89190 AUT18610.1177/1362361313489190AutismSharp et al.

Original Article
Autism

The Autism MEAL Plan: A parent-training 2014, Vol. 18(6) 712­–722

© The Author(s) 2013
Reprints and permissions:
curriculum to manage eating aversions and sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1362361313489190
low intake among children with autism aut.sagepub.com

William G Sharp1,2, T Lindsey Burrell1,3 and David L Jaquess1,2

Abstract
Feeding problems represent a frequent concern reported by caregivers of children with autism spectrum disorders, and
growing evidence suggests atypical patterns of intake may place this population at risk of nutritional and/or related medical
issues, including chronic vitamin and mineral deficiencies, poor bone growth, and obesity. This combination of factors
emphasizes a clear need to identify and disseminate evidence-based treatment of feeding problems associated with autism
spectrum disorders. Behavioral intervention represents an effective treatment for chronic feeding concerns in this population;
however, evidence has largely been established with trained therapists working in highly structured settings. This pilot study
seeks to fill this gap in the literature by describing and evaluating the Autism MEAL Plan, a behaviorally based parent-training
curriculum to address feeding problems associated with autism spectrum disorders. We assessed the feasibility of the
intervention in terms of program content and study protocol (e.g. recruitment and retention of participants, assessment
procedures), as well as efficacy in terms of changes in feeding behaviors. A total of 10 families participated in the treatment
condition, and the program was evaluated using a waitlist control design (n = 9), representing the first randomized-control
study of a feeding intervention in autism spectrum disorders. Results provide provisional support regarding the utility of the
program, including high social validity, parent perception of effectiveness, and reduced levels of caregiver stress following
intervention. Implications, limitations, and future directions for this line of research are discussed.

Keywords
autism, food selectivity, intervention, mealtime problems, parent training, pediatric feeding disorders

Food selectivity (i.e. only eating a narrow variety of foods)
has been widely documented among children with autism
spectrum disorders (ASD), with dietary patterns often
involving strong preferences for starches and snack foods
coinciding with a bias against fruits and vegetables (Ahearn
et al., 2001; Cornish, 2002; Field et al., 2003). Estimates of
atypical intake in ASD reach as high as 90%, suggesting
that feeding problems may occur at epidemic levels in this
population (see Ledford and Gast, 2006; Matson and
Fodstad, 2009; Sharp et al., 2013a for reviews). Children
with ASD often exhibit a strong emotional response when
presented with nonpreferred food, including crying, disrup-
tion, and aggression during meals (Sharp et al., 2013b).
There is also evidence that poor dietary diversity in ASD
may increase the risk of nutritional and/or related medical
issues, including vitamin and mineral deficiencies (Bandini
disease) into adolescence and adulthood (Ho et al., 1997),
both of which are associated with excessive consumption
of snack and fats. Together, such high prevalence of feeding
problems in ASD combined with the risk of long-term
sequelae amplify the need to identify and disseminate evi-
dence-based treatment to remediate these concerns.
Behavioral intervention represents a well-supported
treatment for pediatric feeding disorders (Sharp et al.,
2011a), and there is growing support that this technol-
ogy can be applied to address selective eating patterns
associated with ASD. Retrospective chart reviews
1TheMarcus Autism Center, USA
2Emory University School of Medicine, USA
3Texas Tech University, USA

et al., 2010; Zimmer et al., 2012) and poor bone growth

(Hediger et al., 2008). Selective eating patterns may also Corresponding author:
William G Sharp, Pediatric Psychology and Feeding Disorders Program,
explain increased rates of constipation among children with The Marcus Autism Center, 1920 Briarcliff Road, Atlanta, GA 30329,
ASD (Ibrahim et al., 2009) while also portending elevated USA.
risk of diet-related diseases (e.g. obesity and cardiovascular Email: william.sharp@choa.org
Sharp et al.
conducted by Laud et al. (2009) and Sharp et al. (2011b)
suggest significant improvement in feeding behaviors
following behavioral intervention aimed at expanding
dietary variety among samples of 46 and 13 children
(respectively) treated at intensive feeding programs.
Summaries of single-case reports in the extant litera-
ture also document uniformly positive outcomes asso-
ciated with behavioral intervention to address food
selectivity in ASD, including significant improvements
in the variety and volume of nonpreferred foods (e.g.
vegetables) consumed during meals (Ledford and Gast,
2006). Treatment descriptions typically involve both
antecedent manipulations and consequence-based pro-
cedures combined to maintain the least restrictive envi-
ronment while promoting contact with the sensory
experience of food and exposure to the primary and
secondary reinforcement contingent upon eating (Sharp
et al., 2011a). Consequence-based procedures include
escape extinction (e.g. nonremoval of the spoon or
plate), which involves persisting with a feeding demand
while ignoring refusal behaviors, as well as differential
reinforcement of appropriate mealtime behaviors (e.g.
acceptance and swallowing). Antecedent manipulations
involve stimulus-fading procedures for introducing
variety, texture, and/or volume of food presented dur-
ing meals, as well as simultaneous and sequential pres-
entation of preferred and nonpreferred food items
(Sharp et al., 2011b). Among children with ASD, treat-
ment packages often incorporate these elements with
consideration to the unique cognitive and behavioral
profile associated with ASD (e.g. resistance to change
and heightened sensory defensiveness), such as greater
use of antecedent manipulations and shaping to slowly
introduce new sensory experiences (taste, texture, and
temperature) due to the strong emotional response
associated with the introduction of novel foods (Ledford
and Gast, 2006).
While the available evidence suggests behavioral inter-
vention represents a promising treatment avenue to
address food selectivity among children with ASD, the
extant literature primarily involves children working with
behavioral therapists in highly structured settings (e.g.
inpatient hospital unit and day treatment program), raising
questions whether similar improvements can be achieved
with less intensive approaches. With relatively few pedi-
atric feeding programs spread out geographically, there is
a strong need to develop and evaluate alternative treat-
ment avenues to promote greater access to care. This will
also provide an alternative treatment option for children
with milder feeding concerns whose presentation does not
warrant intensive intervention. Parent training repre-
sents a viable candidate for disseminating this technol-
ogy in such cases, as this mode of treatment has been
successfully applied to other widespread childhood
concerns, including conduct problems (Sanders et al.,
713
2000), toileting issues (Stark et al., 1990), and anxiety
(Creswell and Cartwright-Hatton, 2007). Developing a
parent-directed feeding intervention may also hold eco-
nomical benefits, requiring less frequent and concentrated
clinical contact while increasing the likelihood of applica-
tion and maintenance in the natural environment.
Interventions are more effective with involvement of
family members and teachers rather than specialists
alone (Buschbacher et al., 2004; Horner et al., 2002;
Ingersoll and Dvortcsak, 2006), as parent-directed treat-
ment facilitates generalization to the home environ-
ment, improves parent–child interactions, and maximizes
the amount of intervention children receive (Girolametto
and Tannock, 1994). Research also indicates parents of
children with ASD can effectively implement continuous
treatment, highlighting the potential for parent-directed
feeding intervention to approximate outcomes docu-
mented with behavioral therapists (Koegel et al., 1996).
Caregiver involvement in treatment may also enhance
parent well-being, including increased positive affect
(Koegel et al., 1996; Solomon et al., 2008), reduced stress
(Symon, 2001), and improved self-efficacy (Feldman and
Werner, 2002). This represents a critical issue for caregiv-
ers of children with ASD given the hardships and additional
stressors placed on family members related to the disorder
(Symon, 2001), which include increased stress related to
child dependency and physical limitations (Bouma and
Schweitzer, 1990), disruptions in daily activities, and long-
term child caring responsibilities (Koegel et al., 1992).
Chronic feeding difficulties and related dietary concerns
also represent an additional source of strain on caregivers
(Greer et al., 2008), possible increasing child rearing bur-
den (e.g. preparing multiple menus for each meal), parental
stress, and social isolation. Unfortunately, many families
also lack social support critical for coping with the stressors
in their daily lives (Symon, 2001), suggesting the need to
enhance caregiver skills in managing challenging behavior
while also increasing interaction with peers. With this in
mind, high caregiver involvement in treatment is not only
important for the prognosis of the child but may also hold
benefits for parents as well.
Although parent-directed interventions represent a
promising treatment approach in ASD (Campbell and
Kozloff, 2007), only a handful of studies have docu-
mented parents as the principal agent of change to address
feeding concerns in this population (Najdowski et al.,
2003, 2010). More commonly, generalization of treatment
gains occurs once improvements in dietary variety and
mealtime behaviors have been stabilized with a trained
therapist, as detailed by both Laud et al. (2009) and Sharp
et al. (2011b). In addition, past research has been limited
to individual treatment protocols, raising questions
whether similar positive outcomes can also be achieved in
a group format using a standardized curriculum. Given
that many effective interventions offered for families of
714
children with ASD are costly, time consuming, and may
not initially involve parents (Minjares et al., 2011), treat-
ment offered in a group format may represent a unique,
cost-efficient alternative for disseminating treatment in
the ASD community (McNeil et al., 2005). Given the
critical and ubiquitous role parents play in structuring
meals in the home setting, developing and evaluating a
parent-training curriculum designed to target feeding
concerns in ASD also has potential for broad application
and rapid distribution. With this backdrop in place, the
purpose of this article is to (a) describe a behaviorally
based parent-training program—the Autism MEAL
Plan—a curriculum specifically developed to assist car-
egivers to Manage Eating Aversions and Low intake
among children with ASD; (b) evaluate the feasibility of
the parent-training program including program content
and implementation, recruitment and retention of partici-
pants, and assessment procedures (Leon et al., 2011); and
(c) obtain preliminary outcome data evaluating the social
validity and effectiveness of this curriculum using a wait-
list control design.
Method
Participant recruitment
Potential participants were recruited to take part in a two-
phase project. The first phase of the project focused on
methods for evaluating feeding concerns in ASD through
the use of a multimethod assessment battery described in
detail by Sharp et al. (2013b). Upon completion, the sec-
ond phase of the project evaluated the feasibility and pre-
liminary outcome data of the Autism MEAL Plan, which
is the focus of this study. Recruitment occurred through
local early intervention programs, parent support groups,
and state and local autism organizations through flyers,
list serves, and email. The central inclusion criteria
included an ASD diagnosis (i.e. Asperger’s Disorder,
Pervasive Developmental Disorder–Not Otherwise
Specified, and Autistic Disorder) among children aged
between 3 and 8 years. All participants were diagnosed
by professionals not associated with the program (based
on caregiver report); however, the Social Responsiveness
Scale (SRS) parent report form (Constantino, 2005), a
rating scale measuring the severity of autism spectrum
symptoms (described in the following), was used to con-
firm ASD status. To be included in this study, participants
were required to have a total SRS score in the mild, mod-
erate, or severe range (total standard score (T-score) >
60). The presence of a significant feeding issue was not a
requirement of the study, and we did not screen for feed-
ing issues prior to enrollment in the study; however, we
understood that the subject matter of the parent-training
curriculum was likely to attract families concerned with
their child’s mealtime behaviors.
Autism 18(6)
Measures
The assessment battery included a general questionnaire
gathering background information (e.g. date of birth and gen-
der), feeding concerns or dietary habits, and previously diag-
nosed medical, developmental, or mental health issues.
Nursing staff also obtained anthropometric parameters (i.e.
height and weight) for each child during the first assessment
visit to the clinic. To assess the impact of intervention, we
selected measures likely to be sensitive to detecting proxi-
mate changes in response to intervention as outlined by Sharp
et al. (2013b). This included standardized questionnaires
assessing the following: (a) general mealtime problems spe-
cific to ASD, (b) areas of family life that may be affected by
feeding problems (i.e. parent stress), and (c) dietary diversity
as captured by a Food Preference Inventory (FPI). In addi-
tion, we assessed the social validity of the program as well as
parent perception of improvement with a brief questionnaire
administered at the completion of the study.
SRS–parent report form. The SRS–parent report form (Con-
stantino, 2005) is a 65-item rating scale measuring the
severity of autism spectrum symptoms as they occur in nat-
ural social settings. The instrument yields a total score
(T-score), as well as T-scores on five subscales focusing on
social awareness, social cognition, social communication,
social motivation (e.g. anxiety/avoidance), and autistic
mannerisms (e.g. preoccupations). Scores from 60 to 75
reflect deficiencies that are clinically significant and lead to
mild to moderate interference in everyday social interac-
tions consistent with mild to “high functioning” autism. The
scale has demonstrated adequate reliability and validity.
Brief Autism Mealtime Behavior Inventory. The Brief Autism
Mealtime Behavior Inventory (BAMBI; Lukens and Linsc-
heid, 2008) is a parent report checklist designed to measure
mealtime behavior problems observed in children with
ASD. The 18-item measure employs a Likert scale for
reporting the frequency of behaviors (1 = Never/Rarely to 5
= At Almost Every Meal). The scale yields a total score, as
well as scores on three subscales (i.e. Limited Variety, Food
Refusal, and Features of Autism). Items on the Limited Vari-
ety subscale assess a child’s willingness to try new foods
and food preference by preparation, texture, or type. The
Food Refusal subscale focuses on problem behaviors during
meals (e.g. crying, expelling bite, and disruptions during
meals). Finally, the Features of Autism subscale includes
items that assess inattention, self-injury, and rigid behavior
patterns during meals. The authors reported good internal
consistency, high test–retest reliability, and strong construct
and criterion-related validity in the initial validation study.
FPI. The FPI includes 154 items across 7 food catego-
ries—30 fruits, 28 vegetables, 36 proteins, 27 starches, 8
dairy, 20 miscellaneous/snack (i.e. deserts, fats, and sweets
Sharp et al.
such as cake, cookies, or chips), and 5 combination foods
(e.g. lasagna/ravioli, taco/burrito, or soups/stews). A regis-
tered dietician reviewed the list and classified foods into
each of these categories based on classifications provided
by the United States Department of Agriculture. The inven-
tory employs a Likert-type scale assessing preference for
consumption (e.g. Never, With Prodding, Willing, and
Favorite). Respondents were also given the option of select-
ing “N/A” if an item was not part of the family’s regular
diet or the child lacked exposure or experience with the
food. Consistent with previous research (Sharp et al.,
2013b), we derived a food selectivity score by dividing the
number of foods a caregiver reported the child “never” con-
sumed by total number of items (154) multiplied by 100.
Higher scores reflect greater levels of food selectivity.
Parenting Stress Index–short form. The Parenting Stress
Index–short form (PSI-SF; Abidin, 1995) is a screening
instrument designed to provide an indication of the over-
all level of parenting stress an individual is experiencing
through self-report. A total stress score is derived, as well
as data on three subscales involving 12 items each: Paren-
tal Distress (PD), Parent–Child Dysfunctional Interaction
(P-CDI), and Difficult Child (DC). Analysis in this study
focused exclusively on the total PSI score. Percentile
scores from 81 to 84 are considered in the borderline
range of clinical severity, while scores of 85 or greater are
considered clinically significant. The PSI has been used in
pediatric samples and has exhibited high internal
reliability.
Social validity and parent perception of improvement. We
developed a 10-item posttreatment questionnaire assess-
ing the following: (a) general satisfaction and acceptabil-
ity of the program (e.g. “Overall, how satisfied were you
with the parent group?”) and (b) perceived effectiveness
of the intervention to improve behavior during and out-
side of meals (e.g. “In general, how effective were the
behavioral recommendations at improving your child’s
mealtime behaviors?”). Items were rated on 5-point Lik-
ert-type scale (1 = Quite Dissatisfied/Totally Disagree/
Not at All Effective; 5 = Extremely Satisfied/Totally
Agree/Extremely Effective), with higher scores reflecting
greater levels of satisfaction, perceived improvements/
effectiveness, and acceptance of treatment. All items
included in the scale are presented with outcome date in
the following.
Educational curriculum
The Autism MEAL Plan was modeled after well-estab-
lished, evidence-based training programs successfully
applied to other behavior problems in children (e.g. Neary
and Eyberg, 2002; Sanders et al., 2000). The book Treating
Eating Problems of Children with Autism Spectrum
715
Disorders and Developmental Disabilities: Interventions
for Professionals and Parents by Williams and Fox (2007)
also served as a model for the development of training
activities and handouts, as did treatment protocols used
with children with ASD in our clinical practice (see Sharp
et al., 2011b for a description).
The intervention involved eight, 1-h-long parent-train-
ing group sessions. Sessions were didactic in nature, with
structured content presented during each meeting (See
Table 1 for table of contents). Topics covered in the manual
included general behavior management strategies (e.g. rou-
tine and consistency, positive attending) applied during
meals, specific interventions for feeding problems associ-
ated with ASD (e.g. extinction and stimulus fading), and
strategies for promoting self-feeding (e.g. graduated
prompting and backward chaining). Additional themes
emphasized throughout the curriculum included the follow-
ing: (a) the gradual process of behavior change, (b) the
importance of monitoring behaviors (e.g. ABC charts and
mealtime data sheets), (c) identifying and concretely defin-
ing specific behaviors for intervention (i.e. using opera-
tional definitions), (d) using the child’s behavior to guide
the course of treatment through the use of decision rules,
and (e) the possibility for an increase in problem behavior
following the introduction of treatment (i.e. an extinction
burst). To encourage caregivers to apply new skills in the
home setting, a homework assignment accompanied each
lesson, and the first 10 min of each session involved review-
ing these assignments, addressing questions or concerns,
and providing feedback. We did not, however, collect data
regarding completion of homework or fidelity regarding
assigned activities.
The sequence of topics initially focused on strategies
aimed at establishing conditions favorable for a feeding
intervention, including improving meal structure (time,
location, and length), age-appropriate seating, and general
behavioral management strategies. The families were
encouraged to set up these foundational practices in the
home setting before introducing new foods (a topic not
covered until the sixth week). Families were provided with
a three-ring binder during the first session; however, to
encourage attendance, therapists distributed lessons on a
weekly basis (as opposed to providing the entire manual at
the onset of the group) and provided access to missed les-
sons when a family returned for the next scheduled meet-
ing. Although the curriculum was delivered in a group
format using broad examples and principles, participants
were encouraged to individualize the strategies to meet the
needs of their child. Individualization of intervention was
also emphasized when assigning and reviewing completion
of homework on a weekly basis.
Childcare was provided for families (if needed) during
the educational sessions. The intervention did not involve
live feeding activities with parent–child dyads because
researchers were evaluating the feasibility of delivering a
716 Autism 18(6)

Table 1. The Autism MEAL Plan: overview of sessions.

Session Subject
1 Introduction
2 Structuring meals and monitoring behaviors
3 Ways to increase appropriate behavior
4 Effective communication
5 Ways to decrease inappropriate behavior
during meals
6 Methods of introducing foods
7 Teaching self-feeding skills
8 Monitoring and maintaining progress
Content
Types of eating problems
Antecedents and consequences that affect feeding
Modeling
Parenting stress
Planning an intervention
Homework: selecting a target for intervention
Routine and consistency: meal schedules, location, and length
Monitoring your child’s intake
Tracking your child’s mealtime behavior: ABC charting and behavioral
definitions
Homework: ABC chart—tracking mealtime interactions
Positive attending and specific praise
Tangible rewards
Homework: catching your child’s good behavior during meals
Rules for effective commands
General strategies for effective communication
Homework: changing ineffective commands to effective commands
worksheet
Selective ignoring and differential attention
Withdrawing positive reinforcement
Nonremoval of the spoon and exit criterion
Homework: practicing selective ignoring
Modifying the bite size, meal variety, and food texture
Simultaneous presentation and variety fading
Homework: introducing new foods during a meal
Prompting strategies
Four-step prompting sequence
Backward chaining
Homework: implementing the four-step prompting sequence
Tracking progress
What to look for when problems arise
Dietary considerations

parent-only education group. We provided the educational Design and procedure

curriculum and childcare at no cost and offered two group
sessions of the program each week (i.e. morning and even-
ing) in order to accommodate possible scheduling conflicts
and to maximize participation. Three families attended the
Monday morning session and seven families attended the
Wednesday evening session.
The first author (W.G.S.), a behavioral psychologist
with expertise in the treatment of pediatric feeding disor-
ders who developed the curriculum, conducted the sessions
with assistance from a postdoctoral psychology fellow. A
fidelity checklist was utilized to ensure that the clinician
covered all topics of the curriculum each week. Both thera-
pists also took detailed notes regarding any new topics (e.g.
use of time-out during and outside of meals) introduced
during group discussion to ensure this information was dis-
seminated to both cohorts, as well as to provide a reference
during program evaluation and revision.
This study was conducted in compliance with a university-
based Institutional Review Board (IRB). Participation in
the study was voluntary, with no incentives for participa-
tion outside the educational curriculum. We obtained writ-
ten consent from all participants prior to enrollment. We
evaluated treatment using a waitlist control design. During
preintervention, all participants participated in the assess-
ment battery described by Sharp et al. (2013b) and, when
complete, families were randomly assigned to either receive
the treatment condition or to a waitlist control. During the
second stage of the project, the treatment group participated
in the 8-week educational curriculum. We collected attend-
ance data at each session, including the number of caregiv-
ers present. In an effort to boost retention, the waitlist
control group received email correspondence involving
handouts on nonfeeding-related topics with limited
Sharp et al.
Completed Initial
Evaluation and
Randomized
N = 30
Autism MEAL plan
N = 15
Discontinued
N=5
Completed
N = 10
Figure 1. Flow diagram of parent education versus waitlist control.
behavioral content (e.g. limiting television watching and
toileting recommendations). All participants completed a
postintervention assessment battery following completion
of the educational curriculum, and the waitlist group was
subsequently offered the educational curriculum, followed
by a final evaluation.
Statistical analyses
To determine whether there were preexisting differences
between the treatment and waitlist control groups, as well
as those participants lost to attrition, we compared groups
using analysis of variance (ANOVA) on all preintervention
dependent measures. We also calculated descriptive statis-
tics (means and standard deviations) to provide a qualita-
tive summary of pre- and postintervention scores for the
treatment and waitlist control groups. Finally, to evaluate
the degree of change in parenting stress (as captured by the
PSI) and feeding behaviors (as reflected by the BAMBI
total score and subscales and the FPI selectivity score), we
conducted analysis of covariance (ANCOVA) comparing
postscores between the treatment group and the waitlist
control group on each dependent measure statistically con-
trolling for variation in baseline scores. We also calculated
effect size estimates (Cohen’s d) for each outcome variable
based on differences between groups using pre/post-change
scores and evaluated the magnitude of the treatment effect
using conventional standards (0.2 = small, 0.5 = medium,
0.9 = large; Cohen, 1988).
Results
Feasibility outcomes
Study attrition, description of participants, and session attendance.
The CONSORT diagram (Figure 1) displays the flow of
participants through the study. A total of 30 families com-
pleted the first assessment phase (see Sharp et al. (2013b)
717
Waitlist Control
N = 15
Discontinued
N=6
Completed
N=9
for sample description) and were assigned to either the
treatment condition or the waitlist control condition using
computer-generated block randomization to promote equal
number of participants in each group (n = 15). Participant
attrition, however, impacted the number of families who
completed the study, with 19 participants continuing in the
study following randomization. Ten families participated in
the educational curriculum and completed the second
assessment battery. The five families who discontinued
participation following random assignment (but before
attending a parent-training session) cited scheduling issues
as prohibiting further involvement in the study. In the wait-
list control condition, nine families returned to complete
the second evaluation following the 8-week break. We
were unable to contact two of the families, and the remain-
ing four families cited scheduling as a barrier to continued
participation in the study. The 11 families lost to attrition
were not different on any baseline measures compared to
those who completed the study.
There were no significant differences between the
groups (as well as those lost to attrition) on key baseline
characteristics or feeding behaviors (Table 2). The majority
of the children in the study were male, representing 15 of
the 19 children (79%). The average body mass index (BMI)
for age percentile of the sample, calculated based on Center
for Disease Control and Prevention’s BMI-for-age growth
charts (Kuczmarski et al., 2000) using weight and height
obtained during the initial assessment fell within the healthy
weight range (5th to 85th percentile). This suggests no
growth concerns as a group (i.e. underweight or obesity),
which is consistent with research indicating increased risk
of feeding problems (e.g. extreme tantrums during meals,
severe food selectivity, and ritualistic mealtime behaviors)
in ASD and does not necessarily translate into greater risk
of compromised growth (Sharp et al., 2013a). Notably,
scores on the PSI fell in the clinically significant range for
the treatment group and in the borderline range for the wait-
list control group, suggesting high levels of stress among
718 Autism 18(6)

Table 2. Baseline characteristics by group.

Characteristic Treatment (n = 10) Waitlist (n = 9) Attrition (n = 11) p-value(F)

M (SD)/range M (SD)/range M (SD)/range

Male sex, n (%) 8 (80) 7 (78) 8 (73)
Age (in months) 70.8 (20.5)/36–104 64.8 (16.9)/45–94 70 (15.8)/43–93 .73 (.32)
SRS total score 82.4 (8.4)/70–91 80.6 (7.9)/68–91 86.7 (6.6)/78–91 .19 (1.75)
BMI/age % 47 (32.8)/11–93 66.8 (36.4)/9–99 68 (33.6)/6–99 .31 (1.24)
PSI total score 89.3 (7.8)/75–99 79.9 (20.5)/50–99 89 (5.8)/80–95 .49 (.73)
BAMBI total score 51.1 (7.1)/41–61 52.1 (7.8)/36–62 45.6 (14.7)/18–75 .35 (1.09)
BAMBI limited variety 28.2 (5.1)/22–34 28.2 (5.1)/20–34 24.2 (8.4)/8–34 .27 (1.39)
BAMBI food refusal 12.9 (3.5)/8–18 11.9 (3.3)/8–18 10.4 (4.5)/5–21 .33 (1.15)
BAMBI autism features 10.0 (2.1)/7–13 12.0 (3.54)/7–19 11.0 (4.0)/5–20 .44 (.85)
FPI selectivity score 32.6 (22.3)/5–75 37.2 (17.8)/12–63 41.6 (25.8)/4–81 .67 (.42)

SD: standard deviation; SRS: Social Responsiveness Scale; BMI: body mass index; PSI: Parenting Stress Index; BAMBI: Brief Autism Mealtime Behavior
Inventory; FPI: Food Preference Inventory.

Table 3. Caregiver ratings of the Autism MEAL Plan post intervention.

Item Treatment (n = 6) Waitlist (n = 6)
1. O
 verall how satisfied are you with the Marcus Feeding Parent Group? 4.2 4.2
2. In general, how effective were the behavioral recommendations in improving your 4.2 4.3
child’s mealtime behavior?
3. In general, how effective were the behavioral recommendations in improving your 3.3 4.0
child’s behavior outside of meals?
4. A t home, my family will continue to use the behavioral recommendations from this 4.3 4.8
program.
5. C ompared to when we started the program, my child’s feeding/behavior is much 4.0 4.0
improved.
6. If a friend was in need of a similar help, would you recommend the Parent Group to 4.8 4.8
him/her?
7. S oon after using the intervention, you noticed a change in eating behavior that was 3.0 3.7
much better.
8. T his is an acceptable intervention for my child’s eating behavior. 4.0 4.5
9. T he program improved my child’s eating/target behavior so it is not much different 2.8 3.3
from others.
10. A t a restaurant, we plan to use the behavioral treatment program. 4.0 4.3

caregivers participating in the study. In terms of attendance,
83% of sessions were attended by at least one caregiver
from the 10 families in the treatment group. Mothers repre-
sented the modal participant, attending 80% of the sessions,
while fathers attended 37% of the sessions. Four families
had both parents attend at least three sessions.
Social validity and caregiver satisfaction. All families who
completed the educational curriculum (treatment and wait-
list) were asked to complete the posttreatment satisfaction
questionnaire. We administered this questionnaire to both
treatment and waitlist groups at the same point in time (e.g.
after all participants were exposed to the curriculum),
which (in hindsight) reduced the number of treatment fami-
lies (n = 6) who remained in contact with the study follow-
ing the additional 8 weeks required for the waitlist group to
complete parent training. This resulted in a total of 12
families (63% of the total group) who provided feedback
regarding program satisfaction, treatment gains, and social
acceptability (Table 3). Ratings for the available families
suggest a high degree of social validity associated with
Autism MEAL Plan. We also encouraged caregivers to pro-
vide a narrative summary regarding their perception of the
training program. Feedback included both positive state-
ments (e.g. “Awesome program!”) as well as recommenda-
tions for future modifications to the format of the group
(e.g. hands-on training with child present and email com-
munication between sessions).
Efficacy outcomes
Upon completion of the Autism MEAL Plan, caregivers in
the treatment group reported a significant reduction (F (1,
16) = 7.6, p = .01) in PSI scores compared to the waitlist
Sharp et al. 719

Table 4. Mean (SD) levels of parenting stress and feeding behaviors before and after intervention.

Characteristic Treatment Waitlist p-value (F)

Pre Post Pre Post

PSI total score 89.3 (7.8) 81.0 (14.1) 79.9 (20.5) 80.3 (24.9) .01 (7.6)
BAMBI total score 51.1 (7.1) 47.2 (9.6) 52.1 (7.8) 47.2 (12.6) .79 (.07)
BAMBI limited variety 28.2 (5.1) 26.0 (5.2) 28.2 (5.1) 26.8 (6.6) .55 (.36)
BAMBI food refusal 12.9 (3.5) 12.6 (4.1) 11.9 (3.3) 11.0 (3.0) .51 (.46)
BAMBI autism features 10.0 (2.1) 8.6 (2.0) 12.0 (3.54) 9.5 (3.6) .57 (.34)
FPI selectivity score 32.6 (22.3) 38.8 (27.5) 37.2 (17.8) 37.2 (25.9) .21 (1.7)

SD: standard deviation; PSI: Parenting Stress Index; BAMBI: Brief Autism Mealtime Behavior Inventory; FPI: Food Preference Inventory.

Table 5. Summary of key findings and recommendations for future research.

Findings suggest the Autism MEAL Plan

1. May lead to a reduction in total parenting stress
2. Represents a socially valid means to disseminate this technology
3. Holds potential cost-savings benefits in terms of total clinical contact
Future research in this area should
1. Involve direct observation or live feeding activities to assess procedural fidelity and enhance training
2. Include detailed diagnostic characterization to confirm ASD status and determine intellectual status
3. Identify and incorporate feeding screening measures to quantify feeding concerns and determine appropriateness of intervention
4. Increase the number of children exposed to the curriculum while concurrently identifying means to reduce attrition
5. Conduct larger efficacy studies, including determining the external validity of the curriculum to other settings and therapists

ASD: autism spectrum disorders.

control group after controlling for preintervention PSI lev-
els (Table 4). The magnitude of this effect fell in the large
range by conventional standards (d = 1.1). There were no
significant changes detected in terms of feeding behaviors,
as captured by the BAMBI total score and the three BAMBI
subscales, or dietary variety based on the FPI selectivity
score.
Discussion
This pilot study describes a parent-training curriculum,
entitled the Autism MEAL Plan, aimed at teaching caregiv-
ers of children with ASD to develop and implement behav-
iorally based feeding interventions. This represents a novel
approach to address the ubiquity of feeding problems in
this population, as previous treatment descriptions either
involved intensive intervention occurring at specialized
feeding disorder programs or case studies describing indi-
vidualized protocols. By developing a manual-based cur-
riculum involving minimal clinical contact, we sought to
provide an economical alternative with the potential for
replication and rapid dissemination in the ASD community.
The potential cost-savings benefits of a parent-directed
feeding intervention in a group format is highlighted by the
reduction in clinical contact associated with the program,
with 8 h of group sessions (vs 80 h of individual sessions)
required to serve the 10 families in the treatment condition.
Preliminary results also support the Autism MEAL Plan as
a promising treatment avenue for delivering this technol-
ogy, with caregiver feedback indicating high overall satis-
faction with the program in terms of both content and
format of the intervention. Following treatment, caregivers
also reported a significant reduction in overall stress, which
is consistent with previous research indicating caregiver
involvement in treatment has the potential to enhance par-
ent well-being and may also portend increased likelihood
of follow-through with a parent-directed feeding interven-
tion. Although parenting stress was not the primary target
of intervention, research suggests that the reduction in par-
enting stress can positively influence the effectiveness of
the treatment being implemented, as well as potentially
provide additional benefits such as more positive parent–
child interactions (Brookman-Frazee, 2004) and improved
self-efficacy (Feldman and Werner, 2002). Together, the
feasibility information yielded by this pilot study suggests
that Autism MEAL Plan may represent an economical and
socially valid intervention for feeding problems in ASD,
while also providing important guidance for future research
in this area outlined in the following (See Table 5 for
summary).
720
Despite positive parent feedback and reduced levels of
stress, we did not observe a change in mealtime behaviors
(as captured by the BAMBI) or dietary variety. In fact,
although not statistically significant, the mean number of
foods identified as nonpreferred actually increased for the
treatment group following intervention. The lack of
improvement in feeding outcomes, which represent the
ultimate target of the Autism MEAL Plan, raises important
questions regarding the apparent discordance with high
social validity, parent perception of effectiveness, and
reduced caregiver stress associated with the program.
Limitations in study design, including both statistical power
and timeframe of measurement, likely account for some of
this discrepancy. As noted above, the first half of the inter-
vention focused on foundational behavioral principles (e.g.
routine and consistency, selective ignoring), while methods
for introducing new foods into meals did not occur until the
sixth week. The program also emphasized the use of ante-
cedent manipulations and shaping to slowly introduce new
foods. Long-term follow-up, therefore, may be required to
detect more distal shifts in dietary intake. This may also
help explain the observed increase in the number of non-
preferred foods associated with intervention, as caregivers
may have been more likely to present new foods in the
home setting, but not yet able to achieve stable rates of
acceptance due to the likelihood of an extinction burst and/
or the use of stimulus fading and shaping to gradually shift
dietary preference.
The study was also limited by the lack of a training
manipulation check, raising important questions regarding
whether possible protocol drift influenced outcome varia-
bles, as well as the curriculum’s effectiveness in promoting
skill acquisition. Direct observation represents one method
to assess procedural fidelity, providing information regard-
ing parents’ implementation of behavioral strategies as well
as their child’s response to intervention. Adding live feed-
ing activities to the curriculum was also recommended by
many caregivers during postintervention feedback, sug-
gesting that activities such as coaching, modeling, and/or
role plays would enhance the overall training experience
for participants. Procedural fidelity could also be assessed
by collecting detailed information regarding homework
completion, including the extent to which parents used rec-
ommended skills outside of the sessions and/or completed
weekly assignments. Finally, development of a knowledge
test of program content would permit analysis of parents’
acquisition of behavioral principles and effectiveness of the
current format in disseminating this information.
Future research should also seek to increase the number
of children exposed to the curriculum (and reduce attrition)
in order to enhance power and provide a more robust
assessment of intervention. Identifying additional methods
to maximize retention should be a priority to reduce possi-
ble threats to internal validity. Modifying the recruitment
process represents one potential avenue to reduce attrition.
Autism 18(6)
For example, providing families interested in the study
with predetermined days/times for group sessions prior to
enrollment (vs following the initial assessment) may help
minimize scheduling conflicts. In addition, requiring the
presence of feeding concerns as part of the inclusion crite-
ria—possibly through the use of a severity threshold for
enrollment (e.g. accepts < four fruits or vegetables)—may
help assure participants are highly motivated to participate
in an intervention aimed at addressing mealtime difficulties
in ASD. Once the study begins, retention and attendance
could also be enhanced by increasing engagement with
families throughout the study through phone calls and/or
email, including providing consultation between sessions
as suggested by many participants during postintervention
feedback.
Finally, it will be important to better characterize partici-
pants through standardized developmental diagnostic test-
ing and feeding evaluation to provide more detailed
information regarding ASD symptomatology, level of func-
tioning, and severity of feeding concerns. Enhanced assess-
ment would permit questions regarding what subgroup(s)
of children best benefit from this level of intervention, such
as establishing a threshold for children appropriate for a
parent-directed intervention. This will require the develop-
ment of a frontline feeding screen tool to quantify the mag-
nitude of mealtime difficulties associated with ASD, as lack
of standardized measures represents a significant barrier to
research in this area (Matson and Fodstad, 2009). With
food selectivity representing a primary feeding concern
associated with ASD, dietary diversity, as captured through
measures such as the BAMBI’s Limited Variety subscale or
the number of foods identified as never consumed on the
FPI, should continue to be included in outcome measures of
treatment efficacy. In addition, parent performance, knowl-
edge, and application of behavioral principles should be
accessed through the use of direct and/or indirect assess-
ment methods. Direct observation of parent–child mealtime
interactions would necessitate the identification of a coding
system to track and capture both parent and child behavior,
which could be modeled after procedural integrity proce-
dures outlined in single-subject research (e.g. Najdowski et
al., 2010). Outcomes should also assess more general
changes in the family environment, such as changes in par-
enting stress, family routine (e.g. increased visits to restau-
rants or social gatherings involving food), parent–child
interactions, and child behavior outside the context of
meals (e.g. improvements in externalizing behaviors).
With these limitations in mind, this study represents one
of the few empirical evaluations of a feeding intervention
using a group design (See Sharp et al., 2011a for review)
and the only specifically targeting feeding problems in
ASD. Furthermore, it involves one of the largest cohorts of
children with ASD in the feeding intervention literature,
with the exception of retrospective chart reviews of inten-
sive feeding programs (Laud et al., 2009; Sharp et al.,
Sharp et al.
2011b). This highlights the paucity of research in this area
despite high prevalence estimates and the potential for
long-term health sequelae associated with chronic feeding
problems. It also reflects the need for a greater research
thrust focusing on feeding problems in ASD in order to
address key questions regarding the assessment, long-term
consequences, and remediation of atypical patterns of
intake in ASD (Sharp et al., 2013a).
In sum, preliminary evidence suggests that the Autism
MEAL Plan has the potential to fill an important gap in the
treatment literature, providing an alternative option to ther-
apist-driven, clinic interventions for children with milder
feeding concerns, which do not warrant admission to an
intensive feeding program. Feedback from participants, as
well as therapists’ experience implementing the protocol,
provides guidance regarding potential modifications to
training activities going forward. These include greater use
of supplemental learning activities, such as role-playing,
video examples, and in vivo coaching. We will be revising
and reevaluating the training curriculum with these consid-
erations in mind. Finally, the mechanism(s) responsible for
the observed reduction in stress levels should be more fully
explored, with potential candidates including enhanced
self-efficacy related to skill acquisition and/or increased
levels of social support provided by interaction with the cli-
nicians and/or peers.
Funding
This project was funded by a 2008 Applied Research Grant spon-
sored by the Organization for Autism Research.
References
Abidin RR (1995) Parenting Stress Index (PSI) Manual. 3rd ed.
Lutz, FL: Psychological Assessment Resources, Inc.
Ahearn WH, Castine T, Nault K, et al. (2001) An assessment of
food acceptance in children with autism or pervasive devel-
opmental disorder–not otherwise specified. Journal of Autism
and Developmental Disorders 31: 505–511.
Bandini, LG, Anderson, SE, Curtin, C, Cermak, et al. (2010) Food
selectivity in children with autism spectrum disorders and
typically developing children. Journal of Pediatrics 157(2):
259–264.
Bouma R and Schweitzer R (1990) The impact of chronic child-
hood illness on family stress: a comparison between autism
and cystic fibrosis. Journal of Clinical Psychology 46: 722–
730.
Brookman-Frazee L (2004) Using parent/clinician partnerships in
parent education programs for children with autism. Journal
of Positive Behavior Interventions 6(4): 195–213.
Buschbacher P, Fox L and Clarke S (2004) Recapturing desired
family routines: a parent-professional behavioral collabo-
ration. Research and Practice for Persons with Severe
Disabilities 29(1): 25–39.
Campbell M and Kozloff M (2007) Comprehensive programs for
families of children with autism. In: Briesmeister JM and Charles
E (eds) Helping Parents Prevent and Solve Problem Behaviors
(pp. 67–106). 3rd ed. Hoboken, NJ: John Wiley & Sons Inc.
721
Cohen J (1988) Statistical Power Analysis for the Behavioral
Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates.
Constantino JN (2005) The Social Responsiveness Scale (SRS).
Los Angeles, CA: Western Psychological Services.
Cornish E (2002) Gluten and casein free diets in autism: a study
of the effects on food choice and nutrition. Journal of Human
Nutrition and Dietetics 15: 261–269.
Creswell C and Cartwright-Hatton S (2007) Family treatment
of child anxiety: outcomes, limitations, and future direc-
tions. Clinical Child and Family Psychology Review 10:
232–252.
Feldman MA and Werner SE (2002) Collateral effects of behavio-
ral parent training on families of children with developmental
disabilities and behavioral disorders. Behavioral Interventions
17(2): 75–83.
Field D, Garland M and Williams K (2003) Correlates of specific
childhood feeding problems. Journal of Paediatrics and Child
Health 39: 299–304.
Girolametto L and Tannock R (1994) Correlates of directive-
ness in the interactions of fathers and mothers of children
with developmental delays. Journal of Speech and Hearing
Research 37(5): 1178–1191.
Greer AJ, Gulotta CS, Masler EA, et al. (2008) Caregiver stress
and outcomes of children with pediatric feeding disorders
treated in an intensive interdisciplinary program. Journal of
Pediatric Psychology 33: 612–620.
Hediger ML, England LJ, Molly CA, et al. (2008) Reduced bone
cortical thickness in boys with autism or autism spectrum dis-
order. Journal of Autism and Developmental Disorders 38:
848–856.
Ho HH, Eaves LC and Peabody D (1997) Nutrient intake and
obesity in children with autism. Focus on Autism and Other
Developmental Disabilities 12: 187–192.
Horner RH, Carr EG, Strain PS, et al. (2002) Problem behavior
interventions for young children with autism: a research syn-
thesis. Journal of Autism and Developmental Disorders 32(5):
423–446.
Ibrahim SH, Voigt RG, Katusic SK, et al. (2009) Incidence of gas-
trointestinal symptoms in children with autism: a population-
based study. Pediatrics 124: 680–686.
Ingersoll B and Dvortcsak A (2006) Including parent training in
the early childhood special education curriculum for children
with autism spectrum disorders. Journal of Positive Behavior
Interventions 8(2): 79–87.
Koegel LK, Koegel RL and Suratt A (1992) Language inter-
vention and disruptive behavior in preschool children with
autism. Journal of Autism and Developmental Disorders 22:
141–153.
Koegel RL, Bimbela A and Shreibman L (1996) Collateral effects
of parent training on family interactions. Journal of Autism
and Developmental Disorders 33: 3–13.
Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al.
(2000) CDC growth charts: United States. Advance Data
314: 1–27.
Laud RB, Girolami PA, Boscoe JH, et al. (2009) Treatment out-
comes for severe feeding problems in children with autism
spectrum disorder. Behavior Modification 33: 520–536.
Ledford JR and Gast DL (2006) Feeding problems in children
with autism spectrum disorders: a review. Focus on Autism
and Other Developmental Disabilities 21: 153–166.
722
Leon AC, Davis LL and Kraemer HC (2011) The role and inter-
pretation of pilot studies in clinical research. Journal of
Psychiatric Research 45: 626–629.
Lukens CT and Linscheid TR (2008) Development and valida-
tion of an inventory to assess mealtime behavior problems in
children with autism. Journal of Autism and Developmental
Disorders 38: 342–352.
McNeil CB, Herschell AD and Gurwitch RH (2005) Training fos-
ter parents in parent-child interaction therapy. Education &
Treatment of Children 28(2): 182–196.
Matson JL and Fodstad JC (2009) The treatment of food selectiv-
ity and other feeding problems in children with autism spec-
trum disorders. Research in Autism Spectrum Disorders 3(2):
455–461.
Minjares MB, Williams SE, Mercier EM, et al. (2011) Pivotal
response group treatment program for parents of children
with autism. Journal of Autism and Developmental Disorders
41(1): 92–101.
Najdowski AC, Wallace MD, Doney JK, et al. (2003) Parental
assessment and treatment of food selectivity in natural set-
tings. Journal of Applied Behavior Analysis 36: 383–386.
Najdowski AC, Wallace MD, Reagon K, et al. (2010) Utilizing a
home-based parent training approach in the treatment of food
selectivity. Behavioral Interventions 25: 89–107.
Neary EM and Eyberg SM (2002) Management of disruptive
behavior in young children. Infants and Young Children
14(4): 53–67.
Sanders MR, Markie-Dadds C, Tully LA, et al. (2000) The tri-
ple P-positive parenting program: a comparison of enhanced,
standard, and self-directed behavioral family intervention for
parents of children with early onset conduct problems. Journal
of Consulting and Clinical Psychology 68(4): 624–640.
Sharp WG, Berry RC, McCracken C, et al. (2013a) Feeding prob-
lems and nutrient intake in children with autism spectrum
Autism 18(6)
disorders: a meta-analysis and comprehensive review of the
literature. Journal of Autism and Developmental Disorders.
Epub ahead of print 1 February 2013. DOI: 10.1007/s10803-
013-1771-5.
Sharp WG, Jaquess DL and Lukens CT (2013b) Multi-method
assessment of feeding problems among children with autism
spectrum disorders. Research in Autism Spectrum Disorders
7(1): 56–65.
Sharp WG, Jaquess DL, Morton JF, et al. (2011a) Pediatric
feeding disorders: a quantitative synthesis of treatment out-
comes. Clinical Child and Family Psychology Review 13:
348–365.
Sharp WG, Jaquess DL, Morton JF, et al. (2011b) A retrospective
chart review of dietary diversity and feeding behavior of chil-
dren with autism spectrum disorder before and after admis-
sion to a day treatment program. Focus on Autism and Other
Developmental Disabilities 26: 37–48.
Solomon M, Ono M, Timmer S, et al. (2008) The effectiveness
of parent-child interaction therapy for families of children on
the autism spectrum. Journal of Autism and Developmental
Disorders 38: 1767–1776.
Stark LJ, Owens-Stively J, Spirito A, et al. (1990) Group treat-
ment of retentive encopresis. Journal of Pediatric Psychology
15: 659–671.
Symon JB (2001) Parent education for autism: issues in pro-
viding services at a distance. Journal of Positive Behavior
Interventions 3(3): 160–174.
Williams KE and Fox RM (2007) Treating Eating Problems of
Children with Autism Spectrum Disorders and Developmental
Disabilities. Austin, TX: Pro-Ed.
Zimmer MH, Hart LC, Manning-Courtney P, et al. (2012) Food
variety as predictor of nutritional status among children with
autism. Journal of Autism and Developmental Disorders
42(4): 549–556.