Outcomes for ASD Children with Feeding Problems in a Community Mental Health Setting

Background:  Although not a core feature of Autism Spectrum Disorder (ASD), children with ASD often have restricted food intake problems and disruptive mealtime behaviors. Escape extinction, a common evidence-based practice for restrictive food intake disorder for children with ASD, is controversial and considered unethical for many clinicians because of the aversive nature of the technique. Because of this, researchers and clinicians are striving to create and evaluate new feeding treatments for this population. An alternative feeding intervention with limited support in the ASD population is known as the Sequential Oral Sensory (SOS) feeding program. In an effort to improve the feeding behaviors of preschool children with ASD, a community mental health preschool program implemented a variation of the SOS intervention to determine its effectiveness. The feeding intervention is an intensive 8-month program, consisting of 30-minute sessions, 4 days per week, in a school cafeteria setting.

Objectives:  We report findings from a longitudinal archival dataset collected during the implementation of a feeding intervention as part of a community mental health preschool program children with ASD. We examined the relationship between feeding behaviors and other client characteristics including age, sex, autism symptom severity, and cognitive ability. Discrepancies between clinician and caregiver reports of symptoms were also analyzed.

Methods: Sixty children ages 35 to 59 months diagnosed with ASD received a feeding intervention from a community mental health preschool program for eight months. The archival dataset included monthly administrations of the Brief Autism Mealtime Behavior Inventory (BAMBI) from both caregivers and clinicians. Severity of autism symptoms (Clinical Autism Rating Scale) and general cognitive ability (Psycho-Educational Profile) were also collected at intake and post-treatment. We used Hierarchical Linear Modeling (HLM) techniques to model the changes in BAMBI scores across the 8 months of treatment to assess effectiveness of the feeding intervention. We also used HLM to examine autism severity, age, cognitive ability, and gender of the client as potential predictors of treatment outcomes. Differences between caregiver and clinician reports of feeding behaviors at intake and post-treatment were examined using t-tests.

Results: Caregiver reported BAMBI score indicated that at post-treatment, 3 of the 53 children with significant feeding problems at intake no longer had significant problems post-treatment. This finding was not corroborated on the clinician report. HLM analyses show that reported problematic feeding behaviors over the course of treatment decreased significantly on caregiver report, but not on clinician report. Food selectivity significantly decreased on both caregiver and clinician reports. Cognitive ability at intake significantly predicted decreased food selectivity on the clinician report. Age, autism severity, and gender did not significantly predict treatment outcomes. Clinicians rated the feeding problems as more severe at intake than the caregivers. There were no significant differences between reporters at post-treatment.

Conclusions: The intensive feeding program implemented at a community mental health preschool program for children with ASD showed minimal effectiveness. This study highlights the difficulty of effective implementation of feeding interventions in community mental health settings. Implications, limitations, and future directions for research will be discussed.