Printer Friendly

The use of functional communication training to reduce pica.


Pica is an eating disorder characterized by the ingestion of inappropriate items. The items may range from relatively harmless materials (e.g., food from the floor, grass) to other substances (e.g., glass) with risk for serious health problems or even death. Previous studies have shown that behavioral techniques are effective for decreasing pica, but few studies have assessed this in naturalistic settings. The current paper describes a study conducted in a classroom setting in which functional communication was taught to a 6-year-old boy with autism to replace pica. Differential reinforcement of functional communication behavior was used to teach the boy to request food rather than to engage in pica. This procedure decreased the boy's frequency of pica and of attempted pica.

Keywords: pica, autism; differential reinforcement; functional communication training; manding.


Pica is an eating disorder often displayed by persons with developmental disabilities (Singh, 1983). It is typically defined as either the consumption of non-food items or the compulsive consumption of both food and non-food items (Albin, 1977; Paisey & Whitney, 1989). Many individuals with pica ingest items that are relatively harmless, such as food from the floor, grass, or bugs (Johnson et al., 1994; Fisher et al., 1994). However, some individuals with pica ingest items (e.g., glass) that can lead to serious digestive problems, health issues, and possibly death (Rojahn et al., 1987).

Several literature reviews describe interventions for persons who present with pica (e.g., McAdam, Sherman, Sheldon, & Napolitano, 2004). In one recent review McAdam et al. (2004) summarized the various behavioral interventions that have been demonstrated to produce a clinically significant reduction in pica, including differential reinforcement, overcorrection, contingent aversive presentations, and discrimination training. However, this review also found that the majority of published studies used intervention packages that included a punishment component. Moreover, few of the studies were conducted in natural settings (e.g., public schools, community-based agencies for adults with developmental disabilities).

Three published studies have demonstrated that differential reinforcement as the primary independent variable was successful in decreasing the pica of 6 persons with developmental disabilities. In one of these studies, Goh, Iwata, & Kahng, (1999), used naturalistic observations to assess the function of cigarette pica performed by three adult males and one female with developmental disabilities in a state residential facility. The function of pica for all participants was determined to be automatic reinforcement. That is, each participant's pica was maintained by non-socially mediated variables. Next, a series of preference assessments were conducted to identify the specific reinforcing aspect of the cigarette (e.g., the unsmoked filter) and to identify high preference items that could function as competing reinforcers. Initially, for three of the participants, the authors assessed the affect of satiation by providing preferred edible items every 10 seconds, for 5 minutes, prior to the start of a session. Next, the participants were taught to hand a cigarette to their therapist and all attempts to engage in pica were blocked. For the fourth participant, a punishment procedure was used in addition to non-contingent reinforcement to reduce the participant's pica.

In another study, Donnelly and Olczak (1990) used differential reinforcement of an alternative behavior (DRA) to compete with the cigarette pica of two adult males. Both participants were differentially reinforced for chewing sugarless mint gum. The results demonstrated that gum chewing successfully competed with the participant's pica.

Smith (1987) used verbal prompting ("don't touch") and differential reinforcement of incompatible behavior (DRI) to reduce the pica of an adult male participating in community-based supportive employment. The DRI procedure involved differential reinforcement for touching and manipulating work materials. Results demonstrated that the DRI treatment package significantly reduced the participant's pica. These data are particularly interesting because they represent the first published demonstration that a differential reinforcement intervention package can be used successfully to reduce the pica of persons with developmental disabilities in natural (community-based) setting.

Finally, Kern, Daroshta, and Adelman (2006) demonstrated that differential reinforcement for the exchange of inedible items for edible items was successful in reducing the pica of two boys with developmental disabilities. One participant also was successful in maintaining the reduction in pica across a variety of settings, including the hospital, the classroom, and the cafeteria. The second participant required continued intervention to support generalization across settings.

Although several studies have demonstrated that intervention packages including differential reinforcement can be successful in reducing pica, only one study has examined the efficacy of differential reinforcement in educational settings. To date, no published study has examined the efficacy of differential reinforcement for the production of functionally equivalent communication behavior to replace pica in an educational setting. Thus, the purpose of the present study was to examine whether this type of functional communication training (FCT) could result in functionally equivalent behaviors to replace the pica of a 6-year-old boy with autism in his classroom.



A 6-year-old boy with autism (Richard) served as the single participant in this study. Richard's diagnosis of autism had been established independently by a multi-disciplinary assessment team, including a developmental pediatrician and a psychologist. Additionally, Richard presented with global delays in receptive and expressive language, cognitive skills, and social skills. However, he could communicate his desires and needs through speech using 3- to 4-word utterances. Behaviorally, Richard displayed little awareness of dangerous situations, including the dangers of eating edible and non-edible items (e.g., rock salt) that he picked up from the floor. Since preschool, Richard had participated in classrooms that were structured according to principles of applied behavior analysis (ABA); and he had received early intensive behavioral intervention services prior to entering the preschool.


All observations were conducted in the participant's classroom over a three month period. The classroom included five other students whose educational status ranged from kindergarten through second grade. Throughout most of this study, Richard attended an ABA-based classroom with five other students, located in a district school. It offered a high level of support services (e.g., 1:1 aide for each child, speech/language therapy, occupational therapy) as prescribed by each child's Individualized Education Plan (IEP). During the summer, Richard's classroom was located in a self-contained school for children with disabilities staffed by some of the same paraprofessionals as during the regular school year under the direction of a different teacher.

The sessions were conducted by the authors, who also provided services to the participant in the classroom (i.e., behavioral consultant, speech therapist, associate teacher, and classroom teacher). Sessions were conducted in the context of different classroom activities, and the participant's 1:1 paraprofessional continued to provide support for Richard as appropriate for these activities. Initially the sessions were conducted during structured activities (e.g., discrete-trial teaching). Later, they were conducted at other times throughout Richard's school day (see below).

Preliminary Procedures

Functional assessment: Prior to the intervention, naturalistic observations of the antecedents and consequences of the behavior were conducted by the classroom behavioral consultant. Data collected did not indicate a specific pattern to the behavior (e.g., no social consequence was consistently associated with the participant's pica), and the function was therefore determined to be non-socially mediated (i.e., maintained by automatic reinforcement).

Preference assessment: A preference assessment was conducted in order to identify preferred food item(s) which could be used to "bait" the environment during the study. Bating involved placing preferred food items on the floor as a temptation for Richard to engage in pica. Results indicated that pretzels could be used for this purpose. Additionally, Richard had been observed to eat rock salt from the floor during the winter. For this reason, rock candy (which looked like rock salt) was also used at points later in this study.

Dependent Variable

Data were collected on the latency and frequency of pica and pica attempts. The specific dependent measure was the length of time between the placement of a preferred edible item on the floor near Richard and Richard's reaching for the item. A "+" was scored, for each trial, if the participant did not reach for the items on the floor for the required latency and then verbally requested the item. A "-" was scored if the participant either reached for the item before the required latency or did not request the item. Data were collected on the latency and frequency of pica and pica attempts. The specific dependent measure was the length of time between the placement of a preferred edible item on the floor near Richard and Richard's reaching for the item. A "+" was scored, for each trial, if the participant did not reach for the items on the floor for the required latency and then verbally requested the item. A "-" was scored if the participant either reached for the item before the required latency or did not request the item. The reliability of scoring was assessed by calculating the inter-observer agreement during 70% of the sessions and involving 76% of the trials. The total number of agreements was divided by the total number of agreements plus disagreements. Results indicated 98% inter-observer agreement.

Experimental Design

A single-subject changing criterion design was used to demonstrate experimental control.

Baseline: Three baseline trials were conducted during one session to measure the dependent variable before any treatment was introduced. Each trial began with the placement of preferred food (pretzels) on the floor near Richard and it ended when Richard reached for the item. Results indicated that Richard reached for the pretzels immediately after they were placed (0-second latency) on all baseline trials (i.e., 3 times).

Independent Variables: After baseline, Richard was taught to ask for pretzels. Because functional communication (requesting attention) had been taught to Richard previously as a replacement for self-injury, he was familiar with the process. As before, a verbal prompt was used to teach Richard to request the pretzel; and only one trial was necessary for him to understand the contingencies. After that he began requesting pretzels independently.

Next, Richard was told that a timer would be set and that if he wanted a food item when the timer beeped, he could ask for it. Initially, the timer was set for 1 second, and food items (initially pretzels) were placed on the floor as they were in the baseline condition; and Richard was told "don't touch". If Richard attempted to reach the food items, he was blocked and a "-" was scored for the trial. When the timer began to beep, and if Richard didn't request a pretzel (or other food item) independently, he was verbally prompted to do so. If he verbalized a request, he was reinforced with social praise and given a "clean" food item from a container. A "+" for each trial was only scored if Richard both and verbally requested the item and did not reach for the food item prior to the timer beeping. The length of time during which Richard was required to wait without touching or attempting to reach for a food item on the floor was systematically increased across trials.

Typically no more than 10 trials were conducted in one treatment session or per day. Initially, 1 session of approximately 10 trials was conducted per day for the trials with the terminal criteria of 1 to 59 seconds. On Figure I, below, this included trials 1 through 132. When the wait criteria increased to 60 seconds or more per trial, the number of trials conducted per session varied between 1 and 6 per day, depending on time constraints dictated by the participant's schedule. On Figure I, this included trials 133 through 157.

The distance between the therapist and Richard also varied across sessions. This was done both to replicate the various instructional situations which Richard experienced in his classroom and to ensure that the proximity of the therapist was not influencing his behavior. When the first intervention sessions were conducted, the therapist was seated directly across from Richard. When the latency of the trials was increased to 60 seconds, the therapist began to vary her distance from Richard systematically in order to promote stimulus generalization. During these trials, the therapist placed pretzels on the floor near Richard's feet and then stepped back.

Once the latency criteria increased to 120 seconds, the therapist started to conduct the sessions during different parts of Richard's school day. For example, sessions were conducted when Richard participated in large-group instruction conducted by his classroom teacher and during transitions between classroom activities. These were natural contexts in which Richard had displayed pica prior to the initiation of this study.

Finally, in order to evaluate the external validity of the intervention, sessions were conducted for 30 consecutive minutes across 2 days. The specific times were chosen to correspond with a variety of natural activity contexts during which Richard had been observed to display pica prior to the intervention (e.g., reading lessons, transitioning across the room). During these sessions, Richard's entire classroom was "baited" with a variety of items including crushed rock candy.



The results of this study are summarized in Figure 1, above. Richard participated in 157 trials over a 3-month period. The data show that differential reinforcement of functional communication was effective in reducing the frequency with which he engaged in pica (see Figure 1). Richard only reached for the food on the floor on 4 (3%) of the 157 trials. These occurred once at each of four criterion levels (i.e., 1-second, 4-second, 60-second, and 240-second trials, respectively). More importantly, Richard succeeded on 153 (97%) of the 157 trials in reaching the latency criteria and verbally requesting the food item rather than attempting to engage in pica.


While the results of studies with small numbers of participants must be interpreted with caution, the current results do suggest that functional communication training could be a promising procedure for helping children with autism to use functionally equivalent communication responses as a replacement for pica when their history of pica has involved food items from the floor. Additionally, since anecdotal data suggested that Richard successfully generalized his communication skills across settings (e.g., to the general education class and to transitions between activities), it is possible that other individuals with similar histories of engaging in pica also may also be successful in generalizing these skills.

As indicated above, however, one must be cautious about the generality of these results. There is great variability in skill sets among children with the diagnosis of autism, and there is variability in the environments in which they function. The participant in this study was enrolled in a highly-structured school program, with educational staff trained in ABA, and he had a previous history of using functional communication successfully as a replacement for problem behavior. Further, only specifically identified food items were used as stimuli. Therefore, it is unclear whether the current intervention package would be effective for a person with different entry-level skill sets, with a history of pica involving a variety of non-edible items, or in a classroom with less adult support. Additionally, although anecdotal reports from Richard's staff indicated that the intervention package was successful in settings other than his classroom (e.g., at home, in the community), his performance was not assessed formally in these other settings. Replication and extension of these procedures should be done to assess formally whether these results can be generalized to other participants and other environments.


The authors would like to thank Kim Lazzar, Lindsey McIlvene, Richard, and his family for their support in conducting this study.


Donnelly, D. R., & Olczak, P. V. (1990). The effect of differential reinforcement of incompatible behaviors (DRI) on pica for cigarettes in person with intellectual disability. Behavior Modification, 14, 81-96.

Fisher, W. W., Piazza, C., Bowman, L. G., Kurtz, P. F., Sherer, M. R., & Lachman, S. R. (1994). A preliminary evaluation of empirically derived consequences for the treatment of pica. Journal of Applied Behavior Analysis, 26, 23-36.

Goh, H., Iwata, B. A., & Kahng, S. W. (1999). Multicomponent assessment and treatment of cigarette pica. Journal of Applied Behavior Analysis, 32, 297-316.

Johnson, C., Hunt, F., & Siebert, M. J. (1994). Discrimination training in the treatment of pica and food scavenging. Behavior Modification, 18, 214-229.

Kern, L., Starosta, K., & Adelman, B. E. (2006). Reducing pica by teaching children to exchange inedible items for edibles. Behavior Modification, 30, 135-158.

McAdam, D. B., Sherman, J. A., Sheldon, J. B., & Napolitano, D. A. (2004). Behavioral interventions to reduce the pica of persons with developmental disabilities. Behavior Modification, 28, 45-72.

Rojahn, J., McGonigle, J., Curcio, C., & Dixon, M. (1987). Suppression of pica by water mist and aromatic ammonia: A comparative analysis. Behavior Modification, 11, 65-74.

Singh, N. N. (1983). Behavioral treatment of pica in mentally retarded persons. Psychiatric Aspects of Mental Retardation Newsletter, 2, 33-36.

Smith, M. D. (1987). Treatment of pica in an adult disabled by autism by differential reinforcement of incompatible behavior. Journal of Behavior Therapy and Experimental Psychiatry, 18, 285-288.

Author Contact Information:

Deborah A. Napolitano

University of Rochester School of Medicine

Box 671

601 Elmwood Dr.

Rochester, NY 14642

Phone: 585-273-3315


Lisa A. Blakkman

Monroe #1 BOCES

41 O'Connor Rd.

Fairport, NY 14450

(585) 377-4660

Lori B. Kohl

Monroe #1 BOCES

41 O'Connor Rd.

Fairport, NY 14450

(585) 377-4660

Heather M. Vallese

Monroe #1 BOCES

41 O'Connor Rd.

Fairport, NY 14450

(585) 377-4660

David B. McAdam

University of Rochester School of Medicine

Box 671

601 Elmwood Dr.

Rochester, NY 14642

Phone: 585-273-2759

COPYRIGHT 2006 Behavior Analyst Online
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Napolitano, Deborah A.; Blakkman, Lisa A.; Kohl, Lori B.; Vallese, Heather M.; McAdam, David B.
Publication:The Journal of Speech-Language Pathology and Applied Behavior Analysis
Geographic Code:1U2NY
Date:Dec 22, 2006
Previous Article:Stimulus equivalence as a function of training protocol in a connectionist network.
Next Article:The effects of enthusiasm on skill acquisition by children with autism.

Related Articles
Top business schools see value of communication skills.
Soil ingestion: a concern for acute toxicity in children.
Pica for foam rubber in patients with sickle cell disease. (Case Report).
Converters and students take home PICA trophies.
Bellevue University press release: (June 14, 2006).
The Army TACOM-Picatinny Center for Contracting and Commerce in Picatinny, NJ, plans a solicitation for marketing services support for the Office of...
Association of Equipment Manufacturers (AEM), names Rick Zettler, as its PR Person of the Year.
Steady as she goes: Latin America's medical device sector is keeping pace with increasing demand from the United States, Europe and Asia.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters