Causal Models of Attention-Deficit/Hyperactivity

Disorder: From Common Simple Deficits to Multiple

Developmental Pathways
Edmund J.S. Sonuga-Barke
Until recently, causal models of attention-deficit/hyperactivity disorder (ADHD) have tended to focus on the role of common, simple,
core deficits. One such model highlights the role of executive dysfunction due to deficient inhibitory control resulting from disturbances
in the frontodorsal striatal circuit and associated mesocortical dopaminergic branches. An alternative model presents ADHD as
resulting from impaired signaling of delayed rewards arising from disturbances in motivational processes, involving frontoventral
striatal reward circuits and mesolimbic branches terminating in the ventral striatum, particularly the nucleus accumbens. In the
present article, these models are elaborated in two ways. First, they are each placed within their developmental context by consideration
of the role of person ⫻ environment correlation and interaction and individual adaptation to developmental constraint. Second, their
relationship to one another is reviewed in the light of recent data suggesting that delay aversion and executive functions might each
make distinctive contributions to the development of the disorder. This provides an impetus for theoretical models built around the idea
of multiple neurodevelopmental pathways. The possibility of neuropathologic heterogeneity in ADHD is likely to have important
implications for the clinical management of the condition, potentially impacting on both diagnostic strategies and treatment options.

Key Words: Attention-deficit/hyperactivity disorder, causal models, The Quest for Common Core Dysfunctions in the
executive function, delay aversion, development, dopamine Science of Psychopathology
Philosophers of science have observed that the practice and
ttention-deficit/hyperactivity disorder (ADHD)1 is a debil-

A itating childhood psychiatric condition characterized by

severe and persistent impulsiveness, inattention, and
overactivity, resulting in significant long-term educational and
social disadvantage (Swanson et al 1998). Despite considerable
scientific interest regarding the underlying psychopathophysiol-
ogy of the condition, ADHD remains poorly characterized in this
sense. Traditionally, explanations of ADHD have been based
around simple causal models of single, common core dysfunc-
tions. In this review, we begin by contrasting two such models
that have been influential: 1) executive dysfunction due to
deficient inhibitory control; and 2) impaired signaling of delayed
rewards arising from disturbances in motivational processes. We
highlight the potential limitations of these simple causal para-
digms and the need for them to be elaborated further if they are
to provide full accounts of the clinical phenomenology of ADHD.
One way to achieve this is to place them into a framework that
highlights the role of the child’s social environment in shaping
neurodevelopmental pathways to ADHD. Recent evidence of
neuropsychologic heterogeneity in ADHD supports the likeli-
hood that multiple neurodevelopmental pathways underpin this
disorder and highlights the need for theoretical models of ADHD
to combine motivational and cognitive elements.
From the Developmental Brain-Behaviour Unit, School of Psychology, Uni-
versity of Southampton, Southampton, United Kingdom; and the Child
Study Centre, Department of Psychiatry, New York University, New York,
New York.
Address reprint requests to Edmund J.S. Sonuga-Barke, Ph.D., University of
Southampton, School of Psychology, Southampton S017 1BJ, United
Kingdom; E-mail: ejb3@soton.ac.uk.
Received May 26, 2004; revised August 30, 2004; accepted September 14,
2004.
1
In this review, the term ADHD is used to indicate ADHD combined type,
so that the theoretical challenge is to account for co-occurring
symptoms of inattention, hyperactivity, and impulsiveness.
progress of science is shaped by the assumptions held by
scientists about the nature of the phenomena of interest (Sonuga-
Barke 1998). Although such assumptions are essential to science,
in that they provide a common set of meanings that allow
communication between scientists themselves and with their
audiences, they also constrain research by determining legitimate
scientific questions and acceptable ways of providing answers. In
this way, the classic disease model of mental disorders has been
pivotal in shaping the scientific paradigm and defining the aims
and objectives of scientists in the field of child psychopathology
(Sonuga-Barke 1998). Implicit in this model is the assumption
that mental disorders are discrete disease entities, qualitatively
different from the normal range of functioning, which result from
a dysfunction of neuropsychologic/biologic mechanisms within
the patient. Indeed, DSM-IV translates the assumption of dys-
function into a defining feature of mental disorders (American
Psychiatric Association 1994). Given this, it is not surprising that
much scientific psychopathology seems motivated by a quest to
identify the site of the core dysfunctions that “cause” disorders.
Common Core Dysfunction in Causal Models of ADHD
In the neuroscience of ADHD, this has meant that one
question above all has provided the ultimate challenge for
researchers: where, within the brain/mind of the ADHD child, is
the site of the common core dysfunction that “causes” ADHD
(Sonuga-Barke 1994)?
A Cognitive Dysfunction Model of ADHD: The Mediating Role
of Inhibitory-Based Executive Deficits
Figure 1 illustrates one version of what, until recently, has
been considered the most useful (and is therefore the dominant)
class of explanations of ADHD psychopathology (i.e., cognitive
dysfunction models). This model is described with the notation
of the developmental causal modeling framework proposed by
John Morton and Uta Frith (Morton and Frith 1995). Because this
framework encourages us to think systematically about the
causal processes underpinning disorders, it aides the comparison

0006-3223/05/$30.00 BIOL PSYCHIATRY 2005;57:1231–1238

doi:10.1016/j.biopsych.2004.09.008 © 2004 Society of Biological Psychiatry
1232 BIOL PSYCHIATRY 2005;57:1231–1238
of the distinctive characteristics of, and therefore the specific
predictions made by, different causal models of ADHD. At the
same time, it must be recognized that the model is not theoret-
ically neutral, in that it reflects certain metatheoretical assump-
tions about the nature of disorder. These assumptions can be
seen in the ground rules set out by Morton and Frith, which
include the requirement to start at the neurobiologic level and
proceed to build causal chains across intermediate cognitive or
neuropsychologic and behavioral levels of analysis (e.g., symp-
toms of ADHD). Morton and Frith also emphasize the importance
of ensuring a full account of the disorder in question by
explaining all core clinical characteristics. The particular cogni-
tive dysfunction model presented is adapted from Barkley’s
unified theory of ADHD (Barkley 1997), in which symptoms of
the disorder are considered to be caused by the disruption of
neurocognitive control systems, with brain– behavior relations
fully mediated by deficits in inhibitory-based executive pro-
cesses. Executive functions are higher-order, top-down, cogni-
tive processes that allow appropriate set maintenance and shift
and that facilitate the flexible pursuit of future goals. Deficits on
tasks thought to measure these processes are a frequently
observed characteristic of children with ADHD (Barnett et al
2001; Bayliss and Roodenrys 2000; Clark et al 2000; Cornoldi et al
1999; Pennington and Ozonoff 1996; Seidman et al 1997; Ser-
geant et al 2002), with the substantial amount of evidence
implicating response inhibition deficits thought to be especially
compelling (Nigg 2001). Problems with working memory, plan-
ning, and set shift have also been identified (Karatekin and
Asarnow 1998; McLean et al 2004; Nigg et al 1998). Executive
dysfunction is also implicated in the processes involved in the
distribution of cognitive– energetic resources (i.e., effort) to
activation and arousal systems that are required to meet the
changing demands of different situations and settings and that
seem to be disrupted in ADHD (Douglas 1983; Sergeant 2000;
van der Meere 1996; van der Meere et al 1999). Barkley (1997)
argues that the general pattern of executive impairment associ-
ated with ADHD is grounded in more specific early-appearing
deficits in response inhibition (Nigg 2001; Oosterlaan et al 1998).
Response inhibition refers to the ability to inhibit an inappropri-
ate prepotent or ongoing response in favor of a more appropriate
www.sobp.org/journal
E.J.S. Sonuga-Barke
Figure 1. Schematic representation of a simple cogni-
tive deficit model of attention-deficit/hyperactivity dis-
order (ADHD) (adapted from Barkley 1997) and a simpli-
fied account of associated frontostriatal circuitry
(adapted from Alexander et al 1990). B, C, and S repre-
sent biology, cognition, and symptoms, respectively
(Morton and Frith 1995). The slashed C represents cog-
nitive deficit. NE, norepinephrine; DA, dopamine;
DLPFC, dorsolateral prefrontal cortex.
alternative. It is regarded as a prerequisite for self-control (Mu-
raven and Baumeister 2000), emotional regulation (Eisenberg
2002), and cognitive flexibility (Arbuthnott and Frank 2000). As a
domain of competence, response inhibition seems to be fraction-
ated into conceptually related clusters of functions, with each
cluster sharing common elements but also having key distinctive
features (Nigg 2001; Olson et al 2002; Winstanley et al 2004a).
At a neurobiologic level, there is growing evidence that
inhibitory control and other executive functions are underpinned
by one of a number of the functionally segregated but anatom-
ically proximate basal ganglia–thalamocortical circuits first pro-
posed by Alexander et al (1990; Christakou et al 2004). As shown
in Figure 1, this executive circuit links the prefrontal cortex (Aron
et al 2004a, 2004b) to the dorsal neostriatum (preferentially, the
caudate nucleus; Eagle and Robbins 2003) via excitatory glu-
taminergic cells. Reciprocal pathways pass via inhibitory connec-
tions through a complex of basal ganglia foci to the dorsomedial
thalamus with excitatory glutaminergic cells connecting back to
the prefrontal cortex (Heyder et al 2004). Data from structural
and functional neuroimaging studies support the hypothesis that
deficits in inhibitory-based executive functions in ADHD are
associated with disturbances in this circuit (Bush et al 1999;
Casey et al 2001; Castellanos 1997; Castellanos et al 2002;
Hesslinger et al 2001; Rubia et al 1999). Dopamine, which is
implicated in ADHD on the basis of pharmacologic and genetic
studies (Levy and Swanson 2001), is a key neuromodulator of
this circuit (Nieoullon and Coquerel 2003). Two distinct branches
seem to be involved. The mesocortical branch originates in the
ventral tegmental area of the rostral portion of the brainstem and
terminates in the prefrontal cortex, whereas the second branch
originates in the substantia nigra and terminates in the dorsal
striatum.
An Alternative Motivational Dysfunction Model: Disrupted
Signaling of Delayed Reward
A number of motivation-based dysfunction models have been
proposed as alternatives to cognitive theories of ADHD. These
models shift the focus from core deficits in inhibitory control to
suboptimal reward processes, with the links between neurobio-
logic processes and ADHD symptoms being mediated by deficits
E.J.S. Sonuga-Barke
in these mechanisms (Sagvolden 1991; Sagvolden et al 1998).
Figure 2 presents one such model, which builds on the general
ideas of Sagvolden and colleagues (see Sagvolden et al 1998). In
this, ADHD is the outcome of neurobiologic impairment in the
power and efficiency with which the contingency between
present action and future rewards is signaled. This leads to a
reduction in the control exerted by future rewards on current
behavior, a diminution in their “value,” and an increase in the
extent to which they are discounted (i.e., a steeper delay of
reward gradient). This account is supported by the consistent
finding that children with ADHD often display hypersensitivity to
delay and consequent difficulties in waiting for motivationally
salient outcomes, as well as in working effectively over extended
periods of time (Kuntsi et al 2001; Neef et al 2001; Schweitzer and
Sulzer-Azaroff 1995; Sonuga-Barke et al 1996; Tripp and Alsop
2001). This difficulty in waiting seems to be independent of
inhibitory deficits associated with executive dysfunction (Solanto
et al 2001; Sonuga-Barke et al 1994). A double dissociation
between inhibitory deficits and waiting for valued outcomes is
suggested by the fact that children with ADHD can wait even
when waiting involves inhibition, but they often choose not to
wait even when waiting does not involve inhibition (Sonuga-
Barke 1994).
Interestingly, the neurobiology of impaired signaling of
delayed rewards and inhibitory-based executive deficits seem
to share some common elements. The ability to wait for
delayed rewards seems to be related to alterations in another
of the dopamine-modulated thalamocortical– basal ganglia
circuits mentioned previously (Alexander et al 1990), although
in this context the motivational or affective circuit likely plays
the dominant role (McClure et al 2004). This circuit links the
ventral striatum (in particular, the nucleus accumbens; Zink et
al 2004) to frontal regions (especially the anterior cingulate
[Rogers et al 2004] and orbitofrontal cortex [Kringelbach and
Rolls 2004]), connections that are reciprocated via the ventral
pallidum and related structures through the thalamus (Robbins
and Everitt 1996). The amygdala seems also to be implicated
in this system, possibly playing a role in defining the motiva-
tional significance of incentives (Winstanley et al 2004b).
BIOL PSYCHIATRY 2005;57:1231–1238 1233
Figure 2. Schematic representation of a simple motiva-
tional model of attention-deficit/hyperactivity disorder
(ADHD) (adapted from Sagvolden et al 1998) and a sim-
plified account of frontostriatal circuitry (adapted from
Alexander et al 1990). B, C, and S represent biology,
cognition, and symptoms, respectively (Morton and
Frith 1995). The slashed C represents cognitive deficit.
NE, norepinephrine; DA, dopamine.
Once again, dopamine is a key neuromodulator of the reward-
signaling function of this circuit (Schultz 1998, 2002; Wight-
man and Robinson 2002). On this occasion, however, it is the
mesolimbic branch, originating in the ventral tegmental area
and terminating in the nucleus accumbens and associated
limbic foci, that seems to be of greatest importance. Studies
point to a specific role for this circuit in signaling rewards,
coding incentive value, and regulating other behavioral pro-
cesses involved in the maintenance of responding under
conditions of delayed reward (Burk and Mair 2001; Richardson
and Gratton 1998; Wade et al 2000). For instance, lesions in
the core of the accumbens have been shown to reduce the
ability of rats to wait for large delayed rewards in a self-control
paradigm (Cardinal et al 2001). Animal models of ADHD
highlight the way in which alterations of dopamine activity (of
either hypo- or hyperdopaminergic types; Russell 2002) might
distort or reduce signals of delayed reward (Max et al 2002;
Sadile 2000; Sagvolden 1991; Sagvolden et al 1998; Viggiano et
al 2002). Research that directly examines the role of this circuit
in delayed reward signaling in patients with ADHD is now
required.
From Simple Causal Models to Complex Development
Pathways
In their ground rules of the causal modeling framework,
Morton and Frith (1995) highlight the importance of providing a
“full account” of the disorder under investigation. In this sense,
causal models cannot be regarded as providing explanatory
power unless they capture the broad essence of the disorder as
it is manifest at the level of the individual child. To judge whether
a model provides a full account, it is necessary to specify a
minimum set of characteristics that need to be explained. What
would constitute this minimum set in ADHD? As a starting point,
models of combined-type ADHD need to account for the cardi-
nal symptom domains of impulsiveness, inattention, and hyper-
activity; however, a number of other candidate-defining features
clearly exist. For the purposes of the current analysis, everyday
difficulties with what might be called behavioral self-regulation
www.sobp.org/journal
1234 BIOL PSYCHIATRY 2005;57:1231–1238
(e.g., time management, ordering and organizing activities) will
be included in our minimum set of characteristics that need to be
explained by a good causal model of ADHD. Although these are
currently not diagnostic elements, they seem to be common
clinical manifestations (Abikoff et al 2003).
Do the two simple models of the psychopathophysiology of
ADHD presented above satisfy the “full account” criteria with
regard to this minimum set of characteristics? Although both
models have limitations, these might be most marked in the case
of the simple motivational model. This model certainly provides
a coherent and empirically supported account of impulsiveness
in ADHD. Further elaboration is, however, required if the basic
dysfunction in signaling of delayed rewards is to provide a
convincing explanation for inattentiveness, hyperactivity, and
problems in behavioral self-regulation as they present in ADHD.
One such elaboration has been proposed by Sagvolden et al (in
press), who have combined dysfunctional reinforcement and
extinction processes with broad-based, dopamine-related defi-
cits in motor control and attention to provide a hybrid motiva-
tional model.
An alternative thoroughgoing motivational approach to the
problem is provided by the delay aversion hypothesis. This can
be seen as extending the simple motivational model by placing
the underlying disposition of certain children to discount delayed
reward into its neurodevelopmental context (Sonuga-Barke et al
1994). According to this hypothesis, delay aversion—a negative
emotional reaction to the imposition of delay—is a developmen-
tal consequence of the failure of an impulsive child to engage
effectively with delay-rich environments. This negative emo-
tional response is manifested behaviorally as attempts to avoid or
escape delay, compounding the existing tendency of the child to
choose impulsively in settings in which choices between imme-
diate and delayed rewards are available. In other settings, in
which delay cannot actually be reduced by behaving in this way,
attempts are made to “systematically” reduce the perception of
time spent during delay. According to the model, this is achieved
by attending to interesting and absorbing aspects of the child’s
environment or by acting on that environment to make it more
interesting and absorbing. From a psychopathologic perspective,
such expressions of delay aversion would likely be characterized
as inattention and overactivity. In this way, the emergence of
delay aversion over time is hypothesized to lead to an elabora-
tion of symptoms from impulsiveness to inattentiveness and
overactivity. Although additional research is required to distin-
guish this model of ADHD from the simple motivational model,
evidence already exists that provides some support for it. For
example, results of a recent study in which the dot probe
paradigm was used found cues of the imposition and escape
from delay to selectively capture the attention of children with
ADHD in the same way that motivationally salient stimuli (such
as physical and social threat) have been shown to capture the
attention of other psychopathologic groups (e.g., anxious chil-
dren) (Sonuga-Barke et al 2004). Furthermore, the tendency of
children with ADHD to choose small immediate over large
delayed rewards or to behave impulsively in decision-making
tasks seems most marked when such behaviors reduce overall
delay and time on task rather than just increasing reward or
response immediacy (Sonuga-Barke et al 1994). Children with
ADHD also show an increase in activity and inattention during
delay periods (Antrop et al 2002).
Whereas early versions of the delay aversion model focused
on the impact of impulsiveness on the emergence of delay-
related inattention and overactivity, the cognitive developmental
www.sobp.org/journal
E.J.S. Sonuga-Barke
consequences of impulsiveness and delay aversion have been
the focus of more recent revisions. These have focused on the
way in which the delay-averse motivational style of the child acts
to constrain the normal acquisition of organizational skills, by
reducing the demands for self-organization and limiting the
experience required for the development of such skills (Sonuga-
Barke 2003).
Explanations of the development of delay aversion and
associated deficits in self-organizational skills invoke three re-
lated developmental processes. The first is characterized by child
⫻ environment correlation, whereby the developmentally ante-
cedent impulsive response of the child shapes his or her social
and family environment by, under certain circumstances, elicit-
ing a punitive or negative response from parents and siblings to
his or her failure to engage effectively with, and operate effi-
ciently in, a delay-rich environment. Second are person ⫻
environment interactions, whereby the punitive social environ-
ment, partially created by the behavior of the child, moderates
the links between underlying and early-appearing impulsiveness
and the emergence of a more generalized delay aversion: delay
settings acquire negative connotations and elicit a negative
affective response from which the child tries to escape. The final
developmental process is characterized by individual accommo-
dation to the child’s underlying predisposition toward impulsive-
ness and the constraints this imposes on learning opportunities.
Figure 3 provides an overview of the delay aversion hypoth-
esis as an extended and elaborated motivational account of the
development of ADHD. Five component processes of signifi-
cance are proposed. First, alternations in neurobiologic reward
circuits impair the signaling of delayed rewards (as in the simple
motivational model; Figure 2), leading to impulsiveness. Second,
impulsiveness leads to failures to effectively engage with, and
operate in, delay-rich environments. Third, this failure to engage
has the potential to elicit a negative/punitive response from a
parent or other significant adult (e.g., teacher), which over time
leads to delay aversion. Fourth, the failure to engage also
constrains the experience of managing delay and delay-demand-
ing activities and so reduces the opportunities to develop the
organizational skills and strategies required to do this. Finally,
delay aversion is expressed both as a compounding of existing
impulsiveness and an elaboration of behavioral characteristics
into the domains of attention and activity combined with impov-
erished organizational skills. Over time, these processes rein-
force a pattern of persistent symptomatology and associated
impairment. By this mechanism, a fundamentally motivational
disorder acquires specific cognitive characteristics in the areas of
everyday self-organizational skills. Although not explicitly refer-
enced in Figure 3, two other component processes of likely
significance should be mentioned here. The first relates to
negative parental responses elicited by the child’s disorder,
placing the child at risk of developing oppositionality. The
second relates to the possibility that the child might accommo-
date to the constraints imposed by his or her underlying predis-
position to impulsiveness and delay aversion. For instance, it
might be possible that ADHD children develop strategies to
exploit limited processing time more effectively. It is important to
note that children might become delay averse for a range of
reasons other than those illustrated here. For instance, time might
simply pass more slowly for delay-averse children, or they might
have experienced uncertain and inconsistent environments in
which future promised rewards and events were not delivered,
so that over time delay comes to signal, not failure and associated
E.J.S. Sonuga-Barke
negativity (as in the case of the model proposed in Figure 3), but
uncertainty and disappointment.
There is, of course, no reason to doubt that similar develop-
mental processes play a role in determining the outcome for
children with deficits in inhibitory-based executive dysfunction.
Figure 4 provides a schematic overview of a hypothetical devel-
opmental pathway model based on the simple cognitive model.
In this case, failures to engage effectively with tasks and envi-
ronments that are demanding in terms of inhibitory-based exec-
utive processes are hypothesized to have a dual impact on
motivational and cognitive development. On the one hand,
negative/punitive responses might be elicited from significant
adults, potentially resulting in “executive-task aversion,” which
in turn could lead to attempts to escape or avoid settings
requiring executive effort and skills (e.g., task avoidance and
early premature disengagement). Reduced exposure to execu-
tive-type tasks might limit the opportunities to develop executive
skills, thus compounding and elaborating basic underlying neu-
robiologic executive deficits. At the same time, failure on exec-
utive tasks might also reduce the extent to which tasks are
intrinsically motivating. This will in turn reduce task effort and
engagement, perpetuating the above process. At present, these
models are purely hypothetical and require systematic longitu-
dinal research to explore the specific predictions that they make.
BIOL PSYCHIATRY 2005;57:1231–1238 1235
Figure 3. Schematic representation of a
hypothetical motivational developmental
pathway for attention-deficit/hyperactiv-
ity disorder.
Cognitive and Motivational Models “Head to Head”
In keeping with the idea that ADHD is a discrete disease
resulting from a single core dysfunction, cognitive and motiva-
tional models have traditionally been regarded as competitive
rather than complementary. To address this issue, it is necessary
to contrast the two models and identify distinctive elements that
could provide the basis for their comparative assessment. Al-
though the developmental pathways described in Figures 3 and
4 both potentially provide a full account of ADHD in terms of the
minimum set of characteristics listed (inattention, impulsiveness,
hyperactivity, and difficulties in self-organization and regula-
tion), they clearly have a number of distinctive elements. For
instance, whereas inattentiveness and overactivity are likely to be
fundamental features of ADHD in the executive model, in the
motivational model one might expect inattentiveness to be
displayed only in delay-rich environments and among children
of parents with high expectations about impulse control com-
bined with a punitive parenting style (reflecting the outcome of
gene ⫻ environment correlation and interaction). One might also
expect a specific profile of cognitive problems according to the
motivational model, with problems being particularly pro-
nounced in the area of self-organizational skills and strategies in
delay-rich settings. At the most fundamental level, however, the
Figure 4. Schematic representation of a
hypothetical cognitive developmental
pathway for attention-deficit/hyperactiv-
ity disorder (ADHD).
www.sobp.org/journal
1236 BIOL PSYCHIATRY 2005;57:1231–1238
two models are most distinctive in terms of the presence of
inhibitory deficits (the cognitive model) and impaired respond-
ing to delay (the motivational model). These two elements
should therefore provide the most direct test between the two
models.
A recent head-to-head study of the two models exploited
this distinctiveness in underlying processes (Solanto et al
2001). School-aged children with a diagnosis of combined-
type ADHD performed the stop-signal task, in which a dom-
inant response already initiated is inhibited after a “stop”
signal, and a choice delay task, in which children had to
choose between a small immediate and a large delayed
rewards. The results indicated no association between choices
of the small immediate reward (delay aversion) and stop-
signal reaction time (inhibitory failures), suggesting that in-
hibitory deficits and delay aversion were somewhat indepen-
dent characteristics, overlapping at no more than chance
levels. Furthermore, performance on both tasks was found to
be strongly and independently associated with ADHD. To-
gether, the two measures allowed nearly 90% of children with
ADHD to be correctly classified. This pattern of results has
also been demonstrated in preschool children with ADHD,
aged 3–5 years (Sonuga-Barke et al 2003). In this study,
significant and independent associations were reported be-
tween ADHD and composite scores of executive dysfunction
(including scores for working memory, planning, set shifting,
and impulse control) on the one hand and delay sensitivity on
the other. Interestingly, a study of the mediating pathways
between hydrocephalus and hyperactivity also supports the
distinctive contribution of executive and motivational factors
(Stevenson and Cate 2004).
From Single to Multiple Pathways
Results of such head-to-head trials provide a direct chal-
lenge to the assumption of neuropsychologic homogeneity of
ADHD implicit in the current scientific paradigm. By suggest-
ing that two uncorrelated but theoretically coherent and
neurobiologically plausible processes each make a distinctive
and valuable contribution to predicting disorder, these data
indicate that multiple pathway models might be needed to
fully explain the ADHD phenomena. Although multivariate
head-to-head studies are the best way to test these sorts of
ideas, results of studies testing single models also provide
circumstantial evidence for the likelihood of multiple path-
ways. For instance, although neuropsychological studies of
executive dysfunction in ADHD consistently report an associ-
ation, these are at best moderate in size, with considerable
overlap between cases and controls. For instance, Nigg et al
(2005) have demonstrated in a combined analysis of more
than 1000 ADHD cases that only approximately 50% of
children with ADHD show a deficit in inhibitory control as
measured on the stop-signal test when judged on the basis of
a 90th percentile normal population cut-off. For many other
executive-type problems, that proportion was lower still:
executive dysfunction seemed to be neither a necessary nor a
sufficient condition for the expression of the disorder. Inter-
estingly, when the same criteria were applied to the preschool
study described above (Dalen et al 2004; Sonuga-Barke et al
2003), 29% of ADHD-range children displayed both delay
aversion and executive dysfunction, 27% delay aversion only,
15% executive dysfunction only, and 29% neither problem. It
will be important that future studies not only focus on the
www.sobp.org/journal
E.J.S. Sonuga-Barke
distinctiveness of delay aversion and executive deficits but
also on the way that these characteristics combine to influence
outcome.
On the basis of data from such head-to-head trials, a dual
pathway hypothesis of the psychopathophysiology of ADHD has
been proposed (Sonuga-Barke 2002, 2003). This hypothesis
builds on the idea that alterations within the executive circuit
modulated by mesocortical dopamine and the reward circuit
modulated by mesolimbic dopamine constitute more or less
discrete neuropsychologic bases for dissociable psychological
processes leading to executive/inhibitory deficits and delay
aversion, respectively. Recently, Nigg et al (2004) proposed an
alternative multiple pathway model, in which an executive
deficit pathway, associated with problems with effortful control,
was set alongside pathways associated with the problems with
positive and negative approach.
Although initial results are suggestive of the existence of
multiple independent pathways in ADHD, further study of the
relationship between these processes is obviously required to
properly test this hypothesis. To study these multiple pro-
cesses, a paradigm shift is required in the neuroscientific study
of ADHD. Research should use multivariate approaches with
measures from across a range of candidate cognitive and
motivational domains (e.g., inhibitory-based executive pro-
cesses and delay-related motivational processes) with large
samples of children of different ages in longitudinal and
genetically and neurobiologically informative designs. This
would allow continuities and discontinuities between motiva-
tional and cognitive components of the disorder to be studied.
Research also should explore the interactions between moti-
vational and cognitive features of the disorder during devel-
opment and the possibility that children with ADHD might
develop compensatory skills and strategies. In this review we
have not dealt with the issue of originating causes within
causal pathways to ADHD; however, proper tests of the causal
models set out here will require that the distinctive contribu-
tions of genetic and environmental factors (and their interac-
tions) in determining pathways and the mediating role of the
different basal ganglia thalamocortical circuits be explored.
Functional imaging studies, including executive control and
delayed-reward tasks within the same ADHD populations, and
studies with pharmacological probes of executive (e.g., spe-
cific dopamine 1 receptor agonists and norepinephrine trans-
porter blockers) and reward circuits (e.g., specific dopamine 2
receptor agonists and dopamine transporter blockers) might
represent fruitful strategies for distinguishing between the
neurobiological circuitry involved in the two pathways. Fur-
thermore, because these factors might act more like additive
risks than distinct “insults,” it will be especially important to
explore whether motivational and cognitive processes in
ADHD are best modeled as discrete categories or as continu-
ous traits. General population samples rather than case–
control designs will be important for this.
As Nigg et al (2004) have suggested, a dual pathway account
might of course not be exhaustive, and other pathways might be
hypothesized. For instance, other neuropsychologic mecha-
nisms, such as deficits in state-regulation (Sergeant 2005) and
alerting and orienting processes (Banaschewski et al 2003), with
distinct neurobiologic origins, have been implicated in ADHD.
Research should therefore also explore the neurobiologic alter-
ations underpinning these pathways, the psychological pro-
cesses by which they are mediated, and the environmental
factors that moderate them.
E.J.S. Sonuga-Barke
Clinical Implications
The recognition of neuropsychologic heterogeneity and the
possibility of different neuropsychological subtypes of ADHD
inherent in such multiple-pathway hypotheses raise fundamental
issues for the way that this disorder is identified and treated. First,
it encourages clinicians to adopt procedures that allow the
identification of different neuropsychologic subtypes of ADHD.
This might force a move away from current phenomenological,
symptom-based approaches to diagnosis toward more theoreti-
cally informed models. Information about the context for symp-
tom expression would become more significant, with diagnostic
criteria making increasing reference to contextual factors to
distinguish between children with delay-averse and disinhibited
ADHD. For instance, symptoms of inattention and overactivity
displayed only in delay-rich settings might be indicative of a
delay-averse ADHD subtype, whereas more pervasive expres-
sions might reflect more fundamental cognitive deficits associ-
ated with inhibitory dysfunction. Neuropsychological testing also
might take on a greater significance. Although at present evi-
dence for the diagnostic value of laboratory tests is lacking, tests
of inhibitory deficits and delay aversion might play a key role in
more precise definition of the clinical phenotype. Second, once
the neuropsychological subtypes have been identified, it might
be possible to match them to particular treatment options. This
applies equally to pharmacologic and nonpharmacologic inter-
ventions. Agents that target specific frontostriatal circuits are
likely to be differentially effective in treating inhibitory and
delay-averse subtypes. There also might be potential to tailor
nonpharmacologic treatments to subtypes. For instance, shaping
and fading of delay tolerance leading to desensitization to delay
might be useful for delay-averse children, whereas attention and
cognitive training might be valuable for dysexecutive children.
Aspects of this work were presented at the conference, “Ad-
vancing the Neuroscience of ADHD,” February 28, 2004, in
Boston, Massachusetts. The conference was sponsored by the
Society of Biological Psychiatry through an unrestricted educa-
tional grant from McNeil Consumer & Specialty Pharmaceuti-
cals.
Abikoff H, Gallagher R, Ma J (2003): Measuring and treating organizational,
time management and planning deficits in children with ADHD. J Child
Adolesc Psychopharmacol 13:422– 423.
Alexander GE, Crutcher MD, Delong MR (1990): Basal ganglia-thalamocorti-
cal circuits—parallel substrates for motor, oculomotor, prefrontal and
limbic functions. Prog Brain Res 85:119 –146.
American Psychiatric Association (1994): Diagnostic and Statistical Manual of
Mental Disorders, 4th ed. Washington, DC: American Psychiatric Press.
Antrop I, Buysse A, Roeyers H, Van Oost P (2002): Stimulation seeking and
hyperactive behavior in children with ADHD: A re-analysis. Percept Mot
Skills 95:71–90.
Arbuthnott K, Frank J (2000): Trail making test, part B as a measure of
executive control: Validation using a set-switching paradigm. J Clin Exp
Neuropsychol 22:518 –528.
Aron AR, Monsell S, Sahakian BJ, Robbins TW (2004a): A componential anal-
ysis of task-switching deficits associated with lesions of left and right
frontal cortex. Brain 127:1561–1573.
Aron AR, Robbins TW, Poldrack RA (2004b): Inhibition and the right inferior
frontal cortex. Trends Cogn Sci 8:170 –177.
Banaschewski T, Brandeis D, Heinrich H, Albrecht B, Brunner E, Rothenberger
A (2003): Association of ADHD and conduct disorder— brain electrical
evidence for the existence of a distinct subtype. J Child Psychol Psychiatry
44:356 –376.
Barkley RA (1997): Behavioural inhibition, sustained attention and executive
functions: Constructing a unifying theory of ADHD. Psychol Bull 121:
65–94.
BIOL PSYCHIATRY 2005;57:1231–1238 1237
Barnett R, Maruff P, Vance A, Luk ESL, Costin J, Wood C, et al (2001): Abnormal
executive function in attention deficit hyperactivity disorder: The effect
of stimulant medication and age on spatial working memory. Psychol
Med 31:1107–1115.
Bayliss DM, Roodenrys S (2000): Executive processing and attention deficit
hyperactivity disorder: An application of the supervisory attentional
system. Dev Neuropsychol 17:161–180.
Burk JA, Mair RG (2001): Effects of dorsal and ventral striatal lesions on
delayed matching trained with retractable levers. Behav Brain Res 122:
67–78.
Bush G, Frazier JA, Rauch SL, Seidman LJ, Whalen PJ, Jenike MA, et al (1999):
Anterior cingulate cortex dysfunction in attention-deficit/hyperactivity
disorder revealed by fMRI and the Counting Stroop. Biol Psychiatry 45:
1542–1552.
Cardinal RN, Pennicott DR, Sugathapala CL, Robbins TW, Everitt BJ (2001):
Impulsive choice induced in rats by lesions of the nucleus accumbens
core. Science 292:2499 –2501.
Casey BJ, Durston S, Fossella JA (2001): Evidence for a mechanistic model of
cognitive control. Clin Neurosci Res 1:267–282.
Castellanos FX (1997): Toward a pathophysiology of attention-deficit/hy-
peractivity disorder. Clin Pediatr 36:381–393.
Castellanos FX, Lee PP, Sharp W, Jeffries NO, Greenstein DK, Clasen LS, et al
(2002): Developmental trajectories of brain volume abnormalities in
children and adolescents with attention-deficit/hyperactivity disorder.
JAMA 288:1740 –1748.
Christakou A, Robbins TW, Everitt BJ (2004): Prefrontal cortical-ventral stria-
tal interactions involved in affective modulation of attentional perfor-
mance: Implications for corticostriatal circuit function. J Neurosci 24:
773–780.
Clark C, Prior M, Kinsella GJ (2000): Do executive function deficits differenti-
ate between adolescents with ADHD and oppositional defiant/conduct
disorder? A neuropsychological study using the Six Elements Test and
Hayling Sentence Completion Test. J Abnorm Child Psychol 28:403– 414.
Cornoldi C, Barbieri A, Gaiani C, Zocchi S (1999): Strategic memory deficits in
attention deficit disorder with hyperactivity participants: The role of
executive processes. Dev Neuropsychol 15:53–71.
Dalen L, Sonuga-Barke EJS, Remington RE (2004): Inhibitory deficits, delay
aversion and preschool AD/HD: Implications for the dual pathway
model. Neural Plast 11:1–11.
Douglas VI (1983): Attentional and cognitive problems. In: Rutter M, editor.
Developmental Neuropsychiatry. New York: Guilford 280 –329.
Eagle DM, Robbins TW (2003): Inhibitory control in rats performing a stop-
signal reaction-time task: Effects of lesions of the medial striatum and
d-amphetamine. Behav Neurosci 117:1302–1317.
Eisenberg N (2002): Emotion-related regulation and its relation to quality of
social functioning. Minn Symp Child Psychol 32:133–171.
Hesslinger B, Thiel T, Tebartz van Elst L, Hennig J, Ebert D (2001): Attention-
deficit disorder in adults with or without hyperactivity: Where is the
difference? A study in humans using short echo (1)H-magnetic reso-
nance spectroscopy. Neurosci Lett 304:117–119.
Heyder K, Suchan B, Daum I (2004): Cortico-subcortical contributions to
executive control. Acta Psychol 115:271–289.
Karatekin C, Asarnow RF (1998): Working memory in childhood-onset
schizophrenia and attention-deficit/hyperactivity disorder. Psychiatry
Res 80:165–176.
Kringelbach ML, Rolls ET (2004): The functional neuroanatomy of the human
orbitofrontal cortex: Evidence from neuroimaging and neuropsychol-
ogy. Prog Neurobiol 72:341–372.
Kuntsi J, Oosterlaan J, Stevenson J (2001): Psychological mechanisms in
hyperactivity: I. Response inhibition deficit, working memory impair-
ment, delay aversion, or something else? J Child Psychol Psychiatry 42:
199 –210.
Levy F, Swanson JM (2001): Timing, space and ADHD: The dopamine theory
revisited. Aust N Z J Psychiatry 35:504 –511.
Max JE, Fox PT, Lancaster JL, Kochunov P, Mathews K, Manes FF, et al (2002):
Putamen lesions and the development of attention-deficit/hyperactiv-
ity symptomatology. J Am Acad Child Adolesc Psychiatry 41:563–571.
McClure SM, York MK, Montague PR (2004): The neural substrates of reward
processing in humans: The modern role of fMRI. Neuroscientist 10:260 –268.
McLean A, Dowson J, Toone B, Young S, Bazanis E, Robbins TW, Sahakian BJ
(2004): Characteristic neurocognitive profile associated with adult atten-
tion-deficit/hyperactivity disorder. Psychol Med 34: 681– 692.
www.sobp.org/journal
1238 BIOL PSYCHIATRY 2005;57:1231–1238
Morton J, Frith U (1995): Causal modelling: A structural approach to developmental
psychopathology. In: Cicchetti D, Cohen DJ, editors. A Manual of Developmental
Psychopathology, vol. 1. New York: John Wiley & Sons, 357–390.
Muraven M, Baumeister RF (2000): Self-regulation and depletion of limited
resources: Does self-control resemble a muscle? Psychol Bull 126:247–
259.
Neef NA, Bicard DF, Endo S (2001): Assessment of impulsivity and the devel-
opment of self-control in students with attention deficit hyperactivity
disorder. J Appl Behav Anal 34:397– 408.
Nieoullon A, Coquerel A (2003): Dopamine: A key regulator to adapt action,
emotion, motivation and cognition. Curr Opin Neurol 16:S3–S9.
Nigg J (2001): Is ADHD a disinhibitory disorder? Psychol Bull 127:571–598.
Nigg J, Willcutt E, Doyle A, Sonuga-Barke EJS (2005): Casual heterogeneity in
ADHD: Do we need neuropsychologically impaired subtypes? Biol Psy-
chiatry 57:1224 –1230.
Nigg JT, Goldsmith HH, Sachek J (2004): Temperament and attention deficit
hyperactivity disorder: The development of a multiple pathway model.
J Clin Child Adolesc Psychol 33:42–53.
Nigg JT, Hinshaw SP, Carte ET, Treuting JJ (1998): Neuropsychological corre-
lates of childhood attention-deficit/hyperactivity disorder: Explainable
by comorbid disruptive behavior or reading problems? J Abnorm Psychol
107:468 – 480.
Olson SL, Bates JE, Sandy JM, Schilling EM (2002): Early developmental
precursors of impulsive and inattentive behavior: From infancy to mid-
dle childhood. J Child Psychol Psychiatry 43:435– 447.
Oosterlaan J, Logan GD, Sergeant JA (1998): Response inhibition in AD/HD,
CD, comorbid AD/HD ⫹ CD, anxious and control children: A meta-anal-
ysis of studies with the stop task. J Child Psychol Psychiatry 39:411– 425.
Pennington BF, Ozonoff S (1996): Executive functions and developmental
psychopathology. J Child Psychol Psychiatry 37:51– 87.
Richardson NR, Gratton A (1998): Changes in medial prefrontal cortical do-
pamine levels associated with response-contingent food reward: An
electrochemical study in rat. J Neurosci 18:9130 –9138.
Robbins TW, Everitt BJ (1996): Neurobehavioural mechanisms of reward and
motivation. Curr Opin Neurobiol 6:228 –236.
Rogers RD, Ramnani N, Mackay C, Wilson JL, Jezzard P, Carter CS, Smith SM
(2004): Distinct portions of anterior cingulate cortex and medial prefron-
tal cortex are activated by reward processing in separable phases of
decision-making cognition. Biol Psychiatry 55:594 – 602.
Rubia K, Overmeyer S, Taylor E, Brammer M, Williams SCR, Simmons A, et al
(1999): Hypofrontality in attention deficit hyperactivity disorder during
higher-order motor control: A study with functional MRI. Am J Psychiatry
156:891– 896.
Russell VA (2002): Hypodopaminergic and hypernoradrenergic activity in
prefrontal cortex slices of an animal model for attention-deficit hyperac-
tivity disorder—the spontaneously hypertensive rat. Behav Brain Res
130:191–196.
Sadile AG (2000): Multiple evidence of a segmental defect in the anterior
forebrain of an animal model of hyperactivity and attention deficit.
Neurosci Biobehav Rev 24:161–169.
Sagvolden T (1991): The attention deficit disorder might be a reinforcement
deficit disorder. In: Georgas J, Manthouli M, Besevegis E, Kokkevi A,
editors. Contemporary Psychology in Europe: Theory, Research and Appli-
cation. Gottingen: Hogrefe and Huber 131–143.
Sagvolden T, Aase H, Zeiner P, Berger D (1998): Altered reinforcement mecha-
nisms in attention-deficit/hyperactivity disorder. Behav Brain Res 94:61–71.
Sagvolden T, Borga Johansen E, Aase H, Russell VA (in press): A dynamic
developmental theory of AD/HD predominantly hyperactive/impulsive
and combined subtypes. Behav Brain Sci.
Schultz W (1998): Predictive reward signal of dopamine neurons. J Neuro-
physiol 80:1–27.
Schultz W (2002): Getting formal with dopamine and reward. Neuron 36:
241–263.
Schweitzer JB, Sulzer-Azaroff B (1995): Self-control in boys with attention
deficit hyperactivity disorder: Effects of added stimulation and time.
J Child Psychol Psychiatry 36:671– 686.
Seidman LJ, Biederman J, Faraone SV, Weber W, Ouellette C (1997): Toward
defining a neuropsychology of attention deficit-hyperactivity disorder:
Performance of children and adolescents from a large clinically referred
sample. J Consult Clin Psychol 65:150 –160.
www.sobp.org/journal
E.J.S. Sonuga-Barke
Sergeant J (2000): The cognitive-energetic model: An empirical ap-
proach to attention-deficit hyperactivity disorder. Neurosci Biobehav
Rev 24:7–12.
Sergeant JA (2005): Modelling ADHD: A critical appraisal of the cognitive-
energetic model. Biol Psychiatry 57:1248 –1253.
Sergeant JA, Geurts H, Oosterlaan J (2002): How specific is a deficit of exec-
utive functioning for attention-deficit/hyperactivity disorder? Behav
Brain Res 130:3–28.
Solanto MV, Abikoff H, Sonuga-Barke E, Schachar R, Logan GD, Wigal T, et al
(2001): The ecological validity of delay aversion and response inhibition
as measures of impulsivity in AD/HD: A supplement to the NIMH multi-
modal treatment study of AD/HD. J Abnorm Child Psychol 29:215–228.
Sonuga-Barke EJS (1994): On dysfunction and function in psychological
theories of childhood disorder. J Child Psychol Psychiatry 35:801– 815.
Sonuga-Barke EJS (1998): Categorical models of childhood disorder: A
conceptual and empirical analysis. J Child Psychol Psychiatry 39:115–
133.
Sonuga-Barke EJS (2002): Psychological heterogeneity in AD/HD—a dual
pathway model of behaviour and cognition. Behav Brain Res 130:
29 –36.
Sonuga-Barke EJS (2003): The dual pathway model of AD/HD: An elabora-
tion of neuro-developmental characteristics. Neurosci Behav Rev 27:593–
604.
Sonuga-Barke EJS, Dalen L, Remington B (2003): Do executive deficits and
delay aversion make independent contributions to preschool attention-
deficit/hyperactivity disorder symptoms? J Am Acad Child Adolesc Psychi-
atry 42:1335–1342.
Sonuga-Barke EJS, De Houwer J, De Ruiter K, Ajzenstzen M, Holland S (2004):
AD/HD and the capture of attention by briefly exposed delay-related
cues: Evidence from a conditioning paradigm. J Child Psychol Psychiatry
45:274 –283.
Sonuga-Barke EJS, Houlberg K, Hall M (1994): When is “impulsiveness” not
impulsive? The case of hyperactive children’s cognitive style. J Child
Psychol Psychiatry 35:1247–1253.
Sonuga-Barke EJS, Williams E, Hall M, Saxton T (1996): Hyperactivity and
delay aversion. III: The effect on cognitive style of imposing delay after
errors. J Child Psychol Psychiatry 37:189 –194.
Stevenson J, Cate IP (2004): The nature of hyperactivity in children and
adolescents with hydrocephalus: A test of the dual pathway model.
Neural Plast 11:13–21.
Swanson J, Sergeant J, Taylor E, Sonuga-Barke EJS, Jensen P, Cantwell D
(1998): Attention deficit hyperactivity disorder and hyperkinetic disor-
der. Lancet 351:429 – 433.
Tripp G, Alsop B (2001): Sensitivity to reward delay in children with attention
deficit hyperactivity disorder (ADHD). J Child Psychol Psychiatry 42:691–
698.
van der Meere J, Gunning B, Stemerdink N (1999): The effect of methylpheni-
date and clonodine on response inhibition and state regulation in chil-
dren with ADHD. J Child Psychol Psychiatry 40:291–298.
van der Meere JJ (1996): The role of inattention in hyperactivity disor-
ders. In: Sandberg S, editor. Monographs on Child and Adolescent
Psychiatry: Hyperactivity Disorders. Cambridge: Cambridge University
Press, 106 –145.
Viggiano D, Grammatikopoulos G, Sadile AG (2002): A morphometric evi-
dence for a hyperfunctioning mesolimbic system in an animal model of
ADHD. Behav Brain Res 130:181–189.
Wade TR, de Wit H, Richards JB (2000): Effects of dopaminergic drugs on
delayed reward as a measure of impulsive behavior in rats. Psychophar-
macology 150:90 –101.
Wightman RM, Robinson DL (2002): Transient changes in mesolimbic dopa-
mine and their association with ‘reward’. J Neurochem 82:721–735.
Winstanley CA, Dailey JW, Theobald DEH, Robbins TW (2004): Fractionating
impulsivity: Contrasting effects of central 5-HT depletion on different
measures of impulsive behavior. Neuropsychopharmacology 29:1331–
1343.
Winstanley CA, Theobald DEH, Cardinal RN, Robbins TW (2004): Contrasting
roles of basolateral amygdala and orbitofrontal cortex in impulsive
choice. J Neurosci 24:4718 – 4722.
Zink CF, Pagnoni G, Martin-Skurski ME, Chappelow JC, Berns GS (2004):
Human striatal responses to monetary reward depend on saliency. Neu-
ron 42:509 –517.