Development of a Standardized Protocol for Food Preference Assessment in Autism through Direct Observation.

Background:

Food selectivity is a particular feature of the restrictions and stereotypes of Autism. The majority of studies have investigated food preferences and the factors influencing selectivity using caregiver or parent reports such as CEBI (Children’s Eating Behavior Inventory), BAMBI (Brief Autism Mealtime Behavior Inventory), FPI (Food Preference Inventory) or YAQ (Youth/Adolescent Food Frequency Questionnaire, FFQ (Food Frequency Questionnaire) or others. All of these assessments document the presence of food selectivity in Autism subjects when comparing them to typically developing children. Thus far, no study in the literature proposes a standardized direct observation of feeding behavior protocol, which in principal should guarantee better accuracy; hence the purpose of our study.

Objectives:

In this pilot study, we assess the feasibility of a standardized protocol application to explore and monitor food selectivity by directly observing eating behavior in children and adolescents with autism residing at our Rehabilitation Institution.

Methods:

The study sample consisted of ten children and adolescents affected by Autism. The assessment of autism symptom severity was performed through the ADOS scale. Only patients with primary autism (with no cerebral damage or genetic diseases) were included in the final sample (subjects with ADOS Calibrated Severity Scale > 6). Ten subjects affected by mild-moderate intellectual disability not related to autism but residing at the Institution formed the control group. The caregivers present at each participant meal in the dining halls complied food diaries every day for lunch and dinner, carefully notating which foods the subjects accepted and which ones were refused. The observation period lasted four weeks with 20 days monitored (Monday to Friday of each week). The institution’s general menu during this observation period consisted of 39 different serving selections at lunch and 37 at dinner. A comparison between the scores obtained from dietary choice patterns of the two groups and in particular the scores of refused foods was performed using a Mann-Whitney U test; the level of significance was set at p < 0.05. In autism group the Spearman non-parametric test was performed in order to explore correlations between the variables studied.

Results:

Subjects with autism resulted significantly more selective than controls (lunch: p = 0.016, dinner: p = 0.042). Furthermore, children and adolescents with autism were more at risk of becoming underweight or overweight because of unbalanced dietary intake. We found a negative correlation between: food selectivity and duration of stay ( R = - 0.5848 ), as well as food selectivity and age (R = - 0.6437 ), but a positive correlation between food refusal and disease severity measured with ADOS II scale ( R = 0.4441 ).

Conclusions:

Our data confirm the feasibility of a direct observation monitoring protocol for feeding behavior and the importance of food selectivity in subjects with autism. Younger children are more selective than older ones and the duration of institutional residency seems to positively impact this behavior.