Content uploaded by Sebastiaan Dovis
Author content
All content in this area was uploaded by Sebastiaan Dovis on Oct 13, 2014
Content may be subject to copyright.
ORIGINAL CONTRIBUTION
Prevalence and diagnostic validity of motivational impairments
and deficits in visuospatial short-term memory and working
memory in ADHD subtypes
Sebastiaan Dovis •Saskia Van der Oord •
Hilde M. Huizenga •Reinout W. Wiers •
Pier J. M. Prins
Received: 28 October 2013 / Accepted: 26 August 2014
ÓSpringer-Verlag Berlin Heidelberg 2014
Abstract Deficits in working memory (WM) and rein-
forcement sensitivity are thought to give rise to symptoms
in the combined (ADHD-C) and inattentive subtype
(ADHD-I) of ADHD. Children with ADHD are especially
impaired on visuospatial WM, which is composed of short-
term memory (STM) and a central executive. Although
deficits in visuospatial WM and reinforcement sensitivity
appear characteristic of children with ADHD on a group-
level, the prevalence and diagnostic validity of these
impairments is still largely unknown. Moreover, studies
investigating this did not control for the interaction
between motivational impairments and cognitive perfor-
mance in children with ADHD, and did not differentiate
between ADHD subtypes. Visuospatial WM and STM
tasks were administered in a standard (feedback-only) and
a high-reinforcement (feedback ?10 euros) condition, to
86 children with ADHD-C, 27 children with ADHD-I
(restrictive subtype), and 62 typically developing controls
(aged 8–12). Reinforcement sensitivity was indexed as the
difference in performance between the reinforcement
conditions. WM and STM impairments were most pre-
valent in ADHD-C. In ADHD-I, only WM impairments,
not STM impairments, were more prevalent than in con-
trols. Motivational impairments were not common (22 %
impaired) and equally prevalent in both subtypes. Memory
and motivation were found to represent independent neu-
ropsychological domains. Impairment on WM, STM, and/
or motivation was associated with more inattention
symptoms, medication-use, and lower IQ scores. Similar
results were found for analyses of diagnostic validity. The
majority of children with ADHD-C is impaired on visuo-
spatial WM. In ADHD-I, STM impairments are not more
common than in controls. Within both ADHD subtypes
only a minority has an abnormal sensitivity to
reinforcement.
Keywords ADHD subtypes Working memory
Reinforcement Reward
Introduction
Deficits in executive functioning are proposed to play a
pivotal role in explaining the problems individuals with
ADHD encounter in daily life [1,2]. Executive functions
allow individuals to regulate their behavior, thoughts and
emotions, and thereby enable self-control. Working mem-
ory (WM) is considered a core causal executive process in
ADHD [3], and is described as the ability to maintain,
control and manipulate goal-relevant information [4,5].
Research indeed suggests that WM is one of the most
impaired executive functions in ADHD [6,7,63], and that
WM impairments in children with ADHD may account for
their deficits in attention [8,9], hyperactivity [10], and
impulsivity [11].
According to Baddeley [4] WM is a multicomponent
system consisting of two storage subsystems and a central
executive. The storage subsystems—phonological and
visuospatial short-term memory (STM)—are dedicated to
Electronic supplementary material The online version of this
article (doi:10.1007/s00787-014-0612-1) contains supplementary
material, which is available to authorized users.
S. Dovis (&)S. Van der Oord H. M. Huizenga
R. W. Wiers P. J. M. Prins
Developmental Psychology, University of Amsterdam,
Weesperplein 4, 1018 XA Amsterdam, The Netherlands
e-mail: s.dovis@uva.nl
S. Van der Oord
Clinical Psychology, KU Leuven, Leuven, Belgium
123
Eur Child Adolesc Psychiatry
DOI 10.1007/s00787-014-0612-1
the short-term storage of modality (phonological or visu-
ospatial) specific information. The central executive is a
mental control system with limited attentional resources
that is responsible for supervising, controlling and manip-
ulating information in the STM systems. Studies investi-
gating WM components in children with ADHD indicate
that, on a group-level, both their STM and central execu-
tive are impaired [e.g., 12–15]. Furthermore, meta-analytic
findings suggest that children with ADHD show more
impairment on tasks measuring visuospatial WM than on
tasks measuring phonological WM [e.g., 2,6].
Although impaired visuospatial WM appears charac-
teristic of children with ADHD on a group-level, recent
findings suggest that ADHD is a neuropsychologically
heterogeneous disorder that probably is not characterized
by any single core dysfunction [16–19]. Given that only a
subset of children with ADHD meets criteria for an exec-
utive function deficit [16–23], visuospatial WM deficits on
group-level are probably carried by only a subset of chil-
dren with ADHD [16]. However, despite its obvious sig-
nificance for assessment and treatment, only two studies
(Holmes et al. [24]; Lambek et al. [25]) have attempted to
demarcate this WM-impaired subset within the ADHD
population. These studies found visuospatial WM impair-
ments in 29–47 % of the children with ADHD [25],
1
and
an overall diagnostic hit rate (overall correct classification
of children with and without ADHD) based on visuospatial
WM measures of about 75 % (correctly identifying 84.3 %
of the children with ADHD and 58 % of typically devel-
oping (TD) children [24]). In addition, even less is known
about the individual differences within the ADHD popu-
lation on the components of visuospatial WM: only
Holmes et al. investigated the diagnostic validity of a
visuospatial STM measure. They found this measure to be
less accurate in discriminating between children with and
without ADHD (correctly identifying 81.9 % of the chil-
dren with ADHD, but only 12 % of TD children) than their
measure of visuospatial WM.
Moreover, the results of these prevalence- and diag-
nostic validity studies [24,25] may be confounded by
motivational deficits. Motivational models propose that
children with ADHD are less stimulated by reinforcement
(i.e., reward) than typically developing children (probably
due to a dopaminergic deficit) and therefore require higher
amounts of reward in order to perform optimally [28–32].
Research indeed shows that children with ADHD, in con-
trast to their TD peers, show suboptimal performance on
visuospatial WM- and visuospatial STM tasks under reg-
ular reinforcement conditions (e.g., feedback-only), and
require relatively high incentives (e.g., feedback ?10
euros) to perform to their full abilities [13,33,34]. Holmes
et al. and Lambek et al. did not control for these motiva-
tional deficits in children with ADHD (both studies used
only regular reinforcement conditions), which may have
resulted in an overestimation of the prevalence and diag-
nostic validity of WM and STM impairments in their
ADHD samples.
Furthermore, ADHD can be divided into multiple sub-
types (American Psychiatric Association [43]). The two
most prevalent and valid diagnostic subtypes of ADHD are
the combined subtype (ADHD-C) and the predominantly
inattentive subtype (ADHD-I; [7,64,65]). ADHD-C and
ADHD-I are characterized by distinct patterns of symp-
tomatic behavior, associated features and demographics
[e.g., see 47]. Nevertheless, the studies of Holmes et al.
[24] and Lambek et al. [25] included both children with
ADHD-C and ADHD-I, but did not differentiate between
these subtypes. Moreover, although (on a group level) both
subtypes appear to have equally pronounced motivational
deficits (i.e., abnormal reinforcement sensitivity), evidence
suggests that children with ADHD-I are less impaired on
visuospatial WM than children with ADHD-C and, in
contrast to children with ADHD-C, seem unimpaired on
visuospatial STM (at least when motivational deficits are
taken into account; [30,35–37]; also see [7,38,66]).
Therefore, the findings of Holmes et al. and Lambek et al.
may be neither representative of children with ADHD-C,
nor of children with ADHD-I.
Finally, although an abnormal sensitivity to reinforce-
ment (as defined by Haenlein and Caul [28]) might be
characteristic for children with ADHD on a group-level
(for reviews [39,40]; see also [13,33,34]), the prevalence
and diagnostic validity of this motivational deficit within
the ADHD population is largely unknown. Only one recent
study (de Zeeuw et al. [23]) investigated its prevalence in
children with ADHD, and found that \8 % of these chil-
dren could be classified as having an abnormal sensitivity
to reinforcement. However, de Zeeuw et al. used a small
ADHD sample (n=26) which included all ADHD sub-
types (obviously, subtype comparisons were not possible),
and concluded that the low prevalence rate (e.g., preva-
lence in TD controls was 10 %) was probably related to the
high frequency of positive feedback that was applied dur-
ing their motivation task (80 % of the trials were rewar-
ded), which may have attenuated the impact of the
motivational deficits in their ADHD sample [23]. To our
knowledge, no studies investigated the diagnostic validity
of abnormal reinforcement sensitivity in ADHD.
The current study therefore investigated:(1) the prevalence
and diagnostic validity of visuospatial WM impairments and
1
Loo et al. [26] (in adolescents), Sjo
¨wall et al. [22] and Wahlstedt
et al. [27] generally find a somewhat lower prevalence of working
memory deficits in the ADHD population than Lambek et al. [25].
However, this lower prevalence might be explained by the fact that
these three studies only use a composite score of both visuospatial and
phonological working memory measures.
Eur Child Adolesc Psychiatry
123
visuospatial STM impairments in children with ADHD, tak-
ing their motivational deficits into account; (2) the prevalence
and diagnostic validity of these motivational deficits in chil-
dren with ADHD; and (3) whether the prevalence and diag-
nostic validity of these impairments differ between ADHD
subtypes. Exploratively, we examined the differences
between the neuropsychologically/motivationally impaired
and unimpaired children with ADHD-C and ADHD-I on
behavioral symptoms and other demographic variables (e.g.,
medication use, gender, IQ, etc.).
We investigated these questions by using the task scores
of children with ADHD-C, ADHD-I and TD children on
the visuospatial WM- and STM version of the Chessboard
task [13,33]. To account for, and investigate, the motiva-
tional deficits in the ADHD samples we presented these
tasks in two reinforcement conditions: a feedback-only
condition and a condition with feedback and a large
monetary incentive (10 euros). This 10 euros condition was
previously found to optimize task performance in children
with ADHD-C [33]. The change in performance between
the feedback-only and 10 euros condition was considered
the measure of sensitivity to reinforcement (the reinforce-
ment sensitivity index; see Footnote 3).
We predicted that: (1) visuospatial WM and reinforce-
ment sensitivity would significantly discriminate children
with ADHD (of both subtypes) from TD children, and that
related impairments would be more prevalent in children
with ADHD-C and ADHD-I than in TD children [21,24,
30,33]; (2) visuospatial STM would only discriminate
children with ADHD-C, but not children with ADHD-I,
from TD children [30], and (3) children with ADHD-C and
ADHD-I who were classified as impaired on WM, STM
and/or reinforcement sensitivity would have more behav-
ioral problems and less favorable demographic character-
istics than their unimpaired ADHD-C or ADHD-I peers [7].
Method
Participants
A total of 175 children participated: 86 children with ADHD-
C (aged 8–12 years), 27 children with ADHD-I (aged
9–12 years), and 62 TD children (aged 8–12 years). Children
with ADHD were recruited from outpatient mental-health-
care centers, TD children through elementary schools. Por-
tions of the data were presented elsewhere [13,35].
Inclusion criteria
For all groups: (a) an IQ score C80 established by the short
version of the Dutch Wechsler Intelligence Scale for
Children (WISC-III [41]). Two subtests, Vocabulary and
Block Design, were administered to estimate Full Scale IQ
(FSIQ). This composite score has satisfactory reliability
and correlates highly with FSIQ [42]; (b) absence of any
neurological disorder, sensory (color blindness, vision) or
motor impairment as stated by the parents; (c) not taking
any medication other than methylphenidate.
For the ADHD-C group: (a) a prior DSM-IV-TR [43]
diagnosis of ADHD combined-type and absence of any autism
spectrum disorder (ASD) according to a child psychologist or
psychiatrist; (b) a score within the clinical range (95th–100th
percentile) on the ADHD scales of both the parent and teacher
version of the Disruptive Behavior Disorder Rating Scale
(DBDRS [44, Dutch translation: 45]). The DBDRS contains
four DSM-IV scales; Inattention, Hyperactivity/Impulsivity,
Oppositional Defiant Disorder (ODD), and Conduct Disorder
(CD). Adequate psychometric properties are reported [45];
(c) meeting criteria for ADHD combined-type on the ADHD
section of the Diagnostic Interview Schedule for Children,
parent version (PDISC-IV [46]). The PDISC-IV is a structured
diagnostic interview based on the DSM-IV, with adequate
psychometric properties; (d) absence of CD based on the CD
sections of the PDISC-IV.
For the ADHD-I group: (a) a prior DSM-IV-TR diag-
nosis of ADHD inattentive-type and absence of any ASD
according to a child psychologist or psychiatrist; (b) a score
within the clinical range on the Inattention scale and a
score below the clinical range on the Hyperactivity/
Impulsivity scale of both the parent and teacher version of
the DBDRS; (c) to ensure that the ADHD-I group did not
include any children with subthreshold ADHD-C, we fol-
lowed recommendations made in the benchmark review of
Milich et al. [47, see also 1,30]: children in the ADHD-I
group not only had to meet criteria for ADHD inattentive-
type on the ADHD section of the PDISC-IV, but also had
to have\4 hyperactivity/impulsivity symptoms; (d) no CD
based on the CD sections of the PDISC-IV.
For the control group: (a) a score within the normal
range (\80th percentile) on all scales of both the parent and
teacher version of the DBDRS; (b) absence of a prior
DSM-IV-TR diagnosis of ASD or any other psychiatric
disorder (apart from a mathematics disorder or reading
disorder) as stated by the parents.
Group differences in demographics and characteristics and
are listed in Table 1(including the presence of a DMS-IV-
TR diagnosis of a mathematics disorder or reading disorder
as stated by the parents). Eight children in the ADHD-I group
(30 %) and 61 children in the ADHD-C group (71 %) were
taking methylphenidate,
2
but discontinued medication at
2
This relative difference between the ADHD groups in medication-
use was significant, v
2
(1) =13.814, p\0.001. However, including
medication-use as a covariate in analyses where the ADHD groups
were compared (and covariation was possible) did not change the
pattern of the results.
Eur Child Adolesc Psychiatry
123
least 24 h before the test-session, allowing a complete wash-
out [48].
The Chessboard task: WM and STM
The WM version of the Chessboard task [33] is a visuo-
spatial WM performance measure based on two WM tasks:
the Corsi Block Tapping Task [49] and the subtest Letter–
Number Sequencing from the Wechsler Adult Intelligence
Scale (WAIS-III [50]). The WM task taps the ability to
both maintain and reorganize visuospatial information that
is relevant for the task at hand (see Fig. 1). To ensure that
every presented sequence of stimuli has to be reorganized
(and the central executive is tapped), the order of stimuli
presentation is random with the restriction that in every
sequence at least one blue stimulus is presented before the
last green stimulus.
The STM version of the Chessboard task [13] is a vis-
uospatial STM performance measure tapping the ability to
maintain visuospatial information relevant for the task at
Table 1 Group demographics, parent and teacher ratings and mean performance differences
Measure Group
ADHD-C ADHD-I TD children F/v
2
g
p
2
Group comparison
a, b
(n=86) (n=27) (n=62)
MSD MSD MSD
Gender (M:F) 70:16 -18:9 – 27:35 – 22.9 C =I; TD =C, I
Age (years) 10.4 1.3 11.1 1.1 10.1 1.2 6.9 I [C, TD; C =TD
FSIQ 101 11.2 106 10.5 110 12.6 11.6 I =C, TD; C \TD
DBDRS parent
Inattention 21.6 4.0 19.0 4.5 2.5 2.4 538.3 TD \I\C
Hyperactivity/impulsivity 20.7 4.6 7.0 3.4 2.2 2.3 476.4 TD \I\C
ODD 12.4 5.3 5.5 4.4 1.9 2.2 158.1 TD \I\C
CD 2.6 2.3 0.9 1.4 0.1 0.3 41.2 TD =I; C [TD, I
DBDRS teacher
Inattention 17.2 5.0 15.7 4.9 1.6 1.8 307.5 TD \I, C; I =C
Hyperactivity/impulsivity 15.6 5.7 2.9 2.5 1.0 1.5 261.0 TD =I; C [TD, I
ODD 9.9 5.8 3.9 2.8 0.7 0.9 93.0 TD \I\C
CD 1.6 2.2 0.3 0.8 0.1 0.2 20.4 TD =I; C [TD, I
Weekly spendable income (in euros) 2.5 2.7 2.3 1.6 1.7 1.1 3.1 I =C, TD; C [TD
Mathematics disorder (yes:no) 0:86 -0:27 – 0:62 – – –
Reading disorder (yes:no) 6:80 -0:27 – 2:60 – 2.7 ns (p=0.260)
Medication-use (yes:no) 61:25 -8:19 – – – 13.8 I =C
Working memory (age adjusted)
Feedback-only 5 1.0 5.5 1.0 6.3 0.8 22.4 0.21 C \I\TD
10 euros 5.5 0.8 6 0.8 6.5 0.7 17.4 0.17 C \I\TD
Short-term memory (age adjusted)
Feedback-only 5.3 1.0 5.8 0.9 6.3 0.8 13.0 0.13 TD [C, I; C =I;
10 euros 5.7 0.8 6.2 0.8 6.4 0.6 8.6 0.09 C \TD, I; TD =I
Reinforcement sensitivity index 7.4 % 11.1 % 6.7 % 11.7 % 2.2 % 9.5 % 3.9 0.05 TD \C, I; C =I
IADHD-I, CADHD-C, TD typically developing children, CD Conduct Disorder, DBDRS Disruptive Behavior Disorder Rating Scale, FSIQ Full
Scale IQ, M:F male:female, ODD Oppositional Defiant Disorder
a
MAN(C)OVAs were performed. If the overall group-effect was significant (p\0.05), additional post hoc Tukey tests or additional
MANCOVAs were performed to clarify the group differences (for all significant differences pvalues were\0.01). Nominal data were analyzed
with Chi square tests
b
If an independent measure (i.e., gender, age, FSIQ, parent and teacher ratings, weekly spendable income, or medication use) differed between
certain groups (e.g., ADHD-C vs. TD group) it was subsequently used as covariate in the matching group comparison of the performance indices
(i.e., the working memory, short-term memory, and reinforcement sensitivity index). Therefore, in analyses of the performance indices, IQ,
weekly spendable income and gender were used as covariates when all groups were compared and when ADHD-C was compared to controls;
gender was used as covariate when ADHD-I was compared to controls; and parent-rated inattention, ODD, CD and medication-use were used as
covariates when the ADHD groups were compared. Covarying for these independent measures did not change the pattern of the results
Eur Child Adolesc Psychiatry
123
hand. The STM version is an STM analogue of the WM
task: the stimuli have to be reproduced in the same way as
on the WM task; green stimuli have to be reproduced
before the blue stimuli (see Fig. 1). However, in contrast to
the WM task, on each trial of the STM task all the green
stimuli are presented before the blue stimuli. Therefore,
none of the presented sequences on the STM task have to
be reorganized (and only the storage component is tapped;
for more details see [13]).
The difficulty level of both tasks is adaptive; after two
consecutive correct or incorrect reproductions, the
sequence is increased or shortened by one stimulus. Min-
imal sequence length is two stimuli and there is no maxi-
mum sequence length. Because the difficulty level adapts
to individual performance, the amount of positive and
negative feedback is approximately the same (55 %
reward, 45 % response-cost) for each child and in both task
versions and both reinforcement conditions. Each task
consists of *5 practice trials followed by 30 experimental
trials, and takes about 10 min to complete (for more details
[13]).
Reinforcement conditions
Each participant completed both reinforcement conditions,
and each reinforcement condition contained both the STM
and WM task (see orders of presentation used in counter-
balancing below). In the feedback-only condition, children
were instructed to do their best and respond as accurately
as possible. In the 10 euros condition, children were told
that they could earn 10 euros if they performed well
enough on the task. In both reinforcement conditions,
participants received immediate visual and auditory feed-
back and could monitor their overall performance by
means of a ‘performance bar’ (for a detailed description see
[13] or ESM Appendix 1).
Dependent measures
On both task versions, the first 12 trials are required to
reach the child’s optimal difficulty level and were therefore
excluded from analysis [13,33], and see ESM Appendix 2.
Thus, performance on each task was measured by the mean
sequence length of the last 18 trials. The reinforcement
sensitivity index was defined as the relative difference in
mean (STM and WM) performance between the 10 euros
condition and the feedback-only condition (i.e., the per-
centage difference in mean performance as a result of extra
reinforcement).
3
Fig. 1 A trial on the working memory version of the Chessboard
task. aTo start a trial the arrowhead button in the bottom-right corner
of the screen has to be clicked. bThen the focus screen (a black
screen with a little white cross) is presented. cSubsequently, a
sequence of stimuli (squares that light up) is presented 1 91ona
494 grid with green and blue squares ordered in a chessboard
formation. Each stimulus lights up for 900 ms and is followed by an
inter-stimulus interval of 500 ms. dAfter the stimulus-sequence is
presented the participant responds by mouse-clicking on the squares.
To respond correctly the presented stimuli have to be reproduced in a
reorganized way: the green stimuli have to be reproduced before the
blue stimuli; both in the same order as presented (the numbers in
picture d show an example of a correct reorganization). eAfter a
response feedback is presented. AAfter feedback-presentation, the
participant can start the next trial by clicking on the arrowhead button
[13]. Dimensions of the task (height 9width): 4 94 grid
(14 913.9 cm), individual stimuli (3.4 93.2 cm); distances
between adjacent stimuli: 0.3 cm between horizontally adjacent
stimuli and 0.2 cm between vertically adjacent stimuli (differences
between the height and the width were the result of a small 3D-effect
in the stimuli)
3
Reinforcement sensitivity index =[(WM ?STM 10 euros] –
(WM ?STM FO)] 9[100/(WM ?STM 10 euros)]. WM =age-
corrected mean score on WM task; STM =age-corrected mean score
on STM task; FO =feedback-only condition.
Eur Child Adolesc Psychiatry
123
Procedure
The study was approved by the faculty’s IRB. The partic-
ipating mental-healthcare centers sent recruitment letters to
the parents of all children aged 8–12 years with a DSM-IV-
TR diagnosis of ADHD (all subtypes). The participating
elementary schools sent recruitment letters to the parents of
all children aged 8–12 years (no matching procedure was
applied). If parents were interested in participating they
could contact the researchers for more information and to
sign up for the study. After obtaining informed consent
from the parents (on behalf of the participating children),
parents and teachers completed the DBDRS. If DBDRS
inclusion criteria were met, participants were invited to one
100-min test-session. During this session’s first hour the
two reinforcement conditions (feedback-only and 10 eu-
ros), each containing the WM and STM version of the
chessboard task, were administered, intermitted by a 5-min
break. Thereafter, the WISC-III subtests were adminis-
tered. In parallel, parents of children with ADHD were
interviewed with the PDISC-IV. If the child met the
inclusion criteria (s)he was included in the data set. To
control for order effects, the order of administration of the
reinforcement conditions and the task versions (STM and
WM) were counterbalanced separately within groups
(resulting in 8 orders of presentation).
Orders of presentation used in counterbalancing:
1 FO: STM [WM 10 euros: STM [WM
2 10 euros: STM [WM FO: STM [WM
3 FO: WM [STM 10 euros: WM [STM
4 10 euros: WM [STM FO: WM [STM
5 FO: STM [WM 10 euros: WM [STM
6 10 euros: STM [WM FO: WM [STM
7 FO: WM [STM 10 euros: STM [WM
8 10 euros: WM [STM FO: STM [WM
STM short-term memory, WM working memory, FO feedback-only
No information about the reinforcement conditions was
provided before the test-session (e.g., to avoid expectations
of receiving money). Children and their families were not
compensated for participating in this study over and above
the 10 euros from the high-reinforcement condition. Chil-
dren with ADHD were tested at their mental-healthcare
center, TD children at their school. Testing took place
between 9 a.m. and 5 p.m. Test rooms were quiet and views
from windows were blocked. Specific reinforcement
instructions (e.g. ‘If you perform well enough on this task
you will get these 10 euros’) were given to the child at the
start of each reinforcement condition (for complete
instructions see description of the reinforcement conditions
in ESM Appendix 1). During testing an experimenter was
present, sitting behind the child pretending to read a book.
Data analysis
Given the difference in age range between the ADHD-I
group (aged 9–12 years) and the ADHD-C and TD groups
(aged 8–12 years), task scores were, after checking for
normality and outliers,
4
adjusted for age using a regression
procedure. That is, in the entire sample we regressed task
scores on age, and the discrepancy between observed and
predicted data was taken as the age-adjusted task score.
These age-adjusted task scores were used in all analyses.
Prevalence
On the STM- and WM task children with ADHD were
characterized as impaired if their age-corrected task
score fell below the lowest 10th percentile of scores in
the TD group. Children with ADHD were characterized
as impaired on the reinforcement sensitivity index if
their score fell above the 90th percentile of the TD
group (this 10 % cut-off was also used in [17,19,25]).
5
Group differences were examined using 2-sided Chi
square analyses.
Diagnostic validity
Discriminant analyses were conducted to evaluate the
extent to which age-adjusted scores on STM and WM tasks
in the feedback-only (FO) and 10 euros conditions, and on
the reinforcement sensitivity index accurately discrimi-
nated between ADHD-C and controls, between ADHD-I
and controls, and between both ADHD groups. Differences
were examined using two-sided Chi square analyses.
Finally, analyses were conducted comparing clinical and
demographic variables between children with ADHD who
were classified as either impaired or non-impaired on WM,
STM and/or reinforcement sensitivity (based on the 10 %
cut-off), using MANOVAs or Chi square as appropriate.
Partial Eta squared effect sizes (g
p
2
) are reported for the
MANOVAs: g
p
2
=0.01 is regarded a small effect size, 0.06
a medium effect size, and 0.14 a large effect size [52]. For
Chi square analyses phi (U) or Crame
´r’s (V) effect sizes are
reported (depending on the number of categories): U/
V=0.10 indicates a small effect size, 0.30 a medium
4
Participants were excluded from analyses when both of the
following criteria were met: (1) a standardized residual on any of
the dependent measures with an absolute value [2, and (2) a Cook’s
distance C1[51]. Based on this criterion none of the participants had
to be excluded.
5
For the sake of completeness prevalence using a 5 % cut-off is
reported in ESM Appendix 3.
Eur Child Adolesc Psychiatry
123
effect size, and 0.50 a large effect size [67]. Unless
otherwise stated, analyses had adequate statistical power
(power C0.80) to detect at least medium effects.
Results
Counterbalancing and mean scores
The three groups did not differ in the relative number of
times that each counterbalancing-order was presented,
v
2
(14) =1.83, p=0.999, Crame
´r’s V=0.07, power to
detect a medium effect was 0.72. Also, including count-
erbalancing-order as a covariate did not change the results.
Group demographics and age-adjusted mean scores for
each of the five performance indices (STM performance in
both reinforcement conditions, WM performance in both
reinforcement conditions, and the reinforcement sensitivity
index) are listed in Table 1. For a detailed discussion of
comparable mean results see Dovis et al. [13,35].
Prevalence of impairment
To account for the effect of motivational deficits on per-
formance, only WM- and STM performance in the 10 euros
condition were used to estimate prevalence of WM and
STM impairment (unless otherwise stated). Figure 2pre-
sents the proportion of children with ADHD-C and ADHD-
I who met the 10 % threshold for an impairment on the
WM-, the STM-, and/or the reinforcement sensitivity
index. 75.6 % of the ADHD-C group, 55.6 % of the
ADHD-I group, and 27.4 % of the TD group had an
impairment on any one of these dependent measures, these
group differences were significant (ADHD-C vs. TD,
v
2
(1) =33.82, p\0.001, U=0.48; ADHD-I vs. TD,
v
2
(1) =6.47, p\0.05, U=0.27; ADHD-C vs. ADHD-I,
v
2
(1) =3.99, p\0.05, U=0.19).
Next, prevalence of impairment was examined per per-
formance index (see Fig. 2):
Working memory
58.1 % of the ADHD-C group, 33.3 % of the ADHD-I
group, and 9.7 % of the TD group was impaired on WM.
These group differences were significant (ADHD-C vs.
TD, v
2
(1) =35.97, p\0.001, U=0.49; ADHD-I vs. TD,
v
2
(1) =7.51, p=0.012, U=0.29; ADHD-C vs. ADHD-
I, v
2
(1) =5.07, p=0.024, U=0.21).
STM
40.7 % of the ADHD-C group, 18.5 % of the ADHD-I
group, and 9.7 % of the TD group was impaired on STM.
Except for the difference between the ADHD-I and the TD
group, these group differences were significant (ADHD-C
vs. TD, v
2
(1) =17.31, p\0.001, U=0.34; ADHD-I vs.
TD, v
2
(1) =1.36, p=0.298, U=0.12; ADHD-C vs.
ADHD-I, v
2
(1) =4.42, p=0.036, U=0.20).
Reinforcement sensitivity
22.1 % of the ADHD-C group, 22.2 % of the ADHD-I group,
and 9.7 % of the TD group was classified as having an
abnormal reinforcement sensitivity (motivational impair-
ment). Only the difference between the ADHD-C and the TD
group was significant, but the effect size was comparable to the
effect size of the difference between the ADHD-I and TD
group (ADHD-C vs. TD, v
2
(1) =3.96, p\0.05, U=0.16;
ADHD-I vs. TD, v
2
(1) =2.54, p=0.174, U=0.17; ADHD-
C vs. ADHD-I, v
2
(1) =0.00, p=0.989, U=0.001).
In the ADHD-C group, WM impairments were more
prevalent than STM impairments [v
2
(1) =5.23, p=
0.022, U=0.17], and both were more prevalent than
motivational impairments (WM vs. motivation, v
2
(1) =
23.26, p\0.001, U=0.37; STM vs. motivation, v
2
(1) =
6.91, p=0.009, U=0.20). In the ADHD-I group these
differences were non-significant (WM vs. STM, p=0.214,
U=0.17; WM vs. motivation, p=0.362, U=0.12; STM
vs. motivation, p=0.735, U=0.05, power to detect
medium effects was 0.60).
Overlap of impairments
In both ADHD groups there was significant overlap between
WM and STM deficits [ADHD-C: v
2
(1) =6.32, p=0.01,
U=0.27; ADHD-I: v
2
(1) =6.01, p=0.03, U=0.47;
30.3 % of children with ADHD-C and 14.8 % of children
with ADHD-I were impaired on both indices; see Fig. 2].
However, overlap between the reinforcement sensitivity
index and the memory indices was non-significant [WM and
motivation: ADHD-C, v
2
(1) =0.304, p=0.581, U=0.06;
ADHD-I, v
2
(1) =0.964, p=0.628, U=0.19; STM and
motivation: ADHD-C, v
2
(1) =0.450, p=0.502, U=0.07;
ADHD-I, v
2
(1) =1.75, p=0.555, U=0.16], suggesting
that these impairments are not associated. However, the
power for the analyses of the ADHD-I group was low (power
to detect a medium effect was 0.34).
Prevalence differences between reinforcement conditions
In both ADHD groups prevalence rates of WM- and STM
impairments were not significantly influenced by type of
reinforcement condition [ADHD-C: WM 10 euros (58.1 %
prevalence) vs. WM FO (50 %), v
2
(1) =1.15, p=0.284,
U=0.08; STM 10 euros (40.7 %) vs. STM FO (54.7 %),
v
2
(1) =3.36, p=0.067, U=0.14; ADHD-I: WM 10
Eur Child Adolesc Psychiatry
123
euros (33.3 %) vs. WM FO (37 %), v
2
(1) =0.081,
p=0.776, U=0.04; STM 10 euros (18.5 %) vs. STM FO
(25.9 %), v
2
(1) =0.429, p=0.513, U=0.09, the power
to detect a medium effect for the ADHD-I group was 0.60],
but note the trend for the effect of reinforcement on the
prevalence of STM impairments in the ADHD-C group.
Discriminant analyses
Multiple discriminant analyses were conducted to evaluate
the extent to which the five age-corrected performance
indices could accurately discriminate between the groups
(see Table 2).
ADHD-C vs. TD children
First, the five indices were entered in the analysis together
(see Table 2). The overall Wilks’s lambda was significant
(K=0.61, v
2
(5, N=148) =72.23, p\0.001). Canoni-
cal variate correlation coefficients for the five indices were:
WM FO (0.88), WM 10 euros (0.81), STM FO (0.66), STM
10 euros (0.60), reinforcement sensitivity index (-0.30).
Fig. 2 Proportion of the ADHD-C, ADHD-I and TD children with
visuospatial working memory- (WM), visuospatial short-term mem-
ory- (STM) (based on performance in the high-reinforcement
condition), and motivational deficits (i.e., abnormal reinforcement
sensitivity), and their degree of co-occurrence (totals may not equal
100 % because of rounding)
Eur Child Adolesc Psychiatry
123
The higher the absolute value of the coefficient, the more
the dependent measure contributes to group separation;
positive and negative coefficients contribute to group sep-
aration in opposite ways.
Next, separate discriminant analyses were run to
investigate how useful each single measure was at dis-
criminating between the ADHD-C and TD group. Wilks’s
lambda was significant for each measure,
6
suggesting that
each of these measures significantly discriminates between
the ADHD-C and TD group. Classification rates for these
measures are shown in Table 2. The overall correct clas-
sification rates based on WM performance (WM
FO =78.4 %; WM 10 euros =75 %) or STM perfor-
mance (STM FO =70.3 %; STM 10 euros =71.6 %)
were not significantly influenced by the amount of rein-
forcement [WM FO vs. WM 10 euros, v
2
(1) =0.47,
p=0.492, U=0.04; STM FO vs. STM 10 euros,
v
2
(1) =0.066, p=0.798, U=0.02], suggesting that the
diagnostic validity of WM performance and STM perfor-
mance does not change when motivation is taken into
account.
The reinforcement sensitivity index provided a signifi-
cantly worse overall classification rate (57.4 %) than all
other indices [motivation vs. WM FO, v
2
(1) =14.90,
p\0.001, U=0.22; motivation vs. WM 10 euros,
v
2
(1) =10.21, p=0.001, U=0.19; motivation vs. STM
FO, v
2
(1) =5.28, p=0.022, U=0.13; motivation vs.
STM 10 euros, v
2
(1) =6.51, p=0.011, U=0.15].
ADHD-I vs. TD children
When the five indices were entered in the analysis together,
Wilks’s Lambda was significant (K=0.82, v
2
(5,
N=89) =16.88, p=0.005, power for this analysis to
detect a medium effect was 0.56). Canonical variate cor-
relation coefficients were: WM FO (0.91), WM 10 euros
(0.74), STM FO (0.58), STM 10 euros (0.36), reinforce-
ment sensitivity index (-0.43).
Next, separate discriminant analyses were run for each
single measure. Wilks’s Lambda was not significant for
STM performance in the 10 euros condition (p=0.121),
nor for the reinforcement sensitivity index (although there
was a trend; p=0.06), suggesting that these measures do
not significantly discriminate between the ADHD-I group
and the TD group. For all other measures Wilks’s Lambda
was significant.
7
Classification rates are shown in Table 2.
The overall correct classification rates based on WM per-
formance (WM FO =70.8 %; WM 10 euros =62.9 %) or
STM performance (STM FO =65.2 %; STM 10 eu-
ros =62.9 %) were not significantly influenced by the
amount of reinforcement (WM FO vs. WM 10 euros,
v
2
(1) =1.24, p=0.265, U=0.08; STM FO vs. STM 10
Table 2 Classification rates based on the age-corrected performance measures
Measure(s) included in
discriminant analyses
Group comparison
ADHD-C vs. TD children ADHD-I vs. TD children ADHD-C vs. ADHD-I
(n=148) (n=89) (n=113)
Correct
ADHD-C
classif. (%)
Correct TD
child
classif. (%)
Correct
overall
classif.
(%)
Correct
ADHD-I
classif. (%)
Correct TD
child
classif. (%)
Correct
overall
classif.
(%)
Correct
ADHD-C
classif. (%)
Correct
ADHD-I
classif. (%)
Correct
overall
classif.
(%)
All measures/indices 76.7 83.9 79.7* 66.7 72.6 70.8* 65.1 59.3 63.7
Working memory
Feedback-only 73.3 85.5 78.4* 63 74.2 70.8* 52.3 63 54.9*
10 euros 74.4 75.8 75.0* 55.6 66.1 62.9* 62.8 63 62.8*
Short-term memory
Feedback-only 66.3 75.8 70.3* 70.4 62.9 65.2* 61.6 63 61.9*
10 euros 70.9 72.6 71.6* 59.3 64.5 62.9 62.8 59.3 61.9*
Reinf. sensitivity index 53.5 62.9 57.4* 55.6 61.3 59.6
40.7 55.6 44.2
TD typically developing children, Correct ADHD classif. correctly classified children with ADHD, Correct TD child classif. correctly classified
TD children, Correct overall classif. overall correct classification of group membership; all measures using the five indices; both short-term
memory and working memory tasks in both reinforcement conditions and the motivational index, Reinf. sensitivity reinforcement sensitivity
* Wilks’s lambda was significant (p\0.05);
p=0.06
6
WM FO, K=0.67, v
2
(1, N=148) =59.12, p\0.001; WM 10
euros, K=0.70, v
2
(1, N=148) =52.25, p\0.001; STM FO,
K=0.78, v
2
(1, N=148) =36.25, p\0.001; STM 10 euros,
K=0.81, v
2
(1, N=148) =30.73, p\0.001; reinforcement sensi-
tivity index, K=0.94, v
2
(1, N=148) =8.42, p=0.004.
7
WM FO, K=0.85, v
2
(1, N=89) =14.57, p\0.001; WM 10
euros, K=0.89, v
2
(1, N=89) =9.80, p=0.002; STM FO,
K=0.93, v
2
(1, N=89) =6.26, p=0.012.
Eur Child Adolesc Psychiatry
123
euros, v
2
(1) =0.098, p=0.755, U=0.02), suggesting
that the diagnostic validity of WM performance and STM
performance does not change when motivation is taken into
account.
The overall correct classification rate of the reinforce-
ment sensitivity index (59.6 %) was not significantly dif-
ferent from other indices (motivation vs. WM FO,
v
2
(1) =2.48, p=0.116, U=0.12; motivation vs. WM 10
euros, v
2
(1) =0.21, p=0.644, U=0.04; motivation vs.
STM FO, v
2
(1) =0.60, p=0.439, U=0.06; Motivation
vs. STM 10 euros, v
2
(1) =0.21, p=0.644, U=0.04).
ADHD-C vs. ADHD-I
When the five indices were entered in the analysis together,
Wilks’s Lambda was not significant, K=0.92, v
2
(5,
N=113) =9.38, p=0.095; but note that the power to
detect a medium effect was 0.69. Canonical variate cor-
relation coefficients for the five indices were: WM FO
(-0.70), WM 10 euros (-0.85), STM FO (-0.72), STM 10
euros (-0.84), reinforcement sensitivity index (0.09).
Next, separate discriminant analyses were run for each
single measure. Wilks’s Lambda was not significant for the
reinforcement sensitivity index (p=0.771), suggesting
that this measure did not significantly discriminate between
the ADHD groups. For all other measures Wilks’s Lambda
was significant.
8
Classification rates are shown in Table 2.
The overall correct classification rates based on WM per-
formance (WM FO =54.9 %; WM 10 euros =62.8 %) or
STM performance (STM FO =61.9 %; STM 10 eu-
ros =61.9 %) were not significantly influenced by the
amount of reinforcement (WM FO vs. WM 10 euros;
v
2
(1) =1.48, p=0.224, U=0.08; STM FO vs. STM 10
euros; v
2
(1) =0.00, p=1.00, U=0.00).
Comparing impaired vs. non-impaired children
with ADHD
Of all independent variables (see Table 1), only teacher-
rated inattention on the DBDRS, medication-use, and IQ
differed significantly between impaired
9
and non-impaired
children with ADHD-C (power to detect medium effects
was 0.59). Teacher-rated inattention and medication-use
were higher in impaired children (DBDRS-score =18.2
vs. 15.8, p=0.026, g
p
2
=0.07; medication-use =75.4 vs.
38.1 %, p=0.002), and IQ was lower in impaired children
(99 vs. 107 points; p=0.002, g
p
2
=0.11). Subdividing the
impaired ADHD-C sample into a memory impaired (only
impaired on WM and/or STM) and a motivationally
impaired group did not reveal specific memory- or moti-
vation-related effects (but power to detect medium effects
was 0.60). No differences were found between impaired
and non-impaired children with ADHD-I, but sample sizes
were too small (power to detect medium effects was only
0.13). For correlations between ADHD symptoms and
performance on the indices see ESM Appendix 4.
Discussion
This study investigated (1) the prevalence and diagnostic
validity of visuospatial WM and STM impairments in
children with ADHD when motivational deficits are taken
into account; (2) the prevalence and diagnostic validity of
reinforcement sensitivity deficits in children with ADHD,
and (3) whether the prevalence and diagnostic validity of
these impairments differ between ADHD subtypes. Ex-
ploratively, differences between the impaired (see Footnote
9) and unimpaired children with ADHD-C and ADHD-I
were examined.
The present findings showed that when motivational
deficits of children with ADHD were taken into account,
both WM and STM impairments were more prevalent in
the ADHD-C group than in the ADHD-I and TD group.
In the ADHD-I group, only WM impairments, not STM
impairments, were more prevalent than in the TD group.
In the discriminant analyses the same pattern of results
was found. In general, correct classification- and prev-
alence rates were not significantly affected by the type
of reinforcement condition, except that STM perfor-
mance only discriminated between ADHD-I and TD
children in the feedback-only condition. In both ADHD
groups there was a significant association between WM
and STM impairments, but these memory impairments
were not associated with deficits in reinforcement sen-
sitivity (although power for the analysis in the ADHD-I
group was low). Reinforcement sensitivity deficits were
equally prevalent in both ADHD groups, but only in the
ADHD-C group this deficit was significantly more pre-
valent than in the TD group. In children with ADHD-C,
this motivational deficit was less prevalent than
impairments of WM and STM. The reinforcement sen-
sitivity index only discriminated significantly between
ADHD-C and TD children (although there was a trend
for ADHD-I and TD children), and its predictive power
was significantly lower than that of either WM or STM
performance. Children with ADHD-C who were classi-
fied as impaired (see Footnote 9) had more teacher-rated
inattention symptoms, were more likely to use ADHD
medication, and had lower IQ scores.
8
WM FO, K=0.96, v
2
(1, N=113) =4.71, p=0.030; WM 10
euros, K=0.94, v
2
(1, N=113) =6.96, p=0.008; STM FO,
K=0.96, v
2
(1, N=113) =5.09, p=0.024; STM 10 euros,
K=0.94, v
2
(1, N=113) =6.88, p=0.009.
9
Impaired on WM (10 euros), and/or STM (10 euros), and/or
reinforcement sensitivity; using the 10 % cut-off.
Eur Child Adolesc Psychiatry
123
Memory
With motivation taken into account, 58.1 % of the children
with ADHD-C were found to be impaired on visuospatial
WM. This prevalence rate is somewhat higher than that of
the only other study that examined the prevalence of vis-
uospatial WM in ADHD ([25] where 29–47 % of the
ADHD sample was found impaired). Our findings suggest
that this difference might be related to the fact that Lambek
et al. did not differentiate between ADHD subtypes, since
the prevalence of WM impairments was significantly
higher in ADHD-C than in ADHD-I (58.1 vs. 33.3 %).
Further, our finding suggests that visuospatial WM
impairments are at least as prevalent in children with
ADHD-C as other ‘key’ neuropsychological dysfunctions
(prevalence of inhibition, 45–51 %; reaction time vari-
ability, 44–48 %; delay aversion, 14–56 % [e.g., 17,19,
22,53]), and are more prevalent than phonological WM
impairments (27–35 % impaired [25]). These findings
further suggest that impaired visuospatial WM may indeed
be a core causal executive process for a majority of chil-
dren with ADHD-C [3]. However, at the same time, these
results support models and previous findings which suggest
that ADHD is a neuropsychologically heterogeneous dis-
order that cannot be characterized by a single core dys-
function [16–22,25]. Furthermore, although WM
impairments in ADHD-I were less prevalent than in
ADHD-C, they were more prevalent than in the TD group,
and WM performance significantly discriminated between
ADHD-I and TD children, suggesting that visuospatial
WM deficits may also cause problems in a substantial part
(33.3 %) of the ADHD-I population.
This is the first study to investigate the prevalence of
visuospatial STM impairments in children with ADHD. In
children with ADHD-C, STM impairments were less
common than WM impairments (40.7 vs. 58.1 %
impaired). Furthermore, we found that about half of the
WM-impaired children with ADHD-C could not be clas-
sified as STM-impaired. Since WM capacity is regarded as
the sum of both STM- and central executive capacity [54],
this finding suggests that about half of the cases with vis-
uospatial WM impairments in the ADHD-C population are
not the result of visuospatial STM impairments, but are
solely caused by impairments in their central executive. In
the other half of the cases WM impairments may be the
result of STM impairments only, or of a combination of
STM and central executive impairments. To examine this,
future prevalence studies should include an additional task:
one that only measures central executive performance.
10
Although less prevalent than WM impairments, more
than 40 % of the ADHD-C group was impaired on STM,
and STM performance correctly discriminated between
ADHD-C and TD children in 71.6 % of the cases. This
suggests that STM impairments may give rise to ADHD-
related problems in a substantial part of the ADHD-C
population. In contrast, STM impairments were not more
prevalent in the ADHD-I group than in the TD group, nor
did STM performance significantly discriminate between
these samples (at least not when the confounding effect of
motivation was taken into account). These results are in
line with the theoretical appraisal by Diamond [30] and
with recent studies which suggest that children with
ADHD-I, in contrast to children with ADHD-C, are espe-
cially impaired on the central executive component, but not
on the STM component of WM [30,35,36]. Furthermore,
STM performance only discriminated between ADHD-I
and TD children in the feedback-only condition, not in the
high-reinforcement condition. This suggests that impaired
STM performance in children with ADHD-I results from
insufficient motivation to perform (also see [35]), and
promotes the use of additional incentives in studies that
investigate STM in children with ADHD-I.
Motivation
Although both theory [e.g. 28,29] and research [39,40;
also see 13,33,34] suggest that an abnormal sensitivity to
reinforcement is characteristic of children with ADHD on a
group level, our findings show that this motivational
impairment, apart from being a valid and distinct impair-
ment, is actually not so common among these children
(only 22 % were classified as impaired). De Zeeuw et al.
[23] found an even lower prevalence rate (\8 % impaired),
but this difference in results probably is related to a dif-
ference in reward frequency schedules: It has been sug-
gested that high reward frequency schedules attenuate
reinforcement sensitivity problems in children with ADHD
[23,31] and reward frequency was much higher in the
study of de Zeeuw et al. (80 % of the trials were rewarded,
compared to 55 % of the trials in our study). Although
further expansion of our research design was not possible
in our current study (e.g., increasing testing time would
potentially have impacted the sustained attention, motiva-
tion and performance of our participants), it would be
interesting for future studies to explore the effects of dif-
ferent reward frequencies on the prevalence of reinforce-
ment sensitivity problems in children with ADHD (e.g., by
adding a condition where only a minority of the trials is
rewarded, or a condition without reinforcement).
Reinforcement sensitivity deficits were equally pre-
valent in both ADHD subtypes, but only in the ADHD-C
group this impairment was significantly more prevalent
10
Future studies should be aware that impaired inhibitory perfor-
mance can also have a small impact on the central executive
performance of children with ADHD [see 12].
Eur Child Adolesc Psychiatry
123
than in the TD group, and the reinforcement sensitivity
index did not discriminate significantly between ADHD-I
and TD children. These findings are consistent with theo-
ries stating that motivational abnormalities characterize the
combined subtype only [31], and contradict theories stating
they apply to the inattentive subtype in particular [30].
However, we found a trend towards significance
(p=0.06) for the reinforcement sensitivity index to dis-
criminate between ADHD-I and TD children, which sug-
gests that this difference would have been significant in a
study with higher statistical power. Future studies should
test this hypothesis using a more substantial ADHD-I
sample. Based on our current results, we can conclude that
reinforcement/reward sensitivity deficits are not so com-
mon in children with ADHD (e.g., less common than
memory impairments in ADHD-C), and seem equally
prevalent in both ADHD subtypes.
Memory and motivation
To our knowledge our data provide the first evidence that
impairments in visuospatial WM and STM in ADHD are
dissociable from impairments in reinforcement sensitivity.
This absence of associations across motivational and
memory domains highlights the neuropsychological heter-
ogeneity in ADHD and supports recent evidence suggest-
ing separable neuropsychological subtypes in ADHD [e.g.,
16,19,23]. In this context our findings are especially
strong since they are based on neuropsychological mea-
sures that were probably not confounded by motivational
deficits. Furthermore, the absence of overlap between
memory and reinforcement sensitivity suggests that the
combined assessment of these domains may contribute to
improved neuropsychological differentiation of ADHD.
Nonetheless, it must be noted that this absence of associ-
ations between deficits in motivation and memory was also
found in controls. This suggests that the neuropsychologi-
cal heterogeneity in ADHD may be a derivative of normal
variation (see also [16]).
Correlates of impairments on WM, STM and/
or reinforcement sensitivity
Children with ADHD-C who were classified as impaired on
WM, STM, and/or reinforcement sensitivity had more
teacher-rated inattention symptoms, were more likely to
use ADHD medication (methylphenidate), and had lower
IQ scores than their unimpaired ADHD-C peers (for the
ADHD-I group power was inadequate to interpret this
analysis).
This seems consistent with models that suggest that
inattentiveness results from WM dysfunctions [1,3] and
with previous studies demonstrating that inattention, not
hyperactivity/impulsivity, is associated with neuropsycho-
logical impairment in children with ADHD [7,20,27,55,
56]. However, because this was a cross-sectional study it is
difficult to make causal inferences. Further, it is unclear
why impairment was only associated with teacher-rated
inattention, not with parent-rated inattention.
Impaired children with ADHD-C (on WM, STM and/or
reinforcement sensitivity) were more likely to be treated
with methylphenidate (75.4 vs. 38.1 % medication-use).
This is in line with evidence (in normal adults) suggesting
that the effectiveness of dopaminergic medication can be
predicted by WM performance in an un-drugged state [57],
and might be explained by the finding that WM capacity
predicts baseline levels of dopamine synthesis in the stri-
atum [58]. Future studies should investigate this in ADHD,
using larger samples (particularly for ADHD-I) to better
differentiate between memory and motivational impair-
ments (especially since there is also a strong relationship
between dopamine synthesis and motivation [58]).
Our finding that WM, STM and/or reinforcement sen-
sitivity impairments in ADHD-C are associated with lower
IQ scores is in line with previous ADHD prevalence studies
[21,23], and with findings in TD children [e.g., 59]. Fur-
ther, it supports the assumption that WM is crucial for the
mental activities basic to children’s intelligence [59], and is
consistent with the idea that neuropsychological impair-
ments (e.g., in WM) are responsible for the lower level of
intellectual performance typically found in children with
ADHD [1].
11
Limitations
The sample size of the ADHD-I group was relatively small
(n=27) and as a result some of the analyses (especially
the within-group analyses) were underpowered (i.e., power
was inadequate to detect medium effects). Therefore, the
underpowered null findings in the ADHD-I group should
be interpreted with caution (due to the possibility of type II
error). Although it must be noted that effect sizes of the
underpowered null findings were small, future studies
should use a larger sample size to replicate the findings in
the ADHD-I group.
Another potential limitation may have been the differ-
ence in IQ and weekly spendable income between the
ADHD-C and the TD group, and the difference between
the TD group and the ADHD groups on gender. However,
in ADHD–TD group comparisons, covarying for these
independent variables did not change the pattern of the
mean results (see Table 1). Further, the ADHD groups
differed on parent-rated inattention on the DBDRS.
11
In ADHD-C only the mean IQ score of the impaired subsample
was significantly lower than that of the TD group.
Eur Child Adolesc Psychiatry
123
However, in ADHD group-comparisons, covarying for this
inattention score did not change the mean results (see
Table 1).This suggests that our outcomes were not con-
founded by this difference in inattention. In addition, the
ADHD groups did not differ on teacher-rated inattention.
Although all children discontinued their ADHD medi-
cation at least 24 h before testing (allowing a complete
wash-out), there was a difference between the ADHD
groups in prior medication use: medication use was more
common in ADHD-C. However, since evidence suggests
that performance on WM measures is not influenced by the
chronic use of ADHD medication [60], and because
including medication use as a covariate did not change the
pattern of our mean results, we assume that the outcome of
this study was not confounded by this difference in prior
medication use.
Although all children were screened for externalizing
disorders, ASD, learning disorders (i.e., reading disorder
and mathematics disorder) and intellectual disabilities (i.e.,
an IQ score C80), and control children were only included
in the study if their parents stated they had no prior or
current DSM-IV-TR diagnosis (other than a reading dis-
order or a mathematics disorder), participants were not
specifically screened for internalizing disorders such as
anxiety or depressive disorders. However, evidence sug-
gests that anxiety and depressive disorders can affect WM
performance in typically developing groups [e.g., 66,68–
71], and there is some (although conflicting) evidence
regarding the effect of comorbid anxiety or depression on
the working memory performance of children with ADHD
[e.g., see 66,71–74]. There is also recent evidence sug-
gesting that high levels of anxiety and depression can
differentially modify WM performance according to
ADHD subtype [66]. Interestingly, it is suggested that
emotional states (e.g., anxiety) interact with cognitive
functioning through motivation [75]. However, little is
known about this interaction in children with ADHD [but
see 66]. Therefore, future prevalence studies investigating
ADHD subtype differences in WM, STM and/or motiva-
tional deficits should also assess and examine effects of
symptoms of anxiety and depression.
A strong point of the current study is that we investi-
gated the prevalence and diagnostic validity of WM and
STM impairments in children with ADHD by using mea-
sures that were probably not confounded by motivational
deficits (i.e., as strong incentives were used to optimize
performance
12
). Nonetheless, the prevalence and diagnos-
tic validity of many other important ADHD-associated
neuropsychological dysfunctions are still not examined in
this way. For example, we are unaware of studies that
investigate the prevalence and diagnostic validity of
impairments in inhibition or sustained attention in children
with ADHD by using measures that are not likely to be
confounded by motivational deficits. Future prevalence and
diagnostic validity studies should therefore adapt their
neuropsychological assessment tools to account for these
motivational deficits in children with ADHD.
We did not specifically investigate the extent to which
problems with sustained attention impacted the WM and
STM performance of children with ADHD. However, we
did control for situational factors (e.g., test rooms were quiet
and views from windows were blocked) and cognitive fac-
tors (e.g., the task versions were self-paced for optimal
attention/vigilance) that could provoke lapses of attention.
Moreover, in a previous study [33], where we used the same
WM task, we found that a 10 euros reinforcement condition
(the same as in the current study) normalized the sustained
attention of children with ADHD (i.e., if children with
ADHD were motivated with 10 euros, their mean WM
performance was as stable over time as the WM perfor-
mance of controls). Because the WM and STM-related
results in the current study were mainly based on perfor-
mance in the 10 euros condition, we assume that these results
were not confounded by problems with sustained attention
in children with ADHD. This assumption is also substanti-
ated by the slopes of the figures in ESM Appendix 2.
In the current study, we only investigated the effects of
immediate reinforcement. However, as the prevalence of
delay aversion in children with ADHD might be at least as
high as the prevalence of immediate reinforcement deficits
[e.g., see 19,22], it would be interesting to also investigate
the impact of delayed reinforcement on the prevalence and
diagnostic validity of WM and STM impairments in chil-
dren with ADHD-I and ADHD-C (especially as there might
be some conceptual overlap between delay aversion and
memory; e.g., see [22]; but also see [76]).
Clinical implications
First of all, it should be noted that 24.4 % of the children
with ADHD-C and no less than 44.4 % of the children with
ADHD-I showed no impairment on any of the investigated
indices (WM, STM, or reinforcement sensitivity). Further-
more, clinicians should be aware that although all these
indices discriminated significantly between children with
ADHD-C and TD children, only the WM and STM measures
showed clinically acceptable diagnostic validity, with both
sensitivity and specificity being C70 % (as was recom-
mended by Glascoe and Squires [61]). In addition, based on
these guidelines, none of the indices showed acceptable
diagnostic validity to distinguish children with ADHD-I from
TD children, or to distinguish between the ADHD subtypes.
Moreover, when it comes to distinguishing children with
12
The 10 euros condition was previously found to optimize task
performance in children with ADHD-C [33].
Eur Child Adolesc Psychiatry
123
ADHD-C from TD children, the diagnostic validity of
ADHD rating scales is, at this point, still much better (with
correct overall classification rates of 90–95 % [62]) than that
of any neuropsychological task (including visuospatial WM
or STM measures). As such, measures of visuospatial WM,
visuospatial STM or reinforcement sensitivity are not the
best choice for making DSM-oriented ADHD diagnoses in
children (especially not for diagnosing ADHD-I).That said, a
majority of children with ADHD-C are characterized by a
visuospatial memory and/or motivational impairment, and
assessment of these impairments may (independently) pro-
vide information about possible causal mechanisms of the
ADHD behavior of an individual child (e.g., the association
between his/her low WM and his/her classroom inattention
problems), and can help clinicians choose the best approach
for treatment. For example, it may help clinicians choose the
best treatment approach within behavioral parent- and tea-
cher training
13
(e.g., using reward systems versus techniques
to unburden WM; only a minority of children with ADHD-C
may require an intensive reward system, whereas a majority
of these children require strategies to unburden WM and
have less need for an additional intensive reward system), or
may help determine the relevance of a neuropsychological
training program (like STM or WM training) for an indi-
vidual child with ADHD. In line with this, our results imply
that interventions such as Cogmed working memory training,
of which there is debate as to whether mainly STM is trained
[e.g., 79], should focus more on training the central execu-
tive, especially in children with ADHD-I.
Acknowledgments We are grateful to the participating mental-
healthcare centers [Jeugdriagg Noord Holland Zuid, GGz Noord
Holland Noord (Centrum voor Kinder- en Jeugdpsychiatrie), Re-
gionaal Centrum voor Kinder en Jeugdpsychiatrie Gooi en Vechtst-
reek, Bosman GGz, Stichting De Praktijk, Stichting Kram, PuntP,
Academisch Behandelcentrum UvA Minds, & Kinderpraktijk VIS]
and the participating schools (OBS De Weidevogel, Amsterdam;
OBS De Witte Olifant, Amsterdam; De Dr. E. Boekmanschool,
Amsterdam; OBS Jules Verne, Alkmaar; PCBS Van der Brugghen-
school, Huizen; Montessorischool De Boog, Nieuw-Vennep; and De
Willemsparkschool, Amsterdam), to Jasper Wijnen for programming
the task, to Tim van den Broek, Josje de Bont, Annette Brouwer,
Tycho Dekkers, Lucie van den Eertwegh, Roza van der Heide, Li-
sanne Klink, Astrid Nauta, Inge Meulenberg, Murie
¨l Musa, Pascale
Riaskoff, Elise Tilma, Marije Voermans, Ida de Vries, and Pamina
Warmbrunn for their help with data collection, and to all participating
children and families.
Conflict of interest SvdO has been a paid consultant for Janssen
pharmaceuticals in the development and evaluation of a serious game
‘‘Healseeker’’ aimed at training cognitive functions. Other authors
report no conflicts of interest.
References
1. Barkley RA (2006) Attention-deficit hyperactivity disorder. A
handbook for diagnosis and treatment, 3rd edn. Guilford Press,
New York
2. Nigg JT (2006) What causes ADHD? Understanding what goes
wrong and why. Guilford Press, New York
3. Rapport MD, Chung KM, Shore G, Isaacs P (2001) A conceptual
model of child psychopathology: implications for understanding
ADHD and treatment efficacy. J Clin Child Psychol 30:48–58
4. Baddeley AD (2007) Working memory, thought and action.
Oxford University Press, Oxford
5. Conway ARA, Jarrold C, Kane MJ, Miyake A, Towse J (2007)
Variation in working memory. Oxford University Press, Oxford
6. Martinussen R, Hayden J, Hogg-Johnson S, Tannock R (2005) A
meta-analysis of working memory impairments in children with
attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc
Psychiatry 44:377–384
7. Willcutt EG, Nigg JT, Pennington BF, Solanto MV, Rohde LA,
Tannock R, Lahey BB (2012). Validity of DSM-IV attention
deficit/hyperactivity disorder symptom dimensions and subtypes.
J Abnorm Psychol 121:991–1010
8. Burgess GC, Depue BE, Ruzic L, Willcutt EG, Du YP, Banich
MT (2010) Attentional control activation relates to working
memory in ADHD. Biol Psychiatry 67:632–640
9. Kofler MJ, Rapport MD, Bolden J, Sarver DE, Raiker JS (2010)
ADHD and working memory: The impact of central executive
deficits and exceeding storage/rehearsal capacity on observed
inattentive behavior. J Abnorm Child Psychol 38:149–161
10. Rapport MD, Bolden J, Kofler MJ, Sarver DE, Raiker JS, Al-
derson RM (2009) Hyperactivity in boys with attention-deficit/
hyperactivity disorder (ADHD): a ubiquitous core symptom or
manifestation of working memory deficits? J Abnorm Child
Psychol 37:521–534
11. Raiker JS, Rapport MD, Kofler MJ, Sarver DE (2012) Objec-
tively-measured impulsivity and ADHD: testing competing pre-
dictions from the working memory and behavioral inhibition
models of ADHD. J Abnorm Child Psychol 40:699–713
12. Alderson RM, Rapport MD, Hudec KL, Sarver DE, Kofler MJ
(2010) Competing core processes in ADHD: Do working memory
deficiencies underlie behavioral inhibition deficits? J Abnorm
Child Psychol 38:497–507
13. Dovis S, Van der Oord S, Wiers RW, Prins PJM (2013) What part
of working memory is not working in ADHD? Short-term
memory, the central executive and effects of reinforcement.
J Abnorm Child Psychol 41:901–917
14. Rapport MD, Alderson RM, Kofler MJ, Sarver DE, Bolden J,
Sims V (2008) Working memory deficits in boys with attention-
deficit/hyperactivity disorder (ADHD): The contribution of cen-
tral executive and subsystem processes. J Abnorm Child Psychol
36:825–837
15. Rhodes SM, Park J, Seth S, Coghill DR (2012) A comprehensive
investigation of memory impairment in ADHD and oppositional
defiant disorder. J Child Psychol Psychiatry 53:128–137
16. Fair DA, Bathula D, Nikolas MA, Nigg JT (2012) Distinct neu-
ropsychological subgroups in typically developing youth inform
heterogeneity in children with ADHD. PNAS 109:6769–6774
17. Nigg JT, Willcutt EG, Doyle AE, Sonuga-Barke EJS (2005)
Causal heterogeneity in ADHD: do we need neuropsychologi-
cally impaired subtypes? Biol Psychiatry 57:1224–1230
18. Pineda DA, Puerta IC, Aguirre DC, Garcı
´a-Barrera MA, Kamp-
haus RW (2007) The role of neuropsychologic tests in the diag-
nosis of ADHD. Pediatr Neurol 36:373–381
19. Sonuga-Barke EJS, Bitsakou P, Thompson M (2010) Beyond the
dual pathway model: Evidence for the dissociation of timing,
13
These evidence-based interventions [77,78] aim at improving
behavioral control in children with ADHD by teaching parents and
teachers to use token (reward) systems and techniques to unburden
the WM of these children (e.g., providing reminders and a structured
environment).
Eur Child Adolesc Psychiatry
123
inhibitory control, and delay-related impairments in attention-
deficit/hyperactivity disorder. J Am Acad Child Adolesc Psy-
chiatry 49:345–355
20. Biederman J, Monuteaux MC, Doyle AE, Seidman LJ, Wilens TE,
Ferrero F, Faraone SV (2004) Impact of executive function deficits
and Attention-Deficit/Hyperactivity Disorder (ADHD) on aca-
demic outcomes in children. J Consult Clin Psychol 72:757–766
21. Lambek R, Tannock R, Dalsgaard S, Trillingsgaard A, Damm D,
Thomsen PH (2010) Validating neuropsychological subtypes of
ADHD: how do children with and without an executive function
deficit differ? J Child Psychol Psychiatry 51:895–904
22. Sjo
¨wall D, Roth L, Lindqvist S, Thorell LB (2013) Multiple
deficits in ADHD: executive dysfunction, delay aversion, reaction
time variability, and emotional deficits. J Child Psychol Psychi-
atry 54:619–627
23. De Zeeuw P, Weusten J, van Dijk S, van Belle J, Durston S
(2012) Deficits in cognitive control, timing and reward sensitivity
appear to be dissociable in ADHD. PLoS One 7:e51416
24. Holmes J, Gathercole SE, Place M, Alloway TP, Elliott JG,
Hilton KA (2010) The diagnostic utility of executive function
assessments in the identification of ADHD in children. Child
Adolesc Ment Health 15:37–43
25. Lambek R, Tannock R, Dalsgaard S, Trillingsgaard A, Damm D,
Thomsen PH (2011) Executive dysfunction in school-age chil-
dren with ADHD. J Attent Disord 15:646–655
26. Loo SK, Humphrey LA, Tapio T, Moilanen IK, McGough JJ,
McCracken JT, Smalley SL (2007) Executive functioning among
Finnish adolescents with attention-deficit/hyperactivity disorder.
J Am Acad Child Adolesc Psychiatry 46:1594–1604
27. Wa
˚hlstedt C, Thorell LB, Bohlin G (2009) Heterogeneity in
ADHD: neuropsychological pathways, comorbidity and symptom
domains. J Abnorm Child Psychol 37:551–564
28. Haenlein M, Caul WF (1987) Attention deficit disorder with
hyperactivity: a specific hypothesis of reward dysfunction. J Am
Acad Child Adolesc Psychiatry 26:356–362
29. Sergeant JA, Oosterlaan J, Van der Meere J (1999) Information
processing and energetic factors in attention-deficit/hyperactivity
disorder. In: Quay HC, Hogan AE (eds) Handbook of disruptive
behavior disorders. Kluwer Academic/Plenum Publishers, New
York, pp 75–104
30. Diamond A (2005) Attention-deficit disorder (ADHD without
hyperactivity): a neurobiologically and behaviorally distinct disorder
from ADHD (with hyperactivity). Dev Psychopathol 17:807–825
31. Sagvolden T, Johansen EB, Aase H, Russell VA (2005) A
dynamic developmental theory of ADHD predominantly hyper-
active/impulsive and combined subtypes. Behav Brain Sci 28:
397–419
32. Sonuga-Barke EJS (2011) Editorial: ADHD as a reinforcement
disorder—moving from general effects to identifying (six) spe-
cific models to test. J Child Psychol Psychiatry 52:917–918
33. Dovis S, Van der Oord S, Wiers RW, Prins PJM (2012) Can
motivation normalize working memory and task persistence in
children with attention-deficit/hyperactivity disorder? The effects of
money and computer-gaming. J Abnorm Child Psychol 40:669–681
34. Strand MT, Hawk LW, Bubnik M, Shiels K, Pelham WE,
Waxmonsky JG (2012) Improving working memory in children
with attention-deficit/hyperactivity disorder: the separate and
combined effects of incentives and stimulant medication. J Ab-
norm Child Psychol 40:1193–1207
35. Dovis S, Van der Oord S, Wiers RW, Prins PJM (in press) ADHD
subtype differences in reinforcement sensitivity and visuospatial
working memory. J Clin Child Adolesc Psychol. doi:10.1080/
15374416.2014.895940
36. Martinussen R, Tannock R (2006) Working memory impairments
in children with ADHD with and without comorbid language
learning disorders. J Clin Exp Neuropsychol 28:1073–1094
37. Pasini A, Paloscia C, Alessandrelli R, Porfirio MC, Curatolo P
(2007) Attention and executive functions profile in drug naive
ADHD subtypes. Brain Dev 29:400–408
38. Geurts HM, Verte
´S, Oosterlaan J, Roeyers H, Sergeant JA
(2005) ADHD subtypes: do they differ in their executive func-
tioning profile? Arch Clin Neuropsychol 20:457–477
39. Luman M, Oosterlaan J, Sergeant JA (2005) The impact of
reinforcement contingencies on AD/HD: a review and theoretical
appraisal. Clin Psychol Rev 25:183–213
40. Luman M, Tripp G, Scheres A (2010) Identifying the neurobi-
ology of altered reinforcement sensitivity in ADHD: a review and
research agenda. Neurosci Biobehav Rev 34:744–754
41. Kort W, Compaan EL, Bleichrodt N, Resing WCM, Schittekatte
M, Bosmans M, Verhaeghe P (2002) WISC-III NL Handleiding.
(Dutch manual). NIP, Amsterdam
42. Sattler JM (2001) Assessment of children: cognitive applications,
4th edn. Author, San Diego
43. American Psychiatric Association (2000) DSM-IV-TR. APA,
Washington, DC
44. Pelham WE, Gnagy EM, Greenslade KE, Milich R (1992) Tea-
cher rating of DSM-III-R symptoms for disruptive behavior dis-
orders. J Clin Child Psychol 8:259–262
45. Oosterlaan J, Scheres A, Antrop I, Roeyers H, Sergeant JA (2000)
Vragenlijst voor Gedragsproblemen bij Kinderen (VvGK). Swets
& Zeitlinger, Lisse
46. Shaffer D, Fisher P, Lucas CP, Dulcan MK, Schwab-Stone ME
(2000) NIMH diagnostic interview schedule for children version
IV (NIMH DISC-IV). J Am Acad Child Adolesc Psychiatry
39:28–38
47. Milich R, Balentine AC, Lynam DR (2001) ADHD combined
type and ADHD predominantly inattentive type are distinct and
unrelated disorders. Clin Psychol 8:463–488
48. Greenhill LL (1998) Childhood ADHD: Pharmacological treat-
ments. In: Nathan PE, Gorman J (eds) A guide to treatments that
work. Oxford University Press, New York, pp 42–64
49. Corsi PM (1972) Human memory and the medial temporal region
of the brain. Diss Abstr Int 34:819
50. Wechsler D (1997) Wechsler Adult Intelligence Scale, 3rd edn
(WAIS-III). Harcourt Assessment, San Antonio
51. Field A (2005) Discovering statistics using SPSS, 2nd edn. SAGE
Publications, London
52. Kittler JE, Menard W, Phillips KA (2007) Weight concerns in
individuals with body dysmorphic disorder. Eat Behav 8:115–120
53. Solanto MV, Abikoff H, Sonuga-Barke EJ, Schachar R, Logan
GD, Wigal TL, Turkel E (2001) The ecological validity of delay
aversion and response inhibition as measures of impulsivity in
AD/HD. J Abnorm Child Psychol 29:215–228
54. Engle RW, Tuholski SW, Laughlin JE, Conway ARA (1999)
Working memory, short-term memory, and general fluid intelli-
gence: a latent-variable approach. J Exp Psychol Gen 128:309–331
55. Bauermeister JJ, Barkley RA, Bauermeister JA, Martı
´nez JV,
McBurnett K (2012) Validity of the sluggish cognitive tempo,
inattention, and hyperactivity symptom dimensions: neuropsy-
chological and psychosocial correlates. J Abnorm Child Psychol
40:683–697
56. Tillman C, Eninger L, Forssman L, Bohlin G (2011) The relation
between working memory components and ADHD symptoms from
a developmental perspective. Dev Neuropsychol 36:181–198
57. Cools R, D’Esposito M (2011) Inverted-U-shaped dopamine
actions on human working memory and cognitive control. Biol
Psychiatry 69:113–125
58. Cools R, Gibbs SE, Miyakawa A, Jagust W, D’Esposito M (2008)
Working memory capacity predicts dopamine synthesis capacity
in the human striatum. J Neurosci 28:1208–1212
59. Swanson HL (2008) WM and intelligence in children: what
develops? J Educ Psychol 100:581–602
Eur Child Adolesc Psychiatry
123
60. Coghill DR, Rhodes SM, Matthews K (2007) The neuropsycho-
logical effects of chronic methylphenidate on drug-naive boys
with ADHD. Biol Psychiatry 62:954–962
61. Glascoe FP, Squires J (2007) Issues with the new developmental
screening and surveillance policy statement. Pediatrics 119:861–863
62. Conners CK (1999) Clinical use of rating scales in diagnosis and
treatment of ADHD. Pediatr Clin N Am 46:857–870
63. Kasper LJ, Alderson RM, Hudec KL (2012) Moderators of
working memory deficits in children with attention-deficit/
hyperactivity disorder (ADHD): a meta-analytic review. Clin
Psychol Rev 32:605–617
64. Gomez R, Harvey J, Quick C, Scharer I, Harris G (1999) DSM-
IV AD/HD: confirmatory factor models, prevalence, and gender
and age differences based on parent and teacher ratings of Aus-
tralian primary school children. J Child Psychol Psychiatry
40:265–274
65. Wolraich ML, Hannah JN, Baumgaertel A, Feurer ID (1998)
Examination of DSM-IV criteria for attention deficit/hyperac-
tivity disorder in a county-wide sample. J Dev Behav Pediatr
19:162–168
66. Ferrin M, Vance A (in press) Differential effects of anxiety and
depressive symptoms on working memory components in chil-
dren and adolescents with ADHD combined type and ADHD
inattentive type. Eur J Child Adolesc Psychiatry. doi:10.1007/
s00787-013-0509-4
67. Cohen J (1992) A power primer. Psychol Bull 112:155–159
68. Hadwin JA, Brogan J, Stevenson J (2005) State anxiety and
working memory in children: a test of processing efficiency
theory. Educ Psychol 25:379–393
69. Rose EJ, Ebmeier KP (2006) Pattern of impaired working
memory during major depression. J Affect Disord 90:149–161
70. Walsh ND, Williams SCR, Brammer MJ, Bullmore ET, Kim J,
Suckling J, Fu CHY (2007) A longitudinal functional magnetic
resonance imaging study of verbal working memory in depres-
sion after antidepressant therapy. Biol Psychiatry, 62:1236–1243
71. Vance A, Ferrin M, Winther J, Gomez R (2013) Examination of
spatial working memory performance in children and adolescents
with attention deficit hyperactivity disorder, combined type
(ADHD-CT) and anxiety. J Abnorm Child Psychol 41:891–900
72. Mayes SD, Calhoun SL, Chase GA, Mink DM, Stagg RE (2009)
ADHD subtypes and co-occurring anxiety, depression, and
oppositional-defiant disorder: differences in Gordon diagnostic
system and Wechsler working memory and processing speed
index scores. J Atten Disord 12:540–550
73. Sarkis SM, Sarkis EH, Marshall D, Archer J (2005) Self-regu-
lation and inhibition in comorbid ADHD children: an evaluation
of executive functions. J Atten Disord 8:96–108
74. Schatz DB, Rostain AL (2006) ADHD with comorbid anxiety: a
review of the current literature. J Atten Disord 10:141–149
75. Pessoa L (2009) How do emotion and motivation direct executive
control? Trends Cognit Sci 13:160–166
76. 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 Psychiatry 42:1335–1342
77. Pelham WE, Fabiano GA (2008) Evidence-based psychosocial
treatments for attention-deficit/hyperactivity disorder. J Clin
Child Adolesc Psychol 37:184–214
78. Evans SW, Owens JS, Bunford N (in press) Evidence-based
psychosocial treatments for children and adolescents with atten-
tion-deficit/hyperactivity disorder. J Clin Child Adolesc Psychol.
doi:10.1080/15374416.2013.850700
79. Shipstead Z, Hicks KL, Engle RW (2012) Cogmed working
memory training: does the evidence support the claims? J Appl
Res Mem Cognit 1:185–193
Eur Child Adolesc Psychiatry
123
