carbohydrate processed foods, as well as total carbohydrates and added sugars, and lower in protein, vegetables, and vitamins K and B12. Our findings demonstrate that overall diet quality of youth with full or subthreshold ARFID is inadequate in macro- and micronutrient composition, ultimately increasing the risk for nutritional deficiencies and metabolic disorders.
We demonstrated that participants with full or subthreshold ARFID have a diet high in added sugars, a major component of processed foods, more specifically, foods categorized as refined grains. Refined grains have been milled, which is a process that removes the bran and germ. This results in a finer texture and improves shelf life, but it also removes dietary fiber, iron, and many B vitamins. Frequent consumption of processed foods can increase risk for metabolic sequelae, as they tend to be higher in added sugars, sodium, and solid fats, and lower in fiber and protein. The high palatability and unique macro- and micronutrient composition of processed foods is thought to contribute to gut-brain signaling dysfunction, which may drive intake and reinforce pathological eating behaviors [44
]. Results from a randomized controlled trial found that individuals randomized to a processed food diet had a greater intake of total calories, fat, and carbohydrates, a lower intake of protein, and a significant increase in weight over a two-week period compared to individuals randomized to an unprocessed diet. These findings further support the need to limit consumption of processed foods despite abundant access and ease of purchase and preparation [48
]. In children and adolescents, increased intake of processed foods is associated with overweight and obesity, elevated blood glucose levels, and increased risk of hypertension [49
]. Within our sample, percent of median BMI was not significantly different between groups. However, examining the long-term effects of limited diet variety and intake of processed foods in ARFID is warranted to determine the risk of increasing BMI and associated metabolic dysfunction. Increased consumption of processed foods is related to reduced intake of other food groups including fruits, vegetables, whole grains, and protein [51
]. We found that most of the commonly eaten foods reported by participants with full or subthreshold ARFID include processed foods and refined carbohydrates and that individuals with full or subthreshold ARFID report these foods at a significantly greater frequency than healthy controls. This likely contributes to the greater intake of total carbohydrates, percent of total calories from carbohydrates and added sugars, and reduced intake of protein and percent of total calories from protein that we observed in individuals with full or subthreshold ARFID.
Similar to AN, some individuals with ARFID are low-weight, which is often associated with inadequate intake as defined by limited volume and calories. However, other individuals with ARFID are of normal weight or overweight and report inadequate dietary variety. Other than in low-weight individuals where energy repletion is a critical first step, our findings suggest that therapeutic interventions for ARFID should prioritize the introduction of nutritionally adequate foods over processed foods with known detrimental effects. While patients with very severe presentations of ARFID may require gradual exposure to novel foods, early changes to eating flexibility that favor minor variations in processed foods or the introduction of supplements should ultimately be leveraged to initiate bigger changes such as the introduction of non-processed whole foods later in treatment [2
]. Cognitive-behavioral therapy is one current treatment modality for ARFID, which implements the USDA MyPlate guidelines as well as describes common micronutrient deficiencies to help identify foods for exposure therapy and learning [2
]. In addition, multidisciplinary teams including those in gastroenterology, nutrition, speech language pathology, and occupational therapy adopt similar approaches using the MyPlate guidelines as a primary tool for nutrition education and diet expansion [53
]. Our findings support current cognitive-behavioral therapy techniques and will help guide future multidisciplinary approaches and development of evidence-based guidelines for treatment specific to ARFID.
Micronutrient analysis revealed that youth with full or subthreshold ARFID reported lower vitamin K (as phylloquinone), reflective of reduced vegetable intake, and lower vitamin B12, consistent with reduced consumption of animal-based proteins, compared to healthy controls [43
]. Vitamin K is most commonly known for its role in blood clotting; however, novel functions continue to be identified [55
]. Vitamin B12 is an essential nutrient for brain development throughout childhood and adolescence due to its involvement in DNA synthesis and methylation, and maintaining genomic stability [54
]. Vitamin B12 insufficiency is associated with cognitive impairments in children [56
]. These findings are relevant for individuals with ARFID—whom we found to have insufficient intake of vitamin K and vitamin B12—and further support the need to maintain a varied diet and meet nutrient requirements from food before supplements. We did not find any other micronutrient insufficiency in this group, which may be attributed to consumption of processed foods. Food processing practices, such as enrichment and fortification, improve the nutrient content in foods by incorporating a variety of fat- and water-soluble vitamins and minerals that help individuals meet the DRI [57
]. With respect to our findings, B vitamins, including vitamin B12, are often found in fortified cereals, whereas vitamin K is not used in enrichment or fortification due to its interaction with anticoagulants [43
]. Thus, vegetables and plant-based oils remain the primary source of dietary vitamin K [58
]. In contrast to previous case reports of micronutrient toxicity in individuals with ARFID due to selective eating [2
], we found no indication of excessive micronutrient intake. The USDA Dietary Guidelines for Americans and DRI recommend consumption of nutritionally dense foods via diverse dietary intake to ensure optimal nutrient intake for physiological processes and to reduce disease risk [13
]. Diet diversity is considered imperative as nutrient composition varies among foods, and the likely synergistic effects among the nutrients and compounds within a food matrix may be lost when relying on supplements [60
]. Dependence on nutrient/energy supplements is one way to meet diagnostic criteria for ARFID. While 19% of participants with full or subthreshold ARFID endorsed taking nutrient/energy supplements in general, only six (12%) participants reported actual nutrient/energy supplement use within the four-day food record period, with healthy controls not reporting any use of nutrient/energy supplements. This indicates that supplement use may be sporadic and suggests that providers should maintain vigilance for patient adherence to nutrition-related recommendations. Thus, improving diet diversity in individuals with full or subthreshold ARFID is necessary to ensure nutritional adequacy and to preserve physiological functions independent of a diagnosed nutrient deficiency.
Our findings must be interpreted considering some limitations. First, we chose to use a four-day food record in this first analysis of dietary intake in ARFID; like any self-report instrument, such records are subject to potential bias and misreporting [36
]. Second, extensive analysis of micronutrient intake continues to be a challenge as the food composition databases utilized for nutrient composition can be limited in high-quality data representative of current national food habits and consumption patterns [61
]. Thus, we limited our micronutrient analysis to vitamins and minerals previously evaluated in pediatric populations. Third, dietary intake of macronutrients and micronutrients is not necessarily a reflection of nutritional status, of which we currently lack well-validated biomarkers [62
]. Thus, dietary questionnaires are useful in identifying risk for nutritional insufficiency. The question of how to best evaluate nutritional status using assessments of food intake or biomarkers continues to be a challenge among clinicians and healthcare practitioners working with ARFID [65
]. Our study supports the use of food records in full or subthreshold ARFID as a tool to help guide food exploration and diet diversity within therapeutic interventions.
In summary, in this first study of nutritional intake in youth with full or subthreshold ARFID, we found the diet of youth with full or subthreshold ARFID is higher in refined-carbohydrate processed foods, as well as total carbohydrates and added sugars, and lower in protein, vegetables, and vitamins K and B12. Thus, the severe selectivity and restrictions around eating among individuals with full or subthreshold ARFID appear to increase their risk of nutritional deficiencies. However, a diet high in processed foods may mitigate some micronutrient deficiencies since food processing techniques (e.g., fortification and enrichment) alter the micronutrient content of food. Conversely, a diet high in processed foods increases risk for metabolic diseases. Independent of nutrient status or intake, the selective diet often observed in ARFID may also contribute to psychosocial impairments such as limiting an individual’s ability to engage in social eating with peers, further demonstrating the burden of this disorder. Future studies will be important to examine differences in nutrient intake in individuals with prototypical ARFID presentations (i.e., sensory sensitivity, lack of interest, and fear of aversive consequences) [2
]. Our examination of dietary intake in individuals with full or subthreshold ARFID supports the development of therapeutic interventions—such as those utilizing food exposure methodology or cognitive-behavioral therapy—to not only improve overall intake but reduce risk of nutritional deficiencies and lessen the psychosocial burden of the disorder.