Risk for avoidant/restrictive food intake disorder and related variable in Latin America

Abstract

Background: Avoidant/Restrictive Food Intake Disorder (ARFID) is a chronic childhood eating disorder whose recent inclusion in diagnostic manuals reflects the growing burden of feeding disturbances. This study examined baseline predictors of ARFID risk in Colombian children and tested their predictive value at six-month follow-up. Methods: Caregivers of 101 children aged 5–8 years completed the EDY-Q (ARFID risk), CBCL 4–18 (child emotional and behavioral symptoms), PHQ-9 (caregiver depression), PAFAS (parenting/family adjustment), and a question regarding active play frequency. At six-month follow-up, ARFID risk was reassessed. Two multiple linear regressions were conducted: (1) baseline ARFID risk regressed on 17 variables; (2) follow-up ARFID risk regressed on baseline ARFID risk plus significant predictors from Model 1. Results: Model 1 explained 46.8 % of baseline ARFID risk variance (p < 0.001). Significant predictors were child Somatic Complaints (β = 0.351, p = 0.017), caregiver Anxiety-Depression (β = –0.277, p = 0.018), and Parent-Child Relationship quality (β = 0.297, p = 0.001). Model 2 explained 74.4 % of follow-up ARFID risk variance (p < 0.001). Only baseline ARFID risk (β = 0.828, p < 0.001) and Somatic Complaints (β = 0.187, p = 0.009) remained significant. Conclusion: Baseline somatic complaints and prior ARFID risk robustly predict future ARFID risk, whereas caregiver anxiety/depression and parent-child relationship quality do not predict longitudinally. Early identification of somatic symptoms alongside ARFID screening may inform targeted interventions to reduce long-term risk.