Inter-hemispheric Asymmetry Patterns in the ADHD Brain: A Neuroimaging Replication Study

Abstract

Attention-deficit/ hyperactivity disorder (ADHD) is the most common neurodevelopment disorder in children, and many genetic markers have been linked to the behavioral phenotypes of this highly heritable disease. The neuroimaging correlates are similarly complex, with multiple combinations of structural and functional alterations associated with the disease presentations of hyperactivity and inattentiveness. Thus far, neuroimaging studies have provided mixed results in ADHD patients, particularly with respect to the laterality of findings. It is possible that hemispheric asymmetry differences may help reconcile the variability of these findings. We recently reported that inter-hemispheric asymmetry differences were more sensitive descriptors for identifying differences between ADHD and typically developing (TD) brains (n=849) across volumetric, morphometric, and white matter neuroimaging metrics. Here, we examined the replicability of these findings across a new data set (n=202) of TD and ADHD subjects at the time of diagnosis (medication naive) and after a six week course of either stimulant drugs, non-stimulant medications, or combination therapy. Our findings replicated our earlier work across a number of volumetric and white matter measures confirming that asymmetry is more robust at detecting differences between TD and ADHD brains. However, the effects of medication failed to produce significant alterations across either lateralized or symmetry measures. We suggest that the delay in brain volume maturation observed in ADHD youths may be hemisphere dependent. Future work may investigate the extent to which these inter-hemispheric asymmetry differences are causal or compensatory in nature.

1. INTRODUCTION

Human brains are not symmetric. Normal variation and functional specialization produce asymmetries of structure, function, and behavior that evolve throughout life, and are thought to originate from developmental, hereditary, experiential and pathological factors (Toga and Thompson, 2003). However, certain pathologies can also modulate, exacerbate, or arise from brain asymmetries (Thompson et al., 1998), (Crow, 1997).

Attention itself induces functional asymmetries in healthy individuals. For example, PET and EEG studies suggest that attention directed to global aspects of visual processing are more right lateralized, but local processing is more left hemisphere dominant (Fink et al., 1996), (Yamaguchi et al., 2000). A number of studies have investigated the link between structural brain asymmetries and attention-deficit/hyperactivity disorder (ADHD) (Rubia et al., 2000), (Hale et al., 2015), (Shaw et al., 2009a), and its characteristic behavioral systems such as impulsivity (Gordon, 2016). Early lesion studies suggested a role for unilateral right hemisphere dysfunction in ADHD (Heilman and Van Den Abell, 1980), (Stefanatos and Wasserstein, 2001). However, the patterns of brain asymmetry and its relationship to one’s ability to sustain attention appears to be more complex. For example, decreased asymmetry in ADHD subjects was observed, using homologous points on the cortex for comparison (Shaw et al., 2009a). However this study population consisted mostly of subjects with a history of psychostimulant medication (Shaw et al., 2009a), and it is possible that atypical structural asymmetry may be resolved or altered over the course of pharamacotherapy.

Recently, we examined altered patterns of inter-hemispheric asymmetry in a large cohort of ADHD youths and typically developing (TD) controls. Patterns of altered asymmetry differences were found in white matter mean diffusivity (MD) in cingulum, inferior and superior longitudinal fasciculus, and cortico-spinal tracts (p<0.001) with the effect of stimulant treatment tending to reduce these patterns of asymmetry differences. Gray matter volumes were more asymmetric in medication free ADHD individuals compared to TD in twelve cortical regions and two non-cortical volumes studied (p<0.05). Overall, asymmetry group differences were more significant than lateralized comparisons between ADHD and TD subjects across morphometric, volumetric, and DTI comparisons. In the present work, we sought to test the replicability of these findings using comparable methods on a new data neuroimaging data set.

2. METHODS

2.1. Subjects and Study Design

Parents and participants provided written informed permission and assent, respectively, prior to commencing the study which was approved by the UCLA Institutional Review Board. The study was an 8 week double-blind, randomized control trial the following treatment conditions: placebo only, d-methylphenidate (MPH) extended release (5–20 mg/day) + placebo, guanfacine (GUAN) + placebo, and combination (GUAN+MPH). Subjects aged 7 to 14, consisted of children who met DSM-IV diagnostic criteria for ADHD and age-matched TD youths (n = 202). The subjects were recruited as part of the UCLA Translational Research to Enhance Cognitive Control Center (TRECC). An ADHD diagnosis was determined via a semi-structured diagnostic interview (Kiddie-Schedule for Affective Disorders and Schizophrenia–Present and Lifetime version (K-SADS-PL; Kaufman et al., 1997) and a Clinical Global Impression-Severity (CGI-S) score ≥ 4. Exclusion criteria included: autistic disorder, chronic tic disorder, psychosis, bipolar disorder, structural heart defects, blood pressure >95th or below <5th percentile for age and body mass index, or a medical condition which precluded the use of stimulants or alpha agonists. For additional details, see McCracken et al. 2016. Neuropsychological data was acquired at the baseline visit which included obtaining scores for the Strengths and Weaknesses of ADHD Symptoms and Normal Behavior Rating (SWAN). Subjects also completed stop signal reaction time (SSRT), short working memory (SWM) and go-no-go behavioral tasks.

2.1. Neuroimaging & Preprocessing

All neuroimaging protocols took place at baseline shortly after diagnosis (Time 1), and after an eight weeks course of placebo, MPH+placebo, GUAN+placebo, or combination MPH/GUAN (Time 2). Data were collected on a 3 Tesla Siemens Allegra Scanner at the UCLA Ronald Reagan Medical Center. High resolution T1 weighted MPRAGE anatomical MRI scans were processed with FreeSurfer’s recon-all processing pipeline for whole brain segmentation and automated parcellation (Fischl and Dale, 2000). This generates segmentations for white matter, gray matter, and subcortical volumes. A mesh model of the cortical surface was then divided into 34 cortical brain regions, and for 14 subcortical and non-cortical regions per hemisphere. Total volume (mm³) for each region was used for group comparisons.

The diffusion tensor imaging (DTI) protocol consisted of axial multislice single-shot echo-planar imaging (EPI) sequence with the following parameters: repetition time [TR]/echo time [TE] = 7300/95 msec, flip angle = 90°, field of view (FOV) = 240mm², slice thickness = 2.5mm, no slice gap, number of slices = 55, k-space matrix = 96×96, image matrix = 192×192, voxel size=2.5mm³, with 30 diffusion gradient directions taken from Jones et al. (2002), b=800s/mm², and 5 non-diffusion-weighted volumes with b=0. All imaging data were analyzed using the FMRIB Software Library (FSL). Preprocessing included eddy current distortion correction and brain extraction. Diffusion tensors were estimated at each voxel using dtifit to calculate fractional anisotropy (FA) maps, where were registered to each subjects T1-weighted structural scan, followed by group registration the the Johns Hopkins diffusion tensor imaging (DTI) white matter (WM) atlas http://cmrm.med.jhmi.edu/; Wakana et al., 2004).

2.2. Statistical Analysis

All statistical analyses were consistent with our previously described methods (Douglas et al. 2018). In brief, for each MRI volumetric and DTI measure, we first computed within-hemisphere statistics for group comparisons (e.g., TD right hemisphere caudate volume versus ADHD right hemisphere caudate volume). To assess brain inter-hemispheric symmetry differences between ADHD and TD populations for each structural parameter, we calculated the asymmetry index (AI) as the difference between the left and right hemispheres divided by the mean. We assessed the absolute value of AI for each subject and imaging measure. Crucially, this method allows individual brains to serve as their own anatomic control, diminishing the influence of other variables such as gender and maturation between scanning sessions.

We used a general linear mixed effects model to study the relationship between within-hemisphere and AI differences in ADHD (Bates et al., 2012). Fixed effects included: site of data collection, age, sex, diagnosis, and their interactions, to account for known differences in maturation rates in the ADHD population (Larisch et al., 2006),(Castellanos and Proal, 2009). P-values were obtained by likelihood ratio tests of the full model against the null-model that disregarded the influence of diagnosis. All code for this analysis was implemented in R (R Core Team, 2012), using the linear mixed-effects lme model tool (Bates et al., 2012). All reported p-values were adjusted to correct for multiple comparisons (Benjamini and Hochberg, 1995) within our code. Participant ID was included as a random effect within the model to account for multiple comparisons.

3. RESULTS

3.1. Study Cohort

For Time 1, the mean age for ADHD and TD subjects were 9.8 ± 2.0 and 10.5 ± 2.2, respectively (mean ± standard deviation; see Table 1). Approximately 27% of ADHD and 61% of TD subjects were females. Due to attrition of subjects over the course of the experiment, the demographics were slightly altered in the subset who returned for the followup imaging session. At Time 2, the mean age for subjects in the TD group was 11 ± 2.4 (21% female) and the mean age for subjects in the ADHD group was 9.8 ± 2.1 (29% female). SWAN scores were only assessed at baseline. Mean SWAN scores were 62.9 ± 22.7 and 118.2 ± 27.8 for ADHD and TD, respectively. A summary of the results from the behavioral task is shown in Table 1.

Table 1:

Summary of behavioral results for ADHD and TD groups at baseline and after treatment.

	SWM Accuracy	SWM Response Rate	Stop-Signal SSRT	Stop Signal % Go Trials Correct	Stop Signal % Inhibit Trials Correct
Baseline
ADHD	68.6 ± 11.5	88.4 ± 17.1	308.7 ± 140.7	90 ± 11.2	44.4 ± 16.3
TD	73.6 ± 11.4	92.7 ± 13.2	264.3 ± 116.3	95.8 ± 6.6	49.3 ± 13.7
After Treatment
Methylphenidate	67.6 ± 13.0	92.4 ± 10.0	243.6 ± 113	93.1 ± 11.8	47 ± 13.7
Guanfacine	67.1 ± 13.8	86.5 ± 19.4	204.8 ± 166.5	88 ± 19.5	50.3 ± 15.2
Combined	70.1 ± 9.5	93.3 ± 9.5	204.7 ± 135.2	89.1 ± 19.2	53.3 ± 13.7
All ADHD	68.3 ± 12.7	90.8 ± 13.8	219.5 ± 139	90.2 ± 17	50 ± 14.4
TD	78.1 ± 9.0	90.2 ± 19.3	244.4 ± 120.2	92.1 ± 18	48.4 ± 16

3.2. Volumetric Neuroimaging Findings

After adjustment for age, sex, diagnosis, and their interactions, we observed statistically significant cortical gray matter volume differences for group comparisons between TD and ADHD-Free (medication naïve) subjects in lateralized comparisons (Table 2) including: rostral middle frontal, inferior parietal, inferior temporal (p<0.05), and superior frontal (p<0.01) in the right hemisphere, and lateral orbitofrontal, precentral (p<0.05), rostral middle frontal, superior parietal (p<0.01), and the superior frontal (p<0.001) in the left hemisphere. For Asymmetry Index (AI), the following brain regions were significantly different for group comparisons in the entorhinal, fusiform (p<0.001), and the rostral anterior cingulate (p<0.05). In contrast, absolute asymmetry index was significantly different in nine cortical gray matter volumes with focus in the frontal regions. In all absolute AI comparisons, the mean absolute magnitude of AI was greater in the ADHD-Free group compared to TD subjects (see Table 2).

Table 2:

P-values are shown comparing Typically Developing (TD) and ADHD-Free participants in Time 1 for Cortical Gray Matter volumetric results. These comparisons were done for the following: TD versus ADHD-Free Left Hemisphere, TD versus ADHD-Free Right Hemisphere, TD versus ADHD-Free Asymmetry Index (AI), and TD versus ADHD-Free absolute Asymmetry Index (Abs(AI)). The p-values shown were corrected for multiple comparisons, with significant comparisons in bold. A + sign indicates the measure to be larger in the ADHD group. In all lateralized (left and right) bold cases, the ADHD group had smaller (estimates) volumes.

Cortical Gray Matter Volumetric Results Typically Developing vs. ADHD-Free
Region	Gray Matter Volume	Right	Left	AI	Abs (AI)
Cingulate	Caudal Middle Frontal	0.612	0.598	0.397	0.049 +
	Isthmus cingulate	0.751	0.438	0.224	0.234
	Posterior Cingulate	0.961	0.969	0.662	0.309
	Rostral Anterior Cingulate	0.957	0.471	0.0309	0.206
Frontal	Lateral Orbito-frontal	0.119	0.048	0.290	0.132
	Medial orbitofrontal	0.393	0.485	0.522	0.003 +
	Paracentral	0.648	0.749	0.338	0.202
	Pars Opercularis	0.871	0.483	0.065	0.255
	Pars orbitalis	0.784	0.635	0.286	0.014 +
	Pars Triangularis	0.948	0.603	0.088	0.060
	Rostral Middle Frontal	0.036	0.005	0.161	0.023 +
	Superior Frontal	0.003	0.000	0.109	0.185
	Frontal Pole	0.738	0.897	0.053	0.012 +
Occipital	Cuneus	0.601	0.915	0.213	0.068
	Lateral Occipital	0.794	0.617	0.058	0.616
	Lingual	0.501	0.810	0.246	0.186
	Pericalcarine	0.437	0.588	0.419	0.094
Parietal	Inferior Parietal	0.023	0.153	0.362	0.149
	Postcentral	0.812	0.619	0.206	0.030 +
	Precentral	0.181	0.011	0.102	0.115
	Precuneus	0.103	0.179	0.329	0.139
	Superior Parietal	0.102	0.003	0.095	0.051
	Supramarginal	0.860	0.520	0.120	0.041 +
Temporal	Entorhinal	0.601	0.758	0.006	0.235
	Fusiform	0.644	0.164	0.006	0.440
	Inferior temporal	0.022	0.054	0.626	0.201
	Middle Temporal	0.159	0.206	0.278	0.072
	Superior Temporal	0.232	0.378	0.434	0.007 +
	Temporal Pole	0.861	0.973	0.498	0.016 +
	Transverse Temporal	0.618	0.986	0.057	0.367
	Parahippocampal	0.987	0.873	0.190	0.696
Insular Cortex	Insula	0.774	0.599	0.254	0.159

In a supplementary analysis, we examined presentation type among the ADHD-Free (Time 1) subjects. For AI, statistically significant regions included the entorhinal, fusiform, lateral occipital, parahippocampal and pars opercularis (p<0.05) in the inattentive type, and entorhinal and fusiform (p<0.05) in the combined type. There was an increase in absolute AI for nine regions in the inattentive type with focus in the frontal and temporal regions (pars orbitalis (p<0.001), medial orbitofrontal, rostral middle frontal (p<0.01), pericalcarine, superior parietal, supramarginal, middle temporal, superior temporal, and temporal pole (p<0.05)) and six regions in the combined type (caudal middle frontal, medial orbitofrontal, frontal pole, cuneus, superior temporal, and temporal pole (p<0.05)).

Additionally, we examined results for cortical gray matter volume comparisons between ADHD-Free (Time 1) versus ADHD-Rx (Time 2). The only significant differences were in absolute asymmetry index (AI) across the following regions (Table 3): medial orbitofrontal, pars orbitalis, rostral middle frontal, lateral occipital, lingual, pericalcarine, precentral, superior parietal, middle temporal, superior temporal, parahippocampal (p < 0.05), and supramarginal (p < 0.01). All regions had a decrease in absolute AI after treatment, indicating that medication overall tends to decrease asymmetry.

Table 3.

P-values are shown comparing ADHD participants before treatment and after treatment for Cortical Gray Matter volumetric results. These comparisons were done for the following: ADHD-Free versus ADHD-Rx Treatment Left Hemisphere, ADHD-Free versus ADHD-Rx Treatment Right Hemisphere, ADHD-Free versus ADHD-Rx Treatment Asymmetry Index (AI), and ADHD-Free versus ADHD-Rx Treatment Absolute Asymmetry Index (Abs(AI)). The p-values shown were corrected for multiple comparisons, with significant comparisons in bold. All regions displayed a decrease in Absolute AI for ADHD-Rx.

Before versus After Treatment: Cortical Gray Matter Volumetric Results for ADHD-Free vs. ADHD-Rx
Region	Gray Matter Volume	Right	Left	AI	Abs (AI)
Cingulate	Caudal Middle Frontal	0.582	0.714	0.622	0.226
	Isthmus cingulate	0.559	0.597	0.837	0.083
	Posterior Cingulate	0.861	0.884	0.763	0.081
	Rostral Anterior Cingulate	0.733	0.948	0.143	0.213
Frontal	Lateral Orbito-frontal	0.830	0.499	0.454	0.066
	Medial orbitofrontal	0.655	0.918	0.313	0.018
	Paracentral	0.235	0.376	0.412	0.069
	Pars Opercularis	0.765	0.820	0.303	0.332
	Pars orbitalis	0.993	0.983	0.930	0.040
	Pars Triangularis	0.968	0.911	0.757	0.186
	Rostral Middle Frontal	0.725	0.629	0.563	0.032
	Superior Frontal	0.825	0.983	0.449	0.168
	Frontal Pole	0.602	0.653	0.949	0.744
Occipital	Cuneus	0.354	0.534	0.463	0.092
	Lateral Occipital	0.993	0.855	0.646	0.034
	Lingual	0.278	0.590	0.272	0.015
	Pericalcarine	0.347	0.418	0.816	0.037
Parietal	Inferior Parietal	0.835	0.903	0.698	0.177
	Postcentral	0.926	0.934	0.644	0.069
	Precentral	0.507	0.918	0.349	0.037
	Precuneus	0.829	0.908	0.758	0.091
	Superior Parietal	0.614	0.494	0.609	0.046
	Supramarginal	0.717	0.670	0.957	0.007
Temporal	Entorhinal	0.645	0.798	0.780	0.126
	Fusiform	0.580	0.741	0.472	0.124
	Inferior temporal	0.361	0.673	0.497	0.068
	Middle Temporal	0.621	0.625	0.130	0.041
	Superior Temporal	0.876	0.196	0.241	0.011
	Temporal Pole	0.786	0.951	0.341	0.257
	Transverse Temporal	0.600	0.802	0.273	0.273
	Parahippocampal	0.737	0.756	0.767	0.025
Insular Cortex	Insula	0.909	0.844	0.717	0.162

We performed the same comparisons with subcortical regions and found fewer statistically significant results for lateralized and asymmetry measures between TD and ADHD-Free (Time 1). For the right hemisphere, the cerebellum was lower in volume (p<0.01, Table 6) for ADHD-Free subjects compared to TD subjects, see Table 4.

Table 4.

P-values are shown comparing Typically Developing (TD) and ADHD-Free participants in Time 1 for Subcortical volumetric results. These comparisons were done for the following: TD versus ADHD-Free Left Hemisphere, TD versus ADHD-Free Right Hemisphere, TD versus ADHD-Free Asymmetry Index (AI), and TD versus ADHD-Free absolute Asymmetry Index (Abs(AI)). The p-values shown were corrected for multiple comparisons, with significant comparisons in bold. In the bold case (Right hemisphere), the ADHD group had smaller volume in the cerebellum.

P-values for Subcortical Volumes Typically Developing vs. ADHD-Free
Subcortical Region	Right	Left	AI	Abs (AI)
Amygdala	0.981	0.958	0.963	0.893
Caudate	0.974	0.993	0.545	0.801
Cerebellum-Cortex	0.0088	0.058	0.752	0.960
Choroid plexus	0.959	0.945	0.594	0.691
Hippocampus	0.810	0.822	0.922	0.712
Inferior Lateral Ventricle	0.806	0.908	0.073	0.075
Lateral Ventricle	0.472	0.449	0.909	0.613
Pallidum	0.978	0.968	0.212	0.442
Putamen	0.870	0.791	0.224	0.350
Thalamus	0.805	0.748	0.925	0.946
Ventral DC	0.912	0.961	0.569	0.487
Vessel	0.964	0.938	0.348	0.087

3.2. White Matter Differences in the ADHD Brain

After adjustment for age and sex, significant differences were found for TD and ADHD-Free in the left hemisphere and all AI measures. Left hemisphere differences were significant in the superior fronto-occipital fasciculus for radial (perpendicular) diffusivity. Two tracts were found to be significant for AI, whereas three tracts were found to be significant in absolute AI. Fractional anisotropy (FA) differences were found in the cingulum. Mean diffusivity asymmetry differences were found in the uncinate fasciculus and sagittal stratum. Symmetry differences in radial (perpendicular) diffusivity were found for the superior fronto-occipital fasciculus. The corticospinal tract was significantly different across all AI and DTI measures (Figure 1).

Figure 1.

Illustrative figure of white matter tracts which were significantly different between the ADHD-Free and TD group. Sagittal and axial views are shown for the tracts: (a) uncinate fasciculus, (b) cingulum, (c) corticospinal tract, (d) sagittal stratum (includes inferior fronto-occipital fasciculus and inferior longitudinal fasciculus; Mori et al. 2008), and (e) superior fronto-occipital fasciculus. All images created were created using DSI studio data for visualization purposes.

4. DISCUSSION

Attention-deficit/hyperactivity disorder (ADHD) is a child-onset neurodevelopmental condition characterized by inattentiveness and/or hyperactivity-impulsivity. Meta-analyses report a 5.29% worldwide prevalence of ADHD for youths under the age of 18 (Polancyzk, et al., 2007), and these symptoms persist in about two-thirds of adults (Barkley et al., 2002, Faraone et al., 2006). There is a tremendous need to identify quantitative neuroimaging biomarkers of ADHD, which can inform intervention studies and therapies targeted to baseline phenotypes. Laterality (right versus left hemisphere) is often used as a metric for examining structural and functional MRI alterations of anatomy and physiology, but recently, these metrics have been called into question due to mixed results (i.e., increase and decrease) in laterality as shown in structural and functional brain studies (Emond et al., 2009; Nakao et al., 2011; Ellison-Wright et al., 2008; Paclt et al., 2016; Douglas et al., 2018; Hart et al., 2012).

Functional and structural brain asymmetries are a part of normal development. However, differences in these normal asymmetry patterns may also be pathologic. The symptoms of ADHD have long been associated with right hemisphere abnormalities as shown in behavioral and neuroimaging studies (Stefanatos & Wasserstein 2011; Heilman & Van Den Abell, 1980; Carter et al., 1995; Vance et al., 2007; Monden et al., 2015; Doi & Shinohara, 2017). However, more subtle patterns of atypical interhemispheric and lateralization have been revealed (Hale et al., 2009; Silk et al., 2016; Hasler et al., 2017; Segal et al., 2017). Overall, our work demonstrates that these asymmetries may be more useful for identifying differences in the ADHD brain. We suggest that future studies may examine the functional role of these asymmetry differences. Determining the extent to which these asymmetries evolve over development may be crucial, given considerable evidence that ADHD is associated with a delay in brain volume maturation (e.g., Hoogman et al. 2017). It is possible that this delay may be hemisphere specific. Furthermore, it is possible that multiple combinations of structural and functional brain abnormalities produce similar phenotypic presentations. Therefore, the delay in maturation may be variably lateralized within the ADHD population, yet result in similar issues with integrating sensory information across cortices. Further work may also reveal the if there is a correspondence between the persistence of altered AI and symptoms of ADHD carrying forward into adulthood. Given the complexity of genetic loci which are linked to hyperactivity and inattentiveness (McGough 2012), it may be difficult to determine whether these altered AI patterns play a causal or compensatory role in generating the ADHD behavioral manifestations.