Long-term follow-up study of low-weight avoidant/restrictive food intake disorder and childhood-onset anorexia nervosa: comparison of autistic eating behaviours.

backgroundAnorexia nervosa (AN) is characterized by restrictive eating associated with weight and shape concerns, whereas avoidant/restrictive food intake disorder (ARFID) is not. Evidence supports an association between AN and autism, and emerging data suggest a similar overlap between ARFID and autism. However, long-term data on autistic eating behaviours in these groups remain scarce. This study examined autistic eating behaviours at long-term follow-up in former patients with childhood-onset AN and low-weight ARFID.

methodsParticipants were 56 former patients treated in specialist child and adolescent eating disorder services in Sweden (37 with AN and 19 with ARFID at treatment start) who were assessed at a mean follow-up of 15.9 years. Autistic eating behaviours were measured with the Swedish Eating Assessment for Autism Spectrum Disorders questionnaire (SWEAA; primary measure). Secondary measures were the Morgan-Russell Outcome Assessment Schedule (M-R OAS) and the Eating Disorder Examination Questionnaire (EDE-Q) [21, 27]. Follow-up assessment also included a structured diagnostic interview (SCID) and measured height and weight. ARFID diagnoses were assigned retrospectively using all available clinical information. Mann-Whitney U tests, Kruskal-Wallis tests, descriptive statistics, and linear regression were used for group comparisons and association analyses.

resultsBoth the AN and ARFID groups showed elevated SWEAA scores at follow-up, with no significant difference between groups. Participants with an ongoing eating disorder at follow-up had the highest SWEAA scores, whereas those with no diagnosis had the lowest scores. In both diagnostic groups, the body mass index (BMI) < 20 category showed the lowest SWEAA scores. In the ARFID group, higher BMI was positively associated with SWEAA total score. M-R OAS scores differentiated follow-up diagnostic groups, whereas EDE-Q scores showed no clear pattern across groups.

conclusionsAutistic eating behaviours were common at long-term follow-up in both childhood-onset AN and low-weight ARFID. Higher SWEAA scores were associated with concurrent psychopathology, particularly an ongoing eating disorder, but autistic eating behaviours were not more pronounced in AN than in ARFID. The unexpectedly low SWEAA scores in the low-BMI group warrant further investigation.