Multidisciplinary approach to the diagnosis of avoidant/restrictive food intake disorder in children with autism spectrum disorder: a multistage mixed-methods study

Full Text

INTRODUCTION

Nutrition is one of the key factors determining children's mental and physical development [1]. According to the World Health Organization (WHO)¹, malnutrition encompasses the following disorders: 1) undernutrition (low weight-for-height, low height-for-age, and low weight-for-age); 2) micronutrient deficiencies or excesses (vitamins and minerals); 3) overweight, obesity and diet-related noncommunicable diseases (such as heart disease, stroke, diabetes and cancer). Inadequate intake of vitamins and micronutrients may be accompanied by high-calorie intake, resulting in “hidden hunger” despite excess body weight. In such cases, traditional methods such as body mass index (BMI) determination often fail to detect nutrient deficiencies in overweight individuals, which necessitates a comprehensive health status assessment. The masking of nutrient deficiency by excess body weight complicates diagnosis [2, 3].

Malnutrition increases the risk of chronic diseases and significantly affects children’s development¹. Despite the importance of nutritional status for health throughout life, patterns of selective/restrictive food intake have long been insufficiently studied in clinical practice. The diagnosis of avoidant/restrictive food intake disorder (ARFID) was first included in the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) in 2013 [4]. Subsequently, a corresponding diagnostic category was introduced in the International Classification of Diseases, 11th Revision (ICD-11). In the Russian-language version, however, it is termed “pathological avoidant-restrictive food intake” [5]. Since the term “disorder” already implies a pathological condition and abnormality, the combination of “pathological” and “disorder” in the name is a pleonasm. Accordingly, the present study uses the standard English term for this diagnostic entity, avoidant/restrictive food intake disorder (ARFID). This diagnosis is not included in the ICD-10.

According to the ICD-11 definition, ARFID is an eating disorder characterized by persistent avoidance or restriction of food intake that results in significant weight loss (or failure to achieve expected weight gain), nutritional deficiency, dependence on enteral nutrition or dietary supplements, or marked impairment of psychosocial functioning [5]. Unlike anorexia or bulimia, food restriction is not associated with fear of weight gain or body image disturbances [5]. Food avoidance and dietary restriction, manifested as a persistent refusal to consume a wide range of both familiar and unfamiliar foods, as well as pronounced food neophobia, may occur both during normal development and in pathological conditions, which considerably complicates the differential diagnosis of ARFID [6].

Normal food selectivity typically emerges at around 18 months of age during the transition to independent feeding and usually resolves by 6–7 years of age [7]. In contrast, the pathological form is characterized by a chronic course and clinically significant consequences, including underweight, micronutrient deficiencies, and growth retardation, which may persist beyond the preschool years and, in some cases, throughout life [6–9]. The combination of a chronic course and adverse medical outcomes suggests that ARFID is an independent nosological entity that requires timely diagnosis and therapeutic intervention.

An essential step in the diagnosis of ARFID is to rule out somatic causes of food avoidance, as restricted food intake might be secondary to gastrointestinal disorders, allergic reactions, or metabolic disorders [10–12]. Failure to definitively exclude underlying general disorders undermines the rationale for behavioral interventions and may lead to worsening of the clinical symptoms.

Comprehensive somatic assessment is especially crucial in children with neurodevelopmental disorders, in whom the risk of underrecognition of somatic comorbidities is increased due to a limited ability to verbalize discomfort and atypical symptom presentation. A diagnosis of a primary eating disorder is only permissible after structural causes of food avoidance have been thoroughly evaluated and excluded [5, 10].

Epidemiological data on ARFID remain limited, with prevalence estimates ranging widely from 0.3% to 15.5% [10, 11]. Moreover, studies indicate a high rate of comorbidity between ARFID and autism spectrum disorders (ASD): according to various sources, the prevalence of ARFID among individuals with ASD ranges from 28% to 54.75% [13, 14].

Disturbances in eating behavior in children with mental disorders, including heightened sensitivity to the texture, smell, and taste of food, food neophobia, reduced interest in eating, and stereotyped eating patterns, were described by leading psychiatrists of the last century [15, 16]. In her monograph entitled “Childhood Autism” (1999), Bashina noted that some children with autism “refuse porridge, kissel (fruit starch drinks), and milk”, and that “the appearance of food provokes protest, negative attitudes, and even anxiety” [17]. Children with ASD often exhibit restrictive, stereotyped behaviors, such as consuming products of only one color, refusing foods of a particular shape, or selecting foods based on their packaging² [18, 19].

For a long time, eating disturbances were considered solely within the context of ASD, which contributed to a monocausal approach in the management of these patients² [19]. Given the crucial role of a balanced diet in ensuring optimal physical and neurocognitive development in children [19], the timely diagnosis of ARFID in children with ASD represents a significant clinical challenge. Identification of this concomitant condition enables the prompt initiation of therapeutic interventions aimed at restoring a balanced diet and addressing associated deficiency states and underlying medical conditions.

The aim of this study was to develop and implement a multidisciplinary approach to the diagnosis of ARFID in children with ASD.

METHODS

Study design

A multistage mixed-methods approach was employed to achieve the study aim. In the first stage, the diagnostic criteria for ARFID were clustered. In the second stage, a conventional algorithm for assessing patients with ASD for ARFID was developed. The third stage involved a cross-sectional study in which comprehensive assessments of ASD patients were conducted and approaches to diagnosing ARFID in children with autism were formulated.

Participants

Inclusion criteria:

• the diagnosis of ASD, verified using the standardized Autism Diagnostic Interview – Revised (ADI-R), which provides a full range of information necessary for establishing this diagnosis² [20]. The study included patients diagnosed with childhood autism (F84.0), atypical autism (F84.1), and Asperger syndrome (F84.5);
• children aged 2–17 years and 11 months (the ADI-R diagnostic interview is applicable for children over two years of age);
• written, voluntary informed consent to participate in the study was obtained from the parents or legal representatives for minors under 15 years of age and from minors aged 15 to 18 years.

Exclusion criteria:

• acute infectious diseases;
• medical conditions requiring emergency hospitalization or isolation;
• any other conditions that would interfere with the completion of the medical examinations scheduled by the study protocol.

Sample

A simple random sampling strategy was used in this study.

From September 1, 2021, to September 1, 2023, a research program of comprehensive assessment (including molecular-genetic testing) of children with ASD was conducted at the Research Institute of Pediatrics and Child Health Protection, Petrovsky National Research Centre of Surgery. The diagnosis of ARFID was carried out within the framework of this project.

Patients with a confirmed diagnosis of ASD were initially evaluated by a pediatric psychiatrist, after which they and their parents or legal guardians were invited to participate in the study. After obtaining informed consent, children with ASD underwent a comprehensive evaluation. Thus, the study included children with ASD regardless of the presence of eating difficulties. The sample was not limited to those patients who had already sought medical attention for complaints related to the quantity or quality of their food intake. This approach helped minimize selection bias and allowed for a more representative picture of the prevalence of food avoidance or dietary restrictions among children with ASD.

Procedures

The study included several stages:

1. At the first stage, the diagnostic criteria for ARFID were clustered. Because this diagnosis is not included in the ICD-10, its diagnostic criteria as defined in the ICD-11 were analyzed.
2. In the second stage, a multidisciplinary team of specialists developed a conventional diagnostic algorithm for assessing the presence of ARFID in ASD patients.
3. Third stage consisted of a cross-sectional study aimed at conducting a comprehensive evaluation of patients with ASD and developing approaches to diagnose ARFID in children with ASD.

At the initial visit, patients and their caregivers were provided with detailed information about the study and were given the opportunity to ask any questions and receive clarification.

After providing informed consent, all patients underwent formal confirmation of the ASD diagnosis using the ADI-R diagnostic interview. The ADI-R is a semi-structured, standardized interview with parents or primary caregivers, designed to assess autism symptoms across three main areas: social interaction, communication, and restricted and repetitive patterns of behavior. The interview includes 93 items used to evaluate the patient's mental state. The ADI-R helps differentiate autism from other developmental disorders with a sensitivity and specificity exceeding 90% in individuals with a mental age of 18 months and older [20].

Subsequently, patients were examined according to the conventional algorithm developed during the second stage of the study. Each consulting specialist could order personalized additional assessments for each child based on the clinical picture, current clinical guidelines and standards of care to confirm or exclude any comorbidities.

Data analysis

During the first stage, using a deconstructive approach, the clinical manifestations of ARFID as defined in the ICD-11 were identified and analyzed. Subsequently, using a descriptive method, clusters were defined and formed to which the identified symptoms were assigned.

During the second stage, based on the cluster analysis of ARFID symptoms, a multidisciplinary team determined the required diagnostic procedures for identifying symptoms within each cluster and developed a diagnostic protocol, i.e. a specific algorithm. This protocol included clinical examinations by various specialists (psychiatrist, pediatrician, endocrinologist, gastroenterologist, geneticist, neurologist, allergist/immunologist) as well as laboratory and imaging tests to identify any underlying medical conditions.

During the third stage, anthropometric parameters obtained during the clinical examination were analyzed using WHO AnthroPlus software, developed by the WHO for assessing physical development in children aged 0–5 years and 5–19 years. The assessment was conducted using z-scores (standard deviations, SD) based on WHO Anthro and WHO AnthroPlus reference data.

To determine nutritional status, the following indices were assessed: weight-for-age (WAZ), height-for-age (HAZ), and body mass index-for-age (BAZ).

The WAZ index was interpreted as follows: underweight was diagnosed when the value was less than −2 SD (standard deviation).

Children with a HAZ below −2 SD were considered stunted, whereas those with a HAZ above +2 SD were considered tall. Children with a BAZ above +2 SD at age < 5 years and above +1 SD at age ≥5 years were characterized as having overweight; children with a BAZ above +3 SD at age < 5 years and above +2 SD at age ≥5 years were considered obese. The normal range corresponded to −2 to +2 SD for children < 5 years and −2 to +1 SD for children ≥5 years. Thinness/underweight was defined below −2 SD.

To diagnose allergic diseases that might lead to the avoidance of certain foods, the ImmunoCAP technology was utilized to detect allergen-specific IgE antibodies [21]. This test assesses sensitization to 112 allergen components. It can identify even asymptomatic allergies, help identify the relevant allergen sources, evaluate the risk of cross-reactivity with related allergens, and guide the selection of effective treatment [21, 22].

Descriptive statistics were used to analyze the following: clinical and instrumental symptoms of somatic disorders; levels of vitamins, micronutrients, and macronutrients; indicators of protein, carbohydrate, and lipid metabolism; and liver and pancreatic enzymes. Statistical tables based on the examination results were generated using matplotlib, scipy, pandas, and numpy modules in Python version 3.8 and 3.3 in Anaconda.

Genetic testing was performed using tandem mass spectrometry (acylcarnitine and amino acid profile) as a screening tool for inherited metabolic disorders, including disruptions of amino acid metabolism, organic acidurias, and defects in mitochondrial β-oxidation of fatty acids [23–25].

Ethical considerations

The study was approved by the independent Ethics Committee at the Central Clinical Hospital of the Russian Academy of Sciences (Protocol No. 071N/2021 dated January 15, 2021).

RESULTS

First stage: clustering of diagnostic criteria for avoidant/restrictive food intake disorder

An analysis of the current diagnostic criteria for ARFID, as defined in the ICD-11, revealed that the features of this disorder cluster into two main criteria, with the behavioral and somatic components integrated into a single diagnostic foundation. The first criterion encompasses a limited quantity and/or variety of food intake that does not meet the physiological energy and nutrient requirements, leading to significant somatic consequences. These consequences include substantial weight loss, a clinically significant nutrient deficiency, the need for additional feeding (including tube feeding), or other medical complications. The second criterion describes disruptions in personal, social, familial, educational, and/or professional functioning that are attributable to eating behavior patterns (e.g., avoidance of social activities involving meals or pronounced psycho-emotional distress).

Even within the framework of the first criterion of the ICD-11 diagnostic system, two key components are combined: 1) a behavioral component — refusal to eat or restriction of food intake; 2) a somatic component — the resulting somatic consequences. In clinical practice, these symptoms can be identified only by an interdisciplinary approach, including the use of psychopathological assessment, clinical laboratory tests, and other testing modalities. Diagnosis typically involves a multidisciplinary team of specialists, including a psychiatrist, pediatrician, gastroenterologist, endocrinologist, medical geneticist, and other specialists as needed.

Accordingly, this study proposes a structured clustering of ARFID diagnostic features into three distinct clusters: behavioral, somatic, and functional disorders (Table 1).

 

Table 1. Diagnostic clusters of avoidant/restrictive food intake disorder

Behavioral cluster	Somatic cluster	Functional disorders cluster
Intake of an insufficient quantity or variety of food to meet adequate energy or nutrient needs.	Significant weight loss, clinically significant nutritional deficiency, dependence on enteral feeding or oral nutritional supplements, or other adverse effects on the individual's physical health.	Pronounced distress and/or significant impairment in personal, family, social, educational, occupational, or other important areas of daily life.

 

The behavioral cluster encompasses the avoidant/restrictive food intake pattern. These features are typically identified through history-taking with primary caregivers.

The somatic cluster, in addition to the core conditions, may present as other adverse effects on physical health: macro- and micronutrient deficiencies leading to anemia, osteopenia, rickets, and other conditions; as well as carbohydrate and lipid metabolism disorders with the potential development of obesity, atherosclerosis, diabetes mellitus, and other diseases.

The functional disorders cluster is characterized by deterioration in parent-child relationships and the development of maladaptive feeding-related behavioral patterns (e.g., increased irritability, as well as the emergence of aggression and self-injurious behavior during feeding attempts).

A key methodological decision in the analysis of the diagnostic criteria for ARFID was the delineation of the somatic cluster as a distinct, independent group. This decision is driven by a specific clinical feature of the disorder: when behavioral manifestations dominate the diagnostic focus, somatic consequences often evade clinical detection. Separating the somatic criteria from the behavioral at the analysis stage allowed their independent verification in the subsequent diagnostic algorithm for children with ASD, precluding the risk of them being absorbed by behavioral symptoms during the clinical assessment.

Accordingly, a diagnosis of ARFID requires a combination of symptoms from the first (behavioral) cluster together with features characteristic of the second cluster (somatic complications secondary to the abnormal eating pattern) and/or the third cluster (functional disorders associated with avoidant or restrictive food intake).

Second stage: development of a diagnostic algorithm

The developed conventional diagnostic algorithm for examining patients with ASD to diagnose ARFID (Figure 1) involves several consecutive steps:

1. Medical history-taking. This includes identification of adverse intrauterine factors, assessment of anthropometric data at birth, weight and height trajectories, feeding patterns, and degree of nutrient absorption (regurgitation, vomiting, unformed stool with undigested food fragments, frequent loose stools, constipation). Information is also collected on the child's behavior during meals, emotional reactions, eating schedule and food selectivity, qualitative and quantitative diet composition, and presence of nausea or vomiting during or after eating.
2. Assessment of a food diary. It is completed by parents or caregivers. The diary records the quantity and composition of foods consumed by the child over 24 hours. The obtained quantitative and qualitative characteristics of actual nutrition are compared with age-appropriate physiological norms for nutrient and energy requirements by a pediatrician and a dietitian. During the assessment, specialists also use a checklist (see Table S1 in the Supplementary) designed to record specific manifestations of avoidant or restrictive food intake. Additionally, information is collected on past illnesses, history of hospitalizations, and medical procedures performed (enteral nutrition, endoscopy).
3. Clinical examination of the child with ASD. This step is performed to assess the adverse health effects of selective and restrictive food intake patterns. The pediatrician performs anthropometric measurements (height, weight) to identify possible developmental disorders and body composition characteristics according to WHO criteria (WHO AnthroPlus). Physical examination includes inspection of the skin and its appendages (hair, nails, teeth), mucous membranes, assessment of subcutaneous fat development, and evaluation of the respiratory, cardiovascular, gastrointestinal, nervous, excretory, and endocrine systems, reveals signs of nutritional insufficiency or excess, as well as macro- and micronutrient deficiencies.
4. Laboratory tests and other testing modalities used in children with ASD. All children undergo complete blood count, tests for levels of vitamins (D, B12); a metabolic panel to evaluate protein metabolism (total protein, albumin, creatinine, urea, prealbumin, ferritin, transferrin), as well as metabolism of carbohydrates, lipids, micro- and macronutrients (sodium, potassium, iron), liver and pancreatic enzymes, along with allergy diagnostics [22, 25, 26]. Assessment of thyroid function includes laboratory analysis of blood thyroid-stimulating hormone (TSH) levels and ultrasound examination of the thyroid gland. Comprehensive evaluation of the gastrointestinal tract is performed using abdominal ultrasound, water-siphon test, and blood biochemistry tests. Body composition impedance analysis is used to determine the degree of muscle and fat tissue deficit.
5. A consultation with a medical geneticist. This step is conducted to rule out rare hereditary diseases, including metabolic disorders (phenylketonuria, acyl-CoA dehydrogenase deficiency), in which clinical presentation ASD symptoms are combined with eating behavior peculiarities — selectivity, refusal of certain foods, or specific reactions to their intake.
6. Consultations with other specialists included in the team, who, within their respective competencies, assess the health consequences of dietary patterns and/or rule out possible diseases that may lead to negative changes in somatic status:
   
   • the endocrinologist evaluates metabolic processes in the body (carbohydrates, fats, proteins), thyroid function, and other disorders affecting nutritional status;
   • the allergist-immunologist performs diagnostic evaluation for allergic diseases, determining whether avoidance of certain foods is related to adverse reactions due to food allergy or product intolerance;
   • the gastroenterologist diagnoses gastrointestinal diseases, evaluates their impact on health and nutritional status, and determines whether the identified disorder may be a consequence of inadequate nutrition, including disorders of a selective or restrictive nature [27];
   • the dermatologist identifies possible comorbid conditions manifesting with symptoms characteristic of nutritional deficiency, such as nail brittleness, hair loss, and rash;
   • the dietitian assesses the patient's diet.

•  

 

Figure 1. Conventional assessment algorithm for patients with ASD to diagnose avoidant/restrictive food intake disorder.

Note: ARFID — Avoidant/Restrictive Food Intake Disorder; ASD — Autism Spectrum Disorder.

Source: Ustinova et al., 2026.

 

During the initial examination (Steps 1 and 2), only patterns of selective and restrictive food intake characteristic of the first cluster may be identified. A diagnosis of ARFID can only be established upon the detection of malnutrition-induced somatic complications. These must be identified through subsequent evaluation, specifically through laboratory testing and other testing modalities, which are mandatory for children presenting with avoidant/restrictive eating behaviors.

Third stage: development of diagnostic approaches for avoidant/restrictive food intake disorder

The study included 80 children diagnosed with ASD by a psychiatrist according to ICD-10 criteria and verified using the ADI-R diagnostic interview.

The age of the study participants ranged from 3 years to 17 years and 8 months. The mean age at the time of the diagnostic evaluation was 101 months (46; 177). The study included 59 (74%) boys and 21 (26%) girls. Table 2 presents the manifestations of avoidant/restrictive food intake identified in our group using a checklist.

 

Table 2. Manifestations of avoidant/restrictive food intake pattern in children with autism spectrum disorders (n=80)

Symptoms of ARFID	Abs. value (%)
Negative responses to introduction of new products	16 (20.0%)
Constant refusal to eat certain foods (10 or more)	40 (50.0%)
Exclusive intake of products with certain appearance, type, texture, or temperature	32 (40.0%)
Ritualistic eating	19 (24.0%)
Slow or fast food intake	33 (41.0%)
Absence / significant deficiency of plant foods in the diet	45 (56.0%)
Absence / significant deficiency of protein foods in the diet	32 (40.0%)
Absence / significant deficiency of fatty foods in the diet	21 (26.2%)
Significant predominance of foods with high sugar content in the diet	54 (67.5%)

Note: ARFID — Avoidant/Restrictive Food Intake Disorder.

 

Using the conventional examination algorithm, the behavioral pattern of selective and restrictive food intake was identified in 71 children (88.8%), including 51 boys and 20 girls. Further comprehensive examination of children with this pattern revealed that 66 of them had somatic disorders caused by the eating disorder. This represented 82.5% of the total sample (80 children) and 93% of the children with the identified behavioral pattern (71 children). Five of these children did not have any somatic disorders.

All subsequent analyses are based on the sample of 66 children who, in addition to the primary ASD diagnosis, were found to have the behavioral pattern of ARFID and were diagnosed with somatic disorders caused by the eating disorder. In these patients, in addition to nutritional status assessment, diseases potentially influencing eating behavior were ruled out.

Half of the children (n=33; 50.0%) had normal body weight, 19 children (28.8%) were overweight or obese, and 14 children (21.2%) were underweight.

Changes in body weight parameters were observed in both boys and girls (Figures 2 and 3).

 

Figure 2. Distribution of body weight values in boys (n=46).

Source: Ustinova et al., 2026.

 

Figure 3. Distribution of body weight values in girls (n=20).

Source: Ustinova et al., 2026.

 

Nine children (13.6%) were found to have short stature (Figure 4). All possible causes, apart from diet-related ones, were excluded.

 

Figure 4. Physical development parameters (height) in the study group of children (n=66).

Source: Ustinova et al., 2026.

 

Height value deviations were revealed both in boys and girls (Figures 5 and 6).

 

Figure 5. Height values in boys (n=46).

Source: Ustinova et al., 2026.

 

Figure 6. Height values in girls (n=20).

Source: Ustinova et al., 2026.

 

Vitamin deficiencies were also revealed in the study group (n=66).

Vitamin D levels were as follows:

• significant deficiency (< 10 ng/mL) was identified in 3 children (4.5%);
• moderate deficiency (10–20 ng/mL) — in 18 children (27.3%);
• insufficiency (20–29 ng/mL) — in 22 children (33.4%).

Only 23 participants (34.8%) had normal vitamin D levels (30–80 ng/mL).

Vitamin B12 deficiency was identified in 25 children (37.9%). Twelve children (18.2%) had dyslipidemia. Four children (6.1%) had glucose intolerance. Among other parameters associated with avoidant/restrictive food intake patterns, the following abnormalities were recorded: reduced hemoglobin levels in 7 children (10.6%); iron deficiency in 7 children (10.6%); reduced ferritin levels in 13 children (19.7%); elevated latent serum iron-binding capacity in 5 children (7.6%). No clinically significant deficiencies of zinc, magnesium, or copper were detected in the study group.

None of the study participants had thyroid function disorders. Functional gastrointestinal disorders included dyspepsia in 9 children (13.6%) and constipation in 8 children (12.1%). Gastroenterological symptoms in the form of vomiting and distress in response to food intake were recorded in 9 children (13.6%). The multidisciplinary team concluded that the identified functional gastrointestinal disorders were consequences of an unbalanced diet.

In 21 children (31.8%), allergic diseases with sensitization to pollen, household, epidermal, and food allergen components were identified. However, the patterns of selective and restrictive food intake did not correspond to specific allergens and were not related to allergy.

In all 66 children (100%), regardless of the presence or absence of deviations in physical parameters (body weight), some degrees of malnutrition were identified. Table 3 presents changes in blood biochemistry parameters according to normal and abnormal body weight values.

 

Table 3. Changes in blood biochemistry parameters depending on the body weight in the study cohort of children (n=66)

Abnormalities	Body weight
	Normal body weight (n=33)	Overweight (n=19)	Deficiency (n=14)	χ²	df	p-value
Vitamin D deficiency, n (%)	15 (45.5%)	12 (63.2%)	14 (100.0%)	12.44	2	0.002
Vitamin B12 deficiency, n (%)	13 (39.4%)	2 (10.5%)	10 (71.4%)	12.77	2	0.002
Iron level reduction, n (%)	3 (9.1%)	0 (0.0%)	4 (28.6%)	7.1	2	0.029
Hemoglobin level reduction, n (%)	2 (6.1%)	0 (0.0%)	5 (35.7%)	12.28	2	0.002
Ferritin level reduction, n (%)	6 (18.2%)	0 (0.0%)	7 (50.0%)	12.84	2	0.002
LIBC reduction, n (%)	0 (0.0%)	4 (21.1%)	1 (7.1%)	7.64	2	0.022
Lipid metabolism disorder, n (%)	3 (9.1%)	8 (42.1%)	1 (7.1%)	10.29	2	0.006

Note: LIBC — latent iron binding capacity.

 

As expected, deficiencies were more common in the group of patients with low body weight (vitamin D — 100%, B12 — 71.4%, ferritin — 50%). However, significant abnormalities were also recorded in groups of overweight children as well as those with normal body weight: vitamin D deficiency (45.5% and 63.2%), lipid metabolism disorders in overweight individuals (42.1%), and B12 deficiency in children with normal body weight (39.4%).

Eleven children (16.7%) also had functional disorders associated with selective and restrictive eating (inability to attend preschool due to refusal to eat; refusal of hospitalization because the child can eat only in familiar conditions; refusal of long walks and trips). No concomitant diseases affecting food selectivity, including inherited metabolic diseases, were registered.

DISCUSSION

Based on the multidisciplinary diagnostic evaluation, it was found that out of 80 children with ASD included in the study, 71 (88.8%) exhibited a behavioral pattern of selective and restrictive eating. Additionally, 66 children (82.5%) were diagnosed with somatic disorders caused by behavioral eating disorders. The most common of these include vitamin and microelement deficiencies, metabolic disorders, and functional gastrointestinal disorders.

The analysis showed that identification of somatic consequences associated with selective and restrictive food intake requires a comprehensive clinical and laboratory work-up with the mandatory involvement of specialists from various fields (pediatrician, gastrointestinal specialist, endocrinologist, psychiatrist, etc.). This justifies the proposed separation of behavioral and somatic diagnostic clusters.

It is particularly important to note that the adverse consequences of ARFID may remain undetected in the presence of apparently "satisfactory" anthropometric parameters, e.g., in cases of normal body weight or obesity. Only a comprehensive multidisciplinary assessment can identify nutritional deficiencies and other somatic abnormalities associated with restricted food intake in such children.

Therefore, ARFID should be considered not as an epiphenomenon of ASD but as an independent comorbidity requiring early diagnosis and multidisciplinary management.

A strength of this work is the application of a multidisciplinary assessment algorithm developed in this study, which, to the authors’ knowledge, has been implemented for the first time. This approach enabled the identification of a high prevalence of ARFID in the study sample and the establishment of present comorbid somatic disorders that frequently remain undiagnosed when standard assessment methods are employed.

One limitation of this study is that the sample comprised patients with ASD presenting to a multidisciplinary healthcare facility. Seeking care at a pediatric clinic may indicate the presence of somatic symptoms, which may have contributed to a higher detection rate of ARFID in the study group compared with population-based data or findings from other studies. However, since this study aimed at refining diagnostic approaches to ARFID rather than to estimate the prevalence of the disorder among children with ASD, this factor did not affect the course of the study, the resolution of the stated research objectives, or the formulation of the conclusions.

The adverse effects of an unbalanced diet are widely discussed not only on physical health but also on behavioral manifestations in children with ASD, including increased irritability, aggression, self-injurious behavior, and negativism, as well as the impact on the pace and trajectory of neurodevelopment [22]. The results of this study confirm the high prevalence of ARFID in children with ASD, as previously reported in other publications [13, 14].

In contrast to previous research, this study employed a multidisciplinary assessment algorithm, which enhanced diagnostic accuracy and enabled the identification of occult somatic consequences of eating disorders. This underscores the need for involvement of specialists from multiple disciplines in the diagnosis of ARFID, including psychiatrists, pediatricians, gastroenterologists, dietitians, and other specialists as needed. The findings indicate that accurate diagnosis of ARFID in children with ASD requires the identification of symptoms across each of the diagnostic clusters: behavioral, somatic, and the functional disorders clusters (e.g., avoidance of food intake in social situations). Establishing comorbidities of avoidant/restrictive food intake disorder not only allows for objectification of the problem, but also provides a rationale for additional medical interventions aimed at correcting the identified nutrient deficiencies.

Previous studies indicate that ARFID is more common in children with comorbid ASD [14, 16, 18], which is consistent with the findings of this study. The diets of children with ASD are typically characterized by either excessive or significantly reduced caloric intake, an excess of fats, sugars, and salt, alongside pronounced deficiencies of vitamins and micronutrients [14].

Particular attention in this study is given to the complexity of diagnosing the somatic consequences of ARFID. In some cases, these complications can remain unrecognized despite normal or even excessive body weight. This study has demonstrated that 28.8% of children with ARFID and ASD were obese. However, this did not preclude the presence of significant deficiencies in vitamins, micro- and macronutrients impacting neuropsychiatric development [28–31]. These findings confirm an important clinical conclusion: excess body weight does not rule out the presence of ARFID, particularly in cases where the diet is limited to high-calorie, low-nutrient-value foods (e.g., an excess of simple carbohydrates alongside a deficit of protein, fiber, vitamins, and micronutrients).

Additionally, the expert group raised questions regarding the applicability of weight and height standards developed for the general population to children with ASD. It has been previously shown that children with ASD exhibit specific features of physical development, including a tendency toward tall stature [32, 33], which can mask signs of malnutrition. This highlights the need for further research to determine correct standards for assessing the physical status of this patient cohort.

ARFID should not be diagnosed in cases where eating disorders are attributable to somatic conditions, such as food allergies, infectious or gastrointestinal disorders, and other states leading to reduced appetite, food restriction, or weight loss. Accordingly, the diagnostic process must include exclusion of somatic causes, which requires a comprehensive clinical and laboratory work-up and the involvement of a multidisciplinary team.

The need for further study of ARFID in children with ASD is driven not only by its high prevalence and diagnostic complexity, but also by the potential involvement of shared etiopathogenetic mechanisms. A neurobiological link between ARFID and ASD has been hypothesized, and understanding this connection could form the basis for novel therapeutic approaches.

Effective care for children with ASD requires the implementation of multidisciplinary models for diagnosing and treating ARFID. Accurate assessment of eating behavior patterns, nutritional, and somatic status is associated with significant challenges and necessitates increased competency among physicians of various specialties, particularly psychiatrists and pediatricians.

It is crucial to develop collaboration between psychiatric and pediatric services. As part of the multidisciplinary approach, the psychiatrist should focus on the behavioral aspects of the disorder, while the pediatrician and other specialists should conduct a comprehensive assessment of somatic status and implement appropriate management of identified conditions.

CONCLUSION

During the study, a classification of ARFID symptoms was proposed, comprising three diagnostic clusters: behavioral, somatic, and functional disorders clusters. Among the 80 examined children with ASD, 71 (88.8%) exhibited avoidant/restrictive food intake patterns, and 66 (82.5%) had somatic complications associated with disordered eating. These findings underscore the critical importance of assessing the somatic consequences of avoidant/restrictive food intake patterns when diagnosing ARFID in children with ASD and the need to include the somatic cluster in the diagnosis. The results highlight the importance of a multidisciplinary approach in the diagnosis and management of ARFID. This approach not only allows for the precise identification of all aspects of the disorder but also enables the development of customized interventions aimed at correcting eating behavior and restoring adequate nutritional status.

 

Authors’ contribution: All the authors made a significant contribution to the article, checked and approved its final version prior to publication.

Funding: The research was carried out without additional funding.

Conflict of interest: The authors declare no conflicts of interest.

Generative AI use statement: Nothing to disclose.

Supplementary data

Supplementary material to this article can be found in the online version by doi:

Table S1: 10.17816/CP15472-145132

 

¹ WHO Fact sheet “Malnutrition”. Available from: https://www.who.int/ru/news-room/fact-sheets/detail/malnutrition

² Clinical guidelines “Autism Spectrum Disorders”. Available from: https://base.garant.ru/409578885/ (Russian).

About the authors

Natalia V. Ustinova

Petrovsky National Research Centre of Surgery; Scientific and Practical Center for Mental Health of Children and Adolescents named after G.E. Sukhareva

Author for correspondence.
Email: ust-doctor@mail.ru
ORCID iD: 0000-0002-3167-082X
SPIN-code: 5003-3852

MD, Dr. Sci. (Med.), Head of the Department of social pediatrics and the organization of multidisciplinary support for children, Chief research

Russian Federation, Moscow; Moscow

Elena A. Gorbunova

Petrovsky National Research Centre of Surgery

Email: ust-doctor@mail.ru
ORCID iD: 0009-0001-0440-2715
SPIN-code: 7418-5571
Scopus Author ID: 6506802044
ResearcherId: AAD-2209-2022

MD, Cand. Sci (Med.), Psychotherapist, Pediatrics and Child Health Research Institute

Russian Federation, Moscow

Leila S. Namazova-Baranova

Petrovsky National Research Centre of Surgery; Pirogov Russian National Research Medical University; Shenzhen MSU-BIT University

Email: ust-doctor@mail.ru
ORCID iD: 0000-0002-2209-7531
SPIN-code: 1312-2147
Scopus Author ID: 48761908100
ResearcherId: C-9485-2019

MD, Dr. Sci. (Med.), Academician of the Russian Academy of Sciences, Head of Pediatrics and Child Health Research Institute, Head of the Department of Faculty Pediatrics, Professor, President of the Union of Pediatricians of Russia

Russian Federation, Moscow; Moscow

Anna Ya. Basova

Scientific and Practical Center for Mental Health of Children and Adolescents named after G.E. Sukhareva; Pirogov Russian National Research Medical University

Email: ust-doctor@mail.ru
ORCID iD: 0000-0002-5001-8554
SPIN-code: 3290-5781

MD, Cand. Sci. (Med.), Deputy Director for Research, Associate Professor, Psychiatry and Medical Psychology Department

Russian Federation, Moscow; Moscow

Natalya G. Ovsyanik

Petrovsky National Research Centre of Surgery

Email: ust-doctor@mail.ru
ORCID iD: 0000-0002-7763-5085
SPIN-code: 9942-0547

Pediatric endocrinologist, Nutritionist, Junior Researcher at the Pediatrics and Child Health Research Institute

Russian Federation, Moscow

Elena V. Kaytukova

Petrovsky National Research Centre of Surgery; Pirogov Russian National Research Medical University

Email: ust-doctor@mail.ru
ORCID iD: 0000-0002-8936-3590
SPIN-code: 1272-7036

MD, Cand. Sci (Med.), Gastroenterologist, Deputy Head for Medical Activities, Head of the Consultative and Diagnostic Center for Children, Associate Professor of the Department of Faculty Pediatrics

Russian Federation, Moscow; Moscow

Zoya Y. Suleymanova

Petrovsky National Research Centre of Surgery

Email: ust-doctor@mail.ru
ORCID iD: 0000-0001-8723-0199
SPIN-code: 7521-1803

MD, Cand. Sci (Med.), Associate Professor, Gastroenterologist, Leading Researcher at the Research Institute of Pediatrics and Child Health Protection

Russian Federation, Moscow

Olga S. Gundobina

Petrovsky National Research Centre of Surgery

Email: ust-doctor@mail.ru
SPIN-code: 4363-8042

MD, Cand. Sci (Med.), Gastroenterologist, Leading Researcher at the Research Institute of Pediatrics and Child Health Protection

Russian Federation, Moscow

References

Supplementary files