Smaller volumes of caudate nuclei in prepubertal children with ADHD: Impact of age

Abstract

Objective

Age-related abnormalities in caudate volumes have been reported to differ across the periods of childhood and puberty in children with ADHD. This study assessed caudate volumetric abnormalities across two narrow age clusters within the childhood period.

Method

Three-dimensional manual tracings of the head and body of the caudate nucleus and of the cerebrum were acquired from 26 medication-naïve boys with a diagnosis of ADHD (ages 5.9–10.8 years), and 24 age-matched normal controls.

Results

Boys with ADHD had smaller total caudate volumes relative to controls, F(1,48) = 4.29, p = 0.04. Adjustment of caudate volumes with respect to age demonstrated that this group difference was driven solely by participants in the 5.9–7.3 year range, F(1, 46) = 5.64, p = 0.022, with an effect size of d = 0.69. No Group effect was found in older participants, F(1, 46) = 0.82, p = 0.37.

Conclusions

These novel findings suggest a different pattern of caudate volume abnormalities across narrow age clusters prior to puberty in boys with ADHD. Anatomical differences in brain structures related to ADHD in prepubertal children should be evaluated with respect to the changing developmental trajectory of brain regions within this period of rapid brain growth.

Introduction

Attention Deficit Hyperactivity Disorder (ADHD) is a major public health problem, affecting 3–9% of children (Froehlich et al., 2007). It is conceptualized as a disorder caused by neurodevelopmental deficits or delays (Shaw et al., 2007), and its neuroanatomical basis has become increasingly better defined as a result of structural neuroimaging studies (Durston, 2003, Valera et al., 2007). ADHD-related motor and cognitive impairments have implicated frontal cortical regions and their corresponding subcortical fields in the basal ganglia (Emond et al., 2009). There are, however, inconsistencies in the hypothesized relations between the presumed pathophysiology of cortico-basal ganglia pathways (including the individual nuclei of the caudate, putamen and globus pallidus) and the clinical features of ADHD, probably because of the heterogeneous nature of the disorder, as well as methodological differences among studies.

The caudate nucleus is the brain structure that has been implicated most often in the pathophysiology of ADHD by neuroimaging, genetic and neuropharmacological studies (Silk et al., 2009). It receives a widespread glutamatergic input from the entire cortex, including the prefrontal cortex (Berendse et al., 1992, Haber et al., 1995, Kemp and Powell, 1970, Yeterian and Pandya, 1991). The caudate sends projections back to the prefrontal cortex via the paralaminar part of the mediodorsal nucleus and the ventroanterior nucleus of the thalamus (Nakano, 2000). The nigrostriatal pathway comprises ∼80% of all dopaminergic projections, resulting in the caudate having one of the highest concentrations of dopamine in the brain.

Basal ganglia-thalamocortical feedback loops have been proposed to modulate cognitive functions associated with the prefrontal cortex (Graybiel, 2001, Nakano, 2000), some of which may be dysregulated in ADHD. A detailed analysis of the functional anatomy of these projections may provide insight into the classical symptom clusters of ADHD and its subtypes. Thus, in healthy adults regional differences in striatal projections to cortex have been characterized using diffusion tensor imaging (DTI) (Lehericy et al., 2004). This study demonstrated that caudal parts of the striatum (posterior putamen) project predominantly to motor/premotor cortex, primary sensory cortex, the posterior part of the supplementary motor area (SMA) and the substantia nigra. Anterior striatal regions (head of caudate nucleus) project to association areas of the medial prefrontal cortex, dorsal and ventral prefrontal cortex, the frontal pole and the mesencephalon. Finally, the ventral striatum projects to limbic areas, including medial orbitofrontal cortex, ventromedial frontal pole, the uncus of the temporal lobe and the temporal pole.

Volumetric studies of the caudate in children with ADHD have generated conflicting results. Different studies have reported that the caudate nuclei of children with ADHD are larger than (Garret et al., 2008, Mataro et al., 1997), smaller than (Castellanos et al., 2003, Hynd et al., 1993, Mahone et al., 2011, Qiu et al., 2009, Semrud-Clikeman et al., 2000) or equal to (Hill et al., 2003) those of normal controls. In addition, the typical asymmetry of the caudate nucleus (right > left) has been reported to be either absent or reversed in children with ADHD (Tremols et al., 2008; Uhlikova et al., 2007).

To try to resolve these inconsistent findings, a meta-analysis pooled results across 21 volumetric studies including 565 ADHD and 583 control children (not including voxel-based morphometry) (Valera et al., 2007). The regions of interest (ROIs) that differed significantly between ADHD and control groups were the cerebellum (posterior inferior vermis), bilateral and right cerebral volumes, the splenium of the corpus callosum and the right caudate nucleus. Other structures in which differences did not reach the significance criteria established by the meta-analysis included the whole caudate nucleus (left and right combined), the head of the caudate, and the left and right globus pallidus. The mean age of children included in the meta-analysis was 11.0 years with a range of 9.1–14.6 years. Another meta-analysis (Frodl & Skokauskas, 2012) using stringent criteria failed to find any caudate volume reductions in ADHD children and adolescents from voxel-based volumetric studies, but confirmed smaller left and right caudate volumes in ADHD from data reported in manual segmentation studies.

There are a number of experimental results suggesting that the age range of the target population might be important. Longitudinal volumetric measures of the caudate before and during puberty indicate that it is among the few structures displaying a striking developmental variation across the periods of childhood and adolescence. Caudate volumes increase during childhood until about 10 years of age and display a slight asymmetry (right > left) in typically developing prepubertal children (Giedd et al., 1999a, Giedd et al., 1999b). An extensive longitudinal study of caudate volumes in ADHD (Castellanos et al., 2002) reported that these volumes were smaller in prepubertal ADHD boys and girls relative to typically developing controls, but that the group difference disappeared during adolescence. The authors hypothesized that the normal trajectory of greater prepubertal synaptogenesis in the control group followed by developmentally normative pruning during adolescence was not present in the ADHD group.

Assessment of structural differences in the brains of typically developing children must take into account the fact that normal brain growth may vary significantly across narrow age ranges, raising questions about the appropriateness of pooling data across broad age ranges in both typically developing children and those with ADHD. In fact, the annual increase in brain weight during normal development is 2.3%/year in the age range of 4–7 years, but 3–5%/year during the period of 8–12 years (Dekaban & Sadowsky, 1978).

Given these observations, differences in subcortical structures between ADHD and typically developing children should be assessed relative to the developmentally changing baseline of brain maturation rates in childhood. In this study, we used a 3-dimensional manual segmentation method to assess the volumes of the anterior region of the caudate nuclei (head and body of caudate) in prebubertal, medication-naïve boys with ADHD and in age-matched healthy controls. Data were evaluated both across the full age range studied (5.9–10.8 years) and within narrower age-defined subgroups in order to assess whether group differences were present and whether they were consistent or varied across this developmental range.