Back

Co-occurring weight- and/or shape-motivated restriction in 5,747 adults with probable ARFID

Hog, L.; MacDermod, C.; White, J. P.; Baker, J. H.; Guintivano, J.; Johnson, J. S.; Ortiz, S. N.; Pisetsky, E. M.; Micali, N.; Bulik, C. M.; Thornton, L. M.; Dinkler, L.

2025-07-29 epidemiology
10.1101/2025.07.28.25332146 medRxiv
Show abstract

ObjectiveAccording to DSM-5-TR, avoidant/restrictive food intake disorder (ARFID) cannot be diagnosed alongside anorexia nervosa (AN), bulimia nervosa (BN), or any other body image disturbance. This does not accurately reflect real-world symptomatology and recent research, indicating the potential need to revise DSM-5-TR Criteria. We investigated the co-occurrence of weight- and/or shape-motivated restriction (WSR) in adults who screened positive for ARFID, providing evidence to inform such changes. MethodsThe sample comprised 5,747 adults who consented to participate in the ARFID-Genes and Environment (ARFID-GEN) research study, screened positive for ARFID on the NIAS and PARDI-AR-Q, and completed the EDE-Q. We placed our participants into four groups: groups one and two screened positive for AN (ARFID-AN; n=147) or BN (ARFID-BN; n=193), group three endorsed WSR without meeting AN or BN criteria (ARFID-WSR; n=2,097), and group four endorsed ARFID symptoms only (ARFID-nWSR; n=3,310). We used generalized linear models to test group differences on the NIAS, PARDI-AR-Q and EDE-Q. ResultsARFID-nWSR showed lower scores than all other groups across most ARFID dimensions on the NIAS and PARDI-AR-Q, and lower odds of meeting DSM-5-TR Criteria A1 to A4 (i.e., weight loss; nutritional deficiencies; dependence on nutritional supplements; significant interference with psychosocial functioning). DiscussionThese findings indicate a mixed phenotype with features of both ARFID and WSR associated with more severe ARFID symptomatology. The DSM-5-TR Criteria may not capture complex real-world symptomatology in adults with probable ARFID, potentially precluding those with the most severe symptoms from receiving accurate diagnoses and appropriate care.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.