Key Takeaways
Extended Takeaways
- The clearest longitudinal signal was symptom-specific: decreases in fasting CCK tracked with improvement in PARDI-lack of interest at both Year 1 and Year 2, while CCK change was not associated with overall PARDI-severity.
- By contrast, increases in fasting ghrelin were linked to lower overall ARFID severity at both follow-ups (adj. R2=0.35, F1, 41 =6.00, p=.02; adj. R2=.50, F1, 28 =6.38, p=.02), but ghrelin change did not predict lack of interest severity.
- Cross-sectionally, the only significant between-group hormone difference emerged at Year 1, when youth meeting cutoff for ARFID-lack of interest had higher fasting CCK with a large effect (η2p =.180); no fasting ghrelin differences were detected across timepoints.
- These associations held after adjusting for age, sex, BMI percentile, and baseline PARDI scores, which supports considering appetite hormones as correlates of symptom course rather than simple reflections of weight status or demographic differences.
- Interpretation should account for attrition in biologic sampling: 100 youth were recruited at Baseline, but hormone analyses included 56 participants at Year 1 and 43 at Year 2, largely because of COVID-19 restrictions, blood volume limitations, or refusal at blood draw.
- Because the sample included youth aged 9–23 years with both full (n=88) and subthreshold (n=12) ARFID, the findings may be most useful for dimensional monitoring of symptom change over time rather than assuming a single hormonal pattern across all ARFID presentations.
