Avoidant Restrictive Food Intake Phenotypes: Prevalence, Developmental Characteristics, and Genetic Architecture

Importance: A narrow range of food consumption and/or restricted eating is a core feature of avoidant restrictive food intake disorder. However, there is limited knowledge of developmental characteristics of children with avoidant/restrictive food intake (ARFI) and its etiological influences, which constrains research, prevention, and intervention efforts.Objective: To estimate the prevalence of ARFI phenotypes in a population-based sample, examine developmental characteristics across childhood, and investigate the genetic architecture of ARFI using genome-wide association analyses.Design, Setting, and Participants: This pre-registered study used data from 35,751 children born 1999-2009 in the population-based Norwegian Mother, Father, and Child Cohort Study (MoBa), with mother-reported data on ARFI symptoms at 3 and 8 years and linkage with diagnostic data from population health registries. Data were analyzed from March 2024 to May 2025.Main Outcomes and Measures: Multiple items were used to identify children with broad ARFI. These children were sub-classified into three groups based on symptom persistence: ARFI-Broad Transient (only at age 3), Emergent (only at age 8), and Persistent (age 3 and 8). Children in these groups with one or more indicators of clinical significance (e.g., nutritional deficiency) were further classified into ARFI-Clinical subgroups. Groups were compared across developmental characteristics from 6 months to 14 years. Genome-wide methods were used to examine SNP heritability (SNP-h2), conduct genetic association analyses, and quantify genetic correlations with other phenotypes. Results: Of 35,751 children (51% male), the prevalence of ARFI-Broad Persistent, Transient, and Emergent was 6.0%, 17.7%, and 8.4%. The prevalence of ARFI-Clinical Persistent, Transient, and Emergent was 1.8%, 3.2%, and 1.4% (6.3% overall). Children with ARFI-Broad Persistent exhibited more developmental difficulties compared with children with no ARFI. SNP-h2 ranged from 8% to 16%. We identified two independent genome-wide significant loci. For ARFI-Clinical, a significant association was identified with ADCY3. Small-to-moderate genetic correlations were observed for ARFI-Broad, ARFI-Clinical and mental health, cognitive/educational, anthropometric, food-related, and gastrointestinal disorder phenotypes.Conclusions and Relevance: The prevalence of ARFI in the general pediatric population is substantial and affected children have elevated risk of developmental difficulties across multiple domains. Our findings point to the need for broad support interventions and advance understanding of the genetic underpinnings of ARFI.