Anda di halaman 1dari 14

Psychological Medicine (2014), 44, 1087-1099.

Cambridge University Press 2013

doi:10.1017/S0033291713001761 ORIGINAL ARTICLE

A longitudinal examination of neuropsychological

and clinical functioning in boys with attention
deficit hyperactivity disorder (ADHD):
improvements in executive functioning do not
explain clinical improvement

D.R. CoghiU^, D. Hayward^ S.M. Rhodes^ C. Grimmer* and K. Matthews^

^Division of Neuroscience, Medical Research Institute, University of Dundee, Ninewells Hospital, Dundee, UK
^NHS Tayside, Murray Royal Hospital, Perth, UK
^School of Psychological Sciences and Health, University of Strathclyde, Graham Hills Building, Glasgow, UK
*NHS Fife Child and Adolescent Mental Health Services, Stratheden Hospital, Stratheden, Cupar, Fife, UK

Background. Attention defidt hyperactivity disorder (ADHD) often, but not always, persists into adulthood.
Investigations of the assodations between clinical and biological markers of persistence can shed light on causal path-
ways. It has been proposed that compensatory improvements in executive neuropsychological functioning are assodated
with clinical improvements. This is the first study to test this hypothesis prospectively.

Method. The clinical and neuropsychological functioning of 17 boys with ADHD (mean age 10.45 years at time 1;
14.65 years at time 2) and 17 typically developing (TYP) boys (mean age 10.39 years at time 1; 14.47 years at time 2)
was tested on two occasions, 4 years apart. This was done using a battery of standardized neuropsychological tests
that included tasks with high and low executive demands.

Results. Clinical improvements were observed over time. Neuropsychological performance improvements were also
evident, with ADHD boys developing with a similar pattern to TYP boys, but with a developmental lag. Whilst there
was an assodation between reduced symptoms and superior performance at retest for one task with a high executive
demand (spatial working memory), this was not seen with two further high executive demand tasks [Stockings of
Cambridge and intra-dimensional extra-dimensional (ID/ED) set shifting]. Also, there was no association between
change in executive functioning and change in symptoms. Baseline performance on the ID/ED set-shifting task predicted
better clinical outcome. Only change in performance on the low executive demand delayed matching-to-sample task
predicted better clinical outcome.

Conclusions. These data highlight the importance of longitudinal measurements of cognition, symptoms and treatment
response over time in children and adolescents with ADHD.
Received 6 February 2013; Revised 22 June 2013; Accepted 21 June 2013; First published online 19 July 2013

Key words: Attention defidt hyperactivity disorder, executive functioning, longitudinal studies, memory,
neuropsychology, outcome.

Introduction to meet full diagnostic criteria as adults still suffer

from significant functional impairment (Faraone et al.
It is now generally accepted that both the symptoms
2006). Accordingly, it has become increasingly import-
and impairments associated with attention defidt
ant to identify the correlates and predictors of persist-
hyperactivity disorder (ADHD) often continue both
ence and remission across development. Early research
into adolescence and adulthood. Rates of persistence
identified the number and severity of symptoms and
into adolescence are around 85% (Biederman et al
the presence of conduct disorder as key predictors of
1996a). Depending on definition, between 30 and
persistence into adolescence (Gittelman et al 1985;
80% continue into adulthood (Kessler et al 2006;
Taylor et al 1991). Subsequent prospedive clinic-based
Fayyad et al 2007). Many of those who do not continue
studies have suggested that a family history of ADHD,
co-morbid mental health problems (espedally conduct
* Address for correspondence: D. R. Coghill, Ph.D., Division of disorder) and a history of psychosocial adversity
Neuroscience, Ninewells Hospital, Dundee DDl 9SY, UK. predict persistence into adolescence (Hart et al. 1995;
(Email: Biederman et al 1996fc). Population-based studies
1088 D. R. Coghill et al

have focused on persistence into adulthood and status. They further propose that whilst dysfunction
identified higher symptom levels, co-morbid child- of the various cortical, primarily prefrontal, circuits
hood disorders (especially major depressive disorder), assodated with higher-order 'executive functioning'
parental antisocial personality disorder or paternal may partly explain the manifestation of ADHD
anxiety (Lara et al 2009) and greater impairment as symptoms, this relationship is not causal. In their
indicators of persistence (Weiss et al 1985; Kessler model, when ADHD symptoms improve over time,
et al 2005). Interestingly, the community-based study this is due to the development of compensatory 'top-
of Lara et al (2009) found neither childhood disruptive down' higher-order regulatory and executive control
behaviour disorders, nor childhood adversity, to be functions, such as planning, inhibitory control and
predictors of persistence. executive aspects of working memory. From this pos-
Few studies have investigated potential biological ition it would be predicted that functional measures
predictors of persistence of ADHD. Li et al (2007) of prefrontal cortical functiorung (primarily inhibitory
reported that better adolescent outcomes were associ- and executive measures) will be dimensionaUy related
ated with possession of at least one C alele of the to ADHD symptoms, with continuing deficits on these
1460C>T polymorphism of the monoamine oxidase tasks being more evident in those with persistent
A (MAO-A) gene. No evidence was found of an associ- symptoms than those whose symptoms have reduced
ation between outcome and polymorphisms of mono- significantly. Furthermore, they propose that, irrespec-
amine oxidase B (MAO-B). These are intriguing tive of current symptom status, subcortical brain struc-
findings as, whilst MAO-B is assodated with dopa- ture and lower-order cognitive functioning will remain
mine metabolism, MAO-A is more specifically assod- relatively unchanged during this time.
ated with serotonin. There are some data to support these hypotheses
In a longitudinal case-control neuroimaging study, (Halperin et al 2008; Bedard et al 2010). Here, re-
Shaw et al (2006) demonstrated that ADHD is assod- mission of ADHD was assodated with good perform-
ated with cortical thinning in various regions import- ance on tasks with a higher executive demand (e.g.
ant for attentional control, and that those with a inhibitory control and working memory), whilst both
poorer outcome in adolescence had 'fixed' thinning remitters and persisters demonstrated defidts on
of the left medial prefrontal cortex, which may com- tasks with lower executive demands (e.g. perceptual
promise the anterior attentional network. On the sensitivity and response variability). Unfortunately,
other hand, in those with better adolescent outcomes, whilst the childhood clinical status was well character-
right parietal cortex thickness normalized over time. ized, neuropsyehological measures were only available
Interestingly, possession of the dopamine receptor 4 in adolescence. It is therefore possible that the two
(DRD4) 7-repeat alele, a gene variation consistently groups already differed on key neuropsyehological
linked with ADHD, was assodated with cortical thin- measures at the baseline assessment. Whilst this
ning in the right orbitofrontal/inferior prefrontal and model has been highly influential in the field, not all
posterior parietal cortex, and with improved clinical existing data are supportive of it. Van Lieshout et al
outcome and normalization of the right parietal corti- (2013) systematically reviewed the literature pertaining
cal region (Shaw et al 2007fc). In the same sample, to cognitive predictors of persistence of ADHD. They
Mackie et al (2007) identified loss of volume in the concluded that, regardless of the type of cognitive
superior cerebellar vermis that persisted regardless function measured, cognitive impairments in early
of clinical outcome. Poorer clinical outcome was, childhood appear to predict the development of
however, assodated with changes in the right and ADHD a few years later. They did not, however, find
left inferior-posterior cerebellar lobes over time, which evidence to support the hypothesis that either auto-
during adolescence became progressively smaller rel- matically controlled lower-order cognitive functions,
ative to both comparison and ADHD partidpants who or more consdously controlled higher-order executive
had a better outcome. functions, differentiate persistence of ADHD from
Halperin & Schulz (2006) have proposed that the remitters. Unfortunately, the studies investigating per-
neural and cognitive mechanisms associated with sistence from childhood to adolescence (or adulthood)
cause and amelioration/'recovery' in ADHD are, at had similar design problems to those of the Halperin
least partially, separable. They suggested that ADHD and Bedard studies (Halperin et al 2008; Bedard et al
arises as a consequence of subcortical dysfunction 2010) described above and were either cross-sectional
which manifests early in life and is associated with in nature or, if longitudinal, did not measure cognitive
abnormalities in lower-order cognitive functioning functioning at both time points.
such as recognition memory. They also suggest that Using a battery of tasks with both high and low
these aspects of functioning remain relatively tem- executive demand we have previously described that,
porally stable over time irrespective of current clinical for boys aged between 7 and 14 years of age with
Neuropsychological and clinical functioning in boys with ADHD 1089

ongoing ADHD, neuropsychological development Present and Lifetime (K-SADS-PL) interview sched-
parallels that of healthy children with an average ule (Kaufman et al 1997) and met criteria for both
delay of arovmd 2 years (Coghill, 2010). We have also hyperkinetic disorder (F90 Intemational Statistical
identified several neuropsychological predictors of Classification of Diseases, tenth revision; ICD-10) and
treatment outcome (Rhodes et al 2004, 2006; Coghill ADHD combined subtype (Diagnostic and Statistical
et al. 2007), but are unaware of any previous studies Manual of Mental Disorders, fourth edition; DSM-
that have utilized baseline measures of both clinical IV). The presence of commonly co-morbid conditions
and neuropsychological performance to investigate (oppositional defiant disorder, conduct disorder and
the ongoing relationship between these outcomes in anxiety disorder) did not result in exclusion. At time
ADHD. 2, when a proportion of the partidpants was still
The present study is, therefore, the first to examine receiving medication treatment for ADHD, the clinical
prospectively the development of neuropsychological assessment focused on days when medication was not
and clinical functioning over a 4-year period in taken and the return of symptoms when medication
children and adolescents with ADHD and healthy had worn off.
controls. We hypothesized that all partidpants would
demonstrate improved neuropsychological function-
ing over time as a consequence of continuing deve-
lopment. Based on the developmental theories of Control partidpants were 17 males, matched with
Halperin & Schulz (2006) described above, we further respect to age and general intellectual ability to the
hypothesized that, for the ADHD group, symptom ADHD group and who had participated as part of a
reduction would be assodated with improved per- larger control group (n=70) in the original study.
formance on tasks with high but not low executive This group was recruited from local schools and
function demands. screened using the same methods as for the ADHD
group. Symptom-free [T score <60 on all subscales of
the 48-item Conners' Parent Rating Scale (CPRS-48),
28-item Connors' Teacher Rating Scale (CTRS-28)
Participants and Child Behavior Checklist (CBCL) subscale T scores
<60] participants and their parents were interviewed
We conducted repeat testing comparing two groups
using the K-SADS-PL to confirm health. A previous
of boys: boys with ADHD and typically developing
or current history of any psychiatric disorder led to
(TYP) healthy control boys, who had previously
partidpated in a neuropsychopharmacological study
of ADHD. Detailed descriptions of the initial assess-
ments and study have been published previously Design
(Rhodes et al 2004, 2006; Coghill et al 2007).
The British Picture Vocabulary Scale, second edition
Exclusion criteria for both groups were: history of
(BPVS; Dunn et al 1997) was used to estimate
neurological impairment; intellectual impairment
verbal ability at time 1. The BPVS was chosen for
(intelligence quotient <80); chronic physical illness;
its ease of administration, applicability to children
sensory or motor impairment; current or previous
between 3 and 15 years, and because it is less
exposure to stimulant medication; and abuse of any
heavily confounded with executive function abilities.
illegal drugs. Informed written consent to participate
At time 1, all participants were also tested on seven
in the study was obtained from each child's parent(s)/
tasks selected from the three batteries (working mem-
guardian at time 1 and again at time 2 from either the
ory and planning, visual memory, and attention) of
young person themselves (if aged ^16 years at this
the Cambridge Neuropsychological Test Automated
time) or their parent parent(s)/guardian.
Battery (CANTAB) (Morris et al 1987). At this baseline
testing session, all partidpants were naive to stimulant
ADHD group
Partidpants were 17 males who were initially recruited Partidpants in the present study were re-contacted
as part of a larger sample of 75 boys. This larger approximately 4 years after their participation in
sample was recruited from consecutive male out- the original study-time 2. They were re-consented,
patient referrals aged between 7 and 15 years to the re-interviewed by an experienced psychiatrist using
Tayside Child and Adolescent Psychiatry service. All the K-SADS-PL and re-tested on an identical battery
partidpants were interviewed by an experienced of tasks. Partidpants in the ADHD group who were
child and adolescent psychiatrist using the Kiddie- taking medication at time 2 (n=12, all immediate-
Schedule for Affective Disorders and Schizophrenia release methylphenidate) had a 72-h medication
1090 D. R. Coghill et al

Table 1. Descriptions and order of presentation of CANTAB tasks

Main outcome References for fuller

Task measures Description task description

High executive function demand tasks

Spatial working memory Between search errors, A self-ordered search task that Petiides & Milner (1982)
strategy score assesses working memory for spatial Kempton et al. (1999)
stimuli and requires a subjed to use Rhodes et al (2004)
mnemonic information to work
towards a goal
Stockings of Cambridge Problems solved in Derived from the 'Tower of Hanoi' task, Shallice (1982)
minimum moves measuring spatial planning, Kempton et al (1999)
working memory, and behavioural
Attentional set-shifting Stage reached A test of the ability to focus attention Kempton et al. (1999)
task/intradimensional on specific attributes of compound
extradimensional stimuli (intradimensional stages) and
set shifting to shift attention when required to a
previously irrelevant stimulus
dimension (extradimensional stages)
Low executive demand tasks
Pattern recognition Percentage correct A test of the ability to recognize Kempton et al. (1999)
a previously presented abstract
pattern in a forced-choice procedure
Spatial recognition Percentage correct A test of the ability to recognize the Kempton et al (1999)
spatial locations of target stimuli Rhodes et al (2004)
Delayed matching Percentage correct A test of the ability to remember Kempton et al (1999)
to sample the visual features of a complex, Rhodes et al (2004)
abstract, target stimulus and to
select from a choice of four patterns
after a variable delay
Paired associates teaming Stage reached, total A test of the ability to learn the Sahakian & Owen (1992)
errors, total trials locations of a progressively increasing
number of abstract stimuli. The main
measures in this task are the number
of trials taken to complete the task
and the total number of errors
across all trials

CANTAB, Cambridge Neuropsychological Test Automated Battery.

washout period prior to interview and neuropsycho- Data analysis

logical testing.
Analysis was conducted in three stages. First, we
examined whether participants were representative of
Neuropsychological assessment
the original sample. Second, we investigated group
The same seven tasks and same order were used at differences in clinical and neuropsychological function-
both times 1 and 2. Each partidpant performed all ing and change between times 1 and 2. Lastly, we
tasks in the same order at both testing sessions. All investigated the relationships between clinical and
tasks were presented on a high-resolution colour moni- neuropsychological change. These relationships were
tor with a touch-sensitive screen. A scheduled break of investigated from three perspectives, i.e. was clinical
approximately 10 min was taken midway through the change predicted by: neuropsychological performance
testing session and partidpants were informed that at time 1; neuropsychological performance at time 2
they could take further breaks as required. Tasks are (mirroring the analysis of Halperin et al 2008); change
described in Table 1. in neuropsychological performance between times 1
Neuropsychological and clinical functioning in boys with ADHD 1091

Table 2. Demographic and clinical characteristics

ADHD Typically developing

Timel Time 2 Time 1 Time 2

BPVS percentile rank 44.71 (31.3) N.A. 51.9 (26.9) N.A.

Age, years 10.45 (2.4) 14.65 (2.5) 10.39 (2.6) 14.47 (2.1)
Total ADHD symptoms 15.4 (2.1) 9.5 (4.5) -
Inattentive symptoms 7.8 (1.3) 5.1 (2.5) - _
Hyperactivity/impulsivity symptoms 7.6 (1.4) 4.4 (2.6) - _
ADHD combined type, n (%) 17 (100) 3(18) - -
ADHD inattentive type, n (%) 0 4(23) - -
ADHD hyperactive/impulsive type, n (%) 0 3(18) - _
No ADHD, n (%) 0 7(41) _

ADHD, Attention deficit hyperactivity disorder; BPVS, British Picture Vocabulary Scale; N.A., not available.
Data are given as mean (standard deviation) unless otherwise indicated.

and 2. Specifically, analysis of variance (ANOVA) was 70 TYP) in BPVS percentile rank, age or any of the key
used to examine the between-group differences across clinical and neuropsychological measures (all p>0.05).
the two test periods (times 1 and 2) and also to exam-
ine within-group change in performance. For the neuro- Clinical data (Table 2)
psychological performance data, repeated-measures
ANOVA was conducted on key measures with time For the ADHD group, the total number of ADHD
(time 1, time 2) as a within-subject factor and group symptoms at time 1 was significantly reduced at time
(ADHD, TYP) as a between-subject factor. Measures 2 (Fi 16=31.2, p<0.001). Similar reductions were seen
with varying difficulty levels were conducted with for hyperactive/impulsive (Fi46=22.0, p<0.001) and
an additional within-subject factor of difficulty. inattentive (Fi,i6=19.9, p<0.001) symptom scores. At
Only significant main effects or interactions involv- time 2 the diagnoses for the ADHD group were:
ing the factors of time or group were followed up DSM IV combined type ADHD, three; inattentive
with post-hoc analyses. Stepwise linear regression was type, four; hyperactive impulsive type, three. The
used to explore the relationship between neuropsycho- remainder (seven partidpants) no longer met criteria
logical performance at times 1 and 2 and change in for any form of ADHD. There was no statistically sig-
neuropsychological performance and ADHD symp- nificant assodation between age at time 2 and degree
toms between these two times. Power analyses con- of symptom reduction (r=0.12, p>0.05). None of the
ducted for the various tasks used in the study using TYP group met diagnostic criteria for ADHD at either
our previously collected published and unpublished time.
data indicated that with an a error level of 5% and
a error level of 80%, sample sizes of >17 would be Neuropsychological data at time 2
required. Of the partidpants, one ADHD boy did not complete
the Stockings of Cambridge (SOC) task and four TYP
boys did not complete the delayed matching-to-sample
(DMtS) task at time 2. Therefore data are reported
Results with 16 ADHD and 13 TYP participants for these
tasks. AU main effects of difficulty were significant
Participant characteristics (Table 2)
in the anticipated direction and are, therefore, not
A total of 17 boys with ADHD and 17 TYP boys were reported. A summary of findings and means is pre-
recruited from the cohort that partidpated in the time 1 sented in Table 3.
stijdy (Rhodes et al 2004, 2005) and reassessed
approximately 4 years later - time 2 (mean 4.14 years, Tasks with a high executive demand (Table 3)
S.D.=0.37 years). There were no differences between
the groups with respect to BPVS at time 1, or age at Spatial working memory (SWM.) (Fig. la)
times 1 or 2 (all p>0.05), or between the two follow- Repeated-measures ANOVA on between search errors
up groups and the original study groups (75 ADHD, on the SWM task revealed a main effect of time, with
1092 D. R. Coghill et al

Table 3. Neuropsychologieal performance across both testing sessions

Subject group Time

ADHD Typically developing of time of group interaction

Task Time 1 Time 2 Timel Time 2 F P F P F P

SWM total BSE 56.6 (23.0) 33.5 (19.4) 47.9 (20.8) 29.0 (15.4) 46.4 <0.001 1.4 >0.05 <1 >0.05
SWM strategy 37.2 (4.5) 33.8 (4.9) 34.9 (5.2) 32.6 (5.8) 5.5 0.03 1.9 >0.05 <1 >0.05
SOCMMS 7.0 (2.1) 9.5 (1.5) 6.9 (2.1) 8.9 (1.6) 34.0 < 0.001 <1 >0.05 <1 >0.05
ID/ED stage reached 6.9 (1.4) 8.1 (1.0) 7.8 (0.9) 8.9 (0.5) 23.6 < 0.001 10.3 0.003 <1 >0.05
Pattern recognition, % correct 80.4 (12.2) 92.4 (13.7) 88.5 (12.7) 90.4 (7.9) 10.6 0.003 <1 >0.05 5.4 0.03
Spatial recognition, % correct 66.5 (13.2) 68.2 (8.5) 72.9 (13.6) 79.7 (11.0) 3.7 >0.05 7.1 0.01 1.3 >0.05
PAL total trials 13.6 (3.7) 10.0 (2.7) 11.7 (4.6) 10.7 (3.1) 16.8 < 0.001 <1 >0.05 5.4 0.03
DMtS, % correct simultaneous 91.8 (15.9) 95.3 (8.8) 96.9 (7.5) 96.9 (7.5) <1 >0.05 1.3 >0.05 <1 >0.05
DMtS, % correct total delay 60.0 (22.4) 75.7 (13.0) 71.0 (15.8) 84.1 (16.0) 9.2 0.005 4.9 0.04 >0.05

ADHD, Attention deficit hyperactivity disorder; SWM, spatial working memory; BSE, between search errors; SOC,
Stockings of Cambridge; MMS, minimum move solutions; ID/ED, intra-dimensional extra-dimensional; PAL, paired assodates
learning; DMtS, delayed matching to sample.
Data are given as mean (standard deviation).

reduced errors at time 2. There was no significant main Attentional set shifting (intra-dimensional
effect of group or significant interactions between time extra-dimensional; ID/ED)
and group, or difficulty and group (F3 96=1.29, p>0.05).
Repeated-measures ANOVA on the stage reached
There was a significant interaction between time and
score revealed a main effect of time, with improved
difficulty (F3,96= 17.89, p< 0.001) but no significant
performance at time 2. There was a significant main
three-way interaction between time, difficulty and
effect of group, which refleded poorer performance
group (F3,96<1).
by the ADHD group across both time points. There
Repeated-measures ANOVA on strategy score re-
was no significant interaction between time and group.
vealed a main effect of time, which reflected a greater
use of a strategy at time 2. There was no effect of
Tasks with a low executive demand
group, or any interaction between time and group.
Pattern recognition memory
Repeated-measures ANOVA on percentage correct on
SOC (Fig lb)
the pattern recognition task revealed a significant effect
Repeated-measures ANOVA on the number of mini- of time, with better performance at time 2. There was
mum move solutions (MMS) on the SOC task revealed no effect of group, but there was a time and group
a main effect of time, reflecting an improved effidency interaction. It was revealed by tests that while the per-
of task completion at time 2. There was no effect of formance of TYP boys was superior to that of ADHD
group or an interaction between time and group. boys at time 1, they no longer differed at time 2.
Repeated-measures ANOVA on average number of
moves on the SOC task revealed a significant effed Spatial recognition memory
of time (fi,3o=21.68, p< 0.001), but no significant effect
Repeated-measures ANOVA on percentage correct on
of group (fi 30=1.01, p>0.05). There were no significant
the spatial recognition memory task revealed no sig-
interactions between time and group (fi3o<l), or
nificant effect of fime. There was a significant effect
difficulty and group (F3 9o<l). There was, however,
of group, which reflected poorer accuracy for the
a significant interaction between time and difficulty
ADHD group. There was no time x group interaction.
(^3,90=4.77, p=0.004). A follow-up ANOVA revealed
that the interaction reflected significantly better
Paired associates learning (PAL)
performance on the 3-, 4- and 5-move problems at
time 2 than at time 1. There was no significant three- Repeated-measures ANOVA on the total number of
way interaction between time, group and difficulty trials revealed a main effect of time, with better per-
(F3,9O<1). formance at time 2. There was no main effect of
Neuropsyehological and clinical functioning in boys with ADHD 1093

(a) Table 4. Relationship between neuropsyehological and clinical

40 1 performance: stepwise linear regression investigating the effects of
ADHD Time 1
Controls Time 1 baseline (time 1) neuropsyehological performance on clinical
ADHD Time 2 outcome (change in ADHD symptom count between times 1 and 2f
2 30 A Controls Time 2
b (S.E.)

c Constant -1.11 (0.79)
ID/ED set shifting. 1.55 (0.56) 0.59
stage reached
I 10
ADHD, Attention defidt hyperacfivity disorder; S.E.,
standard error; ID/ED, intra-dimensional extra-dimensional.
3 Box 4 Box 6 Box 8 Box 'R^=0.35.
Difficulty *p<0.05.
-- ADHD Time 1
O- Controls Time 1 Relationship between neuropsyehological functioning
- ^ ADHD Time 2
A- Controls Time 2 and clinical outcome in the ADHD group
Multiple regression was performed with change in
S 6- total ADHD symptoms between time 1 and 2 as the
dependent measure and baseline neuropsyehological
performance adjusted for age and BPVS across the
various tasks as the predictors [SWM (between search
errors and stirategy scores); SOC (MMS); ID/ED shift
(stage reached), spatial recognition (percentage cor-
rect); pattern recogrution (percentage correct); DMtS
2 Move 3 Move 4 Move 5 Move
(percentage correct simultaneous condition and all de-
lays); PAL (total trials)]. Only baseline performance in
Fig. 1. Interaction between performance and difficulty for terms of stage reached on the ID/ED set-shifting task
attention deficit hyperactivity disorder (ADHD) and control predicted clinical outcome, with better baseline per-
groups at times 1 and 2 for (a) spatial working memory formance predicting better clinical outcome (R^=0.35)
(between search errors) and (b) Stockings of Cambridge (Table 4).
(average moves to solution).
Mirroring the analysis of Halperin et al. (2008), a
multiple regression was conducted with change in
group, but there was a significant time x group inter- total ADHD symptoms as the dependent measure and
action. Post-hoc analyses revealed that while the groups time 2 neuropsyehological performance across the
differed at time 1, they did not at time 2. various tasks as the predictors. Only time 2 perform-
ance on the SWM task predicted outcome, with
superior performance on this task predicting a better
clinical outcome (R^=0.59) (Table 5).
Repeated-measures ANOVA on percentage of correct Multiple linear regression was conducted to assess
responses for the simultaneous condition of the whether change in neuropsyehological performance
DMtS task revealed no significant main effects of predicted clinical symptom reduction. Change in
time or group and no time x group interaction. total ADHD symptoms between times 1 and 2 was
Repeated-measures ANOVA conducted on percen- the dependent measure with change scores on the
tage correct for the delay conditions of the DMtS task neuropsyehological tasks as the predictors. The results
revealed a main effect of time, with better performance are shown in Table 6.
at time 2, and a main effect for group, with poorer per- Change in performance on SOC (MMS) and DMtS
formance for ADHD boys. There was no time x group (% correct across all delays) predicted change in
interaction. There were no significant interactions symptoms. Whilst greater improvement on DMtS pre-
between group and difficulty (F2,56=2.45, p>0.05) or dicted a greater symptom reduction (R^=0.25), on the
time and difficulty (F2,56<1), nor was there a three- SOC (R^=0.25) task a smaller enhancement in perform-
way interaction between group, time and difficulty ance predicted increased reduction in symptoms
1.01, p>0.05). between times 1 and 2. Further inspection of the SOC
1094 D. R. Coghill et al

Table 5. Relationship between neuropsychological and clinical Table 6. Relationship between neuropsychological and clinical
performance: stepwise linear regression investigating the effects of performance: stepwise linear regression investigating the effects of
time 2 neuropsychological performance on clinical outcome change in neuropsychological performance between times 1 and 2 on
(change in ADHD symptom count between times 1 and 2)" clinical outcome (change in ADHD symptom count between times
1 and 2J'''

b (S.E.)
Constant -11.37 (1.40)
Spatial working memory, 0.16 (0.04) 0.77** Stepl
between search errors Constant -7.99 (1.35)
Stockings of Cambridge, solved 0.87 (0.40) 0.50*
ADHD, Attention deficit hyperactivity disorder; S.E., in minimum moves
standard error. Step 2
=^=0.59. Constant -7.19 (1.19)
**p<0.001. Stockings of Cambridge, solved 0.97 (0.35) 0.56*
in minimum moves
Delayed matching to sample, -0.07 (0.03) -0.50*
performance measures indicated that this assodation % correct total delay
appeared to reflect a ceiling effect whereby those indi-
viduals who performed best on this task at baseline ADHD, Attention defidt hyperactivity disorder;
were performing near ceiling and therefore were able S.E., standard error.
to make only small improvements in task performance. ^K^=0.25 for step 1; AR^=0.25 for step 2 (p's<0.05).
This group showed greater symptom reductions than *p<0.05.
those with poorer baseline SOC performance, and
greater change in task performance. The results were time 1. Due to the limited sample size we were unable
similar when altemative measures of task performance to take into account inter-individual differences in
for SOC were used (average moves on the 5-move medication history and changes in medication status
problems, total moves). The result for DMtS remained between times 1 and 2. All participants were stimulant
significant when SOC was removed from the analysis. naive at time 1 and those with ADHD were all exposed
to methylphenidate for at least 2 months during the
initial study. Two-thirds of the ADHD group were
taking methylphenidate just prior to their time 2
assessment. For these partidpants, successful medi-
This is the first prospective description of changes in cation treatment of symptoms and/or neuropsycho-
both clinical and neuropsychological presentation of logical functioning may have made an impact on our
ADHD over development. Based on previous studies, findings. It is possible that successful treatment of
we addressed two main hypotheses. First, that over a ADHD symptoms would have reduced the clinical
4-year period, all participants, irrespective of diag- symptom ratings. However, all assessments were con-
nosis, would demonstrate improved neuropsychologi- ducted by experienced clinicians who were skilled in
cal, as well as clinical functioning, over time. This making clinical assessments of individuals taking
hypothesis was supported by the data. Second, based medications for ADHD and the assessment of continu-
on the developmental theories of Halperin & Schulz ing need for treatment in such patients. They were
(2006), we hypothesized that, within the ADHD instructed to focus on days when medication was not
group, a greater reduction in symptoms would be taken (a very common occurrence in adolescents with
assodated with larger improvements in executive func- ADHD) and times of the day when medication had
tions, but not in low executive demand neuropsycho- worn off. It is also possible that these partidpants
logical tasks. This hypothesis was not supported. may have experienced withdrawal effects due to the
The findings reported here should, however, be stopping of treatment prior to their time 2 assessment.
viewed with several limitations in mind. Despite a However, whilst there are no empirical data to
gap of 4 years between testing sessions, it is possible determine the optimal gap between last dose and
that some of the improvements noted across both neuropsychological testing, the 72-h gap between dis-
groups were due to practice effects. Our sample continuation and assessment used in this study was
size was limited and there were some missing data. equivalent to 24 half-lives for immediate-release
Both factors may have reduced our ability to detect methylphenidate and three times the usual length
significant effects. However, both of the subsamples used in other similar studies. We believe that this
remained representative of the original samples at should have been suffident to permit the return of
Neuropsychological and clinical functioning in boys with ADHD 1095

both clinical symptoms and neuropsychological diffi- whilst no such relationship was identified for low
culties and the resolution of any withdrawal/discon- executive demand tasks. They concluded that, in line
tinuation effects. The limited sample size also meant with their previously published theoretical position,
that we were unable to directly compare 'remitters' ADHD is assodated with early-appearing and endur-
with 'persisters'. However, as ADHD is a dimensional ing subcortical (and presumably low executive
rather than categorical disorder, any distinction demand cognitive) dysfunction, whilst recovery over
between persistence and remission is essentially arbi- the course of development is assodated with improve-
trary and the use of a dimensional approach with con- ments in executive functioning. Unfortunately, whilst
tinuous variables that we have followed here is, we these partidpants had clinical assessments at two
believe, more appropriate. time points, their neuropsychological performance
Clearly the finding that, at a group level, ADHD was only measured at follow-up. As ADHD is ex-
symptoms improve over time is not new. Faraone tremely heterogeneous with respect to neuropsycho-
et al (2006) found consistent evidence for an age- logical functioning (Nigg et al 2005; Coghill et al
dependent decline in symptoms over time. Whilst 2007) it is possible that those with a good clinical out-
they found a relatively low rate of diagnostic persist- come in the Halperin et al (2008) sample had always
ence, this rate increases considerably if one focuses demonstrated better executive functioning compared
on continued impairment rather than diagnostic with those with a poorer outcome. Certaiiily, our
status. Our findings with respect to neuropsychologi- data do not support the conclusions from the
cal development were also expected. The ADHD Halperin et al (2008) study. Whilst we did find that
boys improved over time on all tasks vdth the excep- better performance on the SWM task at time 2 pre-
tion of the spatial recognition task and on this task dicted clinical response, we did not find any meaning-
many ADHD boys were already performing at ceiling ful relationship between time 2 performance on the
at time 1. Our finding that, for most tasks, there was no two other CANTAB tasks that have been traditionally
interaction between group and time suggests that, at looked upon as having a 'prominent' executive com-
a group level, and similar to the evidence for brain ponent (SOC and ID/ED set-shifting tasks). Import-
maturation described by Shaw et al {2007a), the devel- antly, there was no association between changes in
opment of neuropsychological functioning in ADHD clinical status and performance on any of the high
parallels that of healthy children but with a develop- executive function demand tasks. We did, however,
mental lag. The group and time interactions that find that superior baseline performance on the ID/ED
were apparent for the pattem recognition and PAL set-shifting task predicted a better clinical outcome
tasks also seem likely to be explained by ceiling effects. and, perhaps more importantly, that a bigger improve-
The control group, but not the ADHD group, were per- ment in performance on the low executive demand
forming near ceiHng at time 1 with little room for DMtS task (a measure of short-term memory function-
further improvement. Future studies would benefit ing) also predicted a better clinical outcome. These
from the inclusion of recognition memory and learning observations are striking in the context of our previous
tasks with higher levels of difficulty. Indeed, the more findings that tWs task was associated with the greatest
demanding DMtS task demonstrated the expected pat- effect size (ADHD versus controls) at baseline, was the
tem of development over time, with parallel changes task most improved by both acute and chronic methyl-
in performance between the two groups. phenidate challenge, and was the strongest predictor of
The lack of an association between symptom clinical changes with methylphenidate (Rhodes et al
reduction and improved executive functions in the 2004, 2005, 2006; Coghill et al 2007), emphasizing
present study was somewhat, but not entirely, unex- that low as well as high executive demand cognitive
pected. Biederman et al (2007) found that the majority functions are likely to play an important role in
of subjects with an executive function defidt at base- ADHD. The recent systematic review of van Lieshout
line continued to have a defidt 7 years later (positive et al (2013) identifies weaknesses in many of the pre-
predictive value 69%), whilst the majority of those vious studies in the field and also concludes that cur-
who did not have an executive function deficit at base- rent evidence does not support the hypothesis put
line also did not at follow-up (negative predictive forward by Halperin & Schulz (2006).
value 75%). Unfortunately, they did not report the for- Our findings also suggest that the relationships
mal associations between changes in neuropsycho- between cognitive and symptomatic aspects of
logical and clinical presentation. Halperin et al (2008) ADHD seem to be more complex than previously
are the only group to have reported an association recognized. Most researchers and clinicians have
between clinical change and neuropsychological func- assumed a linear relationship whereby cognitive defi-
tioning. They found that a good clinical outcome was cits underpin symptoms (e.g. that an inhibitory
associated with better executive functions at follow-up. deficit, or aversions to delay, will result in observable
1096 D. R. Coghill et al


Symptoms Impairment


Fig. 2. Causal models for attention defidt hyperactivity disorder, (a) Traditional causal model, (b) Potential alternative
causal model.

symptoms of impulsivity, or that a defidt in working other disorders, they do not fully describe the pro-
memory wl result in inattention) (Fig. 2a). Our blems associated with ADHD. Indeed, the purpose
findings of a relative lack of association between clini- of the classification systems is to distinguish between
cal and cognitive changes over time, when considered groups rather than fully describe these groups
alongside our previous findings of a lack of association (Coghill & Sonuga-Barke, 2012). For ADHD, the symp-
between clinical and cognitive response to methyl- toms retained in the current classification systems
phenidate (Coghill et al 2007), challenge this notion. represent those that have been shown to respond to
These data suggest instead that cognitive aspects of stimulant treatment and, as we have previously de-
ADHD may sit alongside symptoms at the same monstrated, many of the cognitive deficits assodated
level of analysis within the causal model with both, with ADHD do not show a strong response to methyl-
potentially, making independent contributions to over- phenidate (Coghill et al 2007). Whilst the inclusion of
all impairment (Fig. 2b). This interpretation is consist- symptoms more closely assodated with these cognitive
ent with findings from other groups. For example, defidts may not improve our ability to detect and diag-
although working memory training has been shown nose ADHD, it may lead to a more complete descrip-
to be effective at improving various aspects of mem- tion of the difficulties faced by these children and
ory functioning by several groups (Klingberg et al suggest a broader set of treatment targets. Further
2005; Holmes et al 2010), the effects of this ti-aining examination of these relationships between the cogni-
on ADHD symptoms are much more modest tive and symptom levels of analysis and their relation-
(Sonuga-Barke et al 2013). Whilst it could be argued ship with impairment is required. To date many of the
that these findings question the importance of the cog- studies relevant to this question have been observa-
nitive impairments in ADHD, there are other interpret- tional in nature. It is now time to focus on experimental
ations. It is possible that, whilst the symptoms of studies that examine the relationships between cogni-
ADHD as presented within the various classification tive deficits and impairment and the role of symptoms
systems, appropriately distinguish those with ADHD in this relationship. Should the hypotheses described
from the general population and from those with above hold up to such testing, it will be necessary to
Neuropsychological and clinical functioning in boys with ADHD 1097

rethink our approaches to managing ADHD and inte- institutions they work for have received research sup-
grate both symptoms and cognition as important treat- port or royalties from the companies or organizations
ment targets. This would strengthen the rationale for indicated: D.R.C. (Flynn, Janssen-Cilag, Lilly, Medice,
treatments that focus on cognitive training which, as Novartis, Otsuka, Oxford University Press, Pfizer,
described above, have been demonstrated to be effec- Schering-Plough, Shire, Vifor Pharma); S.M.R. (Lilly,
tive in improving cognition in those with ADHD but Shire); and K.M. (Bristol-Myers-Squibb, Cyberonics
seem to have a much smaller effect on symptoms. Inc., GSK, Medtronic, Merck Serono, Reckitt Benckiser,
Whilst it seems likely that different pharmacological Schering-Plough, St Jude Medical).
and non-pharmacological approaches make an impact
differently on different aspects of cognitive function-
ing, this has not been well studied. A better under- References
standing of the strengths and weaknesses of the Bedard AC, Trampush JW, Newcom JH, Halperin JM (2010).
various treatment approaches may help to individua- Perceptual and motor inhibition in adolescents/young
lize treatment and would introduce a rationale for adults with childhood-diagnosed ADHD. Neuropsychology
assessing the cognitive profile of those with ADHD. 24, 424-434.
Currently testing for specific cognitive deficits does Biederman J, Faraone S, Milberger S, Curtis S, Chen L,
not add much to the assessment or cUnical manage- Marrs A, Ouellette C, Moore P, Spencer T (1996).
Predictors of persistence and remission of ADHD into
ment of those with ADHD but if it is shown that
adolescence: results from a four-year prospective follow-up
spedfic cognitive defidts have an impact on clinical study. Journal of the American Academy of Child and
status independent of symptoms then this raises the Adolescent Psychiatry 35, 343-351.
possibility of treatment approaches being individ- Biederman J, Faraone S, Milberger S, Gitite J, Mick E,
ualized according to a particular profile. It could also Chen L, Mennin D, Marrs A, Ouellette C, Moore P,
stimulate the development of new treatment appro- Spencer T, Norman D, Wilens T, Kraus I, Perrin J (19966).
aches, both pharmacological and non-pharmacological, A prospective 4-year follow-up study of attention-deficit
aimed at providing a more rational and comprehensive hyperactivity and related disorders. Archives of General
approach to treating ADHD. Psychiatry 53, 437-446.
Biederman J, Petty CR, Fried R, Doyle AE, Spencer T,
In conclusion, we now report that cognitive per-
Seidman LJ, Gross L, Poetzl K, Faraone SV (2007).
formance improves over time in boys with ADHD in a
Stability of executive function defidts into young adult
similar way, and at a similar rate, to healthy boys. years: a prospective longitudinal follow-up study of grown
However, as with cortical development, neuropsycho- up males with ADHD. Acta Psychiatrica Scandanavia 116,
logical development in ADHD is associated with a 129-136.
developmental lag. Whilst we did not find evidence to Coghill D, Sonuga-Barke EJ (2012). Annual Research Review:
support the notion that remission of ADHD symptoms categories versus dimensions in the classification and
is associated with improvements in executive functions, conceptualisation of child and adolescent mental
we did find that improvements in low executive demand disorders-implications of recent empirical study. Journal
short-term memory storage, an asped of cognitive of Child Psychology and Psychiatry 53, 469-489.
functioning that has previously been assodated with Coghill DR (2010). Heterogeneity in hyperkinetic disorder.
M.D. Thesis. University of Dundee: Dundee.
a positive clinical response to methylphenidate, were
Coghill DR, Rhodes SM, Matthews K (2007). The
associated with dinical improvement. These data high-
neuropsychological effects of chronic methylphenidate on
light the importance of monitoring cognition, symptoms drug-naive boys with attention-defidt/h5rperactivity
and cognitive aspects of treatment response over time in disorder. Biological Psychiatry 62, 954-962.
children and adolescents with ADHD. Dunn L, Dunn L, Whetton C, Burley J (1997). British Picture
Vocabulary Scale, 2nd edn. NFER-Nelson: London.
Faraone SV, Biederman J, Mick E (2006). The age-dependent
decline of attention defidt hyperactivity disorder: a
We thank the patients and investigators who took part meta-analysis of follow-up studies. Psychological Medicine
in this study. The original study was supported by a 36, 159-165.
local trust through a TENOVUS-Scotiand initiative Fayyad J, de Graaf R, Kessler R, Alonso J, Angermeyer M,
Demyttenaere K, de Girolamo G, Haro JM, Karam EG,
and the follow-up study was funded by the Chief
Lara C, Lpine JP, Ormel J, Posada-Villa J, Zaslavsky AM,
Scientist's Office, Scotiand. Jin R (2007). Cross-national prevalence and correlates of
adult attention-defidt hyperactivity disorder. British Journal
Declaration of Interest of Psychiatry 190, 402-409.
Gittelman R, Mannuzza S, Shenker R, Bonagura N (1985).
The following authors have received compensation for Hyperactive boys almost grown up. I. Psychiatric status.
serving as consultants or speakers, or they or the Archives of General Psychiatry 42, 937-947.
1098 D. R. Coghill et al

Halperin JM, Schulz KP (2006). Revisiting the role of of Medical Cenetics Part B Neuropsychiatrie Genetics 144B,
the prefrontal cortex in the pathophysiology of 430-133.
attention-deficit/hyperactivity disorder. Psychological Mackie S, Shaw P, Lenroot R, Pierson R, Greenstein DK,
Bulletin 132, 560-581. Nugent III TF, Sharp WS, Giedd JN, Rapoport JL (2007).
Halperin JM, Trampush JW, Miller CJ, Marks DJ, Cerebellar development and clinical outcome in attention
Newcom JH (2008). Neuropsyehological outcome in deficit hyperactivity disorder. American Journal of Psychiatry
adolescenfs/young adults with childhood ADHD: profiles 164, 647-655.
of persisters, remitters and confrols. Journal of Child Morris RC, Evendon JL, Sahakian BJ, Robbins TW (1987).
Psychology and Psychiatry 49, 958-966. Computer-aided assessment of dementia: comparative
Hart EL, Lahey BB, Loeber R, Applegate B, Frick PJ (1995). studies of neuropsyehological deficits in Alzheimer-type
Developmental change in attention-deticit hyperactivity dementia and Parkinson's disease. In Cognitive
disorder in boys: a four-year longitudinal study. Journal of Neurochemistry (ed. S. M. Stahl, S. D. Iverson and
Abnormal Child Psychology 23, 729-749. E. C. Goodman), pp. 21-36. Oxford Universify Press:
Holmes J, Gathercole SE, Place M, Dunning DL, Hilton KA, Oxford.
Elliott JG (2010). Working memory deficits can be Nigg JT, Willcutt EG, Doyle AE, Sonuga-Barke EJ (2005).
overcome: impacts of training and medication on working Causal heterogeneity in attention-deficit/hyperactivity
memory in children with ADHD. Applied Cognitive disorder: do we need neuropsychologically impaired
Psychology 24, 827-836. subtypes? Biological Psychiatry 57, 1224-1230.
Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P, Petrides M, Milner B (1982). Deficits on subject-ordered
Williamson D, Ryan N (1997). Schedule for Affective tasks after fronfal- and temporal-lobe lesions in man.
Disorders and Schizophrenia for School-Age Neuropsychologia 20, 249-262.
Children-Present and Lifetime Version (K-SADS-PL): initial Rhodes SM, Coghill DR, Matthews K (2004).
reliability and validity data. Journal of the American Academy Methylphenidate restores visual memory, but not working
of Child and Adolescent Psychiatry 36, 980-988. memory function in attention deficit-hyperkinetic disorder.
Kempton S, Vance A, Maniff F, Luk E, Costin J, Pantelis C Psychopharmacology (Berlin) 175, 319-330.
(1999). Executive function and attention deficit Rhodes SM, Coghill DR, Matthews K (2005).
hjqiieractivity disorder: stimulant medication and better Neuropsyehological functioning in stimulant-naive boys
executive function performance in children. Psychological with hyperkinetic disorder. Psychological Medicine 35,
Medicine 29, 527-538. 1109-1120.
Kessler RC, Adler L, Barkley R, Biederman J, Conners CK, Rhodes SM, Coghill DR, Matthews K (2006). Acute
Demler O, Faraone SV, Greenhill LL, Howes MJ, neuropsyehological effects of methylphenidate in stimulant
Secnik K, Spencer T, Ustun TB, Walters EE, drug-naive boys with ADHD II-broader executive and
Zaslavsky AM (2006). The prevalence and correlates of non-executive domains. Journal of Child Psychology and
adult ADHD in the United States: results from the National Psychiatry 47, 1184-1194.
Comorbidity Survey Replication. American Journal of Sahakian BJ, Owen AM (1992). Computerized assessment in
Psychiatry 163, 716-723. neuropsychiafry using CANTAB: discussion paper. Journal
Kessler RC, Adler LA, Barkley R, Biederman J, Conners CK, of the Royal Society of Medicine 85, 399-402.
Faraone SV, Greenhill LL, Jaeger S, Secnik K, Spencer T, Shallice T (1982). Specific impairments of planning.
Ushin TB, Zaslavsky AM (2005). Patterns and predictors of Philosophical Transactions of the Royal Society London B
attention-deficit/hyperactivity disorder persistence into Biological Sciences 298, 199-209.
adulthood: results from the National Comorbidity Survey Shaw P, Eckstrand K, Sharp W, Blumenthal J, Lerch JP,
Replication. Biological Psychiatry 57, 1442-1451. Greenstein D, Ciasen L, Evans A, Giedd J, Rapoport JL
Klingberg T, Fernell E, Olesen PJ, Johnson M, Gustafsson P, (2007). Attention-deficit/hyperactivity disorder is
Dahlstrom K, Gillberg CG, Forssberg H, Westerberg H characterized by a delay in cortical maturation. Proceedings
(2005). Computerized training of working memory in of the National Academy of Sciences USA 104, 19649-19654.
children with ADHD - a randomized, controlled trial. Shaw P, Gomick M, Lerch J, Addington A, Seal J,
Journal of the American Academy of Child and Adolescent Greenstein D, Sharp W, Evans A, Giedd JN,
Psychiatry 44, 177-186. Castellanos FX, Rapoport JL (2007;). Polymorphisms of the
Lara C, Fayyad J, de Graff R, Kessler RC, Aguilar-Gaxiola S, dopamine D4 receptor, clinical outcome, and cortical
Angermeyer M, Demytteneare K, de Girolamo G, structure in attention-deflcit/hyperactivity disorder.
Haro JM, Jin R, Karam EG, Lpine JP, Mora ME, Ormel J, Archives of Ceneral Psychiatry 64, 921-931.
Posada-Villa J, Sampson N (2009). Childhood predictors of Shaw P, Lerch J, Greenstein D, Sharp W, Ciasen L, Evans A,
adult attention-deficit/hyperactivity disorder: results from Giedd J, Castellanos FX, Rapoport J (2006). Longitudinal
the World Health Organization World Mental Health mapping of cortical thickness and clinical outcome in
Survey Initiative. Biological Psychiatry 65, 46-54. children and adolescents with attention-deficit/
Li J, Kang C, Zhang H, Wang Y, Zhou R, Wang B, Guan L, hyperactivity disorder. Archives of Ceneral Psychiatry 63,
Yang L, Faraone SV (2007). Monoamine oxidase A gene 540-549.
polymorphism predicts adolescent outcome of Sonuga-Barke EJ, Brandeis D, Crtese S, Daley D, Ferrin M,
aftention-deflcit/hyperactivity disorder. American Journal Holtmann M, Stevenson J, Danckaerts M, van der Oord S,
Neuropsychological and clinical functioning in boys with ADHD 1099

Dopfner M, Dittmann RW, Simonoff E, Zuddas A, van Lieshout M, Luman M, Buitelaar J, Rommelse NN,
Banaschewski T, Buitelaar J, Coghill D, Hollis C, Oosterlaan J (2013). Does neurocognitive
Konofal E, Lecendreux M, Wong IC, Sergeant J (2013). functioning predict future or persistence of ADHD?
Nonpharmacological interventions for ADHD: systematic A systematic review. Clinical Psychology Reviews 33,
review and meta-analyses of randomized controlled trials 539-560.
of dietary and psychological treatments. American Journal of Weiss G, Hechtman L, Milroy T, Perlman T (1985).
Psychiatry 170, 275-289. Psychiatric status of hyperactives as adults: a controlled
Taylor E, Sandberg S, Thorley G, Giles S (1991). prospective 15-year follow-up of 63 hjrperactive children.
The Epidemiology of ADHD. Oxford University Press: Journal of the American Academy of Child and Adolescent
New York. Psychiatry 24, 211-220.
Copyright of Psychological Medicine is the property of Cambridge University Press and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.