A COMPARISON OF SENSORY INTEGRATIVE AND BEHAVIORAL ... · a comparison of sensory integrative and behavioral therapies as treatment for pediatric feeding disorders laura r. addison

A COMPARISON OF SENSORY INTEGRATIVE AND BEHAVIORALTHERAPIES AS TREATMENT FOR PEDIATRIC FEEDING DISORDERS

LAURA R. ADDISON

ADDISON BEHAVIORAL RESOURCES, CAMARILLO, CALIFORNIA

CATHLEEN C. PIAZZA

UNIVERSITY OF NEBRASKA MEDICAL CENTER’S MUNROE-MEYER INSTITUTE

MEETA R. PATEL

CLINIC 4 KIDZ, SAUSALITO, CALIFORNIA

MELANIE H. BACHMEYER

KENNEDY KRIEGER INSTITUTE

KRISTI M. RIVAS AND SUZANNE M. MILNES

UNIVERSITY OF NEBRASKA MEDICAL CENTER’S MUNROE-MEYER INSTITUTE

AND

JACKIE ODDO

THE ATLANTA SPEECH SCHOOL, ATLANTA, GEORGIA

We compared the effects of escape extinction (EE) plus noncontingent reinforcement (NCR)with sensory integration therapy as treatment for the feeding problems of 2 children. Resultsindicated that EE plus NCR was more effective in increasing acceptance, decreasinginappropriate behavior, and increasing amount consumed relative to sensory integration forboth children. The results are discussed in terms of the challenges of evaluating sensory-integration-based treatments, and the reasons why component analyses of multicomponenttreatments like sensory integration are important.

Key words: escape extinction, food refusal, food selectivity, feeding disorders, noncontingentreinforcement, pediatric feeding disorders, sensory integration

The term pediatric feeding disorders describesa heterogeneous group of behaviors that may becharacterized by inadequate oral intake, foodrefusal, selectivity by type or texture of food,and expelling or packing (pocketing) food, toname a few. Investigators have proposed anumber of hypotheses to explain the etiology offeeding disorders. One hypothesis is derivedfrom a neurodevelopmental theory proposed byAyres (1972, 1974) known as sensory integrationtheory. Although a complete review of sensoryintegration theory is beyond the scope of this

paper, some of the main tenets that are relevantto the current study are as follows (Kimball,1999). The central nervous system (CNS) has ahierarchical organization such that sensoryinput must be organized by the lower brainfor processing in higher levels of the brain tooccur. All of the sensory systems (i.e., auditory,visual, vestibular, proprioceptive, tactile) inter-act to receive and organize information. Onceorganized, the sensory information can beprocessed by the CNS to produce end-productskills and abilities (e.g., concentration; Kimball,1999). Rather than viewing behavior as deter-mined by specific environmental influences(e.g., antecedents such as the absence of adult

Address correspondence to Cathleen C. Piazza (e-mail: [email protected]).

doi: 10.1901/jaba.2012.45-455

455

JOURNAL OF APPLIED BEHAVIOR ANALYSIS NUMBER 3 (FALL 2012)2012, 45, 455–471

attention), the sensory integrative frameworkproposes that behavior should be analyzed inthe context of the antecedent effects of sensorysystem modulation (Kimball, 1999) or how thebrain integrates sensory information to produceend-product behavior. That is, identification ofa specific antecedent is not necessarily impor-tant (e.g., noisy cafeteria) because a child with asensory integration dysfunction will not be ableto process a variety of sensory stimuli (Kimball,1999). Of interest to behavior analysts is thatthe sensory integration framework ‘‘shift[s]from viewing behavior as learned patterns toseeing symptoms as a whole and reflective ofspecific actions of the CNS that misidentifiesnonnoxious environmental stimuli as irritatingor even dangerous’’ (Kimball et al., 2007, p. 3).

The translation of sensory integration theoryinto practice is that difficulty modulatingsensory information can result in sensorydefensiveness, which will be indicated by anindividual’s under- or overresponsiveness tosensory stimuli (Wilbarger, 1984). There are avariety of overt behaviors that are associatedwith sensory defensiveness, and feeding prob-lems are one example (Cermak, 2001). Sensory-based feeding problems may be diagnosed whena child demonstrates atypical responses tostimulation in and around the mouth such ascoughing, gagging, spitting out, or refusingfoods (Case-Smith & Humphry, 2005).

When a sensory-based feeding problem isidentified, sensory integration therapy would bethe treatment of choice (Case-Smith &Humphry, 2005; Cermak, 2001; Royeen &Lane, 1991; Tarbell & Allaire, 2002; William-son & Anzalone, 1997). From a sensoryintegration perspective, the inappropriate feed-ing behavior is a symptom of the child’sinability to process sensory information tomake an adaptive response (Kimball, 1999).Intervention targets the underlying sensoryprocessing deficits rather than the specificbehaviors (Kimball, 1999). The goal of therapyis to promote sensory modulation, whichshould result in decreased sensory defensiveness.

No empirical studies have described thecomponents of a sensory integration frameworkfor the treatment of feeding problems; however,recommendations for treatment of sensory-based feeding problems appear in textbooksand review articles (e.g., Case-Smith &Humphry, 2005; Cermak, 2001; Cermak &Mitchell, 2006; Lane, 2008; Royeen & Lane,1991; Williamson & Anzalone, 1997). Accord-ing to these sources, treatment involves mod-ification of the child’s sensory diet, which is thesensory input needed by an individual toorganize sensory information effectively (Kim-ball, 1999). Examples of a sensory diet for achild with feeding problems might includerhythm and music activities, proprioceptiveactivities, heavy work, and sensory modulationtechniques (Cermak, 2001; Roley & Schneck,2001). Koomar and Bundy (2002) note that itis not necessary to apply sensory input to theentire body; input to one system will affectother systems. They indicate that their clientsusually prefer sensory input to the arms, legs,and back rather than to the face. Pruzansky andFarrington (1999) note that adaptive respond-ing will be more likely if intervention involvesseveral sensory systems and requires intersenso-ry integration. Although sensory integrationoften is conducted by an occupational therapist(OT), the therapist also may serve as aconsultant to parents and other professionalsby developing a sensory diet that can beimplemented with the child throughout theday (Wilbarger & Wilbarger, 1991).

Sensory integration appears to be a widelyrecommended therapy for feeding problems(Baranek, 2002; Lane, 2008; Mailloux &Roley, 2004; D. Parham & Mailloux, 2005;Schaaf & Miller, 2005; Yack, Sutton, &Aquilla, 2002). Studies on occupational therapypractice suggest that sensory integration is thetherapeutic strategy used most frequently byOTs (Brown, Rodger, Brown, & Roever, 2005;Case-Smith & Miller, 1999; Storch & Eskow,1996; Watling, Deitz, Kanny, & McLaughlin,1999), who are often the first-line clinicians for

456 LAURA R. ADDISON et al.

children with feeding disorders (Caretto, Top-olski, Linkous, Lowman, & Murphy, 2000).

An alternative hypothesis for the etiology offeeding disorders is that inappropriate mealtimebehavior is maintained, at least in part, byenvironmental events. This hypothesis suggeststhat feeding disorders develop as a result of thepairing of eating with aversive events such asgagging, choking, or vomiting (e.g., Di Lorenzoet al., 2005). Children who have these aversiveexperiences may exhibit refusal behavior (e.g.,crying, batting at the spoon) at mealtime, andparental responses such as providing escapefrom spoon presentations may function asreinforcement for refusal behavior (Borrero,Woods, Borrero, Masler, & Lesser, 2010;Girolami & Scotti, 2001; Najdowski et al.,2008; Piazza, Fisher, et al., 2003). In fact,Borrero et al. (2010) showed that parents ofchildren with feeding problems frequentlyremoved the spoon or cup or terminated themeal following refusal behavior.

Reviews of the literature by Kerwin (1999)and Volkert and Piazza (2012) concluded thatprocedures based on the principles of operantconditioning are the only interventions forfeeding problems with empirical support. Morespecifically, results of numerous studies haveshown that escape extinction (EE) is effective astreatment (Ahearn, Kerwin, Eicher, Shantz, &Swearingin, 1996; Bachmeyer et al., 2009;Cooper et al., 1995; Piazza, Patel, Gulotta,Sevin, & Layer, 2003; Reed et al., 2004).Conceptually, EE involves no longer providingescape for inappropriate mealtime behavior.One of the most commonly used EE proce-dures to treat feeding problems is nonremovalof the spoon, in which the spoon remains at thechild’s lips until the feeder can deposit the biteinto the child’s mouth. For example, Reed et al.(2004) compared the effects of noncontingentreinforcement (NCR), NCR plus EE usingnonremoval of the spoon, and EE with fourchildren with feeding disorders. Acceptanceincreased and inappropriate behavior decreasedduring NCR plus EE and EE alone. Noncon-

tingent reinforcement in combination with EEwas associated with lower levels of inappropri-ate behavior or negative vocalizations for somechildren.

Taken together, results of the literature onpediatric feeding disorders suggest that there areat least two potential explanations of theiretiology, which are associated with specificmethods of treatment. Given the widespreaduse of, but lack of empirical support for,sensory integration as treatment for feedingproblems, it seems that evaluation of theapproach is warranted. The logical method ofevaluation would be to compare sensoryintegration to a procedure with empiricalevidence as treatment for pediatric feedingdisorders. Therefore, in the current investiga-tion, we compared the effects of sensoryintegration with an operant-based treatment(NCR plus EE) for two children with feedingproblems. We selected EE because the results offunctional analyses suggested that escape func-tioned as reinforcement for inappropriatebehavior for both children. In addition, EE isthe operant-based treatment with the mostempirical support (Volkert & Piazza, 2012).We added NCR to EE per parental request andthe results of Reed et al. (2004), whichsuggested that NCR plus EE was effective astreatment for feeding problems.

METHOD

ParticipantsTwo children who had been admitted to a

pediatric feeding disorders day-treatment pro-gram participated. The children were includedin the study because their primary presentingproblem was a feeding disorder. In addition,the program OT diagnosed the feeding prob-lems as sensory based and recommendedsensory integration as treatment. Followingthe recommendation of Wilbarger and Wilbar-ger (1991), she based her recommendations onan interview with the parents and directobservation of child behavior, from which shedetermined that each child had difficulties with

SENSORY INTEGRATION AND FOOD REFUSAL 457

sensory processing that contributed to theirfeeding problems. Both of these children hadfood refusal characterized by overreaction(crying, gagging, spitting food out) to presen-tation of food and other indicators of sensoryintegration dysfunction (e.g., overresponsive totags in clothing, difficulties during toothbrushing, aversion to touch around the mouthand face, aversion to walking in sand or grass).These end-product behaviors were indicators ofdysfunction for sensory system modulation,specifically, overregistration of sensory inputfrom oral and tactile stimuli (Kimball, 1999).

Mark was a 1-year-old boy who had beenadmitted for food refusal and failure to thrive.His diagnoses included gastroesophageal refluxdisease, high blood pressure, asthma, and acoarctation of the aorta. His medicationsincluded Lasix, Enalapril, Prilosec, Periactin,Metoclopriamide, Pulmacort, and Xopenex.These medications were stable throughout thestudy. Irma was a 3-year-old girl who had beenadmitted for poor oral intake and foodselectivity by type and texture. Her diagnosesincluded developmental delays, dysphagia, andmitochondrial disorder. Her medications in-cluded Coenzyme Q10 and Carnitor. Bothchildren ate some foods and drank somebeverages, but not enough to sustain theirnutritional, caloric, or hydration needs.

TherapistsThe feeding therapists were predoctoral

interns in psychology or individuals withbachelor’s or master’s degrees in psychology ora related field. The feeding therapists conductedthe feeding sessions, including all sessionsduring the functional analysis and the treatmentevaluation. The OT did not conduct anyfeeding sessions. The feeding therapists andthe OT conducted the sensory integrationprotocol described below. The OT had beenin her profession for more than 20 years, hadextensive training and experience in sensoryintegration therapy and pediatric feeding dis-orders, and was licensed in her profession. Inaddition, she had advanced certification in the

administration and interpretation of the Sen-sory Integration and Praxis Tests.

Setting and MaterialsThe feeding therapist conducted feeding

sessions in rooms (4 m by 4 m) equipped withone-way observation. A high chair, food ordrink, scale, gloves, fluid-resistant bib withcrumb catcher (the bottom of the bib foldedand snapped to form a receptacle), and eatingor drinking utensils were present during allsessions. The feeding therapists and the OTconducted the sensory integration activities inthe sensory integration room (4 m by 4 m) thatwas equipped with floor mats, a mirror, and theitems listed below. Feeding sessions were notconducted in the sensory integration room, andthe sensory integration activities were notconducted in the feeding therapy rooms.

Dependent Variables and Data CollectionTrained observers used laptop computers to

score acceptance and inappropriate behavior.Observers scored acceptance once per presenta-tion. A presentation occurred when the feederplaced the cup or spoon 4 cm from the child’slips, not including the placement of the cup orspoon following re-presentation of expelleddrinks or bites. Acceptance occurred when anyamount of liquid or the entire bolus of food(Irma, eating) passed the child’s lips within 5 sof presentation. Data on acceptance werecollected by dividing the number of acceptancesby the number of presentations and convertingthe ratio to a percentage. Observers scoredinappropriate behavior each time the childturned his or her head 458 or more away fromthe cup or spoon, hit the cup or spoon or thefeeder’s arm or hand, or covered the mouthwhen the cup or spoon was within arm’s reachof the child. Observers scored the length of timein which the cup or spoon was within arm’sreach of the child by activating a timer withinthe data-collection program when the feederpresented the cup or spoon by placing it within4 cm of the child’s lips and turning off thetimer when the feeder moved the cup or spoonapproximately 30 cm from the child (i.e., out ofthe child’s reach). The computer then calculat-


ed the duration of cup or spoon within arm’sreach of the child. The frequency of inappro-priate behavior was converted to a rate bydividing the number of inappropriate behaviorsby the amount of time (in minutes) the cup orspoon was within arm’s reach of the child.Feeding therapists also measured how much thechild drank or ate in grams, using a Tanita1475T scale. To calculate grams consumed, thefeeder placed the cup of liquid or bowl of foodon the scale prior to and after the session andrecorded the pre- and postweights. The feederused paper towels, which weighed 2 g each, toclean up spills. The feeder cleaned spills byusing the paper towels to wipe up any drink orfood that was on the child’s face, bib, tray, orfloor after each session. The feeder thencalculated grams consumed with the formula(presession food weight minus postsession foodweight) minus (paper towels with spill minuspaper towels without spill).

For the functional analysis, a second observersimultaneously but independently collecteddata during 41%, 35%, and 33%, of sessionsfor Mark, Irma (drinking), and Irma (eating),respectively. Interobserver agreement for accep-tance was calculated by partitioning eachsession into 10-s intervals, dividing the numberof intervals during which both observers agreedon the occurrence or nonoccurrence of accep-tance by the total number of intervals, andconverting this ratio to a percentage. Interob-server agreement for inappropriate behavior wascalculated by dividing the number of exactagreements, defined as a 10-s interval in whichboth observers scored the same frequency ofinappropriate behavior, by the number ofintervals and converting this ratio to a percent-age. Mean agreement for acceptance was 99%(range, 85% to 100%) for Mark and 100% forIrma (drinking and eating). Mean agreementfor inappropriate behavior was 90% (range,87% to 100%) for Mark, 92% (range, 97% to100%) for Irma (drinking), and 99% (range,95% to 100%) for Irma (eating).

During the treatment analysis, a secondobserver simultaneously but independently

collected data during 26%, 34%, and 34% ofsessions for Mark and Irma (drinking andeating), respectively. Mean agreement foracceptance was 95% (range, 81% to 100%)for Mark, 98% (range, 71% to 100%) for Irma(drinking), and 97% (range, 83% to 100%) forIrma (eating). Mean agreement for inappropri-ate behavior was 97% (range, 78% to 100%)for Mark, 99% (range, 90% to 100%) for Irma(drinking), and 99% (range, 89% to 100%) forIrma (eating).

DesignDuring Mark’s functional analysis, we used a

reversal design in which we randomly selectedthe order of test conditions (escape andattention) with a control phase following eachphase of a test condition (Piazza, Fisher, et al.,2003). The order of conditions was attention,control, escape, control, escape. For Irma, weused a pairwise design (e.g., Bachmeyer et al.,2009) in which test and control conditions werealternated in each phase. We conductedseparate analyses for eating and drinking,because different oral motor skills are requiredto be successful with eating and drinking, andbecause some children have different abilitieswith each skill.

During the treatment evaluation, we used anABCBC design to compare levels of acceptanceand rates of inappropriate behavior in theescape (A), escape plus sensory integration (B),and EE plus NCR (C) conditions.

General ProcedureA feeding therapist conducted meals five

times a day with approximately 1 to 3 hrbetween the start of each meal (e.g., 9:00 a.m.,10:30 a.m., 12:00 p.m., 2:30 p.m., 4:00 p.m.).Each meal lasted approximately 30 to 45 minand consisted of approximately three to sevenfive-bite sessions, with 1- to 2-min breaksbetween sessions. During phases that includedsensory integration, the therapist implementedthe sensory integration protocol for 10 minprior to the scheduled meal.

During meals, the feeding therapist presented3 cc of whole milk with Carnation Instant


Breakfast in a cut-out cup (a plastic cup with asection cut out in a U shape) to Mark and 8 cc ofLactaid milk thickened with baby oatmeal cerealin a cut-out cup to Irma. The OT recommendedthickening Irma’s liquids due to her inexperienceas an oral feeder. The program dietitianrecommended the liquids for each child basedon the child’s caloric and nutritional needs.Irma’s parents identified 16 foods, four foodsfrom each of the food groups of fruit, protein,starch, and vegetable, to present in sessions. Thefeeding therapist randomly selected four of theparent-identified foods (a fruit, protein, starch,and vegetable) to present to Irma in each session.The order of food presentation was randomlyselected within the session. The foods werepureed texture (table food blended in a chopperuntil smooth). The feeding therapist presentedIrma with a level bolus of food (the therapistfilled the bowl of the spoon and scraped thebowl on the side of the dish to level the bolus)on a small maroon spoon. The OT selected thetextures, bolus sizes, and utensils.

Across all feeding sessions, the feedingtherapist presented the cup or spoon approxi-mately 4 cm from the midline of the child’smouth accompanied by a verbal prompt to‘‘take a drink [bite]’’ approximately every 30 s.The therapist delivered praise (e.g., ‘‘Good jobtaking a drink’’) following acceptance. Thechild’s mouth was checked 30 s after the drinkor bite entered the mouth to determine if thechild had swallowed. The feeding therapistprompted the mouth check by instructing thechild to ‘‘show me, ahhh’’ while modeling anopen mouth. If the child did not open his orher mouth, the therapist inserted a rubber-coated baby spoon into the mouth and turnedthe spoon 908 to open the mouth. The feedingtherapist delivered praise (e.g., ‘‘Good jobswallowing your drink’’) if no liquids or solidslarger than the size of a pea were in the child’smouth during the first check. No praise wasdelivered if the absence of liquids or solids atthe 30-s check was due to expulsion. Thetherapist delivered a verbal prompt (e.g.,

‘‘swallow your drink’’) if liquids or solids largerthan the size of a pea were in the child’s mouthat the check and repeated the prompt every 30 suntil no liquids or solids larger than the size of apea were visible. The therapist then presentedthe next drink or bite.

Functional AnalysisWe conducted a functional analysis to

identify the reinforcers for inappropriate be-havior. During the functional analysis, thefeeding therapist followed the general proce-dures described above in addition to the specificprocedures for each condition described below.Across all conditions, the therapist held the cupor spoon at midline (the position in space inwhich the feeder presented the cup or spooninitially) if the child did not accept the drink orbite and provided no differential consequencefor expulsion or vomiting.

During the control condition, the feedingtherapist provided no differential consequencefor inappropriate behavior and interacted withthe child (e.g., sang) in the presence of toys onthe tray throughout the session. During theescape condition, the therapist removed the cupor spoon for 30 s if the child engaged ininappropriate behavior and presented the nextdrink or bite after the 30-s interval. Thetherapist did not provide any other differentialconsequence following inappropriate behavior(e.g., the child was not reprimanded). Toyswere not available. During the attentioncondition, the therapist provided 30 s ofattention following inappropriate behavior(but the cup or spoon remained at midline).Toys were not available.

Sensory Integration Protocol Development andTraining

The OT developed individualized sensoryintegration programs for each child. Becauseboth children demonstrated high levels ofinappropriate behavior during the presentationof food, the OT developed a sensory integrationprogram outside of the mealtime (Case-Smith& Humphry, 2005). That is, rather thaninitiating therapy in an environment that


resulted in child distress, the OT initiatedsensory activities in the context of play toestablish a relationship of trust with the child(Case-Smith & Humphry, 2005). Not all of thesensory activities involved the mouth because,as indicated above, patients often prefer sensoryinput to the arms, legs, and back rather than theface (Koomar & Bundy, 2002). Because thesensory systems interact, stimulation of onesensory system will affect other sensory systems(Pruzansky & Farrington, 1999). The OTimplemented the sensory integration procedureat regular intervals throughout the day accord-ing to a schedule (Wilbarger & Wilbarger,1991). Regular delivery of an individuallyprescribed sensory diet is hypothesized topromote changes in the CNS that will resultin alterations in end-product behavior (e.g.,improved feeding). The OT then trained thefeeding staff to implement the sensory integra-tion procedure. Consultation and caregivertraining are common methods of deliveringregularly scheduled sensory activities, becausemost OTs are not available to provide therapythroughout the day for multiple consecutivedays with a single child (Kimball et al., 2007;Wilbarger & Wilbarger, 1991).

First, the OT described the protocol verbally,and a written protocol was developed based onher instructions. Next, the OT approved thewritten protocol and demonstrated the protocolto the feeding therapists. She then observed thefeeding therapists implementing the protocolwith the child and provided feedback until thefeeding therapists implemented the protocolcorrectly. Next, the treatment evaluation began.The OT periodically observed each feedingtherapist implementing the sensory integrationprotocol to ensure correct implementationthroughout the study and provided correctivefeedback, if necessary; however, she did notcollect data on the integrity of protocolimplementation. In addition, the OT periodi-cally implemented the protocol herself.

During the sensory integration protocol, thetherapist interacted with the child by prompt-ing activities as described below; providingpositive comments about the activity (e.g., ‘‘the

bug feels great’’) or other topics (e.g., ‘‘what abeautiful day’’); providing periodic, noncontin-gent praise (e.g., ‘‘you are doing a great job’’) inan upbeat, enthusiastic tone of voice; andsmiling and laughing periodically. We attempt-ed to balance child directedness (i.e., allowingthe child to determine the type and timing ofthe activities rather than specifying the activitiesin advance), which is a hallmark of sensoryintegration, in the context of developing aprocedure that was technological. To that end,the OT recommended a sequence of sensoryintegration activities, but the sequence couldchange if the child demonstrated interest in(e.g., pointed to, crawled toward) an alternativeactivity. If the child changed activities, thetherapist followed the child to the alternativeactivity and implemented that activity accord-ing to the protocol (see below). For example, ifthe child was bouncing on the ball and then leftthe ball and began crawling through the tunnel,the therapist went to the tunnel with the childand followed the tunnel protocol. However, wedid not collect data on how often the therapistfollowed the recommended routine and howoften the child initiated an alternative activity.The recommended sequence of sensory inte-gration activities were vibrating bug, tunnel,therapeutic brush, joint compression, ball, andARK toys for Mark; vibrating bug (lips andcheeks only) and ARK toys for Irma (drinking);and vibrating bug, bubbles, and ball for Irma(eating). (ARK toys are resin-covered plastictoys in the shape of animals that vibrate inresponse to pressure that are used for oralstimulation during oral motor and sensoryintegration therapy.)

The sensory integration activities were im-plemented as follows. The therapist placed avibrating bug (a soft, cloth-covered toy shapedlike a bug that vibrated in response to pressure)sequentially on the child’s feet, legs, hands,arms, stomach, back, cheeks, and lips for about2 to 3 s each. The therapist verbally promptedthe child to crawl through a vinyl tunnel thatwas approximately 150 cm in length and 46 cm


in diameter. The therapist could use a variety ofstrategies to motivate the child to crawl throughthe tunnel such as verbal encouragement (e.g.,saying, ‘‘you can do it’’), the mirror (e.g.,placing the mirror at the end of the tunnel),crawling through the tunnel with the child, orphysically assisting the child to crawl throughthe tunnel. The child was encouraged to crawlthrough the tunnel three or four times. Thetherapist stroked the child’s hands, legs, feet,and back four times each with a therapeuticbrush with firm pressure. (A therapeutic brushis a plastic brush with soft, dense bristlesdesigned specifically to provide sensory stimu-lation.) During joint compression, the therapistsequentially supported and provided firmcompression to each joint (elbow, shoulders,knees, and hip) 10 times. The therapist held abubble wand in front of the child andprompted him or her to blow bubbles. If thechild did not blow, the therapist modeled theresponse, and the trial ended if the child did notblow. The therapist prompted the child to blowbubbles approximately once every 20 s for atotal of about 3 min. The therapist placed thechild prone on his or her stomach on a 45-cmtherapy ball and rolled the ball forward untilthe child’s hands touched the ground. The ball-rolling sequence was repeated approximatelyfour times. The surface of the ARK toy hasgrooves and textures to provide different typesof sensory stimulation to the mouth. Thetherapist (a) placed an ARK toy into the child’smouth near the molars (alternating from oneside of the mouth to the other) and promptedthe child to bite down 10 times; (b) pushedlaterally on the child’s tongue three times andthen moved the ARK toy to the child’s cheek,alternating from one cheek to the other; and (c)rotated the ARK toy across the child’s lips in acircular motion three times. The therapist andchild sat on a mat on the floor during thestationary activities. The sequence of activitieswas repeated with each child for approximately10 min prior to the escape with sensoryintegration feeding sessions (described below).

Treatment EvaluationEscape. Procedures were identical to the

escape condition of the functional analysis.Escape plus sensory integration. During the

escape plus sensory integration phase, a feedingtherapist or the OT implemented the sensoryintegration protocol described above for 10 minbefore each meal. After implementation of theprotocol, the feeding therapist immediatelybrought the child from the sensory integrationroom to the feeding room and began thefeeding sessions. The procedures during thefeeding sessions were identical to the escapecondition of the functional analysis and theescape baseline. The only difference betweenthe escape baseline and the escape plus sensoryintegration condition was that the childexperienced one 10-min period of sensoryintegration prior to participating in each 30-to 45-min meal. That is, the child experiencedone 10-min period of sensory integrationfollowed by approximately three to seven five-bite feeding sessions. We used escape baselinecontingencies during the sensory integrationintervention to maintain a trusting relationshipwith the child. Case-Smith and Humphry(2005) note that a child will not trust an adultwho attempts to place food in his or her mouth.They highlight the importance of responding tothe child’s physical cues of discomfort bywithdrawing the oral stimulus. Therefore, whenthe child turned his or her head or batted at thecup or spoon, the therapist responded to thechild’s cue by removing the cup or spoon.

Escape extinction with NCR. The results ofthe functional analyses for both childrensuggested that escape from cup or spoonpresentations maintained inappropriate behav-ior. Therefore, we used EE during treatment.Both parents requested that we interact withtheir children during the session. Therefore, weadded NCR to EE.

The feeding therapist conducted a mini-choice assessment (Roane, Vollmer, Ringdahl,& Marcus, 1998) prior to the meal in whichshe presented the child with the three most


highly preferred toys identified during astimulus preference assessment (Fisher et al.,1992). The chosen toy was available through-out the meal. The therapist also providedattention throughout the meal.

The therapist held the cup or spoon at thechild’s lips and deposited the drink or bitewhen the child opened his or her mouth. Thetherapist re-presented (i.e., scooped up theliquid or solid from the child’s face or bib withthe cup or spoon and placed the drink or biteinto the child’s mouth) expelled drinks or bitesand provided no differential consequence forinappropriate behavior. The session endedwhen the child had no liquid or solid largerthan the size of a pea in his or her mouthfollowing the fifth drink or bite entering thechild’s mouth, except when the absence ofliquid or solid at the 30-s check was a result ofexpulsion. No differential consequence wasprovided for gagging, coughing, or vomiting.

RESULTS

The results of the functional analyses (Figure1) suggested that inappropriate behavior wasmaintained by escape from drink or bitepresentations for Mark and Irma (drinkingand eating). For Mark, mean inappropriatebehavior was 5.8 responses per minute in theattention condition, 3.4 responses per minutein the two phases of the control condition, and9.1 responses per minute in the two phases ofthe escape condition. For Irma (drinking),mean inappropriate behavior was 0.1 and 0.3responses per minute in the pairwise compar-ison of attention and control, respectively, and3 and 0.2 responses per minute in the pairwisecomparison of escape and control, respectively.For Irma (eating), mean inappropriate behaviorwas 7 and 2 responses per minute in thepairwise comparison of escape and control,respectively, and 1 and 0.4 in the pairwisecomparison of attention and control, respec-tively.

For Mark, levels of acceptance (Figure 2,top) were low (M ¼ 39%), and meaninappropriate behavior (Figure 2, second panel)was 11 responses per minute during the initialescape baseline. During escape plus sensoryintegration, mean levels of acceptance (M ¼41%) and rates of inappropriate behavior (M¼11) were equivalent to baseline. Levels ofacceptance increased (M ¼ 94%), and rates ofinappropriate behavior decreased (M ¼ 2)during EE plus NCR. During the return toescape plus sensory integration, levels ofacceptance (M ¼ 64%) were slightly higherand rates of inappropriate behavior wereslightly lower than during the initial escapeplus sensory integration phase (M ¼ 4), butlower (acceptance) and higher (inappropriatebehavior) than during EE plus NCR. Duringthe reintroduction of EE plus NCR, meanacceptance increased to 100%, and meaninappropriate behavior decreased to 0.5 re-sponses per minute. Mean grams consumedwere 4 during baseline, 7.5 during escape plussensory integration, and 23 during EE plusNCR. These data suggest that the biggestincrease in grams consumed occurred duringthe EE plus NCR condition.

For Irma (drinking), levels of acceptance(Figure 2, third panel) were zero during theescape and escape plus sensory integrationphases. Rates of inappropriate behavior (Figure2, bottom) were high during the escape (M ¼33) and escape plus sensory integration (M ¼25) phases. Levels of acceptance increased (M¼94%) and rates of inappropriate behaviordecreased (M¼ 1) during EE plus NCR. Levelsof acceptance decreased and rates of inappro-priate behavior increased during the reversal toescape plus sensory integration. Levels ofacceptance increased (M ¼ 93%) and rates ofinappropriate behaviors decreased (M ¼ 0.7)during the reintroduction of EE plus NCR.Mean grams consumed were 0 during escapeand escape plus sensory integration conditionsand 38 during EE plus NCR. Thus, intake was


Figure 1. Inappropriate behavior per minute for Mark (top), Irma (drinking, middle), and Irma (eating, bottom).


Figure 2. Percentage acceptance and inappropriate behavior per minute during Mark’s and Irma’s drinkingtreatment evaluation.


substantially higher during EE plus NCR thanduring escape with sensory integration.

Figure 3 shows the results of Irma’s eatingtreatment analysis. Mean acceptance (top) was1% and mean inappropriate behavior (bottom)was 7 responses per minute during the initialescape baseline. During escape plus sensoryintegration, mean levels of acceptance wereequivalent to baseline (M ¼ 1%), and rates ofinappropriate behavior increased (M ¼ 22).Levels of acceptance increased (M ¼ 96%) andrates of inappropriate behavior decreased (M¼2) during EE plus NCR. During the return toescape plus sensory integration, levels ofacceptance decreased (M ¼ 0%), and rates of

inappropriate behavior increased. During thereintroduction of EE plus NCR, mean accep-tance increased to 96%, and mean inappropri-ate behavior decreased to 0.9 responses perminute. Mean grams consumed were 0.3during escape, 0.1 during escape plus sensoryintegration, and 6.7 during EE plus NCR.Thus, intake was substantially higher during EEplus NCR than during escape with sensoryintegration.

DISCUSSION

In the current investigation, we implementeda sensory integration intervention for two

Figure 3. Percentage acceptance (top) and inappropriate behavior per minute (bottom) during Irma’s eatingtreatment evaluation.


children who had been diagnosed by an OTwith a sensory-based feeding disorder. The OT,who had extensive training and experience insensory integration, developed the procedures,trained the therapists, and participated inimplementation of the sensory integrationtherapy. The results showed that the sensoryintegration intervention did not produce stableor lasting increases in acceptance and gramsconsumed or decreases in inappropriate behav-ior for either child.

There are a variety of potential reasons whythe sensory integration procedures were noteffective. One challenge we encountered indevelopment of the procedures is that sensoryintegration interventions are supposed toemerge ‘‘dynamically’’ through therapist–childcollaboration (L. D. Parham et al., 2007). Infact, Parham et al. (2007) noted that we cannotbe confident about the results of many existingstudies on sensory integration because therapistsimplemented the procedures in a prescribedsequence (e.g., Humphries, Wright, McDou-gall, & Vertes, 1990) or for a specified period oftime. Parham et al. also criticized studies thatattempted to isolate effective components oftreatment by minimizing the overlap betweensensory integration and the alternative proce-dures in comparison studies (see below; e.g.,Mason & Iwata, 1990). Parham et al.’sconcerns offer a challenge to the scientist whois interested in testing the efficacy of sensoryintegration. That is, how do we developprocedures that are technologically precise andstudies that are well controlled while stilladhering to the constructs of sensory integra-tion? Presumably, too much technologicalprecision and control may contradict the tenetsof sensory integration and too little technolog-ical precision and control may lead to ascientifically inadequate experiment.

This dilemma is exemplified in a study byMason and Iwata (1990), who tested the effectsof sensory integration therapy on the self-injurious behavior (SIB) of three participantswhose SIB was maintained by automatic

reinforcement, attention, and escape fromdemands, respectively. One component of thesensory integration program involved exposingthe individuals to automated physiologicalstimulation (e.g., lights, music) in the absenceof adult attention. Mason and Iwata usedautomated stimulation to assess the effects ofphysiological stimulation in the absence of theconfounding effects of attention. That is, if thetherapist delivered attention, it would not beclear whether the observed effects were afunction of sensory stimulation, adult attention,or both. L. D. Parham et al. (2007) criticizedMason and Iwata, stating,

The research design of this study did not permittherapist–child interaction as a therapeutic strategy.The therapist was present in the room to monitorthe child’s safety but did not otherwise interact withthe child, to allow the child to have continuousaccess to sensory stimulation without the confound-ing effect of ‘‘social stimulation’’ (p. 364). Appar-ently, the researchers assumed that sensoryintegration intervention consisted of sensory stim-ulation only and that the therapist–child relation-ship was not intrinsic to the intervention. (p. 221)

The counterpoint to L. D. Parham et al.’s(2007) criticism, as pointed out by Mason andIwata (1990), is that there may be componentsof sensory integration that have functionaleffects on behavior. For example, the SIB ofone of the participants in the Mason and Iwatastudy was maintained by attention. For thisparticipant, the therapist conducted sensoryintegration with and without adult attention.SIB was low when adult attention was presentbut not when it was absent, which would beexpected based on the results of the functionalanalysis. Mason and Iwata concluded that thelower levels of SIB during sensory integrationwith adult attention were an artifact of thepresence of adult attention rather than an effectof sensory integration per se.

In the current investigation, we specified asequence of activities through which thetherapist prompted the child such that thestudy was sufficiently technological, but thechild also had the opportunity to change or


redirect the activities such that the study wasconsistent with the parameters of sensoryintegration. Our experience was that specifica-tion of the sequence and timing of the activitiesprovided information necessary for replicationand the structure for the sensory integrationsession if the child did not initiate activities onhis or her own, which often did not happen. Italso was necessary to time the sensory integra-tion therapy such that it occurred immediatelybefore the meal, which is a recommendedmethod of implementing this type of therapy(Case-Smith & Humphry, 2005; Glass &Wolf, 1998). That is, the focus of sensoryintegration therapy is to alter the child’s sensoryprocessing capabilities through a regularlyscheduled sequence of sensory-organizing stim-ulation.

Although the sensory integration interven-tion was not effective for the participants in thecurrent investigation, some features of sensoryintegration as described in the literature mayaddress the functional components of feedingproblems for other children. For example,Case-Smith and Humphry (2005) recommend-ed interaction during meals to distract the childand ‘‘looks of delight’’ and meal termination asa reward for appropriate eating. These recom-mendations could be conceptualized in thebehavior-analytic literature as noncontingentpositive (Reed et al., 2004), differential positive(Piazza, Patel, et al., 2003), and differentialnegative (Lalli et al., 1999; LaRue et al., 2011)reinforcement, respectively, and would beappropriate components of treatment forchildren whose inappropriate mealtime behav-ior was maintained by positive (noncontingentand differential positive reinforcement) andnegative (differential negative reinforcement)reinforcement. By contrast, other componentsof sensory integration (e.g., acknowledging thechild’s physical cues by withdrawal of the oralstimulus; Case-Smith & Humphry, 2005) mayactually worsen feeding behavior if inappropri-ate behavior is maintained by negative rein-forcement (Piazza, Fisher, et al., 2003). This

might explain the results of DeGangi, Sickel,Wiener, and Kaplan (1996), who found thatpremature infants who received sensory inte-gration had more feeding problems at follow-up than a control group who did not receivesensory integration. DeGangi et al. randomlyassigned premature infants who did not havereported feeding problems at the start of thestudy to either a sensory integration or a controlgroup, and they followed the two groups overtime to evaluate the effects of the sensoryintegration intervention on a variety of out-comes. It is not clear from the study howcaregivers implemented the sensory integrationactivities. If the caregiver removed the feedingutensil or terminated the meal, as recommend-ed in some sensory integration interventions(e.g., Case-Smith & Humphry, 2005), it ispossible that the inappropriate behavior of thechildren in the sensory integration group mayhave worsened over time if inappropriatebehavior was maintained by escape.

In the current investigation, when the child’sphysical cues of head turning and batting at thespoon were followed by removal of the drink orbite, inappropriate behavior was high andacceptance was unchanged during the escapeand the escape plus sensory integration condi-tions. By contrast, when inappropriate behaviorno longer produced escape during EE plusNCR, inappropriate behavior decreased andacceptance increased. These results are consis-tent with those of Reed et al. (2004) and othersthat have shown that EE in conjunction withpositive reinforcement is an effective treatmentfor feeding disorders (Ahearn et al., 1996;Bachmeyer et al., 2009; Cooper et al., 1995;Piazza, Patel, et al., 2003).

Although EE plus NCR was effective, thesefindings are limited because sensory integrationwas always the first intervention and EE plusNCR always followed sensory integration. It ispossible that sensory integration did have apositive effect on behavior that was notobserved until the EE plus NCR phase (i.e.,the effects of sensory integration were delayed).


This explanation seems unlikely because re-sponding returned to baseline levels during thereversals to escape with sensory integration.Nevertheless, future studies should correct thislimitation.

Another potential limitation of the currentinvestigation is that toys and therapist attentionwere available in EE plus NCR and not inescape plus sensory integration. It was impor-tant to examine sensory integration indepen-dent of other treatment components such asNCR because this is the first study to ourknowledge that systematically examined theeffects of sensory integration as treatment forchildren with feeding disorders. If we hadadded NCR to escape plus sensory integration,the individual effects of sensory integration onthe feeding behaviors would be unclear. Also, itis unlikely that NCR was the reason that EEplus NCR was an effective treatment andsensory integration was not. Previous researchhas shown that NCR alone is not an effectivetreatment for children with feeding disordersand that EE is necessary to increase acceptanceand decrease inappropriate behavior (Reed etal., 2004). In the current investigation, weadded NCR to EE per parental request andbecause the data from Reed et al. (2004)showed that NCR had some beneficial effectsfor some participants when combined with EE.It is possible that NCR contributed to thetreatment effects during EE plus NCR for thechildren in the current investigation. Futureresearch could examine the effects of NCRduring feeding sessions following sensoryintegration therapy.

Another limitation of the current investiga-tion is that it included only two children. It ispossible that the feeding behavior of thesechildren was not representative of the feedingbehavior of children who are commonly treatedwith sensory integration. In addition, thepopularity of sensory integration among first-line clinicians of feeding disorders (e.g., OTs)should provide an incentive for further inves-tigation, particularly collaborative efforts be-

tween experts in sensory integration andbehavior analysis. Such a collaboration wouldensure that therapists implement both sensoryintegration and behavior-analytic treatmentswith high fidelity. The lack of treatment fidelitydata for the sensory integration and behavior-analytic treatments is an additional limitationof the current study. Documentation of thefidelity with which treatments are implementedwould increase the confidence in the findings,which is important as debate continues as to theeffectiveness of sensory integration (Case-Smith& Arbesman, 2008; Gresham, Beebe-Franken-berger, & MacMillan, 1999; Hoehn & Bau-meister, 1994; May-Benson & Koomar, 2010;Schechtman, 2007).

REFERENCES

Ahearn, W. H., Kerwin, M. E., Eicher, P. S., Shantz, J., &Swearingin, W. (1996). An alternating treatmentscomparison of two intensive interventions for foodrefusal. Journal of Applied Behavior Analysis, 29, 321–332.

Ayres, A. J. (1972). Improving academic scores throughsensory integration. Journal of Learning Disabilities, 5,338–343.

Ayres, A. J. (1974). The development of sensory integrativetheory and practice. Dubuque, IA: Kendall/Hunt.

Bachmeyer, M. H., Piazza, C. C., Fredrick, L. D., Reed,G. K., Rivas, K. D., & Kadey, H. J. (2009).Functional analysis and treatment of multiplycontrolled inappropriate mealtime behavior. Journalof Applied Behavior Analysis, 42, 641–658.

Baranek, G. T. (2002). Efficacy of sensory and motorinterventions for children with autism. Journal ofAutism and Developmental Disorders, 32, 397–422.

Borrero, C. S. W., Woods, J. N., Borrero, J. C., Masler, E.A., & Lesser, A. D. (2010). Descriptive analyses ofpediatric food refusal and acceptance. Journal ofApplied Behavior Analysis, 43, 71–88.

Brown, T. D., Rodger, S., Brown, A., & Roever, C.(2005). A comparison of Canadian and Australianpaediatric occupational therapists. Occupational Ther-apy International, 12, 137–161.

Caretto, V., Topolski, K. F., Linkous, C. M., Lowman,D. K., & Murphy, S. M. (2000). Current parenteducation on infant feeding in the neonatal intensivecare unit: The role of the occupational therapist.American Journal of Occupational Therapy, 54, 59–64.

Case-Smith, J., & Arbesman, M. (2008). Evidence-basedreview of interventions for autism used in or ofrelevance to occupational therapy. American Journalof Occupational Therapy, 62, 416–429.


Case-Smith, J., & Humphry, R. (2005). Feedingintervention. In J. Case-Smith (Ed.), Occupationaltherapy for children (pp. 481–520). St. Louis, MO:Mosby.

Case-Smith, J., & Miller, H. (1999). Occupationaltherapy with children with pervasive developmentaldisorders. American Journal of Occupational Therapy,53, 506–513.

Cermak, S. A. (2001). The effects of deprivation onprocessing, play, and praxis. In S. S. Roley, E. I.Blanche, & R. C. Schaaf (Eds.), Understanding thenature of sensory integration with diverse populations(pp. 385–404). San Antonio, TX: Therapy SkillBuilders.

Cermak, S. A., & Mitchell, T. W. (2006). Sensoryintegration. In R. McCauley & M. Fey (Eds.),Treatment of language disorders in children (pp. 435–469). Baltimore, MD: Brookes.

Cooper, L. J., Wacker, D. P., McComas, J. J., Brown, K.,Peck, S. M., Richman, D., et al. (1995). Use ofcomponent analyses to identify active variables intreatment packages for children with feeding disor-ders. Journal of Applied Behavior Analysis, 28, 139–153.

DeGangi, G. A., Sickel, R. Z., Wiener, A. S., & Kaplan,E. P. (1996). Fussy babies: To treat or not to treat?British Journal of Occupational Therapy, 59, 457–466.

Di Lorenzo, C., Colletti, R. B., Lehmann, H. P., Boyle, J.T., Gerson, W. T., Hyams, J. S., et al. (2005).Chronic abdominal pain in children: A technicalreport of the American Academy of Pediatrics and theNorth American Society for Pediatric Gastroenterol-ogy, Hepatology and Nutrition: AAP Subcommitteeand NASPGHAN Committee on Chronic Abdom-inal Pain. Journal of Pediatric Gastroenterology &Nutrition, 43, 249–261.

Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L.P., Owens, J. C., & Slevin, I. (1992). A comparisonof two approaches for identifying reinforcers forpersons with severe to profound disabilities. Journal ofApplied Behavior Analysis, 25, 491–498.

Girolami, P. A., & Scotti, J. R. (2001). Use of analogfunctional analysis in assessing the function ofmealtime behavior problems. Education and Treat-ment in Mental Retardation and DevelopmentalDisabilities, 36, 207–223.

Glass, R. P., & Wolf, L. S. (1998). Feeding and oralmotor skills. In J. Case-Smith (Ed.), Pediatricoccupational therapy and early intervention (2nd ed.,pp. 127–163). Boston, MA: Butterworth Heine-mann.

Gresham, F. M., Beebe-Frankenberger, M. E., &MacMillan, D. L. (1999). A selective review oftreatments for children with autism: Description andmethodological considerations. School Psychology Re-view, 28, 559–575.

Hoehn, T. P., & Baumeister, A. A. (1994). A critique ofthe application of sensory integration therapy tochildren with learning disabilities. Journal of LearningDisabilities, 27, 338–350.

Humphries, T., Wright, M., McDougall, B., & Vertes, J.(1990). The efficacy of sensory integration therapy forchildren with learning disability. Physical and Occu-pational Therapy in Pediatrics, 10, 1–17.

Kerwin, M. E. (1999). Empirically supported treatmentsin pediatric psychology: Severe feeding problems.Journal of Pediatric Psychology, 24, 193–214.

Kimball, J. G. (1999). Sensory integration frame ofreference: Postulates regarding change and applica-tion to practice. In P. Kramer & J. Hinojosa (Eds.),Frames of reference for pediatric occupational therapy(pp. 169–204). Philadelphia, PA: Lippincott.

Kimball, J. G., Lynch, K. M., Stewart, K. E., Williams,M. E., Thomas, M. A., & Atwood, K. D. (2007).Using salivary cortisol to measure the effects of aWilbarger protocol-based procedure on sympatheticarousal: A pilot study. American Journal of Occupa-tional Therapy, 61, 406–413.

Koomar, J. A., & Bundy, A. C. (2002). Creating directintervention from theory. In A. C. Bundy, S. J. Lane,& E. A. Murray (Eds.), Sensory integration: Theoryand practice (pp. 261–302). Philadelphia, PA: Davis.

Lalli, J. S., Vollmer, T. R., Progar, P. R., Wright, C.,Borrero, J., Daniel, D., et al. (1999). Competitionbetween positive and negative reinforcement in thetreatment of escape behavior. Journal of AppliedBehavior Analysis, 32, 285–296.

Lane, S. J. (2008). Sensory integration. In P. J. Accardo(Ed.), Neurodevelopmental disabilities in childhood:Vol. 1. Neurodevelopmental diagnosis and treatment(pp. 673–689). Baltimore, MD: Brookes.

LaRue, R. H., Stewart, V., Piazza, C. C., Volkert, V. M.,Patel, M. R., & Zeleny, J. (2011). Escape asreinforcement and escape extinction in the treatmentof feeding problems. Journal of Applied BehaviorAnalysis, 44, 719–735.

Mailloux, Z., & Roley, S. S. (2004). Sensory integration.In H. Miller Kuhaneck (Ed.), Autism: A comprehen-sive occupational therapy approach (pp. 215–244).Bethesda, MD: AOTA Press.

Mason, S. A., & Iwata, B. A. (1990). Artifactual effects ofsensory-integrative therapy on self-injurious behavior.Journal of Applied Behavior Analysis, 23, 361–370.

May-Benson, T. A., & Koomar, J. A. (2010). Systematicreview of the research evidence examining theeffectiveness of interventions using a sensory integra-tive approach for children. American Journal ofOccupational Therapy, 64, 403–414.

Najdowski, A. C., Wallace, M. D., Penrod, B., Tarbox, J.,Reagon, K., & Higbee, T. S. (2008). Caregiver-conducted experimental functional analyses of inap-propriate mealtime behavior. Journal of AppliedBehavior Analysis, 41, 459–465.

Parham, D., & Mailloux, Z. (2005). Sensory integration.In J. Case-Smith (Ed.), Occupational therapy forchildren (pp. 356–411). St. Louis, MO: Mosby.

Parham, L. D., Cohn, E. S., Spitzer, S., Koomar, J. A.,Miller, J. A., Burke, J. P., et al. (2007). Fidelity insensory integration intervention research. AmericanJournal of Occupational Therapy, 61, 216–227.


Piazza, C. C., Fisher, W. W., Brown, K. A., Shore, B. A.,Katz, R. M., Sevin, B. M., et al. (2003). Functionalanalysis of inappropriate mealtime behaviors. Journalof Applied Behavior Analysis, 37, 187–204.

Piazza, C. C., Patel, M. R., Gulotta, C. S., Sevin, B. S., &Layer, S. A. (2003). On the relative contribution ofpositive reinforcement and escape extinction in thetreatment of food refusal. Journal of Applied BehaviorAnalysis, 36, 309–324.

Pruzansky, C., & Farrington, K. (1999). Case study:Combined sensory integration and neurodevelop-mental treatment frames of reference. In P. Kramer &J. Hinojosa (Eds.), Frames of reference for pediatricoccupational therapy (pp. 589–608). Baltimore, MD:Lippincott, Williams, and Wilkens.

Reed, G. K., Piazza, C. C., Patel, M. R., Layer, S. A.,Bachmeyer, M. H., Bethke, S. D., et al. (2004). On therelative contribution of noncontingent reinforcementand escape extinction in the treatment of food refusal.Journal of Applied Behavior Analysis, 37, 27–41.

Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus,B. A. (1998). Evaluation of a brief stimuluspreference assessment. Journal of Applied BehaviorAnalysis, 31, 605–620.

Roley, S. S., & Schneck, C. (2001). Sensory integrationand visual deficits, including blindness. In S. S. Roley,E. I Blanche, & R. C. Schaaf (Eds.), Understandingthe nature of sensory integration with diverse populations(pp. 313–344). San Antonio, TX: Therapy SkillBuilders.

Royeen, C. B., & Lane, S. J. (1991). Tactile defensivenessand sensory processing. In A. G. Fisher, E. A.Murray, & A. C. Bundy (Eds.), Sensory integrationtheory and practice (pp. 108–133). Philadelphia, PA:Davis.

Schaaf, R. C., & Miller, L. J. (2005). Occupationaltherapy using a sensory integrative approach forchildren with developmental disabilities. MentalRetardation and Developmental Disabilities ResearchReviews, 11, 143–148.

Schechtman, M. A. (2007). Scientifically unsupportedtherapies in young children with autism spectrumdisorders. Psychiatric Annals, 37, 630–645.

Storch, B. A., & Eskow, K. C. (1996). Theory applicationby school-based occupational therapists. AmericanJournal of Occupational Therapy, 50, 662–668.

Tarbell, M. C., & Allaire, J. H. (2002). Children withfeeding tube dependency: Treating the whole child.Infants and Young Children, 15, 29–41.

Volkert, V. M., & Piazza, C. C. (2012). Empiricallysupported treatments for pediatric feeding disorders.In P. Sturmey & M. Hersen (Eds.), Handbook ofevidence based practice in clinical psychology (Vol. 1,pp. 456–481). Hoboken, NJ: Wiley.

Watling, R., Deitz, J., Kanny, E. M., & McLaughlin, J. F.(1999). Current practice of occupational therapy forchildren with autism. American Journal of Occupa-tional Therapy, 53, 498–505.

Wilbarger, P. (1984, September). Planning an adequatesensory diet—Application of sensory processingtheory during the first year of life. Zero to Three,pp. 7–12.

Wilbarger, P., & Wilbarger, J. (1991). Sensory defensivenessin children aged 2–12: An intervention guide for parentsand other caretakers. Santa Barbara, CA: AvantiEducational Programs.

Williamson, G. G., & Anzalone, M. (1997). Sensoryintegration: A key component of the evaluation andtreatment of young children with severe difficulties inrelating and communicating. Zero to Three, 17, 29–36.

Yack, E., Sutton, S., & Aquilla, P. (2002). Building bridgesthrough sensory integration. Las Vegas, NV: SensoryResources.

Received March 18, 2011Final acceptance April 26, 2012Action Editor, Jennifer Zarcone


Documents

A COMPARISON OF SENSORY INTEGRATIVE AND BEHAVIORAL ... · a comparison of sensory integrative and behavioral therapies as treatment for pediatric feeding disorders laura r. addison