evolution of meal time 2021

An evaluation of a renewal-mitigation procedure for inappropriate mealtime behavior

Sarah D. Haney , Cathleen C. Piazza, Kathryn M. Peterson and Brian D. Greer

University of Nebraska Medical Center’s Munroe-Meyer Institute

Renewal, the increase in behavior during extinction following context changes, may be particu- larly concerning during intervention for feeding disorders because context changes are often nec- essary for intervention generality and maintenance (Podlesnik et al., 2017). In the current study, we tested for renewal and evaluated a renewal-mitigation procedure when we transferred intervention from a therapist to a caregiver, from clinic to the home, and changed the foods the feeder presented. We used an ABA arrangement to evaluate the generality of the renewal effect with 7 participants who engaged in inappropriate mealtime behavior. Context A was functional reinforcement. Context B was function-based extinction during the control and mitigation con- ditions and our renewal-mitigation procedure in the mitigation condition. The renewal test was function-based extinction in Context A. We observed renewal of inappropriate mealtime behav- ior in 4 of 7 participants, and our renewal-mitigation procedure was effective for 4 of 4 participants. Key words: avoidant/restrictive food intake disorder, escape extinction, feeding disorder, mit-

igation, renewal, translational research

The diagnosis avoidant/restrictive food intake disorder describes children who do not consume enough calories or nutrients to maintain weight or grow (American Psychiatric Association, 2013). Children with this diagnosis often engage in inappropriate mealtime behavior, like batting at the spoon, to avoid bite or drink presenta- tions. Although researchers have demonstrated the effectiveness of escape extinction for decreas- ing inappropriate mealtime behavior in clinical

settings, less is known about the generalization and long-term maintenance of escape extinction in other contexts. Context is one variable that may affect the

generality and long-term maintenance of extinction (e.g., Kelley et al., 2015; Podlesnik et al., 2017). Context is any stimulus that exerts control over behavior (e.g., Bouton et al., 2011; Kelley et al., 2015; Podlesnik et al., 2017; Todd, 2013). Researchers hypothesize that context changes may be responsible for increases in previously extinguished behavior, a relapse phenomenon called renewal (Bouton et al., 2011; Kelley et al., 2015; Podlesnik et al., 2017). Context theory states that renewal occurs because the context functions as a dis- criminative stimulus for the extinction contin- gencies, and the effects of these discriminative stimuli are specific to the original extinction context. The effects of extinction do not have generality outside the original extinction con- text (Bouton et al., 2011; Podlesnik et al., 2017; Trask et al., 2017; Wathen & Podlesnik, 2018), perhaps because the stimuli that were

Sarah D. Haney is now at the Kennedy Krieger Insti- tute, Baltimore, Maryland. Cathleen C. Piazza and Kathryn M. Peterson are now at Children’s Specialized Hospital, Somerset, New Jersey, and Graduate School of Applied and Professional Psychology, Rutgers University. Brian D. Greer is now at Children’s Specialized Hospital- Rutgers University Center for Autism Research, Educa- tion, and Services, and Rutgers Robert Wood Johnson Medical School.Sarah Haney conducted this investigation in partial fulfillment of the requirements for the degree of Doctor of Philosophy from the University of Nebraska Medical Center. Address correspondence to Cathleen Piazza at:

[email protected] or Sarah Haney at: [email protected] doi: 10.1002/jaba.815

Journal of Applied Behavior Analysis 2021, 54, 903–927 NUMBER 3 (SUMMER)

© 2021 Society for the Experimental Analysis of Behavior

903

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

present in the original context are not present when the context changes (Wathen & Podlesnik, 2018). The study of renewal is particularly relevant

during intervention for children with avoidant/ restrictive food intake disorder for several rea- sons. First, context changes are common because children feed in many environments with different individuals. Second, studies have shown that inappropriate mealtime behavior may reemerge following changes in the feeder (Kelley et al., 2018) and feeder and setting com- bined (Ibañez et al., 2019). Third, inappropriate mealtime behavior that emerges during context changes may contact reinforcement, which may maintain the behavior over time (Kelley et al., 2018; Mitteer et al., 2018). Borrero et al. (2010) showed that the most common conse- quence that caregivers delivered following inap- propriate mealtime behavior was escape in the form of removal of bite or drink presentations or meal termination. Attention was the second most common caregiver consequence. Fourth, results of studies on functional analysis of inap- propriate mealtime behavior have shown that escape functions as negative reinforcement for inappropriate mealtime behavior for most chil- dren with feeding disorders (Bachmeyer et al., 2019; Girolami & Scotti, 2001; Najdowski et al., 2008; Piazza et al., 2003). These data sug- gest that if inappropriate mealtime behavior reemerges, caregivers are likely to deliver escape and possibly attention and these consequences are likely to function as reinforcement. Fifth, Mitteer et al. (2018) showed that undesirable caregiver behavior (e.g., providing the func- tional reinforcer following destructive behavior) may also relapse during context changes, which suggests that caregivers may provide reinforce- ment for inappropriate mealtime behavior dur- ing context changes. This is problematic for children with feeding disorders because insuffi- cient calories or nutrients may have negative short- and long-term consequences like poor growth and behavior and learning problems,

respectively (Freedman et al., 1999; Peterson et al., 2018; Volkert & Piazza, 2012) and cause caregiver anxiety, depression, and stress (Bryant- Waugh et al., 2010; Greer et al., 2007; Kreipe & Palomaki, 2012). Lastly, studying renewal may inform strategies to improve the generality and long-term maintenance of intervention effects (Ibañez et al., 2019; Kelley et al., 2018; Podlesnik et al., 2017; Wathen & Podlesnik, 2018). Our standard clinical practice provides an

opportunity to evaluate renewal in children with avoidant/restrictive food intake disorder and par- allels the ABA arrangement researchers have used to study renewal (Bouton et al., 2011). When we admit a child, we observe caregivers feeding the child in the clinic using the consequences they provide at home, which we define as Context A. Because admitted children have complex feed- ing problems, trained therapists often implement the child’s intervention in the clinic initially, which is Context B. Next, we train caregivers to implement the intervention. Thus, the return to Context A is the caregiver implementing the intervention in the clinic. Caregivers implement the intervention in the home after their imple- mentation integrity in the clinic is high, which is an additional context change. More context changes may be necessary for individual children, such as new settings (e.g., daycare), feeders (e.g., babysitter), foods, or textures. Two studies have demonstrated renewal dur-

ing feeding interventions (Ibañez et al., 2019; Kelley et al., 2018). In Context A in Ibañez et al. (2019), the caregiver provided functional reinforcement for inappropriate mealtime behav- ior in a simulated home setting. In Context B, a therapist implemented function-based extinction in a clinic room. In the return to Context A, the caregiver implemented function-based extinction in a simulated home setting. Renewal of inap- propriate mealtime behavior occurred across par- ticipants even though caregivers implemented extinction with high levels of integrity. Researchers have shown that caregivers pro-

vide escape following inappropriate mealtime

Sarah D. Haney et al.904

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

behavior (Borrero et al. 2010) and that escape functions as reinforcement for inappropriate mealtime behavior (Girolami & Scotti, 2001; Nadjowski et al., 2008; Piazza et al. 2003). Researchers have also shown that inappropriate mealtime behavior (Ibañez et al., 2019; Kelley et al. 2018) and undesirable caregiver behavior (Mitteer et al., 2018) may reemerge following context changes during intervention. Thus, devel- oping and testing procedures to mitigate renewal is important. We could find only one study that assessed renewal-mitigation for inappropriate mealtime behavior. Kelley et al. (2018) used an ABABA+BA arrangement and evaluated a change from therapist to caregiver during function-based extinction for two participants with avoidant/ restrictive food intake disorder. Inappropriate mealtime behavior returned for both participants during the first renewal test when caregivers implemented extinction. During the return to Context B, the therapist implemented extinction. Context A + B was their renewal-mitigation pro- cedure in which the caregiver sat next to the ther- apist while the therapist implemented extinction. During the second renewal test when the care- giver implemented extinction, renewal did not occur for one participant. Rate of inappropriate mealtime behavior was lower and less persistent following the mitigation procedure for the other participant. One limitation was that repeated exposure to the context change, like in the rever- sal design Kelley et al. used, may reduce the likeli- hood that behavior will return (Sweeney & Shahan, 2013; Wacker et al., 2011). Thus, we cannot be certain that the results were due to the renewal-mitigation procedure or to repeated expo- sure to the context change. In the current study, we extended the findings

of Ibañez et al. (2019) by evaluating renewal of inappropriate mealtime behavior during function- based extinction in the control condition and of Kelley et al. (2018) by developing and evaluating a renewal-mitigation procedure in our mitigation condition. Results of previous research have shown that renewal may be less likely when cues for

extinction are present in contexts previously asso- ciated with reinforcement (Bernal-Gamboa, Gámez, & Nieto, 2017; Kelley et al., 2018; Nieto et al., 2017; Todd et al., 2012). Researchers have also shown that increasing the similarity between acquisition and extinction contexts may mitigate renewal when returning to contexts associated with reinforcement (Bandarian-Balooch & Neumann, 2011; Todd et al., 2012). Combining renewal-mitigation procedures may mitigate renewal more effectively than these procedures individually (Bandarian-Balooch & Neumann, 2011; Bernal-Gamboa, Nieto, & Uengoer, 2017; Krisch et al., 2018). For example, Bandarian- Balooch and Neumann (2011) used an ABA renewal arrangement to evaluate renewal of shock expectancy in humans and found that the combi- nation of implementing extinction in multiple contexts (i.e., varying light intensities) and con- text similarity was most effective at mitigating renewal relative to these procedures alone. Our renewal-mitigation procedure used a

therapist–caregiver pairing procedure like Kelley et al. (2018) when we changed the feeder from therapist to caregiver, food-fading when we changed the foods the feeder presented, and con- text similarity when we changed from the clinic to the home. Context A was the caregiver or therapist delivering function-based reinforce- ment for inappropriate mealtime behavior in the control and mitigation conditions. Context B was the therapist implementing function-based extinction in the control and mitigation condi- tions and implementing the renewal-mitigation procedure in the mitigation condition. We eval- uated rate of inappropriate mealtime behavior in a return to Context A in which the caregiver or therapist implemented function-based extinction in the control and mitigation conditions.

Method

Participants Participants were seven children who met the

criteria for avoidant/restrictive food intake disorder

905Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

admitted to an intensive day-treatment program and their caregivers between 2018 and 2020. Par- ticipants attended the program 5 days per week from approximately 9:00 a.m. to 5:00 p.m. Inclu- sion criteria were the participant (a) was between 2 and 18 years old, (b) was safe for oral feeding per a pediatrician or speech pathologist, (c) received 50% or more of daily calories from supplemental feeding or had a diet that was nutri- tionally inadequate per our registered dietitian, and (d) had at least 3 months of outpatient feeding therapy without progress. Exclusion criteria were (a) the participant did not meet the inclusion criteria, (b) the participant engaged in packing, expulsion, or both that did not improve during escape extinction, and (c) acceptance, mouth clean, inappropriate mealtime behavior, or a combination of these behaviors did not improve during escape extinction. We excluded four children from the study, one for persistent packing, one for persistent expulsion, and two because acceptance and inap- propriate mealtime behavior did not improve. Our program’s registered dietitian used the Centers for Disease Control and Prevention (2010) growth chart for ages 2-20 for John, Maisy, Hope, Emilia, Diego, and Julian and the 22q11 deletion syn- drome growth chart for girls ages 2-20 (Tarquinio et al., 2012) for Jade to calculate each child’s body-mass index, height, and weight. Seven chil- dren, three males and four females, between the ages of 3 to 6 met the inclusion criteria and partic- ipated in the study. Table 1 describes participant demographics, nutritional information, and growth parameters upon admission into the Pediatric Feeding Disorders Program.

Setting and Materials Feeders conducted control and mitigation ses-

sions with participants in 4-m by 4-m clinic rooms with adjacent observation rooms with one- way observation and two-way audio and sound. Clinic rooms contained a rectangular table, uten- sils, a food tray, a scale, a chair for the feeder, and weight-appropriate seating for the participant;

and colored bowls, tablecloths, and cards for Emi- lia, Diego, and Julian. Caregivers conducted Jade’s sessions in their home, and therapists observed these sessions using Vidyo, a secure video-conferencing software. John and Emilia sat in a Special Tomato Soft-Touch sitter, and Maisy and Julian sat in a booster seat. We secured the sitter and booster seat to a kitchen chair. Hope sat in a regular kitchen chair. Diego and Jade sat in a highchair. Caregivers selected 16 foods that the partici-

pant currently did not consume, four fruits, four proteins, four starches, and four vegetables with input from the dietitian. Caregivers of John, Maisy, Hope, and Jade also selected an age-, calorically, and nutritionally appropriate liquid as increasing liquid consumption was a target of their treatment.

Feeders Feeders were the participant’s biological mother

(Maisy, Hope, Emilia, Jade, and Diego) and father (John and Hope) and trained clinic therapists. Therapists also served as observers. Training for therapists included didactic modules with posttests, competency training for session and food prepara- tion, and in-vivo training to competency for partic- ipant assessment and treatment and caregiver- training procedures, wherein the therapist must implement the procedures with a confederate child across a variety of child responses with 80% or higher integrity in the absence of corrective feed- back before implementing these procedures with a child or caregiver. We collected data on therapist integrity and provided feedback if necessary when the therapist implemented the procedures with a child or caregiver to ensure that the therapist maintained 80% or higher integrity.

Dependent Variables, Procedural Integrity, and Reliability Observers used BDataPro® to collect data on

inappropriate mealtime behavior, acceptance, and procedural integrity in the control and

Sarah D. Haney et al.906

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

T ab le 1

Pa rt ic ip an t D em

og ra ph ic s, N ut ri tio na lI nf or m at io n,

an d G ro w th

Pa ra m et er su

po n A dm

iss io n in to th e Fe ed in g Pr og ra m

% N ee ds

M et by

M ou th

N ut ri tio

na lI na de qu ac ie s

G ro w th

Pa ra m et er s

N am

e A ge

D ia gn os es

C al or ie s Pr ot ei n

Fl ui ds

H ei gh t W ei gh t

B od y M as s

In de x

Fo od

C on su m ed

Jo hn

6 A ut is m

11 4%

34 8%

13 6%

ch ro m iu m ,i ro n,

m an ga ne se ,

po ta ss iu m ,s el en iu m ,a nd

vi ta m in s

E an d B 3

33 rd

12 th

17 th

Pe di as ur e 1. 0,

yo gu rt ,f ru it ba by

fo od ,a nd

pu dd in g

M ai sy

4 E hl er s D an lo s sy nd

ro m e; hi st or y of

fo od

al le rg ie s, ga st ro st om

y- tu be

de pe nd

en ce ,f ai lu re

to th ri ve ,F

oo d

Pr ot ei n In du

ce d E nt er oc ol iti s

Sy nd

ro m e, an d m ilk

so y pr ot ei n

in to le ra nc e

43 %

20 4%

32 %

ca lc iu m ,c hr om

iu m ,c op pe r,

m ag ne si um

,m an ga ne se ,

po ta ss iu m ,a nd

vi ta m in s B 5 an d D

<1 st

<1 st

1s t

3- 5 oz

of Pu

ra m in o vi a bo tt le ,w

at er ,

be an s, eg gs ,p

ea ch es ,c hi ps ,b

re ad ,

m ac ar on ia nd

ch ee se ,a nd

ho td og

H op e

5 H is to ry

of de ve lo pm

en ta ld

el ay ;

re so lv ed

ne ur ol og ic al de fe ct w ith

fa ci al dr oo pi ng ;a nd

na so je ju na l-,

ga st ro st om

y- ,a nd

ga st ro st om

y- je ju no st om

y- tu be

de pe nd

en ce

28 %

12 6%

13 %

bi ot in ;f ol at e; vi ta m in s A ,B

1, B 2,

B 6,

B 12 ,C

,D ,a nd

E 7t h

3r d

13 th

Ju ic e, ch oc ol at e m ilk ,c hi ck en

nu gg et s, st ea k,

Fr en ch

fr ie s, br ea d,

po ta to es ,p

in ea pp le ,a pp le ,o

ra ng e,

an d ap pl es au ce

E m ili a 5

H is to ry

of ga gg in g an d vo m iti ng ,o

ra l

m ot or

sk ill

de fi ci ts w ith

di ffi cu lty

sw al lo w in g, an d a so y al le rg y

84 %

16 0%

40 %

ca lc iu m ,i ro n,

m an ga ne se ,p

ot as si um

, an d vi ta m in

D 68

th 32

n d

12 th

E ns ur e, Pe di as ur e 1. 0,

ju ic e, ch ip s,

co ok ie s, cr ac ke rs ,d

on ut s, an d ri ce

ca ke s

D ie go

3 T ot al pa re nt er al nu

tr iti on

de pe nd

en ce

an d di ag no se s of

in te st in al fa ilu re

an d sh or t bo w el

sy nd

ro m e

80 %

22 1%

91 %

N on e

19 th

40 th

68 th

59 m lo

f N eo ca te Jr .f ro m

a bo tt le

ap pr ox im

at el y ev er y 2 hr

w hi le

aw ak e, go ld fi sh

cr ac ke rs ,a nd

pe an ut

bu tt er

ba lls

Ja de

4 D iG eo rg e sy nd

ro m e (2 2q 11

de le tio

n sy nd

ro m e) an d hi st or y of

co ng en ita lh

ea rt de fe ct ,

ga st ro st om

y- tu be

de pe nd

en ce ,

dy sp ha gi a, ga st ro es op ha ge al re fl ux

di se as e, N is sa n fu nd

op lic at io n,

an d

or al m ot or

sk ill

de fi ci t

73 %

21 0%

60 %

ca lc iu m ,s el en iu m ,a nd

vi ta m in

D 14

th 7t h

9t h

D id

no t co ns um

e an y fr ui ts ,

pr ot ei ns ,s ta rc he s, or

ve ge ta bl es

or al ly

Ju lia n

5 A ut is m

89 %

18 9%

43 %

ca lc iu m ;c hr om

iu m ;m

ag ne si um

; m an ga ne se ;p

ot as si um

;z in c; an d

vi ta m in s B 5,

B 7,

D ,a nd

E

67 th

72 n d

68 th

Ju ic e, w at er ,c er ea l, cr ac ke rs ,c hi ps ,

co rn do gs ,p

ea nu

t bu

tt er

an d je lly

sa nd

w ic he s, go ld fi sh

cr ac ke rs ,

ba na na ,g ra pe s, fr ui t sn ac ks ,o

re os ,

an d po pt ar ts .

907Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

mitigation conditions (Bullock et al., 2017). During caregiver training, observers scored occurrences and nonoccurrences of procedural integrity on each intervention component using an Excel spreadsheet that generated a graph from the inputted data.

Dependent Variables Observers scored the frequency of inappropri-

ate mealtime behavior when the spoon or cup was within arm’s reach of the participant and each time the participant (a) moved the middle of their mouth 45� or more in any direction or 6.4 cm in any direction away from the utensil; (b) changed direction or paused for 1 s and then moved the head another 45� in any direction or 6.4 cm in any direction relative to the previous position; or if the participant’s hand, arm, or anything in the participant’s hand (e.g., bib) touched the utensil, food, liquid, or any part of the feeder’s hand or arm from the elbow down that the feeder was using to implement the feed- ing procedure; (c) threw the spoon or cup; (d) placed their hand, arm, toy, or bib within 5 cm of the mouth; and (e) moved more than 2.5 cm from and was no longer touching the spoon, cup, feeder’s hand or arm from the elbow down, or the participant’s own mouth and then returned to the spoon, cup, feeder’s hand or arm from the elbow down, or the participant’s own mouth. A presentation occurred the first time the feeder touched the utensil to the participant’s lips to offer the participant the opportunity to accept that bite or drink initially. Observers scored the occurrence of acceptance when the participant (a) opened their mouth and was not crying, screaming, yelling, or making refusal statements or (b) opened their mouth and leaned forward while engaging in negative vocalizations and the feeder deposited the bite or drink within 5 s of the initial presentation. A deposit occurred when the entire bite or drink passed the plane of the wet vermillion of the participant’s lips, and no food or liquid remained on the spoon or in the cup when the feeder removed the utensil

from the participant’s lips. The number of opportunities for observers to score acceptance corresponded to the number of bites or drinks the feeder presented. We defined an entire bite or drink as the volume of solids or liquids the feeder presented on the utensil minus a pea-sized volume or smaller. BDataPro® converted the frequency of inap-

propriate mealtime behavior to responses per minute by dividing the number of inappropriate mealtime behaviors during the session by the duration the utensil was within arm’s reach of the participant. BDataPro® converted acceptance to a percentage after dividing the number of accep- tances by the total number of presentations.

Procedural Integrity Correct-utensil placement assessed procedural

integrity for components specific to the feeder’s utensil use. Observers scored duration of correct-utensil placement in the control and mitigation conditions by activating a duration key on BDataPro® when the feeder met the cri- terion for correct-utensil placement and deactivating the duration key when the feeder did not meet the criterion for correct-utensil placement for 3 s or more. Across all conditions of the study, observers scored correct-utensil placement when the feeder presented the utensil touching the participant’s lips at the scheduled interval; removed the utensil after the bite or drink entered the mouth; and held the utensil to the side of the participant’s mouth if the par- ticipant coughed, gagged, or vomited while the feeder was holding the utensil at the partici- pant’s lips. Observers scored correct-utensil placement during function-based reinforcement if the feeder removed the utensil after the occur- rence of inappropriate mealtime behavior. Observers scored correct-utensil placement dur- ing function-based extinction when the feeder held the spoon touching the participant’s lips until they could deposit the bite or drink into the mouth, or the time-cap elapsed and used the utensil to re-present the expelled bite or drink

Sarah D. Haney et al.908

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

into the mouth within 3 s of expulsion. BDataPro® converted duration of correct- utensil placement to a percentage after divid- ing the duration of correct-utensil placement by the total session time. Mean correct-utensil placement during function-based reinforce- ment in Context A was 99% (range, 99% to 100%) for caregivers and 99% (range, 99% to 100%) for therapists during the control condition, and 99% (range, 91% to 100%) for caregivers and 99% (range, 99% to 100%) for therapists during the mitigation condition. Mean correct-utensil placement during function-based extinction in Context B was 99% (range, 88% to 100%) for therapists dur- ing the control condition, and 99% (range, 88% to 100%) for caregivers and 99% (range, 89% to 100%) for therapists during the miti- gation condition. Mean correct-utensil place- ment during the renewal test in Context A was 99% (range, 98% to 100%) for caregivers and 99% (range, 88% to 100%) for therapists during the control condition, and 99% (range, 99% to 100%) for caregivers and 99% (range, 99% to 100%) for therapists during the miti- gation condition. Correct procedure assessed procedural integ-

rity for components other than those specific to the feeder’s utensil use. Observers scored correct procedure if the feeder completed every compo- nent described below and did not score correct procedure if the feeder did not perform a com- ponent or performed a component incorrectly. Observers scored the occurrence of correct proce- dure in the control and mitigation conditions after the mouth check or at the end of the pre- sentation interval if the feeder (a) used the cor- rect utensil and presented the specified bolus size, (b) delivered a vocal prompt to “Take a bite (drink)” while touching the utensil to the participant’s lips within 5 s of the scheduled presentation, (c) deposited the bite or drink within 5 s of presentation if the participant met criterion for acceptance, (d) delivered vocal, behavior-specific praise within 5 s of acceptance

and mouth clean (i.e., no food in the mouth at the mouth check in the absence of spitting out the bite or drink), (e) delivered a vocal prompt to “Show me, ahh” while conducting a mouth check approximately 30 s after the feeder depos- ited the bite or drink into the participant’s mouth, (f) delivered the vocal prompt to “Swal- low your bite (drink)” within 5 s of the mouth check if the participant packed (i.e., held food or liquid larger than a pea-size volume in the mouth) and delivered the swallow prompt every 30 s if the participant packed on the fifth bite or drink presentation until there was no food or liquid in the participant’s mouth or until 10 min elapsed, (g) provided no differential consequences for coughing, gagging, or vomiting, and (h) presented the next bite or drink within 5 s of the next scheduled presenta- tion interval. Observers scored correct proce- dure for caregivers and therapists separately when therapists fed and caregivers delivered other components of the extinction intervention during the mitigation condition. BDataPro®

converted correct procedure to a percentage after dividing the number of bite or drink pre- sentations with correct procedure by the total number of bite or drink presentations. Mean correct procedure during function-based rein- forcement in Context A was 98% (range, 80% to 100%) for caregivers and therapists during the control condition, and 97% (range, 80% to 100%) for caregivers and 99% (range, 80% to 100%) for therapists during the mitigation con- dition. Mean correct procedure during function-based extinction in Context B was 99% (range, 80% to 100%) for therapists dur- ing the control condition, and 98% (range, 80% to 100%) for caregivers and 100% for therapists during the mitigation condition. Mean correct procedure during the renewal test in Context A was 99% (range, 80% to 100%) for caregivers and 100% for therapists during the control condition, and 97% (range, 80% to 100%) for caregivers and 100% for therapists during the mitigation condition.

909Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

Observers scored occurrences of incorrect attention once during each presentation interval in the control and mitigation conditions when the feeder (a) did not provide behavior-specific praise within 5 s of acceptance and mouth clean, (b) delivered behavior-specific praise if the bite or drink entered the participant’s mouth after 5 s of presentation, or (c) delivered behavior- specific praise if there was food or liquid in the participant’s mouth at the mouth check. Observers also scored incorrect attention during escape baseline and function-based extinction when the feeder delivered attention within 5 s of inappropriate mealtime behavior. Observers scored incorrect attention during the escape and attention baseline when the feeder did not deliver attention within 5 s of inappropriate mealtime behavior. BDataPro® converted incor- rect attention to a percentage after dividing occurrences of incorrect attention by the total number of bite presentations in the session. Mean incorrect attention during function-based reinforcement in Context A was 1% (range, 0% to 20%) for caregivers and 0% for therapists during the control condition, and 0.3% (range, 0% to 20%) for caregivers and 0% for therapists during the mitigation condition. Mean incorrect attention during function-based extinction in Context B was 0% for therapists during the con- trol condition and 0.6 % (range, 0% to 20%) for caregivers, and 0% for therapists during the mitigation condition. Mean incorrect attention during the renewal test in Context A was 0.5% (range, 0% to 20%) for caregivers and 0% for therapists during the control condition, and 1% (range, 0% to 20%) for caregivers and 0% for therapists during the mitigation condition.

Interobserver Agreement Two observers simultaneously and indepen-

dently collected data on a mean of 54% and 66% of sessions in the control and mitigation conditions, respectively. We trained observers to collect data on BDataPro® and on a stream- lined Excel spreadsheet with 80% or higher

interobserver agreement for three consecutive sessions. The BDataPro® software calculated inter-

observer agreement by dividing each session into 10-s intervals. BDataPro® calculated exact agree- ment coefficients for inappropriate mealtime behavior by dividing the number of exact agree- ments (observers scored the same frequency of behavior in an interval) by the number of exact agreements plus disagreements (observers scored different frequencies of behavior in an interval) and converting this ratio to a percentage. Mean interobserver agreement was 94% (range, 73% to 100%) and 96% (range, 71% to 100%) for inappropriate mealtime behavior during the con- trol and mitigation conditions, respectively. BDataPro® calculated total agreement coeffi- cients for acceptance, correct-utensil placement, correct procedure, and incorrect attention by dividing the total number of agreements (i.e., both observers scored the occurrence or nonoccurrence of the behavior in the interval) by the total number of agreements plus disagree- ments (one observer scored and the other observer did not score the occurrence of the behavior in the interval) and converting this ratio to a percentage. Mean interobserver agree- ment during the control condition was 99% (range, 80% to 100%) for acceptance, 99% (range, 90% to 100%) for correct-utensil place- ment, 99% (range, 80% to 100%) for correct procedure, and 99% (range, 80% to 100%) for incorrect attention. Mean interobserver agree- ment during the mitigation condition was 98% (range, 80% to 100%) for acceptance, 99% (range, 85% to 100%) for correct-utensil place- ment, 98% (range, 80% to 100%) for correct procedure, and 99% (range, 80% to 100%) for incorrect attention.

Experimental Design Control and Mitigation Conditions We used a three-phase ABA arrangement to

evaluate the control and mitigation conditions.

Sarah D. Haney et al.910

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

During Context A, the caregivers of John, Maisy, Hope, Emilia, Diego, and Jade and a therapist for Julian delivered function-based reinforcement for inappropriate mealtime behav- ior. During Context B, a therapist implemented function-based extinction of inappropriate meal- time behavior during the control and mitigation conditions. The therapist also implemented the mitigation procedure in the mitigation condi- tion. During the renewal test in Context A, the caregivers of John, Maisy, Hope, Emilia, Diego, and Jade and a therapist for Julian implemented function-based extinction of inappropriate meal- time behavior.

General Procedure Feeders conducted five 40-min meals per

day with at least 40 min between the end of one meal and the start of the next meal. Each session generally consisted of five bite or drink presentations with approximately 1 min between each session during which feeders and observers prepared for the next session. The number of sessions per meal depended on the duration of each session within the meal, which depended on the participant’s behavior. Mean number of sessions per meal was four for Maisy; five for John, Hope, Emilia, Diego, and Julian; and six for Jade. The feeder generally conducted control sessions in some meals and mitigation sessions in other meals. The feeder only presented solids to Emilia, Diego, and Julian. The feeder generally presented solids in some meals and liquids in other meals for John, Maisy, Hope, and Jade. The only time the feeder conducted the control and mitigation conditions in the same meal was if the partici- pant’s behavior met the criterion to change phases and implement the contingencies in the next context, such as a change from Context A to Context B, and we needed to equate the number of sessions between conditions before changing phases. For example, if the participant met criterion in solids sessions during Meal

1 in Context A to change to Context B, then the feeder stopped conducting solids sessions in Meal 1 in Context A so that we could transi- tion to solids sessions in Context B in Meal 2. If 20 min or more remained in Meal 1 and the participant had not met criterion to change to Context A in liquids sessions, then the feeder conducted liquids sessions in Context A for the rest of Meal 1. Feeders presented 1 cc of pureed food, which

was table food blended until smooth with liq- uid added as needed, on a small maroon spoon in solids sessions. Feeders leveled the bolus by scooping pureed food onto the spoon, scraping the bowl of the spoon on the side of the dish to flatten the bolus, and scraping the bottom of the spoon’s bowl on the dish to remove excess puree. The feeder presented eight of the 16 caregiver-selected foods, two fruits, two pro- teins, two starches, and two vegetables, during the control and mitigation conditions for John, Maisy, Hope, Emilia, Jade, and Julian; 16 care- giver-selected foods during the control and mit- igation conditions for Diego. The feeder presented the same foods in every condition in every phase to control for potential differences in participant behavior as a function of food type (Patel et al., 2002) and randomly selected the order of food presentation before the start of the session. Feeders presented 2 cc of Vanilla Pediasure

1.0 with and without fiber in a small pink cut- out cup for Jade and John, respectively. Feeders presented 4 cc of Vanilla Puramino Junior and Peptamin Junior 1.5 in a pink cut-out cup for Maisy and Hope, respectively. During each session, the feeder presented a

bite or drink by touching the utensil to the par- ticipant’s lips while saying, “Take a bite (drink).” The feeder deposited the bite or drink if the participant met criterion for acceptance and delivered behavior-specific praise for accep- tance. The feeder conducted a mouth check approximately 30 s after they deposited the bite or drink into the participant’s mouth by saying,

911Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

“Show me, ahh.” If the participant did not open their mouth within 3 s of the prompt, the feeder inserted a small rubber-coated spoon between the participant’s lips with the bowl of the spoon touching the participant’s lower lip and turned the spoon 90� such that the bowl of the spoon faced the participant’s teeth to prompt the par- ticipant to open their mouth. If the child did not open their mouth within 3 s of this prompt, the therapist inserted the spoon between the participant’s teeth and used the same procedure to prompt the participant to open their mouth. The feeder delivered behavior-specific praise for mouth clean or a verbal prompt to “Swallow your bite (drink)” if the participant was packing and presented the next bite or drink. If the par- ticipant showed the feeder a mouth clean before 30 s elapsed to signal the mouth check, the feeder delivered behavior-specific praise and waited until the mouth-check interval elapsed before presenting the next bite or drink. If the participant had packed food or liquid in their mouth after the fifth bite or drink presentation, the feeder conducted a mouth check and deliv- ered the verbal prompt to swallow every 30 s until no food or liquid was in the participant’s mouth or until 10 min had elapsed from session initiation. The feeder wiped out the partici- pant’s mouth after the session. The feeder deliv- ered no differential consequence for coughing, gagging, or vomiting. The feeder moved the utensil to the side of the participant’s mouth if the participant coughed, gagged, or vomited while the utensil was at the participant’s lips and moved the utensil back to the participant’s lips when the participant stopped coughing, gagging, or vomiting.

Control and Mitigation Conditions Before the current study, we conducted an

unstructured observation in which the caregiver presented preferred and nonpreferred food and liquid to the participant. Next, we conducted a structured observation in which we observed

the caregiver (a) attempt to feed the participant bites of pureed food, (b) attempt to feed the participant liquids from a cup, (c) prompt the participant to self-feed bites of pureed food, (d) prompt the participant to self-feed bites of table-textured food, and (e) prompt the partici- pant to self-feed liquids from a cup. Third, a therapist conducted a paired-choice preference assessment (Fisher et al., 1992) and a functional analysis based on the procedure Bachmeyer et al. (2009) described. One purpose of the study was to assess

whether renewal of inappropriate mealtime behavior occurred when we implemented function-based-extinction intervention for chil- dren with avoidant/restrictive food intake disor- der. The context changes were changing from (a) a therapist to a caregiver as feeder for John, Maisy, Hope, and Emilia in the clinic; (b) from a therapist feeding four foods in the clinic to a caregiver feeding four different foods in the clinic for Diego; (c) from a therapist as feeder in the clinic to a caregiver as feeder at home for Jade; and (d) from a clinic room to a simulated home room for Julian. We compared levels of renewal in the mitigation condition with those of the control condition to determine whether the mitigation procedure was effective. A second purpose was to assess whether our renewal- mitigation procedures would prevent renewal following these context changes in the mitiga- tion condition. The arrangement for the mitiga- tion condition was like the control condition except that we implemented our renewal- mitigation procedure during extinction in Context B. We assessed renewal effects during liquids and solids intervention for John, Maisy, Hope, and Jade and during solids intervention for Emilia, Diego, and Julian. We used colored stimuli to aid Emilia, Diego, and Julian’s dis- crimination of the conditions because the feeder presented pureed food in both conditions, which do not have as many distinctive proper- ties as the same foods at table texture. We also randomly chose the order of sessions during the

Sarah D. Haney et al.912

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

renewal test across the control and mitigation conditions with each participant. We conducted renewal-test sessions in the control condition first and the mitigation condition second with John, Maisy, Emilia, and Jade and vice versa with Hope, Diego, and Julian. We randomly assigned solids for John and

Hope and liquids for Maisy and Jade to the con- trol condition and liquids for John and Hope and solids for Maisy and Jade to the mitigation condition. We randomly assigned one fruit, one protein, one starch, and one vegetable to a food set for Emilia, Diego, and Julian. For Emilia, we randomly assigned Set 1 foods (baked bean, green bean, pancake, pear), orange bowls, and an orange table cloth to the control condition and Set 2 foods (banana, carrot, chicken, potato), blue bowls, and a blue table cloth to the mitigation condition. For Diego, we ran- domly assigned Set 1 foods (banana, cauliflower, egg, pancake), Set 2 foods (applesauce, green bean, hamburger, oatmeal), blue bowls, and a blue 10-cm x 17-cm card to the control condi- tion and Set 3 foods (avocado, broccoli, sweet potato, tuna), Set 4 foods (chicken, potato, strawberry, squash), orange bowls, and an orange 10-cm x 17-cm card to the mitigation condition. For Julian, we randomly assigned Set 1 foods (applesauce, carrot, chicken, potato), orange bowls, an orange table cloth, and an orange 91-cm by 122-cm card to the control condition and Set 2 foods (baked bean, green bean, pear, wheat bread), blue bowls, a blue table cloth, and a blue 91-cm by 122-cm card to the mitigation condition. The feeder started each corresponding phase

of the control and mitigation conditions and transitioned from Context A to Context B on the same day and from Context B to the renewal test on the same day. For example, the feeder conducted the last session of Context B in the control condition and the first session of Context A in the control condition on Tuesday. In addition, we transitioned from Context B to the renewal test in the same meal for John,

Hope, Diego, Jade, and Julian and in consecu- tive meals (e.g., Meal 3 and Meal 4) for Maisy and Emilia. The feeder followed the general procedure described above and the specific pro- cedure described below for the control and miti- gation conditions. We equated the number of sessions in each context across the control and mitigation conditions before changing from context to context to ensure that our manipula- tions, rather than unequal exposure to the experimental conditions, produced the observed effects. For example, if the number of sessions for Context A was 10 in the control condition, then the number of sessions for Context A in the mitigation condition was 10.

Context A Reinforcement (Control and Miti- gation Conditions) The caregivers of John, Maisy, Hope, Emi-

lia, Diego, and Jade and a therapist for Julian delivered functional reinforcement following inappropriate mealtime behavior in the clinic for John, Maisy, Hope, Emilia, and Diego, in the home for Jade, or in a simulated home for Julian. The therapist and observer were in the adjacent observation room during sessions for John, Maisy, Hope, Emilia, Diego; in the clinic observing via a secure video application for Jade; or in the simulated home for Julian. The functional reinforcers were escape and attention for John, Maisy, Hope, Emilia, and Diego; escape and attention during solids and escape during liquids for Jade; and escape for Julian.

Context B Extinction (Control Condition) A therapist conducted sessions with the par-

ticipant in a clinic room. The therapist and observer were in the clinic room during sessions in the control condition and during caregiver training. The therapist implemented escape and attention extinction for John, Maisy, Hope, Emilia, Diego and for Jade during solids and escape extinction for Jade during liquids and for Julian. During escape extinction, the feeder kept the utensil touching the lip until the

913Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

participant opened their mouth such that the feeder could deposit the bite or drink into the mouth or until the 10-min cap elapsed. The feeder used the utensil to scoop expelled bites or drinks and re-presented them into the par- ticipant’s mouth and blocked inappropriate mealtime behavior during presentations and re- presentations with their nonfeeding hand. Feeders maintained high levels of correct- utensil placement during extinction sessions, and no additional blocking from others was necessary. If the participant was expelling at the next scheduled presentation, the feeder re-pres- ented the bite until the participant kept the bolus in their mouth for 3 s and presented the next bite or drink. The feeder re-presented bites following expulsions that occurred during the mouth check for the fifth presentation until the participant had no food or liquid in the mouth or until the 10-min cap elapsed. The feeder started the next session if one was sched- uled. The feeder provided no differential atten- tion following inappropriate mealtime behavior. During the attention-extinction com- ponent, the feeder provided no differential con- sequence for inappropriate mealtime behavior. Caregiver Training (Control Condition).

Therapists used behavioral skills training to teach the caregivers of John, Maisy, Hope, Emilia, Diego, and Jade to implement the intervention before they implemented the inter- vention in Context A. First, the caregiver was in the observation room and watched the thera- pist implement the extinction intervention dur- ing the last three sessions of Context B. We counted the number of sessions required for the participant to meet criteria to change to the renewal test in the control and mitigation con- ditions to identify the last three sessions of Context B. If the participant met criteria to change from Context B to Context A in the mitigation condition first, we used the number of sessions from Context B in the mitigation condition to identify the last three sessions of Context B for the control condition. Second, a

therapist gave the caregiver a written protocol and explained the purpose of the extinction intervention. The caregiver read the interven- tion protocol, and the therapist answered ques- tions. Third, therapists trained the caregiver in a clinic room without the participant present. The therapist narrated and modeled an inter- vention component, and the caregiver role- played the modeled intervention component with another therapist as a confederate partici- pant. This pattern of therapist-narrate and model-caregiver role-play continued until the therapist had narrated and modeled and the caregiver had role-played every intervention component. Finally, the therapist narrated and modeled the entire extinction intervention, and the caregiver role-played the entire extinction intervention. Confederate children used a script that

ensured that they emitted every behavior the caregiver would encounter with their child so the caregiver could observe and practice responses to each possible participant behavior. We also used direct observations of the partici- pant’s behavior in the functional analysis and function-based baseline to program the fre- quency of the scripted confederate-child behav- iors during role-play. For example, if the participant frequently coughed and gagged, the script programmed the confederate child to cough and gag on multiple trials so the caregiver received additional practice for participant- specific behavior. The order of trained and practiced interven-

tion components was (a) consequences for acceptance and mouth clean, (b) prompts at the bite- or drink-presentation interval and the mouth check, and (c) the entire protocol. We used an additive approach during caregiver role-plays with the confederate child. The care- giver implemented the intervention compo- nents they previously mastered during trials with the confederate child when we were teach- ing the caregiver a new intervention compo- nent. For example, after the therapist trained

Sarah D. Haney et al.914

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

the caregiver to deliver consequences, the care- giver continued to deliver consequences while they were learning to deliver the prompts. Ther- apists provided behavior-specific praise for cor- rect protocol implementation and delivered corrective feedback for incorrect protocol imple- mentation during and after each trial. The mas- tery criterion for role-play was that the caregiver implemented the intervention with 80% or higher integrity for one 5-bite or drink session for consequences and prompts and three consec- utive 5-bite or drink sessions for the entire extinction intervention. After the caregiver met the role-play criterion while implementing the entire extinction intervention with the confeder- ate child, the caregiver implemented the extinc- tion intervention with their child during the renewal test in Context A (see below).

Context B Extinction (Mitigation Condition) A therapist fed the participant and

implemented the extinction intervention as described for the control condition. The thera- pist and observer were in the clinic room dur- ing caregiver training and during sessions in the mitigation condition. We also used the proce- dure described below to train the caregiver to implement the extinction intervention and to transition from the therapist to the caregiver as feeder for John, Maisy, Hope, Emilia, Diego, and Jade and to introduce Sets 1 and 3 foods for Diego. The caregiver observed the first three sessions of the extinction intervention from the adjacent observation room and reviewed the intervention protocol with a therapist before we started caregiver fading. Caregiver Training and Mitigation Proce-

dure (Mitigation Condition). We used a train- ing procedure like that described for the control condition except that after the caregiver demon- strated mastery with the confederate child, the therapist and caregiver conducted sessions with the participant. During these sessions, the therapist fed the participant and the caregiver implemented the mastered intervention component(s). The therapist

and caregiver conducted sessions with the partici- pant until the caregiver demonstrated mastery of the intervention component during those sessions with the participant. The therapist and caregiver then repeated the sequence of therapist narrate and model, caregiver role-play with confederate child, therapist and caregiver implement intervention component(s) with the participant until the care- giver demonstrated mastery implementing the entire extinction intervention with the participant with the therapist sitting next to the caregiver. The mastery criterion during sessions with the partici- pant was 80% or higher correct-utensil placement and correct procedure and 20% or lower incorrect attention for three consecutive five-bite or drink sessions for each intervention component. The next step was for the therapist and

observer to fade out of the room. The mastery criterion for therapist-position fading was 80% or higher acceptance, correct-utensil placement, and correct procedure; 20% or lower incorrect attention; and zero instances of inappropriate mealtime behavior for three consecutive five- bite or drink sessions. The therapist moved from sitting approximately 0.3 m from care- giver; to across the table, approximately 0.9 m from caregiver; to next to the clinic-room door, approximately 1.5 m from caregiver. The observer left the room when the therapist moved across the table. The therapist provided immediate corrective feedback for most care- giver errors and occasionally at the end of the session. For example, the therapist reminded the caregiver at the end of the session to present all four foods if the caregiver did not do so dur- ing the session. Therapists provided frequent positive feedback and used an informal method of fading the number of positive feedback state- ments as they faded out of the room. Therapists generally based the number of positive feedback statements on caregivers’ procedural integrity from the previous sessions. Therapists delivered gradually fewer positive feedback statements after sequential sessions with low or decreasing caregiver errors and more positive feedback

915Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

statements after sequential sessions with increased errors. For Diego, we used the same caregiver-

fading procedure described above and faded in Set 3 foods presented in Context A of the miti- gation condition into Context B of the mitiga- tion condition. We randomly selected to fade Diego’s mother in as feeder before fading in Set 3 foods. The therapist and Diego’s mother ini- tially presented Set 4 foods during caregiver fading. Next, Diego’s mother faded in one Set 3 food at a time. She presented one Set 3 food with three Set 4 foods until she had introduced all Set 3 foods after which she faded the ratio of Set 3 foods to Set 4 foods. Diego’s mother always presented one fruit, one protein, one starch, and one vegetable. The criterion for introducing a Set 3 food and fading the ratio of Set 3 foods to Set 4 foods was the same as the mastery criterion for therapist-position fading. We also randomized the bite number in which Diego’s mother introduced each Set 3 food to control for order effects. For example, she introduced tuna on Bite 3 of one session and avocado on Bite 5 of a different session. After Diego’s mother introduced a Set 3 food during one session, she presented the Set 3 food and the same three Set 4 foods in random order during two more sessions. Next, she presented two Set 3 foods and two Set 4 foods for one session, then three Set 3 foods and one Set 4 food for three sessions. For Jade, we used the same caregiver-fading

procedure described above. We also used con- text similarity for Jade and Julian by using a Crosstour video projector to display a picture of the participant’s view from their mealtime chair in the home or simulated home, respec- tively, onto the clinic room wall and placed items from the participant’s home (e.g., toys, lamps, paintings) in the clinic room. We pro- jected a 1.2-m x 2.4-m picture of the family’s kitchen approximately 1.2 m away from Jade on the wall in front of her, which showed a dining table, table centerpiece, clock, kitchen

island with stools, appliances (e.g., refrigerator, stove), and toys. We placed 16 items from Jade’s home including toys (e.g., puzzles, blocks, books) and a table centerpiece in the clinic room. We projected a 1.2-m x 3.4-m pic- ture of the simulated home approximately 0.9 m away from Julian on the wall to his right, which showed a table with a blue table- cloth, blue bowls and a framed picture of his home kitchen; lamp; bookshelf with storage bins, candles, framed picture of people; two paintings; an end table with flowers in a vase and candles; and kitchen appliances (e.g., refrigerator, sink, microwave). We could not project the picture in front of him because of the layout of the clinic room. We placed 18 items, including a lamp, paintings, candles, framed pictures, and an end table with decora- tions shown in the picture in the clinic room. For both participants, we positioned the items in the clinic room to mirror their arrangement in the projected picture.

Context A Extinction (Renewal Test, Control and Mitigation Conditions) The contingencies and contextual arrange-

ments during the renewal tests in Context A were identical in the control and mitigation conditions. The caregivers of John, Maisy, Hope, Emilia, Diego, and Jade and a therapist for Julian implemented the extinction interven- tion as described for the Context B control condition in the clinic with John, Maisy, Hope, Emilia, and Diego; in the home with Jade; and in the simulated home with Julian. Diego’s mother presented Set 1 foods and Set 3 foods during the control and mitigation con- ditions, respectively. We conducted at least five consecutive sessions of the renewal test in the control and mitigation conditions across partic- ipants. The therapist and observer were in the adjacent observation room for John, Maisy, Hope, Emilia, and Diego; in a clinic room observing over Vidyo for Jade, or in the simu- lated home for Julian. Therapists delivered

Sarah D. Haney et al.916

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

corrective feedback as necessary over the two- way audio system from the adjacent observa- tion room for John, Maisy, Hope, Emilia, and Diego and over Vidyo for Jade to ensure care- givers maintained high levels of integrity.

Social Validity We gave caregivers a 5-item questionnaire

and asked them to rate the acceptability of the control and mitigation training procedures. The first author calculated the mean acceptabil- ity rating for the conditions separately by sum- ming ratings across each condition and dividing that sum by the total number of completed questionnaires. Caregiver acceptability of the training procedures was slightly higher for the mitigation condition (4.6 out of 5) relative to the control condition (4.4 out of 5).

Results

The first author reviewed caregiver-integrity data from the first five sessions of the renewal test to ensure that changes in rates of inappro- priate mealtime behavior were not a function of low feeder integrity. During the renewal test in the control and mitigation conditions, mean correct-utensil placement and mean correct procedure was above 98%, and mean incorrect attention was below 1% across feeders. We identified renewal if rate of inappropriate meal- time behavior during any of the first five renewal-test sessions was above the rate of inap- propriate mealtime behavior during the last three sessions of Context B. We identified a mitigation effect in the mitigation condition if we identified renewal during the control condi- tion and if rate of inappropriate mealtime behavior during the first five renewal-test ses- sions was lower than rate of inappropriate mealtime behavior during the last three sessions of Context B. Figure 1 shows mean inappropriate mealtime

behavior per minute during the control (top) and mitigation (bottom) conditions for John

(top left), Maisy (top right), Hope (bottom left), and Emilia (bottom right). We evaluated renewal following a change in the feeder from therapist to caregiver and assessed caregiver fad- ing as our mitigation procedure. During Con- text A of the control condition in which the caregiver delivered function-based reinforce- ment, rate of inappropriate mealtime behavior was high and variable for John’s solids (M = 67; range, 43 to 86), Maisy’s liquids (M = 26; range, 24 to 29), Hope’s solids (M = 25; range, 13 to 40), and Emilia’s Set 1 foods (M = 42; range, 22 to 77). During Context A of the mit- igation condition in which the caregiver deliv- ered function-based reinforcement, rate of inappropriate mealtime behavior was high and variable for John’s liquids (M = 65; range, 23 to 132), Maisy’s solids (M = 25; range, 7 to 45), Hope’s liquids (M = 37; range, 21 to 52), and Emilia’s Set 2 foods (M = 34; range, 9 to 61). During Context B of the control condition in

which the therapist implemented extinction, rate of inappropriate mealtime behavior decreased and was variable for John’s solids (M = 20; range, 0 to 55) and was low and relatively stable for Maisy’s liquids (M = 0.5; range, 0 to 5), Hope’s solids (M = 0.3; range, 0 to 6), and Emilia’s Set 1 foods (M = 0.7; range, 0 to 7). During Context B of the mitigation condition in which the therapist implemented extinction initially and faded the caregiver in as the feeder, rate of inappropriate mealtime behavior decreased and was low and stable for John’s liq- uids (M = 3; range, 0 to 27), Maisy’s solids (M = 0.4; range, 0 to 6), Hope’s liquids (M = 0.3; range, 0 to 4), and Emilia’s Set 2 foods (M = 0.7; range, 0 to 6). During the renewal test of the control condi-

tion in which the caregiver implemented extinction, rate of inappropriate mealtime behavior increased and was low and variable before decreasing to zero for John’s solids (M = 9; range, 0 to 21) and Maisy’s liquids (M = 3; range, 0 to 9) and remained at zero for Hope’s

917Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

Figure 1 Inappropriate Mealtime Behavior Following a Caregiver Context Change

Note. The control condition is displayed on top and the mitigation condition is displayed on bottom for John (top left), Maisy (top right), Hope (bottom left), and Emilia (bottom right).

Sarah D. Haney et al.918

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

solids and Emilia’s Set 1 foods. During the renewal test of the mitigation condition in which the caregiver implemented extinction, rate of inappropriate mealtime behavior increased slightly for one session before decreas- ing to zero for John’s liquids (M = 0.4; range, 0 to 4) and Emilia’s Set 2 foods (M = 0.7; range, 0 to 4), was low and stable before decreasing to zero for Maisy’s solids (M = 0.4; range, 0 to 4), and remained at zero for Hope’s liquids. Overall, we observed renewal for John and Maisy only and observed higher and more persistent rates of responding during the renewal test of the control condition relative to the mitigation condition. Figure 2 shows mean inappropriate mealtime

behavior per minute during the control (top) and mitigation (bottom) conditions for Diego. We evaluated renewal following a change in the

feeder and food from a therapist feeding one set of foods to a caregiver feeding a different set of foods during intervention for solids refusal and assessed caregiver and food fading as our mitigation procedure. During Context A of the control condition with Set 1 foods in which the caregiver delivered function-based reinforce- ment, rate of inappropriate mealtime behavior was high and variable (M = 77; range, 32 to 103). During Context A of the mitigation con- dition with Set 3 foods in which the caregiver delivered function-based reinforcement, rate of inappropriate mealtime behavior was high and variable (M = 90; range, 62 to 115). During Context B of the control condition

with Set 2 foods in which the therapist implemented extinction, rate of inappropriate mealtime behavior was moderately high and var- iable then decreased and was low and stable (M = 4; range, 0 to 52). During Context B of the mitigation condition with Set 4 foods in which the therapist implemented extinction and care- giver and food fading, rate of inappropriate mealtime behavior was moderately high and var- iable then decreased to zero (M = 10; range, 0 to 60). During the renewal test in which the caregiver implemented extinction, rate of inap- propriate mealtime behavior remained at zero in the control condition with Set 1 foods and was low and stable in the mitigation condition with Set 3 foods (M = 1; range, 0 to 7). For Diego, we did not observe renewal in either condition. Figure 3 shows mean inappropriate mealtime

behavior per minute during the control (top) and mitigation (bottom) conditions for Jade. We evaluated renewal following a change in the feeder and setting from therapist in the clinic to caregiver in the home during intervention for liquids refusal (control condition) and solids refusal (mitigation condition). We assessed care- giver fading and context similarity as our mitiga- tion procedure. During Context A in which the caregiver delivered function-based reinforcement in the home, rate of inappropriate mealtime behavior was variable in the control condition

Figure 2 Inappropriate Mealtime Behavior Following a Caregiver and Food Context Change

5 10 15 20 25 30 35 40 45 50 55 60

0

20

40

60

80

100

120

Session

Caregiver Fading

Diego Set 3 & 4 Foods

Context A Reinforcement (Caregiver + Set 3 Foods)

Context A Exinction (Caregiver + Set 3 Foods)

Context B

Extinction + Caregiver and Food Fading (Therapist + Caregiver + Set 4 Foods)

Food Fading

D ay 1

D ay 2

D ay 3

D ay 4

D ay 5

5 10 15 20 25 30 35 40 45 50 55 60

0

20

40

60

80

100

120

Context A Reinforcement (Caregiver + Set 1 Foods)

Context A Exinction (Caregiver + Set 1 Foods)

Context B

Extinction (Therapist + Set 2 Foods)

Diego Set 1 & 2 Foods

D ay 1

D ay 2

D ay 3

D ay 4

D ay 5

In ap p ro p r i a t e M e a lt i m e B e h av io r P er M in u t e

Note. The control condition is displayed on top and the mitigation condition is displayed on bottom.

919Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

with liquids (M = 26; range, 5 to 66) and mitiga- tion condition with solids (M = 18; range, 5 to 58). During Context B of the control condition

with liquids in which the therapist implemented extinction in the clinic, rate of inappropriate mealtime behavior was low and variable then decreased to zero (M = 2; range, 0 to 20). During Context B of the mitigation condition with solids in which the therapist implemented extinction in the clinic, faded the caregiver in as the feeder, and used context sim- ilarity (i.e., arranged the clinic room with items from home), rate of inappropriate mealtime behavior was low and variable then stabilized at zero (M = 1; range 0 to 11). During the renewal test in which the caregiver implemented extinction in the home, rate of inappropriate mealtime behavior increased and was low and variable in the control condition

with liquids (M = 3; range, 0 to 6) and remained zero in the mitigation condition with solids. For Jade, we observed renewal in the control condition only. Figure 4 shows mean inappropriate mealtime

behavior per minute during the control (top) and mitigation (bottom) conditions for Julian. We evaluated renewal following a change in the setting from clinic to a simulated home and assessed context similarity as our mitigation procedure during intervention for solids refusal. During Context A in which the therapist deliv- ered function-based reinforcement in a simu- lated home, rate of inappropriate mealtime behavior was high and variable in the control condition with Set 1 foods (M = 52; range, 33 to 78) and mitigation condition with Set 2 foods (M = 60; range, 43 to 86). During Con- text B of the control condition with Set 1 foods in which the therapist implemented extinction

Figure 3 Inappropriate Mealtime Behavior Following a Caregiver and Setting Context Change

10 20 30 40 50 60 70 80 90 100

0

15

30

45

60

75

Context A Reinforcement (Caregiver +

Home)

Context A Exinction

(Caregiver + Home)

Context B

Extinction (Therapist + Clinic Room)

Jade Liquids

D ay

1

D ay

5D ay

4

D ay

2 D a y

3

D ay

6

D ay

7

D ay

8

10 20 30 40 50 60 70 80 90 100

0

15

30

45

60

75

Session

Context A Reinforcement (Caregiver +

Home)

Context A Exinction

(Caregiver + Home)

Context B Extinction + Caregiver Fading and

Context Similarity (Therapist + Caregiver + Clinic Room)

Caregiver Fading

Jade Solids

D ay

6

D ay

1

D ay

7D ay

2

D ay

3

D ay

3 D ay

5

In ap p ro p ri at e M ea lt im

e B eh av io r P er M in u te

Note. The control condition is displayed on top and the mitigation condition is displayed on bottom.

Figure 4 Inappropriate Mealtime Behavior Following a Setting Con- text Change

10 20 30 40 50 60 70 80 90 100

0

10

20

30

40

50

60

70

80

90

Context A Reinforcement (Simulated Home)

Context A Exinction (Simulated Home)

Context B

Extinction (Clinic Room)

Julian Set 1 FoodsD

ay 1

D ay 4

D ay 2

D ay 3

D ay 5

D ay 6

D ay 7

D ay 8

D ay 9

D ay 1 0

D ay 1 1

D ay 1 2

D ay 1 3

D ay 1 4

D ay 1 5

D ay 1 6

D ay 1 7

10 20 30 40 50 60 70 80 90 100

0

10

20

30

40

50

60

70

80

90

Session

Context A Reinforcement (Simulated Home)

Context A Exinction (Simulated Home)

Context B

Extinction + Context Similarity (Clinic Room)

Julian Set 2 Foods

D ay 1 D ay 2

D ay 3

D ay 4

D ay 5

D ay 6

D ay 7

D ay 8

D ay 9

D ay 1 0

D ay 1 1

D ay 1 2

D ay 1 3

D ay 1 4

D ay 1 5

D a y 1 6

D ay 1 7

In ap p ro p ri at e M ea l t im e B eh av io r P er M i n u t e

Note. The control condition is displayed on top and the mitigation condition is displayed on bottom.

Sarah D. Haney et al.920

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

in a clinic room, rate of inappropriate mealtime behavior decreased and was low and variable (M = 5; range, 0 to 26). During Context B of the mitigation condi-

tion with Set 2 foods in which the therapist implemented extinction in a clinic room and used context similarity, rate of inappropriate mealtime behavior decreased and was low and variable (M = 4; range, 0 to 14). During the renewal test in which the therapist implemented extinction in the simulated home, rate of inap- propriate mealtime behavior remained low and variable in the control condition with Set 1 foods (M = 7; range, 0 to 13) and remained zero in the mitigation condition with Set 2 foods. For Julian, we observed renewal in the control condition only. Figure 5 depicts the proportion of baseline

rate of inappropriate mealtime behavior during

the renewal test in the control and mitigation conditions. The first author divided the mean rate of inappropriate mealtime behavior for the first five sessions of the renewal test by the mean rate of inappropriate mealtime behavior for the entire phase of reinforcement in Context A to calculate proportion of baseline rate. We used the first five sessions of the renewal test because five was the fewest number of sessions we con- ducted for the renewal test across the control and mitigation conditions for all participants. The proportion of baseline rate was between 0 and 0.4 during the control condition and between 0 and 0.2 during the mitigation condi- tion across participants, showing that inappro- priate mealtime behavior was generally higher during the renewal test of the control condition relative to the mitigation condition across participants.

Discussion

The current study extended the findings of Ibañez et al. (2019) by evaluating renewal of inap- propriate mealtime behavior following changes in feeder, food, setting, and a combination of these. We observed renewal of inappropriate mealtime behavior following a change from therapist to caregiver as feeder for two of four participants, John and Maisy, from therapist in clinic to care- giver in the home with one participant, Jade, and from clinic to simulated home for one participant, Julian. These effects occurred even though feeders implemented the extinction intervention with high levels of integrity. These results differ from those of Ibañez et al., who observed renewal of inappropriate mealtime behavior in all three participants. The current study extended the findings of

Kelley et al. (2018) in three ways. First, we eval- uated renewal and renewal-mitigation concur- rently instead of sequentially because repeated exposure to context changes may decrease the likelihood of renewal (Sweeney & Shahan, 2013; Wacker et al., 2011). Second, we

Figure 5 Proportion of Baseline Rate of Inappropriate Mealtime Behavior Across Participants

1 2 3 4 5

0.0

0.1

0.2

0.3

0.4

Mitigation Condition

Sessions of Renewal Test

John

Maisy

Hope

Emila

Julian

Jade

Diego

1 2 3 4 5

0.0

0.1

0.2

0.3

0.4

Control Condition

Jade

John

Maisy

Hope

Emila

Julian

Diego

P ro p o rt io n o f B as el in e

Note. Proportion of baseline rate of inappropriate meal- time behavior during the first five sessions of the renewal test in the control (top) and mitigation (bottom) condi- tions across participants.

921Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

evaluated systematic fading as a renewal- mitigation procedure following a change in feeder from therapist to caregiver. We used behavioral skills training to teach caregivers the intervention. Caregivers practiced an intervention component (e.g., consequences) with a confederate participant before implementing the component with the participant. We continued this procedure (i.e., model, practice with confederate participant, implement with participant) until the caregiver implemented the entire intervention with the par- ticipant. Caregivers may signal reinforcement availability due to their history of providing rein- forcement for inappropriate mealtime behavior (Borrero et al., 2010; Najdowski et al., 2008; Piazza et al., 2003). Introducing the caregiver as feeder is likely an essential context change during intervention; therefore, evaluating renewal- mitigation procedures when the caregiver imple- ments the intervention is important. We gradually increased the number of intervention components the caregiver implemented (i.e., consequences, consequences + praise, entire protocol) to increase opportunities for the caregiver to develop stimulus control for extinction of inappropriate mealtime behavior. Rate of inappropriate mealtime behavior was lower for John and Maisy and persisted across fewer sessions for John in the mitigation condition following the mitigation procedure relative to the control condition. These results suggest that sys- tematic caregiver fading decreased the magnitude and persistence of renewal but may not have prevented it. Interestingly, John’s inappropriate mealtime

behavior decreased more rapidly and maintained at lower rates in Context B of the mitigation condition relative to the control condition. Therefore, we wondered whether the differential efficacy of extinction, mitigation procedure, or both influenced renewal because responding during the renewal test was lower and less per- sistent in the mitigation condition relative to the control condition. By contrast, Maisy’s results suggest that the mitigation procedure was responsible for the difference in renewal

effects between the control and mitigation con- ditions. Caregivers in Kelley et al. (2018) sat next to the therapist while the therapist implemented extinction and only implemented extinction during the renewal test. Kelley et al. observed renewal for one of two participants. However, the magnitude and persistence of renewal decreased following the mitigation pro- cedure for the other participant, as in the cur- rent study. Third, we evaluated the efficacy of context

similarity and the combination of systematic fading and context similarity following changes in feeder and food, feeder and setting, and set- ting. These are common context changes for children we treat with avoidant/restrictive food intake disorder and that caregivers report as challenging. We did not observe renewal fol- lowing a feeder and food change for Diego but did observe renewal following a feeder and set- ting change for Jade and a setting change for Julian. For Jade, we evaluated renewal following

changes in the feeder and setting because Jade engaged in inappropriate mealtime behavior dur- ing sessions in the home but not in the clinic. Therefore, the home was the relevant stimulus condition under which inappropriate mealtime behavior occurred. A combined feeder and setting change is common during intervention for inap- propriate mealtime behavior because children feed in different environments (e.g., home, school) with many individuals (e.g., grandparents, peers). Jade’s results replicate those of Ibañez et al. (2019) and Saini et al. (2018) and demonstrate that changes in the implementer and setting produce renewal of negatively reinforced socially significant behavior. The absence of renewal during the renewal test of the mitigation condition, which was systematic fading and context similarity, replicate those of previous studies on context similarity as an effec- tive mitigation procedure (Bandarian-Balooch & Neumann, 2011; Todd et al., 2012). Results also suggest that renewal-mitigation procedures may be more effective when combined (Bandarian-

Sarah D. Haney et al.922

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

Balooch & Neumann, 2011; Bernal-Gamboa, Nieto, & Uengoer, 2017; Krisch et al., 2018); however, research on renewal mitigation of socially significant behavior is limited. For Julian, we evaluated renewal following a

setting change because Julian’s caregivers could not attend required sessions due to their work schedules. As in Mitteer et al. (2018), who showed renewal of undesirable caregiver behav- ior following a setting change, renewal of inap- propriate mealtime behavior occurred during the renewal test of the control condition for Julian. Julian’s results and those of Mitteer et al. suggest that changes in the setting may evoke renewal of negatively reinforced socially significant behavior. Interestingly, the most robust renewal effects

occurred for Jade and Julian. The commonality between their arrangements was a setting change; therefore, we wondered whether the setting exerts more control over inappropriate mealtime behavior than does the caregiver. Results of our study and previous research show that a stimulus change such as in imple- menter or setting may exert more control over behavior than the contingency (e.g., Ibañez et al., 2019; Saini et al., 2018). Before inter- vention, the occurrence of inappropriate meal- time behavior may have discouraged caregivers from feeding their child outside of the home. Thus, reinforcement for inappropriate meal- time behavior likely occurs primarily in the home. Participants in our study had a relatively short history of reinforcement and a more recent history of extinction of inappropriate mealtime behavior in the clinic. Therefore, the clinic may have signaled extinction of inappro- priate mealtime behavior, which competed with the participant’s history of reinforcement with the caregiver. Alternatively, there may be more stimuli in the home that signal reinforcement availability (e.g., kitchen chair, table) relative to the caregiver as the only signal for reinforce- ment availability in the clinic. However, the current study did not assess to which stimuli

participants attended or which stimuli exerted control over inappropriate mealtime behavior and could be a direction for future research. Results of previous research have suggested

that behavior that is more resistant to extinc- tion initially may be more likely to reemerge following context changes (Podlesnik & DeLeon, 2015). When renewal occurred in the control condition, the number of sessions until the rate of inappropriate mealtime behavior was zero during Context B was greater in the con- trol condition relative to the mitigation condi- tion for John, Maisy, Jade, and Julian. This difference was small for Maisy (i.e., two ses- sions). Researchers have also suggested that behavior that is more persistent during extinc- tion may be more likely to relapse following context changes (Kimball et al., 2018; Podlesnik et al., 2017). In the current study, inappropriate mealtime behavior occurred dur- ing more sessions of extinction in Context B in the control condition relative to the mitigation condition for John, Maisy, Jade, and Julian. This difference was small for Maisy (i.e., three sessions). When we observed renewal of inappropriate

mealtime behavior during the renewal test of the control condition for John, Maisy, and Jade, we observed concomitant decrements in per- centage of acceptance. Researchers have observed similar disruption of intervention effects between responses targeted for extinction and responses targeted for acquisition during ABA renewal arrangements with other socially significant behavior (Ibañez et al., 2019; Kelley et al., 2018; Saini et al., 2018). These results suggest that context changes may produce renewal of a previously extinguished response and decrements in the alternative appropriate response, which is an important consideration for evaluating procedures to enhance the gener- ality of an intervention. We evaluated proportion of baseline rate of

inappropriate mealtime behavior during the renewal test in the control and mitigation

923Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

conditions to quantify renewal and compare our results across participants and with previous research. Proportion of baseline values near 1.0 demonstrate closer approximations to rate of inappropriate mealtime behavior during rein- forcement in Context A. Proportion of baseline rate of inappropriate mealtime behavior during the renewal test was greater during the control condition relative to the mitigation condition across participants except Maisy. We compared our proportion of baseline data to that of Ibañez et al. (2019) and Saini et al. (2018) who evaluated renewal of negatively reinforced socially significant behavior. Our proportion of baseline values were generally lower than those of Ibañez et al. and Saini et al. One limitation of the current study was the

potential for carry-over effects from alternating between the control and mitigation conditions. We tried to increase the discriminability between the conditions by assigning solids or liquids to each condition for John, Maisy, Hope, and Jade and using colored stimuli (i.e., tablecloths, bowls, cards) for Emilia, Diego, and Julian. However, the caregiver’s presence in the mitigation condi- tion may have influenced responding in the con- trol condition and decreased the likelihood of renewal for participants with whom we changed the feeder. Additionally, the conditions may not have been discriminable to Emilia and Diego because feeders presented pureed foods in both conditions, which are not as visually distinctive as the same foods at table texture. For example, pureed potato and cauliflower look identical. Sim- ilarly, the history of contacting reinforcement and extinction in the presence of the same colored stimuli during the control and mitigation condi- tions may have decreased the discriminability between phases in each condition and influenced the likelihood of renewal (Saini et al., 2018). Alternatively, order effects may have influenced responding during the renewal test. We random- ized the order of the renewal tests across partici- pants, but exposure to the renewal test in one condition may have influenced responding during

the other condition. We transitioned from Con- text B to the renewal test in the same meal for John, Hope, Diego, Jade, and Julian and in con- secutive meals (e.g., Context B in last session of Meal 3 and renewal test in first session of Meal 4) for Maisy and Emilia. Research has suggested that time alone may function as a context change (Bouton & García-Gutiérrez, 2006); therefore, the 40-min break between meals may have influenced our results. However, we transitioned from Context B to the renewal test in the same manner (i.e., same meal or consecutive meal) in the control and mitigation conditions for each participant and observed differential responding within and across participants. For example, we observed renewal in the control condition only for Jade and Julian and did not observe renewal in either condition for Hope and Diego despite transitioning from Context B to the renewal test in the same meal. Further, some of our mitigation procedures included multiple components (e.g., caregiver and therapist’s position fading for caregiver fading procedure) and therefore, it is not possible to determine which component(s) were responsible for our outcomes. The results of the current study suggest that

renewal may occur during intervention for inap- propriate mealtime behavior for children with avoidant/restrictive food intake disorder, partic- ularly following changes in the feeder, setting, and a combination of these changes. This is the first study to our knowledge to evaluate system- atic fading and context similarity alone and in combination as renewal-mitigation procedures for socially significant behavior. Our results showed that systematic fading alone decreased the magnitude and persistence of renewal fol- lowing a feeder change and prevented renewal following a feeder and setting change when we used the procedure with context similarity. These findings highlight the importance of using a fading approach to caregiver training during intervention for feeding disorders. A systematic-fading training approach may pro- vide more opportunities for the caregiver to be

Sarah D. Haney et al.924

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

associated with the extinction contingency and reduce the likelihood of generalization failures when the caregiver implements the intervention alone. Training caregivers to implement the intervention while the child is still engaging in problem behavior may provide them with opportunities to practice implementing the intervention and reduce the likelihood that the caregiver will reinforce problem behavior out- side of the clinic setting (e.g., home). Context similarity alone effectively prevented renewal following a setting change, suggesting that pro- gramming common stimuli may prevent gener- alization failures to settings outside of the clinic and enhance intervention generality. Renewal- mitigation procedures such as systematic fading and context similarity alone and in combination may be effective at preventing or decreasing the magnitude of renewal, but we need to further understand the conditions under which renewal of inappropriate mealtime behavior occurs and the conditions under which renewal-mitigation procedures will be efficacious. These findings are important because these context changes are often necessary for intervention generality and maintenance and are changes that caregivers report as most challenging, particularly during intervention for children with avoidant/restric- tive food intake disorder.

REFERENCES

American Psychiatric Association (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Author.

Bachmeyer, M. H., Kirkwood, C. A., Criscito, A. B., Mauzy, C. R., IV, & Berth, D. P. (2019). A compar- ison of functional analysis methods of inappropriate mealtime behavior. Journal of Applied Behavior Analysis, 52(3), 603-621. https://doi.org/10.1002/ jaba.556

Bachmeyer, M. H., Piazza, C. C., Fredrick, L. D., Reed, G. K., Rivas, K. D., & Kadey, H. J. (2009). Functional analysis and treatment of multiply con- trolled inappropriate mealtime behavior. Journal of Applied Behavior Analysis, 42(1), 641–658. https:// doi.org/10.1901/jaba.2009.42-641

Bandarian-Balooch, S. B., & Neumann, D. L. (2011). Effects of multiple contexts and context similarity on the renewal of extinguished conditioned behavior in an ABA design with humans. Learning and Motiva- tion, 42(1), 53-63. https://doi.org/10.1016/j.brat. 2012.06.003

Bernal-Gamboa, R., Gámez, A. M., & Nieto, J. (2017). Reducing spontaneous recovery and reinstatement of operant performance through extinction-cues. Behav- ioural Processes, 135(2), 1-7. https://doi.org/10.1016/ j.beproc.2016.11.010

Bernal-Gamboa, R., Nieto, J., & Uengoer, M. (2017). Effects of extinction in multiple contexts on renewal of instrumental responses. Behavioural Processes, 142 (9), 64-69. https://doi.org/10.1016/j.beproc.2017. 06.003

Borrero, C. S. W., Woods, J. N., Borrero, J. C., Masler, E. A., & Lesser, A. D. (2010). Descriptive analysis of pediatric food refusal and acceptance. Jour- nal of Applied Behavior Analysis, 43(1), 71-88. https://doi.org/10.1901/jaba.2010.43-71

Bouton, M. E., & García-Gutiérrez, A. (2006). Intertrial interval as a contextual stimulus. Behavioural Processes, 71(2-3), 307–317. https://doi.org/10.1016/j.beproc. 2005.12.003

Bouton, M. E., Todd, T. P., Vurbic, D., & Winterbauer, N. E. (2011). Renewal after the extinc- tion of free operant behavior. Learning & Behavior, 39(1), 57-67. https://doi.org/10.3758/s13420-011- 0018-6

Bryant-Waugh, R., Markham, L., Kreipe, R. E., & Walsh, B. T. (2010). Feeding and eating disorders in childhood. International Journal of Eating Disorders, 43(2), 98-111. https://doi.org/10.1002/eat.20795

Bullock, C. E., Fisher, W. W., & Hagopian, L. P. (2017). Description and validation of a computerized behavioral data program: “BDataPro.” The Behavior Analyst, 40(1), 275-285. https://doi.org/10.1007/ s40614-016-0079-0

Centers for Disease Control and Prevention, National Center for Health Statistics (2010). Clinical growth charts. Retrieved from: https://www.cdc.gov/ growthcharts/clinical_charts.htm.

Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., Owens, J. C., & Slevin, I. (1992). A comparison of two approaches for identifying rein- forcers for persons with severe and profound disabil- ities. Journal of Applied Behavior Analysis, 25(2), 491- 498. https://doi.org/10.1901/jaba.1992.25-491

Freedman, D. S., Dietz, W. H., Srinivasan, S. R., & Berenson, G. S. (1999). The relation of overweight to cardiovascular risk factors among children and ado- lescents: The Bogalusa Heart Study. Pediatrics, 103 (6), 1175–1182. https://doi.org/10.1542/peds.103.6. 1175

Girolami, P. A., & Scotti, J. R. (2001). Use of analog functional analysis in assessing the function of meal- time behavior problems. Education and Training in

925Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

Mental Retardation and Developmental Disabilities, 36 (2), 207-223. https://www.jstor.org/stable/23879736

Greer, A. J., Gulotta, C. S., Masler, E. A., & Laud, R. B. (2007). Caregiver stress and outcomes of children with pediatric feeding disorders treated in an inten- sive interdisciplinary program. Journal of Pediatric Psychology, 33(6), 612–620. https://doi.org/10.1093/ jpepsy/jsm116.

Ibañez, V., Piazza, C., & Peterson, K. (2019). A transla- tional evaluation of renewal of inappropriate meal- time behavior. Journal of Applied Behavior Analysis, 52(4), 1005-1020. https://doi.org/10.1002/jaba.647

Kelley, M. E., Jimenez-Gomez, C., Podlesnik, C. A., & Morgan, A. (2018). Evaluation of renewal mitigation of negatively reinforced socially significant operant behavior. Learning and Motivation, 63(13), 133-441. https://doi.org/10.1016/j.lmot.2018.05.003

Kelley, M. E., Liddon, C. J., Ribeiro, A., Greif, A. E., & Podlesnik, C. A. (2015). Basic and translational eval- uation of renewal of operant responding. Journal of Applied Behavior Analysis, 48(2), 390-401. https:// doi.org/10.1002/jaba.209

Kimball, R. T., Kelley, M. E., Podlesnik, C. A., Forton, A., & Hinkle, B. (2018). Resurgence with and without an alternative response. Journal of Applied Behavior Analysis, 51(4), 854-865. https:// doi.org/10.1002/jaba.466

Kreipe, R. E., & Palomaki, A. (2012). Beyond picky eat- ing: Avoidant restrictive food intake disorder. Current Psychiatry Reports, 14(4), 421-431. https://doi.org/10. 1007/s11920-012-0293-8

Krisch, K. A., Bandarian-Balooch, S., & Neumann, D. L. (2018). Effects of extended extinction and multiple extinction contexts on ABA renewal. Learning and Motivation, 63(1), 1-10. https://doi.org/10.1016/j. lmot.2017.11.001

Mitteer, D. R., Greer, B. D., Fisher, W. W., Briggs, A. M., & Wacker, D. P. (2018). A laboratory model for evaluating relapse of undesirable caregiver behavior. Journal of the Experimental Analysis of Behavior, 110(2), 252-266. https://doi.org/10.1002/ jeab.462

Najdowski, A. C., Wallace, M. D., Penrod, B., Tarbox, J., Reagon, K., & Higbee, T. S. (2008). Caregiver-conducted experimental functional analyses of inappropriate mealtime behavior. Journal of Applied Behavior Analysis, 41(3), 459-465. https:// doi.org/10.1901/jaba.2008.41-459

Nieto, J., Uengoer, M., & Bernal-Gamboa. (2017). A reminder of extinction reduces relapse in an animal model of voluntary behavior. Learning & Memory, 24, 76-80. https://doi.org/10.1101/lm.044495.116

Patel, M. R., Piazza, C. C., Santana, C. M., & Volkert, V. M. (2002). An evaluation of food type and texture in the treatment of a feeding problem. Journal of Applied Behavior Analysis, 35(2), 183-186. https://doi.org/10.1901/jaba.2002.35-183

Peterson, K. M., Ibañez, V. F., Kirkwood, C. A., Crowley, J. G., & Piazza, C. C. (2018). Assessment of pediatric feeding disorders. In J. Matson (Ed.), Handbook of childhood psychopathology and develop- mental disabilities assessment (pp. 415-431). Springer.

Piazza, C. C., Fisher, W. W., Brown, K. A., Shore, B. A., Patel, M. R., Katz, R. M., & Blakely- Smith, A. (2003). Functional analysis of inappropriate mealtime behaviors. Journal of Applied Behavior Analysis, 36(2), 187-204. https://doi.org/10.1901/jaba.2003.36-187

Podlesnik, C. A., & DeLeon, I. G. (2015). Behavioral momentum theory: Understanding persistence and improving treatment. In F. D. DiGennaro Reed & D. D. Reed (Eds.), Bridging the gap between science and practice in autism service delivery (pp. 327- 351). Springer.

Podlesnik, C. A., Kelley, M., E., Jimenez-Gomez, C., & Bouton, M. E. (2017). Renewed behavior produced by context change and its implications for treatment maintenance: A review. Journal of Applied Behavior Analysis, 50(3), 675-697. https://doi.org/10.1002/ jaba.400

Saini, V., Sullivan, W. E., Baxter, E. L., DeRosa, N. M., & Roane, H. S. (2018). Renewal during functional communication training. Journal of Applied Behavior Analysis, 51(3), 603-619. https:// doi.org/10.1002/jaba.471

Sweeney, M. M., & Shahan, T. A. (2013). Behavioral momentum and resurgence: Effects of time in extinc- tion and repeated resurgence tests. Learning & Behav- ior, 41(4), 414-424. https://doi.org/10.3758/s13420- 013-0116-8

Tarquinio, D. C., Jones, M. C., Jones, K. L., & Bird, L. M. (2012). Growth charts for 22q11 deletion syndrome. American Journal of Medical Genetics, 158A(11), 2672- 2681. https://doi.org/10.1002/ajmg.a.35485

Todd, T. P. (2013). Mechanisms of renewal after the extinction of instrumental behavior. Journal of the Experimental Analysis of Behavior, 39(3), 193-207. https://doi.org/10.1037/a0032236

Todd, T. P., Winterbauer, N. E., & Bouton, M. E. (2012). Effects of the amount of acquisition and con- textual generalization on the renewal of instrumental behavior after extinction. Learning & Behavior, 40(2), 145-157. https://doi.org/10.3758/s13420-011-0051-5

Trask, S., Thrailkill, E. A., & Bouton, M. E. (2017). Occasion setting, inhibition, and the contextual con- trol of extinction in Pavlovian and instrumental (operant) learning. Behavioural Processes, 137(7), 64- 72. https://doi.org/10.1016/j.beproc.2016.10.003

Volkert, V. M., & Piazza, C. C. (2012). Pediatric feeding disorders. In P. Sturmey & M. Hersen (Eds.), Hand- book of evidence-based practice in clinical psychology, Vol 1: Child and adolescent disorders (pp. 323-337). John Wiley & Sons, Inc. https://doi.org/10.1002/ 9781118156391.ebcp001013

Wacker, D. P., Harding, J. W., Berg, W. K., Lee, J. F., Schieltz, K. M., Padilla, Y. C., Nevin, J. A., &

Sarah D. Haney et al.926

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense

Shahan, T. A. (2011). An evaluation of persistence of treatment effects during long-term treatment of destructive behavior. Journal of the Experimental Anal- ysis of Behavior, 96(2), 261-282. https://doi.org/10. 1901/jeab.2011.96-261

Wathen, S. N., & Podlesnik, C. A. (2018). Laboratory models of treatment relapse and mitigation

techniques. Behavior Analysis: Research and Practice, 18(4), 362-387. https://doi.org/10.1037/bar0000119

Received May 1, 2020 Final acceptance December 14, 2020 Action Editor, Carrie Borrero

927Renewal-Mitigation of Mealtime Behavior

19383703, 2021, 3, D ow

nloaded from https://onlinelibrary.w

iley.com /doi/10.1002/jaba.815 by B

ehavior A nalyst C

ertification B oard, W

iley O nline L

ibrary on [13/01/2023]. See the T erm

s and C onditions (https://onlinelibrary.w

iley.com /term

s-and-conditions) on W iley O

nline L ibrary for rules of use; O

A articles are governed by the applicable C

reative C om

m ons L

icense