Response inhibition and response re-engagement in attention-deficit/hyperactivity disorder, disruptive, anxious and normal children

Abstract

The purpose of this study was to determine whether attention-deficit/hyperactivity disorder (AD/HD) is uniquely related to impairments in two aspects of executive functioning: (1) response inhibition; and (2) response re-engagement. AD/HD (n=10), disruptive (n=11), anxious (n=11) and normal children (n=21) were compared on the change task. Children were in the age range of 8–12 years. The psychopathological groups were recruited from special educational services. Parent, teacher and child questionnaires were used to select children with pervasive disorders. Controls attended normal classes and scored low on all questionnaires. Compared with normal children, both AD/HD and disruptive children showed poor response inhibition, but only AD/HD children exhibited a deficit in the underlying inhibitory process. Some evidence was found for enhanced response inhibition in anxious children. Both AD/HD and disruptive children demonstrated higher variability in the speed of the response re-engagement process and were less accurate. The results suggest that AD/HD involves a more pervasive impairment in cognitive functioning, rather than a deficit restricted to the powers of response inhibition.

Introduction

Executive functions subserve the regulation of goal-directed behaviour. A fundamental aspect of these functions is the ability to inhibit inappropriate responding 5, 12, 22, 35, 51. Abnormalities in response inhibition are a central component in the description and explanation of child psychopathological disorders and, in particular, of attention-deficit/hyperactivity disorder (AD/HD).

Recently, a number of theoretical models have emerged in which response inhibition deficits have been suggested to underlie AD/HD 4, 5, 13, 29, 35, 37, 38, 39, 54. That is, the inattentive, hyperactive and impulsive behaviour that characterises children with AD/HD is explained in terms of a failure to suppress inappropriate responding. Importantly, several theoretical accounts advocate the view that response inhibition deficits are unique to AD/HD 4, 5, 13, 35, 54.

However, disruptive behaviour disorders (i.e. oppositional defiant disorder (ODD) and conduct disorder (CD)) have also been linked to a deficit in response inhibition 14, 28, 29, 37, 38, 40. Thus, response inhibition deficits may not be confined to AD/HD, but rather characterise children with externalising disorders. In contrast to AD/HD and disruptive behaviour disorders, anxiety disorders have been postulated to arise from abnormally high levels of response inhibition 37, 38. Indeed, a strong tendency for response inhibition has been found to predict later anxiety disorders [8].

A plethora of measures has been developed to assess response inhibition. One of these measures is the ‘stop’-signal task 24, 26. This task is based on a well-established theory of response inhibition 23, 24. In contrast to other measures of response inhibition, the stop-signal task affords measurement of the underlying inhibitory process.

Briefly, the stop-signal task requires fast and accurate execution of a reaction time task, denoted as the primary task. Occasionally, a stop-signal is presented, which requires the child to inhibit the response to the primary task. Stop-signals are presented at different intervals before the subject's expected response to the primary task target. The shorter the interval, the more difficult it becomes to inhibit the response. Usually, the intervals are chosen such that the shortest interval will yield a probability of inhibition close to ‘0’, whereas the longest interval will produce a probability of inhibition close to ‘1’.

Several studies have demonstrated the reliability and validity of the stop-signal task as a measure of response inhibition 21, 50.

In the past few years, the stop-signal task has been utilised in a series of studies with child psychopathological groups and in particular in studies with AD/HD children. Schachar and Logan [42]compared AD/HD children, normal children and four other psychopathological control groups: children with CD, with comorbid AD/HD+CD, with emotional disorders (mainly anxiety disorder) and with learning disorder. Only AD/HD children exhibited poor response inhibition. The impairment in response inhibition was traced back to a slow inhibitory process.

Daugherty, Quay and Ramos [10]studied essentially the same groups as Schachar and Logan [42]. Although they failed to demonstrate group differences, measures of AD/HD were significantly negatively correlated with the ability to inhibit responding. Pliszka, Borcherding, Spratley, Leon and Irick [36]examined the impact of comorbid anxiety disorder on response inhibition in AD/HD children. Interestingly, the presence of concurrent anxiety disorder was found to attenuate the response inhibition deficit in children with AD/HD. The results support indirectly the hypothesis that anxiety is associated with enhanced response inhibition.

Although these findings back the proposition of a specific link between AD/HD and poor response inhibition, other studies have urged some nuances in interpretation. Jennings, Van der Molen, Pelham, Debski and Hoza [20]compared AD/HD children, children with AD/HD and concurrent ODD and normal children. Contrary to the results of the other studies, no differences were found between the AD/HD and normal control group. The comorbid group, however, showed impairments in response inhibition and a deficiency in the inhibitory process.

Oosterlaan and Sergeant [32]contrasted AD/HD, disruptive, and anxious children with a group of normal children. Compared with the normal control group, both AD/HD and disruptive children evidenced poor response inhibition and slow inhibitory processes. No differences were found between anxious and control children. The results of this study raise doubts concerning the claim that poor response inhibition is a unique feature of AD/HD.

Two recent studies utilised a modification of the stop-signal task, known as the change task 11, 25. The change task is purported to measure the cognitive processes underlying the ability to change rapidly a current course of action. This ability involves two successive components of executive functioning [35]: (1) the ability to inhibit the on-going action (as in the stop-signal task); and (2) the ability to shift rapidly to an alternate action, i.e. response re-engagement.

The stimuli in the change task are identical to those in the stop-signal task. However, the two tasks differ with respect to the demand exerted by the stop-signal. In both tasks, a stop-signal requires response inhibition. In the change task, the stop-signal, in addition, requires the subjects to re-engage immediately in a different response, i.e. to execute the so-called change response.

Using the change task, Schachar, Tannock, Marriott and Logan [44]demonstrated that, in comparison with their normal peers, AD/HD children were impaired in both response inhibition and response re-engagement. In a subsequent study, Schachar and Tannock [43]compared AD/HD, comorbid AD/HD+CD, CD and normal children on the change task. A diagnosis of AD/HD was associated with slow inhibitory processes and impaired response re-engagement. No such impairments were found in children with CD, suggesting that the findings were unique to a diagnosis of AD/HD. However, inferences for CD children were limited by the small number of cases (n=5).

The current study was designed to replicate and extend the research of Schachar and colleagues with the change task 43, 44. Specifically, our aim was to determine whether deficits in response inhibition and response re-engagement are unique to AD/HD. Four groups of children were compared on the change task: children with AD/HD, disruptive children, anxious children and normal controls. In contrast to the studies by Schachar and colleagues, we used children who were not referred to clinics, but who attended special educational services because of their behavioural problems. The three psychopathological groups showed pervasive disorders both in school and in home situations.

The comparison between AD/HD and disruptive children is of particular interest. On a theoretical and descriptive level, both disorders have been associated with response inhibition deficits 28, 29, 37, 38, 39, 40. Moreover, the two psychopathological conditions show considerable overlap 2, 3, 7, 15, 19, 30, 41. Clearly, if AD/HD and disruptive children can be distinguished either on response inhibition, on response re-engagement, or both, this would considerably advance the divergent validity of AD/HD and disruptive behaviour disorders.

Another rationale for this study was to seek evidence for the hypothesis that anxiety disorders are associated with more efficient response inhibition than controls 37, 38. Previous research with the stop-signal task has failed to support this hypothesis. Since the change task has been found to exert higher demands on the inhibitory process than the stop-signal task 11, 25, it may reveal the hypothesised high levels of response inhibition in anxious children where the stop-signal task has failed until present to do so.