Delineating subtypes of self-injurious behavior maintained by automatic reinforcement

Principal Investigator: Louis Hagopian

Self-injurious behavior (SIB) is one of the most serious problems experienced by individuals with intellectual disabilities. The standard of care for the assessment of SIB involves conducting a "functional analysis." This assessment procedure involves observing the individual under several analog conditions to identify the situations in which SIB occurs consistently. In most cases, functional analysis indicates that SIB is reinforced by caregiver reactions (e.g., caregiver attention). In roughly 25% of cases, SIB levels are unaffected by social consequences - suggesting that the behavior itself produces reinforcement through unspecified processes such as sensory stimulation. The term automatic reinforcement (or "automatic SIB") has been used to describe this amorphous and highly treatment-resistant type of SIB. In contrast to the vast body of research demonstrating the clinical and heuristic value of identifying social subtypes of SIB, little effort has been aimed at delineating subtypes of automatic SIB.

We propose three subtypes of automatic SIB based on distinct clinical features observed during the functional analysis and in the context of treatment. We hypothesize these clinical features reflect distinct underlying factors related to the reinforcing (and aversive) consequences produced by SIB. The model is also informed by research on the biological basis of SIB, including altered pain sensitivity. For one subtype, the rates of SIB decrease as a function of the level of stimulation in the environment (Subtype I). For this group we posit that SIB produces sensory stimulation that is moderately reinforcing in that it can be readily overcome by other sources of reinforcement (e.g., toys). For other individuals, however, SIB is highly persistent and occurs irrespective of the level of environmental stimulation (Subtype II). We propose that for this group, SIB produces more highly potent biologically reinforcing consequences. Finally, some individuals with automatic SIB also engage in self-restraint, a behavior that is incompatible and actively prevents the occurrence of SIB (Subtype III). We propose that for these individuals, SIB also produces aversive (e.g., potentially painful) consequences, which negatively reinforces self-restraint because it prevents SIB.

Analysis of pilot clinical data obtained from 39 cases with automatic SIB provides some preliminary support for the model. The proposed model provides the conceptual basis for subtype- specific and testable hypotheses. Although we cannot directly observe internal processes underlying automatic SIB, we will perform analyses that can help us understand what may underlie these clinically distinct subtypes. We will employ a range of assessment techniques and measures to examine how reinforcing (or aversive) SIB is in 60 individuals with the behavior. The current proposal represents an initial step of subtyping automatic SIB - one that will set the stage for future research aimed at understanding the biological underpinnings of each subtype, and for clinical trials evaluating more targeted behavioral and pharmacological interventions.