Download as pdf or txt
Download as pdf or txt
You are on page 1of 22

Journal of Applied Behavior Analysis 2020, 9999, 1–22 NUMBER 9999 ()

Treating food selectivity as resistance to change in children with


autism spectrum disorder
Jaime G. Crowley, Kathryn M. Peterson, Wayne W. Fisher and
Cathleen C. Piazza
University of Nebraska Medical Center’s Munroe-Meyer Institute

Change-resistant behavior, such as rigid and selective food consumption, is a core symptom of
autism that can have significant negative consequences for the child (Flygare Wallén et al.,
2018; Levy et al., 2019). In the current study, we used a matching-law-based intervention
(Fisher et al., 2019) to treat the change-resistant feeding behavior of 7 young children with
autism. The feeder gave the participant a choice between a change-resistant and an alternative
food during free- and asymmetrical-choice conditions. Alternative-food consumption increased
for 2 participants during asymmetrical choice when the feeder provided a preferred item for con-
suming the alternative food and no programmed consequence for consuming the change-
resistant food. Alternative-food consumption increased for the other 5 participants after the
feeder exposed at least 1 food to single choice in which the feeder guided the participant to put
the bite of alternative food in his or her mouth if he or she did not do so within 8 s of presenta-
tion. Effects of the single-choice contingencies maintained during reversals and generalized to
other alternative foods the feeder did not expose to single choice. These results are important
because participants consumed alternative foods even when their change-resistant foods were
present, which is similar to typical mealtime contexts in which children have choices among
foods.
Key words: autism, avoidant/restrictive food intake disorder, food selectivity, resistance to
change

According to the Centers for Disease Con- symptoms improve (Joseph et al., 2013; Richler
trol and Prevention, 1 in 59 children in the et al., 2010). relevância social

United States is diagnosed with autism spec- Effective intervention for change-resistant
trum disorder (Baio et al., 2018). Resistance to behavior is important because life can be
change, characterized by emotional outbursts in unpredictable, preventing change is nearly
response to even small environmental changes, impossible, and change-resistant behavior can
is a core symptom of autism that occurs fre- have significant short- and long-term negative
quently in many contexts and stays the same or consequences (Fisher et al., 2019; Flygare
worsens without intervention, even when other Wallén et al., 2018; Freedman et al., 1999;
Kamal Nor et al., 2019; Levy et al., 2019).
Jaime G. Crowley is now at The May Institute. Children with autism frequently demonstrate
Kathryn M. Peterson is now at Children’s Specialized change-resistant behavior at mealtime, such as
Hospital. Wayne W. Fisher is now at Children’s Special-
ized Hospital–RUCARES and Rutgers Robert Wood insistence on eating the same few unhealthy
Johnson Medical School. Cathleen C. Piazza is now at foods or engaging in rigid mealtime routines
Children’s Specialized Hospital and Rutgers Graduate (Bandini et al., 2010; Cermak et al., 2010;
School of Applied and Professional Psychology.
Jaime G. Crowley conducted this study in partial fulfill-
Heifert et al., 2016; Levin & Carr, 2001; Post-
ment of the requirements for the degree of Doctor of Phi- orino et al., 2015; Schreck & Williams, 2006;
losophy from the University of Nebraska Medical Center. Sharp et al., 2018; Silbaugh et al., 2018).
Address correspondence to: Kathryn Peterson, Hyman et al. (2012) found that children with
17 Cypress Rd. Somerset, NJ 08873. Email: Kathryn.
peterson1@rutgers.edu autism did not meet recommended intake of
doi: 10.1002/jaba.711 certain nutrients (e.g., Vitamins D, C, E, and
© 2020 Society for the Experimental Analysis of Behavior
1
2 Jaime G. Crowley et al.

K), choline, fiber, phosphorus, potassium, and choice conditions, the researcher presented a
some children consumed excessive sodium. choice between the change-resistant and alter-
The selective diets of children with autism native responses, which were shorts and long
often consist primarily of processed junk foods pants for Kurt, respectively, and pictures of
that are low in nutritional content and high in Glenn not wearing and wearing a hat or gloves,
fat, salt, and sugar (e.g., French fries and respectively. Choosing either produced 20 s of
cookies; Bandini et al., 2010). Selective diets access to a moderately preferred item in the
with poor nutritional content can cause further free-choice condition. Choosing the change-
impairments in behavioral and cognitive func- resistant response produced no differential con-
tioning, and unhealthy weight gain, which sequence and choosing the alternative response
increases the risk for severe health problems produced 20 s of access to the participant’s
(e.g., obesity, Type-2 diabetes, and hyperten- most preferred item during the asymmetrical-
sion; Freedman et al., 1999; Ludwig et al., choice condition. During the singular-choice
1999). In fact, the prevalence of obesity and condition, the researcher (a) only presented the
diabetes among individuals with autism, and alternative response, (b) guided the participant
the prevalence of hypertension in males with to choose the alternative response if he did not
autism is higher than that of typically develop- do so independently, and (c) gave the partici-
ing peers, which is alarming given the signifi- pant his most preferred item for 20 s after the
cant negative health consequences of these participant chose or the researcher guided the
diseases (Flygare Wallén et al., 2018; Levy alternative response. The researcher removed
et al., 2019). Children with autism and the preferred item if Kurt attempted to remove
change-resistant feeding behavior may also the pants or if Glenn attempted to remove the
engage in inappropriate mealtime behavior hat or gloves.
(e.g., aggression, crying, pushing food away, Fisher et al. (2019) hypothesized that
self-injury, spitting food out, and vomiting), change-resistant behavior persists because chil-
which may discourage the caregiver from dren with autism have difficulty predicting and
attempting to change the child’s diet (Borrero controlling events in their environment
et al., 2016; Borrero et al., 2010; Piazza, Fisher (Gomot & Wicker, 2012). Results of fMRI
et al., 2003). studies suggest that children with autism
Despite the potential negative consequences respond differently to novel stimuli (Gomot
of change-resistant behavior, few studies have et al., 2011; Gomot & Wicker, 2012), which
evaluated interventions for resistance to change may cause unexpected changes to be aversive
specifically. In a notable exception, Fisher et al. and may increase the probability of avoidant
(2019) treated the change-resistant behavior of behavior (Fisher et al.). Repeatedly engaging in
four participants with autism using a change-resistant behavior increases the proba-
concurrent-operants arrangement and the con- bility of predictable consequences and allows
ceptual, but not mathematical, framework of the child to avoid the unpredictable conse-
the generalized matching law. Two participants, quences of an alternative behavior. Fisher et al.
Kurt and Glenn (Experiment 2), refused to proposed that predictable consequences and
wear seasonally appropriate clothing during the avoidance of unpredictable consequences func-
winter and engaged in problem behavior tion as automatic reinforcement for change-
(e.g., aggression) when prompted to do so. The resistant behavior.
alternative responses were wearing long pants Fisher et al. (2019) suggested that Kurt and
for Kurt and wearing a hat and gloves for Glenn chose the change-resistant response
Glenn. During the free- and asymmetrical- exclusively during the free-choice condition
Food Selectivity As Resistance To Change 3

because it produced higher quality automatic whether the intervention generalized to non-
reinforcement (i.e., predictable consequences, targeted alternative foods, and (e) using a rever-
avoidance of unpredictable consequences). sal design to assess whether consumption of
During the asymmetrical-choice condition, the alternative foods maintained in phases of the
value of automatic reinforcement for the asymmetrical-choice condition that followed
change-resistant response outweighed the value phases of the single-choice condition.
of the reinforcer for the alternative response for
Kurt, who continued to allocate responding
Method
exclusively to the change-resistant response.
Presumably, the value of the reinforcer for the Participants
alternative response competed with that of the The inclusion criteria for the current study
change-resistant response for Glenn; alternative were (a) a diagnosis of autism spectrum disor-
responses increased but did not maintain at der via best clinical estimate, (b) a diagnosis of
clinically acceptable levels over time. The pur- avoidant/restrictive food intake disorder based
pose of the singular-choice condition was to on the Diagnostic Statistical Manual (5th ed.)
expose participants to the alternative response criteria (American Psychiatric Association,
and its consequences. Alternative responses 2013), (c) age (between the ages of 2 and
increased to 100% during the singular-choice 8 years), (d) regular consumption of between
condition for both participants. Fisher et al. 3 and 20 foods, based on caregiver report,
hypothesized that repeated exposure to the con- (e) consumption of at least 90% of caloric
sequences of the alternative response during the needs by mouth and no dependence on supple-
singular-choice condition increased the predict- mental nutrition (e.g., gastrostomy tube), (f)
ability of the consequences and eliminated the weight for height greater than the 5th percentile
participants’ avoidant responses because the (i.e., no growth failure), and (g) safe oral feed-
consequences of the alternative response were ing, per physician report. Before enrollment in
no longer unpredictable. the study, a behavior analyst or licensed psy-
Like Fisher et al. (2019), we hypothesized chologist interviewed the caregiver about the
that the food selectivity in children with autism participant’s current and past feeding behavior
may be maintained by automatic reinforcement and medical history, measured and recorded
(i.e., the predictable consequences of consum- height and weight, conducted direct observa-
ing certain foods versus others). Thus, the cur- tions of the participant’s chewing skills with
rent study replicated and extended Fisher et al. table-texture foods, and excluded participants
by (a) giving participants with autism and food who did not demonstrate safe and appropriate
selectivity a choice between consuming change- chewing skills.
resistant and alternative foods in a concurrent- Study enrollment lasted approximately
operants arrangement during free- and 20 months between summer 2017 and spring
asymmetrical-choice conditions, (b) giving par- 2019, during which the program admitted
ticipants their most preferred item for consum- 52 children. Most children did not meet study
ing alternative foods during the asymmetrical- inclusion criteria because they (a) did not have
choice condition, (c) repeatedly exposing partici- a diagnosis of autism spectrum disorder,
pants to alternative foods and their consequences (b) were dependent on supplemental nutrition
during the single-choice condition if consump- or did not meet their caloric needs by mouth,
tion of alternative foods did not increase during (c) did not have the appropriate chewing skills
the asymmetrical-choice condition, (d) using a to consume table-texture foods, or (d) had a
multiple baseline across foods design to assess combination of these factors. Four children
4 Jaime G. Crowley et al.

met the study inclusion criteria and enrolled in Response Measurement


but did not complete the study for the follow- The primary dependent variable was con-
ing reasons: caregivers of two children with- sumption of alternative foods. Trials with con-
drew consent before study initiation, caregivers sumption were ones in which the participant
of one child did not respond to schedule study chose the alternative food and had a clean
appointments, and one child began swallowing mouth. Observers used a computer to score
bites of unmasticated food after the study initi- responses on a modified Excel spreadsheet.
ation (i.e., safety concern). Observers scored the occurrence of a choice up
Five males and two females met the inclusion to one time for each trial if the participant
criteria and completed the study. Six participants placed a bite inside his or her mouth within 8 s
received a diagnosis of autism from a university- of presentation. Observers scored whether the
based clinic and one received a diagnosis from a participant chose the change-resistant or alter-
school psychologist with confirmation from a native food during free- and asymmetrical-
psychiatrist. All of the participants ate a few choice conditions, and whether the participant
fruits, most of the participants ate a few proteins chose the alternative food during the single-choice
and starches (mostly processed foods such as condition. Observers scored mouth clean up to
macaroni and cheese), all but one of the partici- one time for each trial when no food larger than a
pants refused vegetables, all but one of the par- grain of rice was in the participant’s mouth 30 s
ticipants engaged in inappropriate mealtime after the bite entered the mouth, not including
behavior (e.g., head turns and pushing the spoon when no food was in the mouth because the par-
away when caregivers presented novel foods), ticipant expelled (spit out) the bite. From this
four of the participants engaged in other rigid point forward, consumption refers to trials in
mealtime routines, and all of the participants which the participant both chose the alternative
engaged in change-resistant behavior food and had a clean mouth. Observers also
(e.g., ritualistic play) outside of mealtime. scored consumption in the guided-choice condi-
tion if Genevieve deposited the bite inside her
mouth from the point at which the feeder
Setting, Materials, Feeders, and Observers guided her hand with the spoon. The first
Feeders conducted sessions at a university- author calculated percentage of alternative-food
based medical center in a session room in a consumption by examining the data for each
pediatric feeding disorders clinic or in an early session, counting the trials in which both a
intervention classroom. Session materials choice and a mouth clean occurred for the
included child-specific reinforcers, computers alternative food, dividing the number of trials
for data collection, food scales, a table and with choices and mouth clean of the alternative
chairs, colored tablecloths for the asymmetrical- food by the number of trials in the session, and
and single-choice conditions, timers, video converting the ratio to a percentage.
cameras, and video-recording computer soft-
ware. Participants used either age-appropriate
(e.g., booster chair) or weight-appropriate seat- Reliability and Intervention Integrity
ing that maintained their safety (e.g., Special Secondary observers collected data in a
Tomato Chair), and age-appropriate utensils room equipped with one-way observation
(e.g., toddler forks and spoons). Feeders and and two-way communication adjacent to the
observers were trained therapists with Bache- session room and one early intervention
lor’s or Master’s Degrees in applied behavior classroom. On some occasions, observers col-
analysis, psychology, or a related field. lected data while seated next to the
Food Selectivity As Resistance To Change 5

participant and feeder in an early interven- for Genevieve. Low interobserver agreement
tion classroom. A second observer indepen- occurred most often when the primary and sec-
dently scored 33% (range, 15% to 50%) of ondary observers were viewing the session from
sessions across participants, foods, condi- different locations (e.g., the feeder was the pri-
tions, and sessions to assess interobserver mary observer and was in the early intervention
agreement for choices and mouth clean. classroom and the second observer was in the
An agreement for choice occurred if both adjacent observation room), and the secondary
observers agreed the participant made a choice observer was not able to discriminate between
and agreed on which food the participant the change-resistant and alternative foods. The
chose, or both observers agreed the participant first author conducted periodic retraining with
did not make a choice. An agreement for the observers when reliability was below 80%
mouth clean occurred if both observers agreed for more than one session. During retraining,
a mouth clean occurred and agreed for which the first author reviewed operational defini-
food the participant had a mouth clean or both tions, modeled data collection, watched the
observers agreed a mouth clean did not occur. observer practice data collection, and ensured
The first author calculated interobserver agree- reliability for practice sessions was 80% or
ment for choice and mouth clean by dividing above before the observer could resume study
the number of trials with an agreement by the data collection.
total number of trials for each session, calculat- The Excel spreadsheet calculated percentage
ing the mean across sessions, and converting of alternative-food consumption for 100% of
the ratio to a percentage. Mean interobserver participants, foods, conditions, and sessions to
agreement for choice was 100% for Steven, assess interobserver agreement. A secondary
99% (range, 80% to 100%) for Titus, 86% author compared the first author’s and the
(range, 0% to 100%) for Alfonzo, 98% (range, Excel spreadsheet’s calculations for percentage
80% to 100%) for Zachary, 97% (range, 80% of alternative-food consumption for 33% of
to 100%) for Micah, 98% (range, 20% to sessions for each participant. An agreement
100%) for Kelsey, and 98% (range, 40% to occurred when the first author and the Excel
100%) for Genevieve. Mean interobserver spreadsheet calculated the same value for per-
agreement for mouth clean was 100% for Ste- centage of alternative-food consumption for a
ven, 98% (range, 60% to 100%) for Titus, session. A disagreement occurred when the first
91% (range, 0% to 100%) for Alfonzo, 97% author and the Excel spreadsheet calculated a
(range, 40% to 100%) for Zachary, 99% different percentage of alternative-food con-
(range, 80% to 100%) for Micah, 98% (range, sumption for a session. The author then
0% to 100%) for Kelsey, and 98% (range, 0% divided the total number of sessions with an
to 100%) for Genevieve. agreement by the total number of sessions used
Interobserver agreement was below 80% for for reliability and converted to a percentage.
choice in one session for Kelsey and three ses- Interobserver agreement was 100% for Zachary
sions for Genevieve. Interobserver agreement and Micah, 96% for Steven, 97% for Titus,
was below 80% for mouth clean in one session Alfonzo, and Kelsey, and 98% for Genevieve.
for Titus, one session for Zachary, one session Observers scored correct intervention integ-
for Kelsey, and four sessions for Genevieve. rity if the feeder implemented all major proto-
Interobserver agreement was below 80% for col steps correctly or incorrect intervention
consumption, meaning both choice and mouth integrity if the feeder made an error on any
clean in the same session for four sessions for step. The major steps of the protocol changed
Alfonzo, one session for Kelsey, and one session depending on the condition but generally
6 Jaime G. Crowley et al.

included assessment of: a presession-rule state- four for Alfonzo, and five for Zachary and Gen-
ment; bite presentation; prompts; reinforcer evieve. Each meal was between 40 to 45 min
placement; consequences for choices, mouth and consisted of multiple, five-trial sessions
clean, expulsion, packing, and inappropriate with a 1- to 2-min break between each session.
mealtime behavior; implementation of hand- The number of sessions per day depended on
over-hand guidance and nonremoval of the the participant’s schedule and behavior during
spoon; and re-presentation of expelled bites. each session. Mean number of sessions per day
Observers scored intervention integrity for was 13 for Steven, 12 for Titus, nine for
47% of sessions across participants. Mean Alfonzo, eight for Zachary, Micah, and Kelsey,
integrity was 99% (range, 80% to 100%) for and 10 for Genevieve.
Steven, 99% (range, 80% to 100%) for Titus, The feeder presented one food in each ses-
100% for Alfonzo and Zachary, 98% (range, sion of the preassessment, one change-resistant
75% to 100%) for Micah, 99% (range, 40% food and one alternative food in each session of
to 100%) for Kelsey, and 98% (range, 40% to the free- and asymmetrical-choice conditions,
100%) for Genevieve. The criterion for correct and one alternative food in each session of the
implementation of nonremoval of the spoon single-choice condition. The feeder presented
required the feeder to keep the utensil touching one bite per trial in the preassessment, one
the participant’s lips, independent of the rate of change-resistant bite and one alternative bite
inappropriate mealtime behavior. Lower levels per trial in the free- and asymmetrical-choice
of intervention integrity occurred primarily conditions, and one alternative bite per trial in
when the participant had high rates of inappro- the single-choice condition. The size of each
priate mealtime behavior (e.g., head turns). bite was 0.6 cm by 0.6 cm by 0.6 cm, except
Although the feeder kept the utensil close to as described in the preassessment of the
the lips, the utensil was not touching the par- change-resistant foods. Data from previous
ticipant’s lips. Intervention integrity was below studies suggest that food variety and consump-
80% in one session for Micah and Kelsey, and tion can be increased safely and effectively
two sessions for Genevieve. The first author using this bite size (Peterson et al., 2019; Peter-
conducted periodic observations with immedi- son et al., 2016). Before the study, caregivers
ate feedback throughout the study and pro- selected four alternative foods for each partici-
vided retraining when intervention integrity pant. Alternative foods were baked beans, broc-
was below 80% for more than one session. coli, sandwich (deli meat, lettuce, bread), and
During retraining, the first author reviewed and strawberry for Steven; chicken, green beans,
modeled the protocol, role-played with the feeder, peas, and potato for Titus; avocado, cauli-
observed the feeder conducting the protocol with flower, chicken, and potato for Alfonzo;
the participant, and immediately checked the chicken, corn, green beans, and lasagna for
intervention-integrity data to ensure it was 80% Zachary; baked beans, peas, pineapple, and
or above before the feeder conducted the protocol white rice for Micah; green beans, peas, potato,
without the first author present. and yam for Kelsey; and carrot, cheese, pan-
cake, and pear for Genevieve.
The feeder stated the contingencies before
General Procedure each session (see below in intervention evalua-
Participants came to the clinic for the study tion). The feeder started each trial by pre-
2 to 5 days per week for one to four meals per senting each bite on a utensil in a bowl within
day. The total number of meals per week was arm’s reach of the participant approximately
10 for Steven, Micah, and Kelsey, 20 for Titus, every 30 s. The feeder provided praise
Food Selectivity As Resistance To Change 7

(e.g., “Good job taking a bite”) for participant Preassessment of Change-Resistant


choices. The feeder removed the bowl(s) after Response. Caregivers identified change-
the participant placed the bite in his or her resistant foods for their child. Children with
mouth, set a timer for 30 s, and conducted a change-resistant feeding behavior often refuse
mouth check when 30 s elapsed by saying, preferred foods presented in a format that is
“Show me aah” and examining the contents of different from that which they typically eat the
the participant’s mouth. The feeder provided food (e.g., pizza cut into small pieces versus a
praise for mouth clean or said, “Swallow your pizza slice). Therefore, we conducted an assess-
bite” if food larger than a grain of rice was in ment to ensure that participants would con-
the participant’s mouth at the mouth check. sume the caregiver-identified change-resistant
The feeder started the next trial after the mouth foods in the study’s presentation format, which
check. The bite size was small enough that a was a 0.6-cm by 0.6-cm by 0.6-cm bite. The
participant could have five bites in his or her purpose of presenting the change-resistant food
mouth simultaneously without compromising in this format was to equate the size of the
his or her safety (i.e., the size of five 0.6-cm by change-resistant and alternative bites during the
0.6-cm by 0.6-cm bites is smaller than the size study.
of one bite a child the same age as study partic- The conditions of the assessment were a
ipants would typically consume). After the fifth free-operant presentation with a fixed-time pre-
trial, the feeder conducted mouth checks every sentation and bite-size fading to a 0.6-cm by
30 s if food larger than a grain of rice was in 0.6-cm by 0.6-cm bite. If the participant con-
the participant’s mouth until no food larger sumed at least 60% of the change-resistant food
than a grain of rice was in the participant’s in one condition, the feeder presented the food
mouth. No participant held food in his or her in the next condition. If the participant con-
moth longer than two 30-s intervals after the sumed 59% or less of the change-resistant food
fifth trial. The feeder did not provide any dif- in any condition, the feeder asked the care-
ferential consequences for coughing, gagging, givers to identify a different change-resistant
negative vocalizations, or vomiting. food. The feeder continued the procedure of
presenting change-resistant foods in the free-
Preassessment operant condition and progressing foods
Preference Assessment. Therapists con- through the conditions or replacing foods based
ducted a paired-stimulus preference assess- on participant consumption until the partici-
ment to identify the participants’ most pant consumed at least two change-resistant
preferred items (Fisher et al., 1992). Edible foods in the 0.6-cm by 0.6-cm by 0.6-cm bite-
items used in the preference assessment did presentation condition.
not constitute a large portion of the partici- During free-operant sessions, the feeder pres-
pants’ daily diets based on caregiver-recorded ented the change-resistant food in the same for-
3-day food logs. The most preferred item was mat the caregivers typically used (e.g., pancakes
an iPad for Steven, Kelsey, and Genevieve; cut into 1.5-cm by 1.5-cm by 1.5-cm triangles
Skittles for Titus, Alfonzo, and Micah; and with syrup and butter). The feeder said, “It’s
Hot Cheetos for Zachery. Participants did time for a snack. You can have as much or as
not have free access to their preferred items little of this snack as you want,” and set the
between sessions. The therapist asked care- timer for 5 to 10 min, based on the amount of
givers to restrict access to the items at home; food presented and on caregiver report of typi-
however, we did not assess compliance with cal meal length. During the fixed-time sessions,
this request. the feeder followed the general procedure, but
8 Jaime G. Crowley et al.

the bite size was 1.5 cm by 1.5 cm by 1.5 cm. in front of you and say, ‘Take one bite’. If you
During bite-size-fading sessions, the feeder take a bite of (alternative food), you get (pre-
followed the general procedure and gradually ferred item).” The contingencies were identical
reduced the bite size across sessions from to free choice with the following modifications:
1.5 cm by 1.5 cm by 1.5 cm to 0.6 cm by the feeder held or placed the preferred item in
0.6 cm by 0.6 cm. Each participant consumed the participant’s sight, approximately 10 cm
at least two change-resistant foods in the behind the bowl with the alternative food; the
0.6-cm by 0.6-cm by 0.6-cm presentation for- feeder provided 30 s of access to the iPad for
mat. These consisted of banana and quesadilla Steven, Kelsey, and Genevieve, one Skittle to
for Steven; chicken nugget, French fries, Titus, Alfonzo, and Micah, and one piece of
Nutella on bread, and apple for Titus; Fruit Hot Cheeto to Zachary if they consumed the
Loops, Captain Crunch, tropical fruit, and pan- bite of the alternative food. The feeder removed
cake for Alfonzo; hot dog and Cheerios for both bowls and the preferred item, and pres-
Zachary; Fruit Loops, Captain Crunch, and ented the next pair of bites if the participant
waffle for Micah; pancake, hot dog, pear, and did not make a choice. The feeder removed the
strawberry for Kelsey; and Fruit Loops, Captain preferred item and followed the general proce-
Crunch, and chicken nugget for Genevieve. dure if the participant chose the change-
resistant food.
Intervention Evaluation Single Choice. The stated contingencies
Free Choice. The feeder used random selec- were, “I’ll place one bite in front of you and
tion and counterbalancing to select an alterna- say, ‘Take one bite’. If you take a bite of (alter-
tive food to present in each session. The feeder native food), you get (preferred item).” The
rotated the pairings of change-resistant and procedure was identical to asymmetrical choice
alternative foods by randomly selecting a with the following modifications. The table
change-resistant food to present in each session. cloth, bowls, and utensils were a different color
The stated contingencies were, “I’ll place both from those in asymmetrical choice. The feeder
bites in front of you and say, ‘Take one bite’. If presented a bite of the alternative food and
you take a bite, I’ll say, ‘Good job.’” The feeder said, “Take one bite.” If the participant did not
presented a bite of the change-resistant food choose the alternative food, the feeder physi-
and a bite of the alternative food and said, cally guided the participant to touch the utensil
“Take one bite.” The feeder followed the gen- to his or her lips and hold the utensil touching
eral procedure and did not provide any differ- the lips until he or she opened his or her
ential consequences for inappropriate mealtime mouth such that the feeder could guide the
behavior or expulsion. The feeder removed the participant to put the bite in his or her mouth,
bowls if the participant did not make a choice. or 10 min from session initiation elapsed
The feeder blocked the participant’s attempt to (Peterson et al., 2016). If the participant did
place a second bite in his or her mouth if the not grip the utensil during hand-over-hand
participant had already placed one of the bites guidance, the feeder held the utensil and
in his or her mouth. followed the same procedure without hand-
Asymmetrical Choice. The feeder placed a over-hand guidance.
colored table cloth (e.g., orange) on the table. The feeder re-presented expelled bites by
The bowls and utensils were the same color as scooping up the bite or retrieving a fresh bite
the table cloth and the color remained the same with the utensil and touching the utensil to the
for this condition throughout the study. The participant’s lips until the feeder could deposit
stated contingencies were, “I’ll place both bites the bite into the participant’s mouth, or
Food Selectivity As Resistance To Change 9

10 min from session initiation elapsed. If the backward-chaining or preferred-item-access


participant was expelling the bite at the mouth- step, the feeder implemented the previous step.
check interval for Trials 1 to 4, the feeder re- The steps of the backward-chaining proce-
presented the bite until it remained in the dure were (a) place bite in mouth, (b) touch
mouth for 3 s and started the next trial. If the utensil to lips, (c) bring utensil 2.5 cm from
participant was expelling the bite at the mouth- lips, (d) bring utensil halfway between bowl
check interval for Trial 5, the feeder re- and lips, (e) bring utensil 2.5 cm from bowl,
presented the bite until the participant had no and (f) place hand on utensil. The feeder pres-
food in the mouth larger than a grain of rice, ented the bite and immediately guided Gen-
or 10 min from session initiation elapsed. evieve to perform the prescribed step in the
The feeder used his or her forearm and an backward-chaining procedure. The feeder dis-
open hand to gently block the participant from continued guidance after Genevieve’s hand was
contacting the utensil. A second therapist in the prescribed position, provided 30 s of
assisted if the feeder could not maintain the iPad access if the bite entered Genevieve’s
participant’s safety during the session (e.g., the mouth within 8 s of presentation, and provided
participant was at risk of falling out of the chair), another 30-s access if Genevieve had a mouth
which occurred for four participants during the clean. The feeder guided Genevieve to place
single-choice condition with their first alternative the bite in her mouth if she did not do so
food. The second therapist stood directly behind within 8 s of presentation and did not provide
the participant and placed his or her outstretched iPad access.
arms with open hands on either side of the partic- Next, the feeder manipulated iPad access by
ipant so that they were hovering above the partici- providing (a) 30 s of iPad access for choices
pant’s arms when the feeder initiated hand-over- and another 30-s access for mouth clean and
hand guidance and discontinued blocking when (b) 30 s of iPad access for choices but not for
the trial ended. The purpose of the second thera- mouth clean. The feeder then implemented the
pist was to minimize the participant’s contact with single-choice contingencies.
the utensil and to prevent the participant from
upending the chair or otherwise harming him or Experimental Design
herself. We removed the blocker after three con- We used a combination of a concurrent mul-
secutive sessions of five or fewer instances of inap- tiple baseline across foods and reversal designs.
propriate mealtime behavior. The feeder implemented the asymmetrical-
Guided Choice. We used guided choice choice contingencies with one alternative food
with Genevieve because the percentage of con- when levels of consumption were low and sta-
sumption was low during the single-choice con- ble across alternative foods during the free-
dition. The procedure was identical to single choice baseline, and implemented the single-
choice with the addition of a backward- choice contingencies if levels of alternative-food
chaining procedure and a manipulation of consumption did not increase during asymmet-
preferred-item access. The feeder implemented rical choice. The feeder reversed to the previous
one step in a session (e.g., guided the bite to contingencies if percentage of alternative-food
Genevieve’s lips on every trial of the session). consumption increased to 80% or above for
When consumption was 80% or greater for three consecutive sessions. When percentage of
four of five sessions for one backward-chaining alternative-food consumption maintained at
or preferred-item access step, the feeder 80% or above in the least effortful intervention
implemented the next step. When consump- condition, the feeder repeated the sequence of
tion was below 80% for five sessions for one conditions necessary to increase consumption
10 Jaime G. Crowley et al.

to 80% or above with the next alternative food. zero (chicken, potato, and peas) or near zero
The feeder repeated this process until percent- (green beans) levels during the reversal to free
age consumption was 80% or above for four choice. Mean percentage consumption
alternative foods. The feeder conducted rever- increased to high levels for all foods in the
sals with subsequent alternative foods as time return to asymmetrical choice. Like Steven, the
permitted. The feeder implemented guided feeder did not implement single choice because
choice with Genevieve when carrot consump- of the high levels of consumption across foods
tion did not increase during single choice but during asymmetrical choice.
otherwise followed the sequence described Figure 3 displays the percentage of trials with
above. The feeder implemented free, asymmet- consumption for Alfonzo for potato, cauli-
rical, guided, and single choice with cheese and flower, avocado, and chicken. During the initial
free, asymmetrical, and single choice with pan- free-choice condition, mean percentage con-
cake and pear. sumption was zero (potato, cauliflower, avo-
cado) or near zero (chicken) for all foods. Mean
percentage of consumption increased to moder-
Results ate (potato) to high levels (cauliflower, avocado,
Figure 1 displays the percentage of trials with and chicken) in the initial exposure to asym-
consumption for Steven for baked beans, straw- metrical choice. Alfonzo’s pattern of responding
berry, broccoli, and sandwich. Steven’s pattern was similar for cauliflower, avocado, and
of responding was similar across foods. During chicken in that consumption increased during
the initial free-choice condition, mean percent- the initial exposure to asymmetrical choice.
age consumption was 0% across foods. Mean Mean percentage of potato consumption then
percentage consumption increased to high, sta- increased to high, stable levels in the single-
ble levels for all foods during the initial expo- choice condition and maintained at 100% in
sure to asymmetrical choice (mean, 93%; the reversal to asymmetrical choice. Mean per-
range, 20% to 100%). Consumption decreased centage consumption decreased for all foods in
to 0% across foods during the reversal to free the reversal to free choice (mean, 17%; range,
choice. Mean percentage consumption 0% to 100%), but returned to high, stable
increased to high, stable levels for all foods in levels in the final asymmetrical choice condi-
the return to asymmetrical choice. The feeder tion (mean, 94%; range, 40% to 100%).
did not implement single choice because levels Figure 4 displays the percentage of trials with
of consumption increased across foods during consumption for Zachary for chicken, lasagna,
asymmetrical choice. green beans, and corn. During the initial free-
Figure 2 displays the percentage of trials with choice condition, mean percentage of con-
consumption for Titus for green beans, sumption was zero (chicken, lasagna, and green
chicken, potato, and pea. Titus’ pattern of bean) or near zero (corn) for all foods. Zachary
responding was similar across foods and like never consumed chicken or lasagna during the
that of Steven. Mean percentage consumption initial exposure to asymmetrical choice, but
was either low or mainly low but variable in percentage of consumption increased (mean,
the initial free choice condition for all foods 71%; range, 0% to 100%) during the subse-
(mean, 8%; range, 0% to 100%). Mean per- quent phase of single choice for both foods.
centage consumption increased to high levels Mean percentage of consumption also
for all foods during the initial exposure to maintained at high levels during the reversal to
asymmetrical choice (mean, 84%; range, 0% to asymmetrical choice (mean, 97%; range, 60%
100%). Consumption decreased to to 100%), decreased during the reversal to free
Food Selectivity As Resistance To Change 11

Figure 1
Percentage of Trials with Consumption for Steven for Baked Beans (First Panel), Strawberry (Second Panel), Broccoli (Third
Panel), and Sandwich (Fourth Panel) in Free and Asymmetrical Choice.

choice (mean, 14%; range, 0% to 80%), and in asymmetrical choice for white rice and pine-
increased to 100% during the final return to apple (mean, 84%; range, 40% to 100%).
asymmetrical choice for both foods. Percentage Mean percentage of consumption increased
of consumption increased during the initial during single choice (mean, 79%; range, 0% to
exposure of green beans and corn (mean, 97%; 100%) and maintained at high levels during
range, 60% to 100%) in the asymmetrical the first reversal to asymmetrical choice (mean,
choice phase. 84%; range, 40% to 100%) for baked beans
Figure 5 displays the percentage of trials with and peas. Percentage of consumption decreased
consumption for Micah for baked beans, pine- during the reversal to free choice across all
apple, white rice, and peas. During the initial foods. Mean percentage of consumption
free-choice condition, mean percentage con- increased in the second return to asymmetrical
sumption was zero (baked beans and pineapple) choice for all foods except baked beans. Baked
or low (white rice and peas) for all foods. beans were exposed to single choice again
Micah never consumed baked beans or peas before maintaining at high, stable levels in the
during the initial exposure to asymmetrical final asymmetrical-choice condition (mean,
choice, but percentage consumption was high 94%; range, 80% to 100%).
12 Jaime G. Crowley et al.

Figure 2
Percentage of Trials with Consumption for Titus for Green Beans (First Panel), Grilled Chicken (Second Panel), Potato
(Third Panel), and Peas (Fourth Panel) in Free and Asymmetrical Choice.

Figure 6 displays the percentage of trials with consumption maintained in the reversal to
consumption for Kelsey for potato, peas, green asymmetrical choice for potato and green bean,
beans, and yam. During the initial free-choice but not for peas. Percentage of peas consump-
condition, mean percentage consumption was tion increased during the return to single
zero (potato, peas, and green beans) or near choice (M = 82%; range, 0% to 100%), and
zero (yam) for all foods. Percentage of con- maintained at high, stable levels during a final
sumption increased to moderate levels during reversal to asymmetrical choice (M = 97%;
the initial exposure to asymmetrical choice for range, 80% to 100%).
potato and peas (mean, 59%; range, 0% to Figure 7 displays the percentage of trials with
100%); remained at zero for green bean; and consumption for Genevieve for carrot, cheese,
increased to high, stable levels for yam (mean, pancake, and pear. During the initial free-
94%; range, 80% to 100%). Mean percentage choice condition, mean percentage consump-
consumption increased to high levels during tion was zero (carrot and cheese) or near
the subsequent phase of single choice for zero (pancake and pear) for all foods. Percent-
potato, peas, and green bean (mean, 86%; age of consumption remained at zero during
range, 40% to 100%). These high levels of the asymmetrical-choice condition for both
Food Selectivity As Resistance To Change 13

Figure 3
Percentage of Trials with Consumption for Alfonzo for Potato (First Panel), Cauliflower (Second Panel), Avocado (Third
Panel), and Chicken (Fourth Panel) in Free, Asymmetrical, and Single Choice.

carrot and cheese but increased to high levels choice (cheese, pancake, pear), all decreased
for pancake and pear (mean, 82%; range, 20% to low levels during this reversal (mean,
to 100%). Percentage of carrot consumption 22%; range, 0% to 60%). During the final
remained at zero during single choice, but reversal to asymmetrical choice, all foods
increased during guided choice and maintained increased to high levels, although percentage
at high, stable levels in a reversal to single of pancake consumption was variable (mean,
choice. Similar to carrot, percentage of 93%; range, 40% to 100%).
cheese consumption increased in guided
choice and maintained in single choice. Per-
centage of cheese consumption also Discussion
maintained in the reversal to asymmetrical The procedures in the current study are
choice, but percentage of carrot consumption novel relative to preparations researchers typi-
did not. Carrot required additional exposure cally use in intervention-outcome studies on
to single choice following the reversal to pediatric feeding disorders. In most studies, a
asymmetrical choice to increase consumption feeder presents individual (e.g., LaRue et al.,
again. Of the foods with a reversal to free 2011; Patel et al., 2002) or multiple bites of
14 Jaime G. Crowley et al.

Figure 4
Percentage of Trials with Consumption for Zachary for Chicken (First Panel), Lasagna (Second Panel), Green Beans (Third
Panel), and Corn (Fourth Panel) in Free, Asymmetrical, and Single Choice.

target foods simultaneously (e.g., Penrod et al., when the change-resistant food was available
2010) to evaluate the effects of the experimen- concurrently. We conceptualized consumption
tal manipulation. However, in typical mealtime as the participant choosing the alternative food
situations, children have choices among foods, by picking up the utensil with the bite of alter-
such as in the school cafeteria or when native food, placing the bite in his or her
searching the pantry for a snack. In these situa- mouth within 8 s of presentation, and chewing
tions, children with food selectivity are likely to and swallowing the bite of alternative food
consume their change-resistant food rather than (mouth clean). The participant had the option
an alternative food. The preparation in the cur- of consuming the change-resistant food in free-
rent study more closely approximated a typical and asymmetrical-choice conditions or making
mealtime situation in which a child has a no response in all conditions instead of con-
choice among foods. In fact, we pitted the suming the alternative food.
alternative food against the participant’s In the current study, we used the prepara-
change-resistant food in a paired-choice tion and conceptual framework of Fisher et al.
arrangement. The goal was to teach the partici- (2019), who based their intervention for
pant to consume the alternative food even change-resistant behavior in children with
Food Selectivity As Resistance To Change 15

Figure 5
Percentage of Trials with Consumption for Micah for Baked Beans (First Panel), Pineapple (Second Panel), White Rice
(Third Panel), and Peas (Fourth Panel) in Free, Asymmetrical, and Single Choice.

autism conceptually, but not mathematically, Fisher et al. (2019) proposed that children with
on the generalized matching law. Basic research autism prefer predictable consequences and avoid
on the matching law shows that individuals unpredictable consequences. These preferred conse-
allocate their responses to options that produce quences function as automatic reinforcement for
relatively more frequent or higher value of rein- change-resistant behavior. Participants in the cur-
forcement. These reinforcement parameters rent study consistently consumed the change-
predict the value of a reinforcing consequence. resistant food when given a choice between a
The matching law states that responses that change-resistant and an alternative food in the free-
produce proportionally more frequent or more choice condition, with a few exceptions. Consump-
valued reinforcement will occur more fre- tion of the change-resistant foods occurred almost
quently than other responses that are available exclusively in free choice because it had a long his-
concurrently (McDowell & Caron, 2010; tory of producing multiple sources of automatic
Petry & Heyman, 1994; Pierce & Epling, reinforcement, and no programmed reinforcement
1983). The systematic relation between relative was available for consuming alternative foods.
rates of responding and relative reinforcer value Consumption of the change-resistant food
is one of the most robust findings in behavior might have persisted until the participant
analysis. repeatedly received a higher quality reinforcer
16 Jaime G. Crowley et al.

Figure 6
Percentage of Trials with Consumption for Kelsey for Potato (First Panel), Peas (Second Panel), Green Beans (Third Panel),
and Yam (Fourth Panel) in Free, Asymmetrical, and Single Choice.

for consuming the alternative food. Providing consumption may be sufficient to change
reinforcement in the asymmetrical-choice con- responding for some participants from almost
dition for consuming alternative foods was suf- exclusive consumption of change-resistant foods
ficient to change responding from the change- to exclusive or increased consumption of alter-
resistant to the alternative foods for Steven and native foods. Alternatively, presession rules may
Titus. Levels of consumption of the first alter- have contributed to or controlled responding
native food presented in asymmetrical choice for Steven and Titus. They were among the
was inconsistent but greater than in free choice oldest participants (i.e., ages 7 and 8), had the
for Alfonzo and Kelsey. These results are novel most sophisticated verbal repertoires, and dis-
because consumption of alternative foods played more adaptive skills (e.g., toileting, self-
increased consistently during asymmetrical dressing) than the other participants. Future
choice for two participants and inconsistently researchers could evaluate whether presession
for two participants, even though the partici- rules contribute to or control responding in
pants’ change-resistant foods were available similar concurrent-operants arrangements.
concurrently. Thus, providing an empirically Percentage consumption of foods in the first
identified reinforcer for alternative-food panel of the multiple baseline, chicken, baked
Food Selectivity As Resistance To Change 17

Figure 7
Percentage of Trials with Consumption for Genevieve for Carrot (First Panel), Cheese (Second Panel), Pancake (Third Panel),
and Pear (Fourth Panel) in Free, Asymmetrical, Single, and Guided Choice.

beans, and carrot, was 0% during the first consumption. The high levels of high quality
exposure to asymmetrical choice for Zachary, reinforcement may have increased the value of
Micah, and Genevieve, respectively. Persistent consequences for alternative-food consumption.
change-resistant responding may have Repeated exposure to the alternative food dur-
prevented participants from contacting rein- ing single choice may also be important because
forcement for alternative-food consumption. exposure is one method by which humans
Thus, reinforcement for alternative-food con- develop preferences for foods (Birch, 1989,
sumption remained low and the consequence 1999; Birch & Marlin, 1982; Birch et al.,
for alternative-food consumption remained 1987; Capaldi, 1996). Genevieve showed a
unpredictable. Levels of chicken and baked similar response pattern but required backward
beans consumption increased for Zachary and chaining and preferred-item manipulation plus
Micah, respectively, during single choice. Pre- single choice to increase carrot consumption.
senting the alternative food in the absence of Although consumption of potatoes increased
the change-resistant food during single choice for Alfonzo, and consumption of potatoes and
exposed the participants to the predictable, peas increased for Kelsey during asymmetrical
favorable consequences of alternative-food choice, levels of alternative-food consumption
18 Jaime G. Crowley et al.

were not clinically valid until the feeder pres- Levels of consumption of the food in the
ented these foods in a single-choice condition. first panel of the multiple baseline did not
Results for Alfonzo, Zachary, Micah, and increase until the feeder implemented single
Kelsey are consistent with previous literature choice for Alfonzo, Zachary, Micah, and Kelsey
demonstrating that nonremoval of the spoon, a or guided choice for Genevieve. However,
component of the single-choice contingencies levels of consumption increased during asym-
in the current study, may be necessary to metrical choice without exposure to single
increase consumption for some participants choice for some alternative foods in subsequent
(Patel et al., 2002; Piazza, Patel et al., 2003; panels of the multiple baseline. Consumption
Reed et al., 2004). Results for Genevieve are increased for cauliflower, avocado, and chicken
similar to those obtained with Bryce in Peter- in panels 2, 3, and 4, respectively, for Alfonzo;
son et al. (2016). Bryce’s acceptance of alterna- for green beans and corn in panels 3 and
tive foods increased only after the researchers 4, respectively, for Zachary; for pineapple and
added backward chaining to nonremoval of the white rice in panels 2 and 3, respectively, for
spoon, suggesting that some participants Micah; and for yam and pear in panel 4 for
require intervention components in addition to Kelsey and Genevieve, respectively. Thus, we
nonremoval of the spoon to change their feed- presented one food for Alfonzo, two foods for
ing behavior. Intervention dosage also could Zachary and Micah, and three foods for Kelsey
have affected Genevieve’s results as her appoint- in single choice before they began consuming
ments were more spaced apart across the week. other food(s) during asymmetrical choice. We
Similarly, Peterson et al. hypothesized that presented carrot and cheese in panels 1 and
intervention dosage contributed to the results 2 in guided choice and pancake in panel 3 in
for individual participants in their study and single choice with Genevieve, and her con-
likely affects outcomes for children with autism sumption of pear in panel 4 increased during
receiving early intensive behavioral intervention asymmetrical choice.
(Lovaas, 1987; McEachin et al., 1993). Note that Micah’s pattern of responding was
The current study adds to the literature because different from that of Alfonzo and Zachary.
the results demonstrated that single choice had After alternative-food consumption increased
beneficial effects even after we reversed those con- during asymmetrical choice with one food in
tingencies. Participants continued to consume the multiple baseline (e.g., panel 2 food) for
alternative foods during asymmetrical-choice con- Alfonzo and Zachary, consumption of alterna-
ditions that followed single-choice conditions even tive foods in subsequent panels (e.g., panels
though discriminative stimuli were paired with 3 and 4) of the multiple baseline also increased
asymmetrical- and single-choice conditions. The during asymmetrical choice. By contrast, the
value of reinforcement for alternative-food con- feeder implemented single choice with peas in
sumption may have increased during single choice panel 4 for Micah, even though he was con-
because alternative-food consumption produced suming foods during asymmetrical choice in
predictable and favorable consequences. The value panels 2 and 3. Micah’s history of vegetable
maintained at the increased level and was higher refusal may not account for the findings
than that of change-resistant-food consumption because he consumed peas in the free-choice
during the return to asymmetrical choice. We rep- baseline. An alternative explanation is that peas
licated this finding with every participant for look like green Skittles, which were a preferred
whom we presented at least one food in the item. He may have consumed peas initially in
single-choice contingencies and for every food the free-choice condition because of their visual
presented in the single-choice contingencies. similarity to Skittles, but pea consumption
Food Selectivity As Resistance To Change 19

extinguished after he discriminated the differ- quantities than foods they do not prefer
ence. Thus, exposure to the reinforcer for pea (Capaldi, 1996).
consumption during single choice was necessary Although the purpose of the current study
to reestablish responding. Titus’ pattern of was to evaluate the effects of a matching-law
baseline consumption of peas, which also based intervention, the ultimate clinical goal is
decreased over time during free choice, and his for participants to consume a diet that is age,
preference for Skittles provides support for this nutritionally, and calorically appropriate (U.S.
explanation. Researchers should investigate Department of Health and Human Services
whether the visual properties of foods affect and U.S. Department of Agriculture, 2015).
consumption (Wadhera & Capaldi Phillips, Thus, a next step would be for researchers to
2014) for children with autism. assess whether intervention effects generalize to
One limitation of the study is that we did other feeders, environments, and maintain over
not test participants’ ability to discriminate the time. Following the conclusion of the current
colored stimuli or biases for the colors associ- study, we taught caregivers to implement
ated with asymmetrical- and single-choice con- single-choice contingencies using the procedure
ditions. We also do not know whether the described by Ibañez et al. (2019) because previ-
preassessment affected participants’ consump- ous literature has demonstrated renewal of
tion of or preference for the change-resistant inappropriate mealtime behavior during context
foods or their resistance to change. The gradual changes (Ibañez et al., 2019; Kelley et al.,
changes made to the change-resistant foods 2018; Kelley et al., 2015).
may have decreased participants’ preferences for Participants received clinical services after the
the change-resistant foods or their resistance to study to progress to age-typical feeding. Steven,
change. Future studies should evaluate partici- Alfonzo, and Kelsey graduated from the clinical
pants’ ability to discriminate between and pref- program consuming age, nutritionally, and
erence for colored stimuli. Future researchers calorically appropriate meals. Caregivers
should also conduct baseline and preference increased the volume of food Titus consumed
assessments with change-resistant and alterna- during outpatient therapy but dropped out of
tive foods before and after the preassessment to therapy before Titus was eating age-appropriate
determine whether the preassessment affects volumes of food. Zachary continued to receive
food preferences, resistance to change, or both. early intervention services, and Micah and
Another limitation is that we did not con- Genevieve continued to receive outpatient ser-
duct a preference assessment with the change- vices; the variety and amount of foods con-
resistant and alternative foods. Consumption of sumed have increased for all three. The goal is
alternative foods was zero or near zero percent to continue until participants are eating age
during free choice, suggesting that participants appropriately.
preferred the change-resistant foods. The cur- The current study replicated and extended
rent study did not evaluate whether partici- Fisher et al. (2019) by evaluating the effects of
pants’ preferences correlated with their an intervention for the change-resistant feeding
decreased consumption of change-resistant behavior of participants with autism conceptu-
foods and increased consumption of alterna- ally based on the generalized matching law.
tive foods. Food preference could contribute This intervention (a) increased every partici-
to the value of the reinforcing consequences of pant’s alternative responses and reduced
change-resistant and alternative-food con- change-resistant responses, (b) maintained
sumption. Said simply, humans consume effects with the simpler, less effortful
foods they prefer more often and in greater asymmetrical-choice condition for participants
20 Jaime G. Crowley et al.

who required the single-choice contingencies, behavior. Behavior Analysis in Practice, 9, 364–379.
and (c) demonstrated generalization to foods https://doi.org/10.1007/s40617-016-0149-5
Borrero, C. S. W., Woods, J. N., Borrero, J. C.,
experimenters did not target in the single- Masler, E. A., & Lesser, A. D. (2010). Descriptive
choice condition. Results suggest that behavior analyses of pediatric food refusal and acceptance.
analysts can use this intervention to change the Journal of Applied Behavior Analysis, 43, 71–81.
https://doi.org/10.1901/jaba.2010.43-71
environment in systematic ways to decrease the Capaldi, E. D. (1996). Conditioned food preferences. In
occurrence of change-resistant behavior E. D. Capaldi (Ed.), Why we eat what we eat: The
(Makrygianni et al., 2018; Reichow, 2012; psychology of eating (pp. 53–80). American Psycholog-
ical Association.
Smith & Iadarola, 2015) and to increase the
Cermak, S. A., Curtin, C., & Bandini, L. G. (2010).
occurrence of behavior that is variable, flexible, Food selectivity and sensory sensitivity in children
and adaptable (Jennett & Hagopian, 2008). with autism spectrum disorders. Journal of the Ameri-
can Dietetic Association, 110, 238–246. https://doi.
org/10.1016/j.jada.2009.10.032
Fisher, W. W., Felber, J. M., Phillips, L. A., Craig, A. R.,
REFERENCES Paden, A. R., & Niemeier, J. J. (2019). Treatment of
American Psychiatric Association. (2013). Diagnostic and resistance to change in children with autism. Journal
statistical manual of mental disorders (DSM-5). APA. of Applied Behavior Analysis, 52, 974–993. https://
Baio, J., Wiggins, L., Christensen, D. L., Maenner, M. J., doi.org/10.1002/jaba.588
Daniels, J., Warren, Z., Kurzius-Spencer, M., Fisher, W. W., Piazza, C. C., Bowman, L. G.,
Zahorodny, W., Rosenberg, C. R., White, T., Hagopian, L. P., Owens, J. C., & Slevin, I. (1992).
Durkin, M. S., Imm, P., Nikolaou, L., Yeargin- A comparison of two approaches for identifying rein-
Allsopp, M., Lee, L-C., Harrington, R., Lopez,M., forcers for persons with severe and profound disabil-
Fitzgerald, R. T., Hewitt, A. … Dowling, N. F. ities. Journal of Applied Behavior Analysis, 25,
(2018). Prevalence of autism spectrum disorder 491–498. https://doi.org/10.1901/jaba.1992.25-491
among children aged 8 years - Autism and Develop- Flygare Wallén, E., Ljunggren, G., Carlsson, A. C.,
mental Disabilities Monitoring Network, 11 sites, Pettersson, D., & Wändell, P. (2018). High preva-
United States, 2014. Morbidity And Mortality Weekly lence of diabetes mellitus, hypertension and obesity
Report-Surveillance Summaries, 67, 1-23. https://doi. among persons with a recorded diagnosis of intellec-
org/10.15585/mmwr.ss6706a1 tual disability or autism spectrum disorder. Journal of
Intellectual Disability Research, 62, 269–280. https://
Bandini, L. G., Anderson, S. E., Curtin, C., Cermak, S.,
doi.org/10.1111/jir.12462
Evans, W., Scampini, R., … Must, A. (2010). Food
Freedman, D. S., Dietz, W. H., Srinivasan, S. R., &
selectivity in children with autism spectrum disorders
Berenson, G. S. (1999). The relation of overweight
and typically developing children. Journal of Pediat-
to cardiovascular risk factors among children and ado-
rics, 157, 259–264. https://doi.org/10.1016//j.jpeds.
lescents: the Bogalusa Heart Study. Pediatrics, 103,
2010.02.013
1175–1182. https://doi.org/10.1542/peds.103.6.
Birch, L. L. (1989). Effects of experience on the modifica- 1175
tion of food acceptance patterns. Annals of the Gomot, M., Blanc, R., Clery, H., Roux, S.,
New York Academy of Sciences, 561, 209–216. Barthelemy, C., & Bruneau, N. (2011). Candidate
https://doi.org/10.1111/j.1749-6632.1989.tb20983.x electrophysiological endophenotypes of hyper-
Birch, L. L. (1999). Development of food preferences. reactivity to change in autism. Journal of Autism and
Annual Review of Nutrition, 19, 41–62. https://doi. Developmental Disorders, 41, 705–714. https://doi.
org/10.1146/annurev.nutr.19.1.41 org/10.1007/s10803-010-1091-y
Birch, L. L., & Marlin, D. W. (1982). I don’t like it; I Gomot, M., & Wicker, B. (2012). A challenging,
never tried it: Effects of exposure on two-year-old unpredictable world for people with autism spectrum
children’s food preferences. Appetite, 3, 353-360. disorder. International Journal of Psychophysiology, 83,
https://doi.org/10.1016/s0195-6663(82)80053-6 240–247. https://doi.org/10.1016/j.ijpsycho.2011.
Birch, L. L., McPhee, L., Shoba, B. C., Pirok, E., & 09.017
Steinberg, L. (1987). What kind of exposure reduces Heifert, T. A., Susi, A., Hisle-Gorman, E., Erdie-
children’s food neophobia? Looking vs. tasting. Appe- Lalena, C., Gorman, G., Min, S. B., &
tite, 9, 171-178. https://doi.org/10.1016/S0195- Nylund, C. M. (2016). Feeding disorders in children
6663(87)80011-9 with autism spectrum disorders are associated with
Borrero, C. S. W., England, J. D., Sarcia, B., & eosinophilic esophagitis. Journal of Pediatric Gastroen-
Woods, J. N. (2016). A comparison of descriptive terology and Nutrition, 63, 69–79. https://doi.org/10.
and functional analyses of inappropriate mealtime 1097/mpg.0000000000001282
Food Selectivity As Resistance To Change 21

Hyman, S. L., Stewart, P. A., Schmidt, B., Cain, U., Ludwig, D. S., Majzoub, J. A., Al-Zahrani, A.,
Lemcke, N., Foley, J. T., … Ng, P. K. (2012). Dallal, G. E., Blanco, I., & Roberts, S. B. (1999).
Nutrient intake from food in children with autism. High glycemic index foods, overeating, and obesity.
Pediatrics, 130, 145–153. https://doi.org/10.1542/ Pediatrics, 103, 26–32. https://doi.org/10.1542/peds.
peds.2012-0900L 103.3.e26
Ibanez, V. F., Piazza, C. C., & Peterson, K. M. (2019). A Makrygianni, M. K., Gena, A., Katoudi, S., & Galanis, P.
translational evaluation of renewal of inappropriate (2018). The effectiveness of applied behavior analytic
mealtime behavior. Journal of Applied Behavior Analy- interventions for children with autism spectrum dis-
sis, 52, 1005–1020. https://doi.org/10.1002/jaba.647 order: A meta-analytic study. Research in Autism Spec-
Jennett, H. K., & Hagopian, L. P. (2008). Identifying trum Disorders, 51, 18–31. https://doi.org/10.1016/j.
empirically supported treatments for phobic avoid- rasd.2018.03.006
ance in individuals with intellectual disabilities. McDowell, J. J., & Caron, M. L. (2010). Matching in an
Behavior Therapy, 39, 151–161. https://doi.org/10. undisturbed natural human environment. Journal of
1016/j.beth.2007.06.003 the Experimental Analysis of Behavior, 93, 415–433.
Joseph, L., Thurm, A., Farmer, C., & Shumway, S. https://doi.org/10.1901/jeab.2010.93-415
(2013). Repetitive behavior and restricted interests in McEachin, J. J., Smith, T., & Lovaas, O. I. (1993).
young children with autism: Comparisons with con- Long-term outcome for children with autism who
trols and stability over 2 years. Autism Research, 6, received early intensive behavioral treatment. Ameri-
584–595. https://doi.org/10.1002/aur.1316 can Journal of Mental Retardation, 97, 359–372.
Kamal Nor, N., Ghozali, A. H., & Ismail, J. (2019). Patel, M. R., Piazza, C. C., Martinez, C. J.,
Prevalence of overweight and obesity among children Volkert, V. M., & Santana, C. M. (2002). An evalua-
and adolescents with autism spectrum disorder and tion of two differential reinforcement procedures with
associated risk factors. Frontiers In Pediatrics, 7, escape extinction to treat food refusal. Journal of
38–38. https://doi.org/10.3389/fped.2019.00038 Applied Behavior Analysis, 35, 363–374. https://doi.
Kelley, M. E., Jimenez-Gomez, C., Podlesnik, C. A., & org/10.1901/jaba.2002.35-363
Morgan, A. (2018). Evaluation of renewal mitigation
Penrod, B., Wallace, M. D., Reagon, K., Betz, A., &
of negatively reinforced socially significant operant
Higbee, T. S. (2010). A component analysis of
behavior. Learning and Motivation, 63, 133–141.
parent-conducted multi-component treatment for
https://doi.org/10.1016/j.lmot.2018.05.003
food selectivity. Behavioral Interventions, 25,
Kelley, M. E., Liddon, C. J., Ribeiro, A., Greif, A. E., &
207–228. https://doi.org/10.1002/bin.307
Podlesnik, C. A. (2015). Basic and translational eval-
uation of renewal of operant responding. Journal of Peterson, K. M., Piazza, C. C., Ibanez, V. F., &
Applied Behavior Analysis, 48, 390–401. https://doi. Fisher, W. W. (2019). Randomized controlled trial
org/10.1002/jaba.209 of applied behavior analysis versus wait-list control
LaRue, R. H., Stewart, V., Piazza, C. C., Volkert, V. M., for food selectivity in children with autism spectrum
Patel, M. R., & Zeleny, J. (2011). Escape as rein- disorder. Journal of Applied Behavior Analysis, 52,
forcement and escape extinction in the treatment of 895–917. https://doi.org/10.1002/jaba.650
feeding problems. Journal of Applied Behavior Analy- Peterson, K. M., Piazza, C. C., & Volkert, V. M. (2016).
sis, 44, 719–735. https://doi.org/10.1901/jaba.2011. A comparison of a modified sequential oral sensory
44-719 approach to an applied behavior-analytic approach in
Levin, L., & Carr, E. G. (2001). Food selectivity and the treatment of food selectivity in children with
problem behavior in children with developmental dis- autism spectrum disorders. Journal of Applied Behav-
abilities: Analysis and intervention. Behavior Modifi- ior Analysis, 49, 1–27. https://doi.org/10.1002/
cation, 25, 443–470. https://doi.org/10.1177/ jaba.332
0145445501253004 Petry, N. M., & Heyman, G. M. (1994). Effects of quali-
Levy, S. E., Pinto-Martin, J. A., Bradley, C. B., tatively different reinforcers on the parameters of the
Chittams, J., Johnson, S. L., Pandey, J., … response-strength equation. Journal of the Experimen-
Kral, T. V. E. (2019). Relationship of weight out- tal Analysis of Behavior, 61, 97–106. https://doi.org/
comes, co-occurring conditions, and severity of 10.1901/jeab.1994.61-97
autism spectrum disorder in the study to explore early Piazza, C. C., Fisher, W. W., Brown, K. A., Shore, B. A.,
development. The Journal of Pediatrics, 205, Patel, M. R., Katz, R. M., … Blakely-Smith, A.
202–209. https://doi.org/10.1016/j.jpeds.2018. (2003). Functional analysis of inappropriate mealtime
09.003 behaviors. Journal of Applied Behavior Analysis, 36,
Lovaas, O. I. (1987). Behavioral treatment and normal 187–204. https://doi.org/10.1901/jaba.2003.36-187
educational and intellectual functioning in young Piazza, C. C., Patel, M. R., Gulotta, C. S.,
autistic children. Journal of Consulting and Clinical Sevin, B. M., & Layer, S. A. (2003). On the relative
Psychology, 55, 3–9. https://doi.org/10.1037/0022- contributions of positive reinforcement and escape
006X.55.1.3 extinction in the treatment of food refusal. Journal of
22 Jaime G. Crowley et al.

Applied Behavior Analysis, 36, 309–324. https://doi. Developmental Disabilities, 27, 353–363. https://doi.
org/10.1901/jaba.2003.36-309 org/10.1016/j.ridd.2005.03.005
Pierce, W. D., & Epling, W. F. (1983). Choice, Sharp, W. G., Postorino, V., McCracken, C. E.,
matching, and human behavior: A review of the liter- Berry, R. C., Criado, K. K., Burrell, T. L., &
ature. The Behavior Analyst, 6, 57–76. https://doi. Scahill, L. (2018). Dietary intake, nutrient status,
org/10.1007/BF03391874 and growth parameters in children with autism spec-
Postorino, V., Sanges, V., Giovagnoli, G., Fatta, L. M., trum disorder and severe food selectivity: An elec-
De Peppo, L., Armando, M., … Mazzone, L. tronic medical record review. Journal of the Academy
(2015). Clinical differences in children with autism of Nutrition and Dietetics, 118, 1943–1950. https://
spectrum disorder with and without food selectivity. doi.org/10.1016/j.jand.2018.05.005
Appetite, 92, 126–132. https://doi.org/10.1016/j. Silbaugh, B. C., Swinnea, S., & Penrod, B. (2018). Syn-
appet.2015.05.016 thesis of applied behavior analytic interventions for
Reed, G. K., Piazza, C. C., Patel, M. R., Layer, S. A., packing in pediatric feeding disorders. Behavior Modi-
Bachmeyer, M. H., Bethke, S. D., & Gutshall, K. A. fication, 42, 249–272. https://doi.org/10.1177/
(2004). On the relative contributions of non- 0145445517724541
contingent reinforcement and escape extinction in Smith, T., & Iadarola, S. (2015). Evidence base update
the treatment of food refusal. Journal of Applied for autism spectrum disorder. Journal of Clinical
Behavior Analysis, 37, 27–42. https://doi.org/10. Child & Adolescent Psychology, 44, 897–922. https://
1901/jaba.2004.37-27 doi.org/10.1080/15374416.2015.1077448
Reichow, B. (2012). Overview of meta-analyses on early
intensive behavioral intervention for young children U.S. Department of Health and Human Services and
with autism spectrum disorders. Journal of Autism U.S. Department of Agriculture (2015). 2015–2020
and Developmental Disorders, 42, 512–520. https:// Dietary guidelines for Americans (8th Ed.). http://
doi.org/10.1007/s10803-011-1218-9 health.gov/dietaryguidelines/2015/guidelines/
Richler, J., Huerta, M., Bishop, S. L., & Lord, C. (2010). Wadhera, D., & Capaldi Phillips, E. D. (2014). A review
Developmental trajectories of restricted and repetitive of visual cues associated with food on food acceptance
behaviors and interests in children with autism spec- and consumption. Eating Behaviors, 15, 132–143.
trum disorders. Development and Psychopathology, 22, https://doi.org/10.1016/j.eatbeh.2013.11.003
55–69. https://doi.org/10.1017/s0954579409990265
Schreck, K. A., & Williams, K. (2006). Food preferences Received April 25, 2019
and factors influencing food selectivity for children Final acceptance February 19, 2020
with autism spectrum disorders. Research in Action Editor, Dorthea Lerman

You might also like