Improved effect of EEG-biofeedback intervention on cognitive function in childhood idiopathic epilepsy with ADHD: A retrospective study Authors: Dazhi Cheng, Yuchen Yang, Xiuxian Yan, Qian Chen Published: 2024 ABSTRACT AbstractBackgroundAlthough childhood idiopathic epilepsy syndromes are frequently associated with a good prognosis, they also have a high prevalence of comorbid attention-deficit hyperactivity disorder (ADHD). The current study aimed to examine whether an intervention using electroencephalogram (EEG) biofeedback can improve cognitive performance in childhood idiopathic syndromes with ADHD.MethodsIn this study, thirty-six childhood idiopathic epilepsy syndrome patients with comorbid ADHD were divided into either the treatment or the control group. All participants’ cognitive functions were assessed by a standardized test battery on processing speed, spatial perception, arithmetic, semantic comprehension, forward/backward verbal working memory, non-verbal matrix reasoning, visual attention and executive function before and after the training.ResultsANOVA analysis showed that the treatment group improved significantly on task performances in forward digit span and backward digit span tasks, but not in other cognitive assessments. The findings indicated that the treatment group receiving the EEG-Biofeedback intervention showed improved verbal working memory.ConclusionsThe findings suggested that EEG biofeedback can improve the core cognitive function in childhood idiopathic epilepsy syndromes with ADHD.Keywords: Electroencephalogram biofeedback, Childhood epilepsy, ADHD, Cognitive function, Working memory Introduction As one of the most common neurological disorders in pediatrics [1], childhood epilepsy is complicated by psychiatric, cognitive, and social comorbidities [2]. Although children with childhood idiopathic epilepsy syndromes have a good prognosis in epileptic seizure prior to adulthood, they still suffer from the burden of comorbidities including cognitive function deficits in attention, memory and language [3]. One of the common cognitive dysfunctions is impaired attention or the behavioral form of attention-deficit hyperactivity disorder (ADHD) [4]. There are many nonpharmacological approaches for treating ADHD besides psychostimulant drugs [5]. Nonpharmacological therapies ranged from behavioral interventions to complementary medicines. There are several types of behavioral interventions such as cognitive training, cognitive behavioral therapy and neurofeedback. In cognitive training, computerized feedback is used to reinforce correct responses and then improve attentional abilities. Cognitive behavioral therapy for ADHD includes behavioral parent training and behavioral classroom interventions. In electroencephalogram (EEG) biofeedback, participants were instructed to consciously modulate their brain activity by operant conditioning [6,7]. The intervention principles of EEG biofeedback are based on a control theory framework that utilizes neurofeedback to modulate brain activity. At the beginning of training, the feedback shows fluctuations reflecting unconditioned neural variability. As training proceeds, the trainee's brain activity is adjusted to the threshold for obtaining rewarding feedback, and an optimal balance of brain waves is achieved through a series of learning events. This process involves neurophysiological mechanisms such as dopamine receptor activation. Previous studies have proved the effectiveness of EEG biofeedback for the treatment of ADHD [[7], [8], [9]]. A recent meta-analysis suggested that the effect of EEG biofeedback intervention on both hyperactivity and inattention symptoms of ADHD could last for 6–12 months [9]. In addition, a number of recent studies found that children with ADHD who were receiving EEG biofeedback showed improvement in executive functioning that was significantly more enduring compared to children in the cognitive training condition. Compared to cognitive training and cognitive behavioral therapy, EEG biofeedback has two major advantages. On one hand, the positive effect of EEG biofeedback on improvement in executive functioning among children with ADHD was shown to be enduring [10]. On the other hand, EEG biofeedback is precisely targeted during regulation of brain activities [11]. In addition, Bakhtadze et al. evaluated the effects of EEG biofeedback on cognitive flexibility in children with ADHD with and without temporal lobe epilepsy and found that executive functioning improved significantly in all groups of patients [12]. Although EEG biofeedback intervention has proved its effectiveness for the treatment of ADHD, its effect on cognitive functions in children with childhood idiopathic epilepsy syndromes with ADHD is unknown. The current study examined whether EEG biofeedback can improve cognitive performance in childhood idiopathic epilepsy syndromes with comorbid ADHD. We expected to observe a positive effect of the EEG biofeedback intervention on the core cognitive ability in childhood idiopathic epilepsy syndromes with comorbid ADHD. To investigate the intervention effect, a battery of neurocognitive tests were administered before and after the training to measure the improvements in cognitive abilities. Materials and methods Participants Patients with a history of childhood idiopathic epilepsy syndromes were enrolled in the study. All participants came from the clinical case in department of neurology in the Capital Institute of Pediatrics in China. Based on the criteria of the International League Against Epilepsy Childhood, idiopathic epilepsy syndromes were diagnosed by a clinical pediatric neurologists [13]. Idiopathic epilepsy syndromes included childhood absence epilepsy (CAE), self-limited epilepsy with centrotemporal spikes (SeLECTS) and focal seizures epilepsy. Patients with CAE were characterized with typical absence seizures induced by hyperventilation, and showed 3–4Hz EEG generalized spike-and-wave complexes. SeLECTS patients were excluded if they had abnormal magnetic resonance imaging (MRI) and a non-rapid eye movement (NREM) sleep discharge index ≥50 %. Focal seizures were diagnosed on the basis of clinical presentation and EEG results, and no significant abnormalities of cortical structures were found by imaging. According to Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V), childhood idiopathic epilepsy syndromes with ADHD was diagnosed by clinical pediatric neurologists based on the presence of clinical symptoms such as hyperactivity and inattention and the diagnostic criteria for ADHD according to the DSM-V. In addition, integrated Visual and Auditory Continuous Performance Test (i.e., IVA-CPT) was used as an auxiliary tools for assessing ADHD symptoms. The final sample included 8 children with childhood absence epilepsy (CAE), 22 children with self-limited epilepsy with centrotemporal spikes (SeLECTS), and 6 children with focal seizures epilepsy. Participants' medications were categorized as monotherapy and duotherapy. Monotherapy include the five medications Levetiracetam, Oxcarbazepine, Topiramate, Lamotrigine, and Sodium Valproate. Duotherapy are combinations of Oxcarbazepine + Levetiracetam or Sodium Valproate, and Sodium Valproate + Lamotrigine. Prior to the EEG biofeedback intervention, all participants receiving antiepileptic drugs were seizure free for the preceding three months at least. Thirty-six participants were randomised into two groups: the treatment group receiving the EEG biofeedback training (10 males and 8 females, age range 6–12 years) and the control group receiving no training (11 males and 7 females, age range 6–12 years). All participants completed the intervention throughout its duration. The intervention procedure including the EEG biofeedback intervention, pre‐ and post-intervention cognitive assessment is shown in Fig. 1. Illustration of intervention procedure. EEG biofeedback intervention In EEG biofeedback intervention, participants were encouraged to control their hyperactive and impulsive behaviors by learning to strengthen their beta waves (13–20 Hz) and suppress the theta waves (4–8 Hz) at the vertex [cortical zero (Cz)]. The ratio of theta and beta was plotted in a bar graph presented on a computer screen. In the training, there was a 45-s baseline testing phase, during which participants were instructed to look at a red dot on the center of the computer screen. During the baseline testing phase, the researcher encourages the child to remove distractions and concentrate by using a uniform instruction phrase, so as to smoothly enter into the EEG biofeedback treatment. The participants were instructed to complete four computerized games that were used to improve ADHD symptoms in an EEG biofeedback machine. The Four training tasks were sustained attention, practical task, short-term memory and discriminability. This is due to the fact that the epileptic child population is often accompanied by cognitive impairment, and epilepsy co-morbid cognitive impairment is mainly manifested as impaired attention, visuospatial and executive functions, which seriously affects the children's academic performance, quality of life and long-term social prognosis [[14], [15], [16]]. Through EEG biofeedback training, it is expected to optimize the brainwave patterns closely related to cognitive function, such as reducing theta-wave activity related to distraction and increasing beta-wave activity related to concentration, working memory and executive function, and to evaluate the efficiency of neural activity in brain regions closely related to cognitive function (e.g., parietal, temporal, and occipital lobes), so as to determine whether EEG biofeedback is able to improve the cognitive function of the patients. During the training tasks, the game on the feedback instrument monitor ran as an animation, at which point the child's activity in both frequency bands was represented by bars on the screen, and the game continued if there was a concomitant (below/above baseline values) decrease in Theta activity/increase in Beta activity, and vice versa. One point was scored for each additional second of play, and the points were displayed in the upper right corner of the screen. The patients' cognitive functions were evaluated. The training session including the four tasks lasted for 20 min each day and was implemented for 40 days during workday hours. The training procedure was conducted by a clinical technician in a training room. Pre‐ and post-intervention cognitive assessment tasks All participants completed a series of neuropsychological test battery both during the week before the first day of intervention and the week following the completion of the intervention. The test battery assessed cognitive abilities in: processing speed, spatial perception, arithmetic, semantic comprehension, verbal working memory, visual attention and executive functioning (see Table 1 and Fig. 2). All tests were carried out using computerized applications on the website (i.e., www.dweipsy.com/lattice) [17]. Participants’ responses were recorded by the computer. Table 1Neuropsychological test battery.Table 1AbilityTestsProcessing speedChoice reaction timeSpatial perceptionMental rotationArithmetic/computational fluencySimple subtractionSemantic comprehensionWord semanticsForward verbal working memoryForward digit spanBackward verbal working memoryBackward digit spanNon-verbal matrix reasoningRaven's Progressive MatricesVisual attentionVisual tracingExecutive functionWisconsin Card Sorting Test Neuropsychological test battery. Examples of neuropsychological tests. Choice reaction time The choice reaction time task was similar to a reaction time test [18]. On each trial, there was a white dot presented on either the left or the right of a cross in front of a black background on the screen. Participants were asked to press the “Q” (left) or “P” (right) key to make a judgment. The task contained 30 trials in total. The dot was presented on the left of the cross in half of the trials. This task was used to evaluate processing speed. Shorter choice reaction time represents more favourable performance. Mental rotation This task was adapted from the previously used mental rotation task [19] and was used to evaluate spatial skills. In the task, there were three three-dimensional images presented on the computer screen. Participants were asked to choose which image matched with the target image from the two alternatives on the bottom of the screen by pressing the “Q” (image on the left) or “P”(image on the right) key. The task was consisted of 180 trials with a 3-min time-limited. Higher scores mean better performance. Simple subtraction There were 92 simple subtraction problems including subtractions of single-digit numbers from either two-digit numbers or other single-digit numbers (e.g., 6–2, 17–8). Participants were asked to press the “Q” or “P” key choose the correct answer from the two candidate answers on the bottom of the screen. This task was used to assess calculation ability with a 2-min time limit. Higher scores mean better numeracy. Word semantics This task was similar to that used by previous researchers [20]. In each trial, a sentence with a missing word was presented on the computer screen. Participants were asked to select the semantically appropriate word from the two candidate words to make the sentence complete. The stimulus were presented on the screen until the participants responded. All sentences were adapted from textbooks used in middle and primary schools in China. The higher the participant's score, the better they performed. Forward or backward digit spans This task was used to assess verbal working memory [21]. A series of digits were verbally presented and each digit lasted for 200 ms. In the forward digit-span condition, participants were asked to input the digits in the exact digit order they have heard, whereas in the backward digit-span condition, participants were instructed to input the digits in a reversed order. A single score would be added upon successful recall of all digits in the correct order, and all other responses would be recorded as incorrect responses. The task lasted for about 10 min. The higher the final score, the better their performance. Visual tracing This task was adapted from a previous test [22]. In the task, several curved lines interweaved with one another were connected to the target number. Participants were asked to trace a specified line using their eyes only and then mark the correct end point. Each trial contained three questions. When participants correctly solved each question, a correct response would be recorded. The task was consisted of 36 trials with a 4-min time limit. The higher a participant's score, the better their performance. Wisconsin Card Sorting Test This task was adapted from a manual version of Wisconsin Card Sorting Test, which assesses the inhibition ability and mental flexibility [23,24]. There were four target cards: (a) one red triangle, (b) two green stars, (c) three yellow crosses, and (d) four blue circles. The task had three principles including color, form, or quantity. Participants were asked to turn over the response card and match them with the target cards according to one of the three principles. When participants made a judgment, a feedback would be given. If the judgement followed the principles, it would be recorded as a correct response. This was a 20-min time-limited task. The more successful matches and higher scores represent better performance. Test procedure The study used an EEG biofeedback therapy machine, model Infiniti 3000C, with the child wearing an electrode cap, three dry electrodes against the skin in the mid-frontal region, and headphones. A baseline test was performed first, lasting 45 s, and participants were instructed to look at a red dot in the center of a computer screen to set thresholds for each training session based on the baseline test level. The battery of tests was administered to patients in two 45-min sessions in an examination room. Test procedures were presented on a computer screen, and instructions were given orally. For each test, an instruction was given first, followed by a practice session. After participants finished the practice session and had no more questions, they could press any key to begin the formal test. The tests were administered in the same order for all participants. Each participant was monitored by one tester. Statistical analysis To control for the effect of guessing, for all, but one test (i.e., the choice reaction time test), corrected scores were computed by subtracting the number of incorrect responses from the number of correct responses [25]. For the choice reaction time test, the median reaction time was calculated. Data were analyzed using a two-way analysis of variance (ANOVA) with test phase (pretest and posttest) as the within-subjects variable and group (treatment group vs. control group) as the between-subjects variable for each test. All p values of the interactions and main effects were corrected by Greenhouse–Geisser method. Significant effects were further analyzed by post hoc pairwise analyses using the Bonferroni correction. Differences in categorical variables were analyzed using the chi-squared test. Analysis of covariance (ANCOVA) was conducted to examine the effect of medication on the intervention. In addition, correlation analyses were performed to investigate the relationships between changes in cognitive performance levels and symptom changes in IVA-CPT. To more fully illustrate the results, we selected scores on the Forward Digit Span and Backward digit span tasks and used a repeated measures ANOVA to further assess the impact of the EEG biofeedback treatment. We also conducted correlation analyses of the psychocognitive test tasks used to assess participants' cognitive functioning to explore the associations between the results of the different tasks. Results Population characteristics Table 2 showed the demographic characteristics for both the treatment and the control groups. There were no significant differences between the treatment group and the control group in age (F = 0.25, p = 0.62), gender (χ2 = 0.22, p = 0.64), Age at onset (F = 0.38, p = 0.54), and Number of seizures (F = 0.15, p = 0.70). Table 2Participant demographics for treatment and control groups.Table 2CharacteristicsTreatment groupControl groupp ValueNo. patients1818/Age, y, mean ± SD (range)8.3 ± 1.9 (6–12)8.1 ± 1.5(6–12)p = 0.62Age at onset, mean ± SD (range)6.2 ± 1.6 (2–10)5.8 ± 2.1(2–12)p = 0.54Number of seizures2.1 ± 1.2 (1–6)2.1 ± 1.3 (1–6)p = 0.70Gender, Male/Female10/811/7χ2 = 0.11Grade2.92.7/Epilepsy classificationCAE1012/SeLECTS62Focal seizures epilepsy24MedicationMonotherapy1414/Duotherapy44Note: CAE, childhood absence epilepsy; SeLECTS, self-limited epilepsy with centrotemporal spikes. Grade, counting from the first grade of primary school. Participant demographics for treatment and control groups. Note: CAE, childhood absence epilepsy; SeLECTS, self-limited epilepsy with centrotemporal spikes. Grade, counting from the first grade of primary school. Cognitive function In ANOVA analysis, significant effects were further analyzed by post hoc pairwise analyses using the Bonferroni correction (see Table 3). Thus, an adjusted alpha of 0.05 corresponded to 0.00625 before correction (0.05/8 = 0.00625). There was a significant interaction between testing phase and group in the forward digit-span task and the backward-digit span task, F (1, 32) = 5.91, p < 0.01, η2= .16; F (1, 32) = 5.45, p < 0.05, η2= .15. Further analyses showed that the treatment group had significant improvements in the forward digit-span task, F (1, 32) = 18.62, p = 0.00014, and the backward digit-span task, F (1, 32) = 11.35, p = 0.0019. For both pretest and posttest, the treatment group did not differ from the control group in working memory performance. The main effect and interaction effect did not reach significant levels in other tasks, all p > 0.05. The ANCOVA demonstrated that after controlling for medication, the interaction effect in the forward digit-span task and the backward digit-span task remained significant, F (1, 31) = 5.32, p < 0.05; F (1, 31) = 5.13, p < 0.05. Correlation analyses showed that changes in working memory performance did not correlate with IVA-CPT symptom changes. Table 3Means (standard deviations), ANOVA of scores and correlation analyses results in cognitive tests for treatment and control groups.Table 3TestsTreatment groupControl groupFrPre-testPost-testPre-testPost-testChoice reaction time612.00(106.75)611.63(57.82)699.39(63.28)623.44(47.17)0.430.09Mental rotation5.38(2.36)9.81(2.08)4.28(1.67)7.17(1.85)0.180.22Simple subtraction13.75(2.83)20.31(2.39)19.00(2.11)26.17(2.14)0.030.30Word semantics5.38(2.77)10.81(3.08)6.39(2.33)13.06(2.47)0.090.17Forward digit span2.5(0.79)6.19(0.63)4.17(0.77)5.00(0.72)5.91∗−0.10Backward digit span1.38(0.52)3.69(0.68)2.72(0.52)2.83(0.55)5.45∗−0.01Visual tracing7.78(1.31)7.72(1.20)4.21(0.76)5.42(1.21)1.07−0.09Wisconsin Card Sorting Test43.11(7.21)54.22(7.63)40.89(6.68)57.26(6.03)0.02−0.04 Means (standard deviations), ANOVA of scores and correlation analyses results in cognitive tests for treatment and control groups. The results of the repeated measures ANOVA showed that the treatment group scored significantly higher on the forward digit span task, with a significant time effect (see Table 4), F(1, 32) = 14.82, p < 0.01, and an equally significant interaction effect between time and group, F(1, 32) = 5.91, p = 0.02. The treatment group also demonstrated elevated scores on the backward digit span test, with a significant time effect, F(1, 32) = 6.60, p = 0.02, and an equally significant interaction effect between time and group, F(1, 32) = 5.45, p = 0.03. The results demonstrated a significantly greater improvement in the treatment group than in the control group, thus further validating the effectiveness of EEG biofeedback therapy. Table 4Repeated measures ANOVA results for forward and backward digit span in treatment and control groups.Table 4Forward digit spanBackward digit spanTreatment groupControl groupTreatment groupControl groupPre-test2.50(3.16)4.17(3.28)1.38(2.06)2.72(2.19)Post-test6.19(2.54)5.00(3.05)3.69(2.70)2.83(2.33)FF(time) = 14.82, F(time∗group) = 5.91F(time) = 6.60, F(time∗group) = 5.45pp(time) < 0.01, p(time∗group) = 0.02p(time) = 0.02, p(time∗group) = 0.03 Repeated measures ANOVA results for forward and backward digit span in treatment and control groups. The results of the correlation analyses are presented in matrix form (see Table 5). There was a moderate correlation between the pretest and posttest for the choice reaction time task (p < 0.05); the mental rotation task showed high agreement between the pretest and posttest (p < 0.05); the word semantics task showed a significant positive correlation between the pretest and posttest (p < 0.05); the digit span tests, especially the backward digit span task, showed a high correlation between pretest and posttest (p < 0.01). Table 5Correlation matrix of cognitive task performance before and after intervention.Table 5123456789101112131415161.Choice reaction time(pre)12.Choice reaction time(post)0.374∗13.Mental rotation(pre)0.055−0.04814.Mental rotation(post)0.142−0.0280.18315.Simple subtraction(pre)0.045−0.350∗0.2510.10816.Simple subtraction(post)−0.340∗−0.3190.0710.359∗0.492∗∗17.Word semantics(pre)−0.046−0.2580.021−0.0060.475∗∗0.399∗18.Word semantics(post)−0.152−0.1430.0310.2890.362∗0.493∗∗0.384∗19.Forward digit span(pre)0.2830.0250.1940.1450.2780.081−0.0190.084110.Forward digit span(post)−0.1010.0000.3200.107−0.0940.077−0.098−0.0850.291111.Backward digit span(pre)0.2870.2060.1180.2140.3160.1960.1050.1840.852∗∗0.147112.Backward digit span(post)−0.0520.0050.1120.257−0.1040.1620.1050.1740.1760.669∗∗0.237113.Visual tracing(pre)−0.1900.2240.204−0.237−0.357∗−0.103−0.147−0.026−0.0540.196−0.1280.248114.Visual tracing(post)−0.2850.072−0.058−0.284−0.185−0.212−0.231−0.047−0.044−0.245−0.032−0.2040.507∗∗115.Wisconsin Card Sorting Test(pre)−0.0570.065−0.060−0.282−0.152−0.028−0.285−0.025−0.1820.049−0.346∗−0.0240.435∗∗0.160116.Wisconsin Card Sorting Test(post)0.0690.228−0.139−0.393∗−0.382∗−0.346∗−0.415∗−0.320−0.184−0.248−0.216−0.3140.1500.410∗0.2501 Correlation matrix of cognitive task performance before and after intervention. Discussion The current study aimed to clarify the effect of EEG biofeedback on cognitive functions in childhood idiopathic epilepsy syndromes with comorbid ADHD. We assessed changes in cognitive performances using a standardized, computerized test battery before and after the EEG-biofeedback training. Our results indicated that the EEG Biofeedback intervention group showed improved working memory. The finding suggested that the EEG biofeedback intervention could improve the core cognitive function in childhood idiopathic epilepsy syndromes with comorbid ADHD. A leading theory of ADHD suggested a core deficit in working memory [26,27]. A meta-analysis examined cognitive performances in children with ADHD by a variety of executive function tasks, and found significant impairments in working memory in children with ADHD [27]. Meanwhile, recent studies suggested that the EEG biofeedback training can enhance working memory. For example, in a controlled experimental study, Escolano et al. found that the EEG-biofeedback treatment significantly enhanced working memory [28]. In addition, EEG-biofeedback training has been proved to be effective in neurocognitive disorders. EEG-biofeedback training improved not only attentiveness but also working memory in children with a reading disorder [29]. In a recent study, children with ADHD showed significant improvement in sustained attention, verbal working memory, and response inhibition after an EEG-biofeedback training [30]. With regard to working memory, researchers explained that the central executive component of working memory is similar to an attentional system. Therefore, the neuronal circuits of working memory could be affected by controlling the attentional system using the EEG-biofeedback training. The present study found that childhood idiopathic epilepsy syndromes with comorbid ADHD improved in verbal working memory after the EEG-biofeedback training. Previous researchers proposed an explanation of the EEG-biofeedback training mechanism as a kind of “dynamic resonant loop” affecting the frontal cortical areas. With the frontal disturbances and executive dysfunction, childhood idiopathic epilepsy syndromes showed delayed long-term development in working memory. The EEG-Biofeedback training could influence the executive network and then lead to improvement in working memory. This idea was supported by the fMRI findings that presented significant activations in the frontal cortex while performing Go/No-Go task in children with ADHD following the EEG-biofeedback training [31,32]. Such idea was also consistent with the viewpoint that the EEG-biofeedback training led to high-level efficiency in storage, access and updating of working memory. The current study confirmed that the EEG biofeedback intervention may be an effective treatment for childhood idiopathic epilepsy syndromes with comorbid ADHD by improving working memory. This finding suggests that the EEG biofeedback could be a helpful approach for improving the cognitive function in epilepsy patients with comorbid ADHD. While the study provided some supports for its effectiveness, there were no indications that the EEG biofeedback intervention was associated with any reduction in ADHD symptom severity. This might be caused by the relatively small sample in the study, but it may also reflect the independence of the improved neuro-cognitive function with respect to the pathophysiology of ADHD. In addition, when discussing the effects of biofeedback training, it is important to consider the number of training sessions. A meta-analysis study suggests that the number of neurofeedback interventions has a positive effect on the response inhibition domain [33]. A randomized controlled trial also showed that 30 to 40 sessions of EEG-Biofeedback training were as effective as methylphenidate in reducing symptoms of inattention and hyperactivity, and were even associated with superior academic performance after treatment [34,35]. Therefore, we used 40 sessions of EEG biofeedback training, which is one of the more effective training volumes [36,37]. This study has some limitations. First, although participants in the treatment group showed a greater improvement in verbal working memory as a result of the intervention when compared to the control group, we did not find superior verbal working memory performance in the post-treatment assessment in the treatment group than the control group. It is possible that this is an effect due to the different cognitive reserves of the patients. As such, the beneficial effects of this treatment remain questionable. Second, the present study enrolled a relatively small sample. Childhood idiopathic epilepsy syndromes with comorbid ADHD are rare in our hospital, and therefore it is difficult to secure a larger sample. Our findings will require confirmation in a larger sample of children. In addition, this study did not provide the long-term effect of the EEG biofeedback intervention. Future studies should include additional follow-up assessments to test the long-term effect of the EEG biofeedback intervention on cognitive functions in childhood idiopathic epilepsy syndromes with comorbid ADHD. Conclusion The current study provided support for the effectiveness of the EEG-Biofeedback intervention in improving verbal working memory in childhood idiopathic epilepsy syndromes with comorbid ADHD. The findings suggested that such intervention may be used to enhance the core cognitive function in children who assemble the characteristics in our sample. Future research on the intervention could focus on its long-term effects and explore its influence on the common comorbidities of childhood epilepsy. AbbreviationsADHDAttention-deficit hyperactivity disorderEEGElectroencephalogramCAEChildhood absence epilepsySeLECTSSelf-limited epilepsy with centrotemporal spikesMRIMagnetic resonance imagingNREMNon-rapid eye movementDSM-VDiagnostic and Statistical Manual of Mental Disorders, Fifth EditionIVA-CPTVisual and Auditory Continuous Performance TestANOVAAnalysis of varianceANCOVAAnalysis of covariance Abbreviations CRediT authorship contribution statement Dazhi Cheng: Writing – original draft, Methodology, Funding acquisition, Formal analysis, Data curation, Conceptualization. Yuchen Yang: Writing – original draft. Xiuxian Yan: Investigation, Data curation. Qian Chen: Writing – review & editing, Funding acquisition, Conceptualization. Ethics approval and consent to participate The study was conducted in accordance with the 1964 Declaration of Helsinki and its later amendments, and was approved Human Research Ethics Committee at the Capital Institute of Pediatrics (approval no.: SHERLL 2015023). Consent to participate was obtained from each patient. Written informed consent was obtained from the parents or guardians of all the patients. Data availability statement All data generated or analyzed during this study are included in this published article. Consent for publication Not applicable. Funding This research was supported by a grant from Natural Science Foundation of China (Grant No. 31700977, 32271122), Natural Science Foundation of Beijing (Grant No. 5212004), and the project supported by Beijing Municipal Science & Technology Commission (Grant No. Z171100000417020). Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements Not applicable. Contributor Information Dazhi Cheng, Email: cheng2023@cnu.edu.cn. Qian Chen, Email: chenqianxhl@163.com. References Associated Data Data Availability Statement All data generated or analyzed during this study are included in this published article.