Grey matter volumes in treatment naïve vs. chronically treated children with attention deficit/hyperactivity disorder: a combined approach
Introduction
Psychostimulants such as methylphenidate or amphetamines are the most commonly used psychotropic medications in children and adolescents (Zuvekas and Vitiello, 2012). In 2008, it was estimated that around 4 million children were treated with stimulants each day in the USA (Zuvekas and Vitiello, 2012). Methylphenidate, in particular, is the first-line treatment for attention deficit/hyperactivity disorder (ADHD) (Biederman and Faraone, 2005). Despite its widespread prescription, an ongoing debate exists regarding its long-term effects on the brain, especially regarding substance abuse liability (Kim et al., 2009, Volkow, 2012). Methylphenidate indeed increases dopamine levels in the nucleus accumbens, a pharmacological effect underlying drug reward that may trigger the neuroadaptations in dopaminergic and glutamatergic signaling associated with addiction (Volkow, 2012). Due in part to their impulsivity, children and adolescent with ADHD are known to be at risk for the development of substance abuse (Levy et al., 2014). In 2006, a report indicated that more than 7 million people in the US had abused ADHD stimulants, and that as many as 750,000 teenagers may show signs of addiction (Kroutil et al., 2006).
In this context, neuroimaging techniques represent valuable tools to track down the potential effects of medication on the human brain. In the last two decades, several cross-sectional structural magnetic brain imaging (sMRI) studies have compared volumes of specific brain regions between children with ADHD who were medication-naïve, children with ADHD who had been using methylphenidate, and typically developing (TD) children. An early study examined the grey matter (GM) volumes of the caudate nucleus, the cerebellum and the four major lobes using automated tracing procedures, and reported no significant effect of medication (Castellanos et al., 2002). Subsequent cross-sectional studies examined GM volumes of the caudate nucleus (Semrud-Clikeman et al., 2006, Sobel et al., 2010), the anterior cingulate cortex (ACC; Semrud-Clikeman et al., 2006), the cerebellum (Bledsoe et al., 2009), the thalamus (Ivanov et al., 2010), the putamen, the globus pallidus (Sobel et al., 2010) and the nucleus accumbens (Hoekzema et al., 2014). These studies generally reported normalizing effects of medication on GM volume deficits found in ADHD patients (Schweren et al., 2013), except for Hoekzema and colleagues who found substantial volumetric declines in the bilateral nucleus accumbens in medicated patients. However, analysis of longitudinal data suggested that this initial reduction was followed by a recovery of volume under maintained treatment (Hoekzema et al., 2014). Finally, in a recent longitudinal study, Shaw and colleagues examined the volumes of the caudate nucleus, the putamen and the globus pallidus in a large cohort of children with ADHD and TD children (Shaw et al., 2014). They reported no significant association between history of psychostimulant treatment and developmental trajectories.
Grey matter volumes are only one among different possible indicators of brain structure, along with cortical thickness, surface area and cortical gyrification (Winkler et al., 2009). One longitudinal study examined change in cortical thickness in medicated and never-medicated adolescents with ADHD when compared to TD adolescents, and found an excessive rate of cortical thinning in the right motor strip, the left middle/inferior frontal gyrus and the right parieto-occipital region in the never-medicated group only (Shaw et al., 2009). In a larger sample encompassing the participants recruited in this study, no significant differences in cortical thickness were detected at study entrance when comparing medicated and never-medicated children with ADHD, except in a small region in the anterior temporal cortex (Shaw et al., 2006).
Taken together, these findings support the view that administration of methylphenidate is not detrimental to the development of the human brain (Schweren et al., 2013). However, this literature is characterized by several limitations. First, despite a preferential activation of prefrontal cortex catecholamine neurotransmission by methylphenidate (Spencer et al., 2012), none of these studies examined medication effects on GM volumes in frontal regions (only the volume of the entire frontal lobe was once examined (Castellanos et al., 2002)). Second, effect of methylphenidate on grey matter volume in the nucleus accumbens was only examined in children medicated for less than 13 months (Hoekzema et al., 2014). Longer treatment durations may be required to identify noticeable structural changes in this structure. Third, the above sMRI studies relied on lobar manual or automated region of interest (ROI) measurement methods. ROI methods yield a single value for the volume of the region examined, obtained after averaging signal over the ROI. This signal averaging can cause a dilution of the measure of the volume difference, especially when this difference is only present in a limited part of the ROI (Voormolen et al., 2010). Voxel-based morphometry (VBM) is an alternative automated approach that employs a general linear model framework to conduct mass voxel wise statistical tests throughout the brain, allowing it to be a particularly sensitive technique for detecting focal differences in GM volumes within all regions of the brain (Ashburner and Friston, 2000). VBM has been shown to outperform ROI methods when detecting focal differences in morphology (Voormolen et al., 2010). However, theoretically, ROI methods remain superior when between-group differences are distributed uniformly over a small ROI, since the ROI analysis at this spatial scale benefits from substantial signal averaging (Voormolen et al., 2010). Accordingly, both methods can provide different types of information and are thus considered as complementary (Giuliani et al., 2005).
While no VBM study to date has examined the effects of methylphenidate on GM volumes in ADHD patients, two meta-analyses investigated potential effects of medication indirectly through meta-regression analyses, based on variations in percentage of patients receiving medication included in published VBM studies of ADHD. These studies reported a ‘normalizing’ effect of psychostimulants in the caudate nucleus, the ACC and the amygdala (Nakao et al., 2011, Frodl and Skokauskas, 2012). Such meta-regression analyses are powerful tools, but due to their indirect nature, they are also characterized by intrinsic limitations. Here, for example, medication was assessed through the percentage of patients receiving medication in past studies, which does not take into account between-study variations in treatment duration, daily dosage, comorbidity rates, gender ratios, pre-processing options and/or statistical thresholding. Such variations in the underlying studies introduce confounds to the meta-analysis. Considering the widespread prescription of methylphenidate, more direct approaches should be used to gather convergent information and improve our understanding of methylphenidate’s effects on brain development.
The present sMRI study is the first to use VBM to directly investigate GM volume differences in never-medicated children with ADHD, medicated children with ADHD and TD children. VBM was used as a first step in an exploratory manner to investigate GM volume differences between groups throughout the whole brain. As a second step, to improve detection of potential diffuse differences in relevant structures of the brain, measurements of three selected ROIs were obtained through automated tracing. Only structures of moderate size were selected (Voormolen et al., 2010), based on previous findings (Nakao et al., 2011, Frodl and Skokauskas, 2012, Hoekzema et al., 2014) and pharmacological plausibility (Volkow, 2012). These ROIs were: the caudate nucleus, the amygdala and the nucleus accumbens. We hypothesized that medicated children with ADHD would exhibit less pronounced grey matter volume reductions than medication-naïve children with ADHD in the caudate nucleus and in the amygdala compared to TD children, that no significant between-group difference would be found in the nucleus accumbens, and that there would be a significant correlation between treatment duration and GM volumes in the caudate nucleus and in the amygdala in medicated children with ADHD.
Section snippets
Participants
Eighty-four children (44 males) aged 7.3−12.9 participated in this study. Children with ADHD were recruited from the outpatient clinic in Erasme Hospital, Université libre de Bruxelles, Belgium. TD participants were recruited from local schools in Brussels or via personal request to professionals working at Erasme Hospital. Out of the 84 participants, six children with ADHD were excluded from the analysis following the discovery of anatomical brain abnormalities, and one child with ADHD was
Demographic characteristics
Analyses revealed no significant differences between the medicated ADHD, never-medicated ADHD and TD groups for gender, IQ, age or SES (Table 1). Medicated and never-medicated children with ADHD did not differ on ratings of ADHD symptoms’ severity (Table 1).
Voxel-based morphometry
ADHD never-medicated versus TD children: when compared with TD children, never-medicated children with ADHD displayed decreased GM volume in the insula and in the middle temporal lobe (Table 2).
ADHD medicated versus TD children: Medicated
Discussion
This sMRI study is the first to use VBM to directly compare GM volumes in medicated and never-medicated children with ADHD, and TD children. When compared to both medicated children with ADHD and TD children, never-medicated children with ADHD presented decreased GM volume in the insula and in the middle temporal gyrus. When compared to TD children, medicated children with ADHD exhibited decreased GM volume in the middle frontal gyrus and in the precentral gyrus. Differences in GM volume
Role of the funding source
This work was supported by a grant from the Belgian National Fund for Scientific Research (FNRS 3.4.516.08.F). Stéphane A. De Brito was supported by a research fellowship from the Swiss National Science Foundation (SNSF PA00P1_139586).The funding sources had no involvement in study design, in the collection, analysis or interpretation of data, in the writing of the report and in the decision to submit the paper for publication.
Contributors
Thomas Villemonteix: analysis, interpretation of data, drafted the first version of the article; Stéphane A. De Brito: Analysis, interpretation of data; Hichem Slama: acquired data, interpretation of data; Martin Kavec: technical conception, acquired data; Danielle Balériaux: technical conception, acquired data; Thierry Metens: technical conception, acquired data. Simon Baijot: acquired data, interpretation of data; Alison Mary: acquired data, interpretation of data; Philippe Peigneux:
Conflict of interest
Isabelle Massat received compensations as a speaker in local educational meeting devoted to ADHD in 2013 and financial aid for registration for the ADHD World Congress in 2014 from Shire Company. Thomas Villemonteix received financial aid for travel expenses during the ADHD World Congress in 2014 from Shire Company. All other authors report no conflict of interest.
Acknowledgments
This work was supported by a grant from the Belgian National Fund for Scientific Research (FNRS 3.4.516.08.F). Stéphane A. De Brito was supported by a research fellowship from the Swiss National Science Foundation (SNSF PA00P1_139586). The authors thank all children and their families for their participation, and Mustapha Nouali and his team for kind help and assistance in MRI data collection.
References (51)
- et al.
Voxel-based morphometry − the methods
NeuroImage
(2000) - et al.
Attention-deficit hyperactivity disorder
Lancet
(2005) - et al.
A magnetic resonance imaging study of the cerebellar vermis in chronically treated and treatment-naïve children with attention-deficit/hyperactivity disorder combined type
Biol. Psychiatry
(2009) - et al.
Methylphenidate treatment during pre- and periadolescence alters behavioral responses to emotional stimuli at adulthood
Biol. Psychiatry
(2003) - et al.
Sensitized nucleus accumbens dopamine terminal responses to methylphenidate and dopamine transporter releasers after intermittent-access self-administration
Neuropharmacology
(2014) - et al.
ADHD, stimulant treatment in childhood and subsequent substance abuse in adulthood − a naturalistic long-term follow-up study
Addict. Behav.
(2014) - et al.
Voxel-based morphometry versus region of interest: a comparison of two methods for analyzing gray matter differences in schizophrenia
Schizophr. Res.
(2005) - et al.
Nonmedical use of prescription stimulants in the United States
Drug Alcohol Depend.
(2006) - et al.
Stimulant actions in rodents: implications for attention-deficit/hyperactivity disorder treatment and potential substance abuse
Biol. Psychiatry
(2005) - et al.
A voxel-based morphometry study of young occasional users of amphetamine-type stimulants and cocaine
Drug Alcohol Depend.
(2014)