Brain iron levels in attention-deficit/hyperactivity disorder: a pilot MRI study.

Abstract

Brain iron deficiency has been supposed to be involved in the pathophysiology of ADHD. Available studies assessing iron in ADHD are based on serum ferritin, a peripheral marker of iron status. To what extent serum ferritin correlates with brain iron (BI) is unclear. The main aim of this study was to compare BI, estimated with magnetic resonance imaging (MRI) in the putamen, pallidum, caudate, and thalamus, between children with and without ADHD. The secondary aim was to assess the correlation between serum ferritin and BI levels.
Thirty-six children (18 with and 18 without ADHD, the latter including nine healthy controls and nine psychiatric controls) completed MRI and blood sampling. Brain iron levels were estimated by imaging T2*.
Children with ADHD showed significantly lower estimated BI in right and left thalamus compared to healthy controls. Estimated BI did not differ significantly between children with ADHD and psychiatric controls. Children with ADHD had significantly lower levels of serum ferritin than healthy as well as psychiatric controls. Serum ferritin and T2* values did not correlate significantly in most regions.
Low iron in the thalamus may contribute to ADHD pathophysiology.

Full text

Page 1

ORIGINAL INVESTIGATION
Brain iron levels in attention-defi cit/hyperactivity disorder:
A pilot MRI study
SAMUELE CORTESE 1,2,3,4 , ROBIN AZOULAY 5 , F. XAVIER CASTELLANOS 4,6 ,
FRAN Ç OIS CHALARD 5 , MICHEL LECENDREUX 1,7 , DAVID CHECHIN 8 ,
RICHARD DELORME1, GUY SEBAG 5 , ANDREA SBARBATI 9 , MARIE-CHRISTINE MOUREN 1 ,
BERNARDO DALLA BERNARDINA 2 & ERIC KONOFAL 1,7,10
1 Child and Adolescent Psychopathology Unit, Robert Debr é Hospital, Paris VII University, Paris, France, 2 Child Neuropsychiatry
Unit, G.B. Rossi Hospital, Department of Life and Reproduction Sciences, Verona University, Verona, Italy, 3 UMR_S INSERM
U 930, CNRS ERL 3106, Fran ç ois-Rabelais University, Child Psychiatry Centre, University Hospital, Tours, France, 4 Phyllis
Green and Randolph Cowen Institute for Pediatric Neuroscience, New York University Child Study Center, NYC, NY, USA,
5 Pediatric Imaging Department, Robert Debr é Hospital, Paris, France, 6 Nathan Kline Institute for Psychiatric Research,
Orangeburg, NYC, NY, USA, 7 Pediatric Sleep Disorders Center, Robert Debr é Hospital, Paris, France, 8 Philips Healthcare,
Suresnes, France, 9 Department of Morphological and Biomedical Sciences, Section of Anatomy and Histology, Verona University,
Verona, Italy, and 10 Sleep Disorders Center, Piti é -Salp ê tri è re Hospital, Paris, France
Abstract
Objective. Brain iron defi ciency has been supposed to be involved in the pathophysiology of ADHD. Available studies
assessing iron in ADHD are based on serum ferritin, a peripheral marker of iron status. To what extent serum ferritin cor-
relates with brain iron (BI) is unclear. The main aim of this study was to compare BI, estimated with magnetic resonance
imaging (MRI) in the putamen, pallidum, caudate, and thalamus, between children with and without ADHD. The second-
ary aim was to assess the correlation between serum ferritin and BI levels. Methods. Thirty-six children (18 with and 18
without ADHD, the latter including nine healthy controls and nine psychiatric controls) completed MRI and blood sam-
pling. Brain iron levels were estimated by imaging T2 ∗ . Results. Children with ADHD showed signifi cantly lower estimated
BI in right and left thalamus compared to healthy controls. Estimated BI did not differ signifi cantly between children with
ADHD and psychiatric controls. Children with ADHD had signifi cantly lower levels of serum ferritin than healthy as well
as psychiatric controls. Serum ferritin and T2 ∗ values did not correlate signifi cantly in most regions. Conclusions. Low iron
in the thalamus may contribute to ADHD pathophysiology.
Key words: ADHD; iron; children; thalamus; magnetic resonance imaging
Introduction
Attention-defi cit/hyperactivity disorder (ADHD) is a
highly prevalent childhood neuropsychiatric condi-
tion, estimated to affect approximately 5% of school-
age children worldwide (Polanczyk et al. 2007).
According to the criteria of the Diagnostic and Sta-
tistical Manual of Mental Disorders-Fourth Edition-
Text Revision (DSM-IV-TR) (American Psychiatric
Association 2000), ADHD is defi ned by a persistent
and age-inappropriate pattern of inattention, hyper-
activity-impulsivity or both. Despite an extensive
worldwide literature (Wolraich 1999), the genetic
(Mick and Faraone 2008) and environmental
(Millichap 2008) aetiological factors, as well as
the pathophysiology underlying ADHD, are not well
understood.
Intriguing albeit preliminary observations suggest
that iron defi ciency may be involved in the pathophys-
iology of ADHD, at least in a subset of patients
(Cortese et al. 2008; Konofal et al. 2008). Iron is an
essential trace metal, which plays a central role in a
multitude of biological processes, including many
Correspondence: Samuele Cortese, MD, PhD, Phyllis Green and Randolph Cowen, Institute for Pediatric Neuroscience, New York
University Child Study Center, 215 Lexington Ave, 14th Floor, New York, NY 10016, USA. E-mail: samuele.cortese@gmail.com
(Received 8 October 2010 ; accepted 21 February 2011 )
The World Journal of Biological Psychiatry, 2011; Early Online, 1–9
ISSN 1562-2975 print/ISSN 1814-1412 online © 2011 Informa Healthcare
DOI: 10.3109/15622975.2011.570376
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.

Page 2

2 S. Cortese et al.
unavailable for clinical investigations. Accordingly, we
selected a classical relaxometry method.
Therefore, the main aim of this study was to com-
pare brain iron levels, estimated by means of MRI
relaxometry, in a sample of children with ADHD
and age-matched controls. Iron levels were estimated
in four moderately-to-highly iron-rich regions (Mor-
ris et al. 1992) involved in the pathophysiology of
ADHD (Dickstein et al. 2006): putamen, globus pal-
lidus, caudate, and thalamus. In order to assess the
potential specifi city of brain iron defi ciency in ADHD
in relation to other psychiatric disorders, we also
included among the controls, beside healthy sub-
jects, a group of children with psychiatric disorders
other than ADHD, in order to compare: (1) children
with ADHD versus healthy controls; (2) children
with ADHD versus children with other psychiatric
disorders; (3) children with ADHD versus children
without ADHD (healthy controls plus children with
other psychiatric disorders).
The secondary aim of the study was to assess the
correlation between peripheral iron status and esti-
mated brain iron levels in the four targeted regions
in children with ADHD as well as in controls.
Methods and materials
Participants
Patients with ADHD (group 1), as well as those with
other non-ADHD psychiatric disorders (group 2),
were recruited from the outpatient consultation and
from the inpatient units of the Child and Adolescent
Psychopathology Service of the Hospital Robert
Debr é in Paris, France, between January 2007 and
October 2008. Healthy controls (group 3) were
recruited from relatives of hospital employees during
the same period. Inclusion and exclusion criteria for
all groups are reported in Table I.
As for the age range, we decided to recruit subjects
between 8 and 14 years, to include as much of the
age range most representative of “ classic ” childhood
ADHD (6 – 14); in our experience, children younger
than 8 years of age are much less likely to yield sat-
isfactory rates of completed MRI data. Right handed-
ness, assessed using the Edinburgh Handedness
Inventory (Oldfi eld 1971) was required to minimize
neurobiological heterogeneity. Diagnosis of ADHD,
as well as of other psychiatric disorders, was made
according to DSM-IV-TR criteria and confi rmed by
semi-structured interview with the Schedule for
Schizophrenia and Affective Disorders for Children –
Present and Lifetime, French version (K-SADS-PL)
(Mouren-Sim é oni et al. 2002). Patients with ADHD
(group 1) were required not to have other comorbid
psychiatric disorders, except oppositional defi ant
essential brain functions (Andrews 1999). The iron
defi ciency hypothesis of ADHD is grounded on sev-
eral lines of evidence. First, iron is a co-factor of
enzymes necessary for the synthesis and catabolism
of the monoaminergic neurotransmitters (Youdim
2000), which are implicated in the pathophysiology
of ADHD (Biederman and Faraone 2005). Second,
iron defi ciency is associated with decreased dop-
amine transporter expression (Beard et al. 1993);
variation in the dopamine transporter gene has been
linked to genetic vulnerability for ADHD (Mick and
Faraone 2008). Third, iron defi ciency may lead to
dysfunction in the basal ganglia (Youdim et al. 1989),
which are believed to play a signifi cant role in the
pathophysiology of ADHD (Kieling et al. 2008).
Fourth, iron defi ciency has been reported in children
with cognitive and behavioral impairments that
prominently include poor attention and hyperactiv-
ity (Lozoff et al. 2006).
To date, seven studies specifi cally assessing iron
status in children with ADHD have been published
(Konofal et al. 2004; Millichap et al. 2006; Oner
et al. 2008, 2010; Cortese et al. 2009; Juneja et al.
2010; Menegassi et al. 2010). All of them assessed
serum ferritin levels, a peripheral marker of iron sta-
tus (i.e. a marker of total body iron, not specifi cally
of brain iron). Five studies reported a signifi cant
inverse correlation between severity of behavioural
ADHD symptoms and serum ferritin levels (Konofal
et al. 2004; Oner et al. 2008, 2010; Cortese et al.
2009; Juneja et al. 2010), while two failed to replicate
this relationship (Millichap et al. 2006; Menegassi
et al. 2010). However, even the positive results should
be considered with caution. Given the limited and
mixed evidence from studies assessing the relation-
ship between estimated brain iron levels and periph-
eral measures of iron status (Argyropoulou et al.
2000; Haba-Rubio et al. 2005; Christoforidis et al.
2007; Godau et al. 2008), the extent to which serum
ferritin correlates with brain iron levels remains
unclear.
Since brain iron is what is expected to impact on
neuronal functions underlying ADHD symptoms,
the assessment of brain iron in ADHD is the next
logical step towards understanding the possible role
of iron defi ciency in the aetiopathophysiology of
ADHD. Magnetic resonance imaging (MRI) may be
used to estimate brain iron. MRI methods can only
provide an indirect estimation of brain iron levels.
MRI methods to indirectly measure brain iron can
be divided into classical ones, based on MRI relaxo-
metry, and novel approaches, such as susceptibility-
weighted imaging, magnitude imaging, phase imaging,
and magnetic fi eld correlation (Haacke et al. 2005).
When this study was designed (2006), these more
recent methods were still under development and
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.

Page 3

Brain iron in ADHD 3
and exclusion criteria were invited to provide a blood
sample and to participate in an MRI session.
Blood sampling
Blood samples were collected by a registered nurse
in the morning from each subject to obtain a com-
plete blood count, serum ferritin, serum iron, and
haemoglobin concentrations. Serum ferritin and iron
levels were determined with commercial kits: ferritin
levels by the Tinaquant method and iron levels by the
Ferrozine method (Roche, Basel, Switzerland), as
reported in previous studies assessing iron status in
ADHD (Konofal et al. 2004; Cortese et al. 2009).
MRI data acquisition
The three primary relaxometry metrics pertinent to
detecting brain iron are the transverse relaxation
rates and their reciprocals (i.e. R2 ? 1/T2, R2 ∗ ? 1/
T2 ∗ , and R2 ’ ? 1/T2 ’ ) (Brass et al. 2006). T2 is the
duration required for 63% of the transverse magne-
tization to be lost (Westbrook et al. 1993). T2 ∗ decay
is a combination of two effects: T2 decay effects and
dephasing due to magnetic fi eld inhomogeneities.
Therefore, T2 ∗ decay is faster than T2 decay
(Chavhan et al. 2009). R2 ’ ( ? 1/T2 ’ ) is defi ned as the
difference between R2 ∗ and R2 and arises from the
transverse relaxation mechanisms that refl ect the
reversible signal losses associated with local fi eld
inhomogeneity (Brass et al. 2006). Since brain iron
shortens transverse relaxation times, it inversely
increases relaxation rates (Brass et al. 2006). There-
fore, relaxometry metrics provide an indirect estima-
tion of regional iron levels. However, each of these
disorder (ODD), which can be associated with
ADHD in up to 84% of cases (Pliszka 2007). Intel-
lectual defi ciency [full scale Intellectual Quotient ? 70
on the Wechsler Intelligence Scale for Children Third
Edition, French version (Wechsler 1991) plus impair-
ment in adaptive function, according to criterion #2
of the DSM-IV-TR] and any neurological diseases
were excluded to minimize neurobiological heteroge-
neity. Drugs (such as antacids) or chronic diseases
(e.g., celiac disorder) that could decrease iron absorp-
tion were excluded. Anemia was defi ned according to
age-related cut-offs (Yip et al. 1984). Infl ammatory
conditions, that could alter iron status, were excluded.
In the absence of empirical evidence that ADHD
drugs alter iron status, prior stimulant treatment was
not exclusionary.
The study was conducted in accordance with the
Declaration of Helsinki (International Committee of
Medical Journal Editors 1989) and was approved by
the local Ethics Committee. Written informed con-
sent was obtained from the parents of all participants
and written assent was obtained from all children.
Procedure
First, children and their parents were administered
a clinical interview and the French version of the
semi-structured K-SADS-PL interview (Mouren-
Sim é oni et al. 2002) to detect inclusion and exclusion
criteria. Intellectual Quotient was assessed by a clin-
ical psychologist using the French version of the
Wechsler Intelligence Scale for Children, Third Edi-
tion (Wechsler 1991). The ADHD-Rating Scale-IV-
Parent was used to assess ADHD symptom severity
(Du Paul et al. 1998). Subjects meeting inclusion
Table I. Inclusion and exclusion criteria for participation in the study.
Subjects with
ADHD (Group 1)
Subjects with other psychiatric
disorders (Group 2)
Healthy controls
(Group 3)
Inclusion criteria
ADHDas per DSM-IV-TR criteria
Age range: 8 – 14 years
Right handedness
Exclusion criteria
Intellectual defi ciency
Any neurological diseases
Drugs and chronic diseases reducing
iron absorption
Current or past iron supplementation
Anaemia
Comorbid psychiatric disorders other
than ADHD, except oppositional
defi ant disorder, as per DSM-IV-TR
criteria
Inclusion criteria
Psychiatric disorders other than ADHD
(as per DSM-IV-TR criteria)
Age range: 8 – 14 years
Right handedness
Exclusion criteria
Intellectual defi ciency
Any neurological diseases
Drugs and chronic diseases reducing
iron absorption
Current or past iron supplementation
Anaemia
ADHD (as per DSM-IV-TR criteria)
Inclusion criteria
No psychiatric disorders
(as per DSM-IV-TR criteria)
Age range: 8 – 14 years
Right handedness
Exclusion criteria
Intellectual defi ciency
Any neurological diseases
Drugs and chronic diseases reducing
iron absorption
Current or past iron supplementation
Anaemia
Any psychiatric disorders (as per
DSM-IV-TR criteria)
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.

Page 4

4 S. Cortese et al.
differences of two points in T2 ∗ (R2 ∗ ). Since we
planned to compare children with ADHD versus
children without ADHD, i.e. healthy controls as well
as children with non-ADHD psychiatric disorders,
we doubled the size of the ADHD group. In order
to account for data loss from excessive subject
motion, we enrolled 28 subjects with ADHD, 14
children with non-ADHD psychopathology, and 14
healthy control subjects.
Statistical tests. Two tailed t -tests were carried out to
compare serum ferritin levels and R2 ∗ values in the
six regions of interest in children with and without
ADHD. Afterwards, in order to compare R2 ∗ values
in the regions of interest and serum ferritin levels
among group 1, 2, and 3, a MANCOVA model was
carried out considering R2 ∗ values and serum ferri-
tin levels as dependent variables, the “ group ” status
(1, 2, or 3) as fi xed variable, and “ gender ” as covari-
ate. Post-hoc analysis to assess intergroup differences
was performed according to the Bonferroni proce-
dure. In order to include in the statistical analysis a
more homogeneous group of children with ADHD,
we repeated the analyses excluding: (1) the subjects
with ADHD inattentive ( N ? 4) and hyperactive-
impulsive ( N ? 1) types; (2) the subjects who were
receiving stimulant treatment when they underwent
MRI scan ( N ? 3); and (3) the subjects with comor-
bid ODD ( N ? 2). Spearman correlations were con-
ducted between serum ferritin and R2 ∗ measures.
Statistical analyses were performed using SPSS
v15.0 (SPSS, Inc., Chicago, IL, USA).
Results
We enrolled 28 children with ADHD, 14 healthy
controls, and 14 psychiatric controls. However, we
could not use imaging data of 10 children with
ADHD, fi ve healthy controls and fi ve psychiatric
controls because of excessive motion. Therefore,
complete data were obtained for 18 children with
ADHD (13 with combined type, four with predom-
inantly inattentive type, and one with predominantly
hyperactive/impulsive type), nine healthy controls,
and nine psychiatric controls (three with generalized
anxiety disorder, two with obsessive-compulsive dis-
order, two with major depressive disorder, one with
panic disorder, and one with psychotic disorder not
otherwise specifi ed). Two children with ADHD pre-
sented with comorbid ODD. Three children with
ADHD were receiving stimulants when recruited;
the others were treatment na ï ve. Children with
ADHD (group 1) presented with a mean score on
the ADHD Rating Scale ? 41.1 ? 7.2 (raw score).
None of the subjects reported having dietary restric-
tions. Subjects ’ data are reported in Table 2.
metrics includes trade-offs: R2 is rendered less sen-
sitive by brain water content, whereas R2 ∗ is suscep-
tible to contributors other than iron, and the
sensitivity of R2 ’ is uncertain (Stankiewicz et al.
2007). An alternative method, fi eld dependent R2
increase, has been reported to have high sensitivity
and good specifi city (Bartzokis et al. 1993). How-
ever, it depends on measuring R2 at two different
fi eld intensities, i.e., two different scanners, which
would have represented a challenge for our hyperac-
tive patients.
After considering available methods, we decided
to use R2 ∗ (inverse of T2 ∗ ) to indirectly estimate
brain iron in our subjects because: (1) while the
other relaxometry metrics have been used in studies
assessing brain iron excess, we found evidence from
one study in animals (rat model) showing that R2 ∗
is reliable to estimate not only iron excess but also
iron defi ciency (Zywicke et al. 2002); (2) two inde-
pendent groups (Aquino et al. 2009; Peran et al.
2009) have demonstrated, in humans, a signifi cant
correlation between R2 ∗ values in specifi c brain
regions and the values of iron measured post-
mortem (Hallgren and Sourander 1958).
Brain imaging was performed with a 1.5 Tesla
Intera scanner (Philips Healthcare, Best, The Neth-
erlands). A multishot echo planar imaging sequence
was modifi ed as described by Dahnke and Schaeffter
(2005) to obtain a multi echo gradient sequence.
The images at different echo times were used to
derive the T2 ∗ maps. The sequence parameters were:
fi eld of view 28 cm, matrix acquisition: 160/160;
number of axial slices: 6; slice thickness: 4 mm; rep-
etition time: 618 ms; 17 echoes, with a time interval
between echo, Delta TE ? 5 ms; fl ip angle: 50 ° ;
acquisition time: 4 min.
The boundaries of the entire right and left thala-
mus, head of the caudate, putamen, and globus pal-
lidus were manually traced on T2 ∗ maps by the
second author (RA) who was blind to diagnostic
group. The sequence was acquired in the axial com-
missural planes that included the basal ganglia and
thalamus. T2 ∗ and R2 ∗ relaxation rates in these
regions were estimated with Levenberg – Marquardt
exponential fi tting as described by Danhke and
Schaeffter (2005) (MRI T2 ∗ Analysis Tools, Philips
Investigational Tools).
Statistical analysis
Sample size and power analysis. Based on data pro-
vided by Allen et al. (2001), we calculated that
including a minimum of 10 patients in each of the
three groups (children with ADHD, children with
other psychiatric disorders, and healthy controls)
would provide 85% power to detect between-group
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.

Page 5

Brain iron in ADHD 5
with ADHD and those with other psychiatric disor-
ders in any brain region examined ( P ? 0.05).
Children with ADHD had signifi cantly lower
serum ferritin levels compared to psychiatric con-
trols ( P ? 0.002).
Results did not change signifi cantly when we
excluded from the analyses: (1) the four subjects
with ADHD inattentive type and the subject with
ADHD hyperactive-impulsive type (post-hoc tests
with Bonferroni procedure: R2 ∗ values in ADHD vs.
healthy controls in right thalamus: P ? 0.006, in left
thalamus: P ? 0.004, serum ferritin levels: P ? 0.004);
(2) the three ADHD subjects who were receiving
stimulant treatment when they underwent MRI scan
(post-hoc tests with Bonferroni procedure: R2 ∗ val-
ues in ADHD vs. healthy controls in right thalamus:
P ? 0.003, in left thalamus: P ? 0.001, serum fer-
ritin levels: P ? 0.007); and (3) the two ADHD sub-
jects with comorbid ODD (post-hoc tests with
Bonferroni procedure: R2 ∗ values in ADHD vs.
healthy controls in right thalamus: P ? 0.006, in left
thalamus: P ? 0.003, serum ferritin levels: P ? 0.006).
It is known that the MANCOVA procedure protects
against Type I errors that might occur if multiple
ANCOVAs were conducted independently (Katz
1999). However, even after performing Bonferroni
correction considering the seven independent vari-
ables tested (0.05/7 ? 0.007), all our fi ndings from
the post-hoc tests would still hold signifi cant.
Ferritin levels did not correlate signifi cantly with
R2 ∗ in the entire sample (except in the right caudate,
r ? 0.33, n ? 36, P ? 0.048), nor within any of the
groups.
Discussion
To our knowledge, this is the fi rst study that esti-
mated brain iron levels in children with ADHD and
Comparisons between ADHD and non ADHD subjects
Preliminary non-corrected analyses by means of two
tailed t -tests revealed that children with ADHD had
signifi cantly lower R2 ∗ values (i.e. lower estimated
iron levels) in right ( P ? 0.005) and left ( P ? 0.008)
thalamus as well as right caudate ( P ? 0.032), along
with a trend in right putamen ( P ? 0.083) compared
to children without ADHD. Children with ADHD
had signifi cantly lower serum ferritin levels than chil-
dren without ADHD ( P ? 0.001). After performing
the stringent Bonferroni correction considering the
seven independent variables tested (six brain regions
of interest plus serum ferritin levels, 0.05/7 ? 0.007),
the difference in the right thalamus and in the serum
ferritin levels still held signifi cant; the difference in the
left thalamus was marginally signifi cant ( P ? 0.008).
Comparisons among ADHD, non-ADHD psychiatric
controls and healthy controls
MANCOVA model revealed no signifi cant interac-
tion between group status and gender (Hotelling ’ s
trace test: P ? 0.072). The variable “ gender ” failed to
have a signifi cant effect in tests of “ between subjects
effect ” for any independent variable. On the other
hand, the variable “ group ” did have a signifi cant
effect for “ R2 ∗ in right thalamus ” ( P ? 0.004), “ R2 ∗
in left thalamus ” ( P ? 0.003), and “ serum ferritin
levels ” ( P ? 0.001). Post-hoc tests with Bonferroni
procedure for multiple comparisons among groups
revealed that children with ADHD ( N ? 18) had
signifi cantly lower R2 ∗ values in right thalamus
( P ? 0.003) and left thalamus ( P ? 0.002), as well
as signifi cantly lower serum ferritin levels ( P ? 0.006)
compared to healthy children ( N ? 8). The right and
left thalamus are indicated in Figure 1 on a standard
brain. Post-hoc tests with Bonferroni procedure
revealed no signifi cant differences between children
Table II. Demographic characteristics, serum ferritin, haemoglobin, and R2 ∗ (s – 1 ) values in the three study groups. Values are means and
standard deviations.
ADHDPsychiatric controls Healthy controls
N
M/F
Age (months)
Serum ferritin (ng/ml)
Haemoglobin (g/dl)
R2 ∗
Right thalamus
Left thalamus
Right globus pallidus
Left globus pallidus
Right putamen
Left putamen
Right caudate
Left caudate
18
16/2
99
5/4 5/4
118.8 (18.2)
32.4 (13.4)
12.7 (1.1)
123.5 (22.6)
54.0 (12.21)
12.2 (1.3)
120.8 (26.3)
51.6 (16.4)
12.9 (1.6)
14.61 (1.16)
14.80 (1.30)
17.49 (1.91)
17.44 (1.87)
15.01 (1.14)
15.00 (1.15)
14.79 (1.42)
14.64 (1.08)
15.76 (1.85)
15.62 (1.86)
17.89 (3.56)
18.78 (3.1)
15.70 (2.37)
15.70 (2.35)
16.07 (1.69)
15.87 (1.76)
17.71 (4.21)
17.71 (3.1)
18.33 (2.82)
17.68 (2.84)
16.39 (2.25)
15.54 (3.30)
16.01 (2.40)
15.33 (2.82)
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.

Page 6

6 S. Cortese et al.
Regionally circumscribed defi cits in brain iron
would be expected to impair myelination and cate-
cholaminergic neurotransmission (Connor and
Menzies 1996). Our results highlight the potential
importance of the thalamus in ADHD, a critical
component of cortico-striato-thalamo-cortical cir-
cuits subserving relevant regulatory processes (Dick-
stein et al. 2006). While neuroimaging studies of
ADHD have not generally focused specifi cally on
thalamic nuclei, thalamic abnormalities have been
noted with increasing frequency with a range of neu-
roimaging techniques (Arcos-Burgos et al. 2010;
Ivanov et al. 2010). Interestingly, the thalamus has
been involved in the regulation of cortical arousal
through thalamo-cortical connections (Montaron
and Buser 1988). In a recent meta-analysis (Cortese
et al. 2009), we concluded that preliminary evidence
from two studies (Lecendreux et al. 2000; Golan
et al. 2004) performed by means of the Multiple
Sleep Latency Test indicates that children with
ADHD have signifi cantly lower mean times to fall
asleep during the day than healthy controls, indicat-
ing that children with ADHD display a tendency to
be sleepier than normal controls during daytime.
Moreover, recent data on cyclic alternating pattern
in subjects with ADHD confi rmed that they may
present with a hypoaroused state (Miano et al. 2006).
This preliminary evidence seems to support the
“ hypoarousal ” theory of ADHD (Weinberg and
Brumback 1990), according to which children with
ADHD, or at least a subsample of them, are sleepier
than controls and might use excessive motor activity
as a strategy to stay awake and alert. Therefore, the
supposed impairment of thalamus functioning, asso-
ciated with iron defi ciency, might contribute to
ADHD symptoms via its impact on alertness.
controls. It is also the fi rst study that and assessed
the correlation between estimated brain iron levels
and serum ferritin levels. We found signifi cantly
lower thalamic iron levels in children with ADHD
compared to healthy children and, in general, to
non-ADHD controls (with a marginal signifi cant
difference for left thalamus after correction for mul-
tiple comparisons). Contrary to histochemical data
(Hallgren and Sourander 1958), iron levels in the
thalamus in our healthy controls were higher than in
our striatal regions of interest. However, Hallgreen
and Sourander assessed adult subjects (30 – 100
years) and young persons who had died in accidents.
Since ADHD individuals are at higher risk than per-
sons without ADHD to be involved in accidents
(Barkley 2002), it is possible that some of these
young persons would have been diagnosed with
ADHD, making this a problematic comparison
group. We also note that R2 ∗ values in our healthy
subjects approximate those reported for children of
similar ages in a recent study (Aquino et al. 2009).
We failed to fi nd signifi cantly lower iron levels in
the other regions we assessed, as we expected on
the base of pathophysiological models of ADHD.
It is possible that this is due, at least in part, to the
statistical test (MANCOVA) with the stringent
correction for multiple analyses that we chose
(Bonferroni). Indeed, in preliminary analyses that
we conducted to compare R2 ∗ among the three
groups with Kruskal – Wallis and Mann – Whitney
tests (non-corrected), we found signifi cantly lower
iron levels also in right caudate in children with
ADHD compared to healthy controls. It is also
possible that a larger sample would reveal signifi -
cantly lower levels of iron in ADHD in other struc-
tures of the basal ganglia.

Figure 1. The regions where R2∗ was signifi cantly lower in ADHD vs. healthy controls (right and left thalamus are reported in yellow
(sagittal, coronal, and transverse planes). The brain is the standard Montreal Neurological Institute (MNI) atlas MNI152 1-mm brain
template. X, Y, Z represent the MNI coordinates.
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.

Page 7

Brain iron in ADHD 7
study, it would be interesting to assess the thalamic
levels of glutamate, as well as of other neurotransmit-
ters possibly altered in ADHD.
Our study should be considered in light of several
limitations. First, the small number of subjects in
our comparison groups limited our statistical power.
Therefore, our fi ndings must be considered tenta-
tive until replicated in larger samples. Second,
although R2 ∗ provides a suitable means of estimat-
ing regional brain iron levels, it is susceptible to
local variations of magnetic fi eld associated with sig-
nal loss unrelated to the internal iron content of the
tissue (Brass et al. 2006). However, the effect of
inhomogeneities on R2 ∗ scales linearly with fi eld
strength (Dahnke and Schaeffter 2005). Therefore,
we expect such effects to be weaker at 1.5 than at
3.0 Tesla. Novel MRI approaches are being devel-
oped that have advantages relative to relaxometry
(Brass et al. 2006). Ongoing work will use magnetic
fi eld correlation methods which estimate brain iron
more reliably (Jensen et al. 2006). Third, the groups
were not precisely matched for sex ratio. The preva-
lence of males was higher in children with ADHD
than in the other two groups, refl ecting the preva-
lence of clinically referred populations (Novik et al.
2006). Regional brain iron levels were found to be
signifi cantly lower in adult women (Bartzokis et al.
2007) in one study but not in another (Xu et al.
2008). At any rate, if ferritin in brain is lower in
female children and adolescents as reported for
older adults (Bartzokis et al. 2007), then matching
for sex would have been expected to enhance our
results. Of note, the effect of gender was not sig-
nifi cant in the MANCOVA model. Fourth, our
study was not powered to examine potential differ-
ences among the ADHD subtypes. However, after
excluding subjects with ADHD inattentive and
hyperactive-impulsive types, our results hold sig-
nifi cant. Finally, we did not take potential volumet-
ric differences into account as we are unaware of a
relationship between R2 ∗ and regional brain vol-
umes and volumetric studies require substantially
larger samples. However, such differences in the
thalamus or striatum would be expected to be rela-
tively minor compared to the magnitude of T2 ∗ and
R2 ∗ differences we observed.
Conclusions
Despite limitations, our fi ndings suggest that thal-
amic iron defi ciency contributes to the pathophysiol-
ogy of ADHD. Replication in larger samples would
provide a rationale for controlled trials of iron sup-
plementation for ADHD associated with regional
brain iron defi ciency, even in the absence of haema-
tological abnormalities such as anaemia. Indeed,
Thalamic differences were found between chil-
dren with ADHD and healthy controls but not
between children with ADHD and psychiatric. How-
ever, children with ADHD had signifi cantly lower
serum ferritin levels than psychiatric controls. There-
fore, we suggest that further larger studies are needed
to assess if iron defi ciency (central and/or peripheral)
is specifi c for ADHD or is also found in other psy-
chiatric disorders.
As for serum ferritin levels, we found that they
were signifi cantly lower in children with ADHD ver-
sus healthy controls as well as versus non-ADHD
psychiatric controls. Our serum ferritin results rep-
licate and extend previous fi ndings (Konofal et al.
2004; Oner et al. 2008, 2010; Cortese et al. 2009;
Juneja et al. 2010). However, although serum ferritin
and R2 ∗ were related, this relationship did not attain
statistical signifi cance in most regions. We speculate
that the lack of a tight relationship between serum
ferritin and brain indices of iron in ADHD may
resemble insights obtained in the study of Restless
Legs Syndrome (RLS). Indeed, a similar lack of sig-
nifi cant correlation between estimated brain iron
and serum ferritin has been reported in RLS (Godau
et al. 2008). Allen and Earley (2007) observed that
patients with RLS appear to have marginal central
nervous system iron levels that can become insuffi -
cient when deprived of normal access to adequate
peripheral iron or may be insuffi cient even with nor-
mal access to adequate peripheral iron. An alteration
in the entry of iron into the brain across the blood –
brain barrier has been hypothesized in RLS (Pic-
chietti 2007) along with the suggestion that it may
account for the lack of signifi cant correlation between
peripheral and central iron values. Accordingly, we
speculate that dysfunction in the blood-brain barrier
or iron transport mechanisms in children with
ADHD may account for the mismatch between
peripheral and central iron. Therefore, we conclude
that research on iron status in ADHD should not
rely exclusively on serum ferritin levels.
Our study lays the groundwork for further imaging
studies aimed at better understanding the effects of
iron defi ciency on neurotransmitters concentration.
In this regard, magnet resonance spectroscopy may
provide useful insights. By means of this technique,
it has been found that iron defi cient mice have higher
striatal levels of glutamate (Ill et al. 2006). Interest-
ingly, a recent systematic review and meta-analysis
(Perlov et al. 2009) reported evidence pointing to
increased levels of glutamate/glutamine in the left
striatum of children with ADHD. Perlov and co-
workers also suggested that other regions poorly
investigated with spectroscopy, such as the thalamus,
should be assessed in future spectroscopic studies in
ADHD. Therefore, considering the results of our
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.

Page 8

8 S. Cortese et al.
effectiveness of stimulant medication. Mol Psychiatry 47(11):
1053–66.
Argyropoulou MI, Metafratzi Z, Kiortsis DN, Bitsis S, Tsatsoulis
A, Efremidis S. 2000. T2 relaxation rate as an index of pituitary
iron overload in patients with beta-thalassemia major. AJR Am
J Roentgenol 175:1567 – 1569.
Barkley RA. 2002. Major life activity and health outcomes associ-
ated with attention-defi cit/hyperactivity disorder. J Clin Psy-
chiatry 63(Suppl 12):10 – 15.
Bartzokis G, Aravagiri M, Oldendorf WH, Mintz J, Marder SR.
1993. Field dependent transverse relaxation rate increase may
be a specifi c measure of tissue iron stores. Magn Reson Med
29:459 – 464.
Bartzokis G, Tishler TA, Lu PH, Villablanca P, Altshuler LL,
Carter M, et al. 2007. Brain ferritin iron may infl uence age- and
gender-related risks of neurodegeneration. Neurobiol Aging
28:414 – 423.
Beard JL, Connor JR, Jones BC. 1993. Iron in the brain. Nutr
Rev 51:157 – 170.
Biederman J, Faraone SV. 2005. Attention-defi cit hyperactivity
disorder. Lancet 366:237 – 248.
Brass SD, Chen NK, Mulkern RV, Bakshi R. 2006. Magnetic
resonance imaging of iron deposition in neurological disorders.
Top Magn Reson Imaging 17:31 – 40.
Chavhan GB, Babyn PS, Thomas B, Shroff MM, Haacke EM.
2009. Principles, techniques, and applications of T2 ∗ -based
MR imaging and its special applications. Radiographics
29:1433 – 1449.
Christoforidis A, Haritandi A, Perifanis V, Tsatra I, Athanassiou-
Metaxa M, Dimitriadis AS. 2007. MRI for the determination
of pituitary iron overload in children and young adults with
beta-thalassaemia major. Eur J Radiol 62:138 – 142.
Connor JR, Menzies SL. 1996. Relationship of iron to oli-
godendrocytes and myelination. Glia 17:83 – 93.
Cortese S, Lecendreux M, Dalla Bernardina B, Mouren MC,
Sbarbati A, Konofal E. 2008. Attention-defi cit/hyperactivity
disorder, Tourette’s syndrome, and restless legs syndrome: the
iron hypothesis. Med Hypotheses 70:1128 – 1132.
Cortese S, Faraone SV, Konofal E, Lecendreux M. 2009a. Sleep
in children with attention-defi cit/hyperactivity disorder: meta-
analysis of subjective and objective studies. J Am Acad Child
Adolesc Psychiatry 48:894 – 908.
Cortese S, Konofal E, Dalla Bernardina B, Mouren MC, Lecen-
dreux M. 2009b. Sleep disturbances and serum ferritin levels
in children with attention-defi cit/hyperactivity disorder. Eur
Child Adolesc Psychiatry 18:393 – 399.
Dahnke H, Schaeffter T. 2005. Limits of detection of SPIO at 3.0
T using T2 relaxometry. Magn Reson Med 53:1202 – 1206.
Dickstein SG, Bannon K, Castellanos FX, Milham MP. 2006. The
neural correlates of attention defi cit hyperactivity disor der: an
ALE meta-analysis. J Child Psychol Psychiatry 47:1051 – 1062.
Du Paul GJ, Power TJ, Anastopoulos AD, Reid R. 1998. ADHD
rating scale-IV: checklists, norms, and interpretation. New
York: Guilford Press.
Godau J, Klose U, Di Santo A, Schweitzer K, Berg D. 2008.
Multiregional brain iron defi ciency in restless legs syndrome.
Mov Disord 23:1184 – 1187.
Golan N, Shahar E, Ravid S, Pillar G. 2004. Sleep disorders and
daytime sleepiness in children with attention-defi cit/hyperactive
disorder. Sleep 27:261 – 266.
Haacke EM, Cheng NY, House MJ, Liu Q, Neelavalli J, Ogg RJ,
et al. 2005. Imaging iron stores in the brain using magnetic
resonance imaging. Magn Reson Imaging 23:1 – 25.
Haba-Rubio J, Staner L, Petiau C, Erb G, Schunck T, Macher JP.
2005. Restless legs syndrome and low brain iron levels in
patients with haemochromatosis. J Neurol Neurosurg Psychia-
try 76:1009 – 1010.
currently there is limited evidence, from an open-
label trial (Sever et al. 1997) and a double-blind pilot
trial (Konofal et al. 2008) that oral iron supplemen-
tation may improve ADHD symptoms. Moreover,
since iron defi ciency may limit the effi cacy of treat-
ments for ADHD (Youdim 2000), we speculate that
indicated iron supplementation might improve treat-
ment outcomes in ADHD, at least in a subset of
children with ADHD.
Acknowledgments
None.
Statement of Interest
Dr Cortese has received fi nancial support to attend
medical meetings from Eli Lilly & Company and
Shire Pharmaceuticals, and has been co-investigator
in studies sponsored by GlaxoSmithKline, Eli Lilly
& Company, and Genopharm. He serves as a con-
sultant for Shire Pharmaceuticals.
Dr Lecendreux has received research funding
from Shire. He has served as a consultant for UCB
and Shire Pharmaceuticals. He has served as an
investigator for Eli Lilly & Company and Shire Phar-
maceuticals.
Dr Chechin is employed by Philips Healthcare.
Dr Konofal has served on advisory boards for
Shire Pharmaceuticals and UCB. He has consulted
for Shire Pharmaceuticals. He has served as a med-
ical writer for Remidica and Janssen-Cilag. He has
served on the speaker ' s bureau of UCB and served
as a principal investigator in clinical trials supported
by Eli Lilly & Company and Janssen-Cilag.
Drs Azoulay, Castellanos, Chalard, Sebag, Sbar-
bati, Delorme, Mouren, and Dalla Bernardina report
no competing interests.
References
Allen RP, Earley CJ. 2007. The role of iron in restless legs syn-
drome. Mov Disord 22(Suppl 18):S440 – 448.
Allen RP, Barker PB, Wehrl F, Song HK, Earley CJ. 2001. MRI
measurement of brain iron in patients with restless legs syn-
drome. Neurology 56:263 – 265.
American Psychiatric Association. 2000. Diagnostic and statistical
manual of mental disorders, 4th ed. Text Revision. American
Psychiatric Association, Washington, DC.
Andrews NC. 1999. Disorders of iron metabolism. New Engl J
Med 341:1986 – 1995.
Aquino D, Bizzi A, Grisoli M, Garavaglia B, Bruzzone MG, Nar-
docci N, et al. 2009. Age-related iron deposition in the basal
ganglia: quantitative analysis in healthy subjects. Radiology
252:165 – 172.
Arcos-Burgos M, Jain M, Acosta MT, Shively S, Stanescu H,
Wallis D, et al. 2010. A common variant of the latrophilin
3 gene, LPHN3, confers susceptibility to ADHD and predicts
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.

Page 9

Brain iron in ADHD 9
Novik TS, Hervas A, Ralston SJ, Dalsgaard S, Rodrigues PR,
Lorenzo MJ. 2006. Infl uence of gender on attention-defi cit/
hyperactivity disorder in Europe – ADORE. Eur Child Adolesc
Psychiatry 15(Suppl 1):I15 – 24.
Oldfi eld RC. 1971. The assessment and analysis of handedness:
the Edinburgh inventory. Neuropsychologia 9:97 – 113.
Oner O, Alkar OY, Oner P. 2008. Relation of ferritin levels with
symptom ratings and cognitive performance in children with
attention defi cit-hyperactivity disorder. Pediatr Int 50:40 – 44.
Oner O, Oner P, Bozkurt OH, Odabas E, Keser N, Karadag H,
et al. 2010. Effects of zinc and ferritin levels on parent and
teacher reported symptom scores in attention defi cit hyperac-
tivity disorder. Child Psychiatry Hum Dev 41:441 – 447.
Peran P, Cherubini A, Luccichenti G, Hagberg G, Demonet JF,
Rascol O, et al. 2009. Volume and iron content in basal ganglia
and thalamus. Hum Brain Mapp 30:2667 – 2675.
Perlov E, Philipsen A, Matthies S, Drieling T, Maier S, Bubl E,
et al. 2009. Spectroscopic fi ndings in attention-defi cit/hyperac-
tivity disorder: review and meta-analysis. World J Biol Psychia-
try 10:355 – 365.
Picchietti D. 2007. Is iron defi ciency an underlying cause of pedi-
atric restless legs syndrome and of attention-defi cit/hyperactiv-
ity disorder? Sleep Med 8:693 – 694.
Pliszka S. 2007. Practice parameter for the assessment and treat-
ment of children and adolescents with attention-defi cit/hyper-
activity disorder. J Am Acad Child Adolesc Psychiatry 46:
894 – 921.
Polanczyk G, de Lima MS, Horta BL, Biederman J, Rohde LA.
2007. The worldwide prevalence of ADHD: a systematic review
and metaregression analysis. Am J Psychiatry 164:942 – 948.
Sever Y, Ashkenazi A, Tyano S, Weizman A. 1997. Iron treatment
in children with attention defi cit hyperactivity disorder. A pre-
liminary report. Neuropsychobiology 35:178 – 180.
Stankiewicz J, Panter SS, Neema M, Arora A, Batt CE, Bakshi R.
2007. Iron in chronic brain disorders: imaging and neurother-
apeutic implications. Neurotherapeutics 4:371 – 386.
Wechsler. 1991. Echelle d’intelligence de Wechsler pour enfants;
troisi è me edition.
Weinberg WA, Brumback RA. 1990. Primary disorder of vigi-
lance: a novel explanation of inattentiveness, daydreaming,
boredom, restlessness, and sleepiness. J Pediatr 116:720 – 725.
Westbrook C, Kaut Roth C, Talbot J. 1993. MRI in practice.
Oxford: Blackwell Publishing.
Wolraich ML. 1999. Attention Defi cit hyperactivity disorder: the
most studied and yet the most controversial diagnosis. Ment
Retard Dev Disabil Res Rev 5:163 – 168.
Xu X, Wang Q, Zhang M. 2008. Age, gender, and hemispheric
differences in iron deposition in the human brain: an in vivo
MRI study. Neuroimage 40:35 – 42.
Yip R, Johnson C, Dallman PR. 1984. Age-related changes in
laboratory values used in the diagnosis of anemia and iron defi -
ciency. Am J Clin Nutr 39:427 – 436.
Youdim MB. 2000. Nutrient deprivation and brain function: iron.
Nutrition 16:504 – 508.
Youdim MB, Ben-Shachar D, Yehuda S. 1989. Putative biological
mechanisms of the effect of iron defi ciency on brain biochem-
istry and behavior. Am J Clin Nutr 50:607 – 615.
Zywicke HA, van Gelderen P, Connor JR, Burdo JR, Garrick MD,
Dolan KG, et al. 2002. Microscopic R2 ∗ mapping of reduced
brain iron in the Belgrade rat. Ann Neurol 52:102 – 105.
Hallgren B, Sourander P. 1958. The effect of age on the non-
haemin iron in the human brain. J Neurochem 3:41 – 51.
Ill AM, Mitchell TR, Neely EB, Connor JR. 2006. Metabolic
analysis of mouse brains that have compromised iron storage.
Metab Brain Dis 21:77 – 87.
International Committee of Medical Journal Editors. 1989. State-
ments from the Vancouver Group. Br Med J 299:1394 – 1395.
Ivanov I, Bansal R, Hao X, Zhu H, Kellendonk C, Miller L,
et al. 2010. Morphological abnormalities of the thalamus
in youths with attention defi cit hyperactivity disorder. Am J
Psychiatry 167:397 – 408.
Jensen JH, Chandra R, Ramani A, Lu H, Johnson G, Lee SP,
et al. 2006. Magnetic fi eld correlation imaging. Magn Reson
Med 55:1350 – 1361.
Juneja M, Jain R, Singh V, Mallika V. 2010. Iron defi ciency in
indian children with attention defi cit hyperactivity disorder.
Indian Pediatr. 47(11):955–8
Katz MH. 1999. Multivariable analysis: a practical guide for clini-
cians. Cambridge: Cambridge University Press. p. 55.
Kieling C, Goncalves RR, Tannock R, Castellanos FX. 2008.
Neurobiology of attention defi cit hyperactivity disorder. Child
Adolesc Psychiatr Clin N Am 17:285 – 307, viii.
Konofal E, Lecendreux M, Arnulf I, Mouren MC. 2004. Iron
defi ciency in children with attention-defi cit/hyperactivity disor-
der. Arch Pediatr Adolesc Med 158:1113 – 1115.
Konofal E, Lecendreux M, Deron J, Marchand M, Cortese S,
Zaim M, et al. 2008. Effects of iron supplementation on atten-
tion defi cit hyperactivity disorder in children. Pediatr Neurol
38:20 – 26.
Lecendreux M, Konofal E, Bouvard M, Falissard B, Mouren-
Simeoni MC. 2000. Sleep and alertness in children with
ADHD. J Child Psychol Psychiatry 41:803 – 812.
Lozoff B, Beard J, Connor J, Barbara F, Georgieff M, Schallert T.
2006. Long-lasting neural and behavioral effects of iron defi -
ciency in infancy. Nutr Rev 64:S34 – S43.
Menegassi M, Mello ED, Guimaraes LR, Matte BC, Driemeier
F, Pedroso GL, et al. 2010. Food intake and serum levels of
iron in children and adolescents with attention-defi cit/hyperac-
tivity disorder. Rev Bras Psiquiatr 32:132 – 138.
Miano S, Donfrancesco R, Bruni O, Ferri R, Galiffa S, Pagani J,
et al. 2006. NREM sleep instability is reduced in children with
attention-defi cit/hyperactivity disorder. Sleep 29:797 – 803.
Mick E, Faraone SV. 2008. Genetics of attention defi cit hyperactiv-
ity disorder. Child Adolesc Psychiatr Clin N Am 17:261 – 8.
Millichap JG. 2008. Etiologic classifi cation of attention-defi cit/
hyperactivity disorder. Pediatrics 121:e358 – e365.
Millichap JG, Yee MM, Davidson SI. 2006. Serum ferritin in
children with attention-defi cit hyperactivity disorder. Pediatr
Neurol 34:200 – 203.
Montaron MF, Buser P. 1988. Relationships between nucleus
medialis dorsalis, pericruciate cortex, ventral tegmental area
and nucleus accumbens in cat: an electrophysiological study.
Exp Brain Res 69:559 – 566.
Morris CM, Candy JM, Oakley AE, Bloxham CA, Edwardson JA.
1992. Histochemical distribution of non – haem iron in the
human brain. Acta Anat (Basel) 144:235 – 257.
Mouren-Sim é oni, Cook S, Flament S, Nedey-Saiag MC, Nollet-
Cl é men ç on C, Pailler-Martinot ML. 2002. Kiddie-SADS-
version vie enti è re (K-SADS-P/L). Edition revis é e et modifi é e.
Paris: Editions INSERM.
World J Biol Psychiatry Downloaded from informahealthcare.com by Nyu Medical Center on 05/18/11
For personal use only.