Neurochemical Correlates of Executive Function in Children with Attention-Deficit/Hyperactivity Disorder

Abstract

Objectives

Attention-Deficit/Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental disorder with no known biomarkers. The objectives of this study were 1) to investigate spectroscopic biomarkers in the right prefrontal cortex (R-PFC) and left striatum; 2) to evaluate Executive Function (EF) performance; and, 3) to examine the clinical relevance of glutamate in EF tasks.

Methods

A total of 21 children with ADHD (M = 10.41 years, SD = 1.41) and 15 controls without ADHD (M = 9.90 years, SD = 1.54 years) were enrolled. Short echo proton magnetic resonance spectroscopy (1H-MRS; TE = 30ms) was used to study the changes in the R-PFC and left striatum. Both groups completed an EF assessment battery, including working memory, inhibition, cognitive flexibility and verbal fluency tasks.

Results

In the R-PFC, independent t-tests found decreased concentration of glutamate (p = 0.009), NAA (p = 0.029) and choline (p = 0.016) in ADHD participants compared to controls. No significant differences were seen in the left striatum. Multivariate analysis of variance did not indicate overall EF challenges in the ADHD sample (p < .05). Positive correlations with glutamate concentration and EF performance in the control group were observed, however, no such correlations were reported in the ADHD group.

Conclusions

The results indicated a subgroup of children with ADHD who presented with hypo-glutamatergic signalling in the R-PFC. Additionally, findings suggested a decoupling effect of glutamate in EF related tasks in children with ADHD compared to controls. Thus, glutamate concentration may be a possible ADHD biomarker and novel treatments target.

Attention-Deficit/Hyperactivity Disorder (ADHD) is a common neurodevelopmental disorder in childhood, with an estimated prevalence of 5% (Brault & Lacourse, 2012). Traditionally, symptoms of ADHD are characterized by developmentally inappropriate levels of inattention or impulsivity and hyperactivity (APA, 2013). While it is common for most young children to be active, energetic, and impulsive, children with ADHD often have significant difficulties in controlling their impulses and struggle to concentrate on monotonous tasks (APA, 2013). These impulsive and inattentive behaviours can affect their academic performance (Wolraich et al., 2005), and social functioning (Hoza, 2007).

Currently, there are no known biomarkers for ADHD. However, neuroimaging research studies have identified some plausible biomarkers. For example, studies have previously implicated the frontal-striatal pathway (FSP) to be associated with ADHD (Bush, 2010; Kelly, Margulies & Castellanos, 2007). Structural Magnetic Resonance Imaging (MRI) studies have also found delayed brain maturation in children with ADHD specifically in the FSP (Shaw et al., 2007; Yang, Carrey, Bernier & MacMaster, 2015).

Interestingly, the regions implicated in the FSP are also regions essential for executive functions (EF). For example, the dorsolateral prefrontal cortex (DLPFC) has been found to be involved with working memory and cognitive flexibility tasks (Forbes et al., 2014). Other areas of the prefrontal cortex (PFC) including ventrolateral and orbital PFC and basal ganglia are involved in emotional processing, acquisition, and reversal of stimulus-reward associations (Cortese et al., 2012; Rubia, 2018). Furthermore, studies have shown that the regions in the subcortical areas such as the caudate and putamen have connections with PFC and are involved in task switching and planning (Hanakawa, Goldfine & Hallett, 2017; Rubia 2018).

Along with neuroimaging findings, research suggests that children with ADHD demonstrate significant EF deficits on a range of performance-based measures (e.g., Willcutt, Doyle, Nigg, Faraone & Pennington, 2005) and parental behaviour rating scales (Toplak, Bucciarelli, Jain, & Tannock, 2009). It is important to mention that EF is an umbrella term used to refer to a complex range of cognitive abilities including goal-directed planning, impulse control, cognitive flexibility, and self-monitoring (Barkley, 2014; Diamond, 2013). As such, inconsistencies regarding EF deficits are commonly reported in the literature (e.g., the seminal work of Willcutt and colleagues (2005) indicated only moderate effect sizes when analyzing 82 studies evaluating EF differences in children with ADHD). These inconsistencies could be due to lack of consensus in the literature regarding the exact definition of EF (Waserman & Waserman, 2012), different sampling procedures, and the diagnostic criteria used to define the ADHD groups (Willcutt et al., 2005).

Generally, ADHD is thought to be a disorder that arises due to dysfunctions in the dopaminergic pathway (Volkow et al., 2009; Volkow et al., 2011). As such, most ADHD stimulant medications for the treatment of ADHD are targeted toward dopamine transporters (Faraone, 2018). However, there has been an increased interest in investigating the link between ADHD and dysfunctions in the glutamatergic pathway (Naaijen et al., 2018; Spencer, Uchida, Kenworthy, Keary & Biederman, 2014).

Glutamate is a major excitatory neurotransmitter in the brain. Studies using a non-invasive neuroimaging technique, Magnetic Resonance Spectroscopy (MRS), have shown dysfunction of glutamate in children with ADHD (Courvoisie, Hooper, Fine, Kwock & Castillo, 2004; MacMaster, Carrey, Sparkes & Kusumakar, 2003; Moore et al., 2006). Specifically, studies have found augmented levels of glutamate in children with ADHD in the right PFC and the left striatum (Courvoisie et al., 2004; MacMaster et al., 2003; Moore et al., 2006). Other studies have found no significant difference in glutamate levels in the left cerebellum (Soliva et al., 2010) and the middle frontal gyrus (Tafazoli et al., 2013).

MRS studies have also investigated changes in other neurometabolite concentrations in children with ADHD. For example, studies have found differences in the N-Acetyl Aspartate (NAA) concentration in children with ADHD compared to controls (Courvoisie et al., 2004; Sparkes, MacMaster & Carrey 2004). Similarly, studies have demonstrated differences in choline related compounds in children with ADHD when compared to controls (Hammerness Biederman, Petty, Henin & Moore, 2012; Sparkes et al., 2004; Tafazoli et al., 2013).

While numerous MRS studies have been conducted in children with ADHD, the current MRS literature in regard to neurochemical changes has been inconsistent (for review, see Altabella, Zoratto, Adriani & Canese, 2014). These differences could be due to methodological differences and advances in MRI technology. For instance, most of the previously published studies have been conducted with 1.5 Tesla MRI, with few recent publications using 3 Tesla MRI machines (Naaijen et al., 2017b; Tafazoli et al., 2013). The improved spectral resolutions can have a significant impact on the quantification of neurometabolites given the reduction of signal to noise ratio with stronger spectral fields (Stagg & Rothman, 2014). Previous studies also included smaller sample sizes (Carrey et al., 2003; MacMaster et al., 2003; Tafazoli et al., 2013). Lastly, earlier studies were conducted with medication naïve children (Carrey, MacMaster, Gaudet & Schmidt, 2007; MacMaster et al., 2003) while more recent studies have included children who had been on medication, which can result in long-term implications on the neurobiology of individuals.

Present Study

Given the changes in neuroimaging methodologies, it is important to replicate previous MRS findings with newer technology. Additionally, to our knowledge, there are no published studies that have investigated the relations between neuropsychological measures of EF and concentration of neurometabolites in the PFC in children with ADHD. As a result, the current study examined the effect of these neurometabolites on multiple EF tasks such as working memory, inhibition, cognitive flexibility and verbal fluency to answer the following research questions:

1. What are the differences in glutamate, NAA, and choline concentrations between children with ADHD and healthy controls in the right PFC and the left striatum?

2. Do children with ADHD demonstrate EF deficits on performance-based measures assessing inhibition, working memory, cognitive flexibility, and verbal fluency when compared to healthy controls?

3. Are there any relations between the neurometabolite glutamate with EF performance-based measures such as inhibition, working memory, cognitive flexibility, and verbal fluency?

Methods

Participants

A total of 21 children with ADHD (average age = 10.41 years, SD = 1.14, males 81%) and 15 healthy controls (average age 9.90 years, SD =1.54, males = 40%) took part in the study. There were no significant age differences between the two groups, p > .05. However, there was a significant gender difference, p < .05. Ethics was obtained from both the university and hospital research ethics board. Parental consents and assent from participants were obtained.

Inclusion Criteria

Participants in the ADHD group had to have a) a standard-of-care health professional to oversee their progress and a diagnosis of ADHD prior to study participation, b) behaviour ratings using the Behavior Assessment System for Children-Second Edition (BASC-2; Reynolds & Kamphaus, 2004) indicating that the child currently met DSM-5 ADHD criteria (APA, 2013), and c) a cognitive screener reporting no intellectual disability (scaled score > 4) on both the Vocabulary Multiple Choice (VCMC) and the Matrix Reasoning (MR) subtests from the Wechsler Intelligence Scale for Children-Fourth Edition Integrated (WISC-IV Integrated; Kaplan, Fein, Maerlander, Morris, & Kramer, 2004).

Neuropsychological measures

Four primary domains of EF were measured in both the ADHD and healthy controls group. Specifically, for measuring working memory, the Digit Span Backwards and Spatial Span Backwards subtests from the WISC-IV Integrated (Kaplan et al., 2004) were used. Commission errors on the Conners’ Continuous Performance Test (CPT II; Conners, 2004) and Delis Kaplan Executive Function System (DKEFS) Colour Word Interference test (Delis, Kaplan, & Kramer, 2001) were utilized to measure inhibition. Trails-Making Test, Part-B (TMT-Part B; Reitan & Wolfson, 1985) and Wisconsin Card Sorting Task (WCST; Heaton, Chelune, Talley, Kay & Curtis, 1993) were used to measure cognitive flexibility. Lastly, DKEFS-Verbal Fluency test (Delis et al., 2001) was utilized to measure verbal fluency.

Magnetic Resonance Spectroscopy (MRS) Acquisition

All MRS scans were conducted at a large Western Canadian children’s hospital using a 32-channel head coil, 3 Tesla GE 750w Scanner. Short echo proton magnetic resonance spectroscopy (1H-MRS) was used to extract spectra from the right PFC (Figure 1) and left striatum (Figure 1; 15mm x 15mm x 20 mm voxels; Carrey et al., 2007; MacMaster et al., 2003). T1-weighted anatomical images (0.8mm) were also collected. Additional spectral parameters were as follows: echo time = 30msec, repetition time = 2000 msec, 96 data points, acquisition = 138 voxels, total time 5 minutes. The MRS data were analyzed using the Linear Combination (LC) Model method (Provencher, 1993; 2001; 2016).

Figure 1.

Right-Prefrontal Cortex (R-PFC) and Left-Striatum (L-Striatum) spectroscopy voxel placement example shown on an anatomical brain scan.

Procedures

Participants had to complete screening measures to ensure their eligibility for the study. Parents were asked to answer ADHD symptomatology questionnaires to gather information about their child’s inattentive and hyperactive/impulsive symptoms. Following the screening measures, parents provided written consent. All eligible participants then completed an additional battery of neuropsychological measures. Lastly, eligible participants were scheduled for a neuroimaging session 1–2 days following the neuropsychological testing.

Data Analysis

The Statistical Package for the Social Sciences (SPSS) version 24.0 was used to conduct all the data analyses. The data were inspected for missing values and outliers prior to running any statistical analyses. The data were also evaluated for normality, linearity, homogeneity of variance, and homoscedasticity to meet the assumptions of parametric analysis. Independent sample t-tests were utilized to investigate the neurometabolite differences in ADHD and healthy controls in the right PFC and the left striatum. Given the significant gender difference, Analysis of Covariance (ANCOVA) was conducted to investigate impact of gender on the different neurometabolites. Multivariate analysis of variance (MANOVA) was performed to examine the differences in performance-based neuropsychological tests between children with ADHD and the healthy control group. Multivariate analysis of covariance (MANCOVA) was also conducted to control for gender differences. Lastly, Pearson correlations for parametric analyses and Spearman Correlations for non-parametric analyses were conducted to examine the relations between EF and glutamate concentrations. All the analyses were corrected for multiple comparisons using Benjamini-Hochbergs Principle (Benjamini & Hochberg, 1995).

Results

Group Differences in Screening Measures

Table 1 presents the sample characteristics regarding their cognitive and behavioural screening measures. Overall, the children in the healthy control group demonstrated higher performance on the verbal (VCMC; p < .05) and non-verbal (MR) subtests (p < .05) compared to the ADHD group. The healthy control group also showed significantly lower attention and hyperactivity concerns compared to the ADHD group as indicated by parent ratings.

Table 1.

Cognitive and Behavioural Group Characteristics at Baseline

Variable	ADHD (n = 21)		Healthy Controls (n =15)		t	Cohen’s d
	M	SD	M	SD
Age (years)	10.41	1.14	9.90	1.54	−1.14	0.38
Cognitive Tasks
WISC-IV-I VCMC (Verbal) (SS)	96.50	10.40	114.33	15.22	4.12**	1.37
WISC-IV-I MR (Non-Verbal) (SS)	100.25	12.41	108.67	11.57	2.04*	0.77
Behaviour Ratings
BASC-2 Attention Problems (t-score)	69.95	6.10	48.80	8.47	−8.73**	2.87
BASC-2 Hyperactivity (t-score)	73.62	14.53	46.13	7.22	−6.74**	2.40

Note. BASC-2 = Behavior Assessment System for Children, Second Edition; MR = Matrix Reasoning; SS = Standard Score (M = 100; SD = 15); T-Score (M = 50; SD = 10); VCMC = Vocabulary Multiple Choice; WISC-IV-I = Wechsler Intelligence Scale for Children, Fourth Edition, Integrated;

^**indicates p <.01,

^*indicates p <.05

Neurometabolite Differences

Table 2 summarizes the neurometabolite findings. Results indicated significantly lower concentrations of glutamate, NAA and choline in the ADHD group compared to healthy controls (p < .04). All of these differences were observed in the right PFC voxel. These group differences were still observed in the right PFC after controlling for gender differences. There were no significant mean group differences in metabolite concentration observed in the left striatum (p > .04). Post-hoc analyses found no significant group differences in other metabolites (creatine, glutamine, inositol, and glutathione) in the right PFC and left striatum.

Table 2.

Magnetic Resonance Spectroscopy Results in ADHD and Healthy Controls at Baseline

	ADHD		Healthy Controls		t	Cohen’s d
	Mean	SD	Mean	SD
Right-Prefrontal Cortex
NAA	9.36	0.45	9.73	0.51	2.89*	0.77
Glu	9.76	1.07	10.82	1.07	2.80*	0.99
Cho	1.82	0.12	1.93	0.15	2.54*	0.81
Cr	2.82	1.07	3.05	2.42	.37	0.12
Gln	4.32	1.21	4.04	0.29	−.66	0.38
Ino	4.81	0.47	5.02	0.56	1.21	0.41
GSH	2.20	1.10	1.87	0.38	−1.07	0.40
Left-Striatum
NAA	9.95	0.67	9.98	0.42	.15	0.05
Glu	10.16	1.37	10.26	1.39	.22	0.07
Cho	1.86	0.18	1.96	0.23	1.36	0.48
Cr	3.91	1.36	3.81	2.07	−.16	0.06
Gln	4.79	1.35	4.98	1.29	.38	0.14
Ino	4.00	0.59	3.62	0.62	−1.87	0.63
GSH	2.27	0.52	2.25	0.40	−.11	0.04

Note. PFC = Prefrontal Cortex, NAA = N-Acetyl Aspartate, Glu = Glutamate, Cho = Choline, Cr = Creatine, Gln = Glutamine, Ino = Inositol, GSH = Glutathione,

^*Benjamini-Hochberg corrected p value < .04.

EF Performance

Table 3 summarizes the EF findings. The MANOVA did not reveal a significant group effect, (p > .05, Wilk’s Lambda Λ = .30). However, when the EF measures were analyzed through univariate analyses, there were significantly better performances by the healthy control group on the Letter Fluency (p <.04) and Category Fluency tasks (p <.04) compared to the ADHD participant group. No significant differences were observed on any other EF related performance tasks. Furthermore, the three EF measures that did not meet the homogeneity of variance assumption were analyzed using non-parametric analysis, Mann-Whitney U Test. The results found significantly higher performances on the Colour Word Interference Task, (U = 86.5, p < .05) and on the Trail Making Test B (U = 78.5, p < .04) by the healthy controls group compared to the ADHD group. No group difference was observed on the Wisconsin Card Sorting Task perseverative error measure (p < .04).

Table 3.

Descriptive Statistics of Executive Functions of Children with ADHD and Healthy Controls at Baseline

Parametric Analysis	ADHD (n = 21)		Healthy Controls (n = 15)		F	partial η 2
	M	SD	M	SD
Working Memory
Digit Span Backwards	98.00	10.56	100.33	16.53	.26 .01
Spatial Span	107.75	14.82	113.33	15.20	1.19	.04
Fluency
Letter	95.00	14.14	105.67	15.10	4.60*	.12
Category	99.00	13.53	110.67	8.84	8.42**	.20
Inhibition
CPT Commissions Errors	52.48	7.89	52.93	7.09	.00	.00
CPT Omissions Errors	57.52	13.62	54.40	16.64	.10	.00
CPT Reaction Time	56.00	8.54	50.73	12.60	1.92	.06
Non-Parametric Analysis	Mean Rank		Mean Rank		χ 2
Response Inhibition
D-KEFS Colour Word Interference (Mean Rank)	14.83		22.23		86.50*
Cognitive Flexibility
Wisconsin Card Sort Task (WCST) Perseverative Errors (Mean Rank)	15.68		21.10		103.50
TMT-Part B Time Standard score (Mean Rank)	14.74		23.77		78.50

^*Note. CPT = Cognitive Performance Task, DKEFS = Delis Kaplan Executive Function System, TMT= Trails Making Test-Part B,

^*indicates p < 0.04 following Benjamini-Hochberg correction

Relations between Glutamate and EF Performance

Pearson and Spearman correlations were conducted with glutamate and the different EF measures (see Tables 4). Results revealed significant positive correlations in the healthy control group between glutamate concentration in the right PFC and performance on the Digit Span Backwards task, r = .59, p < .03, and on the Letter Fluency task, r = .59, p < .03. No significant correlation was found between performance on TMT-Part B task and glutamate concentration, r = .50, p > .03. No other significant correlations between glutamate concentration and EF performance in the healthy controls group were observed. Additionally, there were no significant correlations observed between glutamate concentration and performance on the EF measures in the ADHD group.

Table 4.

Pearson and Spearman Correlations between Glutamate Concentration and EF performance

Measure	ADHD Glutamate Concentration	Healthy Controls Glutamate Concentration
Working Memory
Digit Span Backwards	r = .14	p = .60
Verbal Fluency
Letter	r = .01	p = .97
	r = .59*	p = .03
Category	r = −.19	p = .45
	r = .43	p = .13
Cognitive Flexibility
TMT-Part B Time score	r = −.26	r = .50
	p = .27	p = .07
D-KEFS Inhibition Score	r = .02	p = .36
	p = .93	p =.21

Notes. Glu =Glutamate concentration, Benjamini Hochberg Correction, p value = .03,

^*indicates p <.03..

TMT-Part B= Trails Making Test-Part B, DKEFS = Delis Kaplan Executive Function System.

Discussion

The purpose of the current study was to take an integrated brain-behaviour approach to study spectroscopic biomarkers of ADHD and evaluate the association of these spectroscopic biomarkers with different EF tasks. Specifically, the present study intended to answer these three research questions: (1) What are the differences in glutamate, NAA, and choline concentrations between children with ADHD and healthy controls in the right PFC and the left striatum? (2) Do children with ADHD demonstrate EF deficits on performance-based measures assessing inhibition, working memory, cognitive flexibility, and verbal fluency when compared to healthy controls, and (3) Are there any relations between the neurometabolite glutamate with EF performance-based measures such as inhibition, working memory, cognitive flexibility, and verbal fluency?

Overall findings revealed neurochemical differences in children with ADHD compared to healthy controls. These neurochemical differences in the glutamate, NAA, and choline concentrations were explicitly observed in the right PFC. Study findings also demonstrated specific EF challenges in children with ADHD as measured by performance-based measures. Lastly, the current study demonstrated clinical relevance of the glutamate concentrations in relation to EF tasks. The following sections further describe the results and implications of the findings.

Glutamate

The current study found lower levels of glutamate in children with ADHD compared to healthy controls in the right PFC with no significant differences noted in the left striatum. These findings differ from previous studies, which found higher levels of glutamate in children with ADHD compared to healthy controls (Endres et al., 2015). Given the heterogeneous ADHD symptom presentation which lies along a spectrum of impairment (Kofler et al., 2018), it is likely that a similar range of differences is also observed at the neuronal level. As such, the observed findings of lower glutamate concentration could indicate that the current sample comprised a new theoretical subgroup of children with ADHD who have hypo-glutamatergic activity in their right PFC.

Other explanations for the lower concentration of glutamate in children with ADHD may also be reasonable. The difference in result could be due to the improved methodological rigour and technology used, such as larger sample size compared to previous studies (Carrey et al., 2003; Courvoisie et al., 2004; MacMaster et al., 2003; Sparkes et al. 2004) and improved MRI technology (e.g., 3 Tesla MRI magnet with better spectral resolution and lower signal to noise ratio [Stagg & Rothman, 2014]).

Furthermore, the participants in the current study were not medication naïve whereas some of the previous MRS studies investigating differences in glutamate included medication naïve children with ADHD (Carrey et al., 2007; MacMaster et al., 2003). Given that the current research literature lacks information regarding the long-term adaptation of neurometabolites following usage of methylphenidate, it is possible that the results could be different because of long-term stimulant usage.

No significant difference in glutamate concentration was observed in the left striatum between children with ADHD and healthy controls. Similar results have been found in a recent MRS study with children with ADHD (Naaijen et al., 2017b). It is possible that for participants in the current study, ADHD symptomology was primarily caused by dopaminergic dysfunction, thus making the glutamate dysfunction in the striatum less robust. Additionally, given that the current study found hypo-glutamatergic signalling in the PFC, it is possible that the striatum is driving the ADHD symptoms. Since MRS is unable to capture the concentration of dopamine, the current study was unable to investigate the dopamine concentration in the sample.

NAA

The results from the current study also found significantly lower levels of NAA in the right PFC in children with ADHD compared to healthy controls. No such differences were observed in the left striatum. These findings are consistent with previous MRS studies conducted in children with ADHD (Hesslinger, Thiel, van Elst, Hennig & Ebert, 2001; Tafazoli et al., 2013). NAA is known to be an indicator of a neuronal marker of cell health (Rae, 2014). However, the exact implication of altered NAA is not well understood. Some studies indicate that the lower levels of NAA could be due to cell loss (Rae, 2014; Stagg & Rothman, 2014).

Other studies have suggested that lower levels of NAA could indicate delayed brain maturation and development (Rae, 2014). While it is more likely that the differences in NAA observed in the current study are due to delayed cell development and not cell loss, future studies need to focus on further understanding the role of NAA.

Choline

We found lower levels of choline in the right PFC in children with ADHD compared to healthy controls. This finding is similar to previous MRS studies (Tafazoli et al., 2013; Wiguna, Guerrero, Wibisono & Sastroasmoro, 2014). Similar to the role of NAA, the specific role of altered choline is not well understood. Some studies have reported choline to be a marker of cell density or cell turnover (Rae, 2014). Choline-containing compounds are the main components of cell membranes and products of membrane degradation. Reduced choline concentration could also indicate active demyelination given that phospholipids constitute around 40% of myelin (Stagg & Rothman, 2014). Furthermore, choline is the precursor of acetylcholine and influences neural communication, mediated by transmitters like norepinephrine and dopamine (Rae, 2014). Given that previous neuroimaging studies with children with ADHD have reported cortical thinning in the PFC (Shaw et al., 2007; Yang et al., 2015), it is possible that lower levels of choline could affect their cell density and myelination.

EF Performance

In contrast to initial hypotheses, no significant group differences were found in performance across the different EF domains in children with ADHD as compared to healthy controls. Given the effect size, Wilk’s Lambda Λ = .30; it is possible that larger sample size could have led to significant findings.

When looking at individual domains of EF, the current study found weaker performances on the DKEFS-Colour Word Interference task, Trails-Making Task-Part B and both phonemic and semantic fluency tasks. However, no significant group differences were observed on the CPT Commission Errors, WSCT Perseverative Errors, Digit Span Backwards and the Spatial Span Backwards tasks. These differences in performance could be due to the requirements for sustained attention and concentration as well as the complexity of the tasks. These results are not surprising as the current literature supports variable performance across multiple domains of EF in children with ADHD (Toplak et al., 2009; Willcutt et al., 2005). Specifically, Kofler et al., (2018) indicated that when EF is considered as a single entity, only 39% of their participant sample exhibited EF challenges, whereas this percentage changed to 89% when the different EF components were investigated individually as separate entities.

These findings should also be evaluated in light of the recent limitation of performance-based measures. While numerous studies have indicated that children with ADHD struggle with some aspects of EF (Kofler et al., 2018; Willcutt et al., 2005), neuropsychological measures are often unable to accurately measure this deficit (Toplak et al., 2009). As such, neuropsychological test performance alone is not an indication of disorder-specific challenges and should be complemented with parent and teacher ratings of EF (Toplak et al., 2009).

Relations between glutamate and EF

Interestingly, we did find positive correlations between glutamate concentration in the right PFC and performance on the Digit Span Backwards and the Letter Fluency tasks in the healthy control group. These significant correlations could suggest that although increased levels of glutamate might help with EF performance in healthy controls, it appears to indicate a decoupling effect in children with ADHD, possibly affecting their EF task performance. Naaijen and colleagues (2017a) have previously shown that glutamate signalling-linked genetic variation is associated with increasing hyperactivity/impulsivity symptoms which could potentially affect EF performance (Naaijen et al., 2017a). Another recent study with adults with ADHD found associations between glutamate concentration in the Anterior Cingulate Cortex (ACC) and an increased number of errors during a cognitive control task (Naaijen et al., 2018). As such, it is likely that glutamate concentrations in the right PFC affect EF performance in children with ADHD.

Although higher levels of glutamate concentration can cause cell excitotoxicity, these processes are not well understood (Rae, 2014). It is conceivable that individuals with ADHD have some genetic predispositions that make their cells sensitive to glutamate (Elia et al., 2012), which may affect EF performance. In contrast, participants in the healthy control sample may not possess this genetic predisposition and therefore may not have a similar sensitivity to the glutamate concentration. We can speculate that dysfunctions in the glutamate receptors in individuals with ADHD impact performance in EF tasks and causes this decoupling effect. In sum, the results of the current study indicate the clinical relevance of glutamate in working memory and verbal fluency tasks. Indeed, this report also has implications from the perspective of the Research Domain Criteria (RDoC) approach advocated by the National Institutes of Health (NIMH)(Insel, 2014). Specific aspects of executive function, like Digit Span Backwards and the Letter Fluency, could be explored from the RDoC vantage point, moving from those performance measures through related circuits, to glutamate as observed here, and into receptors (Huang et al., 2019) and specific related genes (Noroozi et al., 2019).

Study Implications

The findings from our study indicate that numerous neurochemical changes are occurring in the brains of children with ADHD when compared to their typically developing peers. It further extends the conceptualization of ADHD as a neurodevelopmental disorder with an underlying biological basis (APA, 2013).

In addition to the neurochemical differences that were observed, our results also showed variable EF performance in children with ADHD, which further support the idea that there are no disorder-specific EF deficits that can be observed in children with ADHD (Kofler et al., 2018; Willcutt et al., 2005). The results also highlight that some individuals with ADHD are resilient, and have learned compensatory strategies to alleviate some of their EF challenges

The main contribution of the current study is the finding of a possible theoretical subgroup of children with ADHD with hypo-glutamatergic signalling. If our conclusions about this new theoretical subgroup of children with ADHD who have hypo-glutamatergic activity in their PFC is replicated in other studies, it would be essential to use alternative treatments options for such individuals.

Limitations and Future Research

The results from this study should be evaluated in the context of some study limitations. The ADHD sample in the current study was based on referrals from primary care physicians and might have included children with more severe ADHD symptomatology compared to community samples. The participants were also not medication naïve. Additionally, the investigation of biological sex differences and the effect of comorbidities on EF performance and MRS outcomes was beyond the scope of this study. Finally, the present study only used neuropsychological measures to evaluate EF. Given that there are discrepancies in the current literature in regard to the utility of neuropsychological measures to assess EF, it would have been beneficial to corroborate parent ratings of EF with neuropsychological findings to gather a better understanding of EF challenges in children with ADHD.

Future research needs to replicate the findings from the current study. Given the findings of a new possible subgroup of ADHD individuals with hypo-glutamatergic activity, future studies may wish to investigate larger cohorts of ADHD individuals to understand other potential subgroups of ADHD based on neurobiological differences and characterize such groups with clinical data. By characterizing such subgroups based on neurobiological findings, it may be possible to provide more targeted interventions. Therefore, forthcoming studies could investigate the efficacy of treatment using neurostimulation in children with ADHD. Given the changes associated with the FSP in ADHD, stimulating areas of the PFC such as superior frontal gyrus, or the dorsolateral prefrontal cortex might lead to changes in these neurotransmitters.

Conclusion

The current study integrated a brain-behaviour approach and found significant differences in glutamate, NAA, and choline concentration in the right PFC in children with ADHD, compared to typically developing peers. The reduced levels of glutamate observed could indicate a new theoretical subgroup of children with ADHD with hypo-glutamatergic signalling. We observed some significant differences in EF task performances. As well, we found positive correlations between glutamate concentration and performance on EF tasks indicating glutamate concentrations in the right PFC may be a possible biomarker of ADHD.