ReviewThe top and the bottom of ADHD: a neuropsychological perspective
Introduction
Several theoretical accounts of Attention Deficit/Hyperactivity Disorder [ADHD] have emerged over the last 10 years. The purpose of clinical research is to understand, diagnose and treat disorders efficiently. To achieve this goal, theoretical models are used as a systematic guide in conducting research. The current models of ADHD have different emphases. Some stress the frontal cortex [1], which we refer to here as the ‘top’ of the system. Others emphasise subcortical and even brain stem loci, which are referred to here as the ‘bottom’ of the system [2]. This variation in which locus of the brain is thought to be the cause of ADHD reflects, in part, the symptomatic heterogeneity of ADHD and the high degree of comorbidity with other disorders. The variation in the models also reflects, in part, the theoretical bias of the model builder. Models of how ADHD should be explained also differ with respect to which point in the development of the disorder they emphasise. Currently, there is no model which incorporates developmental data and can predict the developmental trajectory of ADHD. In addition, the differentiation of ADHD from highly associated (comorbid) disorders such as oppositional defiant disorder (ODD) and conduct disorder (CD) has proved neuropsychologically more difficult than first expected [3]. Likewise, differentiation of the attentional dysfunction in ADHD from highly associated disorders such as dyslexia and higher function autism (HFA) has met with some, but mixed success [4]. Use of contrasting clinical groups with either specific brain damage or known biochemical disorders, such as Phenylketonuria, that allow the comparison of the hypothesised dopamine deficiency in ADHD with conditions with an established dopamine deficiency are currently ongoing. [5], [6]. Critical relationships such as the role of punishment and reward in ADHD and their ensuing influence upon higher cognitive processes have only been modestly established. With this in mind, we briefly review the current models and examine issues which require clarification for further advancement in the field.
Currently, there are five models of ADHD: the Delay aversion model [7], the Behavioural Inhibition/Activation model [8], the Inhibition model [1], the executive function (EF) model [9] and the cognitive-energetic model [2]. These five models can be conceived as varying along a continuum of explanatory power: some models are less and others more comprehensive. In this paper the five models are treated as being conceptually nested within the cognitive-energetic model being considered to be the most comprehensive.
The models differ in how far higher, cognitive processes are considered the key to ADHD. A key term in ‘top-down’ models of ADHD is EF. A second term is working memory (WM), a third concept is attention and a fourth is inhibition. These terms are inter-related and this can be a source of confusion. A function of this paper is to discuss the neuropsychological support for these concepts, in particular for WM, and with reference to ADHD. This discussion may help clinicians unfamiliar with the area to appreciate the significance of this work both in terms of its diagnostic and therapeutic significance. For example, some reflection is required on the validity of the concepts; it is not self-evident whether WM and attention are entirely or only partly independent [10], suggesting that the discovery of a specific process deficit in ADHD will require separating attentional and WM effects in task performance.
Both EF, inhibition and WM are concepts which reflect recent cognitive conceptions of brain function. These terms have been used in ADHD research as mechanisms to explain the ADHD deficit [1], [11]. Is EF a sufficient explanation of ADHD? This question has been answered elsewhere as ‘no’ and will not be dealt with here extensively [2], [12]. It will be concluded that there is evidence of EF deficits in ADHD, although it is unclear whether the EF account of ADHD is process specific enough to differentiate ADHD from other childhood psychopathological disorders, such as HFA.
In contrast to the work focusing on higher cognitive concepts, there is a body of research which has stressed bottom-up processes such as the role of reward-punishment and motivation in ADHD [8], [13], [14], [15]. Barkley [1] noted that several early studies demonstrated that ADHD children become satiated more quickly than controls. Several researchers have suggested that the performance of children with AD/HD is dependent on the presence or absence of response contingencies [16], [17]). Although researchers seem to agree about an unusual sensitivity to reward as a characteristic of children with AD/HD, disagreement exists as to whether these children are over-sensitive or under-sensitive to reward. Some researchers argued and showed that children with AD/HD are less sensitive to reward [17], [18], [19]. However, others have argued that children with AD/HD show an increased tendency to look for immediate reward [20], [21]). There are also studies that have failed to show that reward affects the performance of children with AD/HD differentially [22], [23]. In short, although there is disagreement as to whether children with AD/HD are less or more sensitive to reward, results of several studies have suggested that children with AD/HD react differently to reward than do control children.
Of interest for the separation of ADHD from associated disorders is the finding that CD children are sensitive to reward. For example, Shapiro, Quay, Hogan, and Schwartz [24] showed that children with CD were more sensitive to reward than normal control children. Thus sensitivity to reward may not be specific to ADHD but be a mechanism common to both ADHD and CD.
The effects of reward and punishment have been associated in the cognitive energetic model as being critical to the operation of the effort pool [2], [25]. The energetic component of that model might be considered to be a ‘bottom-up’ system which registers and gives feedback to the orbital frontal cortex on whether a particular stimulus–response relation is satisfying or aversive for the organism [26].
This emphasis has two justifications. First, in the early to preadolescent stage of ADHD, motor restlessness has historically been considered a key element of the disorder [27]. The mechanism of transition by which overactive children become more normo-active as adults is unclear, although overactivity can still distinguish ADHD persons from controls in adult life [28]. While activity in sleep and sleep pattern in ADHD has a chequered history [27], [29], there does seem to be an association of sleep and behavioural overactivity [30]. Consequently, the models that only emphasise higher meta-cognitive dysfunction may be inappropriate for this subgroup of hyperkinetic children. Secondly, the bottom-up system possibly contains interconnections between activity and delay aversion [31]. Schweitzer and Sulzer-Azaroff [32] noted that in contrast to ADHD children, controls adopted a rational strategy of delaying their responses and gaining maximum reward. The ADHD children in that study also showed increased motor activity with task exposure. Various observational studies have shown increased activity of ADHD children in a vigilance task [33], [34] which may depend upon energetic allocation [2]. More recently, reward has been shown to activate the nucleus accumbens, suggesting the ‘bottom-up’ systems may be critical to behavioural deficits in ADHD [35].
Section snippets
Reward, delay and inhibition
The first term of the above heading clearly refers to a bottom-up process [35], [36], [37]. It is unclear whether delay is only a bottom-up process or is also a top-down process [38]. Inhibition is considered the dominant EF function in ADHD [1], whereas Quay [13] would also attribute to it a bottom-up role. This illustrates the need for studies which disentangle the conceptual issues addressed here.
Delay aversion is supported by research which suggests that ADHD children are unwilling to delay
Energetics: when top meets bottom
The cognitive-energetic model draws attention to the fact that ADHD has effects at three levels: cognitive mechanisms such as response output, energetic mechanisms such activation and effort and control systems of EF [2]. The model suggests that disruptive disorders have common deficiencies in EF control systems [42], and may be possibly differentiated either at an energetic level or at specific elementary cognitive stages. Thus the cognitive-energetic model is an attempt to encompass both
WM: maintenance, manipulation and monitoring
In this section, we address evidence concerning the functional anatomy of WM and monitoring. Two types of process in WM may be distinguished: ‘maintenance’ and ‘manipulation’ of information on-line, deal respectively, with the hold function versus the manipulation and/or transposing of information e.g., reversing the order of a list of numbers. The maintenance function of WM includes basic mnemonic processes such as active selection, comparison, and judgement of stimuli held in short-term and
Distinguishing WM and executive attention
WM evokes the association of a mechanism associated with purely memorial process. However, Baddeley and Logie [62] noted that this understandable association is due to how Baddeley and Hitch [63] approached the issue. Had they approached the concept from an attentional point of view, then it would probably have become termed ‘working attention’. The host of models which describe WM, their differences and similarities have been summarised by Miyake and Shah [64]. For present purposes, a working
Discussion
In this paper the main thrust has been to consider in detail the implications of the top-down and the bottom-up processes which may be implicated in ADHD and other childhood psychopathological disorders. Memorial span task requiring no other concurrent function were distinguished from span tasks requiring concurrent processing. In addition, it was noted that concurrent processes such as monitoring or selective attention are intimately associated with WM [51]. Hence, as a working hypothesis, it
Acknowledgements
The first author wishes to express his indebtedness to Drs F.X. Castellanos and E. Sonuga-Barke in discussing their work in relation to this review.
References (94)
- et al.
Sustained attention and inhibition of cognitive interference in treated phenylketonuria: associations with concurrent and lifetime phenylalanine concentrations
Neuropsychologia
(2002) - et al.
The neuropsychological profile of early and continuously treated phenylketonuria: orienting, vigilance, and maintenance versus manipulation functions of working memory
Neurosci Biobehav Rev
(2002) - et al.
How specific is a deficit of executive functioning for attention-deficit/hyperactivity disorder?
Behav Brain Res
(2002) - et al.
Attention deficit disorder with hyperactivity: A specific hypothesis of reward dysfunction
J Am Acad Child Adolesc Psychiatry
(1987) - et al.
Response inhibition and response re-engagement in attention-deficit/hyperactivity disorder, disruptive, anxious and normal children
Behav Brain Res
(1998) - et al.
Additive factor methodology applied to psychopathology with special reference to hyperactivity
Acta Psychol
(1990) Contributions of the prefrontal cortex to the neural basis of human decision making
Neurosci Biobehav Rev
(2002)Psychological heterogeneity in AD/HD-a dual pathway model of behaviour and cognition
Behav Brain Res
(2002)- et al.
Pathologies of brain attentional networks
Neurosci Biobehav Rev iews
(2000) - et al.
Dissociating working memory from task difficulty in human prefrontal cortex
Neuropsychologia
(1997)