Avoidant restrictive food intake disorder (ARFID) in Swedish preschool children: a screening study

Despite its common early onset, little is known about the prevalence and clinical presentation of avoidant restrictive food intake disorder (ARFID) in very young children, hindering early identification and intervention. Differentiating ARFID from normative selective eating is particularly challenging, yet validated parent-reported screening tools are lacking. This study aimed to estimate the point prevalence and describe the clinical characteristics of ARFID in preschoolers. It also evaluated the psychometric properties of the parent-reported ARFID-Brief Screener by assessing its agreement with a diagnostic interview for ARFID.

Parents of 645 children (50.5% male, mean age 3.2 years) completed the ARFID-Brief Screener and a neurodevelopmental screener during 2.5- and 4-year routine check-ups at 21 child health centers in West Sweden. Parents of all screen-positive and of randomly selected screen-negative children were invited to a follow-up diagnostic interview via phone. Additional clinical data were extracted from health records.

Of the 42 children (6.5%) who screened positive for ARFID, 29 were followed up via diagnostic interview, and 21 received an ARFID diagnosis, yielding a positive predictive value of 72%. Negative predictive value, sensitivity, specificity, and overall accuracy of the ARFID-Brief Screener were 94%, 91%, 79%, and 84%, respectively. The estimated point prevalence of ARFID was 5.9%. All diagnosed children exhibited both sensory-based avoidance and low interest in eating. Only 13.5% met ARFID criteria based on weight- or nutrition-related impairment (DSM-5 Criteria A1-A3). Two fifths (39.1%) of children with ARFID exhibited early language delays compared to 13.5% of children without ARFID. More extensive neurodevelopmental problems were associated with greater ARFID severity and with higher scores on the sensory and concern profiles.

ARFID is not uncommon among preschoolers, though prevalence may be slightly overestimated in this study. It is primarily characterized by sensory-based avoidance and low interest in eating, and by psychosocial impairment instead of physical health consequences, underscoring the need to assess impact beyond weight, growth, and nutrition. Early neurodevelopmental difficulties are overrepresented, highlighting their relevance for early detection and intervention. The ARFID-Brief Screener demonstrated promising psychometric properties and may be a valuable tool for routine screening, though follow-up assessments remain necessary to confirm a diagnoses.