| Domain | Best current evidence | Evidence strength/limitations | Key source metadata |
|---|---|---|---|
| Definition & identifiers | ARFID is a DSM-5/ICD-11 feeding/eating disorder defined by persistent restriction/avoidance causing inability to meet nutritional/energy needs with weight/growth effects, nutritional deficiency, supplement/enteral dependence, or psychosocial impairment, without body-image disturbance; MeSH term present in ClinicalTrials derived metadata: “Avoidant Restrictive Food Intake Disorder” (pqac-00000003, pqac-00000027, pqac-00000009) | Strong consensus on nosology; MONDO/OMIM not established in retrieved evidence; mostly disease-level aggregated resources rather than EHR-derived data | Fonseca et al., *J Eat Disord* 2024, published Jun 2024, DOI: https://doi.org/10.1186/s40337-024-01021-z; Sanchez-Cerezo et al., *eClinicalMedicine* 2024, Feb 2024, DOI: https://doi.org/10.1016/j.eclinm.2024.102440; ClinicalTrials.gov NCT06110806 posted 2023-11-01 (pqac-00000003, pqac-00000027, pqac-00000009) |
| Diagnostic presentations & 2024 latent classes | Core presentations: sensory sensitivity, lack of interest/low appetite, fear of aversive consequences; 2024 UK/ROI surveillance LCA of 319 cases identified 4 classes: Fear 7.2% (n=23), Lack of Interest 25.1% (n=80), Sensory 29.5% (n=94), Combined 38.2% (n=122) (pqac-00000003, pqac-00000027) | Strongest recent empirical subtype evidence in pediatric secondary care; may not generalize to adults/community samples | Sanchez-Cerezo et al., *eClinicalMedicine* 2024, Feb 2024, DOI above; Fonseca et al. 2024 review (pqac-00000003, pqac-00000027) |
| Epidemiology | Recent review summarizes prevalence estimates around 0.5–5% in children/adults; pediatric surveillance incidence in Canada reported as 2.02 per 100,000 ages 5–18 years (95% CI 1.76–2.31); mean age often 11.1–14.6 years; males comprise roughly 21–50% in clinical samples (pqac-00000002, pqac-00000026) | Estimates highly heterogeneous by setting and method; no robust incidence data for adults in retrieved evidence | Fonseca et al., *J Eat Disord* 2024; Sanchez-Cerezo et al., *eClinicalMedicine* 2024 (pqac-00000002, pqac-00000026) |
| Etiology, risk factors & comorbidity | Multifactorial model: predisposing neurodevelopmental/medical factors (ASD, ADHD, GI/neurologic disorders, food allergy), precipitating events (vomiting, choking, abdominal pain, bullying, bereavement, medication start), and perpetuating family/behavioral factors; anxiety disorders common (9.1–72%); ASD frequently co-occurs, especially sensory/combined presentations (pqac-00000023, pqac-00000026, pqac-00000017) | Mostly observational and review-level evidence; causal direction often unclear; protective factors not well established in retrieved literature | Fonseca et al. 2024; Sanchez-Cerezo et al. 2024; Nocerino et al., *Nutrients* 2024, Sep 2024, DOI: https://doi.org/10.3390/nu16173034 (pqac-00000023, pqac-00000026, pqac-00000017) |
| Genetics | No monogenic cause established. Review evidence cites a Swedish twin study showing important genetic contribution/high heritability; one review notes a reported locus near **ZSWIM6**, but this was not primary-source validated in retrieved accessible texts. Large-scale ARFID genetics infrastructure is expanding via EDGI2 (pqac-00000004) | Genetic architecture remains early-stage; retrieved evidence does not support causal genes, ClinVar variants, or penetrance estimates for ARFID | Tomaszek et al., *Nutrients* 2025, Jan 2025, DOI: https://doi.org/10.3390/nu17030486; EDGI2 protocol, *BMC Psychiatry* 2025, DOI: https://doi.org/10.1186/s12888-025-06777-5 (not context-cited in table cells beyond accessible ID) (pqac-00000004) |
| Mechanisms / pathophysiology | Best current model is 3-dimensional: altered sensory processing, appetite/homeostatic dysregulation, and negative valence/fear circuitry. Hypothesized regions/signals include insula, orbitofrontal cortex, hypothalamus, amygdala, anterior cingulate, and gut-brain hormones (ghrelin, PYY, CCK, GLP-1). ARFID severity is linked to lower anticipatory pleasure, especially lack-of-interest phenotype; depression partly explains anhedonia findings (pqac-00000003, pqac-00000025, pqac-00000019, pqac-00000020) | Human evidence is still limited and partly hypothesis-driven; little validated omics or pathway-level molecular profiling; no disease-specific GO/CL mappings directly established in retrieved sources | Fonseca et al. 2024 review; Dolan et al., *J Eat Disord* 2023, Nov 2023, DOI: https://doi.org/10.1186/s40337-023-00921-w (pqac-00000003, pqac-00000025, pqac-00000019, pqac-00000020) |
| Phenotypes, complications & QoL | Complications include malnutrition, growth delay, enteral dependence, hospitalization, hypokalemia, fatigue, lethargy, presyncope, constipation, cold intolerance, hypothermia, dry skin, lanugo, alopecia, bradycardia, orthostatic tachycardia, hypotension, pubertal delay/amenorrhea, lower bone mineral density, oral-motor and speech delays; selective eating also impairs social/emotional development and increases family conflict (pqac-00000007, pqac-00000022, pqac-00000017) | Strong clinical face validity; frequency estimates for individual complications are sparse; QoL often described qualitatively rather than with standardized ARFID-specific metrics in retrieved evidence | Fonseca et al. 2024; Nocerino et al. 2024 (pqac-00000007, pqac-00000022, pqac-00000017) |
| Diagnostics & assessment tools | Diagnosis remains clinical/DSM-based with exclusion of food unavailability, cultural practice, anorexia/bulimia, and other medical/psychiatric explanations. PARDI has Cronbach α 0.77–0.89 and diagnostic reliability κ=0.75; NIAS total α=0.84, ω=0.90. Trials also use PARDI-AR-Q, EDA-5, labs (thyroid, celiac), anthropometrics, and sometimes fMRI (pqac-00000007, pqac-00000028, pqac-00000013) | Good early psychometrics for screening/interview tools; no universal gold-standard biomarker; diagnostic workup must exclude medical mimics | Fonseca et al. 2024; Sanchez-Cerezo et al. 2024; ClinicalTrials.gov NCT05954728 (pqac-00000007, pqac-00000028, pqac-00000013) |
| Treatment | Multidisciplinary care is standard. Nutritional rehabilitation prioritizes weight restoration and adequacy of macro/micronutrients, with cautious temporary enteral support when necessary. Psychological approaches with best current support are CBT-AR and family-based treatment (FBT-ARFID). Review-level evidence describes significant reductions in ARFID severity, increased food variety, and weight gain in CBT-AR proof-of-concept studies; small FBT case series also report weight gain and reduced anxiety (pqac-00000001, pqac-00000022, pqac-00000021) | Evidence base still dominated by case series, pilot studies, and nonrandomized designs; no FDA-approved medication for ARFID | Fonseca et al. 2024, DOI above (pqac-00000001, pqac-00000022, pqac-00000021) |
| Pharmacotherapy | Adjunctive medications reported include olanzapine, mirtazapine, fluoxetine, cyproheptadine, and buspirone. Review cites mirtazapine-associated BMI change rising from 0.10 to 0.23/week after initiation and notes olanzapine may improve appetite, anxiety, and rigidity; all evidence is case-based/small series (pqac-00000021, pqac-00000001) | Very low-certainty evidence; no approved drug and no definitive randomized placebo-controlled data in retrieved sources | Fonseca et al., *J Eat Disord* 2024 (pqac-00000021, pqac-00000001) |
| Clinical trials / implementation | Key active/completed trials include: Stanford FBT-ARFID efficacy/mechanism RCT vs non-specific care, ages 6–12, n=98, completed (NCT04450771); earlier Stanford crossover FBT feasibility trial, ages 5–12, n=28, completed (NCT03778216); MGH COUNTERACT RCT of CBT-AR vs nutrition counseling, ages 10–18, n=53, completed (NCT05954728); MGH CBT-AR pilot single-group, ages 10–65, n=35, completed (NCT02963220); Mount Sinai MBIE family-based interoceptive exposure, actual n=12, terminated for risk-mitigation-plan disagreement (NCT06110806) (pqac-00000010, pqac-00000012, pqac-00000013, pqac-00000015, pqac-00000008) | Strong signal of growing implementation research; many results still pending publication or limited to protocol/trial registry detail | ClinicalTrials.gov: NCT04450771, NCT03778216, NCT05954728, NCT02963220, NCT06110806 (pqac-00000010, pqac-00000012, pqac-00000013, pqac-00000015, pqac-00000008) |
| Natural disease in other species / models | No recognized naturally occurring veterinary ARFID entity or validated full-disorder animal model was identified in retrieved evidence; at most, component traits such as sensory aversion, appetite regulation, fear conditioning, or gut-brain signaling can be modeled separately | Important negative finding; avoids over-interpreting feeding phenotypes in animals as DSM-defined ARFID | No disease-specific comparative biology source identified in retrieved ARFID literature (pqac-00000003, pqac-00000024) |


*Table: This compact table summarizes the best available current evidence for avoidant/restrictive food intake disorder across definition, subtypes, epidemiology, mechanisms, diagnostics, treatment, and trials. It emphasizes where evidence is strongest and where important limitations remain.*