Constipation associated with self-injurious and aggressive behavior exhibited by a child diagnosed with autism.

Abstract

A functional analysis was conducted to identify the role environmental variables had on the maintenance of self-injury and aggression. At the outset of the evaluation, an abdominal x-ray showed a moderate to large amount of stool throughout the colon (i.e., constipation). Consequently, medication was administered to promote bowel emptying. Initial high rates of self-injury and aggression reduced to near-zero levels following bowel cleanout, yielding undifferentiated functional analysis results.

**********

Functional analyses frequently result in the identification of variables maintaining problem behavior. However, antecedent conditions that influence behavior can sometimes make interpretation of functional analysis results difficult, or, might require idiosyncratic methods of evaluation. A number of studies have demonstrated that biological events can be one such antecedent variable. For example, O'Reilly (1997) reported episodic self-injurious behavior (SIB) that was exhibited by a young child with intellectual disabilities only when otitis media was present. Similarly, O'Reilly (1995) demonstrated that sleep deprivation resulted in higher levels of escape-maintained problem behavior for one individual. Kennedy and Meyer (1996) reported that biological events on their own, or in combination, can affect functional analysis outcomes. Specifically, an escape function was identified only when allergy symptoms were present with one participant while an increase in escape-maintained problem behavior was observed when a second participant was sleep deprived. A third individual demonstrated a general increase in problem behavior across all conditions of the functional analysis when sleep deprived, with the highest rates occurring in the escape condition. Collectively, three antecedent-behavior relationships were demonstrated: 1) an increase in problem behavior during a specific test condition when a biological condition (i.e., sleep deprivation, otitis media, allergy symptoms) was present, 2) no problem behavior when the biological condition was not present, and 3) problem behavior elevated across functional analysis conditions when the biological condition was present.

One biological condition that might also affect problem behavior exhibited by individuals with intellectual disabilities (ID) is constipation. Prevalence rates for constipation experienced by children with ID ranges from 50 to 75% (Del Giudice et al., 1999; Tse, Leung, Chan, Sien & Chan, 2000), suggesting this condition is especially problematic in the ID population. Previous researchers have implicated constipation as one variable affecting problem behavior. Bosch, VanDyke, Smith, and Poulton (1997) evaluated 25 individuals who exhibited problem behavior and found that 7 had untreated medical problems that are often associated with pain. Of those 7, 6 individuals were diagnosed with a gastrointestinal problem, including 2 with constipation. In the current case example, the effect of constipation on the SIB and aggression exhibited by a young boy with autism was evaluated. Specifically, the levels of SIB and aggression exhibited as a bowel cleanout proceeded were measured.

Method

Participant and Setting

One individual participated in this evaluation. Jerry was 7-years-old and diagnosed with autism, mental retardation, and had a history of severe constipation. He did not use vocal communication strategies and only exhibited a few gestures and signs. Because of the intensity, severity, and frequency of his behavior, an outpatient behavioral clinic team referred Jerry for inpatient intensive behavioral services. At the time of the evaluation, Jerry had been engaging in SIB and aggression several times per day resulting in injury to himself and his parents. In addition, his parents were unable to identify consistent antecedents or consequences related to episodes of problem behavior.

All phases of this evaluation were conducted while Jerry was admitted to an inpatient unit specializing in assessment and treatment of severe problem behavior. Functional analysis sessions took place in a therapy room (6m x 8m) that contained a table, chairs, and stimuli relevant to the various assessment conditions (i.e., preferred and neutral items and task materials). Three to four, 45-min observations were conducted each day (except day 7 of admission; no sessions were run that day to allow the family to observe a religious holiday). Each 45 min observation was further broken into 5 min sessions. With time between each session for preparation (e.g., changing session-related stimuli), each 45 min observation allowed for 3 to 5 sessions.

Data Collection and Interobserver Agreement (IOA)

Aggression consisted of pinching, biting, and head butting the therapist (or, others in the environment). SIB consisted of face slapping and head banging. These responses were measured using 6-s partial interval recording and were recorded as percentage of 6-s intervals. Interobserver agreement was collected during 23% of all sessions. Agreement was calculated by comparing observers' records on an interval-by-interval basis. Total agreements was divided by agreements + disagreements and multiplied by 100%. Combined agreement scores for SIB and aggression averaged 97% (range, 67-100%) across sessions.

Procedures

Assessment and treatment of bowel status. Prior to admission, Jerry was reportedly having daily stools. He was already on a laxative, Miralax 17 g twice daily, at the time of admission. On admission, his abdominal examination was normal. However, given Jerry's history of severe constipation and continued behavior problems, an abdominal x-ray was obtained on the first day of admission and repeated on the seventh day of admission. In addition, the presence of daily stools is not necessarily an indicator of normal bowel status. Afzal et al. (2003) reported no significant differences in the number of bowel motions per week among radiologically constipated children compared with nonconstipated children. Jerry's initial x-ray was read as a moderate amount of soft stool throughout the colon. Because of the amount of stool present, and the fact that he slept better after three large stools that occurred on the first day of his inpatient stay, one tablet of Senokot, a stimulant, was added to the Miralax on the second day of admission. Following this change in the medication regimen, nursing staff reported that Jerry had one to five stools daily which were moderate to large in size and soft in consistency. On discharge, the stool load was decreased. It is of note that the stool volume was not identified based on the history and physical exam.

Assessment of problem behavior. To identify potential reinforcers maintaining problem behavior, a functional analysis was conducted as Jerry's bowel cleanout took place. The functional analysis was based on Iwata et al. (1982/1994) and was conducted using a multielement design. Sessions were 5 min in length and alternated through out a 45-minute observation. Both aggression and SIB resulted in the programmed consequences during the test conditions. Functional analysis conditions included control, contingent attention, contingent tangible, and contingent escape. An attempt was made to assess each condition each day. However, on Day 2, Jerry did not engage with any alternative stimuli in the therapy room, thus the tangible condition was not conducted. During the control condition, Jerry received continuous access to adult attention and had ongoing access to preferred items (previously identified via a free operant preference assessment; Roane, Vollmer, Ringdahl, & Marcus, 1998); no demands were made. During the contingent escape condition, a task based on Jerry's current IEP was presented on an ongoing basis. A 30-s break from the task was provided contingent on the occurrence of aggression or SIB. Neither adult attention nor leisure items were available during breaks. During the contingent attention condition, moderately preferred activities were continuously available and adult attention was initially diverted. Contingent on aggression or SIB, the adults in the room interacted with Jerry for about 30 s. During the contingent tangible condition, low preferred items and adult attention were continuously available. Access to high-preferred tangible items was initially restricted. Contingent on aggression or SIB, Jerry was provided with 30 s of access to the high-preferred items.

Communication Training. Communication Training (CT) took place after the functional analysis concluded and was conducted as a general treatment. During these sessions, Jerry's microswitch activation resulted in access to preferred activities and items, including food. This response-consequence relationship was chosen as the focus of communication training because anecdotal reports from the nursing and behavioral staff indicated that SIB and aggression were sometimes observed when Jerry transitioned past the kitchen/dining area of the inpatient unit and when preferred items and activities were unavailable (e.g., transitioning away from television programs). In addition, given Jerry's limited communication repertoire, this focus was agreed upon by the clinical team and Jerry's parents. Communication training sessions began by allowing Jerry access to preferred items. After a minute or two of free access, items or edibles were restricted and Jerry was prompted to activate a microswitch to obtain the items/edibles. Following appropriate microswitch activation (i.e., microswitch activation in the absence of problem behavior), the preferred item was delivered for 30 s or 1 edible was delivered. If problem behavior occurred during the reward consumption interval, the item was removed. Communication training sessions were alternated with control sessions during which noncontingent access to preferred items was available.

Results

Figure 1 displays the results of Jerry's abdominal x-ray. The x-ray shows a large amount of stool throughout the colon, particularly the descending colon. Consequently, medication was administered to begin bowel cleanout. The figure shows stool entering the colon on the right, proceeding up the ascending colon, across the transverse colon, and down the descending colon. Soft stool is seen in the ascending colon (shown as white/grey and fuzzy in appearance), especially near the hepatic flexure (liver). Firmer stool is noted in the descending colon, including some harder balls (shown as round and white in appearance). Using Barr's score for radiographic assessment of stool load (1979), Jerry had a score of 10 (1) (stool retention) on admission. Using the criteria of Blethyn et al. (1995), Jerry demonstrated Grade II (moderate) retention (2).

Figure 2 displays the results of the early portion of the functional analysis (days 1-6). Intermittent high levels of problem behavior were observed across all conditions. Mean levels of SIB and aggression were 9.8% of intervals during the control condition (range, 0-52% of intervals), 3.8% of intervals during the attention condition (range, 0-32% of intervals), 5.7% of intervals during the escape condition (0-34% of intervals), and 4.8% of intervals during the tangible condition (range, 0-20% of intervals). These results suggested that problem behavior was not sensitive to environmental contingencies and are consistent with an automatic reinforcement function interpretation. Of note, the general trend of problem behavior was downward. Specifically, if the levels of problem behavior during the first 2 days of admission are compared to the 2 days prior to the second x-ray, a reduction of 92% is noted (from an average of 13.6% of intervals across conditions to an average of 1% of intervals across conditions). One potential reason for this decrease in problem behavior was a decrease in Jerry's constipation.

Figure 3 displays the results of the second abdominal x-ray obtained on the eighth day of admission. More air was seen (the black areas) and the stool volume is decreased, and almost completely soft. Using Barr's score for radiographic assessment of stool load (1979), Jerry achieved a score of 5 (no retention) on discharge. Using the criteria of Blethyn et al. (1995), Jerry demonstrated Grade I (mild) retention on the eighth day.

Figure 4 displays the results of the functional analysis, post bowel cleanout (days 8 and 9). Overall, lower levels of problem behavior were observed. Problem behavior was observed during an average of 1.4% of intervals across all conditions. The level of problem behavior in each condition approached and reached 0% by the end of day 9.

Figure 5 displays the results of the communication training portion of Jerry's evaluation. Problem behavior continued to occur at low levels during CT sessions (M = 0.4% of intervals, range = 0-8% of intervals) and control conditions (M = 0.6% of intervals, range = 0-6% of intervals). Low levels of problem behavior during CT maintained over a 4-day period. Microswitch activation was exhibited during the CT sessions in a functional manner.

Figure 6 displays each the entirety of data in a different format. Each bar represents that day's average responding for a particular condition. Displayed in this manner, the overall downward trend across days is apparent.

Discussion

As Jerry's bowel status improved, the overall level of problem behavior decreased, eventually reaching near-zero levels. From a clinical standpoint, this outcome was beneficial because Jerry's behavior was more easily controlled. From a conceptual standpoint, this outcome was troubling because the operant mechanism responsible for maintaining Jerry's problem behavior was not able to be identified, as there was an undifferentiated response pattern during the functional analysis of problem behavior.

These results support recent findings reported by Carr and Owen-DeSchryver (2006) that problem behavior exhibited by individuals with developmental disabilities is increased when certain biological variables are present. Specifically, pain and discomfort were identified as mediating variables for the display of problem behavior such as SIB and aggression. The obtained results are also similar to previous investigations regarding the interaction between biological variables and behavior. Similar to O'Reilly (1995), a change in the presence of the biological condition of constipation was correlated with the presence of problem behavior. However, unlike similar previous studies, the presence of the biological condition did not clarify or otherwise help identify the function of problem behavior. Thus, there is not sufficient evidence to clarify the functional relationship between SIB and aggression, and Jerry's constipation. O'Reilly (1995; 1997) and Kennedy and Meyer (1996) demonstrated that the biological conditions of sleep deprivation, otitis media and allergy symptoms increased the likelihood of escape-maintained behavior. Results such as those described by O'Reilly and Kennedy and Meyer suggest that biological conditions can serve as motivating operations (specifically, establishing operation). Because the specific contingency responsible for maintaining Jerry's problem behavior were never identified, it is difficult to state definitively the function of constipation.

Another limitation of the current study is that, while there was an experimental design in place to evaluate the impact of social reinforcers on Jerry's SIB and aggression, the data with respect to constipation and behavior problems are corelational in nature because constipation was not experimentally manipulated. Evaluating the relationship between biological conditions such as constipation and severe behavior problems is not always conducive to the type of systematic manipulation necessary to achieve experimental control. Unlike sleep deprivation, constipation status will not vary on a day-to-day basis. As well, once the constipation has been identified, medical interventions are typically applied to rectify the situation. The data presented here, then, should be viewed as preliminary. However, given the relatively high prevalence of constipation in the population of individuals with intellectual disabilities, the current data provide evidence that practitioners might want to consider assessing for constipation when working with individuals who exhibit severe behavior problems.

The findings of this study suggest a number of avenues for future research. First, future studies could be focused on adapting single-subject experimental designs to study this particular biological/behavior interaction. One strategy would be to compare functional analyses using a nonconcurrent multiple baseline design (Kennedy, 2005). One drawback to using such a design is that it might be difficult to find a sufficient number of participants experiencing both constipation and severe behavior problems. Second, future studies could continue to focus on identifying whether constipation influences identifiable operant relationships between problem behavior and the environment. For example, it might be that an operant relationship between behavior problems and escape is only observed when the individual is constipated. In such an example, elevated levels of behavior problems might only be observed when two criteria are met: a) the individual is constipated, and b) instructions are delivered. Finally, future studies could study the impact of biological condition such as constipation on treatment effectiveness. Specifically, future research could evaluate whether behavior-based treatments more susceptible to integrity challenges (i.e., more likely to fail) when adverse biological conditions are present.

Acknowledgment

The authors would like to thank the staff of the Biobehavioral Service and nursing staff of the Center for Disabilities and Development for their assistance in the completion of this assessment.

Notes

(1) Barr's scores for radiographic assessment range from 2-25. In general, scores that are < 10 are negative for stool retention, while scores > 10 are positive for stool retention.

(2) Blethyn et al. (1995) grades stool retention on a I-IV level. Grade I indicates feces only in the rectum and caecum (i.e., at the beginning of the ascending colon and in the rectal vault). Grade II indicates the same as Grade I, with the addition of feces on and off throughout the colon. Grade III indicates continuous feces throughout the colon. And, Grade IV indicates continuous feces throughout the colon plus dilated colon and impacted rectum.

References

Afzal, N., Murch, S, Thirrupathy, K., Berger, L., Fagbemi, A., & Heuschkel, R. (2003). Constipation with acquired megarectum in children with autism. Pediatrics, 112, 939-942.

Barr, R., Levine, M., Wilkinson, R., Mulvihill, D (1979). Chronic and occult stool retention: A clinical tool for its evaluation in school-aged children. Clinical Pediatrics 18(11) 674-9.

Blethyn, A., Verrier Jones, K., Newcombe, R., Roberts, G., Jenkins, H. (1995). Radiological assessment of constipation. Archives of Disease in Childhood, 73, 532-33.

Bosch, J., VanDyke, D., Smith, S., & Poulton, S. (1997). Role of medical conditions in the exacerbation of self-injurious behavior: an exploratory study. Mental Retardation, 35, 124-30.

Carr, E. G., & Owen-DeSchryver, J. S. (2006). Physical illness, pain, and problem behavior in minimally verbal people with developmental disabilities. Journal of Autism and Developmental Disorders, 37, 413-424.

Del Giudice, E., Staiano, A., Capano, G., Romano, A., Florimonte, L., Miele, E., et al. (1999). Gastrointestinal manifestations in children with cerebral palsy. Brain and Development. 21, 307-311.

Iwata, B. A., Dorsey, M. F., Slifer, K.J., Bauman, K. E., & Richman, G. S. (1994). Toward a functional analysis of self-injury. Journal of Applied Behavior Analysis, 27, 197-209. (Reprinted from Analysis and Intervention in Developmental Disabilities, 2, 3-20, 1982).

Kennedy, C. H., & Meyer, K. A. (1996). Sleep deprivation, allergy symptoms, and negatively reinforced problem behavior. Journal of Applied Behavior Analysis, 29, 133-135.

Kennedy, C. H. (2005). Single-case designs for educational research. Allyn and Bacon: Boston.

O'Reilly, M. F. (1995). Functional analysis and treatment of escapemaintained aggression correlated with sleep deprivation. Journal of Applied Behavior Analysis, 28, 225-226.

O'Reilly, M.F. (1997). Functional analysis of episodic self-injury correlated with recurrent otitis media. Journal of Applied Behavior Analysis, 30, 165-167.

Tse, P., Leung, S., Chan, T., Sien, A., & Chan, A., (2000). Dietary fibre intake and constipation in children with severe developmental disabilities. Journal of Paediatrics Child Health. 36, 236-239.

Tory J. Christensen, Joel E. Ringdahl, Joni J. Bosch, Terry S. Falcomata, Jeffrey R. Luke, and Marc S. Andelman

University of Iowa 

Correspondence to Tory J. Christensen, 100 Hawkins Drive, Center for Disabilities and Development, Room 340, Iowa City, IA 52242; email: [e-mail]