Comparative Analysis of Self-Injurious Behaviour Topographies in Young Children with and without Developmental Delay

Abstract

Background:

The present study extended the findings of Hoch et al. (2016) by conducting post hoc analyses of specific topographies of self-injurious behaviour (SIB) exhibited by young children with developmental delay (DD) and children with typical development (TD). We conducted these analyses to better understand similarities and differences between the groups from a developmental perspective. No previous study has compared the prevalence, severity and co-occurrence of specific topographies of SIB in young children.

Method:

The participants were parents of two groups of children one with DD (n=49, mean age = 37.5 months) and one with TD (n=49, mean age = 36.6 months). Individual items of the SIB subscale from the Repetitive Behavior Scale – Revised (Bodfish et al., 2000) were used in the analyses.

Results:

Seven of the eight RBS-R SIB categories were reported for both groups. Children in the DD group were significantly more likely to engage in Hits Self against Surface or Object, Hits Self with Body Part, Inserts Finger or Object, Skin Picking and Bites Self. Parental ratings of severity were also significantly greater for the DD group for these five topographies. The DD group engaged in a significantly greater number of SIB topographies than the children in the TD group. Children in the TD group were more likely to exhibit a single SIB topography while the DD group were more likely to engage in two or more topographies. Topographies involving self-hitting were not only more frequent among the children in the DD group but also more likely to be rated as moderate or severe in nature.

Conclusions:

Compared to the TD group, the topographies of SIB exhibited by the DD group were more prevalent, more severe, and co-occurred with greater frequency. Inclusion of a group of children with TD provided an important comparative context for the occurrence of SIB in children with DD.

Despite decades of research, self-injurious behaviour (SIB) remains a challenging clinical problem for people with intellectual disability (Shawler et al., 2019). Longitudinal research attests to the chronicity of SIB as well as its resistance to treatment (Taylor, Oliver, & Murphy, 2011). Current treatment approaches are reactive in nature, initiated long after the behaviour is already established (Richman, 2008), because little is known about the pathway(s) responsible for its development (Richards & Symons, 2018).

Early studies of SIB in children with developmental disabilities focused on populations with severe SIB requiring specialized treatment in medical settings (Hyman, Fisher, Mercugliano, & Cataldo, 1990; Kurtz et al., 2003). A range of SIB topographies was evident among the participants in these studies such as head-banging, head-hitting, self-biting, and body-hitting. Moreover, many of the participants were reported to exhibit multiple topographies of SIB. These studies were important because they showed that severe SIB did occur in childhood (Hyman et al., 1990) and that the mean age of onset was less than two years of age (Kurtz et al., 2003). Other studies focused on precursors of SIB among children with intellectual disability (Berkson, Tupa & Sherman, 2001; Murphy, Hall, Oliver, & Kissi-Debra, 1999; Richman & Lindauer, 2005) or factors associated with the occurrence of SIB in young children at risk for developmental delay (DD) (Kurtz, Huete, Cataldo, & Chin, 2012). Subsequent studies conducted with community samples of children with DD sought to determine the prevalence of SIB in children younger than 5 years of age (Hoch, Spofford, Dimian, Tervo, MacLean, & Symons, 2016; MacLean, Tervo, Hoch, Tervo & Symons, 2010; MacLean & Dornbush, 2012; Mayo-Ortega et al., 2012).

In the only direct comparison of SIB in preschool-aged children with and without DD, 59.1% of the children with DD exhibited at least one form of SIB compared with 28.5% of the children with typical development (TD) (Hoch, et al., 2016). These comparisons were based on the occurrence of any form of SIB and not differences in SIB expression (such as specific topographies) – some of which may occur in both groups and some of which may be unique to a single group (MacLean, Stone & Brown, 1994). Hoch et al. also reported that the total SIB subscale score was significantly greater for the DD group in comparison with the TD group.

Without analyses of specific topographies it is not possible to know whether the differences were apparent for all of the scale items or whether there were particular topographies that contributed to the overall difference. We report a post hoc analysis of the original data (Hoch et al.) specifically investigating SIB topographies (i.e. hitting, biting, skin-picking) in terms of their prevalence, co-occurrence and severity in two groups of children, one with DD and one without.

Method

The participants and measures used in this study have been fully described elsewhere (Hoch et al., 2016). Briefly, we compared parental ratings of a group of children with DD (n = 49) and a contrast group of TD children (n = 49) using data from the Repetitive Behavior Scale – Revised (Bodfish, Symons, Parker, & Lewis, 2000).

Procedures

The SIB subscale of the Repetitive Behavior Scale-Revised (Bodfish, et al., 2000) has eight items that are defined by discrete actions (hits self or bites self), as opposed to specific body locations (e.g., head banging or hand-biting). The items are: Hits Self with Body Part, Hits Self against Surface or Object, Hits Self with Object, Bites Self, Pulls, Rubs or Scratches Self, Inserts Finger or Object, and Skin-picking. Responses are based on observation and interaction with the child over the past month. Items are rated on a four- point scale (behaviour does not occur = 0, behaviour occurs and is a mild problem = 1, behaviour occurs and is a moderate problem = 2, behaviour occurs and is a severe problem = 3). In assigning a score for each item, respondents are instructed to consider: (a) how frequently the behaviour occurs (e.g. weekly versus hourly), (b) how difficult it is to interrupt the behaviour (e.g. can be easily redirected versus becomes distress if interrupted), and (c) how much the behaviour interferes with ongoing events (e.g., easy to ignore or overlook versus very disruptive). Previous psychometric studies have shown that the RBS-R has high rates of internal consistency and interrater reliability for children with developmental disabilities (Lam & Aman, 2007; Rojahn et al., 2013). In the absence of published reliability data for the TD group, we calculated Cronbach’s alpha to assess the internal consistency reliability of the SIB items subscale for that group. The alpha value for the eight items was 0.74 indicating that the SIB subscale items had acceptable internal consistency reliability (Ferketich, 1991).

Results

Prevalence of SIB Topographies:

Figure 1 presents the point prevalence estimates by group for each of the eight SIB topographies. Hits Self against Surface or Object was the most frequently reported for the DD group (n=19) and Rubs or Scratches Self was the most prevalent form for the TD group (n=5). The topography receiving the fewest number of endorsements for the DD group were Hits Self with Object and Bites Self (n=5). None of the children in the TD group were reported to engage in Bites Self.

Figure 1.

Prevalence of parent-reported SIB topography by group, defined as nonzero scores, from the Repetitive Behavior Scale-Revised. Bars reflect the number of mild (white), moderate (gray) and severe (black) problem ratings for each topography. DD denotes the values for the developmental delay group while TD represents the values for the children in the typically developing group. Significant differences are noted with an asterisk.

For five of the eight SIB topographies, the differences in prevalence between groups were statistically significant. In each instance the behaviour was more prevalent in the DD group as compared to the TD group. Those topographies were: Hits Self against Surface or Object (X²=17.51, p<.001, ϕ= 0.422), Hits Self with Body Part (X²=4.34, p=.037, ϕ=0.210), Inserts Finger or Object (X²=4.34, p=.037 ϕ=0.210,), Skin Picking (X²=4.90, p=.026, ϕ=0.223), and Bites Self (X²=5.27, p=0.021, ϕ=.231). Phi (ϕ) is a measure of effect size for a 2X2 contingency table. A value of ϕ= 0.1 is considered to be a small effect, 0.3 a medium effect, and 0.5 a large effect. The prevalence for Hits Self with Object, Pulls Skin or Hair, and Rubs or Scratches Self was not significantly different between the two groups.

Number of topographies.

We also calculated the total number of SIB topographies reported for each child. Our analyses revealed that seven of the children in the DD group and nine of the children in the TD group exhibited one topography whereas 22 of the children in the DD group and five of the children in the TD group exhibited two or more topographies. Children in the TD group were more likely to exhibit a single SIB topography (64%) while children in the DD group were more likely to engage in two or more topographies (76%; X²=6.51, p=.010, ϕ =0.389). The average number of SIB topographies reported for the DD group was 2.52 (range 1 to 7) and 1.43 for the TD group (range 1 to 3). This difference was statistically significant, t(41) = 3.08, p < .001, d = 0.80.

Table 1 presents the topographic distribution of single occurrences of SIB by group. Three differences were noted between groups. First, there was overlap for only two of the eight types of SIB, Hits Self with Body Part and Inserts Finger or Object. Second, single instances of SIB for the DD group frequently involved hitting (5/7) while the TD group engaged in topographies involving pulling, rubbing, scratching or inserting fingers or objects (7/9). Finally, self-biting was the only topography that did not occur in isolation.

Table 1.

Number of Children Engaging in a Single SIB Topography by Group

Topography	Developmental Delay	Typically Developing
Hits self against surface or object	4	0
Skin picking	1	0
Hits self with body part	1	1
Inserts finger or object	1	3
Pulls hair or skin	0	2
Rubs or scratches self	0	2
Hits self with object	0	1
Bites self	0	0
Total	7	9

Table 2 presents the co-occurrence of the eight SIB topographies by group for those children exhibiting two or more topographies. For children exhibiting more than two topographies, all possible combinations are reflected in the table. The DD group values are below the diagonal and co-occurrence among the TD group is above the diagonal line. Each value in the table represents the co-occurrence of two topographies for one child. All of the possible pairwise combinations were evident among the 22 children in the DD group for whom two or more topographies were reported. The most frequently co-occurring behaviours were Hits Self with Body Part and Hits Self against Surface or Object (n=8). Hits Self against Surface or Object also occurred frequently with Inserts Finger or Object (n=6), Bites Self (n=5), and Rubs or Scratches Self (n=5). Three pairings occurred only once. They were Hits Self with Body Part and Inserts Finger or Object, Skin Picking and Hits Self with Object and Skin Picking and Pulls Hair or Skin. Co-occurrence data for the TD group are presented above the diagonal line in Table 2 for comparison. In each instance of co-occurrence only one child in the TD group engaged in both forms of SIB.

Table 2.

Frequency of SIB Co-Occurrence for Individual Children in the DD Group (below diagonal line) and TD Group (above diagonal line)

	Hits Self with Body Part	Hits Self against Surface or Object	Hits Self with Object	Bites Self	Pulls Hair or Skin	Rubs or Scratches Self	Inserts Finger or Object	Skin Picking
Hits Self with Body Part		1	1
Hits Self against Surface or Object	8		1			1
Hits Self with Object	3	4				1
Bites Self	4	5	3
Pulls Hair or Skin	4	4	4	2		1
Rubs or Scratches Self	3	5	2	3	3			1
Inserts Finger or Object	1	6	3	2	3	4
Skin Picking	2	3	1	2	1	5	4

Severity of SIB Topographies:

Figure 1 disaggregates the overall subscale data by topography. For the DD group, the behaviour receiving the greatest percentage of moderate and severe endorsements were Hits Self with Object (80%), Skin picking (71%), Hits Self with Body Parts (60%) and Hits Self against Surface or Object (47%). For the TD group, there were no moderate or severe endorsements for any topography. Table 3 provides the mean scores by group for each topography. Parent ratings were significantly greater for children in the DD Group than those of the TD Group for five of the eight topographies.

Table 3.

Mean RBS-R SIB Item Scores by Group

Topography	DD M (SD)	TD M (SD)	Cohen’s (d)a	t(96)	P
Hits self against surface or object	.63 (.92)	.04 (.20)	.886	4.36	.000
Hits self with body part	.35 (.75)	.06 (.24)	.521	2.53	.014
Skin picking	.29 (.76)	.02 (.14)	.678	2.39	.019
Inserts finger or object	.27 (.60)	.06 (.24)	.459	2.19	.031
Bites self	.16 (.55)	.00 (.00)	.411	2.06	.042
Pulls hair or skin	.22 (.62)	.06 (.24)	.340	1.71	.090
Hits self with object	.22 (.71)	.06 (.24)	.302	1.51	.133
Rubs or scratches self	.27 (.70)	.10 (.30)	.316	1.49	.140

^aCohen’s d statistics of ≥ 0.20, = 0.50, and = 0.80 can be interpreted as small, medium and large effect sizes, respectively.

Discussion

The purpose of this study was to extend the findings of Hoch et al. (2016) by comparing the prevalence, co-occurrence and severity of specific SIB topographies between two groups of children, one with DD and one without. Topographies involving hitting, inserting fingers or objects, skin-picking or self-biting were not only more prevalent among the DD children, these behaviours were also rated as more severe by their parents. Recognizing that difference in prevalence might play a role in the observed differences in severity, we focused on the ratings of moderate or severe. In this analysis, the behaviour receiving the highest proportion of moderate or severe ratings in the DD Group was Hits Self with Object (80%) – a behaviour that did not differ in prevalence from the typically developing group. Skin-picking (71%), Hits Self with Body Parts (60%), and Hits Self against Surface or Object (47%) also received a high proportion of moderate and severe ratings.

Self-biting was the only SIB topography not reported by parents of children in the TD group. In a previous study, 16.3% of TD children were reported to engage in self-biting (Roane, Ringdahl, Vollmer, Whitmarsh, & Marcus, 2007). The children in that study were similar to those in the current study in terms of chronological age (M=37.3 months). However, the measures were substantially different. Roane et al. defined self-biting as “closing lower and upper teeth around the child’s hand, arm, fingers, or other body parts.” The respondents were asked to rate how frequently the behaviour had occurred over the past two months, ranging from rarely to almost hourly. In contrast, the RBS-R requires ratings on a four point scale, ranging from “behaviour does not occur” to “behaviour occurs and is a severe problem.” Given the structure of the RBS-R – it is possible that Self-Biting did occur among the TD children but was not considered a problem by the respondents.

Previous studies have also reported multiple topographies of SIB among young children with or without developmental disabilities. The number of topographies exhibited by young children with developmental disabilities have varied with the type of population. Berkson et al. (2001) and Richman and Lindauer (2005) reported an average of 1.76 and 2.6 for children recruited from early intervention programs while Kurtz et al. (2003) and Kurtz et al. (2012) reported an average of 3 and 4.3, respectively, among young children with a variety of developmental disabilities referred for specialized services. The obtained value for the current study, 2.52 was well within the range of those studies. Similarly the obtained value for the TD Group, 1.43, was comparable to the value of 1.6 reported by Roane et al. We also compared the number of children exhibiting a single topography with the number of children exhibiting two or more topographies for the DD and TD groups. Children in the TD group were more likely to exhibit a single SIB topography while the DD group were more likely to engage in two or more topographies. No previous study has reported such an analysis.

Given that a substantial number of children in each group exhibited two or more SIB topographies, we examined co-occurrence of the behaviour. With regard to the DD group, pairwise comparisons indicated that Hits Self with Body Part and Hits Self against Surface or Object was the most frequent pairing followed by Hits Self against Surface or Object and Inserts Finger or Object, Bites Self, and Rubs or Scratches Self. Although previous studies of children with intellectual disability have reported the number of children that engaged in multiple forms of SIB, co-occurrence was not reported. However, two studies provided the individual data to calculate pairwise co-occurrence and permit comparisons with the current study. Eight of the 21 children reported by Berkson et al. exhibited two or more topographies. Six of those children exhibited Head-hitting and Head banging. In a study by Richman and Lindauer (2005), six children out of 12 exhibited Head Hit and Hand Mouthing, four children exhibited Head Bang and Hand Mouthing, and four children exhibited Head Hit and Head Bang. In the current study, the most common co-occurrence of SIB topographies was Hits Self against Surface or Object and Hits Self with Body Part – both of which may have included the child’s head.

This is the first comparative study of SIB topographies between children with and without developmental delay using a common measure of repetitive behaviour. The analyses suggest three features worthy of further consideration. First, parents of children with DD rated five topographies as more problematic than the parents of typically developing children. Factors that defined ‘problematic’ were how frequently the behaviour occurred, how difficult it is to interrupt the behaviour, and how much the behaviour interferes with ongoing events. Three topographies stand out in this regard because more than half of the ratings were in the moderate or severe category – namely Hits Self with Objects, Skin Picking, and Hits Self with Body Parts. Second, the greatest absolute difference in prevalence between the two groups was for Hits Self against Surface or Object. Nineteen children in the DD group were reported to engage in that behaviour while only two of the typically developing children engaged in that behaviour. Moreover, 47% of the ratings for the DD Group were moderate or severe for that topography. Finally, the concentration of moderate to severe ratings in topographies associated with self-hitting as compared with other topographies was noteworthy.

The findings of the current study have clinical implications that warrant further consideration. First, from a practical standpoint, if SIB occurs and adults (parents, teachers, day-care providers) consider it a moderate or severe problem, then it is time to seek clinical assessment given the observation that degree of concern expressed about a child’s SIB was predictive of increases in SIB in a longitudinal study of young children (Murphy et al., 1999). Second, SIB that is directed toward the head is particularly concerning given the findings reported by Emerson et al. (2001) regarding the persistence of head hitting/banging over time and by inference, its resistance to treatment once established. Third, SIB topographies considered most problematic by parents and other informants on the RBS-R may be those forms that are the most difficult to ignore. By extension, those topographies might be most likely to be reinforced by social consequences and thus more likely to be maintained over time (Emerson et al, 2001; McEvoy & Riechle, 2000). Finally, while many young children are likely to engage in SIB for at least a brief period of time, a behaviour that persists over a few months or produces demonstrable physical injury warrants clinical assessment.

This study shares the same limitations described in Hoch et al. (2016). However, given the explicit focus of the current analyses on SIB topographies, readers will recognize that the RBS-R focuses on discrete actions (e.g. hitting, biting, or scratching) rather than the specific part of the body involved in the behaviour (e.g. head, face, or arm). As a result, it is difficult to compare our results to studies that identify specific body parts as the target of the self-injurious behaviour. Moreover the RBS-R rating system does not include an index of physical consequence (injury) as a measure of severity. Our findings suggest that future researchers should consider alternative tools to measure frequency and severity in a DD and TD population given the limitations of the RBS-R.

Conclusion

Since Tate and Baroff’s (1966) seminal article on SIB, researchers have typically focused on the outcome of the behaviour rather than the behaviour itself. Moreover, there has been a tendency to lump various forms of SIB into a unitary phenomenon – thereby missing an opportunity to expand our understanding and track the emergence rather than the persistence of SIB over time. The results of this study suggest that there may be value in focusing on specific topographies to address fundamental questions regarding the origin and maintenance of behaviour that results in self-injury.