Genetic and environmental etiology of the broad avoidant restrictive food intake disorder phenotype in 6- to-12-year-old Swedish twins

IMPORTANCEAvoidant restrictive food intake disorder (ARFID) is characterized by an extremely limited range and/or amount of food eaten, resulting in the persistent failure to meet nutritional and/or energy needs. Its etiology is poorly understood and knowledge of genetic and environmental contributions to ARFID is needed to guide future research. OBJECTIVETo determine the extent to which genetic and environmental factors contribute to the liability to the broad ARFID phenotype. DESIGN, SETTING, AND PARTICIPANTSA nationwide Swedish twin cohort including 16,951 twin pairs born 1992-2010 whose parents participated in the Child and Adolescent Twin Study in Sweden (CATSS) at twin age 9 or 12 years (49.4% female). CATSS was linked to the National Patient Register (NPR) and the Prescribed Drug Register (PDR). MAIN OUTCOME/MEASURESFrom CATSS, NPR, and PDR, we extracted all parent-reports, diagnoses, procedures, and prescribed drugs between age 6 and 12 that were relevant to the DSM-5 ARFID criteria and developed a composite measure for the ARFID phenotype (i.e., avoidant/restrictive eating with clinically significant impact such as low weight or nutritional deficiency, and with fear of weight gain as an exclusion). In sensitivity analyses, we controlled for autism and medical conditions that could account for the eating disturbance. We fitted univariate liability threshold models to estimate the relative contribution of genetic and environmental variation to the liability to the ARFID phenotype. RESULTSWe identified 667 children (2.0%, 38.2% female) with the ARFID phenotype between age 6 and 12. Variation in the liability to ARFID was largely explained by additive genetic factors (0.79, 95% confidence interval [CI] 0.71-0.86), with significant contributions from non-shared environmental factors (0.21, 95% CI 0.14-0.29). Heritability was very similar when excluding children with autism (0.77, 95% CI 0.67-0.84) or medical illnesses that could account for the eating disturbance (0.80, 95% CI 0.71-0.86). CONCLUSIONS AND RELEVANCEPrevalence and sex distribution of the broad ARFID phenotype were similar to previous studies, supporting the use of existing epidemiological data to identify ARFID. This first study of the genetic and environmental etiology of ARFID suggests that ARFID is highly heritable, encouraging future twin and molecular genetic studies.