INTRODUCTION
The prevalence of attention-deficit/hyperactivity disorder (ADHD) in childhood and adolescence is estimated to range between 5% and 7% [1]. ADHD generally results in chronic impairment, with 60%-85% of diagnosed children continuing to exhibit symptoms into adolescence [2]. More than half of children and adolescents diagnosed with ADHD have one or more additional psychiatric conditions, and over one-quarter have two or more comorbidities [3]. Comorbidity is linked to a higher incidence of psychiatric hospitalization, increased rates of suicide, a poorer quality of life, diminished social functioning, and impaired family functioning, and can increase the likelihood of ADHD symptoms persisting into adulthood [4,5].
Children with ADHD have approximately four times the risk of developing depression compared to typically developing children [6]. ADHD and depression have distinct neurological differences and different genetic predispositions or prefrontal cortical networks are linked to the development of depression in children with ADHD [7,8]. However, ADHD and depression share similar symptoms, such as inattention, mood instability, and impaired executive function [9]. Furthermore, medications used to treat ADHD symptoms can have side effects resembling those of depression, including insomnia, appetite changes, and sleepiness [9]. It is crucial to clearly distinguish between and diagnose ADHD and depression. However, distinguishing depression in children with ADHD can be challenging, and research on discriminating comorbidities is relatively rare. Identifying the neurophysiological differences between ADHD and depression can facilitate differential diagnosis.
Electroencephalography (EEG) has become a promising biomarker for various psychiatric illnesses due to its user-friendly nature, cost-effectiveness, and widespread accessibility [10]. ADHD has been extensively researched using quantitative EEG (qEEG), and approximately 60% of qEEG studies utilize it as a diagnostic aid [11]. Children with ADHD show increased theta activity, decreased alpha activity, and elevated theta/beta ratio during eyes-closed rest [11]. Increased theta activity persists from childhood to adulthood in individuals with ADHD [12]. These results indicate that children with ADHD present hypo-arousal, delayed maturation, and deviations in the development of the central nervous system [11].
qEEG related to depression has been studied as a potential biomarker for depression diagnosis and treatment responsiveness [10]. Several studies that measured power in qEEG frequency bands presented conflicting results between a depression group and healthy controls but consistently showed a relationship between alpha activity and depression [13]. Patients with major depressive disorder exhibit comparatively increased right versus left frontal alpha activity [14]. Reduced right parietal alpha activity is observed in patients with depression because this area is related to regulating emotion-related autonomic arousal [15]. Reduced activity might indicate decreased emotional arousal in the disorder. Another study found that posterior alpha activity is negatively related to depression severity [16]. Some studies have confirmed a relationship between qEEG and ADHD symptoms in adults with major depressive disorder [17,18]. Hyperactivity symptoms are positively correlated with alpha activity in the frontal lobe [17], while inattentive symptoms are negatively correlated with beta activity [18].
To our knowledge, no study has examined differences in qEEG profiles with regard to ADHD symptom severity and the co-occurrence of depression in children diagnosed with ADHD. Exploring qEEG characteristics related to symptom differences and severity of the two disorders in children with depression and ADHD is essential to enhance our understanding of ADHD in children. Therefore, this study investigates qEEG activity profiles concerning ADHD symptom severity and the presence of comorbid depression in children with ADHD. We hypothesized that, in children with ADHD, theta activity would be positively correlated with the severity of ADHD symptoms, beta activity would be negatively correlated, and alpha activity would be further diminished when depression is also present.
METHODS
Participants
This prospective cross-sectional study was conducted at the child and adolescent psychiatric clinic of Soonchunhyang University Seoul Hospital from 1 August 2018 to 31 December 2020. Outpatients who visited the hospital for symptoms of hyperactivity, inattention, and/or impulsivity were included in this study if they had been diagnosed with ADHD combined type based on the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5) [19]. Participants were 87 children between ages 5 and 18 years with ADHD (68 males; mean age±standard deviation [SD]: 10.4±3.3 years). Additionally, major depressive disorder, persistent depressive disorder, other specific depressive disorder, and unspecified depressive disorder were diagnosed according to the DSM-5 and were categorized as depression. Children were classified into three groups based on the severity of their ADHD and depression: mild ADHD (MA), severe ADHD without depression (SA), and severe ADHD with depression (SAD). ADHD and depression severity were categorized based on the cutoff scores of the Korean ADHD rating scale-IV (K-ARS-IV) and the Children’s Depression Inventory (CDI) [20,21]. In this study, children with a K-ARS score of less than 19 points were classified as having mild ADHD, while those with a score of 19 points or more were classified as having severe ADHD. Additionally, a CDI score of 22 points or higher was used to identify the presence of depression. The children received a psychological evaluation and resting-state qEEG. Children did not take psychiatric medications during the qEEG evaluation, and patients who took medications were evaluated with a washout period of at least seven days (19 children). The children in this study had ADHD as their main diagnosis. Exclusion criteria were brain injury, primary diagnosis of a mental disorder other than ADHD, neurological disorder, severe medical condition, or refusal of the child or parent to give consent. The dominant hand of the children was the right in 94.5% (86 children) and the left in 6.5% (6 children). Written informed consent was obtained from all individual children and their parents. This study was conducted following the Declaration of Helsinki and approved by the Institutional Review Board of Soonchunhyang University Seoul Hospital (no. SCHUH 2017-08-003).
EEG
Children received EEG measurement once after enrolling in the study. They were instructed to sit comfortably in a sound-attenuated room while the EEG was recorded and to keep their eyes closed for 5 minutes until clear data were received. EEG was recorded from Ag/AgCL surface electrodes using a 64-channel EEG cap (Neuroscan) utilizing an extended international 10/20-location system and a Neuroscan acquisition system (Neuroscan) with a sampling rate of 1 kHz. Reference electrodes were located on the ears.
The EEG data were processed and analyzed using the opensource software EEGLAB (v2021.1) implemented in MATLAB® 2021b (MathWorks Inc.) [22]. First, the baseline signal was removed from the raw data measured in the EEG test, and a reference was established based on the average of the EEG signal. As the measurement data were all measured in South Korea, power line noise was removed using a 60 Hz notch filter, followed by a bandpass filter (cutoff frequency: 1 to 40 Hz). Then, artifact subspace reconstruction was used to remove noise and bad epochs [23,24]. Nineteen independent components were extracted using independent component analysis. Electrooculogram and electromyograph components were removed from each independent component using a multiple artifact rejection algorithm [25].
Relative spectral power is a method for accurately analyzing EEG signal data. The power spectrum density (PSD) of the preprocessed EEG signal was obtained in each of the 19 channels using Welch’s method [26]. EEG spectral power (μV2) data were averaged into the following 5 bandwidths: delta (1-4 Hz), theta (4-8 Hz), alpha (8-12 Hz), beta (12-25 Hz), and high beta (25-30 Hz). The relative band power was obtained by dividing the total sum of band power for each channel by absolute band power. The Neuroguide system provides eyes-closed and eyes-open normative EEG data of 625 people between the ages of 2 months and 82.6 years [27]. The z-score relative band power was obtained by substituting the measured relative power into a Neuroguide database, standardized by age and gender. The z-score relative band power was used for analysis. This study focused on investigating the correlation between patients’ clinical scales and quantified EEG, using the z-score relative EEG power in the channels frontal (Fz), central (Cz), and parietal (Pz) of the midline, as is typically analyzed [28]. To evaluate the default mode activity (the brain’s intrinsic resting activity), we measured the relative spectral power of each band while children rested with their eyes closed [29]. All EEG data reviews, transformations, and analyses were conducted blind to clinical scale scores.
Psychiatric assessments
ADHD symptom severity of children with ADHD was evaluated using the K-ARS-IV [20,30]. The scale comprises 18 items designed to measure the ADHD diagnostic criteria specified in the DSM-IV. Each item is assessed using a four-point Likert scale that focuses on the frequency of the child’s problematic behavior: “never or rarely” is scored as 0, “sometimes” as 1, “often” as 2, and “very often” as 3. The scale is designed so that the total score of the odd-numbered items indicates inattention, while the total score of the even-numbered items reflects hyperactivity-impulsivity. Higher scores on the K-ARS-IV and its subscales (inattention: K-ARS-In; hyperactivity-impulsivity: K-ARS-H) indicate greater severity of ADHD symptoms. Noting that all children had an existing ADHD diagnosis, those with a total score of 18 points or less were categorized as having mild ADHD, while those with a total score of 19 points or more were classified as having severe ADHD using the cutoff for clinical significance of Jang et al. [30]. The K-ARS-IV standardized by Jang et al. [30] demonstrated good internal consistency (e.g., Cronbach’s alpha ranged from 0.74 to 1) and has demonstrated construct, concurrent, and discriminant validity [31,32].
Depression and anxiety were evaluated with the widely used CDI [21] and State Anxiety Inventory for Children (SAI-C) [33]. As noted, diagnoses of ADHD and depression were also made using DSM-5 criteria, with diagnosable major depressive disorder, persistent depressive disorder, other specific depressive disorder, or unspecified depressive disorder considered to be depression.
The Child Behavior Checklist (CBCL) [34] is a parent-report tool used to evaluate the emotional and behavioral problems of children aged between 6 and 18 years. It includes eight syndrome scales, namely anxious/depressed, withdrawn/depressed, somatic complaints, social problems, thought problems, attention problems, rule-breaking behavior, and aggressive behavior, which group into two higher-order factors: internalizing and externalizing. Additionally, it includes scales that align with DSM disorders, namely affective, anxiety, somatic, ADHD, oppositional/defiant, and conduct problems. A total problem score can also be calculated.
Statistical analysis
Data were analyzed using IBM SPSS 27.0 (IBM Corp.). The mean±SD were computed, and the normality and homogeneity of variances were assessed using the Shapiro-Wilk test and Levene’s test, respectively. Group differences in demographic characteristics, clinical scales, and EEG data were analyzed using one-way analysis of variance and the Kruskal-Wallis test. When comparing EEG, the analysis included covariates of age [35] and SAI-C to exclude the effect of age and anxiety symptoms. Bonferroni-adjusted post hoc tests were conducted to examine the differences between each pair of groups. The relationships between psychological assessments and z-score relative power were assessed using Pearson’s correlation coefficient. For all analyses, the significance level was set at p<0.05.
RESULTS
Demographic and functional characteristics of the children
Table 1 shows that the SAD group was significantly older than the MA and SA groups. There were no differences in the sex ratio, comorbidity, or medication history in any of the three groups. There were significant differences in K-ARS, SAI-C, and CDI scores among groups. Both the SA and SAD groups had similar K-ARS scores, including inattention symptoms and hyperactivity/impulsivity symptoms, but the MA group had lower K-ARS scores. The SAD group had significantly higher scores on the SAI-C and CDI than the MA and SA groups (all differences p<0.05).
For the CBCL scales, the anxious/depressed and withdrawn/depressed scale scores were significantly higher in the SAD group than in the MA group (Table 2). The SA and SAD groups had higher scores than the MA group for the subscales of externalizing and total problems. The SA and SAD groups exhibited significantly higher scores than the MA group for the CBCL DSM scales, including affective, ADHD, oppositional/defiant, and conduct.
Neurophysiological function
Figure 1 presents the relative power of the Fz, Cz, and Pz regions in five frequency bands.
There were significant differences in delta, alpha, beta, and high beta bands in all regions. Similar results were also observed for the z-score relative spectral power (Table 3). Specifically, the SAD group exhibited the lowest delta band power in central and parietal regions compared to the other groups. The SA group also showed lower theta power in the central region compared to the MA group. In contrast, the MA group exhibited the lowest alpha band power among the groups. The neurophysiological features of each group were similarly revealed in the grand average topography of 19-channel quantitative EEG z-score relative power (Figure 2).
Compared to the MA group, alpha band power was greater in the SAD group, whereas a relative decrease was observed in the delta and theta bands.
Correlations of the clinical scales with neurophysiological functions
In Table 4, the K-ARS total score showed a statistically significant positive correlation with the alpha band in all regions (Fz, r=0.40; Cz, r=0.43; Pz, r=0.37, p<0.01). Conversely, the K-ARS total score had a negative correlation with the delta band of Pz (r=-0.25, p<0.05) and theta band of all regions (Fz, r=-0.30; Cz, r=-0.35; Pz, r=-0.33). The CDI score had a negative correlation with the delta band power of all regions (Fz, r=-0.21; Cz, r=-0.24; Pz, r=-0.24, p<0.05); no significant correlation was observed in other bands. Similar to the total KARS score, the CBCL total score and the externalization score exhibited a positive correlation with the alpha band and a negative correlation with the delta and theta bands. Conversely, anxious/depressed scores showed a significant negative correlation in the delta band, similar to the CDI scores.
DISCUSSION
ADHD commonly coexists with depression, leading to a worse prognosis. Accurate diagnosis and treatment of these comorbidities are essential. However, it can be challenging to differentiate ADHD and depression symptoms in children, especially when depression is not easily noticeable. Neurophysiological tests, such as qEEG, may help differentiate between the two disorders due to their distinct physiological mechanisms and pathologies. However, very few studies have examined the relationship between qEEG and depression and other accompanying symptoms in children with ADHD. Therefore, our study explored qEEG according to ADHD symptom severity and depression in children with ADHD. The key findings in this research are as follows: 1) among the three groups, the SAD group (ADHD and depression) had the highest average age of 12.5 years; 2) ADHD symptom severity had a statistically negative correlation with the theta band of all regions; 3) when ADHD symptoms in children with ADHD were more severe, alpha activity tended to be higher; 4) depression severity had a negative correlation with delta band power in all regions; and 5) externalization and aggressive severity exhibited a negative correlation with high beta band power at Fz and Pz.
The SAD group had an average age of 12.5 years, the highest among the three groups and approximately 2 years higher than the other two groups. This relates to the severity of ADHD symptoms and the age of onset of depression [6,36]. Children with ADHD experience increased symptoms of depression, which are related to academic attainment and peer relationships [6,36]. Children who struggle academically and socially are more likely to experience feelings of failure and rejection, leaving them more vulnerable to depression and emotional problems [36,37]. The more severe the symptoms of ADHD, the more difficult it is to achieve academic success and maintain peer relationships, which increases the likelihood of depression. In contrast, depression typically begins in individuals between their mid-teens and late 20s, with approximately half of all cases starting at age 14 years [38,39]. In the current study, the average age of children with severe ADHD and depression was lower than the age of onset of depression typically reported. A multivariate analysis found that very early onset depression, which refers to depressive symptoms developing before the age of 12 years, is highly associated with ADHD when comparing all neuropsychiatric genetic risk scores [7]. Therefore, it is crucial to monitor depression in children with ADHD, as the presence of depression is linked to an unfavorable long-term prognosis for such children [4].
ADHD symptom severity in children with ADHD had a statistically negative correlation with the theta band of all regions. Previous research has consistently reported increased theta activity at Fz, Cz, and Pz of children with ADHD compared to typically developing children [11]. Elevated absolute and relative theta activity persists from childhood through adolescence to adulthood in individuals with ADHD [11]. Abnormally increased theta activity in the resting state is a phenomenon caused by thalamocortical dysrhythmia [40]. It is related not only to brain hypo-arousal but also to other disorders such as attention deficit hyperactivity disorder and depression. However, most previous studies of theta activity have compared typically developing children to those with ADHD, focusing on non-specific signs related to inattention [11]. There is a lack of studies exploring the relationship between ADHD symptom severity and theta activity in children with ADHD. Furthermore, the theta/beta ratio, which has been controversial as an indicator of cortical hypo-arousal, did not show a statistically significant correlation with ADHD symptoms or any other symptom of the CBCL [11].
When ADHD symptoms in children with ADHD are more severe, alpha activity tends to be higher. The CBCL outcomes indicated a positive correlation with external symptoms, attention, and aggression. Previous studies have considered the connection between ADHD and alpha activity, but their results have been inconsistent. Some studies have indicated that the ADHD group displayed either increased or decreased alpha activity compared to the control group, while other studies did not establish a clear relationship between ADHD symptoms and alpha activity [11,41]. In general, alpha activity is considered an inverse marker of arousal and is related to cognitive processing in both children and adults [11,42]. Increased global alpha activity indicates decreased arousal and is related to low performance. As ADHD symptoms become more severe, arousal, concentration, and executive function decline; therefore, alpha activity should increase. Clarke et al.’s studies [11] have consistently reported abnormal arousal mechanisms in children with ADHD, although children with ADHD have low alpha activity compared to typically developing children. In contrast, when depression is present, the decrease in concentration and performance is even more pronounced, accompanied by a further increase in alpha activity. However, in this study, although there was a positive correlation between the severity of depression symptoms and increased alpha activity, the results were not statistically significant. Increasing the sample size and conducting a more detailed examination of alpha activity in future studies is necessary to clarify these findings.
Depression severity had a negative correlation with the delta band power of all regions. Although there have been no studies of depression and delta activity in children with ADHD, studies have been conducted on the association between depression and slow-wave activity in adults [43-45]. The pathophysiology of depressive symptoms in the default mode network (DMN) [46,47] is attributed to reduced and imbalanced connectivity of the fronto-parietal system [48]. In a simultaneous functional magnetic resonance imaging (fMRI)-EEG study, a strong correlation was found between frontal delta power and DMN activity [29,44,45]. The DMN is less active during task-oriented activities and more active during rest, with the level of attenuation possibly being stronger during more demanding tasks. In a study of adults with depression, a decrease in resting-state delta power was observed as psychological pain increased, suggesting that the DMN was less active [43]. In our study, we observed that delta activity decreased as depression worsened in children with ADHD. This decrease may be linked to the reduced and imbalanced connectivity of the DMN. However, this suggestion requires testing via additional qEEG and fMRI studies of children with ADHD, considering the presence of depression, as well as the type and severity of symptoms, to clarify this relationship. Furthermore, we found that delta activity exhibited a negative correlation with inattention symptoms, indicating the need for further research to determine whether this is related to inattention symptoms associated with depression or ADHD.
Externalization and aggressive severity showed a negative correlation with high beta band activity at Fz and Pz. There have been few studies of aggression due to the difficulty of evaluating control groups, but some studies have reported inconsistent results regarding ADHD symptoms and beta activity [10,11,49]. It has been reported that children with ADHD and anxiety disorder present increased beta activity, leading to decreased aggression and externalizing disorders, potentially leading to underestimation of ADHD symptoms [50,51]. Clarke et al. [52] found that a certain subgroup of ADHD showed increased beta activity; this group was more aroused and hyperactive compared to other subgroups of ADHD. In this study, beta activity did not exhibit a correlation with externalizing symptoms and aggression; however, high beta activity was negatively linked to externalizing symptoms and aggression. Some studies have divided beta activity into low (13-21 Hz) and high beta activity (22-30 Hz), or low (12-15 Hz), midrange (15-20 Hz), and high beta activity (18-40 Hz) based on frequency and confirmed a distinction among bands [53,54]. Low beta activity is associated with a healthy oscillation network, such as a state of quiet, focused, and introverted concentration, while high beta activity is linked to an energy-consuming network, represented by significant stress, anxiety, paranoia, high energy, and high arousal. Midrange beta activity is associated with increased energy, anxiety, and performance. High beta activity is associated with significant stress, anxiety, paranoia, high energy, and high arousal [54]. Hence, midrange beta activity suggests that ADHD symptoms may be more manageable due to anxiety, while high beta activity may indicate the potential for ADHD symptoms to worsen due to heightened anxiety and sensitivity, leading to increased aggression. To generalize the findings of this study, it is necessary to analyze the differences in symptoms based on the frequency range of beta activity in a larger sample by categorizing them according to symptoms.
This study has some limitations. First, the differences in brain development among children and adolescents were not considered in this study. Since brain function varies at different developmental stages, it is essential to interpret the results with caution. Future studies will address this issue by including a larger sample size. Second, the study did not include a control group of typically developing children for comparison with children with ADHD. Accurately determining whether EEG differences stem from ADHD, depression symptoms, or developmental variations is challenging. Caution should be exercised when interpreting these results. Third, this study did not differentiate qEEG according to ADHD and subtypes of depression. The severity and the duration of the depression were not verified. Fourth, the study did not analyze parameters other than beta activity in detail. Recently, differences in alpha activity based on symptoms were examined by categorizing such activity into low and high levels. Fifth, we did not explore the relationship between gamma activity and concentration [55,56]. Sixth, a subjective assessment tool was used to measure psychological symptoms. Seventh, 16 children stopped taking the drug for 7 days before the initial evaluation, resulting in a washout period; however, completely excluding the influence of the medication is challenging.
Despite these limitations, this study confirmed the potential neurophysiological markers of qEEG for differentiating the severity of ADHD symptoms and the presence of depression in children with ADHD. Alpha and theta activity might serve as a reliable measure of ADHD symptom severity, while delta activity might function as a reliable measure of depression severity in children with ADHD. Future research is necessary to generalize and specify the findings of this study by incorporating more comprehensive investigations that consider the severity and subtypes of ADHD, associated comorbid disorders, and typically developing children.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
